Pembrolizumab: the value of PDL1 biomarker in head and neck cancer

KEYWORDS:

• Head and neck cancer
• human papilloma virus
• immunotherapy
• pembrolizumab
• programmed cell death (PD-1)

Head and neck cancer is the fifth most common cancer worldwide and despite advances in first-line and second-line treatments, it has still a poor prognosis due to its high incidence and occurrence in sites not amenable to surgery [1,2]. Once the disease has relapsed or becomes metastatic, patients’ median survival time is about 9 months [3], and there are few treatment options. While therapeutic strategies such as surgery, radiation and chemotherapy when applied alone or in combination are useful for managing the disease [4], in most cases, they fail to obtain a cure. The most commonly used treatments for head and neck cancers are platinum agents, taxanes, 5-fluorouracil, methotrexate and cetuximab.

The immune system plays an important role in the development and progression of head and neck cancers, as the development of such tumors strongly depends on their ability to escape immune surveillance. Head and neck cancer tissue contains inflammatory T cells that highly express PD-L1 receptors in both human papillomavirus (HPV)-positive and HPV-negative tumors. Moreover, the therapeutic strategy of creating an immune checkpoint blockade has recently been shown to enable the effective treatment of a variety of tumors [5]. Increased PD-L1 expression has been demonstrated in cases of actinic cheilitis, a premalignant condition of the oral cavity, as well as in respiratory HPV infections, which can produce lesions that progress to laryngeal cancers [6]. Immunosuppressed patients are more likely to develop head and neck cancers [7] and the tumors that occur in these individuals tend to have a poor response to treatment. Moreover, in some instances, the expression levels of immune checkpoint ligands and their receptors have been associated with the clinical prognosis of head and neck cancer patients [8]; however, the overall data are conflicting.

Recent evidence suggests that oral squamous cell carcinoma patients who exhibit high levels of PD-L1 expression are more likely to have a poor clinical outcome and that PD-L1 expression, as determined by immunohistochemical staining, may serve as an independent prognostic marker for a subset of these patients who are male or have a smoking habit [9].

Pembrolizumab (formerly known as lambrolizumab and MK 3575) is a highly selective, humanized, IgG4 monoclonal antibody against PD-1. Pembrolizumab blocks interactions between the PD-1 receptor and its ligands (PD-L1 and PD-L2) and may, thereby, reactivate immune surveillance and elicit antitumor activity (Table 1) [10]. In September 2014, pembrolizumab became the first drug with an anti-PD-1 mechanism of action to be approved by the US Food and Drug Administration for treatment of melanoma and it is currently approved worldwide as monotherapy for advanced (nonresectable or metastatic) melanoma in adults. Moreover, results of the Keynote clinical trial [11] were used to support the accelerated approval of pembrolizumab in the United States for treatment of patients with metastatic non-small-cell lung cancer and patients whose PD-L1-expressing tumors progress during or after platinum-based chemotherapy [12].

Table 1. Immunotherapy trials with pembrolizumab in head and neck squamous cell carcinoma.

Title	Type of cancer	Phase	Treatment line	Experimental arm	Primary end point	Patients N.
Pembrolizumab as first-line treatment for recurrent or metastatic squamous cell cancer of the H&N – NCT02358031	Recurrent or metastatic squamous H&N cancer	III	I	Pembrolizumab or pembrolizumab plus CIS/carboplatin plus 5FU	PFS	750
Pembrolizumab vs. standard treatment for recurrent or metastatic H&N cancer – NCT02252042	Recurrent or metastatic H&N cancer	III	II	Pembrolizumab or cetuximab, docetaxel or methotrexate	PFS, OS	466
Pembrolizumab in platinum cetuximab refractory H&N cancer – NCT02255097	H&N squamous cell carcinoma	II	Adjuvant	Pembrolizumab	Safety, AEs	150
Pembrolizumab in combination with CRT for LA SCCHN – NCT02586207	SCCHN	Ib	III and IVB	Single arm	Safety	39
Re-irradiation plus pembrolizumab in squamous cell carcinoma of the H&N – NCT02289209	Locoregional inoperable recurrence or the second primary squamous H&N cancer	II	Locoregional recurrence	Pembrolizumab	ORR, safety	48
Cisplatin plus radiation combined with pembrolizumab in resected H&N cancer	H&N squamous cell carcinoma	II	Adjuvant	Pembrolizumab	ORR, safety	80

CRT: chemoradiotherapy; LA: locally advanced; SCCHN: squamous cell cancer of the head and neck; H&N: head and neck; PFS: progression-free survival; OS: overall survival; ORR: overall response rate.

The first data concerning the use of pembrolizumab to treat patients with head and neck cancers were presented by Seiwert et al. [13] at the 2014 American Society of Clinical Oncology meeting. The data from the multicenter, non-randomized Phase Ib head and neck squamous cell carcinoma (HNSCC) Keynote 012 study showed that 78% of patients with recurrent metastatic HNSCC expressed PD-L1 in their tumor stroma, regardless of whether HPV was detected. The majority of those patients had previously been treated for their disease and 35.7% had received more than three lines of therapy. Their treatment was continued for 24 months or until disease progression or a drug-related intolerable toxicity occurred [13].

Tumor shrinkage was observed in 26 (51%) of 51 patients, and the best overall response rate (ORR) (19.6%) was observed when pembrolizuamb was given at a dose of 10 mg/kg every 2 weeks to patients who had either HPV infection-associated or non-HPV-associated tumors. Among those patients, 28.7% achieved stable disease and the disease control rate was 61.5%. Furthermore, the responses were durable, as more than 50% of the patients continued to receive treatment after the data cutoff point of the clinical trial.

Pembrolizumab appears to be well tolerated in patients with HNSCC. To date, a low rate of serious drug-related adverse events has been reported and only 16.7% of grade 3–5 adverse events have been drug related. The most common adverse events are fatigue, pruritus, rash and nausea.

Results obtained from patients enrolled in a Keynote 012 continuation study were presented at the 2015 American Society of Clinical Oncology meeting. This cohort included 132 patients who, regardless of their PD-L1 expression level measured with an IgG4 mAb giving a cutoff 1%, received pembrolizumab at a dose of 200 mg every 3 weeks for 24 months or until disease progression or an intolerable toxicity occurred. Approximately 37.9% of the patients had received ≥3 previous treatments. Tumor shrinkage was reported in 57% of the patients and the overall objective response rate was 24.8% (26.3% among HPV-negative patients and 20.6% among HPV-positive patients). Approximately 25% of the patients achieved stable disease and the disease control rate was about 50%. Moreover, 86% of the patients continued their response to treatment at the data cutoff point of the trial and the median follow-up period was 5.7 months. The continuation study confirmed the acceptable safety profile of pembrolizumab, as only 13% of grade 3 and 4 adverse events were drug related, among which none required treatment. The most common adverse events were fatigue, hypothyroidism, decreased appetite and rash. Results from exploratory post hoc analyses conducted on the two previously described cohorts of recurrent/metastatic HNSCC patients in the Keynote 012 trial were presented at the 2015 European Society of Medical Oncology conference. Among all 173 cases analyzed, the best ORR was 23.7%, and 12.4% of the patients experienced at least one drug-related adverse event. This analysis provided the first data concerning the progression-free survival (PFS) times and overall survival (OS) times achieved when using pembrolizumab to treat HNSCC. Although the median PFS time was only 2.2 months, 26.7% and 18% of patients achieved PFS times of 6 and 12 months, respectively. The median OS time was 9.9 months and 47.6% of the patients were still alive at 12 months. For cisplatin-refractory recurrent/metastatic HNSCC, another anti-PD-1 receptor-blocking antibody is being investigated as a single agent in randomized Phase III trials for platinum-refractory HNSCC. In CheckMate 141 randomized Phase III trial, this immunotherapeutic agent demonstrated significant OS benefit compared to standard chemotherapy. Surprisingly, OS and ORR improvement was greater in patients with PD-L1 expression ≥1% [14].

Another anti-PD-L1 antibody (MEDI4736) achieved additional promising data (about 14% response rate scored using RECIST criteria, with 24% response rate in PD-L1-positive patients) in a Phase I trial, warranting the design of a randomized Phase III trial of MEDI4736 alone or in combination with anti-cytotoxic T-lymphocyte-associated antigen 4 (CTLA-4) and compared with standard-of-care agents [15].

PD-L1 is expressed on most HNSCC tumors and it significantly contributes to their ability to escape control by an immune response [16]. Data obtained from the Keynote 012 trial support the hypothesis that PD-L1 is a biomarker for predicting a tumor’s response to treatment with pembrolizumab. High levels of PD-L1 expression, microsatellite instability mutations and immune gene expression signatures are correlated with higher levels of pembrolizumab efficacy [17]. These findings indicate the importance of developing biomarkers capable of identifying patients suitable for anti-PD-1 and anti-PD-L1 therapy. Such information might be useful when seeking to support and rationalize the use of these new and expensive therapies. The positive correlation between pembrolizumab activity in treating HNSCC tumors and PD-L1 expression was confirmed by data obtained in the Keynote 028 Phase 1b study. In that study, a 22% ORR was reported for a cohort of 27 patients with PD-L1-positive advanced nasopharyngeal carcinoma previously pretreated with standard chemotherapy and the overall disease control rate was 78%. Fifteen patients experienced a decrease in the size of their target lesions and the median response period and PFS time were 10 and 5.6 months, respectively. These interesting data may be explained by the fact that high levels of PD-L1 expression are frequently present in cases of Epstein–Barr virus-associated nasopharyngeal cancer and may correlate with a poor prognosis. Cancer immunology is a rapidly evolving field and we have, only recently, begun to understand the complex interactions occurring between cancerous tumors and the host’s immune system. Because tumors employ several strategies to avoid immune surveillance, the use of combination therapy is crucial in order to overcome this barrier.

The prospects of combining pembrolizumab with more established treatments such as chemotherapy and radiotherapy are intriguing, and such an approach provides a synergistic benefit when treating cancers. Chemotherapy and radiotherapy may produce immunogenic effects that synergize with those produced by tumor immunotherapy; this approach would be similar to the case in which the effects of cetuximab, which works through immunologic mechanisms, synergize with those produced by chemotherapy approaches. This latter treatment strategy is being compared with standard treatment in two ongoing Phase 3 studies in patients with recurrent or metastatic head and neck cancer. Additional work is needed to better identify the characteristics of patients who are most likely to benefit from immunologic therapies and then expand that population to encompass more than a minority of patients. Moreover, while preliminary data support the role of PD-L1 expression as a predictive marker, the importance of other immunologic checkpoints and immune-activating receptors remains unknown. Recently, also PD-L2 expression was correlated to response to immunologic therapy. High expression of PD-L2 was associated to improved benefit from atezolizumab, an IgG1 mAb that prevents the interaction of the ligand PD-L1 with its receptor PD-1; however, additional evidence is required to understand the role of PD-L2 in cancer [18].

There are several other molecular markers that may potentially be used as either prognostic or predictive tools in the treatment of head and neck cancer. Epigenetic alterations, vaccinia-related kinase 1 protein, microsatellite instability, loss of heterozygosity, and HPV represent a small part of a growing number of molecular markers that may potentially be used as either prognostic or predictive tools in the treatment of head and neck cancer. However, use of PD-L1 expression as a predictive marker for a patient’s response to treatment raises several issues. First, if PD-L1 expression is a reliable biomarker, it could be argued that all patients should be tested for PD-L1 expression before starting treatment. However, this information alone is not currently sufficient to decide whether immunotherapy will be beneficial for the tested patient. On the other hand, testing for PD-1 expression may provide valuable information concerning which patients are most likely to benefit from immunotherapy when combined with more established treatments such as chemotherapy and radiotherapy. However, we must consider that the use of PD-L1 biomarker has still dark aspects that may influence its validity. It remains unclear how several factors such as the standardization of the cutoff value, the type of samples (archival or fresh), and the sites (primary or metastatic) influence the response to therapy. On the other hand, new clinical data from CheckMate 141 trial confirm that the expression of biomarkers such as PD-L1 may drive the efficacy of these drugs. It can be hypothesized that a combined blockade of the PD-1 and CTLA-4 pathways should provide a broader and more robust tumor response. The use of immunotherapy combined with radiotherapy or chemotherapy to treat malignant tumors is an active area of investigation, and some preliminary safety data have been reported. Those data showed that concurrent combined therapy produced a better response rate than sequential therapy but was associated with a higher incidence of adverse reactions.

It is possible that all HNSCC patients, but more probable that also nasopharyngeal cancer patients, will ultimately benefit from studies designed to prove that a certain level of PD-L1/PD-1 expression can be used to identify individuals most likely to benefit from certain types of cancer treatment. However, this topic will continue to be an important area of ongoing research, and PD-1 represents a promising biomarker that may be crucial when selecting among therapeutic alternatives.

Declaration of interest

G Iuliano is an employee of Merck Sharp Dohme Oncology. R Addeo is an employee of U.O. Oncology while M Caraglia is an employee of the Second University of Naples. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Additional information

Funding

This manuscript has not been funded.