Advertisement

Predictive Biomarkers for Immunotherapy in Lung Cancer: Perspective From the International Association for the Study of Lung Cancer Pathology Committee

Published:September 28, 2022DOI:https://doi.org/10.1016/j.jtho.2022.09.109

      Abstract

      Immunotherapy including immune checkpoint inhibitors (ICIs) has become the backbone of treatment for most lung cancers with advanced or metastatic disease. In addition, they have increasingly been used for early stage tumors in neoadjuvant and adjuvant settings. Unfortunately, however, only a subset of patients experiences meaningful response to ICIs. Although programmed death-ligand 1 (PD-L1) protein expression by immunohistochemistry (IHC) has played a role as the principal predictive biomarker for immunotherapy, its performance may not be optimal, and it suffers multiple practical issues with different companion diagnostic assays approved. Similarly, tumor mutational burden (TMB) has multiple technical issues as a predictive biomarker for ICIs. Now, ongoing research on tumor- and host immune-specific factors has identified immunotherapy biomarkers that may provide better response and prognosis prediction, in particular in a multimodal approach. This review by the International Association for the Study of Lung Cancer Pathology Committee provides an overview of various immunotherapy biomarkers, including updated data on PD-L1 IHC and TMB, and assessments of neoantigens, genetic and epigenetic signatures, immune microenvironment by IHC and transcriptomics, and microbiome and pathologic response to neoadjuvant immunotherapies. The aim of this review is to underline the efficacy of new individual or combined predictive biomarkers beyond PD-L1 IHC and TMB.

      Keywords

      Introduction

      Lung cancer is the leading cause of cancer-related death worldwide.
      • Sung H.
      • Ferlay J.
      • Siegel R.L.
      • et al.
      Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries.
      Its poor prognosis is historically attributed to difficulty in early detection and its low response rates to conventional chemotherapy with or without radiation therapies. Nevertheless, critical progress has been made in the past decade to substantially improve survival of metastatic non-small cell lung cancer (NSCLC). In particular, immune checkpoint inhibitors (ICIs) including the programmed cell death protein-1 (PD-1)/programmed death-ligand 1 (PD-L1) axis blockade, in either first- or second-line settings and irrespective of the histological subtypes of NSCLC, have led to unprecedented prolonged survival for a subset of patients.
      • Reck M.
      • Remon J.
      • Hellmann M.D.
      First-line immunotherapy for non-small-cell lung cancer.
      To date, eight anti–PD-1 or PD-L1 antibodies (pembrolizumab, nivolumab, atezolizumab, durvalumab, cemiplimab-rwlc, sintilimab, tislelizumab, and camrelizumab) have been approved globally or in some countries for treatment of NSCLC, either as monotherapy or in combination with chemotherapy and or an anti–CTLA-4 agent.
      • Reck M.
      • Remon J.
      • Hellmann M.D.
      First-line immunotherapy for non-small-cell lung cancer.
      ,
      • Zhang L.
      • Lin W.
      • Tan F.
      • et al.
      Sintilimab for the treatment of non-small cell lung cancer.
      Although PD-L1 immunohistochemistry (IHC) has been used as a companion diagnostic for anti–PD-1 or PD-L1 monotherapy or nivolumab and ipilimumab combination therapy, it is known to be an imperfect predictive biomarker.
      • Lantuejoul S.
      • Sound-Tsao M.
      • Cooper W.A.
      • et al.
      PD-L1 testing for lung cancer in 2019: perspective from the IASLC pathology committee.
      Similarly, tumor mutational burden (TMB) has multiple technical issues as a predictive biomarker,
      • Sholl L.M.
      • Hirsch F.R.
      • Hwang D.
      • et al.
      The promises and challenges of tumor mutation burden as an immunotherapy biomarker: a perspective from the International Association for the Study of Lung Cancer Pathology Committee.
      although the FoundationOne CDx assay has been approved by the Food and Drug Administration (FDA) as a companion diagnostic for second-line pembrolizumab monotherapy for solid tumors including lung cancer with high TMB.
      • Marcus L.
      • Fashoyin-Aje L.A.
      • Donoghue M.
      • et al.
      FDA approval summary: Pembrolizumab for the treatment of tumor mutational burden-high solid tumors.
      After the publication of a perspective on PD-L1 testing for lung cancer,
      • Lantuejoul S.
      • Sound-Tsao M.
      • Cooper W.A.
      • et al.
      PD-L1 testing for lung cancer in 2019: perspective from the IASLC pathology committee.
      on TMB,
      • Sholl L.M.
      • Hirsch F.R.
      • Hwang D.
      • et al.
      The promises and challenges of tumor mutation burden as an immunotherapy biomarker: a perspective from the International Association for the Study of Lung Cancer Pathology Committee.
      and on pathologic assessment of lung cancer resection specimens after neoadjuvant therapy,
      • Travis W.D.
      • Dacic S.
      • Wistuba I.
      • et al.
      IASLC multidisciplinary recommendations for pathologic assessment of lung cancer resection specimens after neoadjuvant therapy.
      this article from the International Association for the Study of Lung Cancer (IASLC) Pathology Committee provides an overview of various biomarkers that may help in predicting ICI response in NSCLC, featuring updated information on PD-L1 IHC and TMB, and neoantigens, genetic and epigenetic signatures, immune microenvironment assessments by IHC, transcriptomic analyses and novel technologies, and microbiome composition. Assessments of pathologic response to neoadjuvant immunotherapies, which are considered predictive of overall survival, are also discussed. This review emphasizes the notion that a combination of predictive biomarkers outperforms the use of individual marker approach
      • Lee J.S.
      • Ruppin E.
      Multiomics prediction of response rates to therapies to inhibit programmed cell death 1 and programmed cell death 1 ligand 1.
      ,
      • Lu S.
      • Stein J.E.
      • Rimm D.L.
      • et al.
      Comparison of biomarker modalities for predicting response to PD-1/PD-L1 checkpoint blockade: a systematic review and meta-analysis.
      and might help implement a multimodal approach in clinical practice.

      PD-L1 as Predictive Biomarker for Immunotherapy in Lung Cancer

      PD-L1 protein expression has emerged as a clinically useful biomarker for treatment decisions regarding immunotherapy (immuno-oncology [IO]).
      • Doroshow D.B.
      • Bhalla S.
      • Beasley M.B.
      • et al.
      PD-L1 as a biomarker of response to immune-checkpoint inhibitors.
      Several studies have led to approvals by regulatory agencies on the basis of PD-L1 expression (Table 1). Although initially there was a confusion about the comparability of different PD-L1 IHC assays used in clinical trials, several studies including the IASLC “PD-L1 Blueprint Project” compared the different assays and found three assays very similar in expression pattern—22C3 (Dako), 28-8 (Dako), and SP 263 (Ventana)
      • Hirsch F.R.
      • McElhinny A.
      • Stanforth D.
      • et al.
      PD-L1 immunohistochemistry assays for lung cancer: results from phase 1 of the blueprint PD-L1 IHC assay comparison project.
      —and good reproducibility among different observers.
      • Tsao M.S.
      • Kerr K.M.
      • Kockx M.
      • et al.
      PD-L1 immunohistochemistry comparability study in real-life clinical samples: results of blueprint phase 2 project.
      Since then, the three assays have been used interchangeably with similar clinical interpretations of treatment results, despite some preanalytical and analytic variabilities.
      • Tsao M.S.
      • Kerr K.M.
      • Kockx M.
      • et al.
      PD-L1 immunohistochemistry comparability study in real-life clinical samples: results of blueprint phase 2 project.
      ,
      • Tsao M.S.
      • Kerr K.M.
      • Dacic S.
      • Yatabe Y.
      • Hirsch F.R.
      IASLC Atlas of PD-L1 Immunohistochemistry Testing in Lung Cancer.
      Currently, pembrolizumab, cemiplimab-rwlc, and atezolizumab are approved in first-line setting for advanced NSCLC with high PD-L1 expression (≥50% tumor proportion score [TPS] or in ≥50% of tumor cells or PD-L1–stained tumor-infiltrating immune cells ≥ 10% of the tumor area for the latter).

      Food and Drug Administration. FDA approves cemiplimab-rwlc for non-small cell lung cancer with high PD-L1 expression. https://www.fda.gov/drugs/resources-information-approved-drugs/fda-approves-cemiplimab-rwlc-non-small-cell-lung-cancer-high-pd-l1-expression. Accessed February 22, 2021.

      Several studies have revealed that IO monotherapy with pembrolizumab is the treatment of choice in high PD-L1 expressors as retrospective analysis and historical control comparisons have not revealed any difference in outcome between IO alone or in combination with chemotherapy.
      • Paz-Ares L.G.
      • Ramalingam S.S.
      • Ciuleanu T.E.
      • et al.
      First-line nivolumab plus ipilimumab in advanced NSCLC: 4-year outcomes from the randomized, open-label, phase 3 CheckMate 227 part 1 Trial.
      • Sezer A.
      • Kilickap S.
      • Gümüş M.
      • et al.
      Cemiplimab monotherapy for first-line treatment of advanced non-small-cell lung cancer with PD-L1 of at least 50%: a multicentre, open-label, global, phase 3, randomised, controlled trial.
      • Herbst R.S.
      • Giaccone G.
      • de Marinis F.
      • et al.
      Atezolizumab for first-line treatment of PD-L1-selected patients with NSCLC.
      Although the prospective comparison between IO with chemotherapy versus IO alone in different PD-L1 subgroups is currently being studied in the INSIGNA trial (NCT03793179), many guidelines point to chemotherapy-sparing treatment for the population with high PD-L1 expression (approximately 30% of NSCLC). In addition, retrospective analyses have revealed that very high protein expression (TPS ≥90%) should lead to better outcome with IO monotherapy.
      • Aguilar E.J.
      • Ricciuti B.
      • Gainor J.F.
      • et al.
      Outcomes to first-line pembrolizumab in patients with non-small-cell lung cancer and very high PD-L1 expression.
      In contrast, according to the CheckMate 227 study, no or low PD-L1 expression indicated a favorable outcome with PD-L1 with CTLA-4 blockage.
      • Paz-Ares L.G.
      • Ramalingam S.S.
      • Ciuleanu T.E.
      • et al.
      First-line nivolumab plus ipilimumab in advanced NSCLC: 4-year outcomes from the randomized, open-label, phase 3 CheckMate 227 part 1 Trial.
      Table 1Immunotherapies Approved in the Frontline Setting for the Treatment of Advanced/Metastatic NSCLC
      PD-L1 ExpressionRegimen
      Patients should be tested and negative for EGFR and ALK genomic tumor aberrations.
      Pivotal TrialHistology
      PD-L1 ≥ 1%
      Opdivo [PI]. Princeton, NJ: Bristol-Myers Squibb Company; July 23, 2021.
      ,
      Keytruda [PI]. Whitehouse Station, NJ: Merck & Co., Inc.; March 2021.
      Nivolumab + ipilimumabCheckMate 227NSQ/SQ
      Pembrolizumab monotherapyKEYNOTE-024 KEYNOTE-042NSQ/SQ
      PD-L1 ≥ 50%
      Tecentriq [PI]. South San Francisco, CA: Genentech, Inc.; April 2021.
      Libtayo [PI]. Tarrytown, NY: Regeneron Pharmaceuticals, Inc.; Bridgewater, NJ: Sanofi-Aventis U.S. LLC; February 2021.
      U.S. Food and Drug Administration. October 24, 2016.
      U.S. Food and Drug Administration. April 11, 2019.
      Atezolizumab monotherapyIMpower110NSQ/SQ
      Cemiplimab-rwlc monotherapyEMPOWER-LUNG 1NSQ/SQ
      Pembrolizumab monotherapy
      Initially approved as a single agent in patients with metastatic NSCLC and high PD-L1 expression (≥50%). In 2019, this approval was expanded to patients with tumors expressing a PD-L1 TPS of 1% or greater.
      KEYNOTE-024NSQ/SQ
      Regardless of PD-L1 expression
      Opdivo [PI]. Princeton, NJ: Bristol-Myers Squibb Company; July 23, 2021.
      ,
      Keytruda [PI]. Whitehouse Station, NJ: Merck & Co., Inc.; March 2021.
      ,
      Tecentriq [PI]. South San Francisco, CA: Genentech, Inc.; April 2021.
      ,
      China National Medical Products Administration.
      Atezolizumab + bevacizumab, paclitaxel, and carboplatinIMpower150NSQ
      Atezolizumab + paclitaxel protein-bound and carboplatinIMpower130NSQ
      Nivolumab + ipilimumab + 2 cycles platinum-doublet chemotherapyCheckMate 9LANSQ/SQ
      Pembrolizumab + pemetrexed and platinum chemotherapyKEYNOTE-189NSQ
      Pembrolizumab + carboplatin and either paclitaxel or paclitaxel protein-boundKEYNOTE-407SQ
      Sintilimab + pemetrexed and platinum chemotherapyORIENT-11NSQ
      Sintilimab + gemcitabine and platinum chemotherapyORIENT-12SQ
      Tislelizumab + platinum-based chemotherapyRATIONALE 304NSQ
      Tislelizumab + carboplatin-based chemotherapyRATIONALE 307SQ
      Camrelizumab + pemetrexed and carboplatinCameLNSQ
      Camrelizumab + paclitaxel and carboplatinCameL-SqSQ
      NSQ, nonsquamous; PD-L1, programmed death-ligand 1; SQ, squamous; TPS, tumor proportion score.
      a Patients should be tested and negative for EGFR and ALK genomic tumor aberrations.
      b Opdivo [PI]. Princeton, NJ: Bristol-Myers Squibb Company; July 23, 2021.
      c Keytruda [PI]. Whitehouse Station, NJ: Merck & Co., Inc.; March 2021.
      d Tecentriq [PI]. South San Francisco, CA: Genentech, Inc.; April 2021.
      e Libtayo [PI]. Tarrytown, NY: Regeneron Pharmaceuticals, Inc.; Bridgewater, NJ: Sanofi-Aventis U.S. LLC; February 2021.
      f U.S. Food and Drug Administration. October 24, 2016.
      g U.S. Food and Drug Administration. April 11, 2019.
      h Initially approved as a single agent in patients with metastatic NSCLC and high PD-L1 expression (≥50%). In 2019, this approval was expanded to patients with tumors expressing a PD-L1 TPS of 1% or greater.
      i China National Medical Products Administration.
      Whether there is a difference in the predictive value on the basis of histological subtypes, a retrospective analysis of a large number of patients indicates no predictive value of PD-L1 expression in squamous cell carcinoma in contrast to adenocarcinoma.
      • Doroshow D.B.
      • Wei W.
      • Gupta S.
      • et al.
      Programmed death-ligand 1 tumor proportion score and overall survival from first-line pembrolizumab in patients with nonsquamous versus squamous NSCLC.
      Furthermore, in contrast to NSCLC, SCLC has in general low PD-L1 expression.
      • Yu H.
      • Batenchuk C.
      • Badzio A.
      • et al.
      PD-L1 expression by two complementary diagnostic assays and mRNA in situ hybridization in small cell lung cancer.
      So far, no predictive role of PD-L1 expression has been found related to IO in SCLC.
      • Ready N.
      • Farago A.F.
      • de Braud F.
      • et al.
      Third-line nivolumab monotherapy in recurrent SCLC: CheckMate 032.
      Although the above-mentioned predictive paradigm seems relatively simple, many issues remain. First, PD-L1 expression is spatially and temporally heterogeneous and may vary between tumor sites (primary versus metastases). Second, the continuous distribution of PD-L1 expression makes the determination of a reliable binary cutpoint difficult. Last, but not least, it needs to be formally evaluated whether combination of PD-L1 IHC with other markers such as TMB, gene expression profiling (transcriptomic signatures), and multiplex IHC and multiplex immunofluorescence (mIF) for tumor microenvironment (TME) component analyses may be more predictive than PD-L1 IHC alone.
      • Lu S.
      • Stein J.E.
      • Rimm D.L.
      • et al.
      Comparison of biomarker modalities for predicting response to PD-1/PD-L1 checkpoint blockade: a systematic review and meta-analysis.
      ,
      • Hirsch F.R.
      • Walker J.
      • Higgs B.W.
      • Cooper Z.A.
      • Raja R.G.
      Wistuba II. The Combiome hypothesis: selecting optimal treatment for cancer patients.
      For instance, we already know that high PD-L1 expression does not significantly overlap with high TMB and that combination of PD-L1 expression and TMB might give a better prediction than one biomarker alone.
      • Gandara D.R.
      • Paul S.M.
      • Kowanetz M.
      • et al.
      Blood-based tumor mutational burden as a predictor of clinical benefit in non-small-cell lung cancer patients treated with atezolizumab.

      TMB, Neoantigens, Genetic and Epigenetic Signatures

      In September 2020, our group published a review article on the promises and challenges of TMB as an immunotherapy biomarker.
      • Sholl L.M.
      • Hirsch F.R.
      • Hwang D.
      • et al.
      The promises and challenges of tumor mutation burden as an immunotherapy biomarker: a perspective from the International Association for the Study of Lung Cancer Pathology Committee.
      The recognized challenges included the following: (1) identification of therapies whose response was best informed by TMB status; (2) robust definition of a predictive TMB cutpoint; (3) standardization of sequencing panel size and design; and (4) the need for robust technical and informatics rigor to generate precise and accurate TMB measurements across different laboratories. Nevertheless, on June 16, 2020, the FDA granted an accelerated approval of pembrolizumab for the treatment of adult and pediatric patients with unresectable or metastatic solid tumors with high tumor tissue mutational burden (tTMB-H, as defined by ≥10 mutations/megabase [mut/Mb]) determined by an FDA-approved test, who have progressed after previous treatment and who have no satisfactory alternative treatment options.
      • Marcus L.
      • Fashoyin-Aje L.A.
      • Donoghue M.
      • et al.
      FDA approval summary: Pembrolizumab for the treatment of tumor mutational burden-high solid tumors.
      Importantly, the FDA-approved TMB test referred to the FoundationOne CDx assay (Foundation Medicine, Inc.), which was approved on the same day as the companion diagnostic for pembrolizumab in this setting. This approval was based on the results of KEYNOTE-158 (NCT02628067), a phase 2 multicohort, open-label, nonrandomized study involving 10 types of advanced incurable solid tumors including SCLC.
      • Marabelle A.
      • Fakih M.
      • Lopez J.
      • et al.
      Association of tumour mutational burden with outcomes in patients with advanced solid tumours treated with pembrolizumab: prospective biomarker analysis of the multicohort, open-label, phase 2 KEYNOTE-158 study.
      Among 790 patients with assessable tTMB, 102 (13%) had tTMB-H, and their overall response rate was 29% (95% confidence interval [CI]: 21–39) versus 6% (95% CI: 5–8) for the non–tTMB-H group. Impressively, 57% and 50% of patients who responded had response duration of more than or equal to 12 months and 24 months, respectively. Among the 76 patients with SCLC included, 34 (45%) had tTMB-H. The overall response rate for tTMB-H and non–tTMB-H patients with SCLC were 29% and 9.5%, and median overall survivals were 9.4 (95% CI: 5.6–19.1) and 6.3 (95% CI: 3.9–7.7) months, respectively.
      • Marabelle A.
      • Fakih M.
      • Lopez J.
      • et al.
      Association of tumour mutational burden with outcomes in patients with advanced solid tumours treated with pembrolizumab: prospective biomarker analysis of the multicohort, open-label, phase 2 KEYNOTE-158 study.
      This result is consistent with the results of the TMB study on the efficacy of nivolumab or nivolumab plus ipilimumab in previously treated patients with SCLC who were accrued to the CheckMate 032 trial, as discussed in our previous review.
      • Sholl L.M.
      • Hirsch F.R.
      • Hwang D.
      • et al.
      The promises and challenges of tumor mutation burden as an immunotherapy biomarker: a perspective from the International Association for the Study of Lung Cancer Pathology Committee.
      ,
      • Hellmann M.D.
      • Callahan M.K.
      • Awad M.M.
      • et al.
      Tumor mutational burden and efficacy of nivolumab monotherapy and in combination with ipilimumab in small-cell lung cancer.
      Nevertheless, assessment of blood TMB in IMpower133 trial revealed the lack of predictiveness of this marker (at either 10 or 16 mut/Mb) for the efficacy of atezolizumab plus chemotherapy in extensive-stage SCLC.
      • Horn L.
      • Mansfield A.S.
      • Szczesna A.
      • et al.
      First-line atezolizumab plus chemotherapy in extensive-stage small-cell lung cancer.
      To date, assessment of TMB as a predictive marker for ICI therapy in NSCLC trials has largely been retrospective and exploratory, except in CheckMate-227.
      • Hellmann M.D.
      • Paz-Ares L.
      • Bernabe Caro R.
      • et al.
      Nivolumab plus ipilimumab in advanced non-small-cell lung cancer.
      ,
      • Hellmann M.D.
      • Ciuleanu T.E.
      • Pluzanski A.
      • et al.
      Nivolumab plus ipilimumab in lung cancer with a high tumor mutational burden.
      Overall, the data suggested that high tTMB was consistently associated with greater survival benefit from anti–PD-1 or PD-L1 monotherapy versus standard-of-care chemotherapy, whereas such benefit was not apparent in low tTMB patients (Table 2). Nevertheless, this difference was not apparent when ICI is combined with chemotherapy or when dual ICI is compared with single ICI therapy (MYSTIC trial).
      • Rizvi N.A.
      • Cho B.C.
      • Reinmuth N.
      • et al.
      Durvalumab with or without tremelimumab vs standard chemotherapy in first-line treatment of metastatic non-small cell lung cancer: the MYSTIC Phase 3 randomized clinical trial.
      As previously discussed,
      • Sholl L.M.
      • Hirsch F.R.
      • Hwang D.
      • et al.
      The promises and challenges of tumor mutation burden as an immunotherapy biomarker: a perspective from the International Association for the Study of Lung Cancer Pathology Committee.
      despite initial optimistic results in CheckMate 227 that high TMB (≥10 mut/Mb) could be predictive for nivolumab plus ipilimumab combination therapy, the subsequent overall survival data suggested that TMB was prognostic rather than predictive,
      • Hellmann M.D.
      • Paz-Ares L.
      • Bernabe Caro R.
      • et al.
      Nivolumab plus ipilimumab in advanced non-small-cell lung cancer.
      ,
      • Hellmann M.D.
      • Ciuleanu T.E.
      • Pluzanski A.
      • et al.
      Nivolumab plus ipilimumab in lung cancer with a high tumor mutational burden.
      leading to eventual FDA approval of this ICI combination therapy as first-line treatment in patients with metastatic NSCLC without EGFR/ALK aberration on the basis of the PD-L1 (≥1%) status only.

      Food and Drug Administration. FDA approves nivolumab plus ipilimumab and chemotherapy for first-line treatment of metastatic NSCLC. https://www.fda.gov/drugs/resources-information-approved-drugs/fda-approves-nivolumab-plus-ipilimumab-and-chemotherapy-first-line-treatment-metastatic-nsclc. Accessed May 26, 2020

      Table 2Tissue-Based Evaluation of TMB as Predictive Biomarker for ICI Therapies in Phase 3 Trials Involving NSCLC
      Line of TxClinical TrialAnalysisStudy TherapyReference TherapyAssayCutpointSurvivalHazard RatioRef
      High TMBLow TMB
      Anti–PD-1 or PD-L1 vs. chemo
      1LCM-26ENivoChemoWES243 mut/exomePFS0.62 (0.38–1.00)1.82 (1.30–2.55)
      • Carbone D.P.
      • Reck M.
      • Paz-Ares L.
      • et al.
      First-line nivolumab in Stage IV or recurrent non-small-cell lung cancer.
      1LKN-042EPembroChemoWES175 mut/exomePFS0.75 (0.59–0.95)1.27 (1.04–1.55)
      • Herbst R.S.
      • Lopes G.
      • Kowalski D.M.
      • et al.
      Association between tissue TMB (tTMB) and clinical outcomes with pembrolizumab monotherapy (pembro) in PD-L1-positive advanced NSCLC in the KEYNOTE-010 and −042 trials.
      1LMYSTICEDurvaChemoF1CDx10 mut/MbOS0.70 (0.47–1.06)1.26 (0.90–1.77)29
      2LKN-010EPembroChemoWES175 mut/exomePFS0.59 (0.40–0.87)1.09 (0.72–1.63)161
      2/3LPOPLAREAtezoChemoF116.2 mut/MbPFS0.49 (0.19, 1.3)1.28 (0.77, 2.12)
      • Kowanetz M.
      • Zou W.
      • Shames D.
      • et al.
      OA20.01 Tumor mutation burden (TMB) is associated with improved efficacy of atezolizumab in 1L and 2L+ NSCLC patients.
      9.9 mut/MbPFS0.49 (0.25–0.93)2.41 (1.24–4.67)
      Anti–PD1 or PD-L1 + anti–CTLA-4 vs. chemo
      1LCM-227PNivo + IpiChemoF1CDx10 mut/MbPFS0.58 (0.41–0.81)1.07 (0.84–1.35)27,28
      10 mut/MbOS0.68 (0.51–0.91)0.75 (0.59–0.94)
      1LMYSTICEDurva + TremeChemoF1CDx10 mut/MbOS0.72 (0.48–1.09)1.39 (1.00–1.92)29
      Anti-PD1 + chemo vs. chemo
      1LKN-189EPembro + ChemoChemoWES175 mut/exomePFS0.32 (0.21–0.51)0.51 (0.35–0.74)
      • Paz-Ares L.
      • Langer C.
      • Novello S.
      • et al.
      LBA80 - Pembrolizumab (pembro) plus platinum-based chemotherapy (chemo) for metastatic NSCLC: tissue TMB (tTMB) and outcomes in KEYNOTE-021,189, and 407.
      1LKN-407EPembro + ChemoChemoWES175 mut/exomePFS0.57 (0.41–0.81)0.68 (0.48–0.96)
      • Paz-Ares L.
      • Langer C.
      • Novello S.
      • et al.
      LBA80 - Pembrolizumab (pembro) plus platinum-based chemotherapy (chemo) for metastatic NSCLC: tissue TMB (tTMB) and outcomes in KEYNOTE-021,189, and 407.
      1LRATIONALE-307
      Enrolling squamous cell carcinoma only.
      ETisle + ChemoChemoOncoScreen10 mut/MbPFS0.44 (0.27–0.72)0.57 (0.36–0.91)32
      12 mut/Mb0.34 (0.19–0.62)0.61 (0.40–0.93)
      14 mut/Mb0.29 (0.13–0.65)0.57 (0.39–0.82)
      Anti–PD-1 + anti–CTLA-4 + chemo vs. chemo
      1LCM-9LAENivo + Ipi + ChemoChemoF1CDx10 mut/MbPFS0.74 (0.51–1.08)0.75 (0.55–1.02)33
      OS0.49 (0.34–0.70)0.83 (0.63–1.10)
      Anti–PD-L1 + anti–CTLA-4 vs. ICI
      1LMYSTICEDurva + TremeDurvaF1CDx10 mut/MbOS1.00 (0.65–1.54)1.09 (0.79–1.50)29
      Tx, therapy; L, line; E, exploratory; P, prospective; ICI, immune checkpoint inhibitor; chemo, chemotherapy; nivo, nivolumab; pembro, pembrolizumab; durva, durvalumab; atezo, atezolizumab; ipi, ipilimumab; treme, tremelimumab; tisle, tislelizumab; F1, FoundationOne; F1CDx, FoundationOne CDx; PD-1, programmed cell death protein 1; PFS, progression-free survival; TMB, tumor mutational burden; OS, overall survival; PD-1, programmed cell death protein 1; WES, whole exome sequencing; KN, KeyNote; CM, CheckMate.
      a Enrolling squamous cell carcinoma only.
      Limited data are available on the role of plasma or blood TMB (bTMB) as predictive marker for ICIs (Table 3). In the MYSTIC trial,
      • Rizvi N.A.
      • Cho B.C.
      • Reinmuth N.
      • et al.
      Durvalumab with or without tremelimumab vs standard chemotherapy in first-line treatment of metastatic non-small cell lung cancer: the MYSTIC Phase 3 randomized clinical trial.
      although high bTMB (≥20 mut/Mb) is associated with improved OS for durvalumab plus tremelimumab versus chemotherapy, this effect was not found in patients with less than 20 mut/Mb bTMB, but the analysis was retrospective and exploratory. When this cutoff was tested in the BR.34 trial that compared the efficacy of durvalumab plus tremelimumab with versus without platinum-doublet chemotherapy, bTMB was found to have no differential treatment effect on OS.
      • Leighl N.B.
      • Laurie S.A.
      • Goss G.D.
      • et al.
      CCTG BR34: A randomized Phase 2 trial of durvalumab and tremelimumab with or without platinum-based chemotherapy in patients with metastatic NSCLC.
      When ICI combined with chemotherapy was compared with chemotherapy alone, there have been conflicting results as to whether high bTMB can predict differential treatment effects on survival.
      • Wang J.
      • Lu S.
      • Hu C.
      • et al.
      Updated analysis of tislelizumab plus chemotherapy vs chemotherapy alone as first-line treatment of advanced squamous non-small cell lung cancer (SQ NSCLC).
      ,
      • Paz-Ares L.
      • Ciuleanu T.-E.
      • Cobo M.
      • et al.
      First-line nivolumab (NIVO) + ipilimumab (IPI) + 2 cycles chemotherapy (chemo) vs 4 cycles chemo in advanced non-small cell lung cancer (aNSCLC): association of blood and tissue tumor mutational burden (TMB) with efficacy in CheckMate 9LA.
      Table 3Blood-Based Evaluation of TMB as Predictive Biomarker for ICI Therapies in Phase 3 Trials Involving NSCLC
      Line of TxClinical TrialStudy TherapyReference TherapyAssay (Cutpoint)SurvivalHazard Ratio (95% Confidence Interval)Ref
      High TMBLow TMB
      Anti–PD-L1 ± anti–CTLA-4 vs. chemo
      2LOAKAtezoChemoF1CDx (16 mut/Mb)PFS0.65 (0.47–0.92)0.98 (0.80–1.20)23
      OS0.64 (0.44–0.92)0.65 (0.52–0.81)
      1LMYSTICDurvaChemoF1CDx (20 mut/Mb)PFS0.77 (0.52–1.13)1.19 (0.94–1.50)29
      OS0.72 (0.50–1.05)0.93 (0.74–1.16)
      1LIMpower110AtezoChemoF1CDX (16 mut/Mb)PFS0.55 (0.33–0.92)1.00 (0.78–1.29)17
      F1CDX (20 mut/Mb)PFS0.56 (0.30–1.06)0.95 (0.74–1.21)
      1LBFASTAtezoChemoF1CDX (16 mut/Mb)PFS0.77 (0.59–1.00)Not reported
      • Dziadziuszko R.
      • Peters S.
      • Gadgeel S.M.
      • et al.
      Atezolizumab (atezo) vs platinum-based chemo in blood-based tumour mutational burden-positive (bTMB+) patients (pts) with first-line (1L) advanced/metastatic (m)NSCLC: results of the Blood First Assay Screening Trial (BFAST) phase III cohort C.
      F1CDX (16 mut/Mb)OS0.87 (0.64–1.17)Not reported
      1LMYSTICDurva + TremeChemoF1CDx (20 mut/Mb)PFS0.53 (0.34–0.81)1.55 (1.23–1.94)29
      OS0.49 (0.32–0.74)1.16 (0.93–1.45)
      Anti–PD-L1 +/- anti–CTLA-4 vs. anti–PD-L1
      1LMYSTICDurva + TremeDurvaF1CDx (20 mut/Mb)PFS0.76 (0.50–1.15)1.26 (1.02–1.57)29
      OS0.74 (0.48–1.11)1.22 (0.98–1.52)
      Anti–PD-1 ± anti–CTLA-4 + chemo vs. chemo
      1LRATIONALE-307
      Enrolling squamous cell carcinoma only.
      Tisle + ChemoChemoOncoScreen (6 mut/Mb)PFS0.30 (0.13–0.67)0.63 (0.25–1.61)32
      OncoScreen (8 mut/Mb)PFS0.33 (0.14–0.75)0.55 (0.22–1.34)
      OncoScreen (10 mut/Mb)PFS0.30 (0.11–0.82)0.51 (0.23–1.14)
      1LCKM-9LANivo + Ipi + ChemoChemoGuardant OMNI (16 mut/Mb)PFS0.60 (0.42–0.86)0.73 (0.55–0.96)33
      OS0.55 (0.39–0.78)0.78 (0.60–1.00)
      Guardant OMNI (20 mut/Mb)PFS0.54 (0.35–0.84)0.74 (0.57–0.95)
      OS0.48 (0.32–0.73)0.78 (0.62–0.99)
      Anti-PD-L1 ± anti–CTLA-4 + chemo + ICI
      1LBR.34Durva + Treme + chemoDurva + TremeGuardant OMNI (20 mut/Mb)PFS0.96 (0.54–1.71)0.58 (0.42–0.81)31
      OS0.98 (0.53–1.80)0.81 (0.59–1.12)
      Tx, therapy; L, line; chemo, chemotherapy; ICI, immune checkpoint inhibitor; durva, durvalumab; atezo, atezolizumab; treme, tremelimumab; tisle, tislelizumab; nivo, nivolumab; ipi, ipilimumab; F1CDx, FoundationOne CDx; PD-1, programmed cell death protein 1; PD-L1, programmed death-ligand 1; PFS, progression-free survival; TMB, tumor mutational burden; OS, overall survival.
      a Enrolling squamous cell carcinoma only.
      Targeted gene panels are now routinely used for clinical identification of driver mutations using next-generation sequencing technology in many countries, but these panels vary widely in the number and composition of genes covered. Several in silico and actual panel sequencing studies have revealed strong correlation of TMB estimated using these panels compared with whole exome sequencing, and panels of greater than 1.1 or 1.5 Mb have been suggested as most appropriate for clinical TMB measurement.
      • Buchhalter I.
      • Rempel E.
      • Endris V.
      • et al.
      Size matters: dissecting key parameters for panel-based tumor mutational burden analysis.
      • Chalmers Z.R.
      • Connelly C.F.
      • Fabrizio D.
      • et al.
      Analysis of 100,000 human cancer genomes reveals the landscape of tumor mutational burden.
      • Endris V.
      • Buchhalter I.
      • Allgäuer M.
      • et al.
      Measurement of tumor mutational burden (TMB) in routine molecular diagnostics: in silico and real-life analysis of three larger gene panels.
      • Ramos-Paradas J.
      • Hernández-Prieto S.
      • Lora D.
      • et al.
      Tumor mutational burden assessment in non-small-cell lung cancer samples: results from the TMB2 harmonization project comparing three NGS panels.
      • Wei B.
      • Kang J.
      • Kibukawa M.
      • et al.
      Evaluation of the TruSight Oncology 500 assay for routine clinical testing of tumor mutational burden and clinical utility for predicting response to pembrolizumab.
      To develop guidelines for harmonizing TMB across clinical diagnostic platforms, the Friends of Cancer Research TMB Harmonization Consortium has completed and reported the results of their phase 1 (in silico assessment of variation) and phase 2 (analysis of tumor and cell line samples) studies involving 11 and 16 distinct laboratories, respectively. Results of the phase 1 study led to the consortium recommending a series of best practices for panel developers around the following three items: (1) reporting of TMB in mut/Mb to ensure consistency; (2) standardization of analytical validation studies for TMB estimation to include assessment of analytical accuracy, precision, and sensitivity; and (3) assurance of consistency across panels by alignment of panel TMB values to whole exome sequencing–derived universal reference standards.
      • Merino D.M.
      • McShane L.M.
      • Fabrizio D.
      • et al.
      Establishing guidelines to harmonize tumor mutational burden (TMB): in silico assessment of variation in TMB quantification across diagnostic platforms: phase I of the Friends of Cancer Research TMB Harmonization Project.
      The phase 2 study concluded that panel size, gene content, and bioinformatics pipelines may contribute to empirical variability observed in TMB estimation across different panels, and a software tool (http://brb.nci.nih.gov/tmbLab/) was developed and made publicly available to promote reproducibility and comparability across assays.
      • Merino D.M.
      • McShane L.M.
      • Fabrizio D.
      • et al.
      Establishing guidelines to harmonize tumor mutational burden (TMB): in silico assessment of variation in TMB quantification across diagnostic platforms: phase I of the Friends of Cancer Research TMB Harmonization Project.
      Given the limitations of TMB as a single predictive biomarker, there has been, for two main reasons,
      • Merino D.M.
      • McShane L.M.
      • Fabrizio D.
      • et al.
      Establishing guidelines to harmonize tumor mutational burden (TMB): in silico assessment of variation in TMB quantification across diagnostic platforms: phase I of the Friends of Cancer Research TMB Harmonization Project.
      renewed interest in the study of neoantigens, that is, mutated peptides originating from somatic mutations generating antitumor T-cell responses when presented by MHC proteins.
      • Ye L.
      • Creaney J.
      • Redwood A.
      • Robinson B.
      The current lung cancer neoantigen landscape and implications for therapy.
      First, the quantity and quality of the neoantigen response may improve the prediction of ICI treatments
      • Anagnostou V.
      • Smith K.N.
      • Forde P.M.
      • et al.
      Evolution of neoantigen landscape during immune checkpoint blockade in non-small cell lung cancer.
      • Fehlings M.
      • Jhunjhunwala S.
      • Kowanetz M.
      • et al.
      Late-differentiated effector neoantigen-specific CD8+ T cells are enriched in peripheral blood of non-small cell lung carcinoma patients responding to atezolizumab treatment.
      • Forde P.M.
      • Chaft J.E.
      • Smith K.N.
      • et al.
      Neoadjuvant PD-1 blockade in resectable lung cancer.
      • Gettinger S.N.
      • Choi J.
      • Mani N.
      • et al.
      A dormant TIL phenotype defines non-small cell lung carcinomas sensitive to immune checkpoint blockers.
      • Rizvi N.A.
      • Hellmann M.D.
      • Snyder A.
      • et al.
      Cancer immunology. Mutational landscape determines sensitivity to PD-1 blockade in non-small cell lung cancer.
      whereas loss of neoantigens is a mechanism of immune escape and ICI resistance, particularly in lung squamous cell carcinoma.
      • McGranahan N.
      • Furness A.J.
      • Rosenthal R.
      • et al.
      Clonal neoantigens elicit T cell immunoreactivity and sensitivity to immune checkpoint blockade.
      ,
      • Montesion M.
      • Murugesan K.
      • Jin D.X.
      • et al.
      Somatic HLA Class I loss is a widespread mechanism of immune evasion which refines the use of tumor mutational burden as a biomarker of checkpoint inhibitor response.
      Second, neoantigen-based vaccines and adoptive T-cell therapies are being developed.
      • Ye L.
      • Creaney J.
      • Redwood A.
      • Robinson B.
      The current lung cancer neoantigen landscape and implications for therapy.
      Because there is no standardized method for neoantigen identification, current methods rely on bioinformatic pipelines (HLA typing, inference of neopeptides, MHC-binding prediction, and candidate neoantigen selection).
      • Ye L.
      • Creaney J.
      • Redwood A.
      • Robinson B.
      The current lung cancer neoantigen landscape and implications for therapy.
      ,
      • De Mattos-Arruda L.
      • Vazquez M.
      • Finotello F.
      • et al.
      Neoantigen prediction and computational perspectives towards clinical benefit: recommendations from the ESMO Precision Medicine Working Group.
      In NSCLC, the median number of neoantigens per tumor has been predicted to be between 63 and 214.
      • Ye L.
      • Creaney J.
      • Redwood A.
      • Robinson B.
      The current lung cancer neoantigen landscape and implications for therapy.
      Not surprisingly, higher numbers are associated with high PD-L1 expression and smoking-related genomic signatures.
      • Gettinger S.N.
      • Choi J.
      • Mani N.
      • et al.
      A dormant TIL phenotype defines non-small cell lung carcinomas sensitive to immune checkpoint blockers.
      ,
      • Chen Y.P.
      • Zhang Y.
      • Lv J.W.
      • et al.
      Genomic analysis of tumor microenvironment immune types across 14 solid cancer types: immunotherapeutic implications.
      Importantly, nearly half of NSCLC harbor targetable “driver” alterations,
      WHO Classification of Tumours Editorial Board
      WHO Classification oF Tumours.
      and of those, poor response to ICIs in ALK-rearranged or EGFR-mutated patients has been reported,
      • Gainor J.F.
      • Shaw A.T.
      • Sequist L.V.
      • et al.
      EGFR mutations and ALK rearrangements are associated with low response rates to PD-1 pathway blockade in non-small cell lung cancer: A retrospective analysis.
      whereas MET exon 14 skipping mutations could confer ICI sensitivity.
      • Kauffmann-Guerrero D.
      • Tufman A.
      • Kahnert K.
      • et al.
      Response to checkpoint inhibition in non-small cell lung cancer with molecular driver alterations.
      In addition, tumors harboring KRAS and TP53 co-mutations are associated with a T-cell–enriched TME. In contrast, concurrent STK11/LKB1 and or KEAP1 mutations with KRAS mutations harbor a TME characterized by a T-cell exclusion, with low or no PD-L1 expression, down-regulation of MHC class II for KRASMUTSTK11MUT adenocarcinoma, and down-regulation of positive regulators of type I interferon (IFN) and other cytokines for KRASMUTKEAP1MUT adenocarcinoma.
      • Ricciuti B.
      • Arbour K.C.
      • Lin J.J.
      • et al.
      Diminished efficacy of programmed death-(ligand)1 inhibition in STK11- and KEAP1-mutant lung adenocarcinoma is affected by KRAS mutation status.
      ,
      • Skoulidis F.
      • Goldberg M.E.
      • Greenawalt D.M.
      • et al.
      STK11/LKB1 mutations and PD-1 inhibitor resistance in KRAS-mutant lung adenocarcinoma.
      Both mutations confer a worse prognosis and resistance to ICIs in KRAS-mutant tumors.
      • Ricciuti B.
      • Arbour K.C.
      • Lin J.J.
      • et al.
      Diminished efficacy of programmed death-(ligand)1 inhibition in STK11- and KEAP1-mutant lung adenocarcinoma is affected by KRAS mutation status.
      Among other genetic abnormalities, truncating mutations and loss of heterozygosity of B2M are associated with acquired resistance to anti–PD-1 therapy.
      • Zhao Y.
      • Cao Y.
      • Chen Y.
      • et al.
      B2M gene expression shapes the immune landscape of lung adenocarcinoma and determines the response to immunotherapy.
      Epigenetic mechanisms can affect ICI responses mainly through promoter methylation of genes encoding immune checkpoints or tumor-associated antigens and hampering antigen presentation, migration of T cells, and cytokine secretion. A DNA methylation signature called EPIMMUNE has been reported to be associated with ICI response and survival in NSCLC.
      • Duruisseaux M.
      • Martínez-Cardús A.
      • Calleja-Cervantes M.E.
      • et al.
      Epigenetic prediction of response to anti-PD-1 treatment in non-small-cell lung cancer: a multicentre, retrospective analysis.

      Tumor Immune Microenvironment

      The development of anticancer immunotherapies has benefited from advances in our knowledge of cancer immunity. The cancer-immune interaction represents a dynamic spatiotemporal process that involves multiple stimulatory factors and inhibitors acting at different stages of a cycle, as summarized by Chen and Mellman.
      • Chen D.S.
      • Mellman I.
      Oncology meets immunology: the cancer-immunity cycle.
      ICIs, the current mainstay of immunotherapies, aim to enhance killing of cancer cells by reactivating suppressed effector T cells in the TME. From this perspective, the TME can be generally classified into T-cell–inflamed versus non–T-cell–inflamed environments with the former being further classified on the basis of the activation status of the T cells. Notably, activated T-cell infiltration in the TME (T-cell–inflamed TME) is most often accompanied by an IFN-γ–driven adaptive immune resistance phenotype characterized by up-regulation of immune-regulatory pathways including immune-inhibitory receptors, such as PD-1, LAG-3, TIM-3, VISTA, and TIGIT, other inhibitory molecules, such as IDO-1, TGFB1, and INOS, and or expansion of immune-inhibitory populations, such as regulatory T cells (Tregs) and myeloid-derived suppressor cells (MDSCs).
      • Doroshow D.B.
      • Sanmamed M.F.
      • Hastings K.
      • et al.
      Immunotherapy in non-small cell lung cancer: facts and hopes.
      • Karasaki T.
      • Nagayama K.
      • Kuwano H.
      • et al.
      An immunogram for the cancer-immunity cycle: towards personalized immunotherapy of lung cancer.
      • Sanmamed M.F.
      • Eguren-Santamaria I.
      • Schalper K.A.
      Overview of lung cancer immunotherapy.
      The presence of an abundant CD8+ cytotoxic T-cell infiltrate has been one of the most studied biomarkers, with multiple studies on the basis of IHC, mIF, or mRNA signatures.
      • Sholl L.M.
      Biomarkers of response to checkpoint inhibitors beyond PD-L1 in lung cancer.
      • Shepherd D.J.
      • Tabb E.S.
      • Kunitoki K.
      • et al.
      Lymphocyte-activation gene 3 in non-small-cell lung carcinomas: correlations with clinicopathologic features and prognostic significance.
      • Shirasawa M.
      • Yoshida T.
      • Imabayashi T.
      • et al.
      Baseline PD-L1 expression and tumour-infiltrated lymphocyte status predict the efficacy of durvalumab consolidation therapy after chemoradiotherapy in unresectable locally advanced patients with non-small-cell lung cancer.
      • Shirasawa M.
      • Yoshida T.
      • Shimoda Y.
      • et al.
      Differential immune-related microenvironment determines programmed cell death Protein-1/Programmed death-ligand 1 blockade efficacy in patients with advanced NSCLC.
      Unfortunately, each study has applied its own scoring method and cutoffs for high versus low CD8+ T cells, and patient populations were heterogeneous in size and characteristics. Evaluation of the activation state of the cytotoxic cells has also been proposed as an additional marker. For instance, co-expression of CD39, a marker of tumor-antigen-based activation, on CD8+ tumor-infiltrating T cells has been associated with up-regulation of both proliferation and T-cell dysfunction markers and increased response to PD-1 axis blockade.
      • Yeong J.
      • Suteja L.
      • Simoni Y.
      • et al.
      Intratumoral CD39+CD8+ T cells predict response to programmed cell death Protein-1 or programmed death Ligand-1 blockade in patients with NSCLC.
      High expression of PD-1, another antigen-based activation and dysfunction marker, on CD8+ tumor-infiltrating T cells was predictive of both response and survival in NSCLC treated with anti–PD-1 ICIs.
      • Thommen D.S.
      • Koelzer V.H.
      • Herzig P.
      • et al.
      A transcriptionally and functionally distinct PD-1+ CD8+ T cell pool with predictive potential in non-small-cell lung cancer treated with PD-1 blockade.
      Given that PD-1 axis blockade not only induces recovery of dysfunctional PD-1+ CD8+ T cells but also enhances PD-1+ Treg cell-mediated immunosuppression, a profound reactivation of effector PD-1+ CD8+ T cells rather than PD-1+ Treg cells may be necessary for tumor regression.
      • Kumagai S.
      • Togashi Y.
      • Kamada T.
      • et al.
      The PD-1 expression balance between effector and regulatory T cells predicts the clinical efficacy of PD-1 blockade therapies.
      The assessment of T-cell dysfunction could also include evaluation of a variety of markers including those for antigen-based activation (CD28, CD39, CD103, CD137, PD-1), regulation (checkpoint: PD-1, LAG-3, TIM-3), differentiation (CD45RO, TBET, EOMES), proliferation (Ki-67), and apoptosis (FAS, BIM).
      • Datar I.
      • Sanmamed M.F.
      • Wang J.
      • et al.
      Expression analysis and significance of PD-1, LAG-3, and TIM-3 in human non-small cell lung cancer using spatially resolved and multiparametric single-cell analysis.
      • Eiva M.A.
      • Omran D.K.
      • Chacon J.A.
      • Powell Jr., D.J.
      Systematic analysis of CD39, CD103, CD137, and PD-1 as biomarkers for naturally occurring tumor antigen-specific TILs.
      • Gide T.N.
      • Quek C.
      • Menzies A.M.
      • et al.
      Distinct immune cell populations define response to anti-PD-1 monotherapy and anti-PD-1/anti-CTLA-4 combined therapy.
      Some of these markers assessed with mIF have were found to have associations with survival after treatment with PD-1 or PD-L1 inhibitors in patients with NSCLC.
      • Gettinger S.N.
      • Choi J.
      • Mani N.
      • et al.
      A dormant TIL phenotype defines non-small cell lung carcinomas sensitive to immune checkpoint blockers.
      ,
      • Datar I.
      • Sanmamed M.F.
      • Wang J.
      • et al.
      Expression analysis and significance of PD-1, LAG-3, and TIM-3 in human non-small cell lung cancer using spatially resolved and multiparametric single-cell analysis.
      Among the multiple molecules and pathways, some may become predominant at a given time and thus considered as immune evasion “drivers.”
      • Schalper K.A.
      • Carvajal-Hausdorf D.
      • McLaughlin J.
      • et al.
      Differential expression and significance of PD-L1, IDO-1, and B7-H4 in human lung cancer.
      ,
      • Zhang M.L.
      • Kem M.
      • Mooradian M.J.
      • et al.
      Differential expression of PD-L1 and IDO1 in association with the immune microenvironment in resected lung adenocarcinomas.
      For instance, elevated LAG-3 expression was found to be associated with either shorter
      • Datar I.
      • Sanmamed M.F.
      • Wang J.
      • et al.
      Expression analysis and significance of PD-1, LAG-3, and TIM-3 in human non-small cell lung cancer using spatially resolved and multiparametric single-cell analysis.
      or longer progression-free survival (PFS).
      • Shepherd D.J.
      • Tabb E.S.
      • Kunitoki K.
      • et al.
      Lymphocyte-activation gene 3 in non-small-cell lung carcinomas: correlations with clinicopathologic features and prognostic significance.
      The determinants for this apparent contradiction remain unknown, but multiple factors may be involved including the use of different cohorts, tissue imaging platforms applied, and marker cutoffs. In addition, the strong association of PD-L1 and LAG-3 expression observed in one of the studies
      • Shepherd D.J.
      • Tabb E.S.
      • Kunitoki K.
      • et al.
      Lymphocyte-activation gene 3 in non-small-cell lung carcinomas: correlations with clinicopathologic features and prognostic significance.
      suggested that the PD-1/PD-L1 pathway was the predominant “driver” immune-regulatory pathway. The possible role of the recently discovered LAG-3 ligand FGL1, known to be expressed in a fraction of NSCLCs, could be another determinant for these responses.
      • Wang J.
      • Sanmamed M.F.
      • Datar I.
      • et al.
      Fibrinogen-like Protein 1 is a major immune inhibitory ligand of LAG-3.
      The local expansion of immune-suppressive populations in the TME, such as Treg, MDSC, and M2-polarized macrophages, in association with various immunosuppressive cytokines, such as interleukin-6, interleukin-8, and transforming growth factor-β is another mechanism for an adaptive immune resistance phenotype.
      • Sanmamed M.F.
      • Eguren-Santamaria I.
      • Schalper K.A.
      Overview of lung cancer immunotherapy.
      Of those, MDSCs, a heterogeneous immature population of myeloid cells, contribute to resistance to ICIs by targeting effector T cells leading to T-cell dysfunction, promoting tumor angiogenesis, favoring an immunosuppressive network with M2 macrophage polarization, and Treg expansion.
      • Hou A.
      • Hou K.
      • Huang Q.
      • Lei Y.
      • Chen W.
      Targeting myeloid-derived suppressor cell, a promising strategy to overcome resistance to immune checkpoint inhibitors.
      The peripheral blood is also a potential source of biologically relevant information reflecting patient immune status. Circulating peripheral blood mononuclear cells assessed by flow cytometry can predict the efficacy of anti–PD-1 immunotherapy, with changes after treatment in percentage of various immune cells, including CD4+ T cells, CD8+ T cells, MDSCs, regulatory T cells, and PD-1–expressing T cells.
      • Zhuo M.
      • Chen H.
      • Zhang T.
      • et al.
      The potential predictive value of circulating immune cell ratio and tumor marker in atezolizumab treated advanced non-small cell lung cancer patients.
      • Kang D.H.
      • Chung C.
      • Sun P.
      • et al.
      Circulating regulatory T cells predict efficacy and atypical responses in lung cancer patients treated with PD-1/PD-L1 inhibitors.
      • Rogado J.
      • Pozo F.
      • Troule K.
      • et al.
      Peripheral blood mononuclear cells predict therapeutic efficacy of immunotherapy in NSCLC.
      In addition, soluble immune checkpoint-related proteins in the blood are associated with invasion and progression in NSCLC
      • Wang Q.
      • He Y.
      • Li W.
      • et al.
      Soluble immune checkpoint-related proteins in blood are associated with invasion and progression in non-small cell lung cancer.
      and harbor a predictive value in patients treated with immunotherapy.
      • Mildner F.
      • Sopper S.
      • Amann A.
      • et al.
      Systematic review: soluble immunological biomarkers in advanced non-small-cell lung cancer (NSCLC).
      • Mazzaschi G.
      • Minari R.
      • Zecca A.
      • et al.
      Soluble PD-L1 and Circulating CD8+PD-1+ and NK cells enclose a prognostic and predictive immune effector score in immunotherapy treated NSCLC patients.
      • Honrubia-Peris B.
      • Garde-Noguera J.
      • Garcia- Sánchez J.
      • Piera-Molons N.
      • Llombart-Cussac A.
      • Fernández-Murga M.L.
      Soluble biomarkers with prognostic and predictive value in advanced non-small cell lung cancer treated with immunotherapy.

      Transcriptome Signatures

      Many transcriptomic signatures predictive of response or resistance to ICIs have been published in recent years (Table 4). These signatures are based on mRNA data obtained by targeted or whole transcriptome RNA-sequencing and retrieved from either public sources or specific studies. They have been validated on different cohorts and data sets and with different algorithms, including gene set enrichment analysis, to ascertain the gene function or to determine whether a predefined set of genes could be statistically relevant .
      Table 4Main Transcriptomic Signatures Predictive of Response or Resistance to ICIs in NSCLC
      StudiesTumors (No. of Cases With RNAseq Data Available)Platforms or PanelsTranscriptomic Signatures (GEP)Pathways InvolvedClinical Validation p Values, AUC, and/or Hazard Ratios (95% Confidence Interval)
      Ott et al.
      • Ott P.A.
      • Bang Y.J.
      • Piha-Paul S.A.
      • et al.
      T-cell-inflamed gene-expression profile, programmed death ligand 1 expression, and tumor mutational burden predict efficacy in patients treated with pembrolizumab across 20 cancers: KEYNOTE-028.
      20 cancers including SCLC (n = 8)NanoString platform (custom 680-gene panel)18-gene T-cell–inflamed GEP (CD3D, IDO1, CIITA, CD3E, CCL5, GZMK, CD2, HLA-DR, CXCL13, IL2RG, NKG7, HLA-E, CXCR6, LAG3, TAGAP, CXCL10, STAT1, GZMB)T-cell–activated TME, IFN-γ signaling, antigen presentation, chemokines, T-cell cytotoxic activity, and adaptive immunityHigher scores associated with ORR (p = 0.012) and longer PFS (p = 0.017) in patients treated with pembrolizumab (KEYNOTE-028 trial)
      Fehrenbacher et al.
      • Fehrenbacher L.
      • Spira A.
      • Ballinger M.
      • et al.
      Atezolizumab versus docetaxel for patients with previously treated non-small-cell lung cancer (POPLAR): a multicentre, open-label, phase 2 randomised controlled trial.
      NSCLC (n = 224)Fluidigm-based gene expression platform8-Gene T-effector and IFN-γ GEP (CD8A, GZMA, GZMB, IFNγ, EOMES, CXCL9, CXCL10, and TBX21)T-effector and IFN-γ signalingHigher Teff/IFN-γ scores associated improved OS (HR 0·43 [0.24–0.77]) vs. Teff/IFN-γ low (HR 1·10 [0.68–1.76]) in patients treated with atezolizumab (POPLAR trial)
      Higgs et al.
      • Higgs B.W.
      • Morehouse C.A.
      • Streicher K.
      • et al.
      Interferon gamma messenger RNA signature in tumor biopsies predicts outcomes in patients with non-small cell lung carcinoma or urothelial cancer treated with durvalumab.
      NSCLC (n = 97)Pan-transcriptome sequencing4-Gene IFN-γ GEP (IFNγ, CD274, LAG3, and CXCL9)IFN-γ signalingIFN-γ+ (vs. IFN-γ−) scores associated ORR 37.5 (21.7–56.3) vs. 6.2 (2.0–15.8); median OS: 22.7 mo (9.5–NR) vs. 6.5 (4.3–14.2); median PFS: 7.5 mo (2.7–14.6) vs. 1.4 (1.3–2.4) in patients treated with durvalumab
      Damotte et al.
      • Damotte D.
      • Warren S.
      • Arrondeau J.
      • et al.
      The tumor inflammation signature (TIS) is associated with anti-PD-1 treatment benefit in the CERTIM pan-cancer cohort.
      NSCLC (n = 38)NanoStringPanCancer IO 360-gene panel5-Gene TIS (CXCL9, CXL10, CXCL11, TAP1, and PSMB9)IFN-γ signaling and antigen processingHigh TIS scores associated with improved OS (HR = 0.36 [0.14, 0.90], p = 0.02) in PD-1 inhibitor responders
      Hwang et al.
      • Hwang S.
      • Kwon A.Y.
      • Jeong J.Y.
      • et al.
      Immune gene signatures for predicting durable clinical benefit of anti-PD-1 immunotherapy in patients with non-small cell lung cancer.
      NSCLC (n = 21)Oncomine Immune Response Research Assay (395 immune-related gene panel)M1 (CBLBCCR7CD27CD48FOXO1FYB, HLA-BHLAGIFIH1IKZF4LAMP3NFKBIA, and SAMHD1) and peripheral T cell (HLA-DOAGPR18, and STAT1 ) signaturesT-cell activation, antigen presentation, tumor-associated macrophagesLonger PFS associated with high M1 and peripheral T-cell GEP scores (p = 7.84e−5 and p = 8.29e−3) in patients treated with anti–PD-1 monotherapy.

      Positive predictive values (AUC) for peripheral T-cell signature: 0.94.
      Ranganath et al.
      • Ranganath H.
      • Jain A.L.
      • Smith J.R.
      • et al.
      Association of a novel 27-gene immuno-oncology assay with efficacy of immune checkpoint inhibitors in advanced non-small cell lung cancer.
      NSCLC (n = 67)In-house RT-qPCR panel27-Gene IO scoreIO score associated with PFS (HR 0.21 [0.085–0.54], p < 0.001)
      Leng et al.
      • Leng Y.
      • Dang S.
      • Yin F.
      • et al.
      GDPLichi: a DNA damage repair-related gene classifier for predicting lung adenocarcinoma immune checkpoint inhibitors response.
      526 TCGA-LUAD and 438 LUAD cohort data sets (GSE30219, 3121, 50081)Pan-transcriptome sequencing7 Genes (DUT, TYMS, YWHAG, MGMT, POLH, RAD1, and RAD17) from 8 DDR GEPsDDR pathwaysLow-risk score associated with better survival (HR 1.912 [1.421–2573]); positive predictive values (AUC) 0.71
      Jang et al.
      • Jang H.J.
      • Lee H.S.
      • Ramos D.
      • et al.
      Transcriptome-based molecular subtyping of non-small cell lung cancer may predict response to immune checkpoint inhibitors.
      NSCLC (n = 87 LUADs, 101 SCCs, and 35 NSCLCs for validation)Pan-transcriptome sequencing and validation with the Nanostring platform (PanCancer Immune Profiling Panel of 770 mRNA sets)59-Gene signature (IR score)Good TiME (higher expression of immune-regulatory molecules, increased cytolytic activity, higher interferon-g signature, and abundant immune cells) vs. bad TiME“Good-TiME” associated with high response score to PD-1 inhibitors; positive predictive values (AUC) 0.702 (p = 0.039)
      Budczies et al.
      • Budczies J.
      • Kirchner M.
      • Kluck K.
      • et al.
      A gene expression signature associated with B cells predicts benefit from immune checkpoint blockade in lung adenocarcinoma.
      TCGA-LUAD cohort and validation in 43 treated LUADsPan-transcriptome sequencing and validation with the Nanostring platform (PanCancer Immune Profiling Panel of 770 mRNA sets)8-Gene B-cell expression signature (BLK, CD19, FCRL2, MA4A1, TNFRSF17, TCL1A, SPIB, and PNOC)B-cell immune responsesProlonged PFS after ICI treatment associated with abundance of B cells (HR = 0.66, p = 0.00074), CD45+ cells (HR = 0.61, p = 0.01), and total TILs (HR = 0.62, p = 0.025)
      HR, hazard ratio; DDR, DNA damage response; ICI, immune checkpoint inhibitor; GEP, gene expression profiling; LUAD, lung adenocarcinoma; SCC, lung squamous cell carcinoma; PD-1, programmed cell death protein 1; TiME, tumor immune microenvironment; TIS, tumor inflammation signature.
      These predictive transcriptomic signatures consider sets of the genes involved in tumor antigenicity and T-cell priming and activation by interaction with activated dendritic cells, trafficking and infiltration of the T cells into tumors (CXCL9, CXCL10, CCL5 genes, among others), recognition of cancer cells by the T cells (HLA-A, HLA-B, HLA-C, B2M genes), infiltration by the inhibitory cells (such as MDSCs and Tregs) or molecules, and immune checkpoint receptor or ligand-encoding genes.
      Many predictive signatures emphasize the role of T-cell inflammation and dysfunction including IFN-γ–related genes and T effector or immune cytolytic activities. Ayers et al.
      • Ayers M.
      • Lunceford J.
      • Nebozhyn M.
      • et al.
      IFN-γ-related mRNA profile predicts clinical response to PD-1 blockade.
      proposed, in a large cohort of pembrolizumab-treated patients across nine different tumor types, a T-cell–inflamed gene expression profiling consisting of 18 genes involved in T-cell–activated TME, IFN-γ signaling, antigen presentation, chemokines, T-cell cytotoxic activity, and adaptive immunity. This signature has been validated as an independent predictive biomarker of pembrolizumab monotherapy in a pan-cancer clinical trial performed across 20 tumor types, including SCLC (KEYNOTE-028).
      • Ott P.A.
      • Bang Y.J.
      • Piha-Paul S.A.
      • et al.
      T-cell-inflamed gene-expression profile, programmed death ligand 1 expression, and tumor mutational burden predict efficacy in patients treated with pembrolizumab across 20 cancers: KEYNOTE-028.
      ,
      • Trujillo J.A.
      • Sweis R.F.
      • Bao R.
      • Luke J.J.
      T cell-Inflamed versus non-T cell-Inflamed Tumors: A conceptual framework for cancer immunotherapy drug development and combination therapy selection.
      A high T-effector-IFN-γ–associated gene expression was also associated with improved outcomes with atezolizumab and durvalumab.
      • Fehrenbacher L.
      • Spira A.
      • Ballinger M.
      • et al.
      Atezolizumab versus docetaxel for patients with previously treated non-small-cell lung cancer (POPLAR): a multicentre, open-label, phase 2 randomised controlled trial.
      ,
      • Higgs B.W.
      • Morehouse C.A.
      • Streicher K.
      • et al.
      Interferon gamma messenger RNA signature in tumor biopsies predicts outcomes in patients with non-small cell lung carcinoma or urothelial cancer treated with durvalumab.
      Damotte et al.
      • Damotte D.
      • Warren S.
      • Arrondeau J.
      • et al.
      The tumor inflammation signature (TIS) is associated with anti-PD-1 treatment benefit in the CERTIM pan-cancer cohort.
      also reported a tumor inflammation signature associated with benefits from anti–PD-1 monotherapy. This signature was based on five genes related to IFN-γ signaling and antigen processing (CXCL9, CXL10, CXCL11, TAP1, and PSMB9), and its expression was significantly higher in responders to PD-1 inhibitors. Similarly, macrophage M1 signature, peripheral T-cell signature, and high mRNA expression levels of CD137 and PSMB9 were more predictive of response to anti–PD-1 monotherapy than PD-L1 IHC, TMB, or tumor-infiltrating lymphocytes (TILs).
      • Hwang S.
      • Kwon A.Y.
      • Jeong J.Y.
      • et al.
      Immune gene signatures for predicting durable clinical benefit of anti-PD-1 immunotherapy in patients with non-small cell lung cancer.
      Very recently, a 27-gene IO test score (IO score) independent of PD-L1 TPS and TMB helped in identifying patients benefiting from IO in the clinical setting.
      • Ranganath H.
      • Jain A.L.
      • Smith J.R.
      • et al.
      Association of a novel 27-gene immuno-oncology assay with efficacy of immune checkpoint inhibitors in advanced non-small cell lung cancer.
      Another group has investigated the predictive performance of DNA Damage Response (DDR) gene expression profiles along with markers of immune dysfunction and exclusion.
      • Leng Y.
      • Dang S.
      • Yin F.
      • et al.
      GDPLichi: a DNA damage repair-related gene classifier for predicting lung adenocarcinoma immune checkpoint inhibitors response.
      Using a signature based on seven genes from eight DDR pathways, lung adenocarcinomas with low expression of DDR genes such as DUT, TYMS, and YWHAG but high expression of MGMT, POLH, RAD1, and RAD17 harbored better survival. They were also considered more sensitive to ICIs on the basis of the Tumor Immune Dysfunction and Exclusion algorithm proposed by Jiang et al.
      • Jiang P.
      • Gu S.
      • Pan D.
      • et al.
      Signatures of T cell dysfunction and exclusion predict cancer immunotherapy response.
      These tumors also exhibited increased expressions of immune-inhibitory genes, such as TIM-3, IDO1, LAG3, PD-L2, TIGIT, CD276, CD160, VEGFA, VEGFB, SLAMF7, KIR2DL3, and IL1B, and decreased levels of B cells, CD8+ T cells, hematopoietic stem cells, and myeloid dendritic cells.
      • Leng Y.
      • Dang S.
      • Yin F.
      • et al.
      GDPLichi: a DNA damage repair-related gene classifier for predicting lung adenocarcinoma immune checkpoint inhibitors response.
      Jang et al.
      • Jang H.J.
      • Lee H.S.
      • Ramos D.
      • et al.
      Transcriptome-based molecular subtyping of non-small cell lung cancer may predict response to immune checkpoint inhibitors.
      have proposed a comprehensive signature of 59 genes from the Nanostring nCounter PanCancer Immune Profiling Panel of 770 mRNA set. Tumors with a “bad TiME” (Tumor immune Micro Environment) responsible for poor responses to ICIs were characterized by decreased expressions of MHC-I and MHC-II molecules, including HLA-A, -B, -C, and -DRB1, decreased tumor-associated dendritic cells and CD8+ T cells, an increased number of Tregs, and a low M1/M2 macrophage ratio.
      • Jang H.J.
      • Lee H.S.
      • Ramos D.
      • et al.
      Transcriptome-based molecular subtyping of non-small cell lung cancer may predict response to immune checkpoint inhibitors.
      In contrast, tumors with a “good TiME,” indicating good response to anti–PD-1 treatment, were found to have up-regulation of PD-L1, PD-L2, CD8A, PD-1, CTLA-4, BTLA, CD40, CD40L, LAG3, TIGIT, TIM3, and VISTA. This good TiME also presented a high cytolytic activity and an increased IFN-γ signaling pathway expression. Memory B cells were also abundant, in contrast to Tregs. Interestingly, the predictive role of the B cells, complementary to total TILs, was also found with a B-cell expression signature, consisting of eight marker genes (BLK, CD19, FCRL2, MA4A1, TNFRSF17, TCL1A, SPIB, and PNOC), associated with response to ICIs.
      • Budczies J.
      • Kirchner M.
      • Kluck K.
      • et al.
      A gene expression signature associated with B cells predicts benefit from immune checkpoint blockade in lung adenocarcinoma.

      New Technologies for Evaluation of TME

      Given the plethora of immune-cell markers and cytokines involved in prediction of response or resistance to ICIs, multiplexed platforms have been widely used to simultaneously evaluate those predictive markers and determine the activation or dysfunction status of effector T cells, “driver” immune-inhibitory pathways, and extent of immune-inhibitory populations in the TME. Novel methods in development in immuno-oncology include the following: (1) liquid biopsy using multimodal cell-free or circulating tumor DNA (ctDNA) analysis; (2) highly multiplexed spatially resolved tissue analysis; and (3) high-dimensionality single-cell proteomic and or transcriptomic studies from cell suspensions. In general, these technologies require sophisticated instrumentation and relatively complex informatics for their analysis and interpretation. Therefore, they have been largely used in the research setting and for biomarker discovery. Nevertheless, the rapid advancement in their commercial availability, standardization, and progressive reduction in costs have allowed their incorporation into clinical trials.
      The biomarker role of ctDNA in immuno-oncology has been extensively discussed elsewhere,
      • Cabel L.
      • Proudhon C.
      • Romano E.
      • et al.
      Clinical potential of circulating tumour DNA in patients receiving anticancer immunotherapy.
      • Goldberg S.B.
      • Narayan A.
      • Kole A.J.
      • et al.
      Early assessment of lung cancer immunotherapy response via circulating tumor DNA.
      • Nabet B.Y.
      • Esfahani M.S.
      • Moding E.J.
      • et al.
      Noninvasive early identification of therapeutic benefit from immune checkpoint inhibition.
      • Zhang Q.
      • Luo J.
      • Wu S.
      • et al.
      Prognostic and predictive impact of circulating tumor DNA in patients with advanced cancers treated with immune checkpoint blockade.
      and targeted ctDNA panels are currently used in the clinic for identification of gene mutations, measurement of TMB, and monitoring of disease response after treatment (e.g., minimal residual disease or assessment of “molecular response”). Newer modalities with biomarker potential beyond mutation analysis include the assessment of ctDNA methylation marks and fragmentation patterns and identification of oncogenic viral sequences that seem to provide additional and complementary information.
      • Barefoot M.E.
      • Loyfer N.
      • Kiliti A.J.
      • McDeed 4th, A.P.
      • Kaplan T.
      • Wellstein A.
      Detection of cell types contributing to cancer from circulating, cell-free methylated DNA.
      • Cristiano S.
      • Leal A.
      • Phallen J.
      • et al.
      Genome-wide cell-free DNA fragmentation in patients with cancer.
      • Jiang P.
      • Sun K.
      • Peng W.
      • et al.
      Plasma DNA end-motif profiling as a Fragmentomic marker in cancer, pregnancy, and transplantation.
      • Keller L.
      • Belloum Y.
      • Wikman H.
      • Pantel K.
      Clinical relevance of blood-based ctDNA analysis: mutation detection and beyond.
      • Lianidou E.
      Detection and relevance of epigenetic markers on ctDNA: recent advances and future outlook.
      The mIF enables simultaneous mapping of eight to 10 markers in one formalin-fixed, paraffin-embedded section. This technique remains limited compared with highly multiplexed spatially resolved technologies but seems more adaptated to predictive marker validation in clinics, provided that they meet the guidelines of the Society for Immunotherapy of Cancer.
      • Taube J.M.
      • Akturk G.
      • Angelo M.
      • et al.
      The Society for Immunotherapy of Cancer statement on best practices for multiplex immunohistochemistry (IHC) and immunofluorescence (IF) staining and validation.
      In contrast, highly multiplexed, spatially resolved technologies are dedicated to biomarker discovery and can accommodate from 10 to 1000 markers, using a wide array of chemistries and signal detection systems. Examples include iterative cycling immunofluorescence protocols using primary fluorescence-conjugated antibodies and computational image alignment/integration (e.g., Cell DIVE/MultiOmyx
      • Gerdes M.J.
      • Sevinsky C.J.
      • Sood A.
      • et al.
      Highly multiplexed single-cell analysis of formalin-fixed, paraffin-embedded cancer tissue.
      and t-CyCIF
      • Lin J.R.
      • Izar B.
      • Wang S.
      • et al.
      Highly multiplexed immunofluorescence imaging of human tissues and tumors using t-CyCIF and conventional optical microscopes.
      ); cycling protocols using antibodies conjugated to DNA barcodes for signal detection (e.g., CODEX
      • Goltsev Y.
      • Samusik N.
      • Kennedy-Darling J.
      • et al.
      Deep profiling of mouse splenic architecture with CODEX multiplexed imaging.
      ); mass spectrometry-based analysis of metal-conjugated antibodies and focal high-energy tissue ionization (e.g., imaging mass cytometry
      • Giesen C.
      • Wang H.A.
      • Schapiro D.
      • et al.
      Highly multiplexed imaging of tumor tissues with subcellular resolution by mass cytometry.
      and multiple ion beam imaging
      • Angelo M.
      • Bendall S.C.
      • Finck R.
      • et al.
      Multiplexed ion beam imaging of human breast tumors.
      ); localized mRNA or protein signal detection and counting using primary antibodies conjugated with photocleavable fluorescent DNA tags (e.g., digital spatial profiler/GeoMx
      • Amaria R.N.
      • Reddy S.M.
      • Tawbi H.A.
      • et al.
      Neoadjuvant immune checkpoint blockade in high-risk resectable melanoma.
      ) (Fig. 1); and spatial transcriptomics (sometimes also including proteins) using in situ mRNA hybridization and capture on tumor sections coupled to individual positional transcript barcoding and next-generation sequencing (e.g., VISIUM,
      • Ståhl P.L.
      • Salmén F.
      • Vickovic S.
      • et al.
      Visualization and analysis of gene expression in tissue sections by spatial transcriptomics.
      SlideSeq,
      • Rodriques S.G.
      • Stickels R.R.
      • Goeva A.
      • et al.
      Slide-seq: A scalable technology for measuring genome-wide expression at high spatial resolution.
      Dbitseq,
      • Liu Y.
      • Yang M.
      • Deng Y.
      • et al.
      High-spatial-resolution multi-omics sequencing via deterministic barcoding in tissue.
      and SeqScope
      • Cho C.S.
      • Xi J.
      • Si Y.
      • et al.
      Microscopic examination of spatial transcriptome using Seq-Scope.
      ). As an example, a recent analysis of 39-plex immune-related proteins on the GeoMx platform of 53 NSCLC tissue microarrays revealed that the level of stromal CD56-positive natural killer (NK)/NK T cells was significantly associated with better survival after PD-1 axis blockade.
      • Zugazagoitia J.
      • Gupta S.
      • Liu Y.
      • et al.
      Biomarkers associated with beneficial PD-1 checkpoint blockade in non-small cell lung cancer (NSCLC) identified using high-plex digital spatial profiling.
      Another study that used a 37-plex imaging mass cytometry panel, including tumor- and immune-cell markers (Fig. 2), to evaluate 84 tissue microarrays of pretreatment NSCLCs revealed a prediction of durable clinical benefit with 97.3% mean accuracy and consistency using multidimensional markers and spatial patterns.
      • Lin Y.E.
      • Shnitzer T.
      • Talmon R.
      • et al.
      Graph of graphs analysis for multiplexed data with application to imaging mass cytometry.
      Figure thumbnail gr1
      Figure 1Examples of DSP protein analysis. Selected ROIs for molecular quantification are revealed for a NSCLC tumor sample (pan-cytokeratin in green, CD45 in red). We can design ROIs containing only immune cells (ex ROI001), only tumor cells (not found), or regions that contain both immune cells and tumor cells (ex ROI004). In that case, an automatic segmentation, on the basis of the expression of panCK and CD45, allows the identification of two regions (in light green and purple, respectively, on the mixed ROI image) that will be analyzed independently. The selected proteins will be quantified in each region and the results used for differential expression analysis (among ROIs with different localization, different composition, among samples, etc.). DSP, digital spatial profiler; panCK, pancytokeratin; ROI, region of interest.
      Figure thumbnail gr2
      Figure 2(A) Representative images revealing a human FFPE NSCLC sample stained with multiplex immunofluorescence using DAPI, AE1-AE3 panCK, and CD3 with metal-conjugated antibodies (left panels) and subsequently analyzed using imaging mass cytometry (right panels). (B) Representative images from NSCLC stained simultaneously with a 37-marker IMC panel including tumor- and immune-cell markers. Selected markers are indicated within each caption. Bar = 100 μm. DAPI, 4′,6-diamidino-2-phenylindole; FFPE, formalin-fixed, paraffin-embedded; IMC, imaging mass cytometry; panCK, pancytokeratin.
      Highly dimensional, single-cell proteomic or transcriptomic studies from tumor cell extracts or liquid and fluid patient samples (e.g., blood and peripheral blood mononuclear cells, pleural effusion cells), although lacking spatial context, can capture numerous events. Multiple protein markers and single-cell phenotypes can be obtained using time-of-flight mass spectrometry (CyTOF platform) for simultaneous detection of approximately 30 to 40 metal-conjugated antibodies in fresh specimens. Recent studies using this platform coupled to mIF identified CD8+ effector T-cell exhaustion in human primary NSCLCs and revealed its association with reduced sensitivity to PD-1 axis blockade.
      • Datar I.
      • Sanmamed M.F.
      • Wang J.
      • et al.
      Expression analysis and significance of PD-1, LAG-3, and TIM-3 in human non-small cell lung cancer using spatially resolved and multiparametric single-cell analysis.
      ,
      • Sanmamed M.F.
      • Nie X.
      • Desai S.S.
      • et al.
      A burned-out CD8+ T-cell subset expands in the tumor microenvironment and curbs cancer immunotherapy.
      Another study using a 32-marker CyTOF panel to study peripheral blood mononuclear cells from nine patients with advanced NSCLC treated with immunotherapy revealed an increased proportion of the NK cells associated with the treatment response.
      • Cho Y.H.
      • Choi M.G.
      • Kim D.H.
      • et al.
      Natural killer cells as a potential biomarker for predicting immunotherapy efficacy in patients with non-small cell lung cancer.
      Single-cell RNA-sequencing detects the nuances in single-cell gene expression and enables to evaluate the specific cell subsets, pharmacodynamic changes, and intratumoral heterogeneity,
      • Sanmamed M.F.
      • Nie X.
      • Desai S.S.
      • et al.
      A burned-out CD8+ T-cell subset expands in the tumor microenvironment and curbs cancer immunotherapy.
      ,
      • Cho Y.H.
      • Choi M.G.
      • Kim D.H.
      • et al.
      Natural killer cells as a potential biomarker for predicting immunotherapy efficacy in patients with non-small cell lung cancer.
      in contrast to bulk-level transcriptomic, which provides averaged gene expression profiles of both tumor and TME populations. Individual single-cell RNA-sequencing platforms differ in their RNA capture methods, cDNA generation strategies, throughput, target coverage, and length of transcripts.
      • Adil A.
      • Kumar V.
      • Jan A.T.
      • Asger M.
      Single-cell transcriptomics: current methods and challenges in data acquisition and analysis.
      A recent study identified, in resectable NSCLC treated with neoadjuvant PD-1 ICI, tumor-antigen specific CD8+ T cells with tissue resident memory expression profiles, an incomplete effector program, increased immune-inhibitory signals, and expression of transcription factors associated with T-cell exhaustion.
      • Caushi J.X.
      • Zhang J.
      • Ji Z.
      • et al.
      Transcriptional programs of neoantigen-specific TIL in anti-PD-1-treated lung cancers.
      In preliminary explorations, T-cell exhaustion was associated with lack of benefit from the PD-1 axis blockade.
      Another emerging technology that is worth mentioning here is artificial intelligence (AI) and machine learning. A recent study has revealed correlation of AI-powered spatial TIL assessment with tumor response and PFS in patients with advanced NSCLC treated with ICI suggesting its role as a complementary biomarker to PD-L1 IHC.
      • Park S.
      • Ock C.Y.
      • Kim H.
      • et al.
      Artificial intelligence-powered spatial analysis of tumor-infiltrating lymphocytes as complementary biomarker for immune checkpoint inhibition in non-small-cell lung cancer.

      Microbiome

      The microbiome is a dynamic collection of commensal organisms colonizing sites of the human body and has been found to have an impact on anticancer immunity and response to immunotherapy.
      • Ocáriz-Díez M.
      • Cruellas M.
      • Gascón M.
      • et al.
      Microbiota and lung cancer. Opportunities and challenges for improving immunotherapy efficacy.
      The commensal bacteria provide protection from pathogenic organisms and regulate host immunity by “crosstalk” with immune cells in the mucosa through pattern recognition receptors to a significant extent. The innate immune system is stimulated by the host microbiome that regulates local inflammation and helps to inform the adaptive immune response,
      • Zheng D.
      • Liwinski T.
      • Elinav E.
      Interaction between microbiota and immunity in health and disease.
      although the exact mechanism by which altered microbiomes affect response to immunotherapy is unclear.
      Several preclinical studies in mice have suggested that the gut microbiome plays a role in response to ICIs in solid tumors.
      • Sivan A.
      • Corrales L.
      • Hubert N.
      • et al.
      Commensal Bifidobacterium promotes antitumor immunity and facilitates anti-PD-L1 efficacy.
      ,
      • Vétizou M.
      • Pitt J.M.
      • Daillère R.
      • et al.
      Anticancer immunotherapy by CTLA-4 blockade relies on the gut microbiota.
      Higher microbial diversity and higher proportions of “good” bacteria (“eubiosis”) are associated with treatment response in patients with melanoma treated with PD-1 inhibitors.
      • Gopalakrishnan V.
      • Spencer C.N.
      • Nezi L.
      • et al.
      Gut microbiome modulates response to anti-PD-1 immunotherapy in melanoma patients.
      ,
      • Matson V.
      • Fessler J.
      • Bao R.
      • et al.
      The commensal microbiome is associated with anti-PD-1 efficacy in metastatic melanoma patients.
      A landmark study by Routy et al.
      • Routy B.
      • Le Chatelier E.
      • Derosa L.
      • et al.
      Gut microbiome influences efficacy of PD-1-based immunotherapy against epithelial tumors.
      in 2018 revealed that patients with NSCLC with higher diversity of gut microbiome had improved response to ICI treatment. Akkermansia muciniphila and Enterococcus hirae were enriched in responders, and memory T-cell responses against A. muciniphila or E. hirae were found in patients with a better clinical response. Patients with NSCLC treated with antibiotics, known to alter the gut microbiome (“intestinal dysbiosis”), had reduced survival when given within 3 months of PD-1 blockade.
      • Routy B.
      • Le Chatelier E.
      • Derosa L.
      • et al.
      Gut microbiome influences efficacy of PD-1-based immunotherapy against epithelial tumors.
      Results of follow-up in vivo studies revealed fecal microbiota transplantation or oral supplementation with A. muciniphila overcame antibiotic-induced resistance to PD-1 blockade in mice.
      • Routy B.
      • Le Chatelier E.
      • Derosa L.
      • et al.
      Gut microbiome influences efficacy of PD-1-based immunotherapy against epithelial tumors.
      Small phase 1 clinical trials in patients with melanoma were found to have some benefit from fecal microbiome transplantation in overcoming ICI resistance,
      • Baruch E.N.
      • Youngster I.
      • Ben-Betzalel G.
      • et al.
      Fecal microbiota transplant promotes response in immunotherapy-refractory melanoma patients.
      ,
      • Davar D.
      • Dzutsev A.K.
      • McCulloch J.A.
      • et al.
      Fecal microbiota transplant overcomes resistance to anti-PD-1 therapy in melanoma patients.
      and feasibility studies of this approach are ongoing in NSCLC.
      • McLean L.
      • Leal J.L.
      • Solomon B.J.
      • John T.
      Immunotherapy in oncogene addicted non-small cell lung cancer.
      Other studies have revealed differences in gut microbiome between ICI responders and nonresponders in NSCLC with Firmicutes and higher microbial diversity associated with improved outcomes.
      • Jin C.
      • Lagoudas G.K.
      • Zhao C.
      • et al.
      Commensal microbiota promote lung cancer development via γδ T cells.
      In 37 Chinese patients with advanced NSCLC treated with nivolumab as part of CheckMate-078 and CheckMate-870, longer PFS was found in patients with high gut microbiome diversity.
      • Jin C.
      • Lagoudas G.K.
      • Zhao C.
      • et al.
      Commensal microbiota promote lung cancer development via γδ T cells.
      There were also differences in the microbiome composition between responders and nonresponders.
      • Jin C.
      • Lagoudas G.K.
      • Zhao C.
      • et al.
      Commensal microbiota promote lung cancer development via γδ T cells.
      In this study, previous antibiotic therapy did not affect outcome, but the study cohort was small. High microbiome diversity was associated with increased memory CD8+ T cells and NK cell signatures by peripheral blood flow cytometry.
      • Jin C.
      • Lagoudas G.K.
      • Zhao C.
      • et al.
      Commensal microbiota promote lung cancer development via γδ T cells.
      Microbiome studies in lung cancer (and other solid tumors) have almost all focused on the gut microbiome; however, preclinical mouse studies have revealed that the lower respiratory tract microbiome affects local immunity and could potentially be more predictive than the gut microbiome for IO in lung cancer.
      • Tsay J.J.
      • Wu B.G.
      • Sulaiman I.
      • et al.
      Lower airway dysbiosis affects lung cancer progression.
      • Beck J.M.
      • Young V.B.
      • Huffnagle G.B.
      The microbiome of the lung.
      • Dickson R.P.
      The lung microbiome and ARDS. It is time to broaden the model.
      An altered lung microbiome leading to a dysbiotic signature with increased oral commensals has been associated with tumor progression and poor prognosis in patients with NSCLC.
      • Tsay J.J.
      • Wu B.G.
      • Sulaiman I.
      • et al.
      Lower airway dysbiosis affects lung cancer progression.
      Interestingly, modulation of the lung microbiome by aerosolized antibiotics has also been reported to promote immunity against lung metastasis in patients with melanoma.
      • Le Noci V.
      • Guglielmetti S.
      • Arioli S.
      • et al.
      Modulation of pulmonary microbiota by antibiotic or probiotic aerosol therapy: a strategy to promote immunosurveillance against lung metastases.
      Studies to date have focused on the diversity and abundance of the bacterial microbiome using metagenomic shotgun sequencing, 16S ribosomal RNA gene sequencing, and quantitative polymerase chain reaction techniques for selected bacteria in stool or respiratory tract samples.
      • Gopalakrishnan V.
      • Spencer C.N.
      • Nezi L.
      • et al.
      Gut microbiome modulates response to anti-PD-1 immunotherapy in melanoma patients.
      ,
      • Routy B.
      • Le Chatelier E.
      • Derosa L.
      • et al.
      Gut microbiome influences efficacy of PD-1-based immunotherapy against epithelial tumors.
      Nevertheless, the optimal method of evaluating the gut (or respiratory) microbiome has not been determined, and the potential impact of intrinsic and extrinsic factors such as race, diet, smoking, antibiotics, or other environmental exposures is poorly understood. Microbiome-wide association studies may provide more clarity in the future. Apart from bacteria, commensal fungi, viruses, and protozoans are integral components of the human microbiome that play an important role in tumor immunosurveillance.
      • Aykut B.
      • Pushalkar S.
      • Chen R.
      • et al.
      The fungal mycobiome promotes pancreatic oncogenesis via activation of MBL.
      ,
      • Hannigan G.D.
      • Duhaime M.B.
      • Ruffin 4th, M.T.
      • Koumpouras C.C.
      • Schloss P.D.
      Diagnostic potential and interactive dynamics of the colorectal cancer virome.
      Because the microbiome affects anticancer immune response with a “healthy” diverse microbiome favoring response to ICIs,
      • Gopalakrishnan V.
      • Spencer C.N.
      • Nezi L.
      • et al.
      Gut microbiome modulates response to anti-PD-1 immunotherapy in melanoma patients.
      • Matson V.
      • Fessler J.
      • Bao R.
      • et al.
      The commensal microbiome is associated with anti-PD-1 efficacy in metastatic melanoma patients.
      • Routy B.
      • Le Chatelier E.
      • Derosa L.
      • et al.
      Gut microbiome influences efficacy of PD-1-based immunotherapy against epithelial tumors.
      the gut and lung microbiome could potentially be a predictive biomarker of response to IO, and its modulation may improve response to ICIs. More studies are needed to increase our understanding of the complex cancer-microbiome-immune axis and interplay between microbial ecology and host immunity to determine whether the microbiome could be a clinically useful biomarker.

      Major Pathologic Response to IO Neoadjuvant Therapy

      Major pathologic response (MPR) and complete pathologic response (CPR) have been used as an end point for a surrogate of clinical outcome in patients with lung cancer treated with various types of neoadjuvant therapy, including IO therapy. A number of studies have suggested that CPR after neoadjuvant chemotherapy may be a surrogate for overall survival.
      • Junker K.
      • Langner K.
      • Klinke F.
      • Bosse U.
      • Thomas M.
      Grading of tumor regression in non-small cell lung cancer : morphology and prognosis.
      • Pataer A.
      • Kalhor N.
      • Correa A.M.
      • et al.
      Histopathologic response criteria predict survival of patients with resected lung cancer after neoadjuvant chemotherapy.
      • Qu Y.
      • Emoto K.
      • Eguchi T.
      • et al.
      Pathologic assessment after neoadjuvant chemotherapy for NSCLC: importance and implications of distinguishing adenocarcinoma from squamous cell carcinoma.
      • Weissferdt A.
      • Pataer A.
      • Vaporciyan A.A.
      • et al.
      Agreement on major pathological response in NSCLC patients receiving neoadjuvant chemotherapy.
      Nevertheless, it has been reported to be achieved in only 4% to 12% cases after chemotherapy making it a less optimal end point,
      • Gilligan D.
      • Nicolson M.
      • Smith I.
      • et al.
      Preoperative chemotherapy in patients with resectable non-small cell lung cancer: results of the MRC LU22/NVALT 2/EORTC 08012 multicentre randomised trial and update of systematic review.
      ,
      • Pisters K.M.
      • Kris M.G.
      • Gralla R.J.
      • Zaman M.B.
      • Heelan R.T.
      • Martini N.
      Pathologic complete response in advanced non-small-cell lung cancer following preoperative chemotherapy: implications for the design of future non-small-cell lung cancer combined modality trials.
      whereas the recent clinical trial (CheckMate 816) reported a CPR rate of 30.5% after neoadjuvant therapy with nivolumab plus platinum-based chemotherapy.
      • Forde P.M.
      • Spicer J.
      • Lu S.
      • et al.
      Neoadjuvant nivolumab plus chemotherapy in resectable lung cancer.
      In contrast, MPR, defined as 10% or less of viable tumor cells at the primary tumor site in a surgically resected specimen, is more achievable; thus, there has been an increasing interest in its adoption. Although many studies revealed the prognostic value of MPR, there are several technical issues in its assessment which have not been addressed until recently. Importantly, although MPR has been standardized across different types of specimens and therapies in other cancer types, such as breast, a standardized approach to postneoadjuvant surgically resected lung cancer specimens has until recently been lacking. The main questions include standardized approach to gross processing, histologic criteria of MPR, significance of different types of neoadjuvant therapy on histologic features of response (chemotherapy, targeted therapies, IO therapy), assessment of response in lymph nodes, and cutoff values for different histologic tumor types.
      • Weissferdt A.
      • Pataer A.
      • Swisher S.G.
      • et al.
      Controversies and challenges in the pathologic examination of lung resection specimens after neoadjuvant treatment.
      ,
      • Saqi A.
      • Leslie K.O.
      • Moreira A.L.
      • et al.
      Assessing pathologic response in resected lung cancers: current standards, proposal for a novel pathologic response calculator tool, and challenges in practice.
      Currently, there have been two main attempts to address those issues including the IASLC multidisciplinary recommendation for pathologic assessment of lung cancer resection specimens after neoadjuvant therapy and pan-tumor pathologic scoring of response to PD-(L)1 blockade reported by Stein et al.
      • Lee J.S.
      • Ruppin E.
      Multiomics prediction of response rates to therapies to inhibit programmed cell death 1 and programmed cell death 1 ligand 1.
      ,
      • Stein J.E.
      • Lipson E.J.
      • Cottrell T.R.
      • et al.
      Pan-tumor pathologic scoring of response to PD-(L)1 blockade.
      The IASLC approach is specific for lung cancer and therapy agnostic, whereas the latter is tumor type agnostic but specific for immunotherapy, although the definitions of MPR and CPR are the same (10% or less and no residual viable tumor cells in the specimen, respectively). The question is whether the proposed scoring systems are interchangeable and applicable to different types of neoadjuvant therapies.
      In postneoadjuvant therapy resections, proliferative fibrosis, neovascularization, cholesterol clefts, dense TILs, and tertiary lymphoid structures can be found in addition to reduction of viable tumor cells.
      • Cottrell T.R.
      • Thompson E.D.
      • Forde P.M.
      • et al.
      Pathologic features of response to neoadjuvant anti-PD-1 in resected non-small-cell lung carcinoma: a proposal for quantitative immune-related pathologic response criteria (irPRC).
      These are considered to represent a state of immune activation in responders and, collectively, have been referred to as a “regression bed.” It is worth mentioning that these features can also be found in tumors treated with anti–PD-(L)1 combined with other agents such as chemotherapy or targeted therapies or chemotherapy alone, or, in a small proportion of cases, in tumors never exposed to neoadjuvant treatment. Importantly, the difference in interpreting stromal reaction is reflected to the difference in scoring % residual tumor volume between the above-mentioned two scoring systems. The pan-tumor pathologic scoring system includes intratumoral stroma as residual viable tumor, if features of “regression” are not present in immunotherapy-treated lung cancers, whereas the IASLC proposal considers viable tumor cells only as residual viable tumor. There are limited data to suggest that pathologists can consistently distinguish features of treatment response from preexisting stromal features.
      To best of our knowledge, direct comparison studies between the two scoring systems have not been reported. Furthermore, although inter- and intraobserver concordance in scoring of postneoadjuvant surgically resected lung cancer specimens seems good after dedicated pathologist training,
      • Merino D.M.
      • McShane L.M.
      • Fabrizio D.
      • et al.
      Establishing guidelines to harmonize tumor mutational burden (TMB): in silico assessment of variation in TMB quantification across diagnostic platforms: phase I of the Friends of Cancer Research TMB Harmonization Project.
      ,
      • Cottrell T.R.
      • Thompson E.D.
      • Forde P.M.
      • et al.
      Pathologic features of response to neoadjuvant anti-PD-1 in resected non-small-cell lung carcinoma: a proposal for quantitative immune-related pathologic response criteria (irPRC).
      and AI-powered digital assessment on pathology response seems to be strongly correlated with manual assessment,
      • Dacic S.
      • Travis W.T.
      • Giltnane J.M.
      • et al.
      Artificial intelligence (AI)–powered pathologic response (PathR) assessment of resection specimens after neoadjuvant atezolizumab in patients with non-small cell lung cancer: results from the LCMC3 study.
      these need to be further confirmed.
      Gross processing is the first critical step in the assessment of MPR. The above-mentioned approaches revealed some differences in the gross processing of postneoadjuvant lung cancer resection specimens, although complete sampling of a representative section of the tumor bed is the most often used approach.
      Published studies in chemotherapy and chemoradiotherapy revealed prognostic significance of MPR with a cutoff of 10% or less viable tumor.
      • Junker K.
      • Langner K.
      • Klinke F.
      • Bosse U.
      • Thomas M.
      Grading of tumor regression in non-small cell lung cancer : morphology and prognosis.
      ,
      • Pataer A.
      • Kalhor N.
      • Correa A.M.
      • et al.
      Histopathologic response criteria predict survival of patients with resected lung cancer after neoadjuvant chemotherapy.
      ,
      • Blaauwgeers J.L.
      • Kappers I.
      • Klomp H.M.
      • et al.
      Complete pathological response is predictive for clinical outcome after tri-modality therapy for carcinomas of the superior pulmonary sulcus.
      ,
      • Junker K.
      • Thomas M.
      • Schulmann K.
      • Klinke F.
      • Bosse U.
      • Müller K.M.
      Tumour regression in non-small-cell lung cancer following neoadjuvant therapy. Histological assessment.
      More recent studies suggested the best cutoff of 10% viable tumor for squamous cell carcinoma and that of 65% for adenocarcinoma.
      • Qu Y.
      • Emoto K.
      • Eguchi T.
      • et al.
      Pathologic assessment after neoadjuvant chemotherapy for NSCLC: importance and implications of distinguishing adenocarcinoma from squamous cell carcinoma.
      ,
      • Liu X.
      • Sun W.
      • Wu J.
      • et al.
      Major pathologic response assessment and clinical significance of metastatic lymph nodes after neoadjuvant therapy for non-small cell lung cancer.
      ,
      • Zens P.
      • Bello C.
      • Scherz A.
      • et al.
      A prognostic score for non-small cell lung cancer resected after neoadjuvant therapy in comparison with the tumor-node-metastases classification and major pathological response.
      Preliminary data suggest that a cutoff of 10% or less viable tumor is a predictor of overall survival in patients treated with IO.
      • Forde P.M.
      • Chaft J.E.
      • Smith K.N.
      • et al.
      Neoadjuvant PD-1 blockade in resectable lung cancer.
      ,
      • Cascone T.
      • William Jr., W.N.
      • Weissferdt A.
      • et al.
      Neoadjuvant nivolumab or nivolumab plus ipilimumab in operable non-small cell lung cancer: the phase 2 randomized NEOSTAR trial.
      It is possible that other cutoffs will emerge for lung cancer and perhaps for other tumor types as more data become available. Histologic features of response to IO are also emerging.
      Histologic features of response found in the primary tumor can also be found in the lymph nodes. Nevertheless, the response to neoadjuvant treatment may be different between the primary tumor and lymph node metastases, and currently it is uncertain on how to define MPR in cases with minimal viable tumor at the primary site and substantial viable metastases. Recently published, retrospective studies of lung cancers treated with neoadjuvant chemotherapies revealed that nodal disease is a key determinant of outcomes and only patients with MPR and no residual nodal disease were found to have a survival benefit.
      • Corsini E.M.
      • Weissferdt A.
      • Pataer A.
      • et al.
      Pathological nodal disease defines survival outcomes in patients with lung cancer with tumour major pathological response following neoadjuvant chemotherapy.
      ,
      • Pataer A.
      • Weissferdt A.
      • Vaporciyan A.A.
      • et al.
      Evaluation of pathologic response in lymph nodes of patients with lung cancer receiving neoadjuvant chemotherapy.
      Currently, clinical responses are also evaluated radiologically using the Response Evaluation Criteria in Solid Tumors criteria as a standard method. Many studies on neoadjuvant chemotherapy revealed that a significant number of tumors with histologic CPR were radiologically evaluated as stable disease or partial response, and radiological response was not an independent prognostic factor.
      • Pataer A.
      • Kalhor N.
      • Correa A.M.
      • et al.
      Histopathologic response criteria predict survival of patients with resected lung cancer after neoadjuvant chemotherapy.
      The similar, discrepant assessments have been reported in clinical trials with neoadjuvant nivolumab or nivolumab plus ipilimumab,
      • Cascone T.
      • William Jr., W.N.
      • Weissferdt A.
      • et al.
      Neoadjuvant nivolumab or nivolumab plus ipilimumab in operable non-small cell lung cancer: the phase 2 randomized NEOSTAR trial.
      with IO plus chemotherapy,
      • Provencio M.
      • Nadal E.
      • Insa A.
      • et al.
      Neoadjuvant chemotherapy and nivolumab in resectable non-small-cell lung cancer (NADIM): an open-label, multicentre, single-arm, phase 2 trial.
      ,
      • Shu C.A.
      • Gainor J.F.
      • Awad M.M.
      • et al.
      Neoadjuvant atezolizumab and chemotherapy in patients with resectable non-small-cell lung cancer: an open-label, multicentre, single-arm, phase 2 trial.
      or with a new IO agent.
      • Gao S.
      • Li N.
      • Gao S.
      • et al.
      Neoadjuvant PD-1 inhibitor (Sintilimab) in NSCLC.
      The discordant results are likely attributed to the stromal immune activation and pose a further question as to which assessment is more representative for estimating the clinical response in the treatment using ICIs. Importantly, ctDNA clearance was associated with CPR in the CheckMate 816 clinical trial implying ctDNA analysis may serve as a predictive marker for neoadjuvant IO response.
      • Forde P.M.
      • Spicer J.
      • Lu S.
      • et al.
      Neoadjuvant nivolumab plus chemotherapy in resectable lung cancer.

      Summary and Conclusion

      Although there are a plethora of data on single diverse predictive biomarkers currently available in the IO space, head-to-head comparisons between multiple single biomarkers have been limited, and PD-L1 IHC remains the most often used predictive biomarker. Given the complex interaction between the immune system and tumor cells, multiple steps involved in the cancer immunity cycle and imperfect nature of individual biomarkers, a global assessment involving various biomarkers may be warranted to improve prediction of response to ICIs. Although some evidence suggests that a combination of biomarkers gives a better prediction of response and outcomes, the predictive performance of selected combinations needs to be tested and compared in prospective studies, ideally in clinical trials.
      Furthermore, predictive biomarkers need to be practical. In that regard, new technologies may identify new biomarkers that can be easily implemented in clinical practice for a better adaptation of immunotherapies along with still incompletely understood mechanisms of primary and secondary resistance.

      CRediT Authorship Contribution Statement

      Mari Mino-Kenudson: Conceptualization, Methodology, Resources, Writing (original), Writing (editing), Supervision.
      Kurt Schalper, Wendy Cooper, Sanja Dacic, Fred R. Hirsch, Deepali Jain, Fernando Lopez-Rios, Ming Sound Tsao, Yasushi Yatabe, Mary Beth Beasley, Hui Yu, Lynette M. Sholl, Elizabeth Brambilla, Teh-Ying Chou, Ignacio Wistuba, Keith M. Kerr: Resources, Writing (editing).
      Casey Connolly: Project management.
      Sylvie Lantuejoul: Conceptualization, Methodology, Resources, Writing (original), Writing (editing), Supervision.

      Acknowledgments

      The authors thank the International Association for the Study of Lung Cancer Executive Committee for their critical reviews on the manuscript.

      References

        • Sung H.
        • Ferlay J.
        • Siegel R.L.
        • et al.
        Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries.
        CA Cancer J Clin. 2021; 71: 209-249
        • Reck M.
        • Remon J.
        • Hellmann M.D.
        First-line immunotherapy for non-small-cell lung cancer.
        J Clin Oncol. 2022; 40: 586-597
        • Zhang L.
        • Lin W.
        • Tan F.
        • et al.
        Sintilimab for the treatment of non-small cell lung cancer.
        Biomark Res. 2022; 10: 23
        • Lantuejoul S.
        • Sound-Tsao M.
        • Cooper W.A.
        • et al.
        PD-L1 testing for lung cancer in 2019: perspective from the IASLC pathology committee.
        J Thorac Oncol. 2020; 15: 499-519
        • Sholl L.M.
        • Hirsch F.R.
        • Hwang D.
        • et al.
        The promises and challenges of tumor mutation burden as an immunotherapy biomarker: a perspective from the International Association for the Study of Lung Cancer Pathology Committee.
        J Thorac Oncol. 2020; 15: 1409-1424
        • Marcus L.
        • Fashoyin-Aje L.A.
        • Donoghue M.
        • et al.
        FDA approval summary: Pembrolizumab for the treatment of tumor mutational burden-high solid tumors.
        Clin Cancer Res. 2021; 27: 4685-4689
        • Travis W.D.
        • Dacic S.
        • Wistuba I.
        • et al.
        IASLC multidisciplinary recommendations for pathologic assessment of lung cancer resection specimens after neoadjuvant therapy.
        J Thorac Oncol. 2020; 15: 709-740
        • Lee J.S.
        • Ruppin E.
        Multiomics prediction of response rates to therapies to inhibit programmed cell death 1 and programmed cell death 1 ligand 1.
        JAMA Oncol. 2019; 5: 1614-1618
        • Lu S.
        • Stein J.E.
        • Rimm D.L.
        • et al.
        Comparison of biomarker modalities for predicting response to PD-1/PD-L1 checkpoint blockade: a systematic review and meta-analysis.
        JAMA Oncol. 2019; 5: 1195-1204
        • Doroshow D.B.
        • Bhalla S.
        • Beasley M.B.
        • et al.
        PD-L1 as a biomarker of response to immune-checkpoint inhibitors.
        Nat Rev Clin Oncol. 2021; 18: 345-362
        • Hirsch F.R.
        • McElhinny A.
        • Stanforth D.
        • et al.
        PD-L1 immunohistochemistry assays for lung cancer: results from phase 1 of the blueprint PD-L1 IHC assay comparison project.
        J Thorac Oncol. 2017; 12: 208-222
        • Tsao M.S.
        • Kerr K.M.
        • Kockx M.
        • et al.
        PD-L1 immunohistochemistry comparability study in real-life clinical samples: results of blueprint phase 2 project.
        J Thorac Oncol. 2018; 13: 1302-1311
        • Tsao M.S.
        • Kerr K.M.
        • Dacic S.
        • Yatabe Y.
        • Hirsch F.R.
        IASLC Atlas of PD-L1 Immunohistochemistry Testing in Lung Cancer.
        Editorial Rx Press, Colorado Springs, CO2017
      1. Food and Drug Administration. FDA approves cemiplimab-rwlc for non-small cell lung cancer with high PD-L1 expression. https://www.fda.gov/drugs/resources-information-approved-drugs/fda-approves-cemiplimab-rwlc-non-small-cell-lung-cancer-high-pd-l1-expression. Accessed February 22, 2021.

        • Paz-Ares L.G.
        • Ramalingam S.S.
        • Ciuleanu T.E.
        • et al.
        First-line nivolumab plus ipilimumab in advanced NSCLC: 4-year outcomes from the randomized, open-label, phase 3 CheckMate 227 part 1 Trial.
        J Thorac Oncol. 2022; 17: 289-308
        • Sezer A.
        • Kilickap S.
        • Gümüş M.
        • et al.
        Cemiplimab monotherapy for first-line treatment of advanced non-small-cell lung cancer with PD-L1 of at least 50%: a multicentre, open-label, global, phase 3, randomised, controlled trial.
        Lancet. 2021; 397: 592-604
        • Herbst R.S.
        • Giaccone G.
        • de Marinis F.
        • et al.
        Atezolizumab for first-line treatment of PD-L1-selected patients with NSCLC.
        N Engl J Med. 2020; 383: 1328-1339
        • Aguilar E.J.
        • Ricciuti B.
        • Gainor J.F.
        • et al.
        Outcomes to first-line pembrolizumab in patients with non-small-cell lung cancer and very high PD-L1 expression.
        Ann Oncol. 2019; 30: 1653-1659
        • Doroshow D.B.
        • Wei W.
        • Gupta S.
        • et al.
        Programmed death-ligand 1 tumor proportion score and overall survival from first-line pembrolizumab in patients with nonsquamous versus squamous NSCLC.
        J Thorac Oncol. 2021; 16: 2139-2143
        • Yu H.
        • Batenchuk C.
        • Badzio A.
        • et al.
        PD-L1 expression by two complementary diagnostic assays and mRNA in situ hybridization in small cell lung cancer.
        J Thorac Oncol. 2017; 12: 110-120
        • Ready N.
        • Farago A.F.
        • de Braud F.
        • et al.
        Third-line nivolumab monotherapy in recurrent SCLC: CheckMate 032.
        J Thorac Oncol. 2019; 14: 237-244
        • Hirsch F.R.
        • Walker J.
        • Higgs B.W.
        • Cooper Z.A.
        • Raja R.G.
        Wistuba II. The Combiome hypothesis: selecting optimal treatment for cancer patients.
        Clin Lung Cancer. 2022; 23: 1-13
        • Gandara D.R.
        • Paul S.M.
        • Kowanetz M.
        • et al.
        Blood-based tumor mutational burden as a predictor of clinical benefit in non-small-cell lung cancer patients treated with atezolizumab.
        Nat Med. 2018; 24: 1441-1448
        • Marabelle A.
        • Fakih M.
        • Lopez J.
        • et al.
        Association of tumour mutational burden with outcomes in patients with advanced solid tumours treated with pembrolizumab: prospective biomarker analysis of the multicohort, open-label, phase 2 KEYNOTE-158 study.
        Lancet Oncol. 2020; 21: 1353-1365
        • Hellmann M.D.
        • Callahan M.K.
        • Awad M.M.
        • et al.
        Tumor mutational burden and efficacy of nivolumab monotherapy and in combination with ipilimumab in small-cell lung cancer.
        Cancer Cell. 2018; 33 (e4): 853-861
        • Horn L.
        • Mansfield A.S.
        • Szczesna A.
        • et al.
        First-line atezolizumab plus chemotherapy in extensive-stage small-cell lung cancer.
        N Engl J Med. 2018; 379: 2220-2229
        • Hellmann M.D.
        • Paz-Ares L.
        • Bernabe Caro R.
        • et al.
        Nivolumab plus ipilimumab in advanced non-small-cell lung cancer.
        N Engl J Med. 2019; 381: 2020-2031
        • Hellmann M.D.
        • Ciuleanu T.E.
        • Pluzanski A.
        • et al.
        Nivolumab plus ipilimumab in lung cancer with a high tumor mutational burden.
        N Engl J Med. 2018; 378: 2093-2104
        • Rizvi N.A.
        • Cho B.C.
        • Reinmuth N.
        • et al.
        Durvalumab with or without tremelimumab vs standard chemotherapy in first-line treatment of metastatic non-small cell lung cancer: the MYSTIC Phase 3 randomized clinical trial.
        JAMA Oncol. 2020; 6: 661-674
      2. Food and Drug Administration. FDA approves nivolumab plus ipilimumab and chemotherapy for first-line treatment of metastatic NSCLC. https://www.fda.gov/drugs/resources-information-approved-drugs/fda-approves-nivolumab-plus-ipilimumab-and-chemotherapy-first-line-treatment-metastatic-nsclc. Accessed May 26, 2020

        • Leighl N.B.
        • Laurie S.A.
        • Goss G.D.
        • et al.
        CCTG BR34: A randomized Phase 2 trial of durvalumab and tremelimumab with or without platinum-based chemotherapy in patients with metastatic NSCLC.
        J Thorac Oncol. 2022; 17: 434-445
        • Wang J.
        • Lu S.
        • Hu C.
        • et al.
        Updated analysis of tislelizumab plus chemotherapy vs chemotherapy alone as first-line treatment of advanced squamous non-small cell lung cancer (SQ NSCLC).
        Annals Oncol. 2020; 31: S754-S840
        • Paz-Ares L.
        • Ciuleanu T.-E.
        • Cobo M.
        • et al.
        First-line nivolumab (NIVO) + ipilimumab (IPI) + 2 cycles chemotherapy (chemo) vs 4 cycles chemo in advanced non-small cell lung cancer (aNSCLC): association of blood and tissue tumor mutational burden (TMB) with efficacy in CheckMate 9LA.
        J Thorac Oncol. 2021; 16: S750-S751
        • Buchhalter I.
        • Rempel E.
        • Endris V.
        • et al.
        Size matters: dissecting key parameters for panel-based tumor mutational burden analysis.
        Int J Cancer. 2019; 144: 848-858
        • Chalmers Z.R.
        • Connelly C.F.
        • Fabrizio D.
        • et al.
        Analysis of 100,000 human cancer genomes reveals the landscape of tumor mutational burden.
        Genome Med. 2017; 9: 34
        • Endris V.
        • Buchhalter I.
        • Allgäuer M.
        • et al.
        Measurement of tumor mutational burden (TMB) in routine molecular diagnostics: in silico and real-life analysis of three larger gene panels.
        Int J Cancer. 2019; 144: 2303-2312
        • Ramos-Paradas J.
        • Hernández-Prieto S.
        • Lora D.
        • et al.
        Tumor mutational burden assessment in non-small-cell lung cancer samples: results from the TMB2 harmonization project comparing three NGS panels.
        J Immunother Cancer. 2021; 9e001904
        • Wei B.
        • Kang J.
        • Kibukawa M.
        • et al.
        Evaluation of the TruSight Oncology 500 assay for routine clinical testing of tumor mutational burden and clinical utility for predicting response to pembrolizumab.
        J Mol Diagn. 2022; 24: 600-608
        • Merino D.M.
        • McShane L.M.
        • Fabrizio D.
        • et al.
        Establishing guidelines to harmonize tumor mutational burden (TMB): in silico assessment of variation in TMB quantification across diagnostic platforms: phase I of the Friends of Cancer Research TMB Harmonization Project.
        J Immunother Cancer. 2020; 8e000147
        • Ye L.
        • Creaney J.
        • Redwood A.
        • Robinson B.
        The current lung cancer neoantigen landscape and implications for therapy.
        J Thorac Oncol. 2021; 16: 922-932
        • Anagnostou V.
        • Smith K.N.
        • Forde P.M.
        • et al.
        Evolution of neoantigen landscape during immune checkpoint blockade in non-small cell lung cancer.
        Cancer Discov. 2017; 7: 264-276
        • Fehlings M.
        • Jhunjhunwala S.
        • Kowanetz M.
        • et al.
        Late-differentiated effector neoantigen-specific CD8+ T cells are enriched in peripheral blood of non-small cell lung carcinoma patients responding to atezolizumab treatment.
        J Immunother Cancer. 2019; 7: 249
        • Forde P.M.
        • Chaft J.E.
        • Smith K.N.
        • et al.
        Neoadjuvant PD-1 blockade in resectable lung cancer.
        N Engl J Med. 2018; 378: 1976-1986
        • Gettinger S.N.
        • Choi J.
        • Mani N.
        • et al.
        A dormant TIL phenotype defines non-small cell lung carcinomas sensitive to immune checkpoint blockers.
        Nat Commun. 2018; 9: 3196
        • Rizvi N.A.
        • Hellmann M.D.
        • Snyder A.
        • et al.
        Cancer immunology. Mutational landscape determines sensitivity to PD-1 blockade in non-small cell lung cancer.
        Science. 2015; 348: 124-128
        • McGranahan N.
        • Furness A.J.
        • Rosenthal R.
        • et al.
        Clonal neoantigens elicit T cell immunoreactivity and sensitivity to immune checkpoint blockade.
        Science. 2016; 351: 1463-1469
        • Montesion M.
        • Murugesan K.
        • Jin D.X.
        • et al.
        Somatic HLA Class I loss is a widespread mechanism of immune evasion which refines the use of tumor mutational burden as a biomarker of checkpoint inhibitor response.
        Cancer Discov. 2021; 11: 282-292
        • De Mattos-Arruda L.
        • Vazquez M.
        • Finotello F.
        • et al.
        Neoantigen prediction and computational perspectives towards clinical benefit: recommendations from the ESMO Precision Medicine Working Group.
        Ann Oncol. 2020; 31: 978-990
        • Chen Y.P.
        • Zhang Y.
        • Lv J.W.
        • et al.
        Genomic analysis of tumor microenvironment immune types across 14 solid cancer types: immunotherapeutic implications.
        Theranostics. 2017; 7: 3585-3594
        • WHO Classification of Tumours Editorial Board
        WHO Classification oF Tumours.
        5th ed. Thoracic Tumours. 5. IARC Publications, Lyon, France2021
        • Gainor J.F.
        • Shaw A.T.
        • Sequist L.V.
        • et al.
        EGFR mutations and ALK rearrangements are associated with low response rates to PD-1 pathway blockade in non-small cell lung cancer: A retrospective analysis.
        Clin Cancer Res. 2016; 22: 4585-4593
        • Kauffmann-Guerrero D.
        • Tufman A.
        • Kahnert K.
        • et al.
        Response to checkpoint inhibition in non-small cell lung cancer with molecular driver alterations.
        Oncol Res Treat. 2020; 43: 289-298
        • Ricciuti B.
        • Arbour K.C.
        • Lin J.J.
        • et al.
        Diminished efficacy of programmed death-(ligand)1 inhibition in STK11- and KEAP1-mutant lung adenocarcinoma is affected by KRAS mutation status.
        J Thorac Oncol. 2022; 17: 399-410
        • Skoulidis F.
        • Goldberg M.E.
        • Greenawalt D.M.
        • et al.
        STK11/LKB1 mutations and PD-1 inhibitor resistance in KRAS-mutant lung adenocarcinoma.
        Cancer Discov. 2018; 8: 822-835
        • Zhao Y.
        • Cao Y.
        • Chen Y.
        • et al.
        B2M gene expression shapes the immune landscape of lung adenocarcinoma and determines the response to immunotherapy.
        Immunology. 2021; 164: 507-523
        • Duruisseaux M.
        • Martínez-Cardús A.
        • Calleja-Cervantes M.E.
        • et al.
        Epigenetic prediction of response to anti-PD-1 treatment in non-small-cell lung cancer: a multicentre, retrospective analysis.
        Lancet Respir Med. 2018; 6: 771-781
        • Chen D.S.
        • Mellman I.
        Oncology meets immunology: the cancer-immunity cycle.
        Immunity. 2013; 39: 1-10
        • Doroshow D.B.
        • Sanmamed M.F.
        • Hastings K.
        • et al.
        Immunotherapy in non-small cell lung cancer: facts and hopes.
        Clin Cancer Res. 2019; 25: 4592-4602
        • Karasaki T.
        • Nagayama K.
        • Kuwano H.
        • et al.
        An immunogram for the cancer-immunity cycle: towards personalized immunotherapy of lung cancer.
        J Thorac Oncol. 2017; 12: 791-803
        • Sanmamed M.F.
        • Eguren-Santamaria I.
        • Schalper K.A.
        Overview of lung cancer immunotherapy.
        Cancer J. 2020; 26: 473-484
        • Sholl L.M.
        Biomarkers of response to checkpoint inhibitors beyond PD-L1 in lung cancer.
        Mod Pathol. 2022; 35: 66-74
        • Shepherd D.J.
        • Tabb E.S.
        • Kunitoki K.
        • et al.
        Lymphocyte-activation gene 3 in non-small-cell lung carcinomas: correlations with clinicopathologic features and prognostic significance.
        Mod Pathol. 2022; 35: 615-624
        • Shirasawa M.
        • Yoshida T.
        • Imabayashi T.
        • et al.
        Baseline PD-L1 expression and tumour-infiltrated lymphocyte status predict the efficacy of durvalumab consolidation therapy after chemoradiotherapy in unresectable locally advanced patients with non-small-cell lung cancer.
        Eur J Cancer. 2022; 162: 1-10
        • Shirasawa M.
        • Yoshida T.
        • Shimoda Y.
        • et al.
        Differential immune-related microenvironment determines programmed cell death Protein-1/Programmed death-ligand 1 blockade efficacy in patients with advanced NSCLC.
        J Thorac Oncol. 2021; 16: 2078-2090
        • Yeong J.
        • Suteja L.
        • Simoni Y.
        • et al.
        Intratumoral CD39+CD8+ T cells predict response to programmed cell death Protein-1 or programmed death Ligand-1 blockade in patients with NSCLC.
        J Thorac Oncol. 2021; 16: 1349-1358
        • Thommen D.S.
        • Koelzer V.H.
        • Herzig P.
        • et al.
        A transcriptionally and functionally distinct PD-1+ CD8+ T cell pool with predictive potential in non-small-cell lung cancer treated with PD-1 blockade.
        Nat Med. 2018; 24: 994-1004
        • Kumagai S.
        • Togashi Y.
        • Kamada T.
        • et al.
        The PD-1 expression balance between effector and regulatory T cells predicts the clinical efficacy of PD-1 blockade therapies.
        Nat Immunol. 2020; 21: 1346-1358
        • Datar I.
        • Sanmamed M.F.
        • Wang J.
        • et al.
        Expression analysis and significance of PD-1, LAG-3, and TIM-3 in human non-small cell lung cancer using spatially resolved and multiparametric single-cell analysis.
        Clin Cancer Res. 2019; 25: 4663-4673
        • Eiva M.A.
        • Omran D.K.
        • Chacon J.A.
        • Powell Jr., D.J.
        Systematic analysis of CD39, CD103, CD137, and PD-1 as biomarkers for naturally occurring tumor antigen-specific TILs.
        Eur J Immunol. 2022; 52: 96-108
        • Gide T.N.
        • Quek C.
        • Menzies A.M.
        • et al.
        Distinct immune cell populations define response to anti-PD-1 monotherapy and anti-PD-1/anti-CTLA-4 combined therapy.
        Cancer Cell. 2019; 35: 238-255.e6
        • Schalper K.A.
        • Carvajal-Hausdorf D.
        • McLaughlin J.
        • et al.
        Differential expression and significance of PD-L1, IDO-1, and B7-H4 in human lung cancer.
        Clin Cancer Res. 2017; 23: 370-378
        • Zhang M.L.
        • Kem M.
        • Mooradian M.J.
        • et al.
        Differential expression of PD-L1 and IDO1 in association with the immune microenvironment in resected lung adenocarcinomas.
        Mod Pathol. 2019; 32: 511-523
        • Wang J.
        • Sanmamed M.F.
        • Datar I.
        • et al.
        Fibrinogen-like Protein 1 is a major immune inhibitory ligand of LAG-3.
        Cell. 2019; 176 (e12): 334-347
        • Hou A.
        • Hou K.
        • Huang Q.
        • Lei Y.
        • Chen W.
        Targeting myeloid-derived suppressor cell, a promising strategy to overcome resistance to immune checkpoint inhibitors.
        Front Immunol. 2020; 11: 783
        • Zhuo M.
        • Chen H.
        • Zhang T.
        • et al.
        The potential predictive value of circulating immune cell ratio and tumor marker in atezolizumab treated advanced non-small cell lung cancer patients.
        Cancer Biomark. 2018; 22: 467-476
        • Kang D.H.
        • Chung C.
        • Sun P.
        • et al.
        Circulating regulatory T cells predict efficacy and atypical responses in lung cancer patients treated with PD-1/PD-L1 inhibitors.
        Cancer Immunol Immunother. 2022; 71: 579-588
        • Rogado J.
        • Pozo F.
        • Troule K.
        • et al.
        Peripheral blood mononuclear cells predict therapeutic efficacy of immunotherapy in NSCLC.
        Cancers (Basel). 2022; 14: 2898
        • Wang Q.
        • He Y.
        • Li W.
        • et al.
        Soluble immune checkpoint-related proteins in blood are associated with invasion and progression in non-small cell lung cancer.
        Front Immunol. 2022; 13887916
        • Mildner F.
        • Sopper S.
        • Amann A.
        • et al.
        Systematic review: soluble immunological biomarkers in advanced non-small-cell lung cancer (NSCLC).
        Crit Rev Oncol Hematol. 2020; 153102948
        • Mazzaschi G.
        • Minari R.
        • Zecca A.
        • et al.
        Soluble PD-L1 and Circulating CD8+PD-1+ and NK cells enclose a prognostic and predictive immune effector score in immunotherapy treated NSCLC patients.
        Lung Cancer. 2020; 148: 1-11
        • Honrubia-Peris B.
        • Garde-Noguera J.
        • Garcia- Sánchez J.
        • Piera-Molons N.
        • Llombart-Cussac A.
        • Fernández-Murga M.L.
        Soluble biomarkers with prognostic and predictive value in advanced non-small cell lung cancer treated with immunotherapy.
        Cancers (Basel). 2021; 13: 4280
        • Ayers M.
        • Lunceford J.
        • Nebozhyn M.
        • et al.
        IFN-γ-related mRNA profile predicts clinical response to PD-1 blockade.
        J Clin Invest. 2017; 127: 2930-2940
        • Ott P.A.
        • Bang Y.J.
        • Piha-Paul S.A.
        • et al.
        T-cell-inflamed gene-expression profile, programmed death ligand 1 expression, and tumor mutational burden predict efficacy in patients treated with pembrolizumab across 20 cancers: KEYNOTE-028.
        J Clin Oncol. 2019; 37: 318-327
        • Trujillo J.A.
        • Sweis R.F.
        • Bao R.
        • Luke J.J.
        T cell-Inflamed versus non-T cell-Inflamed Tumors: A conceptual framework for cancer immunotherapy drug development and combination therapy selection.
        Cancer Immunol Res. 2018; 6: 990-1000
        • Fehrenbacher L.
        • Spira A.
        • Ballinger M.
        • et al.
        Atezolizumab versus docetaxel for patients with previously treated non-small-cell lung cancer (POPLAR): a multicentre, open-label, phase 2 randomised controlled trial.
        Lancet. 2016; 387: 1837-1846
        • Higgs B.W.
        • Morehouse C.A.
        • Streicher K.
        • et al.
        Interferon gamma messenger RNA signature in tumor biopsies predicts outcomes in patients with non-small cell lung carcinoma or urothelial cancer treated with durvalumab.
        Clin Cancer Res. 2018; 24: 3857-3866
        • Damotte D.
        • Warren S.
        • Arrondeau J.
        • et al.
        The tumor inflammation signature (TIS) is associated with anti-PD-1 treatment benefit in the CERTIM pan-cancer cohort.
        J Transl Med. 2019; 17: 357
        • Hwang S.
        • Kwon A.Y.
        • Jeong J.Y.
        • et al.
        Immune gene signatures for predicting durable clinical benefit of anti-PD-1 immunotherapy in patients with non-small cell lung cancer.
        Sci Rep. 2020; 10: 643
        • Ranganath H.
        • Jain A.L.
        • Smith J.R.
        • et al.
        Association of a novel 27-gene immuno-oncology assay with efficacy of immune checkpoint inhibitors in advanced non-small cell lung cancer.
        BMC Cancer. 2022; 22: 407
        • Leng Y.
        • Dang S.
        • Yin F.
        • et al.
        GDPLichi: a DNA damage repair-related gene classifier for predicting lung adenocarcinoma immune checkpoint inhibitors response.
        Front Oncol. 2021; 11733533
        • Jiang P.
        • Gu S.
        • Pan D.
        • et al.
        Signatures of T cell dysfunction and exclusion predict cancer immunotherapy response.
        Nat Med. 2018; 24: 1550-1558
        • Jang H.J.
        • Lee H.S.
        • Ramos D.
        • et al.
        Transcriptome-based molecular subtyping of non-small cell lung cancer may predict response to immune checkpoint inhibitors.
        J Thorac Cardiovasc Surg. 2020; 159: 1598-1610.e3
        • Budczies J.
        • Kirchner M.
        • Kluck K.
        • et al.
        A gene expression signature associated with B cells predicts benefit from immune checkpoint blockade in lung adenocarcinoma.
        Oncoimmunology. 2021; 101860586
        • Cabel L.
        • Proudhon C.
        • Romano E.
        • et al.
        Clinical potential of circulating tumour DNA in patients receiving anticancer immunotherapy.
        Nat Rev Clin Oncol. 2018; 15: 639-650
        • Goldberg S.B.
        • Narayan A.
        • Kole A.J.
        • et al.
        Early assessment of lung cancer immunotherapy response via circulating tumor DNA.
        Clin Cancer Res. 2018; 24: 1872-1880
        • Nabet B.Y.
        • Esfahani M.S.
        • Moding E.J.
        • et al.
        Noninvasive early identification of therapeutic benefit from immune checkpoint inhibition.
        Cell. 2020; 183 (e13): 363-376
        • Zhang Q.
        • Luo J.
        • Wu S.
        • et al.
        Prognostic and predictive impact of circulating tumor DNA in patients with advanced cancers treated with immune checkpoint blockade.
        Cancer Discov. 2020; 10: 1842-1853
        • Barefoot M.E.
        • Loyfer N.
        • Kiliti A.J.
        • McDeed 4th, A.P.
        • Kaplan T.
        • Wellstein A.
        Detection of cell types contributing to cancer from circulating, cell-free methylated DNA.
        Front Genet. 2021; 12671057
        • Cristiano S.
        • Leal A.
        • Phallen J.
        • et al.
        Genome-wide cell-free DNA fragmentation in patients with cancer.
        Nature. 2019; 570: 385-389
        • Jiang P.
        • Sun K.
        • Peng W.
        • et al.
        Plasma DNA end-motif profiling as a Fragmentomic marker in cancer, pregnancy, and transplantation.
        Cancer Discov. 2020; 10: 664-673
        • Keller L.
        • Belloum Y.
        • Wikman H.
        • Pantel K.
        Clinical relevance of blood-based ctDNA analysis: mutation detection and beyond.
        Br J Cancer. 2021; 124: 345-358
        • Lianidou E.
        Detection and relevance of epigenetic markers on ctDNA: recent advances and future outlook.
        Mol Oncol. 2021; 15: 1683-1700
        • Taube J.M.
        • Akturk G.
        • Angelo M.
        • et al.
        The Society for Immunotherapy of Cancer statement on best practices for multiplex immunohistochemistry (IHC) and immunofluorescence (IF) staining and validation.
        J Immunother Cancer. 2020; 8e000155
        • Gerdes M.J.
        • Sevinsky C.J.
        • Sood A.
        • et al.
        Highly multiplexed single-cell analysis of formalin-fixed, paraffin-embedded cancer tissue.
        Proc Natl Acad Sci U S A. 2013; 110: 11982-11987
        • Lin J.R.
        • Izar B.
        • Wang S.
        • et al.
        Highly multiplexed immunofluorescence imaging of human tissues and tumors using t-CyCIF and conventional optical microscopes.
        ELife. 2018; 7e31657
        • Goltsev Y.
        • Samusik N.
        • Kennedy-Darling J.
        • et al.
        Deep profiling of mouse splenic architecture with CODEX multiplexed imaging.
        Cell. 2018; 174 (e15): 968-981
        • Giesen C.
        • Wang H.A.
        • Schapiro D.
        • et al.
        Highly multiplexed imaging of tumor tissues with subcellular resolution by mass cytometry.
        Nat Methods. 2014; 11: 417-422
        • Angelo M.
        • Bendall S.C.
        • Finck R.
        • et al.
        Multiplexed ion beam imaging of human breast tumors.
        Nat Med. 2014; 20: 436-442
        • Amaria R.N.
        • Reddy S.M.
        • Tawbi H.A.
        • et al.
        Neoadjuvant immune checkpoint blockade in high-risk resectable melanoma.
        Nat Med. 2018; 24: 1649-1654
        • Ståhl P.L.
        • Salmén F.
        • Vickovic S.
        • et al.
        Visualization and analysis of gene expression in tissue sections by spatial transcriptomics.
        Science. 2016; 353: 78-82
        • Rodriques S.G.
        • Stickels R.R.
        • Goeva A.
        • et al.
        Slide-seq: A scalable technology for measuring genome-wide expression at high spatial resolution.
        Science. 2019; 363: 1463-1467
        • Liu Y.
        • Yang M.
        • Deng Y.
        • et al.
        High-spatial-resolution multi-omics sequencing via deterministic barcoding in tissue.
        Cell. 2020; 183 (e18): 1665-1681
        • Cho C.S.
        • Xi J.
        • Si Y.
        • et al.
        Microscopic examination of spatial transcriptome using Seq-Scope.
        Cell. 2021; 184 (e22): 3559-3572