Endo-SPONGE for treating low rectal anastomotic leak

The clinical experts advised the committee that Endo‑SPONGE is a 'niche' technology that is only suitable for treating anastomotic leak in a small selection of people. They explained that several key factors decided how anastomotic leak was treated. These included the anatomy of the anastomosis, the location and accessibility of the leak, and the patient's clinical condition (specifically sepsis severity and their general health status). The clinical experts explained that, in their clinical experience, Endo‑SPONGE would be considered if:

• the anastomotic leak was in the low colorectal area

• the leak cavity was accessible through the anus

• the leak remained localised with no abdomen or peritoneum contamination

• the patient was clinically stable enough to have the procedure.
  
  These anatomical and patient-related factors are likely to inform clinicians' decision making for treating anastomotic leaks in general. But, they do not give any insight about who will benefit most from the procedure. The committee noted that there is no evidence that clearly defines the criteria for patient selection but it was aware of The Association of Coloproctology of Great Britain and Ireland's (ACPGBI) guidance on the prevention, diagnosis and management of colorectal anastomotic leakage, which describes the treatment options. It concluded that it is important to understand which patient population might benefit from Endo‑SPONGE. Collecting real-world evidence from its use in the NHS would help to develop this understanding.

The definition of treatment success after Endo‑SPONGE varied between studies. It was most frequently defined as closure of the leakage cavity to less than 1 cm, or complete granulation and resolution of the cavity. Also, the reported stoma reversal rates varied widely between studies. Experts advised this was an important outcome to measure the clinical effectiveness of Endo-SPONGE relative to other treatments. The committee agreed that there is some evidence that Endo‑SPONGE may improve healing of an anastomotic leakage cavity and increase stoma reversal. However, the evidence is low quality with considerable variation in important clinical endpoints between studies.

The clinical experts advised that Endo‑SPONGE is likely to improve patients' quality of life. This is because it offers the possibility of stoma reversal and restoration of bowel function. However, only 2 studies reported patient outcomes that included patient acceptability (Riss et al. 2009) and functional bowel recovery (Huismann et al. 2019). In the clinical experts' experience, pain and discomfort are the 2 most reported adverse symptoms. Endo‑SPONGE treatment is stopped because of pain in a small number of their patients. The committee concluded that there is uncertainty about the tolerability of Endo‑SPONGE in the wider population. More real-world evidence is needed to understand the effect of Endo‑SPONGE on health-related quality of life and residual bowel function.

The committee concluded that the overall quality of the current evidence is low with a high risk of bias. This is because of the retrospective design of most studies, limited comparators and small sample sizes. The clinical experts explained that the patient groups for whom Endo‑SPONGE might be suitable are small and need to be carefully selected. So, it is unlikely that it would be practical to do a randomised controlled trial. They suggested that using a national database or clinical registry could help evaluate the clinical benefits of Endo‑SPONGE and define the most appropriate patient population. The committee agreed that further research with observational and real-world data would strengthen the evidence.

The clinical experts noted that the rate of anastomotic leak after colorectal surgery in the UK is relatively low (reported to be between 4% and 12%). The clinical experts recognised that there have been improvements in techniques for colorectal surgery, such as stapling and robotics. This could help reduce the incidence of anastomotic leak. However, it remains a serious complication after colorectal surgery in some people. The clinical experts explained that the treatment care pathway for people with anastomotic leak varies across the NHS. It depends on local clinicians' experience and the facilities and resources available. The committee concluded that managing anastomotic leak is made more challenging because there is not a clearly defined care pathway.

The clinical experts advised that specific training is needed for the Endo‑SPONGE procedure but it is easy to learn. The company provides free on-site training. The main challenge of getting clinical experience for this technology is the small number of patients for whom it can be used. A clinical expert explained that, in their organisation, Endo‑SPONGE may only be suitable for about 4 to 5 people per year. Support from the company in the form of training such as simulation training may help to resolve this issue. The committee concluded that training is needed to do the Endo-SPONGE procedure.

The cost modelling done by the external assessment centre (EAC) compared Endo‑SPONGE with percutaneous drainage for treating anastomotic leak. However, the clinical experts advised that this comparison may not be appropriate. They explained that alternative comparators such as the placement of a trans‑rectal or trans-anal drain may also be used for leaks after a low rectal anastomosis. People having these different treatments are likely to have different clinical and anatomical characteristics. The committee concluded from the consultation comments and expert advice that comparators for Endo‑SPONGE in the care pathway may vary depending on patient selection, and percutaneous drainage is likely to be used in a different clinical scenario.

There were 3 clinical scenarios modelled by the EAC. Of these, the clinical experts agreed on a scenario that best reflected current clinical practice. This was the one in which the first assessment and Endo‑SPONGE insertion was done in an operating theatre under general anaesthesia, with subsequent sponge changes done in an outpatient setting under local anaesthesia or light sedation. The clinical experts also added that, in their experience, endoscopy is not necessarily needed to insert Endo‑SPONGE, because of how close the leakage cavities are to the anal verge. The committee noted the EAC's additional cost modelling used other non-surgical comparators. This showed a cost saving of £298 per person over 1 year and £2,230 per person over 10 years. The committee noted that the main cost drivers were reoperation rates and rates of avoiding costs associated with a permanent stoma. However, the studies reported a wide range of values for these important clinical parameters. The committee noted that comparative cost modelling is therefore difficult with the available evidence. It concluded that there are significant uncertainties about the cost consequences of using Endo‑SPONGE. Collecting real-world data would be helpful to inform uncertainties around patient selection, Endo‑SPONGE's place in the care pathway, and clinical and cost outcomes.

The committee concluded that Endo‑SPONGE shows promise for treating anastomotic leak and further studies will help define the clinical and cost benefits. However, doing comparative research is likely to be challenging because of the small number of people with low colorectal anastomotic leak in the NHS each year, and the lack of a clearly defined care pathway. The committee was advised that real-world data, such as from a national registry, would be useful. It could help resolve uncertainties around the optimal use of this technology in clinical practice, including:

• the selection criteria for people who could benefit from Endo‑SPONGE

• the comparative rate of stoma reversal and bowel function recovery using Endo‑SPONGE compared with other treatments

• patient-reported outcome measures such as health-related quality of life

• the cost of Endo‑SPONGE compared with other treatments for anastomotic leak.

NICE commissioned an independent feasibility assessment to consider the potential for further data collection to address the uncertainties in the clinical evidence identified by the committee. The feasibility assessment highlighted that the best approach would be to establish a new national anastomotic leak registry to collect data on all patients with the condition, not just those having Endo‑SPONGE treatment. There are significant cost and resource implications to establish such an NHS-wide register, to collect patient data and produce the required analyses.