Sarcopenia prior to and following chemotherapy to predict morbidity in patients undergoing post-chemotherapy retroperitoneal lymphadenectomy (PC-RPLND).

Abstract

381 Background: Patients undergoing post-chemotherapy retroperitoneal lymphadenectomy (PC-RPLND) are subject to variable but substantial risks. Unlike those patients undergoing other extirpative GU surgeries, patients undergoing PC-RPLND are usually younger and possess fewer co-morbidities. Despite this, the burden of metastatic disease and the effects of pre-operative chemotherapy may leave patients poorly conditioned prior to undergoing surgery. Sarcopenia has previously been demonstrated to predict adverse clinical outcomes in a variety of abdominal surgeries. We hypothesized that sarcopenia would be similarly predictive of morbidity and mortality in patients undergoing PC-RPLND. Methods: The records of all patients undergoing post-chemotherapy retroperitoneal lymphadenectomy for the treatment of metastatic germ cell tumors at both a public safety net hospital and an academic tertiary medical center were reviewed. Sarcopenia was assessed by measuring cross sectional area of the psoas muscle at the middle of the third lumbar vertebral body on pre-chemotherapy and preoperative computerized tomography. Psoas Muscle Index (PMI) was calculated by adjusting total psoas area for patient height (cm2/m2). Univariate and multivariate analysis was performed to assess the predictive value of sarcopenia for morbidity and mortality following PC-RPLND. Results: 90 patients underwent PC-RPLND from the year 2006-2019, of whom 81 patients had both pre-chemo and preoperative cross sectional imaging available. Prior to chemotherapy the mean PMI for this cohort was 7.32 cm2/m2, and this did not significantly change following chemotherapy mean PMI = 7.06 cm2/m2 (p = 0.44). Lower PMI both prior to (p = 0.05) and following chemotherapy (p = 0.03) were predictive of a higher risk of post-operative complication. There was a trend towards longer hospital length of stay in patients with more significant sarcopenia, however this was not clinically significant (p = 0.09). Conclusions: Sarcopenia was predictive of morbidity in patients undergoing PC-RPLND. Although sarcopenia did worsen following chemotherapy this was not statistically significant. Further assessment of sarcopenia and pre-operative nutritional status in this population may provide opportunities to reduce morbidity following PC-RPLND.