Melan-A/MART-1 immunity in a EWS-ATF1 translocated clear cell sarcoma patient treated with sunitinib: a case report

A female patient aged 28 years presented in 2007 with a lesion arising from the deep soft tissue of the left foot, covered by a healthy skin. Prior clinical history was negative for melanoma. TNM classification at presentation was: stage 3, e.g. primary lesion arising from soft tissue of the left foot with positivity of the ipsilateral inguinal sentinel LNs. On whole body CT scan there was no evidence of secondary lesions. Initial treatment consisted of wide excision of the primary tumor (surgery 2007) with diagnosis of clear cell sarcoma (CCS), and confirmed by the positivity of the FISH analysis for EWS-ATF1. The patient received sequential hyperthermic limb perfusion with melphalan and TNF and the surgical dissection of the left inguinal-iliac-obturator LNs. Three of the LNs removed were positive for disease. Local recurrence and inguinal LN relapse was detected in July 2011 and treated with doxorubicin plus dacarbazine for 5 cycles with response. Given the evidence of a new disease progression (a loco-regional and inguinal LN relapse, as confirmed by CT scan and PET) and based on preliminary evidence of sunitinib possible activity in CCS [15], in January 2012 sunitinib was started at the dose of 37.5 mg/day, with a tumor partial response (classified according to RECIST, version 1.1 [17]) to the lesion located on left foot and a complete response to metastasis on upper left leg. The response was confirmed by PET and CT scan (Figure 1). In April 2012, patient underwent left leg amputation, with evidence of pathologic response to sunitinib in the surgical specimen. In May 2012, sunitinib was restarted and maintained at the same dosage till January 2013. During these months of treatment, sunitinib was repeatedly stopped due to toxicity, with evidence of rapid disease progression following treatment interruption and of a new response after restoring treatment. From January 2013, sunitinib was reduced to 12.5 mg/day due to grade 3 cardiac toxicity. After initial disease stabilization, disease progression occurred marked by a re-growth of previously responsive tumor lesions and by the evidence of new lesions to the soft tissues of the left leg and pelvic LNs. Sunitinib was definitively interrupted in April 2013. Patient died of disease in February 2014.

Immune-related analysis were performed at the tumor site and in the peripheral blood of this patient (Table 1). The expression of the MITF regulated melanocytic antigens (HMB-45/gp100 and Melan-A/MART-1, Figure 2A) and S-100 (data not shown) was assessed by immunohistochemistry on pre- (surgery December-2010 and November-2011) and post-sunitinib tumor specimens (surgery April-2012). Pre-treatment tumor lesions displayed a clear positivity for all of the analyzed antigens. Conversely, tumor specimen removed after treatment with sunitinib displayed a selective loss of MART-1 expression, while it retained the positivity for HMB-45 and S-100 (Figure 2A). Post-sunitinib tumor was heavily infiltrated by CD3⁺ T cells that contained a significant proportion of CD8⁺ T cells. Areas of pathological regression were clearly evident in association with lymphocyte infiltration (Figure 2B). No T cells infiltration were detected in the pre-treated lesion (data not shown). The in vivo generation of the MART-1 loss antigen variant paralleled the presence of anti-MART-1 systemic immunity in the blood of this CCS patient. Patient’s peripheral blood mononuclear cells (PBMCs) isolated in the course of sunitinib treatment and before surgery (surgery April-2012), sensitized in vitro with the immunogenic HLA-A*0201 restricted peptide Melan-A/MART-1_[27L] displayed the presence of a remarkable frequency of MART-1 specific CD8⁺T cells (7,72%), as monitored by pentamer staining (Figure 3). These anti-MART-1 specific T cells were functionally active. MART-1 sensitized PBMC released IFNγ when stimulated with the target cells loaded with Melan-A/MART-1-epitope (modified and native) and, importantly, they recognized in a MHC restricted fashion HLA-A*0201⁺MART1⁺, but not HLA-A*0201⁺MART1⁻ and HLA-A*0201⁻MART1⁺ tumor cells as evaluated by ELIspot assay (Figure 3). Conversely, no T cells specific for the HLA-A*0201- gp100_[210M] peptide were detected in post-sunitinib PBMCs of the patient applying the same procedure. All together these evidences strongly support the hypothesis that the post-sunitinib MART-1 negative tumor variant is the in vivo outcome of a T cell-mediated immune selection occurring in CCS patient during sunitinib treatment. The anti-MART-1 systemic immunity in post-sunitinib CCS patients was associated with low frequency of circulating immunosuppressive CD14⁺CD11b⁺HLADR^neg/low monocytic myeloid-derived suppressor cells (mMDSCs), a population expanded in cancer patients, including melanoma [18-21]. Multi-parametric flow cytometry indicate that PBMCs collected during sunitinib treatment displayed a frequency of mMDSCs, comparable to that of healthy donors (HD) (Figure 4). Moreover, this low percentage of mMDSCs correlates with functional active convenctional T lymphocytes measured ex vivo as IL-2 and IFN-γ produced by CD3⁺ cells upon TCR stimulation (Figure 4). A strong increase in the number of circulating mMDSC and functionally impaired T cells was detected at the time of disease progression. Conversely, reduced frequency of CD3⁺CD4⁺CD25^hiFoxp3^hi regulatory T cells (Tregs) comparable to that of HD persisted all along the drug treatment (data not shown).