Abstract Background and Aims Minimal change disease (MCD) and focal segmental glomerulosclerosis (FSGS) may occur in patients with solid cancer but this rare association has never been extensively investigated. Here, we report the first series of patients presenting this association, also named paraneoplastic podocytopathy with nephrotic syndrome (PPNS) . Previous study has demonstrated the specific induction in both podocytes and Reed-Sternberg cells of C-Maf Inducing Protein (CMIP) in patients with MCD and Hodgkin lymphoma. Here, we examined whether this molecular relationship was also relevant in patients developing PPNS in the setting of solid cancer. We have investigated the potential CMIP expression in kidney biopsies and in tumor tissues. Method We collected data from patients who presented with histologically proven cancer and concomitantly developed a nephrotic syndrome (± 3 months) related to MCD or FSGS Follow-up data were collected after 3, 6, 12 months and at last visit. We identified a control group of idiopathic nephrotic syndromes (INS) matched on sex, age and histological type, in order to compare the characteristics of these controls with patients with PPNS. Expression of CMIP was studied by immunohistochemistry and by immunofluorescence confocal microscopy on kidney biopsy and tumor samples. We used kidney biopsies from INS and tumors from patients without associated nephrotic syndrome as controls. Results Twenty-one PPNS patients and 21 controls with INS without cancer were identified. The median age of PPNS patients was higher (68 years) than usually observed in INS. Most of them (76%) had a MCD. A broad spectrum of cancers was found (bronchopulmonary, prostatic, colorectal…) . Baseline characteristics were similar between PPNS patients and INS controls, with a median urinary protein creatinine ratio of 865 and 900 mg/mmol, blood albumin t 15 and 15.3 g/L. Estimated glomerular filtration rate were 48 and 60 ml/min/1.73 m² respectively (no significant difference) . At presentation, PPNS patients were more likely to have hematuria (p = 0.031) , a lower median hemoglobin level than controls (p = 0.039) . Moreover, PPNS patients were less often treated with immunosuppressive drugs (p = 0.004) . After 12 months of follow-up, PPNS patients had a less favorable outcome than INS controls, with a mortality rate three times higher (30% vs 9.5%, p = 0.098) and a much lower rate of complete remission of nephrotic syndrome (25% vs 73.7%, p = 0.008, Table 1). CMIP labeling was performed on 13 tumor samples and 16 kidney biopsies in PPNS patients, and on 5 kidney biopsies with INS, and on 5 control tumor samples. Podocyte labeling of CMIP is observed on all but one kidney biopsy. CMIP labeling was positive on all case tumors, but only a scarce staining was detected in control patients with solid cancer without nephrotic syndrome (Fig. 1). In PPNS patients, confocal microscopy analysis on tumoral tissue showed high expression of CMIP in some epithelial cells and in immune infiltrating cells.. In addition, we found that CMIP was specifically express in CD4 T lymphocyte subset (Fig. 2), but not in CD20, FOXP3 or PD1 lymphocytes, defining a novel FOXP3−/PD1− CMIP high CD4+T lymphocyte subpopulation, within immune cell infiltrates. Conclusion We report the first series of patients with PPNS, who have a more severe renal and overall prognosis than controls. We show that patients with PPNS display higher induction of CMIP, both in the tumor and in podocytes. Strikingly, we identified within immune infiltrates a new subset of T-cell harboring a particular phenotype (CD4+, CMIPhigh, FOXP3−, PD1−) , which is not detected in patients with solid tumor without nephrotic syndrome and which could play a key role in the pathophysiology of PPNS.