Mohamed H El-Farra, Liang Ji, Jonathan Lee, Jenny Cevallos, Marc Abou Assali, David Row, Sharon Lum, Raja R Narayan, Rhami Khorfan
Differences in patient selection exist for robotic rectal cancer resection based on hospital volume. HVIs demonstrated better oncologic and perioperative outcomes, but there was no difference in long-term overall survival based on volume. There is an opportunity for low- and medium-volume institutions to achieve appropriate long-term oncologic outcomes, but further work is needed to investigate patient selection and early post-operative mortality.
INTRODUCTION: Robotic surgery has been increasingly adopted for rectal cancer. However, the majority of studies investigating robotic surgery have focused on high-volume institutions. This study aims to compare the perioperative and oncologic outcomes of robotic rectal cancer surgery at low-, medium-, and high-volume institutions.
METHODS: Adults undergoing robotic proctectomy for primary rectal cancer were identified from the National Cancer Database (2010-2022). Institutions reporting ≥ 2 years of cases were included. A generalized additive model was used to determine volume thresholds classifying hospitals into low, medium, and high volume based on their rates of conversion to open surgery. Patient characteristics and outcomes were evaluated with chi-squared analysis. Overall survival was assessed with multivariable Cox proportional hazards modeling.
RESULTS: Among 327 institutions, thresholds for low, medium, and high were 1-4, 5-13, and > 13 cases/year. The total number of patients and facilities for low, medium, and high were 10,945/142, 17,361/147, and 8,828/38, respectively. High-volume institutions (HVIs) operated more frequently on cN1-2 tumors (46.0, 50.0, 53.5%, p < 0.0005), but fewer poorly differentiated tumors (10.2, 9.8, 8.4%, p < 0.0003). HVIs had lower rates of conversion to open (7.9, 5.2, 3.2%, p < 0.0001), lower 90-day mortality (2.2, 1.4, 1.3, p < 0.0001), and more frequently resected ≥ 12 lymph nodes (76.8, 80.5, 85.7%, p < 0.0001). On multivariable analysis, there was no difference in long-term overall survival based on institutional volume (medium vs. low: HR 0.97 [0.88-1.07], p = 0.52; high vs. low: HR 0.92 [0.82-1.03], p = 0.13).
CONCLUSIONS: Differences in patient selection exist for robotic rectal cancer resection based on hospital volume. HVIs demonstrated better oncologic and perioperative outcomes, but there was no difference in long-term overall survival based on volume. There is an opportunity for low- and medium-volume institutions to achieve appropriate long-term oncologic outcomes, but further work is needed to investigate patient selection and early post-operative mortality.