P1.01-031 Utilization and Timing of Foundation Medicine (FMI) Testing in U.S. Advanced Non-Small Cell Lung Cancer (aNSCLC) Patients


      Actionable insights generated by FMI’s hybrid capture-based next-generation sequencing (NGS) comprehensive genomic profiling services are increasingly important for navigating cancer care in aNSCLC patients. FMI and other NGS platforms support treatment decisions by detecting a variety of genetic alterations implicated in oncogenesis. We describe: 1) the characteristics of aNSCLC patients receiving FMI testing and 2) the utilization patterns and timing of FMI testing in relation to treatment and other molecular tests using a real world oncology electronic health record (EHR) database.


      Flatiron Health has a longitudinal, demographically and geographically diverse database containing EHR data, reflecting routine clinical practice, from over 265 cancer clinics in the US. Inclusion criteria were aNSCLC diagnosis and ≥2 clinic visits within the Flatiron network on or after January 1, 2011. Data pertaining to molecular testing was available on 5 biomarkers (EGFR, ALK, KRAS, ROS1, PDL1) and used to identify 3 mutually exclusive testing groups: FMI, other NGS and non-NGS.


      As of March 31, 2017, the aNSCLC cohort included 33,473 patients. Of 1,395 patients with FMI testing, 738 (53%) also had ≥1 non-FMI test (43% EGFR, 40% ALK, 20% ROS1, 17% PDL1, 16% KRAS). In FMI-tested patients, 45% received results before starting a first line of therapy (vs. 57% of other NGS tested and 79% of non-NGS tested patients). Table 1 details patient and testing characteristics for FMI tested patients, along with first treatments received after FMI testing.


      Patients with FMI testing tended to be younger, non-smokers, and have squamous histology compared to patients receiving non-FMI tests. Nearly 50% of all FMI testing occurred prior to first treatment. Patients receiving FMI testing earlier were less likely to have a non-FMI biomarker test beforehand. Regardless of when FMI testing occurred, ∼20-30% of patients received a NCCN-recommended targeted therapy immediately after the FMI test.


      Table 1Patient, Testing and Treatment Characteristics for Patients Receiving FMI testing
      FMIOther NGSNon-NGSNo tests
      Patient count (%) (Total N=33,473)1,395 (4%)1,946 (6%)17,002 (51%)13,128 (39%)
      Patient and Testing characteristics
      Age at aNSCLC diagnosis (years) (median, [IQR])67 [59, 73]68 [60, 75]69 [61, 76]
      No history of smoking335 (24%)376 (19%)2,731 (16%)
      Squamous cell histology194 (14%)195 (10%)1,532 (9%)
      No. of tests
      11,224 (88%)1,587 (82%)n/a
      ≥2171 (12%)359 (18%)
      Year of testinga
      <20110 (0%)1 (0%)137 (1%)
      20110 (0%)5 (0%)1,215 (7%)
      20122 (0%)16 (1%)2,368 (14%)
      201359 (4%)117 (6%)2,894 (17%)
      2014185 (13%)235 (12%)3,263 (19%)
      2015377 (27%)560 (29%)3,099 (18%)
      2016634 (45%)818 (42%)2,921 (17%)
      2017 (through March 31, 2017)123 (9%)186 (10%)670 (4%)
      Number and Timing of Other Tests in patients with an FMI testFMI tested prior to start of 1st LoTFMI tested during 1st and before start of 2nd LoTFMI tested during 2nd and before start 3rd LoTFMI tested during or after 3rd LoT
      Patient count (%) (Total N=1,386)b626 (45%)489 (35%)157 (11%)114 (8%)
      Other tests conducted before FMI testingc,d
      ALK111 (18%)188 (38%)100 (64%)86 (75%)
      EGFR133 (21%)204 (42%)110 (70%)86 (75%)
      KRAS51 (8%)60 (12%)25 (16%)29 (25%)
      PDL141 (7%)44 (9%)23 (15%)22 (19%)
      ROS153 (9%)92 (19%)40 (26%)32 (28%)
      Other tests conducted after FMI testingc,d
      ALK48 (8%)34 (7%)8 (5%)3 (3%)
      EGFR53 (9%)38 (8%)9 (6%)5 (4%)
      KRAS35 (6%)30 (6%)7 (5%)2 (2%)
      PDL149 (8%)33 (7%)7 (5%)3 (3%)
      ROS137 (6%)28 (6%)6 (4%)1 (1%)
      First Treatment immediately following FMI testing
      Patient count (%) (Total N=802)e424 (68%)252 (52%)77 (49%)49 (43%)
      NCCN-recommended targeted treatment for aNSCLC (N=196)f,i105 (25%)56 (22%)20 (26%)15 (31%)
      NCCN-recommended immunotherapy (N=227)g,i82 (19%)110 (44%)22 (29%)13 (27%)
      Non NCCN-recommended targeted treatment for aNSCLC (N=13)h,i1 (0%)7 (3%)1 (1%)4 (8%)
      a If a patient had more than 1 FMI test, the year of the first FMI test is shown; similarly if a patient had more than 1 other NGS (ie. non-FMI), the first test is shown. If an FMI tested patient had both an FMI and non-FMI NGS test, the year of the first FMI test is shown; similarly if a other NGS tested patient has both an other NGS and non-NGS test, the year of the first other NGS test is shown. b 9 patients where the date of FMI test (ie. test result date or sample collection date) in relation to dates of treatment and line of therapy were missing or unknown are not included in this table. c Total of 738 patients (53% of all FMI tested patients) also received a non-FMI test in addition to their FMI test. d Percentages may not add up to 100% because of patients received more than one test. e Based on 802 total patients who received FMI testing and where treatment data was reported following their FMI test. f NCCN-recommended targeted treatments received included the following: erlotinib (N=63), gefitinib (N=10), afatinib (N=56), crizotinib (N=36), ceritinib (N=4), alectinib (N=8), trastuzumab (N=1), vemurafenib (N=2), dabrafenib (N=2), osimertinib (N=19), cabozantinib (N=0); it may be possible for patients to receive regimens containing more than 1 NCCN-recommended targeted treatment. g NCCN-recommended immunotherapy (immune checkpoint inhibitors) received included the following: nivolumab (N=189), pembrolizumab (N=29), atezolizumab (N=9); it may be possible for patients to receive regimens containing more than 1 NCCN-recommended immunotherapy. In addition, 2 patients received ipilimumab, an immune checkpoint inhibitor currently not recommended by NCCN for aNSCLC. hNon NCCN-recommended targeted treatments received included the following: olaparib (N=2), necitumumab (N=2), cetuximab (N=2), palbociclib (N=1), pazopanib (N=1), temsirolimus (N=1), trametinib (with no dabrafenib) (N=4) ; it may be possible for patients to receive regimens containing more than 1 non NCCN-recommended targeted treatment. i Based on the National Comprehensive Cancer Network (NCCN) guidelines for NSCLC, version 6.2017 LoT: line of therapy as recorded in the Flatiron data, IQR: inter-quartile range. Patients with missing data are excluded from the table; Percentages are rounded to closest decimals.