Introduction Proctectomy can be complicated by pelvic collection adversely affecting recovery; pelvic drains may reduce the incidence and/or severity of pelvic collections. Redo pelvic surgery is more complex due to chronic pelvic infection after anastomotic leak or cancer recurrence. The incidence of pelvic collection following redo proctectomy after neoadjuvant chemoradiation therapy (CRT) is unknown. This study evaluated the incidence of symptomatic pelvic collections in patients undergoing redo proctectomy compared to primary surgery after CRT.

Method After IRB approval, data from patients with rectal cancer undergoing open redo proctectomy following neoadjuvant CRT between 2006–2013 were reviewed. The “redo” (Group A) was matched based on clinical characteristics and surgical strategies [Abdominoperineal resection (APR) vs low anterior resection with diverting loop ileostomy (LAR)] with patients undergoing primary proctectomy (Group B). All patients had pelvic drain at the time of surgery. Incidence of postoperative symptomatic pelvic collections was compared between groups. Clinical relevance of pelvic collection was evaluated by presence of leukocytosis, tachycardia, fever and positive fluid culture at diagnosis.

Results 42 patients in Group A were matched with 42 patients in Group B based on age (58 ± 11 vs 60.9 ± 10 years), Body Mass Index (25.4 ± 5 vs 26.9 ± 4.7 kg/m²) ASA score, Charlson Comorbidity Index (2.8 ± 1.2 vs 2.7 ± 1) and surgical strategy (21 APR/21 LAR Group A vs 18 APR/24 LAR Group B). Postoperative symptomatic collections occurred in 10/42 (24%) patients in Group A vs 4/42 (9%) in Group B (p < 0.05). Collections were treated with percutaneous drain placement. Collections were associated with anastomotic leak in 3/5 vs 2/3 cases (p = ns). Mean interval between surgery and surgical drain removal was 15 ± 12 vs 5 ± 2 days while time from surgery to pelvic collection diagnosis was 15 ± 5 vs 13 ± 8 days in Group A and B, respectively. In Group A, 6/10 patients had a collection diagnosed after surgical drain removal while 4/10 patients developed pelvic collection despite presence of a surgical drain. All patients (4/4) in Group B had drain removal at the time of diagnosis. Tachycardia was the most common symptom indicating pelvic collection formation (p < 0.05). Bacterial growth was noted in 8/10 vs 2/4 cases.

Conclusion Redo pelvic surgery following neoadjuvant CRT is associated with a significantly higher risk of pelvic collection development compared to de novoindex proctectomy, despite drain placement during surgery. Up to 40% of patients required a second percutaneous drain while the initial surgical drain was in place.

Disclosure of interest None Declared.