Abstract
Background/Aim: The efficacy of programmed cell death 1 (PD-1) inhibitor therapy for patients with recurrent and/or metastatic salivary gland carcinoma (R/M SGC) remains unclear. Patients and Methods: We retrospectively analyzed 36 patients with R/M SGC treated with PD-1 inhibitor. The expression of programmed cell death ligand 1 (PD-L1) and mismatch repair (MMR) proteins was also analyzed. Results: The objective response rate (ORR) was 11.1%. The histopathological subtypes of patients who achieved complete response or partial response were salivary duct carcinoma (SDC) in three patients and poorly differentiated carcinoma in one patient, all of whom showed a positive PD-L1 expression. The expression of MMR proteins was not associated with the efficacy of PD-1 inhibitors. Conclusion: Although the efficacy of PD-1 inhibitor therapy in R/M SGC is limited, certain patients may respond and achieve long-term disease control. There is a potential therapeutic effect in SDC patients with positive PD-L1 expression.
Salivary gland carcinoma (SGC) is rare and accounts for only 0.3% of all malignancies and 6% of head and neck malignancies (1). It is classified into 21 histopathological types according to the latest World Health Organization (WHO) classification, and their malignant potential and clinical behavior considerably vary according to histopathological type (1). The first treatment option for SGC is surgery, with the addition of postoperative radiation therapy in cases with adverse features (intermediate or high histopathological grade, local progression, close or positive surgical margins, lymph node metastases, neural/perineural invasion, lymphatic/vascular invasion) (2, 3). However, standard treatment strategies and effective chemotherapy regimens for distant metastases or unresectable cases have not been established. The latest National Comprehensive Cancer Network (NCCN) guidelines also lack mention of a preferred regimen for patients with recurrent and/or metastatic SGC (R/M SGC), so we are forced to choose from among a selection of available regimens, including those mentioned in clinical trials, based on the histopathological type and molecular characteristics of each case (3).
Programmed cell death 1 (PD-1) inhibitors, such as nivolumab and pembrolizumab, have shown clinical efficacy in the treatment of patients with recurrent and/or metastatic head and neck squamous cell carcinoma (R/M HNSCC), and their use in clinical practice is expanding due to their sustained efficacy and relatively low incidence of serious adverse events (4-6). However, there are limited data regarding the efficacy and safety for patients with R/M SGC receiving PD-1 inhibitor therapy (7-9). In cases of R/M HNSCC, programmed cell death ligand 1 (PD-L1) expression has been reported to be associated with the therapeutic effect of PD-1 inhibitors (4, 6). In previous reports, PD-L1 expression is frequently observed in high-grade histological types of SGC and has been suggested to be a poor prognostic factor; however, the relationship between PD-L1 expression and the efficacy of PD-1 inhibitors in SGCs remains to be elucidated (10, 11). Furthermore, little is known on the clinical significance of microsatellite instability (MSI) in SGC, which is known to be associated with the efficacy of PD-1 inhibitors in malignancies of other organs. Therefore, it is important to identify predictors of response to PD-1 inhibitors in order to establish personalized treatment strategies for R/M SGCs.
In the present study, we investigated the efficacy of PD-1 inhibitors in R/M SGC patients in real-world clinical practice and assessed the relationship between histopathological type, expression of PD-L1 and mismatch repair (MMR) proteins as a surrogate for MSI high, and the efficacy of PD-1 inhibitors.
Patients and Methods
Study design and patient selection. This was a non-interventional, retrospective study conducted at eight research facilities, including Kyushu University Hospital. We reviewed the medical records of 36 patients with R/M SGC treated with PD-1 inhibitors (nivolumab or pembrolizumab) at the Kyushu University Hospital (Fukuoka, Japan) or other participating Institutions from March 2017 to March 2021. In addition to SGC cases with major salivary gland primary lesions, cases with minor salivary gland primary lesions, such as in the oral cavity, pharynx, or paranasal sinuses, were also included. Surveyed patients received nivolumab (240 mg/body, once every 2 weeks) or pembrolizumab (200 mg/body, once every 3 weeks) as a treatment in real-world clinical practice until the confirmation of obvious disease progression or occurrence of unacceptable toxicity. The response to the treatment was evaluated using the Response Evaluation Criteria in Solid Tumors version1.1 (RECIST 1.1), based on the findings of computed tomography performed every 8 to 12 weeks (12). We then evaluated and categorized the best overall response (BOR) of all patients during the period of PD-1 inhibitor monotherapy as a complete response (CR), partial response (PR), stable disease (SD), or progressive disease (PD). The objective response rate (ORR) corresponded to the percentage of patients with CR and PR, and the disease control rate (DCR) corresponded to the percentage of patients with CR, PR, and SD.
For the survival analysis, the overall survival (OS) was defined as the time from the start of PD-1 inhibitor treatment to death from any cause, and the progression-free survival (PFS) was defined as the time from the start of PD-1 inhibitor treatment to the diagnosis of disease progression or death. To assess the safety of treatment, we also investigated the occurrence of immune-related adverse events (irAEs). In addition to analyzing clinical information, we analyzed the expression of PD-L1 and MMR proteins immunohistochemically using the residual specimens of tissue samples collected during routine medical procedures (biopsies and surgeries) and evaluated the relationship between the expression of these molecules and the response to PD-1 inhibitors.
The institutional review boards of Kyushu University Hospital (No. 2020-492) and each participating institution approved this study, which was conducted according to the Declaration of Helsinki. Although informed consent was not obtained from each patient, they were given the opportunity to decline to have their clinical records or tissue samples used for research (opt-out consent provision).
Immunohistochemical procedures and evaluations. An immunohistochemical examination was feasible in 35 cases, except for in 1 case with a poorly preserved specimen. Sections (4-μm-thick) were prepared from formalin-fixed paraffin-embedded tissue materials and analyzed immunohistochemically using a high-sensitivity EnVision FLEX system (Dako, Glostrup, Denmark) with the following primary antibodies: MLH1 (clone G168-15; BD Biosciences, San Jose, CA, USA), MSH2 (clone FE11; Calbiochem, Gibbstown, CA, USA), PMS2 (clone A16-4; BD Biosciences), and MSH6 (clone EP49; Dako). An immunohistochemical examination for PD-L1 was performed by LSI Medience Corporation (Tokyo, Japan) using the Dako PD-L1 IHC 22C3 pharmDx assay.
We assessed the PD-L1 expression as follows: tumor proportion score (TPS), percentage of PD-L1 positive viable tumor cells; combined positive score (CPS), proportion of all PD-L1 positive cells (tumor cells, lymphocytes and macrophages) among tumor cells in the same area. Cases with the above scores ≥1 were judged to be positive for PD-L1 expression. The expression of MMP proteins was judged as lost when there was a complete absence of nuclear staining in tumor cells while the surrounding non-tumor cells consistently exhibited nuclear staining. In cases with expression loss of ≥1 MMR protein, the tumor was considered mismatch repair-deficient (dMMR).
Statistical analyses. All calculations were performed using the JMP 14 software program (SAS Institute, Cary, NC, USA). Pearson’s χ2 test was used to analyze the associations between clinicopathological features and the efficacy of PD-1 inhibitors. The patient survival was analyzed using the Kaplan-Meier method and assessed using a log-rank test. All statistical analyses were considered significant for p-values <0.05.
Results
Patient clinicopathological characteristics. The clinicopathological characteristics and treatment status of 36 patients with R/M SGC are summarized in Table I. The median follow-up period for all patients was 6.8 (range=1-31) months. The primary site of the tumor was major salivary glands in 25 patients (69.4%) and minor salivary glands in 11 patients (30.5%). Histopathologically, 29 (80.5%) patients had non-adenoid cystic carcinoma (non-ACC)-type tumors, and 7 (19.4%) had ACC-type tumors. Twenty-five patients (69.4%) had received prior radiation therapy, and 18 (50.0%) had received prior treatment with cetuximab. Regarding PD-1 inhibitor therapy, 28 patients (77.7%) were treated with nivolumab, and 8 (22.2%) were treated with pembrolizumab. The median number of cycles of PD-1 inhibitor therapy was 7 (range=1-52).
Clinicopathological characteristics and treatment status of patients with R/M SGC (n=36).
Response to PD-1 inhibitors and histopathological types. The response to PD-1 inhibitor therapy for R/M SGCs are summarized in Table II. The best overall response (BOR) for PD-1 inhibitors in all cases was CR in 1 (2.7%), PR in 3 (8.3%), SD in 7 (19.4%), and PD in 25 (69.4%). The ORR and DCR were 11.1% and 30.5%, respectively. The histopathological type of patients who responded to PD-1 inhibitor therapy (CR or PR) was non-ACC type in four cases, of whom three had SDC (one CR and two PR) and one poorly differentiated carcinoma (PR). In contrast, none of the patients with the ACC histopathological type showed any objective response to PD-1 inhibitor therapy. The 1- and 2-year OS/PFS rates were 57.8%/25.0% and 46.2%/25.0%, respectively, as shown in Figure 1.
Treatment efficacy of PD-1 inhibitor therapy associated with histopathological type (n=36).
A Kaplan–Meier analysis of (A) overall survival (OS) and (B) progression-free survival (PFS) after PD-1 inhibitor therapy. The 1- and 2-year OS/PFS rates were 57.8%/25.0% and 46.2%/25.0%, respectively.
Expression of PD-L1 and MMR protein and treatment efficacy of PD-1 inhibitors. The expression of PD-L1 and MMR proteins is summarized in Table III. A total of 11 (31.4%) and 24 (68.5%) of 35 cases examined showed a positive PD-L1 expression with TPS ≥1 and CPS ≥1, respectively. Regarding the histopathological type, PD-L1 expression was confirmed in 39.2% (TPS ≥1) and 85.7% (CPS ≥1) of patients with non-ACC histopathological subtypes, including SDC, while none of the patients with ACC showed a positive expression for PD-L1. There was a statistically significant difference in the PD-L1 expression rate between non-ACC-type patients and ACC-type patients (p=0.045; TPS, and p<0.0001; CPS). The histopathological type of 2 patients with a high PD-L1 expression (both TPS and CPS ≥20) was SDC. Representative examples of immunohistochemical staining for PD-L1 in SGCs are shown in Figure 2A and B. None of the cases examined (n=35) showed the loss of expression of any MMR proteins, regardless of the histopathological type, so all cases were judged as mismatch repair-proficient (pMMR). Representative examples of immunohistochemical staining for MMR proteins in SGCs are shown in Figure 2C-F.
Expression status of the PD-L1 and MMR proteins associated with histopathological type (n=35).
The expression of programmed cell death ligand 1 (PD-L1) and mismatch repair (MMR) proteins in salivary gland carcinoma. Representative examples of negative and positive PD-L1 expression in the tumor types as follows: (A) adenoid cystic carcinoma (negative) and (B) salivary duct carcinoma (positive). Representative examples of (C) MLH1, (D) MSH2, (E) MSH6, and (F) PMS2 expression in a case with salivary duct carcinoma. The expression of MMR proteins was retained in all cases, regardless of the histopathological type.
The association between PD-L1 expression and treatment efficacy of PD-1 inhibitors is shown in Table IV. All four patients who responded to PD-1 inhibitor therapy (CR or PR) showed a positive PD-L1 expression (both TPS and CPS ≥1). The clinicopathological features of these four cases are summarized in Table V. One patient with the SDC histopathological type accompanied by pulmonary metastases (Case 1) showed a high PD-L1 expression with TPS/CPS of 90/95 and achieved complete resolution of all pulmonary metastatic lesions after 7 cycles of pembrolizumab therapy (Figure 3).
Distribution of the PD-L1 expression score associated with the treatment efficacy of PD-1 inhibitor therapy (n=35).
Clinicopathological features of the cases with an objective response to PD-1 inhibitors (n=4).
Computed tomography findings of Case 1. (A) Before initiation of pembrolizumab therapy and (B) after seven cycles of pembrolizumab therapy. The disappearance of all pulmonary metastatic lesions was observed.
irAEs. Eight of 36 patients (22.2%) showed a total of 10 irAEs, including 4 endocrine disorders (hypothyroidism in 3, adrenal insufficiency in 1), 2 hepatobiliary disorders, 2 gastrointestinal disorders (enterocolitis), 1 skin disorder, and 1 case of interstitial pneumonia. No new safety signals were identified in this study. Although four of the eight patients who experienced irAEs were unable to continue treatment with PD-1 inhibitors due to adverse events, none of the patients died from treatment-related adverse events.
Discussion
Previous reports on the efficacy of PD-1 inhibitors in the treatment of patients with R/M SGC have reported ORRs ranging from 4.2% to 13.6%, although the number of included patients and histopathological subtypes vary among reports (7-9). The ORR of PD-1 inhibitors for patients with R/M SGC in this study was 11.1%, which was consistent with the results in previous reports. In terms of tumor histopathology, all patients who responded to PD-1 inhibitors in this study were of the non-ACC type (mostly SDC), suggesting that non-ACC patients, especially those with SDC, may benefit from PD-1 inhibitor therapy. In contrast, no objective therapeutic effect of PD-1 inhibitors was observed in patients with ACC in this study, and previous reports have similarly failed to demonstrate a therapeutic benefit in ACC patients, suggesting that PD-1 inhibitor therapy is unlikely to be effective in ACC patients (13).
Among SGCs, PD-L1 expression is reportedly high in high-grade histopathological types, including SDC, while it is extremely low in ACC cases (10, 11, 14-16). It has also been suggested that in SGCs, a high expression of PD-L1 may be associated with an advanced tumor stage and poor clinical prognosis (10, 14, 16, 17). Similarly, in the present study, the PD-L1 expression was high in non-ACC patients, including those with SDC, while no PD-L1 expression was found in ACC patients. Although the relationship between the PD-L1 expression and the efficacy of PD-1 inhibitor therapy in SGCs has not been clarified, all four patients who responded to PD-1 inhibitor therapy showed positive PD-L1 expression in the present study. In particular, 1 SDC patient with a high PD-L1 expression (TPS 90/CPS 95) achieved CR. This suggests that PD-1 inhibitors may be effective in patients with a high PD-L1 expression in SGCs, as with malignant tumors of other organs. In contrast, there were many patients who had a positive PD-L1 expression but did not respond to PD-1 inhibitors in this study. Further case accumulation and validation are needed to conclude on the relationship between PD-L1 expression and the efficacy of PD-1 inhibitors in R/M SGCs.
Marabelle et al. examined the somatic tumor mutational burden (TMB) status in patients with various advanced solid tumors treated with pembrolizumab, including SGCs, and reported that pembrolizumab treatment was more likely to be effective in patients with a high TMB status than in those with a low TMB status (18). In SGCs, it has been reported that SDC cases have much higher TMB levels than other histopathological types, while ACC cases have very low levels of TMB (19, 20). These results also suggest that PD-1 inhibitors are likely to exert a therapeutic benefit in SDC cases. However, the present study did not analyze the TMB status. To identify potential responders to PD-1 inhibitors in R/M SGC patients, we believe it is necessary to determine the TMB status in the future. In contrast, a high MSI, which is generally considered to be involved in the response to immune checkpoint inhibitors, was not identified in any SGC patients in the present study, regardless of the histopathological type, as in previous reports (11, 21, 22). Therefore, it may be of little significance to measure the MSI status when determining the indication of PD-1 inhibitor therapy in R/M SGC patients.
In SDC patients, the expression of therapeutic target molecules, such as human epidermal growth factor receptor 2 (HER2) and androgen receptor (AR), is observed in many cases, and relatively high response rates to treatment targeting these molecules have been reported in recent years (ORR 70.2% with trastuzumab plus docetaxel therapy, and ORR 41.7% with combined androgen blockade therapy, respectively) (23, 24). The present results suggest that the therapeutic efficacy of PD-1 inhibitors for R/M SGCs, including SDC, is limited. Given the response rate, it is recommended that targeted therapy based on molecular characteristics such as HER2 and AR be the first-line treatment of choice for patients with locally advanced and/or recurrent or metastatic SDC. PD-1 inhibitor therapy may be a useful option for patients with SDC who have failed to respond to therapy targeting these molecules and have developed further progression, after measuring the PD-L1 expression and TMB status.
Gutschenritter et al. reported that male sex and perineural invasion are poor prognostic factors in salivary gland carcinoma, and that distant metastasis is the predominant form of recurrence (25). Therefore, it is significant to continue to validate the efficacy of PD-1 inhibitors in R/M SGC patients, for whom effective therapies are currently limited. However, identifying potential therapeutic responders to PD-1 inhibitors in SGC patients is challenging due to the rarity and histopathological, biological diversity of the tumors. It has been reported that monitoring the neutrophil-to-lymphocyte ratio (NLR) may be a useful predictor of anticancer efficacy of nivolumab in R/M HNSCC patients (26). In order to identify patients who will benefit from PD-1 inhibitors in R/M SGC, we believe that the utility of various factors, including NLR, as predictors of response needs to be examined in a larger number of patients in the future.
In conclusion, although the efficacy of PD-1 inhibitor therapy in R/M SGC is limited, some patients may respond and achieve long-term disease control. In particular, there is a potential therapeutic effect in SDC patients with positive PD-L1 expression, while the efficacy of PD-1 inhibitor therapy in ACC patients is estimated to be low. For the identification of potential responders to PD-1 inhibitors in SGC patients, the PD-L1 expression and TMB level may be predictors of treatment response.
Acknowledgements
The English usage in this article was reviewed by Japan Medical Communication. Inc. This work was supported by a Grant-in-Aid for Scientific Research from the Japan Society for the Promotion of Science, Tokyo, (JSPS KAKENHI; Grant Number 17K16921; K Hashimoto).
Footnotes
Authors’ Contributions
KH, RY and HY designed the study. RK, TH, TM and HY performed the experiments. KH, RY and HY performed the analysis and wrote the draft of the article. MM, YK, TN, RT and TN reviewed and edited the article. MM, TW, RJ, ST, MY, HU, AT, MT, TT and TY performed collection of cases. All Authors read and approved the final article.
Conflicts of Interest
The Authors have no potential conflicts of interest to disclose.
- Received November 23, 2021.
- Revision received December 11, 2021.
- Accepted December 18, 2021.
- Copyright © 2022 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
