Does HubBLe spell trouble for HAL?

www.thelancet.com Vol 388 July 23, 2016 311 Tens of thousands of surgical procedures are performed every year in the UK to resolve symptoms of anal bleeding and prolapse caused by haemorrhoids. A steady stream of novel therapeutic approaches may promise more convenient, comfortable, and ultimately eff ective treatment for this common complaint, but innovation is generally associated with increased cost and complexity compared with standard outpatient rubber band ligation (RBL). New surgical devices require systematic evaluation to support development of eff ective patient pathways that make effi cient use of precious health-care resources. Haemorrhoidal artery ligation (HAL) uses Doppler equipped proctoscopy to pinpoint branches of the haemorrhoidal artery that supply blood to haemorrhoids. These feeding vessels are sutured to inhibit further bleeding, while a second stitch fi rmly fi xes the pile within the anal canal to prevent prolapse. Advocates of HAL suggest that it is much more eff ective than standard RBL as a fi rst-line treatment of haemorrhoids, justifying the increased cost and use of general or spinal anaesthesia. To our knowledge, no randomised controlled trials have compared HAL with RBL until now. In The Lancet, Steven Brown and colleagues present HubBLe, a randomised, non-blinded, multicentre, phase 3 study comparing standard RBL versus HAL for fi rst-line or selected second-line treatment of symptomatic haemorrhoids that prolapse and bleed following defecation (ie, grades II and III). This is a clinically relevant patient population that constitutes a large proportion of the surgical workload. Individuals with a history of haemorrhoid surgery or repeated outpatient treatments in the preceding 3 years were excluded. Recurrence of haemorrhoids or haemorrhoidal symptoms at 12 months constituted the primary endpoint. This composite assessment was based on the presence of any one of the following: (1) patients reporting “unchanged or worse” symptoms, rather than “cured or improved”, during a telephone assessment, (2) further therapy being administered, or (3) symptoms suggestive of haemorrhoids being recorded in the medical notes and verifi ed by two investigators blinded to treatment allocation. The study was powered to detect a 50% reduction in haemorrhoid recurrence at 12 months, from 30% with RBL to 15% with HAL. Secondary endpoints assessed at 6 weeks and 12 months included symptom severity score, pain score, surgical complications, further treatment, and health state utility using EQ-5D. Between September, 2012, and May, 2014, 372 patients entered the study, of whom 340 received treatment. Assessment of the primary outcome was possible in 337 patients using 256 patient-reported outcomes, 236 GP forms, and 337 consultant reports. Groups were generally well balanced with marginally more severe haemorrhoids allocated to HAL, and more previously treated patients allocated to RBL. The major fi nding of this study was that haemorrhoid recurrence at 12 months was signifi cantly lower following HAL (48 [30%] of 161 patients) than following RBL (87 [49%] of 176 patients). Actual recurrence rates were much higher than predicted, possibly refl ecting use of a sensitive patient-reported outcome. Further analysis of the composite primary endpoint revealed that this treatment eff ect was accounted for by patients receiving multiple RBL who were classifi ed as having recurred. In our opinion, patients frequently require a course of RBL to achieve optimal symptom control, and the investigators might have inadvertently designated residual haemorrhoids as recurrences. Evaluation of patient-reported dissatisfaction at 12 months revealed no diff erence between RBL and HAL (29%), suggesting that multiple RBL had similar eff ectiveness than did HAL from the patients’ perspectives. Surprisingly, HAL was associated with increased pain and a persistent reduction in EQ-5D compared with RBL, contrary to the widely held belief that it represents a painless approach to management of haemorrhoids. Costs were also higher. The HubBLe trial will undoubtedly change practice within the UK and other health-care systems with fi nite resources. It is unique in directly comparing fi rst-line haemorrhoidal interventions considered of most utility by colorectal surgeons and most acceptable by patients. This study has provided valuable information about the natural history of haemorrhoidal symptoms following treatment, and established a “course” of banding as a real, safe, and more cost-eff ective means of alleviating prolapse and bleeding in the majority of patients. The National Institute for Health and Care Excellence (NICE) will probably wish to update their recommendations on HAL. Does HubBLe spell trouble for HAL?