Summary
Several studies have demonstrated that the short-term weight loss achieved by Roux-en-Y gastric bypass is greater than that achieved by laparoscopic adjustable gastric banding. This notion is supported by Puzziferri et al., who compared the weight loss that these two techniques achieved during the first 2 years after surgery. The real need in this field, however, is for long-term data with > 10 years of follow-up, or in the absence of such data, medium-term data with 3-8 years follow-up. The aim of obesity therapy is to achieve sustainable weight loss, yet the published literature on bariatric surgery is dominated by the presence of short-term data. To convince a skeptical community of the value of bariatric surgery, data that reports the medium-to-long-term outcomes of these approaches must be presented with complete follow-up of all patients, or with sufficient statistical power to allow for those lost to follow-up. Only then can a relevant comparison of various bariatric procedures be conducted.
Commentary
Obesity is a chronic disease that will continue for the lifetime of the vast majority of affected individuals, generate many health problems and reduce quality of life and life span. Bariatric surgery seems to be effective for some patients; however, the weight-loss benefit of surgical approaches must be durable for such techniques to be broadly accepted. Surgical approaches must, therefore, provide substantial weight loss over at least the medium term (i.e. for 3-10 years), but preferably also in the long term (i.e. > 10 years).
Most obese individuals can lose a substantial amount of weight in the short term with nonsurgical methods. An insurmountable challenge for medical weight-loss programs has been their inability to maintain sustained weight loss. Bariatric surgery must overcome this challenge for it to claim a central place in the treatment of obesity.
Puzziferri and colleagues have generated a careful and detailed comparison of early weight loss after Roux-en-Y gastric bypass (RYGB) and laparoscopic adjustable gastric banding (LAGB) in a prospective, observational study of 1,518 obese individuals. Their principal finding was that the weight loss produced by RYGB was more rapid, greater and more consistent than that produced by LAGB during the first 24 months after surgery. Despite the prospective study design and the large number of patients, however, this paper has numerous limitations. Two aspects that attract particular attention are the study's sole focus on early outcomes and the lack of information on patients lost to follow-up.
The authors have not provided data beyond the first 24 months after surgery, despite the fact that they initiated RYGB in 1997 and LAGB in 2001. The authors must, therefore, know the medium-term outcomes for many patients, but have chosen not to share them. At a minimum, 5-year data could be provided, but the authors focus only on the first 2 years. As a consequence, this study has not succeeded in providing new data. The excellent early weight loss achieved by RYGB and the slower, steady, progressive weight loss achieved by LAGB has already been well described by multiple observational studies and confirmed by several systematic reviews. In 2006, the Centre for Obesity Research and Education provided a comprehensive, systematic review of all major bariatric procedures, which included up to 10 years of follow-up data. The findings of this review during the first and second years after surgery were very similar to those of Puzziferri et al., as significantly greater weight loss occurred in patients who underwent RYGB than in those who underwent LAGB. Most importantly, however, by the third year after surgery, the difference in weight loss achieved by the two approaches was not significant, and no difference in weight loss was observed between the approaches in subsequent years.
The basis for the different patterns of weight loss that these techniques elicit is easy to understand. RYGB is not an adjustable technique, and the surgical procedure itself provides the only opportunity to achieve the correct anatomical connections. The weight loss achieved at the end of the first 12 months, therefore, will probably be the maximum attainable. The patient's weight will then remain stable for 1 year before a gentle, but progressive, regain of weight commences.
What, therefore, is the point of focusing on the early outcome of bariatric approaches? A commonly cited, systematic review of the literature on bariatric surgery focused only on the 12-month data. These data were collected at the optimal time point for measuring the maximal effect of RYGB, but at a most unsuitable one at which to measure the effects of LAGB, as the optimal effects of this approach are only just beginning to be reached. What relevance, also, do 12-month or 24-month outcomes have for the management of a chronic disease that has detrimental effects over a patient's lifetime? The medical and general communities need, at a bare minimum, assurance of the favorable outcomes of bariatric approaches over the medium term.
Loss to follow-up severely weakens any bariatric surgery data and has long been a part of the definition of failure for clinical trials. Individuals who have not responded to treatment often do not want to continue with follow-up. This statement is particularly true for studies involving gastric stapling procedures, in which usually nothing further can be done for patients who do not respond favorably. Information on the number of patients lost to follow-up is needed for all clinical studies, because poor outcomes of interventions will be missed if patients who discontinue the study are not included in the analyses.
Puzziferri et al. conclude by stating that a need remains for randomized comparisons of LAGB and RYGB with at least a 5-year follow-up. This point is valid, but it is important to note that any such clinical comparison requires three factors: at a minimum, future trials must have medium-term follow-up that lasts 3-8 years, they must include an optimal aftercare program for patients who have received LAGB, and, by following the principles of intention-to-treat analyses, patients who are lost to follow-up must be counted among the treatment failures to avoid overestimation of the effectiveness of the procedure.