(P 238) INTRA-TUMORAL DC-LAMP EXPRESSION IS ASSOCIATED WITH PROGNOSIS IN SOFT TISSUE SARCOMA

Objective: Soft tissue sarcomas have variable responsiveness to immune checkpoint blockade (ICB). We previously found that a subset of primary undifferentiated pleomorphic sarcoma (UPS) and myxofibrosarcoma (MFS) tumors contained high levels of tumor-infiltrating lymphocytes (TILs) and expressed PD-L1and that better survival was associated with high PD-L1 expression in UPS but not in MFS (Wunder et al, OncoImmunology). Ingenuity pathway analysis identified Th1 pathway activation in UPS, but not in MFS. A group of Th1 genes are known to distinguish tumors with high (HEV) versus low endothelial venules and the presence of HEVs correlates with improved survival in a variety of cancers. HEVs are specialized blood vessels normally found in lymph nodes and their presence in tumors suggests that they may facilitate shuttling of lymphocytes into tumors. HEVs are also found within tumoral tertiary lymphoid structures (TLS) which have been identified in some solid tumors and recently shown to be associated with improved survival in STS. TLS are also implicated in supporting ICB efficacy. LAMP3 is a lysosomal associated membrane protein involved in antigen processing and presentation and is highly expressed in mature dendritic cells. LAMP3 positive dendritic cells (DC-LAMP) have been identified within TLS and immunohistochemical staining of DC-LAMP correlated with the density of tumor HEVs, lymphocyte infiltration and favorable outcomes in some cancers. We previously observed that LAMP3 was overexpressed in PD-L1 high compared to PD-L1 low UPS tumors. Based on these observations, we sought to evaluate the presence and clinical significance of HEVs and DC-LAMP in a subset of primary soft tissue sarcoma tumors.

Methods: After institutional review board approval, cases identified as myxofibrosarcoma (MFS) or undifferentiated pleomorphic sarcoma (UPS) were selected from a retrospective registry of patients who underwent surgical management of an extremity sarcoma. Matched histologic slides/blocks were obtained from the Department of Pathology. 35 UPS and 48 MFS cases were included. LAMP3 positive dendritic cells and high endothelial venules were identified in primary tumor specimens by immunohistochemistry of DC-LAMP and MECA-79, respectively. Tumors were categorized as DC-LAMP Hi or Lo and MECA Hi or Lo based on study population mean expression. Furthermore, DC-LAMP localization (central vs peritumoral) was noted. MECA staining level and DC-LAMP localization and staining intensity were associated with clinical parameters and oncologic outcome. Overall survival was measured from date of surgical resection until last follow up or death.

Results: DC-LAMP staining was detected in 61% of evaluated specimens (51/83). Intra-tumoral localization was characterized as central in 18.8% of MFS and 5.3% of UPS, peritumoral in 10.4% of MFS and 15.8% of UPS, and both central and peritumoral in 35.4% of MFS and 31.6% of UPS tumors. There were no differences in size or grade in tumors with Hi vs Lo DC-LAMP. As seen in the K-M plot in Figure 1A, Hi peri-tumoral DC-LAMP staining tended to be associated with better overall survival in the UPS cohort but not in the MFS group (Figure 1B). In multivariate analysis peri-tumoral DC-LAMP staining was an independent predictor of outcome in UPS (HR 4.19, P-value = 0.030), but not MFS ( p=0.6) (Table 1A). In contrast DC-LAMP staining in the tumor center tended to be associated with overall survival in MFS (p=0.056), but not in UPS (p=0.38) (Table 1B). MECA-79 staining was not associated with clinical characteristics or outcome (Figure 2A-B).

Conclusion: DC-LAMP staining was identifiable in a large proportion of UPS and MFS tumors and had a variable pattern of localization. Both central and peritumoral DC-LAMP localization was detectable in a subset of UPS and MFS tumors. Interestingly, the clinical significance of DC-LAMP localization was distinct in the two histologies, with Hi peritumoral staining being a positive prognostic factor in UPS, and Hi central staining being a poor prognostic factor in MFS. While genetically similar entities, the clinical behavior and intra-tumoral immune landscape is continuing to be characterized as quite distinct between UPS and MFS. Our previous work identified that high PDL1 expression was a positive predictor for overall survival in UPS, but not MFS. Moreover, recent clinical trials evaluating ICB efficacy in soft tissue sarcoma have demonstrated strong responses in UPS, but not MFS. The current study adds additional information to the heterogeneity and clinical significance of the immune landscape in these related sarcoma histologies. TLS have been implicated as important in supporting ICB efficacy. Whether intra-tumoral TLS localization might impact the dichotomous responses to ICB observed in STS will be important area of investigation in future work. Furthermore, MECA staining, a surrogate for HEVs, was present in soft tissue sarcoma but demonstrated no significance association with oncologic outcome. While the mere presence or absence of HEVs alone may not be associated with patient survival, the intra-tumoral cellular constituents and site specific immunologic architecture which HEVs facilitate in establishing within tumors could be an important factor in a STS tumor’s biology and response to treatment. Taken together, our work adds to a growing body of literature examining the functional significance of the heterogenous immunologic architecture of STS tumors.