MAML3-fusions modulate Vascular and Immune Tumor Microenvironment and Confer High Metastatic Risk in Pheochromocytoma and Paraganglioma

Abstract

Background Pheochromocytomas and paragangliomas (PPGLs) are rare neuroendocrine tumors that encompass a genetically heterogeneous disease. Approximately 20-25% of diagnosed cases develop metastases, for which there is an absence of predictive markers and therapeutic stratification strategies. MAML3-fusions in PPGL are associated with increased metastatic risk; however, neither the processes underlying disease progression, nor targetable vulnerabilities have been addressed so far and its prevalence remains unclear. Methods We compiled a total of 779 patients, through the combination of publicly-available and novel data from 10 different series. Omic data, FISH and PD-L1 IHC, were used to identify MAML3-fusion positive PPGLs, which were validated by a custom NGS panel and PCR assays. Differential expression and gene set enrichment analyses were conducted to elucidate distinctive features of MAML3-tumors. CD31 IHC analysis was used to study vascular phenotype, and a classification system was generated according to criteria of homogeneity, number, length and branching of vessels. The immune infiltration of pro-tumor M2 macrophages was examined using lymphocytes infiltration. Results Fusion prevalence stood at 4% (34/779), being the largest MAML3 series reported so far. Patients with MAML3-fusion are mainly single noradrenergic pheochromocytomas, which tend to accumulate secondary events in ATRX. However, we also found two patients with multiple MAML3-related tumors, suggesting a post-zygotic fusion event. MAML3-tumors exhibit a significantly shorter time to metastasis compared to other genotypes, supported by an increased expression of neuroendocrine-to-mesenchymal transition genes and MYC targets. Moreover, these tumors display a unique vascular architecture linked to a characteristic extracellular matrix profile. These tumors present a distinctive immune profile, characterized by PD-L1 and CD40 overexpression, and infiltrating pro-tumor macrophages and NK/cytotoxic cells, making them different from other “immune-cold” metastatic PPGLs. Conclusions Our study highlights the relevance of MAML3-fusions in the context of metastatic PPGLs. We uncovered the presence MAML3-tumor-specific vulnerabilities, such as the Wnt-pathway dysregulation, the rich vascular network, making them susceptible to respond to Wnt-inhibitors and anti-angiogenic therapies. The immune cell infiltration profile suggests that targeting CD40 may be a therapeutic option for these patients, and the clear overexpression of PD-L1 offers an opportunity to reopen clinical trials with MAML3-fusion patients as the ideal candidates.