Abstract

Immune-checkpoint inhibitors targeting the PD-1/PD-L1 axis have brought significant clinical benefit in many solid cancer types, including gastrointestinal malignancies. However, it has been estimated that only 20-40% of patients respond to treatment. The pattern of expression and potential predictive value of PD-L1 as an immunohistochemical biomarker has been extensively studied in gastrointestinal neoplasms. Until now, its predictive value has been demonstrated, and is currently in use only in upper gastrointestinal malignancies (gastroesophageal adenocarcinoma and esophageal squamous cell carcinoma). In this Review, we describe the technical aspects and challenges related to PD-L1 immunohistochemical assays, the current role of PD-L1 as a biomarker in clinical practice and we outline the main studies and clinical trials analyzing the prognostic and predictive value of PD-L1 in gastrointestinal cancers.

Citation

How to cite this article: Mastracci L, Grillo F, Parente P, et al. PD-L1 evaluation in the gastrointestinal tract: from biological rationale to its clinical application. Pathologica 2022;114:352-364.

Introduction

Rudolf Virchow, the father of modern pathology, described the “lymphoreticular infiltrate” in neoplastic tissues in 1863, and hypothesized that there was a connection between cancer and inflammation ¹. On this basis, more than a century of research has shed light on the complex interaction between cancer and the host immune system, as the latter exerts antitumor activity by activation of the innate and adaptive response. However, during the process of tumor immune editing, cancer cells develop several methods of escaping the host immune response, establishing an immunosuppressive tumor microenvironment (TME) ².

Landmark studies have demonstrated that programmed death 1 (PD-L1)/programmed ligand death 1 (PD-L1)-mediated immune checkpoint has a crucial role in tumor immune escape ³. PD-1 was first described in 1992 by a group from Kyoto University as an apoptosis-associated gene ⁴. Further studies from the same group identified that PD-1 expression was found on the surface of T and B lymphocytes and was involved in the inhibition of immune response ^5,6.

PD-1 is a checkpoint protein and suppressor T-cell receptor and is part of the CD28 family. It is expressed by T and B cells, monocytes and dendritic cells. PD-L1 is a transmembrane glycoprotein of the B7 ligand family commonly expressed on the surface of antigen-presenting cells and cancer cells. Multiple signaling pathways have been identified as regulators of PD-L1 expression on tumor cells, including NFκB, MAPK, mTOR, STAT and c-Myc ⁷.

The binding of PD-L1 to PD-1 on T cells causes dephosphorylation of the T-cell receptor SHP-1/2, which in turn results in phosphorylation of the downstream proteins spleen tyrosine kinase (Syk) and phospholipid inositol-3-kinase (PI3K), inhibiting downstream signaling. PD-1/PD-L1 axis activation: i) reduces T cell-mediated immune surveillance, ii) diminishes tumor-infiltration of CD4+/CD8+ T cells, and iii) reduces cytokines including tumor necrosis factor (TNF), interferon-γ (IFN-γ) and Interleukin-2 (IL-2). Overall, this interaction leads to T cell exhaustion and apoptosis, allowing tumor cells to escape immune surveillance ⁷.

In 2010 a landmark clinical study showed striking effects of immunotherapy for the first time: in a compassionate-use trial in patients with advanced refractory melanoma, the use of Ipilimumab, a monoclonal antibody that antagonizes CTLA-4, resulted in significant clinical benefit ⁸.

Monoclonal antibody inhibitors of the PD-1/PD-L1 axis (i.e., immune checkpoint inhibitors [ICIs]) have induced remarkable clinical benefits at advanced stages in various cancer types, becoming the backbone of cancer immunotherapeutic strategies. There are currently 5,683 ongoing/completed clinical trials testing anti-PD1/PD-L1monclonal antibodies, alone or in combination with other therapeutic agents ⁹.

Until now, the Food and Drug Administration (FDA) has approved six monoclonal antibodies targeting PD-1 (Nivolumab, Pembrolizumab and Cemiplimab) or PD-L1 (Atezolizumab, Durvalumab and Avelumab) for the treatment of various cancer types, including melanoma, non-small cell lung cancer, breast cancer, gastroesophageal and colorectal cancer ¹⁰. However, only a subset of patients benefits from PD-1/PD-L1 blockade, with response rates lower than 40%. Most patients show primary resistance to single-agent ICI therapy, and longer follow-up of clinical trial populations is now revealing the development of acquired resistance ¹¹. Thus, future efforts should focus on elucidating resistance mechanisms, exploring effective predictive biomarkers, and developing novel combinational therapies.

PD-L1 testing in gastrointestinal cancers

At present, microsatellite instability (MSI)/mismatch repair protein deficiency (MMRd) and PD-L1 expression are the only predictive biomarkers approved for the use of immunotherapy in the context of gastrointestinal malignancies ¹².

Pembrolizumab is currently in use for the treatment of patients with unresectable or metastatic, MSI/MMRd solid tumors that have progressed following prior treatment and who have no satisfactory alternative treatment options. In colorectal and gastroesophageal adenocarcinomas, ICIs are used as first-line treatment in patients with metastatic or unresectable MSI/MMRd tumors ^13,14.

The pattern of expression and potential predictive value of PD-L1 has been extensively studied in gastrointestinal neoplasms. However, its role as a predictive biomarker has been demonstrated and is currently in use only for upper gastrointestinal tumors, namely: esophageal squamous cell carcinoma, esophageal and gastroesophageal junction adenocarcinoma and gastric adenocarcinoma ¹⁴. For these cancer types, the expression of PD-L1 should be evaluated using the Combined Positive Score (CPS), which consists in dividing the number of positive tumor cells, lymphocytes and macrophages, by the total number of viable tumor cells multiplied by 100 ¹⁵. Of note, tumor cells with cytoplasmic staining, neutrophils, eosinophils, plasma cells, stromal cells, necrotic cells and cellular debris should be excluded from the numerator when calculating the CPS. The alternative method of calculating PD-L1 score, currently in use for non-small cell lung cancer, is the Tumor Proportion Score (TPS), which is defined as the number of positive tumor cells divided by the total number of viable tumor cells multiplied by 100% (Fig. 1). In gastroesophageal adenocarcinomas, the CPS has been shown to outperform TPS in predicting response to ICIs ^15,16; however, a recent study by Doki et al. ¹⁷ demonstrated that both TPS and CPS can be useful predictive biomarkers in esophageal squamous cell carcinoma and should thus be included in the pathology report. The cut-off value of CPS to identify patients who might benefit from immunotherapy is subject to change. While the previous cut-off value for the use of ICIs in gastroesophageal adenocarcinomas was set to CPS ≥ 1, recent subgroup analyses of clinical trials support the use of Pembrolizumab for CPS ≥ 10 and Nivolumab for CPS ≥ 5. For this reason, the pathologist should indicate in the pathology report the exact CPS score when assessing PD-L1 expression in gastroesophageal adenocarcinoma specimens or should specify a clinically meaningful interval (i.e. CPS < 1; 1-4; 5-9; ≥ 10) in order to allow the oncologist to choose the best therapeutic option for the patient (Fig. 2) ¹⁴. For esophageal carcinoma, the TPS evaluation should be also reported, according to recent data.

PD-L1 expression can be assessed by using different immunohistochemical (IHC) assays, each consisting of a specific anti-PD-L1 antibody clone, a platform, and a scoring system. The pathology report should indicate all this information, to ensure that PD-L1 evaluation in clinical practice is comparable to the assessment within the related clinical trials ¹⁸. Of note, a distinctive requirement of the different PD-L1 IHC assays is the use of different autostainers. FDA-approved PD-L1 assays are classified as companion diagnostics. The FDA defines an assay as a companion diagnostic if it provides information that is “essential for the safe and effective use of a corresponding drug or biological product”. The FDA has approved the use of three PD-L1 IHC assays as companion diagnostics: Dako 22C3 for Pembrolizumab in patients with several solid tumors, including gastroesophageal adenocarcinoma; Ventana SP142 for Atezolizumab in patients with urothelial carcinoma, triple-negative breast cancer or non-small-cell lung cancer; and Dako 28-8 for the combination of Ipilimumab and Nivolumab in patients with NSCLC ¹⁹. Although a study provided evidence for the potential interchangeability of 22C3 PD-L1 clone (used in the KEYNOTE-061 study) and 28-8 PD-L1 clone (used in the CheckMate 649 study) ²⁰, another recent study showed only moderate concordance rate between the two assays, with higher sensitivity of the 28-8 PD-L1 clone ²¹. Moreover, several studies provided evidence that different PD-L1 assays (SP142, E1L3N, 28-8, SP263, 22C3) have only low to moderate concordance rates ^22-24. Therefore, the predictive value of different PD-L1 clones is still controversial.

PD-L1 evaluation can be challenging under certain circumstances and the pathologist must be aware of potential pitfalls. First of all, a correct evaluation of PD-L1 expression can be hampered by intra-tumoral heterogeneity. A study that used tissue microarray to analyze multiple cores from the same gastric cancer specimen found inconsistency in PD-L1 expression ²⁵. The results of the same study indicated that at least five biopsies are required to achieve a good representation of the tumor in terms of PD-L1 expression ²⁵. Additionally, when evaluating PD-L1 expression, areas of ulceration or chronic inflammatory processes (i.e., chronic gastritis) should be excluded from the CPS evaluation. A recent study has investigated PD-L1 expression patterns in gastroesophageal dysplastic lesions and found a relatively high prevalence of PD-L1 positivity among these lesions, stressing the importance to make a proper distinction between pre-invasive lesion and invasive carcinoma ²⁶. A robust interobserver agreement is necessary to guarantee the reproducibility of PD-L1 assay. The manufacturers of the 22C3 assay reported an overall percentage agreement of 96.6% when assessing gastroesophageal cancer samples ¹⁵. However, results from an independent group showed less-robust reproducibility, with intraclass correlation coefficients (ICCs) of 0.39 and 0.26 for a CPS of ≥ 1 and of 0.23 and 0.14 for a CPS of ≥ 10, using the 22C3 and SP263 clones, respectively ²⁷. These data stress the need for additional training for practicing pathologists and for additional studies on the reproducibility of the PD-L1 assays and scoring system. Another issue to consider is the lack of consistency of PD-L1 immunohistochemistry using formalin-fixed paraffin-embedded tissue older than five years.

PD-L1 in gastroesophageal cancers

Recent clinical trials have shown that esophageal, esophagogastric junction, and gastric cancer benefit from treatment with ICIs ¹⁴. These emerging data have led to the recent approval of ICIs for the treatment of oesophageal, gastroesophageal and gastric cancer by the FDA and the European Medicines Agency (EMA). Accordingly, the European Society for Medical Oncology (ESMO) has recently integrated checkpoint inhibitors targeting PD-1, namely Nivolumab and Pembrolizumab, into the current clinical guidelines, as standard-of-care for esophageal, gastroesophageal and gastric cancer ^28,29. A summary of the current recommendations by the ESMO Clinical Practice Guidelines is presented in Table I. Besides these recommendations, the FDA has recently approved the use of Pembrolizumab for advanced unresectable or metastatic HER2 positive esophagogastric/gastric cancer patients in the first line setting, in combination with chemotherapy and Trastuzumab therapy ³⁰.

In most cases, PD-L1 expression should be evaluated using the CPS scoring method. An exception is represented by the evaluation of PD-L1 in advanced/metastatic unresectable esophageal cancer, where the use of the TPS is recommended. PD-L1 positivity is defined as CPS ≥ 1 and TPS ≥ 1 ¹⁵. For predictive purposes, the threshold to select patients for immunotherapies is different according to tumor location, type of drug to be used and line of treatment: for advanced/metastatic unresectable esophageal cancer in first-line treatment the threshold to select patients for immunotherapy is TPS ≥ 1% for Nivolumab therapy, but CPS ≥ 10 if Pembrolizumab therapy is considered. For advanced/metastatic unresectable upper GI adenocarcinoma in first-line treatment, the cut-off used is CPS ≥ 5 (Tab. I) ^28,29.

PD-L1 expression occurs in approximately 50-60% of gastroesophageal or gastric adenocarcinoma cases ^15,31 and 20-80% of esophageal squamous cell carcinomas ^32-34. PD-L1 upregulation is more frequent in immune cells of the tumor microenvironment, especially at the invasive margin, rather than in tumor cells in gastric and esophagogastric adenocarcinoma. If only tumor cells ^35,36 are considered, the percentage of positive gastroesophageal and gastric cancer cases drops to 10-20% ^24,37. However, in esophageal squamous cell carcinoma cases, PD-L1 expression is more frequent in tumor cells rather than in immune or stromal cells ³⁴. The value of PD-L1 expression as a prognostic biomarker is still controversial ³⁸.

Gastric cancer is a heterogenous tumor from the morphological and molecular standpoint and tumor heterogeneity should be taken into account for the evaluation of predictive biomarkers. Indeed, PD-L1 expression may vary across distinct morphological and molecular subgroups.

From a morphological point of view, PD-L1 is more frequently expressed in intestinal and mixed type gastric adenocarcinomas (54% and 56%, respectively), as compared to diffuse gastric cancer (32%) ³⁹. Within the rare histopathological variants recognized by the WHO classification, gastric carcinoma with lymphoid stroma (also known as medullary carcinoma or lymphoepithelioma-like carcinoma) is of particular interest when considering morphological biomarkers for targeted immunotherapies: the majority of gastric carcinomas with lymphoid stroma show expression of PD-L1 in tumor (33-68%) and/or immune (77-92%) cells ^40,41.

Within the molecular subtypes proposed by The Cancer Genome Atlas (TCGA) research network ⁴², microsatellite instability (MSI) and Epstein-Barr virus (EBV) infection represent potential molecular hallmarks of response to immunotherapy and are the molecular subgroups more frequently associated with PD-L1 expression ^39,43. PD-L1 expression occurs in tumor and immune cells in up to 33% and 46%, respectively, of MSI gastric cancers ^44,45. Regarding EBV-associated gastric carcinomas, genomic amplification of the chromosomal region 9p24.1, the locus of genes encoding PD-L1 and PD-L2, has been identified in about 15% of cases ^42,45. Although the highest levels of PD-L1 mRNA and protein expression have been identified in EBV-associated cases with 9p24.1 amplification, PD-L1 expression also occurs in EBV-associated cases with no amplification, suggesting that alternative mechanisms may induce PD-L1 expression in this molecular subgroup ^40,44. Overall, PD-L1 expression was observed in tumor and immune cells in up to 50% and 94%, respectively, of EBV-positive gastric carcinomas ^44,45.

When clinical responses to Pembrolizumab therapy were analyzed according to TCGA molecular subtypes, patients harboring MSI-H (at present defined as MSI) and EBV-associated tumors achieved dramatic responses, as compared to genomically stable and chromosomal unstable subtypes ⁴⁶.

The morphological features of gastric carcinoma with lymphoid stroma are also relevant to identifying these specific molecular subtypes, as EBV infection and MSI status are identified in up to 80% of cases ^40,41,47. Based on these findings, some authors have suggested systematically performing EBER in situ hybridization and immunohistochemistry for mismatch repair proteins in gastric cancers presenting the morphologic features of gastric cancer with lymphoid stroma, for a cost-effective molecular characterization and selection of patients for targeted immunotherapies ^40,48.

PD-L1 in small bowel adenocarcinoma

Small bowel adenocarcinoma (SBA) is a rare tumor, and, in virtue of this rarity, its treatment is hindered by a lack of robust data concerning treatment efficacy. Until recently, the National Comprehensive Cancer Network^© (NCCN) guidelines suggested applying to SBA the same treatment protocol of colorectal cancer (CRC). However, SBA shares only limited similarities with CRC from a biological and molecular standpoint. In the same fashion, no strong and reliable data exist regarding PD-L1 in SBA, and its impact on both treatment and prognosis.

Variable percentages of PD-L1 positivity have been reported in the literature, ranging between 25% and 70% ^49-54. A study by Giuffrida et al. ⁵¹ reported higher percentages of PD-L1 positivity in SBA associated with celiac disease or Crohn’s disease rather than in sporadic SBA (35% vs 35% vs 5%, respectively). In another study, Thota et al. ⁴⁹ reported that SBA with pure mucinous histology or even the mucinous component of otherwise PD-L1-positive tumors did not show PD-L1 staining.

CPS is the most used score to evaluate PD-L1 expression ^49-51,55 with values ≥ 1 considered positive; the ZEBRA trial ⁵³, the largest multicenter phase II trial for Pembrolizumab use in advanced SBA, used the Modified Proportion Score (MPS), a score analogue to CPS, to evaluate PD-L1 in enrolled patients.

All these studies are characterized by relatively small study populations, owing to the rarity of the disease; moreover, some of these studies also include carcinomas of the ampulla of Vater in the group of SBA ⁵⁵, while others only include non-ampullary duodenal carcinomas.

Regarding the impact of PD-L1 expression on prognosis, Klose et al. ⁵⁰, showed a direct correlation between PD-L1 positivity and survival, in contrast with data from several other solid tumors. Giuffrida et al.,⁵¹ reported that PD-L1 expression was associated with more favorable prognosis in univariate analysis, but no statistical significance was found in multivariate analysis.

Two large trials evaluated the efficacy of immunotherapy in SBA. The first one was KEYNOTE-158 ⁵⁶ in 2019, which enrolled patients with MSI and non-colonic primary tumors. Nineteen patients with a diagnosis of SBA were enrolled, and the results showed a complete or partial response in approximately half of MSI SBA patients treated with Pembrolizumab (8/19; 42%).

The ZEBRA phase II trial ⁵³ was a large multicenter trial for Pembrolizumab use in advanced SBA, whose results came out in 2021. The results of this trial did not report any complete response, and only three partial responses (3/40; 7.5%) were reported within the study population. Overall, these results were not satisfactory in the SBA unselected population but suggested that specific subgroups of SBA patients – MSI/dMMR and patients with a tumor mutation burden (TMB) > 10mt/Mb – may benefit from treatment.

Taking into account these results, Pembrolizumab and/or Nivolumab are currently recommended as a second-line treatment for MSI SBA patients ⁵⁷.

PD-L1 in colorectal cancer

Although early studies had dismissed checkpoint inhibitors as an effective therapy in CRC, Pembrolizumab has been shown to give significant and durable responses in MSI CRCs both in pre-treated and in the naïve population by the KEYNOTE trials ^58,59. However, contrary to other solid cancers (i.e., non-small cell lung cancer), PD-L1 positive immunostaining is not required for the patient to be eligible to receive the therapy; the value of PD-L1 as a biomarker has been evaluated only in the KEYNOTE-016 trial ⁶⁰ and data on its predictive impact are very limited.

The use of MSI status as the stand-alone surrogate marker for eligibility for ICIs therapy is probably at the basis of the lack of standardization in the approach to PD-L1 evaluation in CRC. Different studies have used either the 1% or the 5% cut-offs to define positivity, while some studies have combined both intensity and percentage of positivity to grade PD-L1 immuno-reactivity, thus resulting in a plethora of papers that are scarcely comparable one to another in terms of results. This is also reflected by the variable percentage of reported PD-L1 positivity that can be found in the literature, which ranges from 9% ⁶¹ to 89% ⁶² on tumor cells and from 5% to 61% ^62,63 on stromal cells.

There is also controversy whether MSS and MSI CRC differ in PD-L1 expression, with several studies reporting higher percentages (77-100%) of PD-L1 expression in MSI CRC ^61,64,65, other studies reporting a statistically higher expression of PD-L1 in MSS CRCs ⁶⁶, and other contributions reporting no significant differences whatsoever ⁶⁷. However, given that several characteristics are shared by MSI and PD-L1 CRCs (i.e., right colon involvement, increased presence of tumor-infiltrating lymphocytes [TILs], mucinous and medullary histology), we can infer that MSI tumors may be more represented among the PD-L1 positive CRCs. Indeed, expression of PD-L1 on CRC cells has been shown to correlate with the increased presence of CD8-positive TILs ^61,66-69 and TBET-positive TILs ⁶¹. FOXP3-positive T-regulatory lymphocytes (Tregs) number among TILs has been reported to be either increased ⁷⁰ or decreased by tumor cell expression of PD-L1 ⁶².

PD-L1 expression has also been reported to be associated with poor tumor differentiation ^61,63,67, higher T, N and M stages at diagnosis ^63,67,71, higher tumor budding ^67,72, epithelial-mesenchymal transition (EMT) ⁶⁷, right-colon tumors ^63,68 and mucinous ^65,68 or medullary histology ⁶¹.

Furthermore, Rosenbaum et al. ⁶¹, reported a trend, albeit not statistically significant, toward BRAF mutations (a molecular alteration often found underlying MSI phenotype) being associated with PD-L1 positivity; it must be noted, however, that Omura et al. ⁷¹ reported inverse results, with BRAF mutations found more often in PD-L1-negative CRC. Most likely, the variance in these results can be attributed to different enrollment criteria and histological material (tissue microarrays vs whole-slide).

The prognostic impact of PD-L1 expression is controversial in CRC. PD-L1 expression is associated with a better prognosis in some studies ^65,66, and with a worse one in others ^61,73. A direct impact on overall survival (OS) has not been consistently shown, but some studies have reported a positive correlation between PD-L1 expression and longer disease-free survival (DFS) ^67,69. A study by Wyss et al. ⁶³ reported better OS and DFS in CRC showing stromal positivity for PD-L1, with no prognostic impact reported for tumor cell PD-L1 expression.

PD-L1 in anal squamous cell carcinoma

Anal squamous cell carcinoma (AnSCC) may express PD-L1 on tumor cells or tumor-related immune cells or both. PD-L1 expression by immune cells strongly correlates with CD8+ T cell density, suggesting a prevalent adaptive mechanism of PD-L1 expression driven by tumor-infiltrating lymphocytes in AnSCC ⁷⁴. However, constitutive (i.e., in absence of significant intra-tumoral immune cell infiltration) and mixed patterns of expression have also been observed ⁷⁵. The PD-L1 positivity rate is wide, ranging from 22% to 85% and this is probably partly due to the fact that various scoring systems (TPS and CPS), positivity cut-offs, and antibody clones (E1L3N, 22C3, SP263, etc.) have been used in different studies. A recent study on 62 AnSCCs, using E1LN3 antibody, reported PD-L1 expression (CPS ≥ 1) in 32% of cases ⁷⁶. No difference in PD-L1 positivity rates has been noted between HIV+ and HIV-negative patients or between HPV+ and HPV-negative cases ^74,76. In the study by Monsrud et al., HIV-positive patients with higher CD4 count were more likely to express PD-L1 on tumor tissue ⁷⁷.

Also in this tumor, the impact of PD-L1 expression on prognosis is still debated. In some studies, PD-L1 (TPS or CPS) has been associated with worse OS, especially in HPV-negative AnSCCs ^76-79, whereas in other papers PD-L1 positivity was associated with better survival ^75,80,81.

A predictive role of PD-L1 in AnSCC has been suggested ⁸². In phase II of the NCI9673 study, which was the first trial to establish the clinical benefits of immunotherapy in AnSCC, responders had higher PD-L1 expression on tumor cells ⁸³, while in the phase II KEYNOTE-158 study treatment response was observed in 15% of AnSCC patients with CPS ≥ 1 (67% of cases) versus 7% of cases with CPS < 1 ⁵⁶.

PD-L1 in pancreatic carcinoma

In pancreatic ductal adenocarcinoma (PDAC), the prognostic value of PD-L1 expression is still unclear and literature data about response to PD-1/PD-L1 inhibitors are not encouraging ⁸⁴. The peculiar milieu in PDAC, which is the continuous interaction between the glandular neoplastic component and tumor microenvironment (TME), could be the main factor affecting the poor response rate to ICI. In addition to cytokines and growth factors secreted by the TME promoting tumoral invasion, migration and angiogenesis, the intricate crosstalk between PDAC cells and TME also involves immune elements ⁸⁵. In PDAC, tumor-associated macrophages switch towards a M2 ‘immunosuppressive’ phenotype, promoting tumor immunity and tumor progression; TILs produce high levels of PD-1 and interact with PDAC cells overexpressing PD-L1, resulting in T lymphocyte depletion. In this scenario, it is imperative to remember that the TME is an ever-changing non-static system ⁸⁵. Moreover, refractoriness of PDAC to PD-1/PD-L1 inhibition could be explained by technical issues such as the fact that the intensity and extent of PD-L1 staining were not taken into account when enrolling patients in clinical trials, by the use of at least four different diagnostic IHC assays, one for each of the currently available anti–PD-1/PD-L1 therapeutics, hindering the reproducibility and uniformity of testing and reporting results, and by the use of TPS instead of CPS. Finally, preanalytical variables (sample collection, processing and storage), heterogeneity of expression of PD-L1 and adjunctive therapies may affect the results and interpretation of PD-L1 tests ⁸⁶.

Interestingly, some differences in PD-L1 expression between neoplastic cells and immune cells have been found in pancreatic cancer; moreover, the prognostic value of PD-L1 differs according to tumoral histotype/molecular subtype. In the ‘usual’ PDAC, a positive PD-L1 expression rate ranging between 19% and 55%, correlation with poor tumor differentiation, more advanced tumor stage, and worse prognosis than PD-L1–negative PDAC are reported ⁸⁶. Enrichment for PD-L1 expression in frequency and extent, in comparison with conventional PDAC, was demonstrated in the undifferentiated histotype ⁸⁶; in adenosquamous histotype ⁸⁷; in sarcomatoid ⁸⁸ and in anaplastic carcinoma with osteoclast-like giant cells ⁸⁹. In the latter subtype, PD-L1 expression has been associated with poor prognosis. No association was found between PD-L1 expression and Tumor Mutational Burden (TMB)/MSI profile in PDCA, suggesting an alternative pathway and expanding the pool of potential immunotherapy recipients ⁹⁰. Finally, PD-L1 mRNA is upregulated in squamous ‘molecular’ subtype vs other subtype in the Bailey’s data set ⁹¹.

PD-L1 in biliary tract carcinoma

Assessing the prognostic and predictive value of PD-L1 in biliary-tract cancers (BTCs) is complex due to their heterogeneous nature. In fact, BTCs can be classified in intrahepatic, extrahepatic (hilar and distal) cholangiocarcinoma, and gallbladder cancer (GBC). These anatomical subtypes, moreover, have been further molecularly and genomically characterized demonstrating significant differences in mutations in IDH1/2, FGFR2, PIK3CA, ERRB2, KRAS and BRAF genes ^92,93. Several studies have tried to shed light on the complexity of the tumor microenvironment (TME) of BTCs. In fact, the immune system in BTCs is affected by chronic inflammation which plays an important role in their carcinogenesis ⁹⁴; moreover, CD8+ and CD4+ cells in neoplastic background have been demonstrated to be related with improved outcomes in BTC ⁹⁵. Conversely, data is available which shows that PD-1 expression in TILs contributes to an immunosuppressive environment ⁹⁶.

As in pancreatic cancer, reproducibility and uniformity of results in different studies are affected by the different clones of PD-L1 used and by the use of TPS, CPS, and other scoring systems. Moreover, TME in BTC has been extensively investigated but the studies targeting PD-L1 have shown modest results in clinical activity in a small number of patients ⁹⁴.

We selected the most recent studies performed on BTC samples dealing with PD-L1 expression and its relative used immunostaining scores.

Deng et al. ⁹⁷ studied PD-L1 expression in tumor cells and TILs in 69 intrahepatic cholangiocarcinomas. Their results showed that the expression of PD-L1 in cancer cells, but not in TILs, was highly correlated with a poorer prognosis, suggesting that a ‘high’ level of expression of PD-L1 is associated with poor prognosis, and PD-L1 expression is a potential predictor and therapeutic target for intrahepatic cholangiocarcinoma.

Fontugne et al. ⁹⁸ investigated PD-L1 expression in tumor cells and TILs in 41 perihilar and 58 intrahepatic cholangiocarcinomas. The authors showed a PD-L1 positivity rate of only 9% among all cases, although this increased to 46% when including PD-L1 positive surrounding inflammatory cells, suggesting a role for ICIs in therapy.

Gani et al. ⁹⁹ examined PD-L1 in intrahepatic cholangiocarcinoma from 54 patients. Interestingly, in absence of PD-L1 expression in neoplastic cells, scoring was focused on tumor-associated macrophages (TAMs) and further subclassified in TAMs positive at the invasive tumor front (TF). The expression of PD-L1 within the TME, specifically within the TF, was found to be associated with an almost 60% shorter survival compared to patients whose tumors did not express PD-L1 in the tumor front.

Mody et al. studied the correlation between PD-L1 expression and molecular profiling in BTCs. PD-L1 positivity was found in 12% of GBC, 7% of intrahepatic, and 5% of extrahepatic cholangiocarcinoma. An increased frequency of PD-L1 expression was found in BRCA2, KRAS, and BRAF mutated BTC, suggesting the need study regimens combining specific targeted therapies with immunotherapies ⁹⁴.

Concerning the relationship between PD-L1 and molecular features, Yoon et al. ¹⁰⁰ proposed two resistance factors to PD-1/PD-L1 blockade in BTCs: KRAS alteration and Chromosomal Instability (CIN). According to their study, 95.0% of patients (19/20) having these factors did not show clinical benefit from anti-PD-1/PD-L1 agents in a PD-L1-positive cohort. Moreover, the authors directly demonstrated a suppressive immune TME with low TIL density in KRAS-altered or CIN tumors, suggesting that assessment of KRAS alteration and CIN status, combined with PD-L1 expression, could be a useful approach, and patient selection based on these factors may improve the efficacy of PD-1/PD-L1 blockade.

In conclusion, in the pancreatobiliary system, the role of PD-L1 expression in assessing the eligibility of patients for immunotherapy is limited and no agreement on PD-L1 clone and scoring system has been achieved.

Conclusions

A growing number of clinical trials are investigating the role of ICIs, alone or in combination, as therapeutic agents against several types of solid tumors, including gastrointestinal neoplasms. The use of PD-L1 IHC assays for stratification of patients and the identification of those who might benefit from ICIs is currently expanding and evolving. In this setting, the pathologist plays a central role in therapeutic decision-making. In fact, accurate biomarkers assessment is essential to ensure the best therapeutic option for the patient. PD-L1 evaluation does not come without challenges, due to the use of different companion diagnostic assays and scoring systems and the high levels of inter-assay variability. In the setting of gastrointestinal cancers, larger studies are needed to develop novel means of assessing PD-L1 expression over time and to establish the real prognostic and predictive value in bowel and pancreatobiliary neoplasms. Future perspectives include the use of digital pathology and automation, the incorporation of other biomarkers into the workflow to better reflect the tumor immune microenvironment, and the simplification of the regulatory landscape.

CONFLICTS OF INTEREST

The authors declare no conflict of interest.

FUNDING

This article was supported by an unrestricted grant from Bristol Myers Squibb.

AUTHORS’ CONTRIBUTIONS

Conceptualization: MF, LM, AV; methodology, GP and FG; data curation, AV and PP; writing-original draft preparation, AV, PP, CR, IG, MLS, MC; writing-review and editing, AV, MF, LM, FG. All authors have read and agreed to the published version of the manuscript.

Figures and tables

Figure 1.Examples of PD-L1 immunohistochemistry scoring using Combined Positive Score (CPS) and Tumor Proportion Score (TPS).

Figure 2.PD-L1 expression staining patterns by Combined Positive Score (CPS) and relative therapeutic indications in gastroesophageal adenocarcinoma. *At present, Nivolumab is indicated (Checkmate 649). **At present, Nivolumab (Checkmate 649) or Pembrolizumab (Keynote 590) are indicated.

References

1. Balkwill F, Mantovani A. Inflammation and cancer: back to Virchow?. Lancet (London, England). 2001; 357(9255):539-545. DOI
2. Grivennikov SI, Greten FR, Karin M. Immunity, inflammation, and cancer. Cell. 2010; 140(6):883-899. DOI
3. Topalian SL, Drake CG, Pardoll DM. Targeting the PD-1/B7-H1(PD-L1) pathway to activate anti-tumor immunity. Curr Opin Immunol. 2012; 24:207-212. DOI
4. Ishida Y, Agata Y, Shibahara K, Honjo T. Induced expression of PD-1, a novel member of the immunoglobulin gene superfamily, upon programmed cell death. EMBO J. 1992; 11:3887-3895. DOI
5. Agata Y, Kawasaki A, Nishimura H. Expression of the PD-1 antigen on the surface of stimulated mouse T and B lymphocytes. Int Immunol. 1996; 8:765-772. DOI
6. Nishimura H, Okazaki T, Tanaka Y. Autoimmune dilated cardiomyopathy in PD-1 receptor-deficient mice. Science. 2001; 291:319-322. DOI
7. Gou Q, Dong C, Xu H. PD-L1 degradation pathway and immunotherapy for cancer. Cell Death Dis. 2020; 11:955. DOI
8. Ku GY, Yuan J, Page DB. Single-institution experience with ipilimumab in advanced melanoma patients in the compassionate use setting: lymphocyte count after 2 doses correlates with survival. Cancer. 2010; 116:1767-1775. DOI
9. Upadhaya S, Neftelinov ST, Hodge J, Campbell J. Challenges and opportunities in the PD1/PDL1 inhibitor clinical trial landscape. Nat Rev Drug Discov. 2022; 21:482-483. DOI
10. Hargadon KM, Johnson CE, Williams CJ. Immune checkpoint blockade therapy for cancer: An overview of FDA-approved immune checkpoint inhibitors. Int Immunopharmacol. 2018; 62:29-39. DOI
11. Yarchoan M, Hopkins A, Jaffee EM. Tumor Mutational Burden and Response Rate to PD-1 Inhibition. N Engl J Med. 2017; 377:2500-2501. DOI
12. Marcus L, Lemery SJ, Keegan P, Pazdur R. FDA Approval Summary: Pembrolizumab for the Treatment of Microsatellite Instability-High Solid Tumors. Clin cancer Res an Off J Am Assoc Cancer Res. 2019; 25:3753-3758. DOI
13. Ganesh K, Stadler ZK, Cercek A. Immunotherapy in colorectal cancer: rationale, challenges and potential. Nat Rev Gastroenterol Hepatol. 2019; 16:361-375. DOI
14. Booth ME, Smyth EC. Immunotherapy in Gastro-Oesophageal Cancer: Current Practice and the Future of Personalised Therapy. BioDrugs. 2022; 36:473-485. DOI
15. Kulangara K, Zhang N, Corigliano E. Clinical Utility of the Combined Positive Score for Programmed Death Ligand-1 Expression and the Approval of Pembrolizumab for Treatment of Gastric Cancer. Arch Pathol Lab Med. 2019; 143:330-337. DOI
16. Lei M, Siemers NO, Pandya D. Analyses of PD-L1 and Inflammatory Gene Expression Association with Efficacy of Nivolumab ± Ipilimumab in Gastric Cancer/Gastroesophageal Junction Cancer Clin cancer Res an Off J Am Assoc Cancer Res. 2021; 27:3926-3935. DOI
17. Doki Y, Ajani JA, Kato K. Nivolumab Combination Therapy in Advanced Esophageal Squamous-Cell Carcinoma. N Engl J Med. 2022; 386:449-462. DOI
18. Marletta S, Fusco N, Munari E. Atlas of PD-L1 for Pathologists: Indications, Scores, Diagnostic Platforms and Reporting Systems. J Pers Med. 2022; 12(7)DOI
19. List of Cleared or Approved Companion Diagnostic Devices (In Vitro and Imaging Tools).
20. Ahn S, Kim K-M. PD-L1 expression in gastric cancer: interchangeability of 22C3 and 28-8 pharmDx assays for responses to immunotherapy. Mod Pathol an Off J United States Can Acad Pathol Inc. 2021; 34:1719-1727. DOI
21. Yeong J, Lum HYJ, Teo CB. Choice of PD-L1 immunohistochemistry assay influences clinical eligibility for gastric cancer immunotherapy. Gastric cancer Off J Int Gastric Cancer Assoc Japanese Gastric Cancer Assoc. 2022; 25:741-750. DOI
22. Heo YJ, Kim B, Kim H, Kim S, Jang MS, Kim K-M. PD-L1 expression in paired biopsies and surgical specimens in gastric adenocarcinoma: A digital image analysis study. Pathol Res Pract. 2021; 218:153338. DOI
23. Kim S-W, Jeong G, Ryu M-H, Park YS. Comparison of PD-L1 immunohistochemical assays in advanced gastric adenocarcinomas using endoscopic biopsy and paired resected specimens. Pathology. 2021; 53:586-594. DOI
24. Ma J, Li J, Qian M. PD-L1 expression and the prognostic significance in gastric cancer: a retrospective comparison of three PD-L1 antibody clones (SP142, 28-8 and E1L3N). Diagn Pathol. 2018; 13:91. DOI
25. Ye M, Huang D, Zhang Q. Heterogeneous programmed death-ligand 1 expression in gastric cancer: comparison of tissue microarrays and whole sections. Cancer Cell Int. 2020; 20:186. DOI
26. Fassan M, Brignola S, Pennelli G. PD-L1 expression in gastroesophageal dysplastic lesions. Virchows Arch. 2019. DOI
27. Tsao MS, Kerr KM, Kockx M. PD-L1 Immunohistochemistry Comparability Study in Real-Life Clinical Samples: Results of Blueprint Phase 2 Project. J Thorac Oncol Off Publ Int Assoc Study Lung Cancer. 2018; 13:1302-1311. DOI
28. Obermannová R, Alsina M, Cervantes A. Oesophageal cancer: ESMO Clinical Practice Guideline for diagnosis, treatment and follow-up. Ann Oncol Off J Eur Soc Med Oncol. 2022. DOI
29. Lordick F, Carneiro F, Cascinu S. Gastric cancer: ESMO Clinical Practice Guideline for diagnosis, treatment and follow-up. Ann Oncol Off J Eur Soc Med Oncol. 2022. DOI
30. Janjigian YY, Kawazoe A, Yañez P. The KEYNOTE-811 trial of dual PD-1 and HER2 blockade in HER2-positive gastric cancer. Nature. 2021; 600(7890):727-730. DOI
31. Knief J, Lazar-Karsten P, Hummel R, Wellner U, Thorns C. PD-L1 expression in carcinoma of the esophagogastric junction is positively correlated with T-cell infiltration and overall survival. Pathol Res Pract. 2019; 215:152402. DOI
32. Yagi T, Baba Y, Ishimoto T. PD-L1 Expression, Tumor-infiltrating Lymphocytes, and Clinical Outcome in Patients With Surgically Resected Esophageal Cancer. Ann Surg. 2019; 269:471-478. DOI
33. Guo W, Wang P, Li N. Prognostic value of PD-L1 in esophageal squamous cell carcinoma: a meta-analysis. Oncotarget. 2018; 9:13920-13933. DOI
34. Jiang Y, Lo AWI, Wong A. Prognostic significance of tumor-infiltrating immune cells and PD-L1 expression in esophageal squamous cell carcinoma. Oncotarget. 2017; 8:30175-30189. DOI
35. Thompson ED, Zahurak M, Murphy A. Patterns of PD-L1 expression and CD8 T cell infiltration in gastric adenocarcinomas and associated immune stroma. Gut. 2017; 66:794-801. DOI
36. Derks S, Nason KS, Liao X. Epithelial PD-L2 Expression Marks Barrett’s Esophagus and Esophageal Adenocarcinoma. Cancer Immunol Res. 2015; 3:1123-1129. DOI
37. Weinberg BA, Xiu J, Hwang JJ. Immuno-Oncology Biomarkers for Gastric and Gastroesophageal Junction Adenocarcinoma: Why PD-L1 Testing May Not Be Enough. Oncologist. 2018; 23:1171-1177. DOI
38. Khoshghamat N, Jafari N, Moetamani-Ahmadi M. Programmed cell death 1 as prognostic marker and therapeutic target in upper gastrointestinal cancers. Pathol Res Pract. 2021; 220:153390. DOI
39. Kim DH, Bae GE, Suh KS. Clinical Significance of Tumor and Immune Cell PD-L1 Expression in Gastric Adenocarcinoma. In Vivo. 2020; 34:3171-3180. DOI
40. Gullo I, Oliveira P, Athelogou M. New insights into the inflamed tumor immune microenvironment of gastric cancer with lymphoid stroma: from morphology and digital analysis to gene expression. Gastric cancer Off J Int Gastric Cancer Assoc Japanese Gastric Cancer Assoc. 2019; 22:77-90. DOI
41. Pyo J-S, Kim NY, Son BK, Lee HY, Oh IH, Chung KH. Clinicopathological Features and Prognostic Implication of Gastric Carcinoma with Lymphoid Stroma. Gastroenterol Res Pract. 2020; 2020:6628412. DOI
42. Comprehensive molecular characterization of gastric adenocarcinoma. Nature. 2014; 513(7517):202-209. DOI
43. Gu L, Chen M, Guo D. PD-L1 and gastric cancer prognosis: a systematic review and meta-analysis. PLoS One. 2017; 12:e0182692. DOI
44. Derks S, Liao X, Chiaravalli AM. Abundant PD-L1 expression in Epstein-Barr Virus-infected gastric cancers. Oncotarget. 2016; 7:32925-32932. DOI
45. Kawazoe A, Kuwata T, Kuboki Y. Clinicopathological features of programmed death ligand 1 expression with tumor-infiltrating lymphocyte, mismatch repair, and Epstein-Barr virus status in a large cohort of gastric cancer patients. Gastric Cancer. 2017; 20:407-415. DOI
46. Kim ST, Cristescu R, Bass AJ. Comprehensive molecular characterization of clinical responses to PD-1 inhibition in metastatic gastric cancer. Nat Med. 2018; 24:1449-1458. DOI
47. Lü B-J, Lai M, Cheng L, Xu J-Y, Huang Q. Gastric medullary carcinoma, a distinct entity associated with microsatellite instability-H, prominent intraepithelial lymphocytes and improved prognosis. Histopathology. 2004; 45:485-492. DOI
48. Setia N, Ahn S, Han HS, Park DY, Lauwers GY. Predictive value of WHO classification for PD-L1 and Her2/Neu expression and distinct associations with protein expression based classification in gastric carcinoma. Hum Pathol. 2019; 94:64-70. DOI
49. Thota R, Gonzalez RS, Berlin J. Could the PD-1 Pathway Be a Potential Target for Treating Small Intestinal Adenocarcinoma?. Am J Clin Pathol. 2017; 148:208-214. DOI
50. Klose J, Lasitschka F, Horsch C. Prognostic relevance of programmed death-ligand 1 expression and microsatellite status in small bowel adenocarcinoma. Scand J Gastroenterol. 2020; 55:321-329. DOI
51. Giuffrida P, Arpa G, Grillo F. PD-L1 in small bowel adenocarcinoma is associated with etiology and tumor-infiltrating lymphocytes, in addition to microsatellite instability. Mod Pathol. 2020; 33:1398-1409. DOI
52. Pedersen K, Smyrk TC, Harrington S, McWilliams RR. Programmed death-ligand 1 (PD-L1) expression in small bowel adenocarcinomas (SBA). J Clin Oncol. 2015; 33:3619. DOI
53. Pedersen KS, Foster NR, Overman MJ. ZEBRA: A Multicenter Phase II Study of Pembrolizumab in Patients with Advanced Small-Bowel Adenocarcinoma. Clin cancer Res an Off J Am Assoc Cancer Res. 2021; 27:3641-3648. DOI
54. Alvi MA, McQuaid S, Wilson R. Microsatellite instability and PDL1 expression in small bowel adenocarcinoma - potential for immune checkpoint inhibitor therapies. Ann Oncol. 2016; 27:vi368. DOI
55. Cardin DB, Gilbert J, Whisenant JG. Safety and Efficacy of Avelumab in Small Bowel Adenocarcinoma. Clin Colorectal Cancer. 2022. DOI
56. Marabelle A, Le DT Ascierto PA. Efficacy of Pembrolizumab in Patients With Noncolorectal High Microsatellite Instability/Mismatch Repair–Deficient Cancer: Results From the Phase II KEYNOTE-158 Study. J Clin Oncol. 2019; 38:1-10. DOI
57. Benson AB, Venook AP, Al-Hawary MM. Small Bowel Adenocarcinoma, Version 1.2020, NCCN Clinical Practice Guidelines in Oncology. J Natl Compr Canc Netw. 2019; 17:1109-1133. DOI
58. Le DT, Kim TW, Van Cutsem E. Phase II Open-Label Study of Pembrolizumab in Treatment-Refractory, Microsatellite Instability-High/Mismatch Repair-Deficient Metastatic Colorectal Cancer: KEYNOTE-164. J Clin Oncol Off J Am Soc Clin Oncol. 2020; 38:11-19. DOI
59. André T, Shiu K-K, Kim TW. Pembrolizumab in Microsatellite-Instability-High Advanced Colorectal Cancer. N Engl J Med. 2020; 383:2207-2218. DOI
60. Le DT, Uram JN, Wang H. PD-1 Blockade in Tumors with Mismatch-Repair Deficiency. N Engl J Med. 2015; 372:2509-2520. DOI
61. Rosenbaum MW, Bledsoe JR, Morales-Oyarvide V. PD-L1 expression in colorectal cancer is associated with microsatellite instability, BRAF mutation, medullary morphology and cytotoxic tumor-infiltrating lymphocytes. Mod Pathol an Off J United States Can Acad Pathol Inc. 2016; 29:1104-1112. DOI
62. Masugi Y, Nishihara R, Yang J. Tumour CD274 (PD-L1) expression and T cells in colorectal cancer. Gut. 2017; 66:1463-1473. DOI
63. Wyss J, Dislich B, Koelzer VH. Stromal PD-1/PD-L1 Expression Predicts Outcome in Colon Cancer Patients. Clin Colorectal Cancer. 2019; 18:e20-e38. DOI
64. Gatalica Z, Snyder C, Maney T. Programmed cell death 1 (PD-1) and its ligand (PD-L1) in common cancers and their correlation with molecular cancer type. Cancer Epidemiol Biomarkers Prev. 2014; 23:2965-2970. DOI
65. Al-Jussani GN, Alsughayer A, Yousuf MS. The clinicopathological features of programmed death ligand-1 expression in colorectal carcinoma. Int J Biol Markers. 2022;3936155221104122. DOI
66. Droeser RA, Hirt C, Viehl CT. Clinical impact of programmed cell death ligand 1 expression in colorectal cancer. Eur J Cancer. 2013; 49:2233-2242. DOI
67. Secinti IE, Ozgur T, Dede I. PD-L1 Expression in Colorectal Adenocarcinoma Is Associated With the Tumor Immune Microenvironment and Epithelial-Mesenchymal Transition. Am J Clin Pathol. 2022. DOI
68. Jung DH, Park HJ, Jang HH. Clinical Impact of PD-L1 Expression for Survival in Curatively Resected Colon Cancer. Cancer Invest. 2020; 38:406-414. DOI
69. Huang C-Y, Chiang S-F, Ke T-W. Clinical significance of programmed death 1 ligand-1 (CD274/PD-L1) and intra-tumoral CD8+ T-cell infiltration in stage II-III colorectal cancer. Sci Rep. 2018; 8:15658. DOI
70. Zhao L, Li C, Zhang R. B7-H1 and B7-H4 expression in colorectal carcinoma: correlation with tumor FOXP3(+) regulatory T-cell infiltration. Acta Histochem. 2014; 116:1163-1168. DOI
71. Omura Y, Toiyama Y, Okugawa Y. Prognostic impacts of tumoral expression and serum levels of PD-L1 and CTLA-4 in colorectal cancer patients. Cancer Immunol Immunother. 2020; 69:2533-2546. DOI
72. Martinez-Ciarpaglini C, Oltra S, Roselló S. Low miR200c expression in tumor budding of invasive front predicts worse survival in patients with localized colon cancer and is related to PD-L1 overexpression. Mod Pathol. 2019; 32:306-313. DOI
73. Shi S-J, Wang L-J, Wang G-D. B7-H1 expression is associated with poor prognosis in colorectal carcinoma and regulates the proliferation and invasion of HCT116 colorectal cancer cells. PLoS One. 2013; 8:e76012. DOI
74. Yanik EL, Kaunitz GJ, Cottrell TR. Association of HIV Status With Local Immune Response to Anal Squamous Cell Carcinoma: Implications for Immunotherapy. JAMA Oncol. 2017; 3:974-978. DOI
75. Iseas S, Golubicki M, Robbio J. A clinical and molecular portrait of non-metastatic anal squamous cell carcinoma. Transl Oncol. 2021; 14:101084. DOI
76. Zhu X, Jamshed S, Zou J. Molecular and immunophenotypic characterization of anal squamous cell carcinoma reveals distinct clinicopathologic groups associated with HPV and TP53 mutation status. Mod Pathol. 2021; 34:1017-1030. DOI
77. Monsrud AL, Avadhani V, Mosunjac MB. Programmed Death Ligand-1 Expression Is Associated With Poorer Survival in Anal Squamous Cell Carcinoma. Arch Pathol Lab Med. 2021. DOI
78. Govindarajan R, Gujja S, Siegel ER. Programmed Cell Death-Ligand 1 (PD-L1) Expression in Anal Cancer. Am J Clin Oncol. 2018; 41:638-642. DOI
79. Zhao Y-J, Sun W-P, Peng J-H. Programmed death-ligand 1 expression correlates with diminished CD8+ T cell infiltration and predicts poor prognosis in anal squamous cell carcinoma patients. Cancer Manag Res. 2018; 10:1-11. DOI
80. Balermpas P, Martin D, Wieland U. Human papilloma virus load and PD-1/PD-L1, CD8(+) and FOXP3 in anal cancer patients treated with chemoradiotherapy: Rationale for immunotherapy. Oncoimmunology. 2017; 6:e1288331. DOI
81. Wessely A, Heppt M V, Kammerbauer C. Evaluation of PD-L1 Expression and HPV Genotyping in Anal Squamous Cell Carcinoma Cancers (Basel). 2020; 12(9)DOI
82. Jácome AA, Morris VK, Eng C. The Role of Immunotherapy in the Treatment of Anal Cancer and Future Strategies. Curr Treat Options Oncol. 2022; 23:1073-1085. DOI
83. Morris VK, Salem ME, Nimeiri H. Nivolumab for previously treated unresectable metastatic anal cancer (NCI9673): a multicentre, single-arm, phase 2 study. Lancet Oncol. 2017; 18:446-453. DOI
84. Fassan M, Scarpa A, Remo A. Current prognostic and predictive biomarkers for gastrointestinal tumors in clinical practice. Pathologica. 2020; 112:248-259. DOI
85. Parente P, Parcesepe P, Covelli C. Crosstalk between the Tumor Microenvironment and Immune System in Pancreatic Ductal Adenocarcinoma: Potential Targets for New Therapeutic Approaches. Gastroenterol Res Pract. 2018; 2018:7530619. DOI
86. Lehrke HD, Graham RP, McWilliams RR. Undifferentiated Pancreatic Carcinomas Display Enrichment for Frequency and Extent of PD-L1 Expression by Tumor Cells. Am J Clin Pathol. 2017; 148:441-449. DOI
87. Lee SM, Sung CO. PD-L1 expression and surgical outcomes of adenosquamous carcinoma of the pancreas in a single-centre study of 56 lesions. Pancreatology. 2021; 21(5):920-927. DOI
88. Silvestris N, Argentiero A, Brunetti O. PD-L1 and Notch as novel biomarkers in pancreatic sarcomatoid carcinoma: a pilot study. Expert Opin Ther Targets. 2021; 25:1007-1016. DOI
89. Luchini C, Cros J, Pea A. PD-1, PD-L1, and CD163 in pancreatic undifferentiated carcinoma with osteoclast-like giant cells: expression patterns and clinical implications. Hum Pathol. 2018; 81:157-165. DOI
90. Vanderwalde A, Spetzler D, Xiao N. Microsatellite instability status determined by next-generation sequencing and compared with PD-L1 and tumor mutational burden in 11,348 patients. Cancer Med. 2018; 7:746-756. DOI
91. Bailey P, Chang DK, Nones K. Genomic analyses identify molecular subtypes of pancreatic cancer. Nature. 2016; 531(7592):47-52. DOI
92. Simbolo M, Fassan M, Ruzzenente A. Multigene mutational profiling of cholangiocarcinomas identifies actionable molecular subgroups. Oncotarget. 2014; 5:2839-2852. DOI
93. Normanno N, Martinelli E, Melisi D. Role of molecular genetics in the clinical management of cholangiocarcinoma. ESMO open. 2022; 7:100505. DOI
94. Mody K, Starr J, Saul M. Patterns and genomic correlates of PD-L1 expression in patients with biliary tract cancers. J Gastrointest Oncol. 2019; 10:1099-1109. DOI
95. Goeppert B, Frauenschuh L, Zucknick M. Prognostic impact of tumour-infiltrating immune cells on biliary tract cancer. Br J Cancer. 2013; 109:2665-2674. DOI
96. Ahmadzadeh M, Johnson LA, Heemskerk B. Tumor antigen-specific CD8 T cells infiltrating the tumor express high levels of PD-1 and are functionally impaired. Blood. 2009; 114:1537-1544. DOI
97. Deng M, Li S-H, Fu X. Relationship between PD-L1 expression, CD8+ T-cell infiltration and prognosis in intrahepatic cholangiocarcinoma patients. Cancer Cell Int. 2021; 21:371. DOI
98. Fontugne J, Augustin J, Pujals A. PD-L1 expression in perihilar and intrahepatic cholangiocarcinoma. Oncotarget. 2017; 8:24644-24651. DOI
99. Gani F, Nagarajan N, Kim Y. Program Death 1 Immune Checkpoint and Tumor Microenvironment: implications for patients with intrahepatic cholangiocarcinoma. Ann Surg Oncol. 2016; 23:2610-2617. DOI
100. Yoon JG, Kim MH, Jang M. Molecular Characterization of Biliary Tract Cancer Predicts Chemotherapy and Programmed Death 1/Programmed Death-Ligand 1 Blockade Responses. Hepatology. 2021; 74:1914-1931. DOI