Bariatric Surgery and Type 2 Diabetes Mellitus: Assessing Factors Leading to Remission. A Systematic Review

Type 2 Diabetes Mellitus (T2DM) is a health problem of paramount proportions and is associated with significant morbidity and mortality. Our study aims to review data published on the effects of different types of bariatric surgeries on T2DM remission, compared to lifestyle and medical intervention (LMI) exclusively, along with a comprehensive finding of numerous preoperative factors that lead to remission. We used PubMed, PubMed Central (PMC), and MEDLINE to search for literature. Our criteria included peer-reviewed, English language articles published in 2010 and onwards, consisting of adults with T2DM and a body mass index (BMI) of >30 kg/m2 as the population of interest. Twenty-four articles with 5,411 patients were selected for this systematic review, which included nine randomized controlled trials (RCTs) and 15 observational studies. The primary endpoint was T2DM remission. Based on the review, bariatric surgery is superior to LMI in inducing remission in T2DM, especially when employing the Roux-en-Y Gastric Bypass (RYGB) technique. Lower age of onset and shorter duration of T2DM, along with a high BMI are some of the factors that can lead to greater remission rates. Further research in RCTs is needed by incorporating double/triple-blind protocols, a standard definition of T2DM remission, long follow-up periods to evaluate for relapses in remission and any side effects, with a focus on inflammatory markers (eg, osteopontin), scoring systems (eg, DiaRem), and benefits of One-Anastomosis Gastric Bypass (OAGB) over other modalities, to advance our understanding of T2DM remission.


Introduction And Background
Obesity is one of the leading global causes of preventable deaths and has rapidly become a health issue of epidemic proportions, impacting people of all ages, from children and adolescents to adults [1,2,3]. The connection between obesity and diabetes has been well researched and documented [4]. In the preceding decade, there has been a significant rise in the incidence of Type 2 Diabetes Mellitus (T2DM), with numbers expected to reach 439 million by the year 2030 [5].
The pathophysiology of T2DM consists of impaired insulin production by the pancreas and the ineffective use of insulin (insulin resistance) in the liver and periphery [6]. Ninety percent of all T2DM patients have excess body weight, which is associated with an increased likelihood of developing complications such as hypertension and cardiovascular disease [4]. The risk of developing T2DM is directly related to an individual's body mass index (BMI). This risk can range from 2%, for BMI 25 to 29.9 kg/m 2 , to 13%, for BMI >35 kg/m 2 [7]. Management of T2DM includes weight loss and pharmacotherapy. However, many oral hypoglycemic agents and insulin therapy cause further weight gain as an adverse effect [8]. The desired goals for the treatment are hemoglobin A1c (HbA1c) <7.0%, low-density lipoprotein (LDL) cholesterol <100 mg/dL, and systolic blood pressure <130 mmHg [9]. Attaining these goals for preventing longterm complications, however, is unsuccessful in as many as 90% of T2DM patients [10].
Bariatric surgery was originally intended for weight reduction only but was added later as a treatment option by the American Diabetes Association (ADA) for the management of obese diabetics in 2009 [11]. By 2011, it was recognized by the International Diabetes Foundation (IDF) within their position statement as a treatment option for T2DM and obesity [12]. The general criteria for bariatric surgery is obese patients with a BMI >40 kg/m 2 or >35 kg/m 2 with obesity-related comorbidities [13]. Various types of bariatric surgical techniques have been studied and are used for weight reduction and treatment of comorbidities. These include Gastric Banding (GB), Roux-en-Y Gastric Bypass (RYGB), Biliopancreatic Diversion (BPD), and Sleeve Gastrectomy (SG) [14].
Numerous studies have shown that bariatric surgery improves glycemic control in T2DM, with many individuals even showing remission [15]. Randomized controlled trials (RCTs) have compared the effectiveness of bariatric surgery against traditional T2DM management, with results indicating a higher likelihood of T2DM remission after surgery, as opposed to lifestyle and medical intervention (LMI) [16]. There is also evidence that bariatric surgery decreases the number of deaths from diabetes, with up to 92% decrease in mortality in patients with T2DM [17].
A problem with data available for bariatric surgery, for resolution of T2DM, is the lack of sufficient long-term follow-up studies for better assessment of remission rates and to check for relapse [18]. Another shortcoming is the insufficient availability of RCTs that are free from inherent bias, and adopt double-or triple-blind protocols. In this systematic review, we aim to gather information on bariatric surgery, and its role in inducing remission in T2DM, while drawing inference on the numerous factors that lead to this remission. These factors include preoperative predictors, comparison of different types of bariatric surgeries, and the difference in follow-up times and how they affect remission rates.

Search Strategy
This systematic review was conducted using the guidelines of 2009 Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) [19]. We performed a comprehensive search of published literature in PubMed, PubMed Central (PMC), and MEDLINE. The keywords used to search the literature within these databases were as follows: "Bariatric Surgery", "Type 2 Diabetes Mellitus", "Remission", and "Diabetes". These keywords were applied separately and/or in conjunction with one another ( Table 1). A total of 349 articles were selected after the removal of duplicates. We screened through each of those articles based on title and then by reviewing the abstracts. Next, we read through the full texts, applied the inclusion and exclusion criteria, and subsequently, we were left with 24 articles that we included in our review ( Figure 1

Inclusion and Exclusion Criteria
We pursued only those studies that were related to our topic of remission in T2DM postbariatric surgery. For this systematic review, we selected only peer-reviewed, English language articles. Our search criteria included literature published in 2010 and onwards. The desired study designs included systematic reviews and meta-analysis, RCTs, and observational studies. The demographic targeted within the researched literature included adults with T2DM who had a BMI >30kg/m 2 . Exclusion criteria applied towards our search included individuals with Type 1 Diabetes Mellitus, pre-diabetes, BMI <30kg/m 2 , pediatric population, and studies with followup of less than a year. In addition, grey literature, case reports, comments, and animal studies were not included.

Results
Upon reviewing the articles for suitability, we included 24 studies in our paper. Nine of these are RCTs and 15 are observational studies. A thorough quality assessment was performed by using the Cochrane Risk of Bias Tool for RCTs and The Newcastle-Ottawa Scale for observational studies. All of the studies satisfied the quality check. Six RCTs and one observational study compared bariatric surgery with medical and lifestyle interventions. Two RCTs and nine observational studies compared different types of bariatric surgeries. One RCT compared RYGB with usual diabetes care and exenatide in patients with T2DM and hypertension. Moreover, three observational studies focused on various pre-operative remission predictors. One observational study researched the role of osteopontin, and one studied the efficacy of the DiaRem score as a predictor of remission of T2DM after bariatric surgery. The primary endpoint in most of the studies was remission in T2DM. The total number of patients for our review is 5,411 ( Table 2).

Discussion
The standard care for T2DM patients has always been lifestyle changes, which include a healthy diet and weight loss, along with regular physical exercise and smoking cessation. If these measures are inadequate in controlling blood glucose levels, pharmacotherapy is introduced, which includes medications like metformin, insulin secretagogues, and eventually, insulin therapy. The problem with insulin therapy and oral hypoglycemic agents is that most of them are known to cause further weight gain, with the exception of dipeptidyl-peptidase-4 (DPP-4) inhibitors and glucagon-like peptide-1 analogues [25]. Bariatric surgery is a modality primarily used for weight reduction in obese patients with comorbidities. Its safety and effectiveness in inducing remission in T2DM should be evaluated.

Bariatric Surgery Versus Lifestyle and Medical Intervention
Numerous studies have shown that bariatric surgery is superior to LMI alone in achieving remission in T2DM [8,16,[21][22][23][24]35]. We conducted reviews of six RCTs, as well as an observational study that evaluate the comparison between LMI and bariatric surgery. The referenced articles ranged from the years 2012 until 2016, and sample sizes starting from 32 to 343. The primary endpoint within each of these studies was remission in T2DM [8,16,[21][22][23][24]35].
The complete remission criteria set by the ADA are fasting plasma glucose (FPG) <100 mg/dL and HbA1c <6%, without the use of any hypoglycemic medication for one year [38]. However, many studies use different explanations of what they define as "T2DM remission". It can be concluded that patients who underwent bariatric surgery achieved T2DM remission in higher numbers compared to patients that were provided just LMI. A significant limitation that we found was that there was no uniform definition of T2DM remission that the studies used.
Having a consistent definition is of utmost importance. This mitigates the ability in comparing the different studies, and the efficacy of the methods used for treating T2DM. Even though the ADA has set criteria, many studies were found to be using other definitions. Our suggestion would be to have a consistent, uniform definition of T2DM remission to increase the accuracy of results. We also noticed that many papers did not discuss the possible long-term adverse effects of bariatric surgery. More studies are needed that highlight this topic alongside the benefits of the surgery. A reason for this could be that many studies do not have a follow-up period that would allow a factor as such to be investigated. In this section, the only study that has a 15-year-long follow-up period is Sjöström et al. [35]. Therefore, more studies are needed that follow patients for a long period.

Comparison of Different Types of Bariatric Surgeries
There are many different types of bariatric surgeries. These include RYGB, SG, GB, BPD, and One-Anastomosis Gastric Bypass (OAGB). RYGB and SG have been greatly studied for their efficacy in remitting T2DM (Figure 2).

FIGURE 2: Roux-en-Y Gastric Bypass
Several observational studies focused on just a single type of bariatric surgery and provided comprehensive findings for its efficacy. Studies conducted by Dasgupta et al. and Nedelcu et al. focused entirely on SG. The conclusion achieved within both of these studies was consistent in that sleeve gastrectomy resulted in T2DM remission in over 50% of patients [7,37] The Adams et al. study gives an insightful look into the realities of long follow-up periods and their effect on T2DM remission rates. It displayed this with 75% remission at two years, 62% at six years, and consistently declining to 51% at 12 years [29]. Similarly, the Madsen et al. study had 74% remission that patients showed at the one-year follow-up, however, 27% of those patients relapsed after five years [26]. This seems to be consistent with other studies where patients are assessed at different intervals postoperatively and the remission rates decrease.
A 2019 RCT serves as evidence of which method is more reliable and effective. This RCT was conducted by Hofsø et al. in Norway and assessed remission at one year postoperatively. They compared SG and RYGB, and their impact on T2DM remission rates. The study results proved that RYGB is more effective than SG. The reasons why results from the Hofsø et al. study are accurate and more representative is because it is an RCT, which is inherently higher on the pyramid of evidence and is a more recent study with a good sample size. Additionally, it is one of the few studies that incorporated a triple-blind protocol, creating an environment ideal for mitigating bias within research [15].
An observational study from 2015 by Yska et al. compared three different surgical methods. This study included RYGB and SG, and a third procedure, GB. By assessing the overall effect of each surgery method on the test group, Yska et al. concluded that two years after surgery, RYGB had the greatest decrease in HbA1c and blood glucose, followed by SG and then GB [32]. Another similar observational study by Wazir et al. was conducted in 2019. This study also included a comprehensive view of GB, SG, and RYGB. This study resembled the 2015 Yska et al. study as it concluded with RYGB being superior to the other methods with greater remission rates at the two-year mark [14].
Similarly, the observational study by Purnell et al. in 2016 compared the effects of RYGB and GB. After a three-year follow-up period, the study concluded that 68.7% of patients in the RYGB group and 30.2% in the GB group were in T2DM remission. Therefore, patients within the RYGB group achieved higher remission rates. An interesting finding was that during the annual follow-ups, the number of RYGB patients in remission decreased as the years progressed, from 71% in year one to 68.7% in year three, compared to GB patients who slightly rose from 29.9% in year one to 30.2% in year three. This infers that regardless of the initial remission rate, patients can maintain sustainable remission after GB in the postoperative years [31]. Further studies with longer follow-up periods are needed to assess if this is true on a long-term scale.
BPD is another type of bariatric procedure, which was included in the observational study by Fernández-Soto et al. conducted in 2017. Alongside BPD, they assessed the efficacy of SG and RYGB. They had a sample size of 49 patients. Contrary to other studies, they concluded that there was no difference in T2DM remission rates after the three procedures, and a substantial 80% of patients had remitted at a one-year follow-up. They mention that this finding could be because only morbidly obese (BMI >50.2kg/m 2 ) patients were included in their study, as opposed to other studies (which include patients with lower BMIs) that show a clear superiority of RYGB over other procedures [28]. This finding from the Fernández-Soto et al. study can be compared with results from Behbehani et al., where participants had a BMI >50.3kg/m 2 and RYGB was still the dominant procedure for attaining remission. There was also a larger sample size and a longer follow-up period in the Behbehani et al. study [36]. However, it should be noted that the research by Fernández-Soto et al. is relatively more recent.
A recent 2019 study conducted by Ruiz-Tovar et al., including 546 patients, investigated the differences in remission attained through SG, RYGB, and OAGB. The findings from this study differed from other research, as they concluded that OAGB was superior to RYGB and SG in remitting T2DM at a five-year follow-up. There was also a greater decrease in BMI after OAGB, compared to RYGB and SG. Furthermore, T2DM remission in RYGB was not significantly higher than with SG, a unique discovery not found in other studies. Due to its study design as an RCT, the Ruiz-Tovar et al. study had more credibility, as patients were randomly assigned to intervention type [20].
Overall, RYGB conclusively proves more effective for patients with BMI >35kg/m 2 . A limitation found while searching for data was the shortage of RCTs specifically catered towards the comparison of different bariatric surgical methods, especially those that incorporate a double/triple-blind protocol. Additionally, there is a shortage of RCTs targeting bariatric surgery for morbidly obese T2DM patients with BMI >50. The OAGB procedure seems to be a promising alternative to treating T2DM, however, more RCTs should focus on this particular modality to check for remission rates. In addition, larger sample sizes and longer follow-up periods are needed.

Preoperative Predictive Factors
To ensure that bariatric surgery is performed only on those patients that could see possible remission, we need to research different factors resulting in a good post-op prognosis. These factors affect not only the efficacy of surgery, but also the long-term implications. We researched different observational studies to determine these clinical factors.
A study conducted by Zenti et al., with 105 patients, showed that the best outcome seen after bariatric surgery was having a <two-year-long diabetes duration, exhibiting good glycemic control (HbA1c <7%), and age of <45 years [34]. patients that were treated with insulin before bariatric surgery. The initial remission rates, at a median of 11.7 months, amounted to only 27%, of which 65% relapsed at a median of 26.7 months. Therefore, due to the large number of individuals who relapsed post-surgery, it can be concluded that pre-op insulin use is an indicator of a worse prognosis after bariatric surgery [30].
Alongside the many different factors that play a role in inducing remission after bariatric surgery, one noteworthy factor is the level of an inflammatory marker, osteopontin. This was researched by a study in 2019 by Carbone et al. They included 41 patients in their study and arrived at the conclusion that at the three-year follow-up, high levels of osteopontin correlated with greater T2DM remission. They also discovered that baseline osteopontin levels predicted remission independent of weight loss, baseline BMI and duration of diabetes. Serum osteopontin levels of >46.57 mg/mL predicts possible remission [27].
Honarmand et al. focused their research on the DiaRem score, which is a scoring system to predict T2DM remission post-bariatric surgery. The baseline variables that are assessed in this scoring system are the age of the individual, HbA1c level, insulin therapy, and use of oral antidiabetic medication like insulin sensitizers or sulfonylureas. Nine hundred patients were included in this study to check for the efficacy of the DiaRem score. The findings showed that at one-year follow-up after RYGB, 37% of individuals had remitted, and the clinical factors within the DiaRem score were associated with this remission. A helpful DiaRem cutoff score of <5 (71.8% sensitivity, 71.3% specificity) was established by this study, and they concluded that this scoring system had a moderate predictive value at best [1].
In light of the evidence available from these studies, it is safe to conclude that age of <60 years, duration of T2DM of <5 years, BMI >40 kg/m 2 , HbA1c <7%, fasting C-peptide levels of >3, osteopontin level >46.57 mg/mL, no preoperative insulin treatment, and a DiaRem score of <5 indicate good prognosis after bariatric surgery. A limitation that was felt while searching for predictive factors was that not many studies were RCTs, especially the ones that specifically focus on preoperative clinical factors. More double-and triple-blind RCTs should focus on this crucial factor. Another suggestion that we would like to include is that scoring systems like the DiaRem should be researched more in RCTs and used widely by physicians in choosing the right individuals for the intervention. In addition, the correlation of levels of osteopontin and T2DM remission should be studied more using a large sample size with a prolonged follow-up period to come to a definitive conclusion about this correlation. In addition, more inflammatory markers like this one should be researched as possible indicators of remission in RCTs.

Limitations
One of the limitations of our study was a scarcity of double/triple-blind RCTs. Within the selected RCTs, the patients were randomly assigned but were not blind to the intervention. In addition, many studies did not use a consistent definition for T2DM remission. We also only chose English language papers, which hindered our ability to find quality literature in other languages. Furthermore, our research did not include Type 1 Diabetes Mellitus patients, individuals with mild obesity, or pre-diabetes. As a result, we were unable to acquire a complete understanding of the effects of bariatric surgery in those populations, and how that compares with effects on T2DM.

Conclusions
This systematic review focused on comparing the effects of different types of bariatric surgeries with LMI on T2DM remission, along with preoperative factors that help achieve this remission. Based on the review, it appears that bariatric surgery is superior to LMI in inducing remission in T2DM patients, and especially the RYGB modality. The perfect candidate would be an individual aged <60 years with a BMI >40kg/m 2 , having a <5-year long T2DM duration, an HbA1c level of <7%, a fasting C-peptide level of >3, and no prior insulin treatment. Along with these factors, having an osteopontin level of >46.57 mg/mL and a DiaRem score of <5 is a great indicator of remission postoperatively. Given the high prevalence of T2DM and the associated complications, our systematic review will be very useful for physicians in selecting appropriate individuals for surgery, who will truly benefit from the intervention. More double-or tripleblind RCTs with a uniform definition of T2DM remission, large sample sizes, and prolonged follow-up periods (to assess for long term adverse effects and relapses in remission) are needed.