Current Adjuvant Therapy Landscape for Resected Stage III and IV Melanoma

Immune checkpoint blockade (ICB)-based immunotherapy (anti-CTLA-4 and anti-PD-1 antibodies) and tyrosine kinase inhibitor-based targeted therapy (BRAF/MEK) have revolutionized the treatment of patients with unresectable stage III or IV melanoma. The resultant improved survival outcomes have provided the rationale to investigate the role of these novel agents in the post-surgical adjuvant setting in patients with resectable high-risk stage III nodal and oligometastatic stage IV disease. The results of three randomized trials demonstrating disease-free survival advantages with the post-resection use of BRAF/MEK combination targeted therapy in stage III patients with BRAF-mutated tumors (Combi-AD1) and single-agent anti-PD-1 immunotherapy in stage III (CheckMate 2382 and Keynote-0543) and stage IV (CheckMate 2384) patients regardless of BRAF mutational status provide the basis for the current US FDA-approved standards of care. A fourth adjuvant trial, ImmuNED5, accrued patients exclusively with resected stage IV disease and used a three-arm randomization schema of placebo, single-agent anti-PD-1, or combination checkpoint blockade (anti-PD-1 plus anti-CTLA-4). A recent updated report showed improved survival outcomes with either of the two immunotherapy regimens compared with placebo and best outcome with the standard high-dose combination ICB regimen.

While these results have been transformational, survival outcomes are still unfavorable, particularly for patients with resected palpable regional lymph node metastases and stage IV disease. Furthermore, it is important to recognize that the realized survival outcomes are likely somewhat inflated as they are not reflective of the true recurrence rate of the entire at-risk population undergoing resection with curative intent. Since the randomizations occurred after resection, some patients, thought to be eligible for adjuvant therapy, were ineligible for trial entry because of early relapse;6 these patients missed the opportunity to potentially benefit from adjuvant therapy. These patients harbor the most biologically aggressive disease and are thought to represent 10–15% of patients with resectable metastases. Improving survival for the entire high-risk patient population, beyond what has been achieved with the current standards, represents an important unmet need. CheckMate 915,7 attempting to build on the success of the CheckMate 238 trial, randomized a similar patient population (resected stage III and IV) to receive single-agent anti-PD-1 or a modified, less toxic, combination ICB regimen. An updated 2022 report demonstrated no outcome benefit with combination ICB over single agent. Another strategy was to explore the role of various regimens or combination therapies, with proven efficacy, in the neoadjuvant setting in patients with measurable, resectable disease, acknowledging there are advantages and disadvantages to both neoadjuvant and adjuvant therapy approaches (Fig. 1).

Fig. 1
figure 1

Neoadjuvant versus adjuvant approaches for immunotherapy in melanoma. Advantages and disadvantages exist for both approaches; however, SWOG S1801 demonstrated improved event-free survival for neoadjuvant-adjuvant immunotherapy over adjuvant immunotherapy

The Rationale for and Early Results from Neoadjuvant Strategies

After the achievement of high clinical response rates and improved survival outcomes with both targeted and ICB therapy in patients with disseminated stage IV and unresectable stage III disease, interest developed in investigating these therapies in the neoadjuvant setting in patients with resectable, measurable nodal metastases so as to (1) determine the extent of clinical and pathological responses; (2) correlate clinical outcomes with extent of response or resistance; (3) prevent early progression of disease by initiating effective systemic therapies earlier to eradicate micrometastatic disease; and (4) establish a translational research platform for biomarker discovery to elucidate mechanisms of response and resistance, since access to tumor is more straightforward in these patients. Single-arm and randomized phase II trials were undertaken and demonstrated high clinical and pathological response rates with combination targeted and ICB regimens (single agent and combination). Significantly improved survival outcomes correlated with complete and major (< 10% viable tumor) pathological responses compared with patients with less responsive or non-responding tumors. The encouraging results of these early neoadjuvant trials and the need to improve upon the outcomes of the phase III adjuvant trials motivated investigators to formalize a multidisciplinary collaborative effort that led to the establishment of the International Neoadjuvant Melanoma Consortium (INMC).8 One of the major contributions of the INMC was a pooled analysis of 184 patients participating collectively in six prospective clinical trials utilizing a variety of targeted or ICB neoadjuvant regimens. This analysis demonstrated that pathological complete response (pCR) occurred in 41% of patients overall and 33% of patients treated with immunotherapy, with an increased pCR rate observed in those receiving combination ICB therapy.9 While pCR was achieved with a higher frequency with targeted therapy and correlated with very favorable survival outcomes, any pathological response less than pCR was associated with a rapid decline in recurrence-free survival (RFS). Remarkably, pCR generated by ICB was associated with nearly 100% RFS, and, in contrast to targeted therapy, even a partial response correlated with high and durable RFS outcomes. In an attempt to compare the impact of neoadjuvant immunotherapy with adjuvant immunotherapy, data from the pooled analysis demonstrated that neoadjuvant immunotherapy was associated with an RFS of 75% at 2 years, which compared favorably with an RFS of 60% at 2 years obtained from adjuvant immunotherapy trials, without accounting for the patients who were not captured in adjuvant trials due to early recurrence. Additional evidence suggesting that neoadjuvant immunotherapy might lead to superior outcomes compared with adjuvant immunotherapy was reported from the OpaCIN trial,10 which studied the use of neoadjuvant versus adjuvant combination ICB, albeit with a relatively small number of patients. Beyond the survival curve analysis demonstrating trends of improved outcomes with neoadjuvant immunotherapy, translational analysis demonstrated a larger number of tumor-resident T cells in the peripheral blood of patients in the neoadjuvant versus adjuvant treatment arm.10 These results are consistent with findings from preclinical models showing improved outcomes with neoadjuvant therapy.9,11 While the aforementioned reports of improved survival outcomes with neoadjuvant immunotherapy in high-risk patients were interesting and even provocative, the relatively few patients in the study and the limitations inherent in these types of analyses categorized the findings as hypothesis-generating. A well-designed, narrowly focused, prospective randomized trial was needed to support substantive changes to the current standard clinical practice of adjuvant single-agent ICB.

SWOG S1801: Design, Results, Impact, and Questions

The SWOG S1801 trial was crafted to evaluate whether pembrolizumab administered both before and after surgery (3 doses neoadjuvantly and 15 doses adjuvantly) compared with pembrolizumab administered after surgery only (adjuvantly), with both treatment arms receiving a total of 18 doses, would increase event-free survival (EFS) in patients with resectable stage IIIB-IVm1C melanoma.12 Importantly, randomization occurred prior to the initiation of either therapy (surgery or pembrolizumab), to ensure that all patients at risk for relapse were included and evenly distributed between the two arms. Power calculations were based on an anticipated EFS of 64% in the adjuvant arm (RFS from Checkmate-238) and 74% in the neoadjuvant-adjuvant arm (from an INMC neoadjuvant systemic therapy pooled analysis). Ultimately, 154 patients were randomized to the neoadjuvant-adjuvant arm and 159 to the adjuvant arm. At 2 years, the EFS was 72% for the neoadjuvant-adjuvant group and 49% in the adjuvant-only group (hazard ratio 0.58, 95% confidence interval 0.39–0.87; p = 0.004). A deeper dive into the results reveals that the reduction in events observed in the neoadjuvant arm (improved EFS) occurred almost completely during the adjuvant phases of the trial (Fig. 2). Given the simplicity in trial design, with only a single perturbation between the two arms (sequence of modalities), one can conclude that administration of pembrolizumab in the context of intact tumor burden primed the immune response to generate more efficacious and tumor-specific T cells, especially the adjuvant phase. In terms of the potential for preventing early relapse, a smaller absolute number of patients experienced an event during the neoadjuvant phase prior to surgery compared with those who relapsed in the surgery arm prior to initiating adjuvant therapy. However, the number of these events overall was surprisingly small. A significant reduction of these early events would not necessarily be expected given the current understanding of the time required for a PD-1-induced anti-tumor response that prevents or thwarts relapse to take place. In contrast, extrapolating from the Combi-AD BRAF/MEK adjuvant trial,1 which demonstrated a significant reduction in early events compared with a placebo control, along with the recognition that targeted therapy has a direct tumoricidal effect, a neoadjuvant targeted therapy trial would more likely be expected to prevent early relapses during neoadjuvant therapy.

Fig. 2
figure 2

Results of the SWOG S1801 trial. a Event-free survival curves for the neoadjuvant versus adjuvant treatment arms. b Timeline of events in each treatment arm. CI confidence interval

The results of the S1801 trial may be readily applied: treatment with neoadjuvant ICB should be offered to patients with high-risk resectable stage III or oligometastatic stage IV melanoma. Decisively, the results from S1801 eliminated concerns about delaying surgery for patients receiving neoadjuvant therapy due to progression that would prevent an attempt at a surgical cure. Additionally, these results will most certainly intensify efforts in the neoadjuvant space to address critically important questions.

  1. (1)

    What is the most appropriate type and duration of neoadjuvant treatment for patients with resectable metastatic melanoma? The INMC provides recommendations to address this question in their white paper.4 Specifically, they recommend a duration of neoadjuvant therapy of 6–8 weeks, given the risk of systemic and/or regional disease progression becoming unresectable. They similarly advise that trials include a post-surgical, adjuvant component until the field has determined how best to employ the individualized pathological response assessment to determine the need for and type of additional post-surgical therapy. Regarding the composition of neoadjuvant therapy, multiple options exist, from targeted therapy (BRAF/MEKi) to anti-PD-1 plus anti-CTLA-4 ICB to intralesional therapy. Notably, neoadjuvant immunotherapy has been shown to induce superior RFS with pCR compared with BRAF/MEK inhibition,9 suggesting that BRAF/MEKi may be a strategy best suited to the adjuvant setting. In patients with unresectable stage III or IV melanoma even with BRAFV600E/K mutation, the DREAMseq trial was stopped early due to the 2-year overall survival (OS) benefit with nivolumab/ipilimumab followed by dabrafenib/trametinib, calling into question whether BRAF/MEKi should be considered in lieu of immunotherapy in the neoadjuvant setting.13 Talimogene laherparepvec (T-VEC), an intralesional oncolytic immunotherapy approved for the treatment of unresectable melanoma, demonstrated improved RFS and OS in stage IIIB-IVM1a patients who received T-VEC followed by surgery versus surgery alone, which notably persisted at the 3-year analyses.14 LAG-3 inhibition has been approved in combination with anti-PD-1 in the unresectable or metastatic setting,15 and has also demonstrated promise in the neoadjuvant setting, with a high pCR rate and excellent safety profile.16

  2. (2)

    How good of a surrogate is pathologic response for long-term outcomes after treatment with neoadjuvant immunotherapy (and targeted therapy)? Preclinical data and small neoadjuvant immunotherapy studies9 suggest that pCR and likely pathologic major response will portend favorable RFS and OS.9,11 Preliminary data from the SWOG S1801 trial reveal a 21% pCR with neoadjuvant pembrolizumab, but we need to wait for final response assessment data to better correlate pCR, partial response, and non-response to outcomes. The above-described combination ICB regimens will likely be further evaluated in an attempt to increase the rates of pCR.

  3. (3)

    Can we limit the extent of surgical resection (and subsequent adjuvant therapy) in patients who have an effective response to neoadjuvant systemic therapy? The PRADO trial attempted to address this question. Ninety-nine patients with stage IIIB-D melanoma were treated with 6 weeks of neoadjuvant combination ICB (flipped dosing schedule of nivolumab/ipilimumab). Subsequent surgery and adjuvant therapy were determined by the pathological response assessment of the resected index lymph node (ILN, largest metastatic lymph node at baseline): major pathologic response (mPR, ≤ 10% viable tumor) had both therapeutic lymph node dissection (TLND) and adjuvant therapy omitted; pathologic partial response (pPR, > 10 to ≤ 50% viable tumor) underwent TLND only; and pathologic non-response (pNR, > 50% viable tumor) underwent TLND, adjuvant systemic therapy (targeted if BRAF mutated or nivolumab if wild-type) ± synchronous radiation. Relapse-free survival and distant metastasis-free survival were as follows: mPR 93 and 98%, pPR 64 and 64%, and pNR 71 and 76%, respectively. Not only did this study suggest that it may be safe to de-escalate treatment in patients with mPR but also that escalation in a non-responding patient could improve RFS.

  4. (4)

    If neoadjuvant immunotherapy is the ‘new normal’, what considerations must surgical oncologists have related to immune checkpoint blockade toxicity? Surgical oncologists must be attuned to subtle symptoms of immune-related adverse events (irAEs) and refer patients for treatment. As Helmink and others outlined, fatigue and poor energy symptoms that may be attributed to metastatic disease may be presenting symptoms of hypophysitis and adrenal insufficiency.17 They propose a thorough history, including bowel habit changes to uncover possible autoimmune colitis, complete skin examination for dermatologic irAEs, and complete blood work to uncover endocrinopathies preoperatively. There are no studies to date that suggest increased wound-healing complications or anastomotic leaks linked to ICB. However, Helmink et al. warn about the surgical implications of high-dose steroids used for the treatment of irAEs preoperatively as a risk factor for postoperative wound complications and stress-induced adrenal insufficiency.

Bottom Line?

The implications of SWOG S1801 on the treatment of resectable stage III and IV melanoma are paradigm-shifting. The translational studies from SWOG S1801 and future neoadjuvant trials will provide critical insight into predicting recurrence, determining prognosis, and personalizing treatment according to the pathologic response.