Nature Communications | Immune Mechanisms of Perioperative Pembrolizumab Plus Lenvatinib in Mucosal Melanoma

This study reveals the critical roles of CD8⁺ T cell activation status and TCR clonality in treatment response for mucosal melanoma, providing an actionable biomarker framework for optimizing perioperative immunotherapy combined with anti-angiogenic therapy, with direct implications for individualized treatment design in patients with this refractory malignancy.

 

Literature Overview

The article titled 'A phase II peri-operative study of pembrolizumab plus lenvatinib for mucosal melanoma,' published in Nature Communications, systematically investigates the efficacy and dynamic changes in the immune microenvironment of resectable mucosal melanoma (MM) treated with neoadjuvant pembrolizumab combined with lenvatinib. By integrating pathological assessment with multi-omics analyses, the study reveals significant differences between responders and non-responders in baseline immune profiles, TCR repertoire remodeling, and spatial interactions, offering a high-resolution map to understand the mechanisms of anti-PD-1 combined with VEGFR inhibition in MM.

Background Knowledge

Mucosal melanoma (MM) is a malignant melanoma originating from mucosal tissues, accounting for ~20% of melanoma cases in Asian populations. It is highly aggressive, prone to metastasis, and exhibits low response rates to single-agent anti-PD-1 therapy (ORR ~13–19%), necessitating more effective treatment strategies. Although the combination of VEGFR inhibitors with immune checkpoint inhibitors (ICI) has shown some efficacy in advanced MM, its application in the perioperative setting remains exploratory, particularly lacking in-depth mechanistic insights into treatment response. Key challenges include: How can potential responders be identified? Why do some patients exhibit primary resistance? This study addresses these questions by leveraging a perioperative design, enabling longitudinal comparisons of pre- and post-treatment tumor specimens, and integrating multi-omics technologies—including whole-exome sequencing (WES), spatial transcriptomics, and TCR-seq—to systematically map the trajectory of therapy-induced immune microenvironment remodeling and identify key prognostic and predictive biomarkers.

 

 

Research Methods and Experiments

The study employed a single-arm, phase II clinical trial design (NCT04622566), enrolling 26 patients with resectable MM who received two cycles of neoadjuvant pembrolizumab plus lenvatinib, followed by surgery and continued adjuvant pembrolizumab. The primary endpoint was the pathological complete response (pCR) rate. Paired tumor tissues collected before and after neoadjuvant therapy were analyzed, and a mouse PDX model was established for pharmacodynamic validation. The study systematically performed whole-exome sequencing (WES), digital spatial profiling (DSP), multiplex immunohistochemistry (mIHC), and TCR sequencing (TCR-seq) to dissect molecular and immune features.

Key Conclusions and Perspectives

• Among 21 patients who underwent surgery, the pCR rate was 9.5%, the major pathological response (MPR) rate was 19.0%, and the overall pathological response rate (PRR) reached 38.1%. The median recurrence-free survival (RFS) was 14.8 months, with a 1-year RFS rate of 61.9%. Although the pCR rate did not meet the prespecified target, the combination therapy induced substantial pathological responses, indicating clinical activity in MM.
• At baseline, responders exhibited a 'hot' tumor immune microenvironment (TIME), characterized by significantly higher infiltration of CD8⁺ and CD4⁺ T cells, along with enrichment of IFN-γ signaling and antigen presentation pathways, whereas non-responders displayed a 'cold' phenotype. This suggests that baseline T cell infiltration levels may serve as a potential pre-screening biomarker.
• Post-treatment, non-responders showed increased T cell infiltration, vascular normalization (increased proportion of CD31⁺αSMA⁺ vessels), and upregulated intrinsic tumor antigen presentation and chemokine signaling, indicating that lenvatinib can remodel 'cold' tumors into 'inflamed' phenotypes. This provides a rationale for extending neoadjuvant duration or combining with tumor-infiltrating lymphocyte (TIL) therapy.
• In responders, there was an increase in the proportion of high-abundance TCR clones and greater clonal persistence. Spatial analysis revealed shorter distances between CD4⁺ and CD8⁺ T cells, suggesting enhanced immune synergy. This indicates that effective responses depend on the expansion and maintenance of antigen-specific T cell clones rather than nonspecific expansion.
• Post-treatment scores reflecting activated CD8⁺ and CD4⁺ T cell features were significantly associated with longer RFS and distant metastasis-free survival (DMFS), and these findings were validated in an independent cohort. This establishes T cell activation status as a robust, cross-cohort, cross-therapy prognostic biomarker, supporting its use in guiding the duration of individualized therapy.

Research Significance and Prospects

This study provides clear mechanistic targets for drug development: combination strategies that enhance CD8⁺ T cell infiltration and activation and promote TCR clonal persistence should be prioritized, such as combining with cellular therapies or novel co-stimulatory agonists.

In terms of clinical monitoring, baseline TIME classification and post-treatment T cell activation scores have the potential to serve as dynamic biomarkers for early identification of responders, guiding decisions on surgical timing or adjuvant therapy, thereby advancing precision perioperative management.

Regarding disease modeling, the study emphasizes the need to develop MM models that recapitulate both 'cold' and 'hot' TIMEs, such as humanized PDX models, to reproduce vascular normalization and T cell recruitment in vitro, accelerating drug screening and mechanistic validation.

 

 

Conclusion

This study systematically elucidates the immunological mechanisms of perioperative pembrolizumab plus lenvatinib in mucosal melanoma, highlighting the central roles of baseline immune microenvironment status, T cell activation, and TCR clonal dynamics in determining treatment outcomes. Although the pCR rate did not meet expectations, the relatively high pathological response rate and acceptable safety profile support further exploration of this regimen. The study’s major contribution is the proposal of an 'immune remodeling trajectory' model: in 'cold' tumors, the combination therapy induces vascular normalization and immune activation, partially bridging the immunological gap; in 'hot' tumors, it amplifies pre-existing immune responses and promotes clonal persistence. This finding provides critical guidance for future trial designs—treatment strategies should be tailored based on baseline TIME classification, such as extending neoadjuvant cycles or combining with cellular therapies for 'cold' tumors. Moreover, activated CD8⁺ T cell features, as a robust prognostic biomarker, hold promise for guiding individualized treatment duration. Overall, this work establishes a mechanistic foundation for improving MM therapy and advances the shift from a 'one-size-fits-all' to an 'immune-informed' paradigm in perioperative treatment, marking a significant step toward precision immunotherapy.

 

Lili Mao, Yumei Lai, Hong Zheng, Jie Dai, and Lu Si. A phase II peri-operative study of pembrolizumab plus lenvatinib for mucosal melanoma. Nature Communications.