术后淋巴渗出液作为肌层浸润性膀胱癌的近端液体活检来源

Introduction: Patients with muscle-invasive bladder cancer (MIBC) experience recurrence rates of up to 50% following radical cystectomy 1,2 . Current monitoring involves imaging scans or peripheral blood ctDNA 3,4 , however, there is unmet need for a sensitive molecular test that can identify recurrence early enough to more precisely tailor adjuvant therapy. Our team has pioneered the use of lymphatic exudate collected via surgical drains (“lymph”) as a proximal liquid biopsy fluid. We have previously shown that lymph collected 24 hours after surgery identified molecular residual disease (MRD) in head and neck squamous cell carcinoma (HNSCC) in the immediate post-surgical window 5,6 . Here, we evaluate the feasibility of postoperative lymph to enable sensitive detection of MRD and characterize the dynamics of circulating tumor DNA (ctDNA) in lymph from MIBC patients for 96 hours after surgery. Methods: Lymph, tumor and whole blood were prospectively collected from 8 MIBC patients undergoing radical cystectomy: 4 patients with disease recurrence (REC) and 4 with no evidence of disease (NED) with >1 year of follow-up. Tumor and peripheral blood were collected at surgery; lymph at 24 ± 6 hours after surgical close. For the serial cohort (n=10), lymph was collected every 8 ± 4 hours for 96 hours after surgery. Next-generation sequencing libraries were prepared from genomic and cfDNA using a 500 gene pan-cancer hybrid capture panel; deduplicated coverage was >250X for tumor and blood and >2500X for lymph. Somatic mutations were identified in tumor with matched blood. Tumor-specific variants were directly genotyped in lymph using a custom bioinformatic pipeline. Mutation calls were filtered by a base-specific error model to eliminate artifacts. Results: At 24 hours post-surgery, lymph ctDNA was detected in 4 out of 8 patients (50%). Significantly higher variant allele fraction (VAF) was observed in REC lymph than NED (p = 0.00038). We classified patients as positive (>1) or negative (<1) for detected mutations and performed a KM survival analysis (sensitivity = 75%, specificity = 75%, Hazard Ratio (HR) = 4.3, p = 0.17). To understand the impact of collection timing, serial lymph samples were analyzed. There were no differences across timepoints for either mutation count or mean VAF (repeated measures ANOVA p = 0.33 and 0.91, respectively). Conclusion: We demonstrate the feasibility of analyzing ctDNA in post-surgical lymph in MIBC. ctDNA levels were stable across the first 96 hours after surgery, a pattern consistent with ongoing shed from residual tumor as opposed to decay of DNA from resected tumor debris. While non-significant in this small cohort, the HR and sensitivity in this feasibility pilot are consistent with published larger studies in HNSCC. If these results generalize, postoperative lymph MRD testing has the potential to provide more personalized adjuvant treatment decision-making in MIBC patients.