Virtual Library
Start Your Search
Julia R Naso
Author of
-
+
P33 - Pathology - Immunotherapy Biomarker (ID 101)
- Event: WCLC 2020
- Type: Posters
- Track: Pathology, Molecular Pathology and Diagnostic Biomarkers
- Presentations: 1
- Moderators:
- Coordinates: 1/28/2021, 00:00 - 00:00, ePoster Hall
-
+
P33.17 - Incidence of PD-L1 Expression at Clinically Significant Cut Offs in NSCLC EGFR Mutated Population Compared to Wild Type (ID 2710)
00:00 - 00:00 | Author(s): Julia R Naso
- Abstract
Introduction
EGFR mutation can induce PD-L1 expression through p-ERK1/2/p-c-Jun pathway. The total incidence of PD-L1 positivity and percentage of PD-L1 immunoexpression at clinically significant cut off levels is not well studied. We performed a retrospective analysis of EGFR mutated NSCLC patients and correlated their tumors PD-L1 status. These results were compared to PD-L1 levels of NSCLC wild type patients.
Methods
In the province of British Columbia Canada, patients with advanced stage IIIB/IV NSCLC are tested for PD-L1 expression using the 22C3 PharmaDx kit. Cut off levels of <1%, 1-49% and >50% are reported, as per the pembrolizumab clinical trials. In British Columbia, mutational testing for activating mutations is performed after PD-L1 IHC, and therefore all patients with known EGFR mutation have a known PD-L1 status. A retrospective analysis was done to identify patients with advanced NSCLC who had PD-L1 testing between January 2017 and May 2018, and had EGFR testing results.
Results
174 patients were identified with tumors positive for an EGFR mutation. Baseline characteristics were median age 67, female 49%, 99% non-squamous and 1% squamous histology. Of the 174 tumors identified, 38% (n=66) were PD-L1 <1%; 36% (n=63) were PD-L1 1-49%; and 26% (n=45) were PD-L1 ≥50%. 771 patients were identified with EGFR wild type tumors. Baseline characteristics were median age 68, female 56%, 99% non-squamous and 1% squamous histology. Of the 771 tumors identified, 35% (n=266) were PD-L1 <1%; 22% (n=172) were PD-L1 1-49%; and 43% (n=333) were PD-L1 ≥50%. Fewer EGFR M+ samples compared to EGFR wildtype samples scored PD-L1 <50% (Chi-square statistic 17.76 p< 0.05) which was statistically significant.
Conclusion
Patients with tumors who are EGFR M+ have lower percentage of high PD-L1 > 50% expression (26% vs 43%) and higher percentage of no expression (38% vs 35%) than patients with wild type NSCLC. Significant fewer EGFR M+ samples compared to EGFR wildtype samples scored PD-L1 <50%. At present PD-L1 expression does not guide treatment decisions with immunotherapy in EGFR M+ patients, as single checkpoint inhibitors have shown little efficacy. The IMpower150 trial did show a signal in EGFR M+ tumors treated with atezolizumab, bevacizumab, carboplatin and paclitaxel. Future trials may show a role for measuring the PD-L1 expression level, and knowing the PD-L1 incidence by different cut off levels in the NSCLC EGFR mutant population may guide future clinical trial design.
-
+
P37 - Pathology - Biomarker Testing (ID 107)
- Event: WCLC 2020
- Type: Posters
- Track: Pathology, Molecular Pathology and Diagnostic Biomarkers
- Presentations: 1
- Moderators:
- Coordinates: 1/28/2021, 00:00 - 00:00, ePoster Hall
-
+
P37.07 - Discordance on Repeat PD-L1 Testing in Non-Small Cell Lung Carcinoma (ID 3224)
00:00 - 00:00 | Presenting Author(s): Julia R Naso
- Abstract
Introduction
PD-L1 expression may be used to predict response to immunotherapy, but may be heterogeneous within a tumor and between tumor sites. Re-testing for PD-L1 on a second specimen may be requested to assess for changes in PD-L1 status. We sought to determine the frequency with which PD-L1 tests on different tissue specimens from the same patient have discordant results, and to assess clinical and pathologic factors that correlate with discordance.
Methods
We retrospectively assessed 22C3 PharmDx PD-L1 assays performed at BC Cancer on 2,959 patients with advanced stage NSCLC between Feb 1, 2017 and Jan 31, 2019. Patients who had PDL1 testing on more than one tissue sample were identified. Patients clinically suspected to have two separate primaries were excluded. PD-L1 scores were considered discordant between two tests if they were in different tumor proportion score categories (<1%, 1-49%, ≥50%). Associations between discordant results and clinical and pathologic factors was calculated using chi squared statistics and Student’s t-tests, with P<0.05 considered significant.
Results
Seventy-seven patients (70 with adenocarcinoma, 5 with squamous cell carcinoma, and 2 with NSCLC not otherwise specified) had PD-L1 tests clinically requested on two different tissue samples. The initial samples had a similar proportion of PD-L1 scores in each category (27 scored <1%, 26 scored 1-49%, and 24 scored ≥50%). Repeat PD-L1 test results were discordant with initial test results for 28 out of 77 patients (36%). The rate of discordance differed significantly depending on the initial score (initial score 1-49%: 62% discordance; initial score 1%: 33% discordance; initial score ≥50%: 13% discordance; P=0.001). Scores rarely differed across both thresholds (i.e. change from <1% to ≥50% occurred in 3/27 cases, and from ≥50% to <1% occurred in 1/24 cases). Among those with an initial 1-49% score, 8 increased to ≥50% and 8 decreased to <1% on repeat testing. There was no significant association between discordance on repeat testing and the length time between specimen collection, histologic diagnosis, EGFR mutation status or whether different pathologists scored the tests. Specimen pairs differing in type (i.e. cytology, biopsy or resection) or site (i.e. lung, pleural fluid, lymph node or distant metastasis) were no more likely to have discordant scores than specimens matched for these factors.
Conclusion
The likelihood of a discordant score on repeat testing was associated with the initial test score, but not the time between specimen collection or differences in specimen site or type. There was an overall 32% chance of scoring in a higher category on repeat testing, which may allow a patient to qualify for immunotherapy treatment. Our study highlights the clinical relevance of investigation into the predictive value of a PD-L1 score above the eligibility threshold on only the second test.
