Immuno-oncology in the perioperative setting: the next frontier in resectable head and neck cancers – a systematic review Authors: Akhil Kapoor, Anuj Gupta, Bipinesh Sansar, Bal Krishna Mishra, Nandini Menon, Minit Shah, Vanita Noronha, Kumar Prabhash Published: 2026 ABSTRACT AbstractPerioperative immunotherapy—particularly immune checkpoint inhibitors (ICIs)—has emerged as a promising strategy to reduce recurrence in resectable head and neck squamous cell carcinoma (HNSCC). Despite standard multimodal therapy, nearly half of patients experience relapse within five years. ICIs such as nivolumab and pembrolizumab have transformed recurrent/metastatic HNSCC, and their integration into curative settings is under active investigation.We conducted a PRISMA-compliant systematic review of PubMed, Embase, Cochrane Library, and major oncology conference proceedings (2018–2026) to identify clinical trials evaluating ICIs in the perioperative setting of HNSCC. Two reviewers independently screened and extracted data. Outcomes of interest included safety, pathological response, event-free or disease-free survival (EFS/DFS), and overall survival (OS). Meta-analysis was not performed due to heterogeneous study designs.Twenty-five studies (18 published, 7 conference abstracts) met inclusion criteria. Neoadjuvant ICI monotherapy was safe but showed modest major pathological response (MPR) rates (~ 6%). Combination strategies were more effective: the IMCISION trial reported 35% MPR with nivolumab plus ipilimumab vs. 17% with nivolumab alone. Neoadjuvant therapy did not delay surgery. In the adjuvant setting, the NIVOPOSTOP trial showed improved 3-year DFS with nivolumab plus postoperative chemoradiotherapy (63.1% vs. 52.5%). The KEYNOTE-689 trial showed perioperative pembrolizumab significantly improved EFS (51.8 vs. 30.4 months), leading to FDA approval in PD-L1 CPS ≥ 1 tumors.Pathologic response correlated with improved survival. Ongoing trials are refining biomarkers, combinations, and sequencing. Perioperative immunotherapy is reshaping curative-intent management of HNSCC, establishing a new multidisciplinary treatment paradigm.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1186/s12885-026-16304-8.Keywords: Head and neck squamous cell carcinoma, Perioperative immunotherapy, Checkpoint inhibitors, Neoadjuvant therapy, Nivolumab and pembrolizumab Introduction Head and neck squamous cell carcinoma (HNSCC) is the sixth most common cancer worldwide, with over 700,000 new cases and 350,000 deaths annually [1]. A majority of these cases present as locally advanced disease requiring multimodal therapy (surgery and/or radiation, often with chemotherapy) [2]. Despite aggressive standard treatment, long-term outcomes remain suboptimal – locoregional recurrence or distant metastases occur in up to 50% of patients within 2–3 years of definitive therapy. For high-risk resected HNSCC (e.g. those with extranodal extension or positive margins), the addition of postoperative concurrent chemotherapy (cisplatin) to radiotherapy two decades ago was the last major advance, yielding an absolute improvement in 5-year survival of only ~ 8% [3]. Clearly, new strategies are needed to further reduce recurrence and improve cure rates in HNSCC. Immuno-oncology has transformed the management of recurrent/metastatic HNSCC in recent years. Immune checkpoint inhibitors (ICIs) targeting the PD-1/PD-L1 axis (nivolumab and pembrolizumab) have become standard of care for advanced HNSCC, based on trials showing improved overall survival compared to chemotherapy [4, 5]. Nivolumab prolonged survival in platinum-refractory recurrent HNSCC (CheckMate 141) [4], and pembrolizumab (± chemotherapy) demonstrated superior survival as first-line treatment in PD-L1–expressing metastatic HNSCC (KEYNOTE-048) [5]. These successes raise the question of whether introducing ICIs earlier in the disease course – i.e. around the time of curative surgery – could eradicate micro metastatic disease, induce anti-tumor immune memory, and thereby prevent relapse in high-risk patients. The perioperative setting is an attractive window for immunotherapy. Neoadjuvant (preoperative) immunotherapy may reduce tumor burden and prime an immune response while the full tumor antigen repertoire is present [6]. This could potentially lead to tumor-specific T cell expansion that continues to seek and destroy residual cancer cells after surgery [7]. Furthermore, pathological response to neoadjuvant therapy (e.g. extent of tumor necrosis in the resection specimen) can serve as an early surrogate marker of efficacy, which is predictive of long-term outcomes in other cancers such as melanoma and non-small cell lung cancer. On the other hand, adjuvant (postoperative) immunotherapy can target minimal residual disease and modulate the post-surgical tumor microenvironment to suppress regrowth [8]. Key challenges include timing ICIs relative to surgery (to avoid impeding wound healing or increasing complications) and integrating immunotherapy with standard radiation ± chemotherapy. Early pilot studies over the past 5–7 years have explored these approaches [9], and recently two large randomized trials (KEYNOTE-689 and NIVOPOSTOP) reported positive results, suggesting perioperative immunotherapy may indeed improve outcomes in locally advanced HNSCC [10, 11]. In this systematic review, we critically evaluate the current evidence for immunotherapy in the perioperative setting of HNSCC (focusing on squamous cell carcinoma of oral cavity, pharynx, and larynx). We summarize the results of neoadjuvant, adjuvant, and combined perioperative immunotherapy trials, including recent high-impact conference presentations (ASCO 2023–2025). We also discuss ongoing clinical trials in this arena and consider practical implications, such as patient selection and toxicity management. Our review adheres to PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) 2020 guidelines for transparent reporting [12]. Methods Search strategy and selection criteria We conducted a comprehensive literature search to identify clinical studies evaluating immune checkpoint inhibitors (ICI) administered in the perioperative setting for resectable HNSCC. The search covered PubMed (Medline), Embase, Cochrane Library, and Scopus databases from January 2018 through January 2026. This date range was chosen to capture the modern immunotherapy era (following initial FDA approvals of nivolumab/pembrolizumab in HNSCC) and to include the most recent trials. We also searched major conference proceedings (American Society of Clinical Oncology, European Society for Medical Oncology, American Association for Cancer Research, and American Society for Radiation Oncology annual meetings from 2018 to 2025) for relevant abstracts or presentation slides, given that some pivotal data (e.g. KEYNOTE-689, NIVOPOSTOP) were initially reported at conferences. The search combined terms for head/neck cancer with terms for immunotherapy and perioperative interventions. A sample PubMed query was: “(head and neck OR oral OR oropharynx OR larynx OR hypopharynx) AND (immunotherapy OR PD-1 OR PD-L1 OR checkpoint) AND (neoadjuvant OR preoperative OR induction OR adjuvant OR postoperative)”. Similar search logic was adapted for other databases. Studies not available with English translation were excluded. We excluded studies focused on nasopharyngeal carcinoma or other non-squamous head and neck malignancies, as well as purely definitive (non-surgical) treatment studies (e.g. ICIs with primary chemoradiation) since our focus was on surgical management. All titles/abstracts were screened, and then full texts of potentially relevant studies were retrieved. Two investigators independently assessed eligibility according to pre-defined criteria: (1) patients with resectable HNSCC (any T/N stage that underwent surgery with curative intent), (2) intervention included an FDA-approved ICI (e.g. pembrolizumab, nivolumab, durvalumab, etc.) given either before surgery (neoadjuvant), after surgery (adjuvant, either concurrently with or after radiation), or both, and (3) reported outcomes on feasibility, safety, pathological response, or survival. Both prospective trials and retrospective series were included if they met these criteria, but single-case reports were excluded. Disagreements on inclusion were resolved by consensus. Figure 1 illustrates the study selection process in a PRISMA flow diagram. Study selection procedure Data extraction and quality assessment Data were extracted from each included study on: study design and phase, sample size, patient population and risk factors, immunotherapy regimen (agent, dose, timing relative to surgery and standard therapy), primary endpoints (pathologic response rates, survival outcomes), and key results. For conference abstracts lacking full manuscripts, we relied on reported summary data (hazard ratios, survival rates, etc.) as available from presentations or press releases. We contacted authors of two conference papers for additional details, receiving one response which provided clarifications on toxicity grading. A formal meta-analysis was not performed due to substantial heterogeneity across included studies with respect to trial phase, patient selection, treatment sequencing, immunotherapy combinations, pathological response definitions, and survival endpoints. Additionally, several pivotal datasets were available only in conference abstract form without complete numerical reporting, limiting statistical pooling reliability. This represents an important limitation of the present review. Instead, we performed a qualitative synthesis. The Cochrane ROBINS-I tool was considered for assessing risk of bias in non-randomized studies, and the Cochrane Risk of Bias 2 tool for randomized trials [13, 14]. However, due to many studies being early-phase (single-arm) or available only as abstracts, we present a narrative assessment of potential biases. For example, we note if a trial was terminated early, or if certain results are from abstract data without peer review (which may limit certainty). The overall quality of evidence is discussed in the context of consistency of findings across studies. This review followed PRISMA 2020 reporting guidelines [12]. A PRISMA checklist is provided (Supplementary Table 1), detailing how each item is addressed in our report. The review protocol was not registered in PROSPERO (as it is a non-meta-analysis systematic review). No funding was obtained for this study, and all authors declare no conflicts of interest influencing this work. Results Our search identified 1,258 records, of which 1,120 were unique after duplicates were removed. After title/abstract screening, 102 records were selected for full-text review. Ultimately, twenty-five studies (18 published, 7 conference abstracts) met inclusion criteria (Fig. 1). Key characteristics and outcomes of major trials are summarized in Table 1. Below, we organize the results by timing of immunotherapy: neoadjuvant, adjuvant, and combination perioperative approaches. We then highlight ongoing trials in Table 1. Table 1Key Trials of Perioperative Immunotherapy in Resectable HNSCCTrial (Phase)Population & DesignPerioperative ICI RegimenPathologic ResponseSurvival OutcomeRefCheckMate 358 (Phase I/II)n = 52, resectable Stage III/IVA HNSCC (HPV + and HPV–); single-arm, neoadjuvant onlyNivolumab 240 mg IV ×2 doses before surgerypCR 0%; MPR ~ 6% (1/17 each HPV + and HPV); no surgical delays2-yr RFS ~ 85% in HPV+; ~70% in HPV– (single-arm)[15]Uppaluri et al. (Phase II)n = 36, resectable HPV– HNSCC (oral cavity mostly); single-arm, neoadj + adj in high-riskPembrolizumab 200 mg IV ×1 pre-op; if high-risk, adjuvant pembro ×6 after CRTpCR 0%; any pathologic tumor response in 44%; 1-yr relapse rate 16.7% in high-risk pts1-yr DFS 83% (high-risk subset); no control arm (historical 65–70%)[9]IMCISION (Phase Ib/II)n = 32, resectable Stage II–IV HNSCC; randomized to neoadj nivo ± ipi (non-comparative)Nivolumab 3 mg/kg IV ×2; +/− ipilimumab 1 mg/kg IV ×1 (combo) before surgeryMPR 35% with nivo + ipi vs. 17% with nivo alone; pCR 0% both arms; no surgery delays2-yr DFS ~ 65% (combo) vs. 57% (nivo) – not powered for survival[16]Camrelizumab+Apatinib(Phase I)n = 20, resectable oral cavity SCC; single-arm, neoadjuvant onlyCamrelizumab 200 mg Q2w ×3 + Apatinib (VEGFR2 TKI) daily ×6 weeks pre-opMPR 40% (8/20); pCR 5% (1/20); all CPS > 10 cases achieved MPR18-mo OS 95%, DFS 89% (single-arm)[17]KEYNOTE-689 (Phase III)n = 714, resectable Stage III–IVA HNSCC (mixed sites, 26% p16+); randomizedPembrolizumab 200 mg IV Q3w ×2 neoadj; then adjuvant pembro Q3w ×15 combined with RT±cisplatinMPR 17.3% vs. 7.4% (Δ + 9.9%); pCR ~ 4% vs. 2%; ORR 23% vs. 20% (ns) neoadj imaging3-yr EFS 59% vs. 52% (HR 0.73, P = 0.004); OS HR 0.83 (ns at interim)[10]NIVOPOSTOP (GORTEC) (Phase III)n = 680, resected LA-HNSCC with high-risk features; randomized adjuvant trialNivolumab 240 mg IV Q2w ×1 starting 2 wks pre-RT, then Q2w during RT, then Q2w ×6 mos post-RT (10 doses)(Neoadj not given; no pathologic response data by design) – Adjuvant only3-yr DFS 63.1% vs. 52.5% (HR ~ 0.76, P = 0.01); 3-yr OS ~ 70% vs. 66% (HR ~ 0.80, not mature)[11]Abbreviations: adj adjuvant, neoadj neoadjuvant, CRT chemoradiotherapy, RT radiotherapy, nivo nivolumab, ipi ipilimumab, ORR objective response rate, ns not significant Key Trials of Perioperative Immunotherapy in Resectable HNSCC pCR 0%; MPR ~ 6% (1/17 each HPV + and HPV) ; no surgical delays 1-yr DFS 83% (high-risk subset) ; no control arm (historical 65–70%) Abbreviations: adj adjuvant, neoadj neoadjuvant, CRT chemoradiotherapy, RT radiotherapy, nivo nivolumab, ipi ipilimumab, ORR objective response rate, ns not significant Neoadjuvant immunotherapy trials (preoperative ICI) Initial phase I–II studies examined single-dose or short-course neoadjuvant PD-1/PD-L1 inhibitors prior to surgery in resectable HNSCC. These studies confirmed that neoadjuvant immunotherapy is feasible and safe, with high rates of on-schedule surgery and no unexpected surgical complications reported [15]. For example, the phase I/II CheckMate 358 trial treated 52 patients (26 HPV-positive, 26 HPV-negative HNSCC) with two doses of neoadjuvant nivolumab monotherapy over 4 weeks before surgery [15]. All but 4 patients proceeded to timely surgery (no surgery delays > 4 weeks attributable to treatment). Grade 3–4 immune-related adverse events (irAEs) occurred in ~ 15% of patients, consistent with known nivolumab toxicity. Although radiologic tumor shrinkage was modest (objective response ~ 10%), this trial provided important pathological data: no patients achieved pathological complete response (pCR), and major pathologic response (MPR) – typically defined as ≤ 10% viable tumor cells remaining – was observed in only 1 of 17 evaluable tumors (5.9%) in each of the HPV + and HPV– cohorts. An additional ~ 15–20% of patients had partial pathologic responses (< 90% but ≥ 50% necrosis). These results indicated that neoadjuvant PD-1 monotherapy can induce immune-mediated tumor cell death in a subset of patients, but monotherapy may be insufficient to induce pCR in most resectable HNSCC, especially HPV-negative disease. Subsequent trials explored intensified neoadjuvant regimens to increase tumor response rates. One strategy was adding a CTLA-4 inhibitor to PD-1 blockade. In the IMCISION trial (phase Ib/II) in the Netherlands, 32 patients with stage II–IV HNSCC were given neoadjuvant nivolumab ± ipilimumab before surgery [16]. Two doses of nivolumab were given 2 weeks apart; the combination arm received a single low dose of ipilimumab added to the first nivolumab dose. All patients underwent surgery on schedule (primary feasibility endpoint met). The addition of ipilimumab significantly boosted tumor regression: MPR was achieved in 35% (8/23) of patients receiving nivolumab + ipilimumab, compared to 17% (1/6) with nivolumab alone [15, 16]. Notably, none of the patients with MPR experienced recurrence during a 2-year follow-up. However, the combination also led to higher immune-related toxicity (38% grade ≥ 3 irAEs vs. 33% with nivolumab alone). This suggests a PD-1/CTLA-4 combo can induce deeper pathological responses in a meaningful fraction of patients, at the cost of moderately increased toxicity. Another approach combined immunotherapy with chemotherapy in the neoadjuvant setting. Rosenberg et al. (abstract, 2020) treated 20 patients with high-risk locally advanced HNSCC with neoadjuvant nivolumab in combination with weekly carboplatin and nab-paclitaxel for 6 weeks [18]. Although detailed results are limited to abstract form, the regimen was reported to be tolerable and showed a pathologic treatment response > 50% in 44% of patients (and > 10% necrosis in 88%) [18]. This suggests chemo-immunotherapy can synergistically increase tumor kill. Similarly, a pilot study in China investigated three cycles of camrelizumab (anti–PD-1) plus apatinib (VEGFR2 inhibitor) before surgery in oral cavity SCC [17]. This combination yielded a 40% MPR rate (8/20 patients) – notably, all 5 patients with PD-L1 CPS > 10 achieved MPR. The regimen was well tolerated without surgical delays. These results underscore the potential of combination strategies (either dual checkpoint blockade or immune + targeted therapy) to improve neoadjuvant efficacy. In summary, neoadjuvant immunotherapy in HNSCC is safe and does not compromise surgical outcomes. Pathologic response rates with ICI monotherapy are modest (MPR on the order of 5–15%, pCR ~ 0%), but are markedly enhanced by combination approaches, reaching ~ 30–40% MPR in small studies. Pathologic responders appear to have excellent short-term outcomes, with many remaining disease-free at 1–2 years. However, it remains unclear if neoadjuvant therapy alone translates into long-term survival benefit, especially since many trials were single-arm. Notably, no neoadjuvant trial to date (as of 2023) showed a significant improvement in survival when ICI was added, likely due to limited sample sizes and short follow-up. This provided rationale for large trials testing perioperative (neoadjuvant + adjuvant) immunotherapy, discussed below. Adjuvant immunotherapy trials (postoperative ICI) Adjuvant immunotherapy trials in HNSCC have aimed to improve outcomes in patients with high-risk pathological features after surgery. Prior to ICIs, multiple attempts (e.g. with EGFR inhibitors or intensification of chemotherapy) failed to improve on standard postoperative chemoradiotherapy for high-risk disease [19]. Two important phase III trials have evaluated ICIs in the postoperative setting: GORTEC 2018-01 (NIVOPOSTOP) This randomized phase III trial enrolled 680 patients with resected stage III/IVA head and neck squamous cell carcinoma (HNSCC) with high-risk pathological features, including extracapsular nodal extension and/or positive surgical margins [11]. All patients received standard postoperative radiotherapy with concurrent high-dose cisplatin (100 mg/m² every 3 weeks for three cycles) as indicated. Patients in the experimental arm additionally received nivolumab initiated two weeks prior to radiotherapy, continued concurrently with chemoradiotherapy, and subsequently administered as maintenance therapy for a total treatment duration of approximately 8 months. The control arm received standard adjuvant chemoradiotherapy alone. The recently published results demonstrated a statistically significant improvement in disease-free survival (DFS) with the addition of nivolumab. The 3-year DFS improved from 52.5% in the control arm to 63.1% in the nivolumab arm, corresponding to an approximately 24% reduction in the risk of recurrence or death (HR 0.76) [11]. Importantly, the DFS benefit appeared consistent across predefined subgroups, including according to PD-L1 status. Treatment compliance remained comparable between both arms, with high radiotherapy completion rates and similar cisplatin dose intensity. Grade ≥ 3 adverse events were largely attributable to concurrent chemoradiotherapy, while immune-related adverse events associated with nivolumab were manageable and consistent with the established safety profile of PD-1 inhibition, without new safety signals. At the time of publication, overall survival (OS) data remained immature, although an early favorable trend toward improved OS was observed with nivolumab. While longer follow-up is required to determine the durability of benefit and impact on OS, NIVOPOSTOP represents the first randomized study in decades to demonstrate improved outcomes beyond standard postoperative chemoradiotherapy in high-risk resected LA-HNSCC. These findings strongly support perioperative/adjuvant immunotherapy as a promising emerging therapeutic strategy, although we await further maturation of survival data and biomarker-driven refinement. ADJUVANT pembrolizumab trial In parallel, Merck’s phase III KEYNOTE-689 (discussed in the next section) included an adjuvant pembrolizumab component. More recently, an EORTC trial (ADHERE, NCT03673735) investigating durvalumab (anti–PD-L1) added to postoperative chemoradiotherapy was also terminated early due to slow accrual and the evolving standard of care (it is listed as withdrawn) [20]. However, a randomized phase II study (ECOG-ACRIN EA3132 presented at ESMO 2022) did compare adjuvant pembrolizumab vs. observation in resected intermediate-risk HNSCC (patients without indications for chemotherapy) [21]. In that trial, adjuvant pembrolizumab did not significantly improve 2-year disease-free survival versus observation in intermediate-risk patients (DFS ~ 78% vs. 74%, P = 0.66), suggesting that not all post-operative settings confer benefit from immunotherapy (particularly if baseline recurrence risk is lower). Overall, the strongest evidence for adjuvant ICI benefit is currently in the high-risk subset as demonstrated by NIVOPOSTOP. In summary, adjuvant immunotherapy (concurrent with or immediately after standard post-operative therapy) is now proven to improve outcomes in high-risk resected HNSCC. The NIVOPOSTOP trial’s positive DFS result is especially noteworthy given that multiple prior targeted agents failed in this setting. These findings strongly support the potential role of adjuvant nivolumab in high-risk resected HNSCC. However, overall survival data remain immature, longer follow-up is needed, and validation across additional studies and biomarker-defined subgroups will be important before universal adoption as a standard-of-care approach. Furthermore, treatment decisions should remain individualized rather than employing a “one-size-fits-all” strategy, particularly given the heterogeneity of PD-L1 expression, HPV status, comorbidity burden, and recurrence risk. It should be noted that the benefit in NIVOPOSTOP was achieved on top of cisplatin-based chemoradiation, indicating immunotherapy can further intensify effective systemic therapy without obvious additive toxicity. Whether certain subsets (e.g. HPV-negative tumors, which comprised most of the trial) benefit more than others will be elucidated with longer follow-up and subgroup analyses. For now, adjuvant nivolumab represents a significant advance for patients with adverse features who historically faced a recurrence rate near 50%. Perioperative combination trials (neoadjuvant plus adjuvant ICI) The theoretical advantages of both neoadjuvant and adjuvant immunotherapy can be harnessed by a combined approach – treating before and after surgery. The landmark KEYNOTE-689 trial embodies this strategy. KEYNOTE-689 (also known as NCT03765918) was a global phase III trial that evaluated perioperative pembrolizumab in resectable locally advanced HNSCC. It included 714 patients with newly diagnosed stage III–IVA HNSCC of the oral cavity, oropharynx (p16-positive or negative), larynx, or hypopharynx, who were candidates for primary surgery. Patients were randomized to receive: Pembrolizumab arm: 2 doses of neoadjuvant pembrolizumab (200 mg Q3W) prior to surgery, followed by standard adjuvant therapy plus pembrolizumab. Standard adjuvant therapy was risk-adapted: patients with high-risk pathology (positive margin and/or extranodal extension) received postoperative cisplatin+radiotherapy (CRT) [10], whereas low-risk patients received radiotherapy alone (or observation if indicated). Pembrolizumab was given concurrently with radiotherapy (and cisplatin if used) and continued every 3 weeks for up to 15 adjuvant doses. Thus, pembrolizumab treatment spanned the perioperative period (~ 1year total). Pembrolizumab arm: 2 doses of neoadjuvant pembrolizumab (200 mg Q3W) prior to surgery, followed by standard adjuvant therapy plus pembrolizumab. Standard adjuvant therapy was risk-adapted: patients with high-risk pathology (positive margin and/or extranodal extension) received postoperative cisplatin+radiotherapy (CRT) [10], whereas low-risk patients received radiotherapy alone (or observation if indicated). Pembrolizumab was given concurrently with radiotherapy (and cisplatin if used) and continued every 3 weeks for up to 15 adjuvant doses. Thus, pembrolizumab treatment spanned the perioperative period (~ 1year total). Control arm: Surgery followed by the same risk-adapted standard adjuvant therapy without pembrolizumab (placebo was not used, this was open-label). Control arm: Surgery followed by the same risk-adapted standard adjuvant therapy without pembrolizumab (placebo was not used, this was open-label). The primary endpoint was event-free survival (EFS), defined as time to progression, recurrence, or death. At the first interim analysis (median follow-up ~ 38 months) the trial showed a significant improvement in EFS with the pembrolizumab regimen. In the intent-to-treat population, 2-year EFS was ~ 71% with perioperative pembrolizumab vs. ~ 53% with standard therapy, and median EFS was 51.8 vs. 30.4 months, respectively. This translated to a hazard ratio of 0.73 (95% CI 0.58–0.92, P = 0.004) favoring pembrolizumab. Notably, the benefit was even more pronounced in patients with PD-L1 expression: for tumors with CPS ≥ 1, HR ~ 0.70, and for CPS ≥ 10, HR ~ 0.66, indicating a ~ 34% risk reduction in the highest PD-L1 group. Conversely, in PD-L1–negative cases (CPS < 1), no EFS benefit was observed (though this subgroup was small). These data suggest that PD-L1 status may modulate the efficacy of perioperative pembrolizumab, in line with its predictive value in the metastatic setting. Pembrolizumab also significantly increased pathologic tumor response rates. The rate of MPR (≤ 10% viable tumor) in the primary tumor was 17.3% with pembrolizumab vs. 7.4% in controls (an absolute difference of + 9.9%, P < 0.0001). Although pathologic complete response (pCR, 0% viable tumor) was rare (a few cases in the pembro arm, none in control), achieving an MPR was associated with markedly better EFS. Nodal pathologic response was also improved: more patients had extensive nodal tumor regression in the pembrolizumab arm. While overall survival data are still immature, a favorable trend was seen in PD-L1–positive patients (e.g. HR 0.72 for CPS ≥ 10). Toxicity in KEYNOTE-689 was manageable: the addition of immunotherapy did not compromise surgery scheduling (over 88% proceeded to surgery in both arms) and did not significantly increase post-op complications (the surgical complication rate was ~ 32% in pembrolizumab vs. 29% in control, not statistically different.Immune-related adverse events occurred at expected frequencies; ∼10% of patients had grade ≥ 3 irAEs with pembrolizumab, mostly endocrine or hepatic, with no new safety signals [10]. In summary, KEYNOTE-689 demonstrates that a neoadjuvant + adjuvant pembrolizumab strategy can significantly improve outcomes over surgery and standard adjuvant therapy alone. This establishes proof-of-concept that perioperative immunotherapy provides additive benefit beyond what either neoadjuvant or adjuvant ICI alone had shown in smaller studies. It is worth noting that KEYNOTE-689 included patients regardless of HPV status and tumor site, and benefit was seen across most subgroups (though was less clear in PD-L1–negative tumors). Based on these results, in June 2025 the FDA approved pembrolizumab for resectable locally advanced HNSCC (stages III–IVA) as perioperative therapy, specifically for tumors expressing PD-L1 CPS ≥ 1. This approval marks a new multi-modality standard for appropriate patients. Table 1 (below) compares the key efficacy outcomes of representative neoadjuvant, adjuvant, and perioperative trials. In brief, the highest pathologic response rates have come from combination neoadjuvant regimens (e.g. PD-1 + CTLA-4 or PD-1 + VEGFR inhibitor, achieving ~ 30–40% MPR) [17] while the largest improvements in survival endpoints have come from trials that integrated immunotherapy into both pre- and post-operative settings (KEYNOTE-689: +22% 3-yr EFS or post-operatively (NIVOPOSTOP: +10.6% 3-yr DFS. These findings collectively herald immunotherapy’s role as the next frontier in curative HNSCC treatment. Ongoing and future trials The success of recent trials has spurred numerous ongoing studies to further refine perioperative immunotherapy in HNSCC (Table 2). Key questions being addressed include: the role of combination immunotherapy (beyond PD-1/PD-L1 monotherapy), the addition of other novel agents, optimal patient selection (by HPV status or PD-L1 expression), and de-intensification of other therapies in responders. Table 2Selected Ongoing Clinical Trials of Perioperative Immunotherapy in HNSCC Trial (Identifier) Regimen Population / Phase Status Ref NCT03700905 (IOR/EORTC) – Nivolumab ± Ipi NeoadjuvantNivolumab ± low-dose ipilimumab before surgery; adjuvant standard therapyResectable Stage III–IV (Phase II)Active (accrual complete)[22]NCT04080804 (RELATIVITY-HN) – Nivo CombosNeoadjuvant nivolumab alone vs + relatlimab (LAG-3 inhibitor) ± ipilimumab (3 arms) → surgery; adjuvant therapy as per SOCResectable LA-HNSCC (Phase II)Active, recruiting[23]NCT03708224 (Atezolizumab Adjuvant)Surgery → postoperative RT ± atezolizumab (1 year) vs RT alone (randomized)LA-HNSCC unfit for cisplatin (Phase III)Completed (Negative)[24]NCT06102395 (EFFECT-NEO) – Pembro + Chemo InductionInduction TPF chemotherapy ± pembrolizumab → surgery → adjuvant SOC (randomized)Resectable oral cavity/oropharynx (Phase III)Recruiting (2025)[25]CTRI/2024/04/065574 (TMC-i study)Neoadjuvant immunotherapy (low dose Nivolumab) + metronomic chemo (TMC) → surgery → adjuvant therapy (experimental arm)undisLocally advanced oral cavity SCC (Phase II) – Tata Memorial Centre (India)Active, recruiting[26]NCT03576417 (GORTEC NIVOPOSTOP) – Nivolumab AdjuvantSurgery → adjuvant cis-RT ± nivolumab (as in NIVOPOSTOP trial)LA-HNSCC high-risk (Phase III) Completed (Positive) [11]Others (e.g., NCT05156978, NCT05633415) Novel ICI or cell therapies in periop setting Various (early phase) Ongoing Trials in italics have reported resultsSOC = standard-of-care, cis-RT = cisplatin chemoradiotherapy, LA-HNSCC = locally advanced HNSCC, TMC = Triple Metronomic Chemotherapy Selected Ongoing Clinical Trials of Perioperative Immunotherapy in HNSCC Trials in italics have reported results SOC = standard-of-care, cis-RT = cisplatin chemoradiotherapy, LA-HNSCC = locally advanced HNSCC, TMC = Triple Metronomic Chemotherapy Several trials are examining ICIs with novel combinations. For example, an ongoing multi-arm phase II trial (NCT04080804) is evaluating neoadjuvant nivolumab in combination with other checkpoint inhibitors like relatlimab (anti–LAG-3) and ipilimumab (anti–CTLA-4) in resectable HNSCC [27]. This trial will provide insight into whether dual or triple checkpoint blockade can further increase cure rates. Another ongoing study platform is evaluating immune-modulating combination approaches alongside neoadjuvant checkpoint inhibition, including combinations incorporating cytokine-mediated immune activation and dual checkpoint blockade strategies. The use of immunotherapy plus induction chemotherapy is being formally tested in the phase III EFFECT-NEO trial (NCT06102395), which began in 2024. EFFECT-NEO randomizes patients to standard induction chemotherapy (e.g. TPF – docetaxel, cisplatin, 5-FU) ± pembrolizumab, followed by surgery and risk-adjusted adjuvant therapy [25] .This will clarify if adding ICIs to induction chemo improves pathologic response and survival. Another frontier is de-intensification for responders. The ongoing phase II trial HN003 (NRG-HN003) is examining whether patients who achieve pCR or MPR to neoadjuvant pembrolizumab can safely omit postoperative radiotherapy, thus personalizing therapy intensity (NCT03021993) [27]. Similarly, trials are looking at using pathological response to guide the need for adjuvant treatment (for example, the planned PRIME trial will give neoadjuvant pembrolizumab and then omit cisplatin in adjuvant RT for those with good responses). Importantly, the field is also focusing on special populations. HPV-positive oropharyngeal cancer, which has a better prognosis, is being studied to see if immunotherapy could allow avoidance of chemotherapy or radiation. The ongoing NCT03708224 trial is adding atezolizumab (anti–PD-L1) to surgery and postoperative radiation (without chemo) in HPV-negative HNSCC to see if immunotherapy can substitute for cisplatin in cisplatin-ineligible patients [27]. Additionally, multiple investigator-initiated trials in Asia are evaluating perioperative ICIs in oral cavity cancers, given the high incidence in that region. Lastly, combinations of ICIs with other novel agents are under exploration: e.g., anti-TGFβ antibodies, tumor vaccines (like HPV E6/E7 vaccines pre-surgery for HPV+ tumors, such as in NCT04369937), and oncolytic viruses in the neoadjuvant period to enhance immunogenic cell death. These ongoing studies (and others not listed due to space) will further clarify the optimal integration of immunotherapy in curative HNSCC. For instance, the results of RELATIVITY-HN will indicate if dual checkpoint blockade yields sufficiently higher cure rates to warrant the added toxicity, and EFFECT-NEO will tell us if an initial immunochemotherapy approach can downstage tumors and perhaps allow less intensive local therapy. The field is rapidly evolving, with perioperative immunotherapy likely to be a component of future risk-adapted treatment algorithms. Discussion The advent of immuno-oncology in the perioperative management of HNSCC represents a paradigm shift in how we approach locally advanced disease. Our systematic review finds that incorporating ICIs around the time of surgery can produce meaningful improvements in pathological and survival outcomes, without irreversibly harming the curative intent treatment course. The recently reported phase III trials (KEYNOTE-689 and NIVOPOSTOP) provide Level I evidence that perioperative immunotherapy improves event-free and disease-free survival beyond our longstanding standards [10, 11]. This is a remarkable development considering that for over 20 years, no adjuvant treatment had surpassed the cisplatin-chemoradiation backbone. One of the most striking findings is the depth of pathologic response achievable with combination immunotherapy. Major pathologic responses (≥90% tumor regression) were rare with single-agent PD-1 blockade (e.g. 5–10% in CheckMate 358) [15], but rose to ~35–40% in trials adding CTLA-4 or VEGF-R inhibitors [16, 17]. This mirrors experiences in melanoma and lung cancer where dual checkpoint blockade or ICI+antiangiogenic therapy has yielded higher pCR rates than PD-1 monotherapy. The correlation between pathologic response and eventual outcomes is encouraging – in IMCISION, none of the patients with an MPR relapsed in 2 years [16], suggesting pathologic response is a strong surrogate of disease control in HNSCC too. It is worth noting, however, that pCR is still uncommon (even the best regimens rarely exceeded 10–20% pCR). Unlike in cancers such as esophageal or breast cancer where neoadjuvant chemo can produce pCRs in a substantial subset, HNSCC appears to be more resistant to systemic therapy-induced complete sterilization. This underlines that surgery remains critical, and also raises the question of whether additional novel agents (e.g. T cell agonists, oncolytic viruses, etc.) might be needed to further boost pCR rates in the future. From a safety and feasibility perspective, nearly all studies converged on the finding that ICIs do not cause frequent surgical delays or significant wound healing issues. Immune-related toxicities in the perioperative context were generally manageable with standard protocols (steroids, etc.), and most importantly, did not lead to excess perioperative mortality or morbidities in the reported trials. For instance, KEYNOTE-689 reported no increase in postoperative complication rates with pembrolizumab, while NIVOPOSTOP showed that adding immunotherapy did not diminish adherence to radiation or chemotherapy. There were isolated cases of severe irAEs (e.g. one patient in a neoadjuvant trial died of immune-related myocarditis, reported anecdotally), but such occurrences are very rare. Nonetheless, as ICIs move into curative settings, vigilance in toxicity monitoring is paramount. Multidisciplinary care must be coordinated – for example, surgeons and medical oncologists should have plans to promptly evaluate new symptoms post-immunotherapy, and consider delaying surgery if a patient has active high-grade toxicity (e.g. colitis) at the planned surgery time. Our review suggests that with proper scheduling (typically allowing 2–3 weeks after the last neoadjuvant ICI dose before surgery) [15], these treatments can be integrated without compromising surgical outcomes. One critical aspect is patient selection. Additional considerations for it included baseline performance status, autoimmune comorbidities, nutritional status, smoking history, and surgical resectability. Patients with bulky nodal disease, extranodal extension, or high PD-L1 expression may derive greater benefit from intensified perioperative strategies, whereas frail or immunologically “cold” tumours may require alternative approaches or biomarker-guided escalation. The FDA approval of pembrolizumab in the perioperative setting is currently restricted to PD-L1 CPS ≥1 tumours mouse, reflecting the greater EFS benefit observed in that subgroup. PD-L1 expression has been a consistent predictor of ICI benefit in recurrent/metastatic HNSCC; it appears to hold relevance in the perioperative setting as well, though perhaps to a lesser degree (since even the CPS ≥1 population in KEYNOTE-689 had HR 0.70). Interestingly, NIVOPOSTOP found benefit regardless of PD-L1, potentially because the trial was not stratified or powered to detect interaction by PD-L1 (and nivolumab’s effect might be more uniform in the adjuvant context without tumor present). HPV status is another factor – HPV-positive oropharyngeal cancers are very immune-responsive in R/M disease and tend to have tumour-infiltrating lymphocytes; however, they also inherently have better prognosis, so trials like KEYNOTE-689 included both but were not focused on HPV-positive specifically. Subgroup analysis from KEYNOTE-689 hinted that even HPV-positive patients might gain some benefit (though absolute improvement may be smaller due to high control arm outcomes). Future trials and longer follow-up will clarify if certain low-risk groups (like HPV-positive, PD-L1–high) might achieve excellent outcomes with immunotherapy allowing de-escalation of other therapy, or conversely if PD-L1–negative, HPV-negative patients need combination immunotherapies for sufficient benefit. Another topic of debate is optimal timing and duration of perioperative immunotherapy. Despite the encouraging progress, several clinically important questions regarding perioperative immunotherapy in HNSCC remain unresolved. One major uncertainty is the optimal timing and duration of treatment. KEYNOTE-689 used a short neoadjuvant course followed by prolonged adjuvant pembrolizumab for up to one year; however, the relative contribution of neoadjuvant versus adjuvant therapy to the observed survival benefit remains unclear, as direct comparative trials are lacking. Nevertheless, the pathological responses observed with preoperative treatment suggest that neoadjuvant immune priming may play an important role. Neoadjuvant immunotherapy may offer biological advantages by activating tumor-specific T-cell responses while the intact tumor and antigenic repertoire remain present [6–9]. This provides the rationale for combined perioperative approaches, where neoadjuvant immunotherapy initiates immune activation, surgery reduces tumor burden, and adjuvant therapy maintains long-term immune surveillance. However, prolonged adjuvant treatment also raises concerns regarding cumulative toxicity, cost, and overtreatment, particularly in patients achieving major pathological response. Optimal patient selection also remains challenging. Although PD-L1 CPS appears predictive in KEYNOTE-689, reliable biomarkers for identifying immunoresistance are still lacking. Emerging biomarkers such as tumor-infiltrating lymphocytes, circulating tumor DNA, and genomic immune profiling may improve future risk stratification. In addition, standardized response assessment remains an unmet need, as conventional imaging may inadequately distinguish immune-related inflammation from residual viable tumor. Whether pathological response can safely guide de-escalation of surgery, radiotherapy, or chemotherapy also remains investigational. These unanswered questions highlight the need for longer follow-up, biomarker-driven studies, and adaptive clinical trial designs before perioperative immunotherapy can be fully individualized in HNSCC. Our review has several limitations. Some included data were available only in conference abstract form at the time of analysis, carrying a risk of incomplete reporting and future updates. Although findings were cross-verified across available sources, subsequent peer-reviewed publications may provide additional analyses and longer follow-up data. Additionally, the review protocol was not prospectively registered in PROSPERO, which may introduce methodological bias in study selection and reporting. Considerable heterogeneity existed across included studies in terms of patient populations, treatment strategies, and endpoints, limiting direct comparisons and precluding formal meta-analysis. Consequently, findings were synthesized qualitatively, and the absence of pooled quantitative estimates should be acknowledged when interpreting comparative efficacy. Furthermore, several included studies were early-phase single-arm trials, limiting formal risk-of-bias assessment. Finally, although recent phase III trials have shown encouraging results, follow-up remains relatively short and overall survival data are still immature. Longer follow-up and confirmatory studies will be necessary to establish the long-term curative impact of perioperative immunotherapy in HNSCC. One must also acknowledge that not all immunotherapy trials in locally advanced HNSCC have yielded positive results. The differing outcomes between KEYNOTE-689 and KEYNOTE-412 are particularly informative and likely reflect important biological and therapeutic distinctions between perioperative and definitive treatment paradigms. KEYNOTE-412 evaluated pembrolizumab administered concurrently with definitive chemoradiotherapy in unresected disease and failed to meet its pre-specified significance threshold for event-free survival improvement (HR 0.83; p=0.043). In contrast, KEYNOTE-689 incorporated pembrolizumab both before and after surgery and demonstrated a significant improvement in event-free survival. The neoadjuvant exposure in KEYNOTE-689 may have enabled more effective priming of anti-tumor immunity while the intact tumor microenvironment and full antigenic repertoire were still present, thereby enhancing systemic immune activation and long-term immune surveillance [6, 7]. In contrast, definitive chemoradiotherapy alone may induce profound lymphopenia and immune exhaustion, potentially limiting the effectiveness of concurrent immune checkpoint inhibition. Furthermore, the surgical setting itself may contribute to improved immunologic outcomes. Surgical resection reduces tumor burden and may decrease immunosuppressive signaling within the tumor microenvironment, thereby improving the efficacy of subsequent adjuvant immunotherapy. The patient populations also differed substantially between the studies with respect to disease biology, treatment intensity, and mechanisms of locoregional control, reinforcing that HNSCC is a heterogeneous disease in which therapeutic efficacy may vary according to treatment context. Therefore, the contrasting findings between KEYNOTE-689 and KEYNOTE-412 should not be interpreted as conflicting evidence regarding immunotherapy efficacy, but rather as evidence that timing, sequencing, and integration with surgery are critical determinants of benefit in HNSCC. Thus, while perioperative ICIs represent a promising advancement, their implementation should remain thoughtful and individualized. Multidisciplinary tumor boards will need to carefully integrate these evolving strategies into clinical decision-making. For example, a patient with multiple positive lymph nodes and high PD-L1 CPS may derive substantial benefit from perioperative pembrolizumab aimed at achieving major pathological response, whereas patients with low CPS scores, significant comorbidities, or limited expected immunologic benefit may be better served with conventional standard therapy approaches to avoid unnecessary toxicity. However, caution remains warranted. Although KEYNOTE-689 demonstrated a significant improvement in event-free survival, overall survival data remain immature, and longer follow-up together with confirmatory studies will be necessary before definitive conclusions regarding long-term cure benefit and universal adoption can be established. Another area for discussion is the impact on surgical and radiotherapy practice. Neoadjuvant therapy sometimes leads to tumor shrinkage that can simplify surgery (organ preservation). However, it can also induce fibrosis or treatment effect that makes pathological interpretation challenging and surgery potentially trickier (though surgeons in these trials did not report undue difficulty). Pathologists need standardized definitions for treatment response in HNSCC – currently MPR and pCR are used, but given the variability (e.g. presence of peritumoral inflammation, etc.), consistent scoring criteria are important. For radiation oncologists, the question arises: if a patient has a pCR in both primary and nodes after neoadjuvant ICI, could we de-escalate adjuvant radiation? Trials are ongoing, but until proven, standard adjuvant therapy should still be delivered regardless of pathologic response (outside of clinical trials). In terms of trials in progress, it is exciting that many are global and collaborative (e.g. an ongoing Tata Memorial Centre trial in India, TMC-i, is evaluating immunotherapy in oral cancer in a resource-adapted manner). These will ensure that findings are applicable across different healthcare settings. Cost and accessibility of ICIs are considerations, especially in low-middle income countries. If perioperative ICIs become standard, efforts to identify predictive biomarkers (beyond PD-L1) to select patients will be critical to avoid over-treatment. Trials examining circulating tumor DNA (ctDNA) dynamics with neoadjuvant ICI are very interesting – early data suggests that clearance of ctDNA after surgery and adjuvant immunotherapy is associated with much lower recurrence rates. This could potentially guide who truly needs adjuvant immunotherapy or extended treatment. Finally, immunotherapy in HNSCC perioperative setting is still in its infancy. The next frontier could include cellular therapies (TILs or CAR-Ts) given before or after surgery, personalized cancer vaccines post-surgery to consolidate immunologic memory, and modulation of the tumor microenvironment (e.g. using therapies to reduce Tregs or myeloid suppressor cells that are abundant in HNSCC). Combining ICIs with such approaches might further improve cure rates. Unresolved questions and future directions Despite encouraging advances, several clinically relevant questions remain unanswered in perioperative immunotherapy for resectable LA-HNSCC. The optimal duration and sequencing of neoadjuvant and adjuvant immunotherapy remain uncertain, including whether both components are necessary for therapeutic benefit or whether one phase contributes disproportionately to outcomes. Similarly, reliable biomarkers for patient selection are still evolving. Although PD-L1 expression appears prognostically relevant in some studies, its predictive utility remains inconsistent. Additional factors including HPV status, smoking history, tumor immune microenvironment, circulating tumor DNA (ctDNA), and molecular immune signatures may further refine patient selection strategies. The mechanisms and timing of primary and acquired immunoresistance also require further characterization. Another unresolved issue involves optimal response assessment following neoadjuvant therapy, including the role of functional imaging modalities, pathological response evaluation, and radiologic-pathologic concordance. Importantly, whether major pathological responses may eventually permit treatment de-escalation or modulation of surgical extent remains investigational and requires prospective validation. Longer follow-up and biomarker-integrated translational analyses from ongoing phase III trials are essential to address these questions. Emerging evidence suggests that gut and tumor microbiome signatures may serve as predictive biomarkers for immunotherapy response and resistance. Future perioperative strategies in LA-HNSCC may incorporate microbiome profiling for treatment personalization and biomarker-driven patient selection [28]. Limitations This review has several limitations. First, the review protocol was not prospectively registered in PROSPERO. Second, due to substantial heterogeneity among included studies with respect to trial design, patient populations, endpoints, treatment sequencing, and therapeutic regimens, a formal quantitative meta-analysis was not feasible. Third, several pivotal studies included in this review have relatively short follow-up durations, and mature overall survival data remain unavailable for multiple perioperative immunotherapy strategies. Conclusions The introduction of immune checkpoint blockade into the perioperative management of HNSCC represents a promising advance in a disease where progress has historically been limited. Neoadjuvant immunotherapy can induce meaningful tumour regression without compromising surgical outcomes, while adjuvant approaches have demonstrated encouraging improvements in disease-free survival. Perioperative pembrolizumab has now gained regulatory approval in selected patients, and adjuvant nivolumab remains a promising emerging strategy. However, longer follow-up, mature overall survival data, and biomarker-driven patient selection are needed before widespread adoption. Taken together, the current evidence supports cautious optimism that perioperative immunotherapy may play an important role in the management of selected patients with locally advanced HNSCC. At the same time, ongoing studies continue to refine its optimal use. Supplementary Information Supplementary Material 1. Acknowledgements The authors acknowledge the invaluable support of their colleagues in the Head and Neck Disease Management Group for their contributions to this review. Authors' contributions AK conceived the study, designed the review, and drafted the manuscript.AK, AG, BS performed the literature search, data extraction, and quality assessment.AK, NM, MS, VN and KP contributed to interpretation of findings, and manuscript revision.All authors read and approved the final version of the manuscript. Funding This work received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors. No funding was received for the research. Data availability All data generated or analyzed during this review are included in this published article and its supplementary information files. Declarations Ethics approval and consent to participate This study is a systematic review of previously published articles. Ethics approval and consent to participate were not required. Consent for publication Not applicable. Competing interests Akhil Kapoor Research Funding: AstraZeneca (Inst), Novartis (Inst), Lilly (Inst), Biocon (Inst), Bayer (Inst), Johnson & Johnson/Janssen (Inst), BMS (Inst), Pfizer (Inst). Anuj Gupta Research Funding: AstraZeneca (Inst), Lilly (Inst), Biocon (Inst), Bayer (Inst). Bipinesh Sansar Research Funding: AstraZeneca (Inst), Johnson & Johnson/Janssen (Inst), BMS (Inst). Bal Krishna Mishra Research Funding: AstraZeneca (Inst), Pfizer (Inst). Vanita Noronha Research Funding: AstraZeneca (Inst), Merck (Inst), Novartis (Inst), Roche India (Inst), Johnson & Johnson/Janssen (Inst), Nanobiotix (Inst). Travel, Accommodations, Expenses: ASCO Kumar Prabhash Research Funding: Biocon (Inst), Dr Reddy’s Laboratories (Inst), Fresenius Kabi (Inst), Alkem Laboratories (Inst), NATCO Pharma (Inst), BDR Pharmaceutics (Inst), Roche (Inst), Aurigene Oncology Limited (Inst), Johnson and Johnson (Inst). No other potential conflicts of interest were reported. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Material 1. Data Availability Statement All data generated or analyzed during this review are included in this published article and its supplementary information files.