Advertisement

A Validation Study for the Use of ROS1 Immunohistochemical Staining in Screening for ROS1 Translocations in Lung Cancer

Open ArchivePublished:May 11, 2016DOI:https://doi.org/10.1016/j.jtho.2016.03.019

      Abstract

      Introduction

      The presence of ROS proto-oncogene 1, receptor tyrosine kinase gene (ROS1) rearrangements in lung cancers confers sensitivity to ROS kinase inhibitors, including crizotinib. However, they are rare abnormalities (in ∼1% of non–small cell lung carcinomas) that are typically identified by fluorescence in situ hybridization (FISH), and so screening using immunohistochemical (IHC) staining would be both cost- and time-efficient.

      Methods

      A cohort of lung tumors negative for other common mutations related to targeted therapies were screened to assess the sensitivity and specificity of IHC staining in detecting ROS1 gene rearrangements, enriched by four other cases first identified by FISH. A review of published data was also undertaken.

      Results

      IHC staining was 100% sensitive (95% confidence interval: 48–100) and 83% specific (95% confidence interval: 86–100) overall when an h-score higher than 100 was used. Patients with ROS1 gene rearrangements were younger and typically never-smokers, with the tumors all being adenocarcinomas with higher-grade architectural features and focal signet ring morphologic features (two of five). Four patients treated with crizotinib showed a partial response, with three also showing a partial response to pemetrexed. Three of four patients remain alive at 13, 27, and 31 months, respectively.

      Conclusion

      IHC staining can be used to screen for ROS1 gene rearrangements, with patients herein showing a response to crizotinib. Patients with tumors that test positive according to IHC staining but negative according to FISH were also identified, which may have implications for treatment selection.

      Keywords

      Introduction

      Lung cancer remains a common cancer with high mortality, with patients typically presenting with advanced stage disease that until recently has been primarily treated with platinum-based chemotherapy. However, the past decade has seen a shift in how these patients are managed owing to the advent of targeted therapy that leads to extended survival if such mutations are present.
      • Morgensztern D.
      • Campo M.J.
      • Dahlberg S.E.
      • et al.
      Molecularly targeted therapies in non-small-cell lung cancer annual update 2014.
      • Kris M.G.
      • Johnson B.E.
      • Berry L.D.
      • et al.
      Using multiplexed assays of oncogenic drivers in lung cancers to select targeted drugs.
      The most frequently recognized actionable mutations are epidermal growth factor receptor gene (EGFR) mutations and anaplastic lymphoma receptor tyrosine kinase gene (ALK) translocations, and most laboratories have protocols for handling specimens to ensure timely testing for these abnormalities, with fluorescence in situ hybridization (FISH) currently the test of choice for ALK translocations.
      • Lindeman N.I.
      • Cagle P.T.
      • Beasley M.B.
      • et al.
      Molecular testing guideline for selection of lung cancer patients for EGFR and ALK tyrosine kinase inhibitors: guideline from the College of American Pathologists, International Association for the Study of Lung Cancer, and Association for Molecular Pathology.
      However, FISH testing can be time-consuming, is complex to interpret, and is comparatively expensive, so validated immunohistochemical (IHC) staining has been recommended as a quicker and more cost-efficient method of screening for ALK translocations,
      • Lindeman N.I.
      • Cagle P.T.
      • Beasley M.B.
      • et al.
      Molecular testing guideline for selection of lung cancer patients for EGFR and ALK tyrosine kinase inhibitors: guideline from the College of American Pathologists, International Association for the Study of Lung Cancer, and Association for Molecular Pathology.
      not least as the incidence of ALK translocations is no more than 2% in non–small cell lung cancers (NSCLCs).
      Similarly, FISH testing has been used to decide whether targeted therapy with crizotinib should be given when ROS proto-oncogene 1, receptor tyrosine kinase gene (ROS1) rearrangements (c-ros oncogene 1, located at 6q22)
      • Birchmeier C.
      • Sharma S.
      • Wigler M.
      Expression and rearrangement of the ROS1 gene in human glioblastoma cells.
      • Charest A.
      • Lane K.
      • McMahon K.
      • et al.
      Fusion of FIG to the receptor tyrosine kinase ROS in a glioblastoma with an interstitial del(6)(q21q21).
      • Gu T.L.
      • Deng X.
      • Huang F.
      • et al.
      Survey of tyrosine kinase signaling reveals ROS kinase fusions in human cholangiocarcinoma.
      are identified,
      • Mazieres J.
      • Zalcman G.
      • Crino L.
      • et al.
      Crizotinib therapy for advanced lung adenocarcinoma and a ROS1 rearrangement: results from the EUROS1 cohort.
      • Cheng H.
      • Ye L.
      • Xue L.
      [Detection of ROS1 gene rearrangement by FISH and analysis of its clinical features in non-small cell lung cancer patients].
      • Kim H.R.
      • Lim S.M.
      • Kim H.J.
      • et al.
      The frequency and impact of ROS1 rearrangement on clinical outcomes in never smokers with lung adenocarcinoma.
      • Shaw A.T.
      • Ou S.H.
      • Bang Y.J.
      • et al.
      Crizotinib in ROS1-rearranged non-small-cell lung cancer.
      • Bergethon K.
      • Shaw A.T.
      • Ou S.H.
      • et al.
      ROS1 rearrangements define a unique molecular class of lung cancers.
      but with a rate of only 1% to 2% in adenocarcinomas, this methodology is an expensive proposition, especially in a screening setting. Using IHC staining as a screening tool for ROS1 rearrangements may be a feasible option, and antibodies that identify the ROS1 protein in tumor cells are now available.
      • Sholl L.M.
      • Sun H.
      • Butaney M.
      • et al.
      ROS1 immunohistochemistry for detection of ROS1-rearranged lung adenocarcinomas.
      Herein, we present our findings from a validation study designed to determine whether IHC staining could be used as a screening tool for ROS1 gene rearrangements. We examined a selected group of mutation-negative lung cancers enriched with known positive cases from our diagnostic archive. We also reviewed published data to date on using IHC staining to screen for ROS1 translocations.

      Materials and Methods

      Given the relative rarity of the translocation and the fact the most driver mutations occur in isolation, a sequential test cohort of 103 patients was selected from 170 patients recruited from our institution to phase 1 of the Cancer Research UK Stratified Medicine Programme who were identified as negative for EGFR, Kirsten rat sarcoma viral oncogene homolog gene (KRAS), and/or B-Raf proto-oncogene, serine/threonine kinase gene (BRAF) mutations, as well as for ALK translocations.
      Cases that did not harbor any of the aforementioned molecular abnormalities were then screened using IHC staining. Sections were dewaxed and heat-induced antigen retrieval was carried out using a Dako PT Link module (PT10126) (Dako, Glostrup, Denmark). Slides were placed in preheated (65°C) target retrieval solution (Envision FLEX High pH TRS [pH9], DM828, diluted 1:50 from concentrate [Dako]). Staining was carried out at room temperature using Dako Envision FLEX reagents and a Dako Link 48 Autostainer (AS48030) (Dako). Slides were incubated for 30 minutes with the ROS1 antibody (D4D6 [Cell Signaling Technology, Danvers, MA] in a 1:300 dilution) before signal amplification (15-minute incubation with FLEX+ Rabbit Linker Solution [K8009] [Dako]). Cases were scored semiquantitatively as negative, weakly positive, moderately positive, or strongly positive along with the percentage of positive cells, after which an h-score was generated, with a score >100 viewed as positive
      • Rogers T.M.
      • Russell P.A.
      • Wright G.
      • et al.
      Comparison of methods in the detection of ALK and ROS1 rearrangements in lung cancer.
      • Cha Y.J.
      • Lee J.S.
      • Kim H.R.
      • et al.
      Screening of ROS1 rearrangements in lung adenocarcinoma by immunohistochemistry and comparison with ALK rearrangements.
      (this being a score of 0–3 intensity multiplied by the percentage of positive cells [range 0–300]).
      FISH analysis was performed directly on 2-μm formalin-fixed, paraffin-embedded (FFPE) tissue sections mounted on charged slides. We used the Cytocell ROS1 Dual Colour Break-Apart FISH Probe (Cytocell Ltd., Cambridge, United Kingdom) comprising a 206-kb Texas red–labeled probe covering the distal part of ROS1 and fluorescein isothiocyanate–labeled probes (green) situated proximal to the ROS1 gene to evaluate ROS1 translocations. FFPE slides were deparaffinised, treated with protease, co-denatured with FISH probe in a Thermobrite slide processor (Abbott Laboratories, Des Plaines, Illinois), washed, counterstained with 4′,6-diamidino-2-phenylindole, and analyzed using the Zeiss AxioImager Z2 fluorescence microscope (Carl Zeiss AG, Jena, Germany). Images were captured and analyzed using Metasystems ISIS software (MetaSystems GmbH, Altlussheim, Germany). As in previous studies, ROS1 FISH test results above a threshold of 15% in FFPE NSCLC tissue specimens were reported.
      • Mescam-Mancini L.
      • Lantuejoul S.
      • Moro-Sibilot D.
      • et al.
      On the relevance of a testing algorithm for the detection of ROS1-rearranged lung adenocarcinomas.
      • Davies K.D.
      • Le A.T.
      • Theodoro M.F.
      • et al.
      Identifying and targeting ROS1 gene fusions in non-small cell lung cancer.
      Cases showing at least 15% split signals were classified as positive, and the sensitivity and specificity of positive staining for ROS1 was generated. Agreement was reported as sensitivity and specificity with 95% confidence intervals (CIs) using FISH as the reference. Statistical analysis was performed with Stata 13 software (StataCorp LP, College Station, TX). Specimens were FISH-tested without knowledge of their ROS1 IHC status.
      This study was approved as a service development study by the local research and development committee (reference number 1240).

      Results

      From 170 patients recruited from our institution into phase 1 of the Cancer Research UK Stratified Medicine Programme, a total of 103 were wild type for EGFR, KRAS, BRAF, and ALK aberrations. Of these patients, 90 were wild type for all four genetic abnormalities, nine had failed analyses for one mutation, and four had failed analyses for two mutations. Of the 103 cases, 39 were adenocarcinomas, 39 squamous cell carcinomas, five small cell carcinoma, two adenosquamous carcinoma, three pleomorphic carcinomas, four large cell carcinoma, two large cell neuroendocrine cell carcinoma, three NSCLC according to biopsy results, and six carcinoids (Table 1).
      Table 1Demographics and Tumor Type in Cases Screened for ROS-1 Protein Using IHC Staining
      Tumor TypeNo.F:MAverage Age (Range)% with Positive h-Score (>100) Using IHC StainingFISH +VE
      ADC3915:2443–83 (66)2 of 391 of 33 (3 failed, tissue not available for 3)
      ADSQ20:270–76 (73)00 of 2
      SQCC3917:2253–87 (72)0Not tested
      SCLC52:347–74 (61)0Not tested
      PLEO CA31:255–74 (61)0Not tested
      LCC43:168–73 (71)0Not tested
      LCNEC20:271–81 (76)0Not tested
      Carcinoid61:525–74 (57)0Not tested
      NSCLC according to biopsy30:347–81 (63)0Not tested
      F:M, ratio of females to males; IHC, immunohistochemical; FISH, fluorescence in situ hybridization; +VE, positive; ADC, adenocarcinoma; ADSQ, adenosquamous carcinoma; SQCC, squamous cell carcinoma; SCLC, small cell lung cancer; PLEO CA, pleomorphic carcinoma; LCC large cell carcinoma; LCNEC, large cell neuroendocrine carconoma; NSCLC, non–small cell lung cancer.
      All cases classified as tumors other than adenocarcinoma were negative using IHC staining and they were not considered for further analysis, apart from two adenosquamous carcinomas (see Table 1). In the adenocarcinoma group, five cases showed moderately intense staining for ROS1 protein (Fig. 1), with positivity in 90% of cells in two cases and positivity in 20%, 40%, and 50% of cells in the other three cases. A further five cases showed weak staining in less than 50% of cells (<5%, <5%, 10%, 20%, and 30%). Therefore, two cases in total were classified as positive because of an h-score higher than 100.
      Figure thumbnail gr1
      Figure 1(A) A cell block from a transbronchial needle aspirate of a lymph node shows a micropapillary architecture. (B) Immunohistochemical analysis for ROS proto-oncogene 1, receptor tyrosine kinase (ROS1) protein shows variably granular cytoplasmic staining of mainly moderate intensity in 80% of cells within micropapillary clusters. (C) A biopsy specimen shows an adenocarcinoma showing focal signet ring cell morphologic features. (D) This case shows moderate staining in 90% of cells for ROS1 protein, with focal submembranous accentuation of staining. (E) A nodal biopsy specimen shows replacement by an adenocarcinoma with mixed micropapillary and cribriform patterns. (F) This case shows strong staining in 90% of cells for ROS1 protein.
      Only cases with sufficient tissue available, which were classified as adenocarcinoma (n = 36) and adenosquamous carcinoma (n = 2), were subjected to FISH testing (n = 38). Of these 38 cases, one case of adenocarcinoma was positive (78% of tumor cells showing a rearrangement). FISH testing was negative in 34 cases (32 adenocarcinomas and two adenosquamous carcinomas) with scores of 1% to 8%, and three cases failed.
      The one positive case according to FISH testing (signal pattern characteristic of an unbalanced ROS1 translocation resulting in loss of the red probe) was also positive according to IHC staining (>90% of cells, moderate staining) (Fig. 2A). Given the low number of cases within the screened cohort that showed ROS1 gene rearrangement using FISH, the archive was reviewed for cases found to be positive using FISH as part of clinical management.
      Figure thumbnail gr2
      Figure 2(A) Fluorescence in situ hybridization testing of case 5 shows a signal pattern characteristic of an unbalanced ROS proto-oncogene 1, receptor tyrosine kinase gene (ROS1) translocation resulting in loss of the red probe. (B) Fluorescence in situ hybridization testing shows 3’ (red) and 5’ (green) regions of the ROS1 gene separated by rearrangement.
      A further four cases were identified from the diagnostic archive of cases that had tested positive for a ROS1 gene rearrangement using FISH (Fig. 2B). Of these, one case showed strong staining in 90% of tumor cells, two cases showed moderate staining in 80% and 100% of cells, and one case showed weak staining in 60% of cells and moderate staining in 30% of cells (h-scores of 270, 200, 160, and 120, respectively). Overall, when the screening cohort of adenocarcinomas and adenosquamous cell carcinomas was combined with these additional four cases, the specificity was 83% (95% CI: 86–100) with a sensitivity of 100% (95% CI: 48–100). When positive, staining was cytoplasmic and variably granular, with one case (case 4) showing focal submembranous accentuation to the staining (Fig. 1D).
      Upon reviewing the five FISH-positive/IHC staining–positive cases, they were all adenocarcinomas, with the resection being cribriform predominant and also including solid and micropapillary areas, the two biopsy samples showing predominant solid or micropapillary architectures, and two cytologic samples comprising micropapillary clusters. Two cases had a focal signet ring cell morphologic pattern (see Fig. 1C). There were three male and two female patients, and their average age was 41 years (Table 2) compared with 68 years for the entire screened cohort and 66 years for the screened patients with adenocarcinomas. Of the five patients, four were known to be never-smokers and one was an ex-smoker of only 3 pack-years.
      Table 2Patient Demographics and IHC Staining, FISH, Histologic, Treatment, Response, and Outcome Data in Patients Positive for ROS-1 Gene Rearrangements
      AgeSexSmoking HistoryIHCFISHHistologic DiagnosisArchitectural Patterns
      Presence of signet ring morphologic pattern in parentheses.
      Treatment (Drug and Duration)Response (RECIST)Outcome
      27FNeverWeak staining in 60% of cells, moderate staining in 30% (h-score = 120)60% positive cellsADCMPCrizotinib for 5 cycles, followed by 6 cycles of pemetrexed and cisplatin, followed by maintenance pemetrexedPartial response to crizotinib and subsequently to pemetrexedAlive with disease at 13 mo
      40MNeverModerate staining in 100% of cells (h-score = 200)>50% of positive cellsADCMPPemetrexed for 34 cycles, followed by crizotonib for 10 cycles, still ongoingPartial response to pemetrexed and subsequently to crizotinibAlive with disease at 31 mo
      41MNeverStrong staining in 90% of cells (h-score = 270)>15% of positive cellsADCSolid/CRIB/MPCisplatin and gemcitabine followed by cyberknife of brain metastasis. ROS translocation identified and treatment with crizotinib for 6 cyclesPoor response to cisplatin and gemcitabine, and partial response to crizotininbDied after relapse at 14 mo
      43FNeverModerate staining in 80% of cells (h-score = 160)20% positive cellsADCSolid (signet ring 40%)Pemetrexed for 5 cycles, followed by crizotinib for 21 cycles, still ongoingPartial response to pemetrexed and subsequently to crizotinibAlive with disease at 27 mo
      53M3.5–pack-year ex-smokerModerate staining in 90% of cells (h-score = 180)78% positive cellsADCCRIB/solid/MP (signet ring <5%)N/AN/AAlive without relapse

      (pT2a N2)
      IHC, immunohistochemical; FISH, fluorescence in situ hybridization; ROS1, ROS proto-oncogene 1, receptor tyrosine kinase gene; RECIST, Response Evaluation Criteria in Solid Tumors; F, female; ADC, adenocarcinoma; MP, micropapillary; M, male; CRIB, cribiform; N/A, not applicable.
      a Presence of signet ring morphologic pattern in parentheses.
      In relation to outcome, four of the patients received sequential treatment with crizotinib and conventional chemotherapy, and one patient underwent resection, with no recurrence to date. All patients showed a partial response to crizotinib, when treated. One patient experienced a relapse after five cycles, was switched to pemetrexed and cisplatin with a partial response, and currently has stable disease and is receiving pemetrexed maintenance therapy. One patient remains alive at 31 months after having initially received five cycles of pemetrexed before being switched to crizotinib (Fig. 3). One patient had a ROS1 gene rearrangement identified at the time of a brain relapse with a subsequent partial response to crizotinib, but he experienced a relapse after six cycles of crizotinib. He was then switched to pemetrexed, again with a partial response, but he died 14 months after diagnosis. One patient was treated with pemetrexed for five cycles followed by crizotinib with a partial response, remains alive 27 months after diagnosis, and is continuing to receive crizotinib therapy.
      Figure thumbnail gr3
      Figure 3Mediastinal lymph nodes (AC) and tumor in the left hilum (DF) show persistent a response to crizotinib over 2.5 years.
      Several cases also showed positive staining of entrapped pneumocytes (Fig. 4) and alveolar macrophages, making scoring problematic in some adenocarcinomas.
      Figure thumbnail gr4
      Figure 4Tumor cells are negative for ROS proto-oncogene 1, receptor tyrosine kinase protein, but there is positive staining of background pneumocytes.

      Discussion

      This study confirms that IHC staining for ROS1 protein within the cytoplasm of tumor cells can potentially be used as a screening tool for ROS1 gene rearrangements, with excellent efficacy using crizotinib in three of four patients with advanced disease who remain alive at 13, 27, and 31 months, respectively. The fourth patient died of disease at 14 months. Sensitivity and specificity using an h-score of 100 were 100% and 83%, respectively, which is in keeping with most of the published literature (Table 3).
      • Sholl L.M.
      • Sun H.
      • Butaney M.
      • et al.
      ROS1 immunohistochemistry for detection of ROS1-rearranged lung adenocarcinomas.
      • Cha Y.J.
      • Lee J.S.
      • Kim H.R.
      • et al.
      Screening of ROS1 rearrangements in lung adenocarcinoma by immunohistochemistry and comparison with ALK rearrangements.
      • Mescam-Mancini L.
      • Lantuejoul S.
      • Moro-Sibilot D.
      • et al.
      On the relevance of a testing algorithm for the detection of ROS1-rearranged lung adenocarcinomas.
      • Boyle T.A.
      • Masago K.
      • Ellison K.E.
      • Yatabe Y.
      • Hirsch F.R.
      ROS1 immunohistochemistry among major genotypes of non-small-cell lung cancer.
      • Shan L.
      • Lian F.
      • Guo L.
      • et al.
      Detection of ROS1 gene rearrangement in lung adenocarcinoma: comparison of IHC, FISH and real-time RT-PCR.
      • Yoshida A.
      • Tsuta K.
      • Wakai S.
      • et al.
      Immunohistochemical detection of ROS1 is useful for identifying ROS1 rearrangements in lung cancers.
      • Warth A.
      • Muley T.
      • Dienemann H.
      • et al.
      ROS1 expression and translocations in non-small-cell lung cancer: clinicopathological analysis of 1478 cases.
      • Rimkunas V.M.
      • Crosby K.E.
      • Li D.
      • et al.
      Analysis of receptor tyrosine kinase ROS1-positive tumors in non-small cell lung cancer: identification of a FIG-ROS1 fusion.
      • Fu S.
      • Liang Y.
      • Lin Y.B.
      • et al.
      The frequency and clinical implication of ROS1 and RET rearrangements in resected stage IIIA-N2 non-small cell lung cancer patients.
      Table 3Review of Literature on Patients Screened for ROS-1 Gene Rearrangements Using IHC Staining
      StudySensitivitySpecificityPlatformAntibodyTreatmentIHC +VE CriteriaNo. +VE Cases by FISHTotal%Cohort EnrichmentTumor TypeHistologic Predominant Pattern
      Signet ring noted, when stated.
      Response to TreatmentAge and

      Sex
      Sholl et al.
      • Sholl L.M.
      • Sun H.
      • Butaney M.
      • et al.
      ROS1 immunohistochemistry for detection of ROS1-rearranged lung adenocarcinomas.
      10092Not statedD4D6; 1:1000Signal stain boost2+, 3+653Preselected cases
      21671.2% of ADCs (2 of 165): 2.2% of pan-WT ADCsNo
      8220ADC 100%- Solid

      - Cribriform

      - Mucinous (signet ring)
      1 of 1 with responseYounger, never-smokers
      Shan et al.
      • Shan L.
      • Lian F.
      • Guo L.
      • et al.
      Detection of ROS1 gene rearrangement in lung adenocarcinoma: comparison of IHC, FISH and real-time RT-PCR.
      10094Not statedD4D6; 1:50Not stated>10%, 2+, 3+136021.7% of ADCsYes (prescreened with ROS1 IHC)ADC 100%- Acinar

      - Papillary

      - Micropapillary
      Not statedYounger, never-smokers
      Lee et al.
      • Lee S.E.
      • Lee B.
      • Hong M.
      • et al.
      Comprehensive analysis of RET and ROS1 rearrangement in lung adenocarcinoma.
      N/AN/ANot stated3287 (Cell Signaling Technology) 1:40EnVision+

      (Dako)
      2+, 3+99410% of 94Triple negative for EGFR, ALK, and KRAS mutationsADC 100%- 2 solid (1 signet ring),

      - 4 acinar

      - 3 mucinous

      - 3 papillary
      Not statedYounger, never-smokers
      Cha et al.
      • Cha Y.J.
      • Lee J.S.
      • Kim H.R.
      • et al.
      Screening of ROS1 rearrangements in lung adenocarcinoma by immunohistochemistry and comparison with ALK rearrangements.
      10073Ventana Benchmark XTD4D6; 1:50Optiview detection kith-score (≥100)51114.5% of ADCsYes, ADC onlyNot statedYounger age
      8219Yes, never- smokers
      13330ADC 100%- Mucinous (signet ring)

      - Cribriform
      Boyle et al.
      • Boyle T.A.
      • Masago K.
      • Ellison K.E.
      • Yatabe Y.
      • Hirsch F.R.
      ROS1 immunohistochemistry among major genotypes of non-small-cell lung cancer.
      10083Ventana Benchmark XTD4D6; 1:100 and 1:250XT Ultraview DAB Kith-score (≥100)5335 of 20 ADCsPreselected cases from FISH studiesADC 100%- 50% solid,

      - 50% mucinous (33% signet ring)
      Not statedNot stated
      Mescam-Mancini et al.
      • Mescam-Mancini L.
      • Lantuejoul S.
      • Moro-Sibilot D.
      • et al.
      On the relevance of a testing algorithm for the detection of ROS1-rearranged lung adenocarcinomas.
      10097Ventana Benchmark XTD4D6; 1:50XT Ultraview DAB Kit>40% 2+, 3+91216%Triple negative for EGFR, ALK, and KRAS mutationsADC 100%- Solid

      - Acinar

      - Cribriform
      2 partial responses after 2 mo: 1 of 2 progressed after 5 mo, 1 of 2 had toxicity (reduced dose)White, advanced stage, nonsmoker, older age (61 y)
      Rogers et al.
      • Rogers T.M.
      • Russell P.A.
      • Wright G.
      • et al.
      Comparison of methods in the detection of ALK and ROS1 rearrangements in lung cancer.
      3399Ventana Benchmark UltraD4D6; 1:50Optiview detection kith-score33220.6% of NSCLCsNoADC 100%Not statedNot statedNot stated
      Fu et al.
      • Fu S.
      • Liang Y.
      • Lin Y.B.
      • et al.
      The frequency and clinical implication of ROS1 and RET rearrangements in resected stage IIIA-N2 non-small cell lung cancer patients.
      N/AN/AVentana Benchmark XTD4D6; 1:50Optiview detection kit>75% of cells, 2+, 3+32041.5% of NSCLCs, 2.2% of ADCsStage IIIA N2+ NSCLCADC 100%- Acinar (1 mixed mucinous and nonmucinous)Not statedYounger, 75% smokers
      Yoshida et al.
      • Yoshida A.
      • Tsuta K.
      • Wakai S.
      • et al.
      Immunohistochemical detection of ROS1 is useful for identifying ROS1 rearrangements in lung cancers.
      94%98%Not statedD4D6; 1:100EnVision-FLEX+ (Dako)h-score (>150)172533.5% of EGFR negative ADCsFurther enriched with 8 additional positive casesADC 100%Not statedNot statedNot stated
      Warth et al.
      • Warth A.
      • Muley T.
      • Dienemann H.
      • et al.
      ROS1 expression and translocations in non-small-cell lung cancer: clinicopathological analysis of 1478 cases.
      100%15%Dako autostainerD4D6; 1:100Pretreatment at pH8Any positive (1+, 2+, 3+)914780.6% of resected NSCLCsNoADC (8)

      LCC (1)
      Lepidic (1 of 6)

      Acinar (3 of 6)

      Solid (2 of 6)
      Not statedFemales, lower-stage resections
      Rimkunas et al.
      • Rimkunas V.M.
      • Crosby K.E.
      • Li D.
      • et al.
      Analysis of receptor tyrosine kinase ROS1-positive tumors in non-small cell lung cancer: identification of a FIG-ROS1 fusion.
      100%87.5%Not statedD4D6; 1:100EnVision-FLEX+ (Dako)Any positive (1+, 2+, 3+)95561.6% of NSCLCsNoADC (8)

      LCC (1)
      Lepidic (1 of 8)

      Acinar (3 of 8)

      Solid (2 of 8)

      Papillary (1 of 8)

      Signet ring (1 of 8)
      Not statedNot stated
      ROS1, ROS proto-oncogene 1, receptor tyrosine kinase gene; IHC, immunohistochemical; +VE, positive; FISH, fluorescence in situ hybridization; ADC, adenocarcinoma; WT, wild type; ROS1, ROS proto-oncogene 1, receptor tyrosine kinase; EGFR, epidermal growth factor receptor gene; ALK, anaplastic lymphoma receptor tyrosine kinase gene; KRAS, Kirsten rat sarcoma viral oncogene homolog gene; NSCLC, non–small cell lung cancer; N/A, not applicable; LCC, large cell carcinoma.
      a Signet ring noted, when stated.
      This study also highlights the importance of individual laboratory validation of a screening program, given the variation in platforms, antibody dilutions, and optimization techniques reflecting the real-life variation in laboratory facilities. Even our chosen thickness of slides for FISH varied from that of other studies (2 μm versus 3–4 μm), this having been internally validated as not affecting FISH results but saving tissue. Scoring systems are a further variation, although we chose an h-score of 100 as a cutoff as this was the choice of most groups using this methodology; plus, it gave us the best specificity. Nevertheless, despite these variations and the fact that specificities would be lower if the cutoff level were reduced, the fact that nearly all studies show a sensitivity of 100% suggests that IHC staining for ROS1 protein can be used to identify potential cases for confirmation by FISH testing, reducing the cost of screening significantly with a low likelihood of cases being missed. This IHC staining utility is similar to that identified for ALK fusions, with many centers now adopting ALK IHC staining as a standard screening tool.
      • Lindeman N.I.
      • Cagle P.T.
      • Beasley M.B.
      • et al.
      Molecular testing guideline for selection of lung cancer patients for EGFR and ALK tyrosine kinase inhibitors: guideline from the College of American Pathologists, International Association for the Study of Lung Cancer, and Association for Molecular Pathology.
      • Cha Y.J.
      • Lee J.S.
      • Kim H.R.
      • et al.
      Screening of ROS1 rearrangements in lung adenocarcinoma by immunohistochemistry and comparison with ALK rearrangements.
      • Choi I.H.
      • Kim D.W.
      • Ha S.Y.
      • Choi Y.L.
      • Lee H.J.
      • Han J.
      Analysis of histologic features suspecting anaplastic lymphoma kinase (ALK)-expressing pulmonary adenocarcinoma.
      • Le Quesne J.
      • Maurya M.
      • Yancheva S.G.
      • et al.
      A comparison of immunohistochemical assays and FISH in detecting the ALK translocation in diagnostic histological and cytological lung tumor material.
      • Just P.A.
      • Cazes A.
      • Audebourg A.
      • et al.
      Histologic subtypes, immunohistochemistry, FISH or molecular screening for the accurate diagnosis of ALK-rearrangement in lung cancer: a comprehensive study of Caucasian non-smokers.
      • Hutarew G.
      • Hauser-Kronberger C.
      • Strasser F.
      • Llenos I.C.
      • Dietze O.
      Immunohistochemistry as a screening tool for ALK rearrangement in NSCLC: evaluation of five different ALK antibody clones and ALK FISH.
      • Conklin C.M.
      • Craddock K.J.
      • Have C.
      • Laskin J.
      • Couture C.
      • Ionescu D.N.
      Immunohistochemistry is a reliable screening tool for identification of ALK rearrangement in non-small-cell lung carcinoma and is antibody dependent.
      Our data and the published literature also raise the question of how to manage those patients for whom there is a positive result for ROS1 with IHC staining but a negative FISH test result. This phenomenon is described in some patients with abnormalities in relation to the ALK gene in whom a clinical response to crizotinib has still been found.
      • Pekar-Zlotin M.
      • Hirsch F.R.
      • Soussan-Gutman L.
      • et al.
      Fluorescence in situ hybridization, immunohistochemistry, and next-generation sequencing for detection of EML4-ALK rearrangement in lung cancer.
      To our knowledge, there have not been any patients treated for ROS1-targeted agents with discordant ROS1 assessment results (ROS1 IHC staining–positive and ROS1 FISH–negative and vice versa), although it is only a matter of time before this happens as ROS1 IHC assessment becomes more routinely implemented. An alternative option for such patients would be to undertake additional polymerase chain reaction–based testing or next-generation sequencing–based testing
      • Ou S.H.
      • Chalmers Z.R.
      • Azada M.C.
      • et al.
      Identification of a novel TMEM106B-ROS1 fusion variant in lung adenocarcinoma by comprehensive genomic profiling.
      because, as in the case of ALK fusions, some tumors have been found to show IHC staining and polymerase chain reaction positivity for ROS rearrangements whereas the FISH test results have proven negative.
      • Boyle T.A.
      • Masago K.
      • Ellison K.E.
      • Yatabe Y.
      • Hirsch F.R.
      ROS1 immunohistochemistry among major genotypes of non-small-cell lung cancer.
      Perhaps surprisingly, the number of cases identified within the quadruple- or triple-negative group did not show the kind of increased frequency seen in some publications in which the positive rate was raised to 6%,
      • Mescam-Mancini L.
      • Lantuejoul S.
      • Moro-Sibilot D.
      • et al.
      On the relevance of a testing algorithm for the detection of ROS1-rearranged lung adenocarcinomas.
      although our rate of 2.5% is comparable with that for other “enriched” cohorts.
      In ROS1 IHC staining–positive cases, both our own data and the literature suggest that nearly all cases are adenocarcinomas and that ROS1 gene rearrangements are frequently seen in patients with solid, micropapillary, and cribriform histologic patterns, as well as in tumors containing tumor cells with a signet ring morphologic pattern.
      • Lee S.E.
      • Lee B.
      • Hong M.
      • et al.
      Comprehensive analysis of RET and ROS1 rearrangement in lung adenocarcinoma.
      This is similar to the findings for patients with ALK gene rearrangements,
      • Choi I.H.
      • Kim D.W.
      • Ha S.Y.
      • Choi Y.L.
      • Lee H.J.
      • Han J.
      Analysis of histologic features suspecting anaplastic lymphoma kinase (ALK)-expressing pulmonary adenocarcinoma.
      • Popat S.
      • Gonzalez D.
      • Min T.
      • et al.
      ALK translocation is associated with ALK immunoreactivity and extensive signet-ring morphology in primary lung adenocarcinoma.
      • Yoshida A.
      • Tsuta K.
      • Watanabe S.
      • et al.
      Frequent ALK rearrangement and TTF-1/p63 co-expression in lung adenocarcinoma with signet-ring cell component.
      • Pareja F.
      • Crapanzano J.P.
      • Mansukhani M.M.
      • Bulman W.A.
      • Saqi A.
      Cytomorphological features of ALK-positive lung adenocarcinomas: psammoma bodies and signet ring cells.
      although histologic pattern does not seem to be a good enough positive or negative predictor to direct ROS1 IHC screening and is unlikely to be sufficiently specific or sensitive in relation to ROS1 gene rearrangements. Nevertheless, it is notable that these three histologic patterns are ones associated with poorer prognosis,
      although this does vary between published studies (see Table 3). Patients also seem to be younger and never-smokers, which is consistent with what might be expected with driver mutations.
      Pathologists need to be aware that staining of native pneumocytes and macrophages as ROS1 protein can be expressed in full-length form in the normal lung, which is an important difference when making comparisons with ALK screening.
      • Yoshida A.
      • Tsuta K.
      • Wakai S.
      • et al.
      Immunohistochemical detection of ROS1 is useful for identifying ROS1 rearrangements in lung cancers.
      In this study, several cases showed weak to moderate positive staining, and careful review of the samples is required to ensure that false-positive results are avoided. Indeed, some true positive cases show differences in staining compared with what is seen in background staining in that there is sometimes submembranous accentuation, described particularly in relation to the ezrin gene (EZR)-ROS1 fusion,
      • Yoshida A.
      • Tsuta K.
      • Wakai S.
      • et al.
      Immunohistochemical detection of ROS1 is useful for identifying ROS1 rearrangements in lung cancers.
      but we were unable to look for specific fusion partners in this study. However, our numbers are small and whether these differences in the staining characteristics could be used to define which cases should be sent for FISH confirmation requires further study.
      The main limitation of the study is that the number of cases proving positive for the ROS1 gene rearrangement was low, necessitating enrichment from the diagnostic archive to support the high level of sensitivity. However, only one prior publication has more than 10 cases, highlighting the rarity of this gene rearrangement and the need for cost-efficient screening.
      In conclusion, with a high sensitivity rate and relatively high specificity rate, IHC screening to identify patients who might harbor ROS1 gene rearrangements is feasible and would be less expensive and time-consuming than FISH testing, which could be reserved for a confirmatory second step. Given the relative rarity of ROS1 gene rearrangements, further refinement of the screening population should be considered, limiting it to patients with a histologic diagnosis of adenocarcinoma and further enriching it by assessing only those patients with tumors that are wild type for EGFRALK, and other more routinely assessed molecular abnormalities. Limiting screening to patients presenting at a younger age and those who are never-smokers are further options that could be considered.

      Acknowledgments

      This project was supported by the National Institute of Health Research Respiratory Disease Biomedical Research Unit at the Royal Brompton and Harefield NHS Foundation Trust and Imperial College London. Dr. Popat acknowledges National Health Service funding to the Royal Marsden Hospital/Institute of Cancer Research National Institute of Health Research Biomedical Research Centre.

      References

        • Morgensztern D.
        • Campo M.J.
        • Dahlberg S.E.
        • et al.
        Molecularly targeted therapies in non-small-cell lung cancer annual update 2014.
        J Thorac Oncol. 2015; 10: S1-S63
        • Kris M.G.
        • Johnson B.E.
        • Berry L.D.
        • et al.
        Using multiplexed assays of oncogenic drivers in lung cancers to select targeted drugs.
        JAMA. 2014; 311: 1998-2006
        • Lindeman N.I.
        • Cagle P.T.
        • Beasley M.B.
        • et al.
        Molecular testing guideline for selection of lung cancer patients for EGFR and ALK tyrosine kinase inhibitors: guideline from the College of American Pathologists, International Association for the Study of Lung Cancer, and Association for Molecular Pathology.
        J Thorac Oncol. 2013; 8: 823-859
        • Birchmeier C.
        • Sharma S.
        • Wigler M.
        Expression and rearrangement of the ROS1 gene in human glioblastoma cells.
        Proc Natl Acad Sci U S A. 1987; 84: 9270-9274
        • Charest A.
        • Lane K.
        • McMahon K.
        • et al.
        Fusion of FIG to the receptor tyrosine kinase ROS in a glioblastoma with an interstitial del(6)(q21q21).
        Genes Chromosomes Cancer. 2003; 37: 58-71
        • Gu T.L.
        • Deng X.
        • Huang F.
        • et al.
        Survey of tyrosine kinase signaling reveals ROS kinase fusions in human cholangiocarcinoma.
        PLoS One. 2011; 6: e15640
        • Mazieres J.
        • Zalcman G.
        • Crino L.
        • et al.
        Crizotinib therapy for advanced lung adenocarcinoma and a ROS1 rearrangement: results from the EUROS1 cohort.
        J Clin Oncol. 2015; 33: 992-999
        • Cheng H.
        • Ye L.
        • Xue L.
        [Detection of ROS1 gene rearrangement by FISH and analysis of its clinical features in non-small cell lung cancer patients].
        Zhonghua Zhong Liu Za Zhi. Oct 2014; 36 ([in Chinese]): 751-754
        • Kim H.R.
        • Lim S.M.
        • Kim H.J.
        • et al.
        The frequency and impact of ROS1 rearrangement on clinical outcomes in never smokers with lung adenocarcinoma.
        Ann Oncol. 2013; 24: 2364-2370
        • Shaw A.T.
        • Ou S.H.
        • Bang Y.J.
        • et al.
        Crizotinib in ROS1-rearranged non-small-cell lung cancer.
        N Engl J Med. 2014; 371: 1963-1971
        • Bergethon K.
        • Shaw A.T.
        • Ou S.H.
        • et al.
        ROS1 rearrangements define a unique molecular class of lung cancers.
        J Clin Oncol. 2012; 30: 863-870
        • Sholl L.M.
        • Sun H.
        • Butaney M.
        • et al.
        ROS1 immunohistochemistry for detection of ROS1-rearranged lung adenocarcinomas.
        Am J Surg Pathol. 2013; 37: 1441-1449
        • Rogers T.M.
        • Russell P.A.
        • Wright G.
        • et al.
        Comparison of methods in the detection of ALK and ROS1 rearrangements in lung cancer.
        J Thorac Oncol. 2015; 10: 611-618
        • Cha Y.J.
        • Lee J.S.
        • Kim H.R.
        • et al.
        Screening of ROS1 rearrangements in lung adenocarcinoma by immunohistochemistry and comparison with ALK rearrangements.
        PLoS One. 2014; 9: e103333
        • Mescam-Mancini L.
        • Lantuejoul S.
        • Moro-Sibilot D.
        • et al.
        On the relevance of a testing algorithm for the detection of ROS1-rearranged lung adenocarcinomas.
        Lung Cancer. 2014; 83: 168-173
        • Davies K.D.
        • Le A.T.
        • Theodoro M.F.
        • et al.
        Identifying and targeting ROS1 gene fusions in non-small cell lung cancer.
        Clin Cancer Res. 2012; 18: 4570-4579
        • Boyle T.A.
        • Masago K.
        • Ellison K.E.
        • Yatabe Y.
        • Hirsch F.R.
        ROS1 immunohistochemistry among major genotypes of non-small-cell lung cancer.
        Clin Lung Cancer. 2015; 16: 106-111
        • Shan L.
        • Lian F.
        • Guo L.
        • et al.
        Detection of ROS1 gene rearrangement in lung adenocarcinoma: comparison of IHC, FISH and real-time RT-PCR.
        PLoS One. 2015; 10: e0120422
        • Yoshida A.
        • Tsuta K.
        • Wakai S.
        • et al.
        Immunohistochemical detection of ROS1 is useful for identifying ROS1 rearrangements in lung cancers.
        Mod Pathol. 2014; 27: 711-720
        • Warth A.
        • Muley T.
        • Dienemann H.
        • et al.
        ROS1 expression and translocations in non-small-cell lung cancer: clinicopathological analysis of 1478 cases.
        Histopathology. 2014; 65: 187-194
        • Rimkunas V.M.
        • Crosby K.E.
        • Li D.
        • et al.
        Analysis of receptor tyrosine kinase ROS1-positive tumors in non-small cell lung cancer: identification of a FIG-ROS1 fusion.
        Clin Cancer Res. 2012; 18: 4449-4457
        • Fu S.
        • Liang Y.
        • Lin Y.B.
        • et al.
        The frequency and clinical implication of ROS1 and RET rearrangements in resected stage IIIA-N2 non-small cell lung cancer patients.
        PLoS One. 2015; 10: e0124354
        • Choi I.H.
        • Kim D.W.
        • Ha S.Y.
        • Choi Y.L.
        • Lee H.J.
        • Han J.
        Analysis of histologic features suspecting anaplastic lymphoma kinase (ALK)-expressing pulmonary adenocarcinoma.
        J Pathol Transl Med. 2015; 49: 310-317
        • Le Quesne J.
        • Maurya M.
        • Yancheva S.G.
        • et al.
        A comparison of immunohistochemical assays and FISH in detecting the ALK translocation in diagnostic histological and cytological lung tumor material.
        J Thorac Oncol. 2014; 9: 769-774
        • Just P.A.
        • Cazes A.
        • Audebourg A.
        • et al.
        Histologic subtypes, immunohistochemistry, FISH or molecular screening for the accurate diagnosis of ALK-rearrangement in lung cancer: a comprehensive study of Caucasian non-smokers.
        Lung Cancer. 2012; 76: 309-315
        • Hutarew G.
        • Hauser-Kronberger C.
        • Strasser F.
        • Llenos I.C.
        • Dietze O.
        Immunohistochemistry as a screening tool for ALK rearrangement in NSCLC: evaluation of five different ALK antibody clones and ALK FISH.
        Histopathology. 2014; 65: 398-407
        • Conklin C.M.
        • Craddock K.J.
        • Have C.
        • Laskin J.
        • Couture C.
        • Ionescu D.N.
        Immunohistochemistry is a reliable screening tool for identification of ALK rearrangement in non-small-cell lung carcinoma and is antibody dependent.
        J Thorac Oncol. 2013; 8: 45-51
        • Pekar-Zlotin M.
        • Hirsch F.R.
        • Soussan-Gutman L.
        • et al.
        Fluorescence in situ hybridization, immunohistochemistry, and next-generation sequencing for detection of EML4-ALK rearrangement in lung cancer.
        Oncologist. 2015; 20: 316-322
        • Ou S.H.
        • Chalmers Z.R.
        • Azada M.C.
        • et al.
        Identification of a novel TMEM106B-ROS1 fusion variant in lung adenocarcinoma by comprehensive genomic profiling.
        Lung Cancer. 2015; 88: 352-354
        • Lee S.E.
        • Lee B.
        • Hong M.
        • et al.
        Comprehensive analysis of RET and ROS1 rearrangement in lung adenocarcinoma.
        Mod Pathol. 2015; 28: 468-479
        • Popat S.
        • Gonzalez D.
        • Min T.
        • et al.
        ALK translocation is associated with ALK immunoreactivity and extensive signet-ring morphology in primary lung adenocarcinoma.
        Lung Cancer. 2012; 75: 300-305
        • Yoshida A.
        • Tsuta K.
        • Watanabe S.
        • et al.
        Frequent ALK rearrangement and TTF-1/p63 co-expression in lung adenocarcinoma with signet-ring cell component.
        Lung Cancer. 2011; 72: 309-315
        • Pareja F.
        • Crapanzano J.P.
        • Mansukhani M.M.
        • Bulman W.A.
        • Saqi A.
        Cytomorphological features of ALK-positive lung adenocarcinomas: psammoma bodies and signet ring cells.
        Cancer Cytopathol. 2015; 123: 162-170
      1. Travis W.D. Brambilla E. Burke A.P. Marx A. Nicholson A.G. WHO Classification of Tumours of the Lung, Pleura, Thymus and Heart. 4th ed. IARC Press, Lyon, France2015