Background: The immune response plays a critical role in shaping the tumor microenvironment in many cancers. Tertiary lymphoid structures (TLS) are organized aggregates of immune cells that may be sites of antigen presentation within tumors. TLS have been described in several solid tumor types (e.g. melanoma, NSCLC) and we have previously reported their intratumoral presence in well differentiated / dedifferentiated (WD / DD) liposarcoma (Tseng et al., Am J Surg Path 2012; Tseng et al., Sarcoma 2015). This disease is unique in that it frequently has low grade (WD) and high grade (DD) components juxtaposed in the same tumor. We sought to further characterize TLS in WD / DD liposarcoma.
Methods: Clinically-available H&E slides were retrieved for consecutive cases of WD / DD liposarcoma resected at our institution between 2011-2016. Each case was scored for overall TLS density (none, rare, few, some, many, inflammatory) and immunologic maturity (none, aggregate, follicle, germinal center). Demographic, clinicopathologic and outcome data were also collected. Associations between the data were analyzed for statistical significance using a two sided Cochran-Mantel-Haenszel test (for ordered categorical variables) and Spearman correlation coefficients (for continuous variables). Clinical outcome data were represented with Kaplan-Meier plots and analyzed using the logrank test.
Results: In total, 43 cases were studied, 51% of which had evidence of high grade disease (DD). The presence of TLS was noted in 72% of all cases, more commonly in DD than WD tumors (82% vs. 62%, p = 0.19). There was no significant association with TLS density scores and histology (WD versus DD); however higher TLS maturity scores were seen in DD tumors compared to WD tumors (p = 0.012). Tumors found in deep body locations (retroperitoneum, mediastinum) had both higher TLS density (p = 0.032) and maturity (p = 0.027) scores compared to tumors in superficial locations (extremity, trunk). After stratifying by histology and then presence of TLS, tumors with TLS were more likely to recur, but this pattern was not statistically significant (p > 0.25); however in a post hoc analysis comparing DD tumors with TLS to all others, this subset was particularly likely to recur (p = 0.002).
Conclusions: TLS are commonly found in WD / DD liposarcoma and may play a role in the disease biology. TLS appear to be specifically associated with high grade disease (DD) and seem to confer a greater likelihood of recurrence. These unique aggregates of immune cells may have potential utility as a clinical biomarker. Further investigation of the intratumoral immune response in WD / DD liposarcoma is warranted. This disease may also serve as a "cancer model" to study the dynamic interaction of tumor- and immune cells in the microenvironment of other solid tumors.