2018-19 Section 7-Neoplastic and Inflammatory Diseases of the Head and Neck eBook

Head and neck oncology

arms while improved during nivolumab treatment [14].

Combined Proportional Score (CPS, in which PD-L1 was measured on tumor and stromal cells) at least 1%, 26.6% with Tumor Proportional Score (TPS, in which PD-L1 was measured on tumor cells only) at least 50%. Pembrolizumab resulted better tolerated than chemotherapy as nivolumab as well [16]. All grades pembrolizumab-related adverse events were lower than those reported in SOC arm (63.3 vs. 83.8%, respectively). The G3–4 adverse events were 13.4% for pembrolizumab and 36.3% for standard chemotherapy. ATEZOLIZUMAB AND DURVALUMAB Atezolizumab [17] and durvalumab [18], both anti- PD-L1 agents, are the most recent immune check- point inhibitors that have demonstrated activity as single-agent in R/M HNSCC. Atezolizumab was studied in 32 patients obtaining results in line with those from all other mentioned immunotherapies: ORR ¼ 22%, PFS ¼ 2.6 months and median OS ¼ 6.0 months. No data were reported about different results between HPV-positive vs. HPV-negative patients. In the atezolizumab study, PD-L1 expres- sion was also confirmed to be correlated to immune response: ORR was lower in PD-L1 less than 5% (14%) than in PD-L1 more than 5% (24%). It is worthwhile to highlight that the assay used to detect PD-L1 expression in this trial [17] was Ventana SP142, that, when compared with other 3 assays (Dako/Agilent 28–8, Dako/Agilent 22c3, Ventana SP263), showed lower sensitivity rate [19]. The importance of PD-L1 expression is clear in durvalu- mab study [18], where only PD-L1 þ less than 25% HNSCC patients were enrolled. Similar results were met in terms of ORR (16.2%), PFS (2.1 months), median OS (7.1 months), and 1y-OS (33.6%). In durvalumab trial, it was also confirmed the benefit from immunotherapy for HPV þ subgroup (ORR: 26.5% in HPV þ vs. 7.9% in HPV ). Immunotherapy can be also exploited for a loco- regional use in R/M HNSCC, based on the impor- tance of local disease control and its related complications (e.g. bleeding, superimposed infec- tions, airways obstruction, dysphagia, etc.). To date, the most widely studied agent is Talimogene Laher- parepvec (T-VEC), an oncolytic virus, which is being investigated in combination with systemic immu- notherapy, such as pembrolizumab, in R/MHNSCC [20]. Table 1 summarized all other ongoing clinical trials exploring loco-regional immunotherapy- based approaches in the same context. LOCO-REGIONAL IMMUNOTHERAPY-BASED APPROACH

PEMBROLIZUMAB In the literature, the first available results (on July 2016) for immunotherapy in HNSCC were those from KEYNOTE-012 study (60 patients, all PD-L1 > 1%) in which pembrolizumab, an anti- PD1 agent, showed 18% of ORR (25% in HPV þ vs. 14% in HPV ), with a median duration of response of 12 months, and 13 months of OS [15 && ]. Based on that, since August 2016, FDA approved pembrolizu- mab for second-line therapy pending phase III trial (KEYNOTE-040) results. At ESMO 2017, the prelimi- nary results of KEYNOTE-040 were presented [16]. In this study, 495 HNSCC patients, progressed within 6 months from the last platinum, were randomized (1 : 1) to pembrolizumab (200mg every 3 weeks) or investigator’s choice (SOC: Taxane or Methotrexate or cetuximab). Results of pembrolizumab were not significant from the statistical point of view, since median OS (8.4 months), 1-year survival (37.3%, HR 0.81), and ORR (14.6%) were better than what observed in the SOC arm (7.1 months, 27.2%, and 10.1%, respectively). Time to response was longer than needed for nivolumab (4.5 vs. 2.1 months) but response duration was longer (18.4 vs. 12.5 months). Taken together, these data showed very similar results with nivolumab in the same cancer patient setting. The reasons of unmet statistical advantage for immunotherapy (pembrolizumab) in KEYNOTE-040 in respect to what observed in CheckMate-141 (nivolumab) could be multiple and pembrolizumab is currently under evaluation for confirmation of FDA approval. First of all, the percentage of patients continuing pembrolizumab beyond PD was lower than in nivolumab study (11 vs. 42%). Second, as mentioned above, cross-over to immunotherapy in pembrolizumab study was not permitted but 32% of patients enrolled in SOC arm did crossover after progression on chemotherapy, exploiting concomitant FDA approval of nivolu- mab. In fact, at ESMO 2017 [16], significant improvement in terms of OS for pembrolizumab (37 vs. 23.2% of SOC, HR 0.72) was demonstrated, censoring SOC population at the time of first sub- sequent immune checkpoint inhibitor administra- tion. Notably, in a subgroup analysis, only patients treated in USA (vs. European/Russian and Asian Pacific) did not benefit from pembrolizumab, possi- bly because only in North America immunotherapy (nivolumab) was already approved at the time of KEYNOTE-040. ORR was progressively higher according to PD-L1 expression [16]: 14.6% in overall population, 17.3% considering only patients with

www.co-otolaryngology.com

Volume 26 Number 2 April 2018

225