P1.09-04 Optimization of PD-L1 Testing Specimen Flow in the Greater Hamilton, Ontario Region

Abstract

Immunotherapy targeted at the programmed cell death 1 (PD-1) receptor and its ligand (PD-L1) is a new treatment option for non-small-cell lung cancer (NSCLC). Immunohistochemistry (IHC) for the PD-L1 protein has been shown to predict response. The 22C3 IHC assay is the only clinically validated PD-L1 test. We present the Hamilton, Ontario, Canada experience of local PD-L1 analysis using the 22C3 assay including both histology and cytology specimens. All data for requests for PD-L1 testing from within Hamilton were collected for one year. Unstained slides were cut for IHC analysis. Both histology and formalin fixed cytology specimens were accepted. Slides were sent for PD-L1 staining centrally at Dynacare in Bowmanville, Ontario, Canada. IHC interpretation was done in Hamilton. The assay was positive if ≥50% of tumour cells (TCs) had any intensity staining. The assay was negative if no TCs had staining. The assay was interpreted as low positive if 1-49% TCs had any intensity staining. Samples with less than 100 cells were considered inadequate. Turn around-time was defined as the accession date to PD-L1 sign out date. 401 samples were evaluated; 108 cytology(C) and 293 histology(H). 36% of samples tested positive (43%:C;33%:H); 20% of samples tested low positive (14%:C;23%:H); 39% of samples tested negative (29%:C;42%:H); 5% were insufficient for evaluation (15%:C;1%:H). Chi-squared analysis identified a statistically significant (χ2 < 0.02) difference in the distribution of test results comparing histology and cytology. The mean turn-around-time (TAT) was 28.9 days (range 12-144). TAT varied by hospital of origin. TAT for PD-L1 results from the hospital with dedicated thoracic pathologists was on average 10 days shorter than from other Hamilton hospitals. TAT in our hospital with dedicated thoracic pathologists improved with time from a mean of 49 days to a mean of 22 days while TAT from other hospitals did not improve over the course of the year. Our cohort mirrors findings in the literature and demonstrates that the 22C3 assay for PD-L1 can be done on both histology and cytopathology specimens; however, the insufficient rates are higher for cytopathology. Cytopathology specimens have a higher PD-L1 positivity rate, a finding that may reflect differences in tumour biology and/or stage in this subgroup. Turnaround times were different based on the hospital of origin, and suggest centralized specimen collection or use of dedicated thoracic pathologists may be advantageous. Correlation with clinical outcomes on our cytology cases will be presented in a separate abstract.