Follow-up and late complications of bariatric surgery

Quality of available follow-up data

There is a paucity of level 1 data on the long-term durability, complications, and re-operation rates after BS, which makes healthcare planning difficult. At the same time, several RCTs comparing SG with RYGB have now reported 5-year outcomes and two RCTs have even reported 10-year outcomes – one comparing metabolic surgery (RYGB or biliopancreatic division [BPD]) with conventional medical therapy, and another comparing GaB with RYGB.

Although a number of non-randomised studies report on longer term outcomes after BS, the majority are compromised of the poor numbers available for follow-up. This introduces follow-up bias. A recent systematic review examining 10-year outcomes after all bariatric procedures included 57 datasets with only two RCTs – one GaB versus medical therapy and another GaB versus RYGB. The quality of most studies was reported to be ‘low’, with many gaps in the data. Also, there are few data on the long-term cost-effectiveness of surgery. For all these reasons, it is not possible yet to distinguish between those patients for whom surgery should be positively recommended, as opposed to those who are eligible for surgery according to agreed thresholds. More randomised studies with adequate 5-year, 10-year, and even longer follow-ups could answer these questions.

It is further recognised that there is significant variation in reporting methods and the definition of outcomes measured. This poses problems for attempts of systematic synthesis of data in meta-analyses. The bariatric community has responded to these challenges by attempting to standardise reporting of outcomes after BS and develop a core outcome set that should be used in every large RCT. At the same time, to power any individual study adequately to answer all of these potential outcome measures is not an easily surmountable challenge.

Framework for follow-up

Although it is generally agreed that BS patients need lifelong follow-up, there is a complete lack of robust evidence to inform the frequency and nature of the possible interventions. It is further unclear whether the responsibility should lie with the bariatric team or the general practitioner (GP).

Lack of patient education, adherence to lifelong follow-up, and provision of healthcare funding for follow-up are further challenges that need to be overcome. Recent guidelines ^, by national societies on follow-up and monitoring of patients after BS are driven by clinical experience and some scientific evidence.

In the UK, the recommendations by NICE form the framework for follow-up after BS ( Box 19.1 ). NICE lays the responsibility for the follow-up during the first 2 years with the bariatric team and for lifelong annual follow-up thereafter with primary care, as part of a shared care model of chronic disease management. These guidelines are an aspiration, given the current preparedness of primary care for bariatric patient care. Also, since a range of serious issues requiring specialist expertise can develop over the course of the lifetime after surgery, it is important that a workable shared care chronic disease model is developed between BS teams and primary care.

Guidelines for follow-up

The NHS England Obesity Clinical Reference Group commissioned a multi-professional group to undertake the daunting task of developing post-BS follow-up guidelines. The group, which also had patient representation, published its recommendations along with the strength of the evidence. Four different shared clinical models were proposed and examined. The common features included annual review, the ability for a GP to refer back to the specialist centre, and submission of follow-up data to the national database. The group recommended that:

• 1.

  ‘All multidisciplinary bariatric surgery teams should follow-up patients at regular intervals post-surgery and offer a minimum of 2 years follow-up.’

• 2.

  All patients should have ‘routine monitoring of blood tests’ (Grade A evidence). The exact nature and frequency of such monitoring as enumerated in the British Obesity and Metabolic Surgery Society (BOMSS) guidelines is based on clinical judgement rather than more robust scientific evidence.

• 3.

  ‘Gastric band patients should have annual follow-up indefinitely’ (Grade C evidence). This is challenging since most GPs have not been taught to do band adjustments. However, GPs could use the annual visit as a prompt to refer a patient to the surgical team for advice on band adjustment.

• 4.

  ‘Patients should have lifelong monitoring to ensure optimum nutrition is maintained’ (based on NICE recommendation). This recommendation reinforces the need for bariatric teams to work with GPs to make local protocols for which tests are needed and red flags that would alert GPs to refer patients back.

A pragmatic solution

Despite the lack of level 1 evidence, the arbitrary 2-year follow-up led by the BS multidisciplinary team (MDT) of surgeon, physician, dietician, specialist nurse, mental health professional, and others seems appropriate since weight loss in most patients will have reached a plateau by then and therefore be in a ‘steady state’. During the 2 years, joint care with GPs is also appropriate, in preparation for care being mainly under the GPs after this time. Annual follow-up visits with bariatric teams and subsequently with GPs should be used to review existing medications as well as to look for potential signs of long-term nutritional or surgical complications.

Weight loss outcomes

There is robust evidence to suggest that BS results in sustainable weight loss in the long term. A recent systematic review looking at 5-year outcomes from five RCTs concluded that both RYGB and SG result in sustained weight loss. This meta-analysis found a significantly greater percentage of excess weight loss (%EWL) following RYGB compared with SG (65.7% vs 57.3%, P < 0.001).

Diabetes outcomes

There is a wealth of data from high-quality studies showing a durable effect of BS on T2DM. A recent meta-analysis of seven RCTs found that the chance of remission of T2DM was significantly higher after BS compared with medical management after at least 2 years’ follow-up (risk ratio [RR] 10, 95% CI 5.5–17.9, P < 0.001). Another recent systematic review and meta-analysis examining 5-year outcomes reported that the resolution of T2DM was 37.4% and 27.5% after RYGB and SG, respectively. The difference was not significant, probably because the individual RCTs were not powered to evaluate the differences in diabetes resolution rates.

More recently, a smaller RCT compared 10-year outcomes in patients with obesity and T2DM who underwent metabolic surgery (RYGB or BPD) or received conventional medical therapy. The 10-year remission rates for T2DM were significantly higher in the surgical group.

Late relapse of T2DM may be seen in up to a third of patients after initial successful remission. However, this should not be seen as a ‘failure’ of BS, as these patients still benefit from improved glycaemic control with fewer anti-diabetic medications and reduced risk of other comorbidities secondary to T2DM. A recent RCT reported that RYGB was significantly more effective than medical therapy in inducing remission of albuminuria and early-stage chronic kidney disease (CKD) in patients with T2DM and obesity.

Other comorbidity outcomes

BS has been shown to improve multiple obesity-related comorbidities such as hypertension (HTN), obstructive sleep apnoea (OSA), non-alcoholic fatty liver disease (NAFLD), gastro-oesophageal reflux disease (GORD) and Barrett’s oesophagus (with RYGB), dyslipidaemia, and ischaemic heart disease (IHD). A recent RCT showed significant improvement in blood pressure (BP) control and remission of HTN 3 years after RYGB compared to patients only on medical therapy.

Two recent systematic reviews found that BS was effective in reducing nocturnal hypoxaemia in patients with OSA. ^, Fakhry et al. published a systematic review reporting significant improvement in all stages of NAFLD following BS. Adil et al. reported significant regression of Barrett’s oesophagus and improvement in GORD after RYGB.

A recent systematic review of large population-based cohort studies including 269,818 patients who underwent BS and 1,270,086 control patients found that BS significantly reduced all-cause mortality ( P < 0.001) and cardiovascular mortality ( P < 0.001). BS was also associated with a reduced incidence of several comorbidities such as T2DM ( P = 0.010), HTN ( P < 0.001), dyslipidaemia ( P = 0.010), and IHD ( P = 0.001).

Gastro-oesophageal reflux disease

GORD has a strong link with SG, most likely due to the changes in anatomy and physiology. A recent systematic review including 10,718 patients from 46 studies found that 19% of patients experience postoperative GORD, 23% develop de novo GORD, and 4% of all patients require revision to RYGB due to severe acid reflux. Another recent systematic review reported the pooled prevalence of Barrett’s oesophagus after SG to be 11.6% ( P < 0.001), and the risk of oesophagitis increased by 13% every year after SG. At the same time, proximal migration of gastric cardia, where the intestinal metaplasia is commonly seen, after SG can lead to diagnostic confusion with Barrett’s oesophagus. Adequately designed endoscopic studies are, therefore, needed to draw firm conclusions.

Stricture

Patients with stricture or stenosis of the sleeve usually present with intermittent vomiting often associated with difficulty in consuming solid foods. Diagnosis is easily confirmed by a contrast study or endoscopy. A number of management options – endoscopic dilatation, laparoscopic sero-myotomy, or wedge resection – have been described and may be successful. A conversion to RYGB with the formation of a pouch above the stricture is however the best treatment option for these patients and guaranteed to bring relief.

Twist or kink

Patients with SG can develop a number of other mechanical problems with the sleeve, such as a twist or kink, with a functional hold-up. Patients present with persistent regurgitation and/or vomiting. The diagnosis is usually easily established on a contrast study that shows hold-up at the incisura angularis, despite there being a way through on endoscopy. Conversion to RYGB is usually successful.

Internal hernia

Herniation of the small bowel through internal defects created during either GB method is a recognised complication with an incidence that may be cumulative over time. If not diagnosed promptly, it can lead to massive gut infarction with disastrous consequences. Internal hernias usually become symptomatic years after surgery when the patient has lost some weight with the consequent opening of potential internal spaces. Many patients will have recurrent colicky abdominal pain long before the diagnosis is made by laparoscopy.

A recent systematic review of six observational studies ( n = 10,031) and two RCTs ( n = 2609) suggested that closure of mesenteric defects in RYGB may be associated with lower risks of internal herniation and re-operation for small-bowel obstruction (SBO) compared with non-closure of the defects. Although less common as there is only one defect, internal herniae can also happen after OAGB. Until robust evidence develops, the authors recommend closure of Petersen’s space in these patients too.

There is no high-quality evidence to guide the closure technique, but a non-absorbable running suture or clips (in one or two layers) is recommended. The use of glue should be regarded as investigational at present.

Anastomotic stricture

Strictures can develop at both gastrojejunal and jejuno-jejunal anastomoses. Gastrojejunostomy strictures appear to be commoner with the use of circular staplers and lend themselves easily to endoscopic diagnosis and management. Unresponsive ones may need a surgical revision of the anastomosis. Jejuno-jejunostomy strictures, in contrast, are uncommon and may be difficult to diagnose in the earlier stages until the patient presents with a bowel obstruction, which may be closed loop obstruction. Treatment involves refashioning of the anastomosis. A stricture of the gastrojejunal anastomosis with OAGB will need conversion to a Roux-en-Y configuration.