Evaluation of an online training tool for scoring programmed cell death ligand-1 (PD-L1) diagnostic tests for lung cancer.
ABSTRACT: BACKGROUND:Numerous studies indicate that higher tumour programmed cell death ligand-1 (PD-L1) expression is associated with greater response to anti-programmed cell death-1 (PD-1)/PD-L1 immunotherapy in non-small cell lung cancer (NSCLC). In the era of precision medicine, there is a need to provide reliable, standardised training for pathologists to improve their accuracy of interpretation and scoring, as the results are used directly to inform clinical decisions. Here we present findings regarding reader reproducibility of PD-L1 tumour cell (TC) staining scoring for NSCLC using a PD-L1 e-trainer tool as part of a PD-L1 immunohistochemistry reader training course. METHODS:The PD-L1 training course was developed based on the use of VENTANA PD-L1 (SP263) and Dako PD-L1 IHC PharmDx 22C3 stained NSCLC samples in combination with a PD-L1 e-trainer tool. Five-hundred formalin-fixed, paraffin-embedded archival samples were obtained from commercial sources and stained for PD-L1. Slides were scored by two expert pathologists, then scanned to produce digital images and re-scored. Thirty-three cases were selected and sorted into three sets: a training set and two self-assessment tests (pre-test and 'competence' test). Participants (all selected board-certified pathologists) received face-to-face training including use of an e-trainer tool. Statistical analyses were performed using the competence test set. Overall percentage agreement (OPA) was assessed between the participant pathologists' registered scores and the reference scores assigned by expert pathologists at clinically relevant PD-L1 cut-offs (?1%, ?25% and???50%). RESULTS:Seven sessions were held and 69 participant pathologists completed the training. Inter-reader concordance indicated high OPA (85-95%) for PD-L1 TC scoring at clinically relevant cut-offs, with Fleiss' Kappa >?0.5. CONCLUSIONS:Use of this web-based training tool incorporated into classroom-style training was associated with an overall moderately good level of inter-reader reproducibility at key cut-offs for TC PD-L1 expression testing in NSCLC. Overall, the online training tool offers a means of standardised training for practising pathologists in a clinical setting.
Project description:Cancer immunotherapies, such as atezolizumab, are proving to be a valuable therapeutic strategy across indications, including non-small cell lung cancer (NSCLC) and urothelial cancer (UC). Here, we describe a diagnostic assay that measures programmed-death ligand 1 (PD-L1) expression, via immunohistochemistry, to identify patients who will derive the most benefit from treatment with atezolizumab, a humanized monoclonal anti-PD-L1 antibody. We describe the performance of the VENTANA PD-L1 (SP142) Assay in terms of specificity, sensitivity, and the ability to stain both tumor cells (TC) and tumor-infiltrating immune cells (IC), in NSCLC and UC tissues. The reader precision, repeatability and intermediate precision, interlaboratory reproducibility, and the effectiveness of pathologist training on the assessment of PD-L1 staining on both TC and IC were evaluated. We detail the analytical validation of the VENTANA PD-L1 (SP142) Assay for PD-L1 expression in NSCLC and UC tissues and show that the assay reliably evaluated staining on both TC and IC across multiple expression levels/clinical cut-offs. The reader precision showed high overall agreement when compared with consensus scores. In addition, pathologists met the predefined training criteria (≥85.0% overall percent agreement) for the assessment of PD-L1 expression in NSCLC and UC tissues with an average overall percent agreement ≥95.0%. The assay evaluates PD-L1 staining on both cell types and is robust and precise. In addition, it can help to identify those patients who may benefit the most from treatment with atezolizumab, although treatment benefit has been demonstrated in an all-comer NSCLC and UC patient population.
Project description:The level of PD-L1 expression in immunohistochemistry (IHC) assays is a key biomarker for the identification of Non-Small-Cell-Lung-Cancer (NSCLC) patients that may respond to anti PD-1/PD-L1 treatments. The quantification of PD-L1 expression currently includes the visual estimation by a pathologist of the percentage (tumor proportional scoring or TPS) of tumor cells showing PD-L1 staining. Known challenges like differences in positivity estimation around clinically relevant cut-offs and sub-optimal quality of samples makes visual scoring tedious and subjective, yielding a scoring variability between pathologists. In this work, we propose a novel deep learning solution that enables the first automated and objective scoring of PD-L1 expression in late stage NSCLC needle biopsies. To account for the low amount of tissue available in biopsy images and to restrict the amount of manual annotations necessary for training, we explore the use of semi-supervised approaches against standard fully supervised methods. We consolidate the manual annotations used for training as well the visual TPS scores used for quantitative evaluation with multiple pathologists. Concordance measures computed on a set of slides unseen during training provide evidence that our automatic scoring method matches visual scoring on the considered dataset while ensuring repeatability and objectivity.
Project description:<h4>Background</h4>Several anti-programmed cell death-1 (PD-1) and anti-programmed cell death ligand-1 (PD-L1) therapies have shown encouraging safety and clinical activity in a variety of tumor types. A potential role for PD-L1 testing in identifying patients that are more likely to respond to treatment is emerging. PD-L1 expression in clinical practice is determined by testing one tumor section per patient. Therefore, it is critical to understand the impact of tissue sampling variability on patients' PD-L1 classification.<h4>Methods</h4>Resected non-small cell lung cancer (NSCLC), head and neck squamous cell carcinoma (HNSCC) and urothelial carcinoma (UC) tissue samples (five samples per tumor type) were obtained from commercial sources and two tumor blocks were taken from each. Three sections from each block (~?100 ?m apart) were stained using the VENTANA PD-L1 (SP263) assay, and scored based on the percentage of PD-L1-staining tumor cells (TCs) or tumor-infiltrating immune cells (ICs) present. Each section was categorized as PD-L1 high or low/negative using a variety of cut-off values, and intra-block and intra-case (between blocks of the same tumor) concordance (overall percentage agreement [OPA]) were evaluated. An additional 200 commercial NSCLC samples were also analyzed, and intra-block concordance determined by scoring two sections per sample (?70 ?m apart).<h4>Results</h4>Concordance in TC PD-L1 classification was high at all applied cut-offs. Intra-block and intra-case OPA for the 15 NSCLC, HNSCC or UC samples were 100% and 80-100%, respectively, across all cut-offs; intra-block OPA for the 200 NSCLC samples was 91.0-98.5% across all cut-offs. IC PD-L1 classification was less consistent; intra-block and intra-case OPA for the 15 NSCLC, HNSCC or UC samples ranged between 70 and 100% and between 60 and 100%, respectively, with similar observations in the intra-block analysis of the 200 NSCLC samples.<h4>Conclusions</h4>These results show the reproducibility of TC PD-L1 classification across the depth of the tumor using the VENTANA PD-L1 (SP263) assay. Practically, this means that treatment decisions based on TC PD-L1 classification can be made confidently, following analysis of one tumor section. Although more variable than TC staining, consistent IC PD-L1 classification was also observed within and between blocks and across cut-offs.
Project description:The expression of programmed death (PD) ligand 1 (PD-L1) protein expression assessed by immunohistochemistry (IHC) has been correlated with response and survival benefit from anti-PD-1/PD-L1 immune checkpoint inhibitor therapies in advanced non-small cell lung carcinoma (NSCLC). The efficacy of several agents appears correlated with PD-L1 expression. It remains controversial whether PD-L1 is prognostic in NSCLC. We assessed the prognostic value of PD-L1 IHC and its predictive role for adjuvant chemotherapy in early stage NSCLC.Tumor sections from three pivotal adjuvant chemotherapy trials (IALT, JBR.10, CALGB 9633) using the E1L3N antibody were studied in this pooled analysis. PD-L1 staining intensity and percentage in both tumor cells (TCs) and immune cells (ICs) were scored by two pathologists. The average or consensus PD-L1 expression levels across intensities and/or percent cells stained were correlated with clinicopathological and molecular features, patient survivals and potential benefit of adjuvant chemotherapy.Results from 982 patients were available for analysis. Considering staining at any intensities for overall PD-L1 expression, 314 (32.0%), 204 (20.8%) and 141 (14.3%) tumor samples were positive for PD-L1 staining on TCs using cut-offs at??1%, ?10% and??25%, respectively. For PD-L1 expressing ICs, 380 (38.7%), 308 (31.4%) and 148 (15.1%) were positive at???1%, ?10% and 25% cut-offs, respectively. Positive PD-L1 was correlated with squamous histology, intense lymphocytic infiltrate, and KRAS but not with TP53 mutation. EGFR mutated tumors showed statistically non-significant lower PD-L1 expression. PD-L1 expression was neither prognostic with these cut-offs nor other exploratory cut-offs, nor were predictive for survival benefit from adjuvant chemotherapy.PD-L1 IHC is not a prognostic factor in early stage NSCLC patients. It is also not predictive for adjuvant chemotherapy benefit in these patients.
Project description:Programmed cell death ligand 1 (PD-L1) immunohistochemistry is used to determine which patients with advanced non-small-cell lung cancer (NSCLC) respond best to treatment with PD-L1 inhibitors. For each inhibitor, a unique immunohistochemical assay was developed. This systematic review gives an up-to-date insight into the comparability of standardised immunohistochemical assays and laboratory-developed tests (LDTs), focusing specifically on tumour cell (TC) staining and scoring. A systematic search was performed identifying publications that assessed interassay, interobserver and/or interlaboratory concordance of PD-L1 assays and LDTs in tissue of NSCLC patients. Of 4294 publications identified through the systematic search, 27 fulfilled the inclusion criteria and were of sufficient methodological quality. Studies assessing interassay concordance found high agreement between assays 22C3, 28-8 and SP263 and properly validated LDTs, and lower concordance for comparisons involving SP142. A decrease in concordance, however, is seen with use of cut-offs, which hampers interchangeability of PD-L1 immunohistochemistry assays and LDTs. Studies assessing interobserver concordance found high agreement for all assays and LDTs, but lower agreement with use of a 1% cut-off. This may be problematic in clinical practice, as discordance between pathologists at this cut-off may result in some patients being denied valuable treatment options. Finally, five studies assessed interlaboratory concordance and found moderate to high agreement levels for various assays and LDTs. However, to assess the actual existence of interlaboratory variation in PD-L1 testing and PD-L1 positivity in clinical practice, studies using real-world clinical pathology data are needed.
Project description:BACKGROUND:A high-quality programmed cell-death ligand 1 (PD-L1) diagnostic assay may help predict which patients are more likely to respond to anti-programmed cell death-1 (PD-1)/PD-L1 antibody-based cancer therapy. Here we describe a PD-L1 immunohistochemical (IHC) staining protocol developed by Ventana Medical Systems Inc. and key analytical parameters of its use in formalin-fixed, paraffin-embedded (FFPE) samples of non-small cell lung cancer (NSCLC) and head and neck squamous cell carcinoma (HNSCC). METHODS:An anti-human PD-L1 rabbit monoclonal antibody (SP263) was optimized for use with the VENTANA OptiView DAB IHC Detection Kit on the automated VENTANA BenchMark ULTRA platform. The VENTANA PD-L1 (SP263) Assay was validated for use with FFPE NSCLC and HNSCC tissue samples in a series of studies addressing sensitivity, specificity, robustness, and precision. Samples from a subset of 181 patients from a Phase 1/2 study of durvalumab (NCT01693562) were analyzed to determine the optimal PD-L1 staining cut-off for enriching the probability of responses to treatment. The scoring algorithm was defined using statistical analysis of clinical response data from this clinical trial and PD-L1 staining parameters in HNSCC and NSCLC tissue. Inter-reader agreement was established by three pathologists who evaluated 81 NSCLC and 100 HNSCC samples across the range of PD-L1 expression levels. RESULTS:The VENTANA PD-L1 (SP263) Assay met all pre-defined acceptance criteria. For both cancer types, a cut-off of 25 % of tumor cells with PD-L1 membrane staining of any intensity best discriminated responders from nonresponders. Samples with staining above this value were deemed to have high PD-L1 expression, and those with staining below it were deemed to have low or no PD-L1 expression. Inter-reader agreement on PD-L1 status was 97 and 92 % for NSCLC and HNSCC, respectively. CONCLUSIONS:These results highlight the robustness and reproducibility of the VENTANA PD-L1 (SP263) Assay and support its suitability for use in the evaluation of NSCLC and HNSCC FFPE tumor samples using the devised ?25 % tumor cell staining cut-off in a clinical setting. The clinical utility of the PD-L1 diagnostic assay as a predictive biomarker will be further validated in ongoing durvalumab studies. TRIAL REGISTRATION:ClinicalTrials.gov: NCT01693562.
Project description:Programmed death-ligand 1 (PD-L1) expression on tumor cells (TC) or tumor-infiltrating immune cells (IC) correlated in several studies with PD-L1/programmed death-1 (PD-1) checkpoint inhibitor efficacy. Since June 2018, a positive PD-L1 status is required for atezolizumab or pembrolizumab treatment of patients with advanced or metastasized urothelial bladder cancer, who are ineligible for cisplatin-containing therapy. We examined technical comparability and inter-reader agreement of four clinically developed PD-L1 assays in locally advanced disease. Archived, formalin-fixed, paraffin-embedded sections from 30 patients (73.3% cystectomies, 26.7% transurethral resections) were stained by PD-L1 immunohistochemistry using VENTANA SP142, VENTANA SP263, DAKO 22C3, and DAKO 28-8 at two sites per manufacturers' protocols and scored blinded at five sites for PD-L1 expression on IC (% per tumor area) and TC (%). Small, non-significant inter-assay differences were observed for IC. For TC, SP142 showed significantly lower staining percentages. Pairwise comparisons revealed -?0.3 to 1.6% differences in adjusted means between assays for IC, and for TC, -?10.5 to -?7.8% (SP142 versus others) and -?1.9 to 2.7% (other comparisons). Inter-reader and inter-assay agreement was moderate to high for both IC and TC. Allocation to binary cutoffs (1%, 5%, 10%) showed substantial to high Kappa agreement scores (0.440-0.923) for IC and TC between assays for each reader. This first multicenter study, with five independent readers blinded with respect to the assay used, suggests that all four currently clinically relevant assays are analytically similar for evaluation of PD-L1-stained IC and three (SP263, 22C3, and 28-8) for PD-L1-stained TC. Inter-observer agreement for trained readers in scoring of both IC and TC positivity was generally high.
Project description:BACKGROUND:In non-small cell lung cancer (NSCLC), PD-L1 expression on either tumor cells (TC) or both TC and tumor-infiltrating immune cells (IC) is currently the most used biomarker in cancer immunotherapy. However, the mechanisms involved in PD-L1 regulation are not fully understood. To provide better insight in these mechanisms, a multiangular analysis approach was used to combine protein and mRNA expression with several clinicopathological characteristics. PATIENTS AND METHODS:Archival tissues from 640 early stage, resected NSCLC patients were analyzed with immunohistochemistry for expression of PD-L1 and CD8 infiltration. In addition, mutational status and expression of a selection of immune genes involved in the PD-L1/PD-1 axis and T-cell response was determined. RESULTS:Tumors with high PD-L1 expression on TC or on IC represent two subsets of NSCLC with minimal overlap. We observed that PD-L1 expression on IC irrespective of expression on TC is a good marker for inflammation within tumors. In the tumors with the highest IC expression and absent TC expression an association with reduced IFN? downstream signaling in tumor cells was observed. CONCLUSIONS:These results show that PD-L1 expression on TC and IC are both independent hallmarks of the inflamed phenotype in NSCLC, and TC-negative/IC-high tumors can also be categorized as inflamed. The lack of correlation between PD-L1 TC and IC expression in this subgroup may be caused by impaired IFN? signaling in tumor cells. These findings may bring a better understanding of the tumor-immune system interaction and the clinical relevance of PD-L1 expression on IC irrespective of PD-L1 expression on TC.
Project description:Programmed death-ligand 1 (PD-L1) expression on tumor cells (TCs) by immunohistochemistry is rapidly gaining importance as a diagnostic for the selection or stratification of patients with non-small cell lung cancer (NSCLC) most likely to respond to single-agent checkpoint inhibitors. However, at least two distinct patterns of PD-L1 expression have been observed with potential biological and clinical relevance in NSCLC: expression on TC or on tumor-infiltrating immune cells (ICs). We investigated the molecular and cellular characteristics associated with PD-L1 expression in these distinct cell compartments in 4,549 cases of NSCLC. PD-L1 expression on IC was more prevalent and likely reflected IFN-?-induced adaptive regulation accompanied by increased tumor-infiltrating lymphocytes and effector T cells. High PD-L1 expression on TC, however, reflected an epigenetic dysregulation of the PD-L1 gene and was associated with a distinct histology described by poor immune infiltration, sclerotic/desmoplastic stroma, and mesenchymal molecular features. Importantly, durable clinical responses to atezolizumab (anti-PD-L1) were observed in patients with tumors expressing high PD-L1 levels on either TC alone [40% objective response rate (ORR)] or IC alone (22% ORR). Thus, PD-L1 expression on TC or IC can independently attenuate anticancer immunity and emphasizes the functional importance of IC in regulating the antitumor T cell response.
Project description:Inhibition of the PD-1/PD-L1 pathway may induce anticancer immune responses in non-small cell lung cancer (NSCLC). Two PD-L1 immunohistochemistry (IHC) assays have been approved as companion diagnostic tests for therapeutic anti-PD-1 antibodies. However, many aspects of PD-L1 prevalence and association with genetically defined subtypes have not been addressed systematically. Here, we analyzed PD-L1 expression in 436 genetically annotated NSCLC specimens enriched for early stages using PD-L1 antibody 5H1. Expression of PD-L1 was detected in the tumor cells (TC) (34% of cases) and in associated immune cells (IC) (49%) across all stages of NSCLC, either alone or in combination. PD-L1 IHC-positive TC, but not IC showed significantly higher PD-L1 RNA expression levels. Expression in TC was associated with TP53, KRAS and STK11 mutational status in adenocarcinomas (AD) and with NFE2L2 mutations in squamous cell carcinomas (SQ). No correlations with histological subtype, clinical characteristics and overall survival were found. The presence of PD-L1-positive IC was significantly associated with patients' smoking status in AD. The findings are in agreement with the emerging concept that tumors with high mutational burden are more likely to benefit from immunotherapy, since TP53 and KRAS mutations are linked to smoking, increased numbers of somatic mutations and expression of neoantigens. Current clinical studies focus on stage IIIB and IV NSCLC; however, PD-L1 expression occurs in earlier stages and might be a predictive biomarker in clinical trials testing (neo-) adjuvant strategies.