Journal of Clinical Medicine Article Genetic Testing to Predict Weight Loss and Diabetes Remission and Long-Term Sustainability after Bariatric Surgery: A Pilot Study Andreea Ciudin 1,2, *,† , Enzamaria Fidilio 1,† , Angel Ortiz 1 , Sara Pich 3 , Eduardo Salas 3 , Jordi Mesa 1,2 , Cristina Hernández 1,2 , Olga Simó-Servat 1,2 , Albert Lecube 2,4 and Rafael Simó 1,2, * 1 Institut de Recerca Vall d’Hebron, Universitat Autònoma de Barcelona (VHIR-UAB), 08035 Barcelona, Spain 2 CIBER de Diabetes y Enfermedades Metabólicas Asociadas, Instituto de Salud Carlos III, 08950 Barcelona, Spain 3 Scientific Department, Gendiag.exe, Joan XXIII, 10, Esplugues de LLobregat, 08950 Barcelona, Spain 4 Endocrinology and Nutrition Department, Hospital Universitari Arnau de Vilanova, IRBLleida, Universitat de Lleida, 25198 Lleida, Spain * Correspondence: aciudin@vhebron.net (A.C.); rafael.simo@vhir.org (R.S.); Tel.: +34-934894172 (A.C. & R.S.); Fax: +34-934894032 (A.C. & R.S.) † Theses authors have contributed equally to this work.  Received: 26 June 2019; Accepted: 28 June 2019; Published: 3 July 2019  Abstract: Introduction: The aim of this pilot study was to assess genetic predisposition risk scores (GPS) in type 2 diabetic and non-diabetic patients in order to predict the better response to bariatric surgery (BS) in terms of either weight loss or diabetes remission. Research Design and Methods: A case-control study in which 96 females (47 with type 2 diabetes) underwent Roux-en-Y gastric by-pass were included. The DNA was extracted from saliva samples and SNPs were examined and grouped into 3 GPS. ROC curves were used to calculate sensitivity and specificity. Results: A highly sensitive and specific predictive model of response to BS was obtained by combining the GPS in non-diabetic subjects. This combination was different in diabetic subjects and highly predictive of diabetes remission. Additionally, the model was able to predict the weight regain and type 2 diabetes relapse after 5 years’ follow-up. Conclusions: Genetic testing is a simple, reliable and useful tool for implementing personalized medicine in type 2 diabetic patients requiring BS. Keywords: diabetes; obesity; bariatric surgery 1. Introduction Obesity represents a major public health problem and it is associated with a significant economic burden on the health systems of developed countries, mainly due to the associated co-morbidities. Among these co-morbidities, type 2 diabetes (T2D) is one of the most important. Bariatric surgery (BS) is a successful treatment for morbid obesity and leads to a dramatic improvement in obesity-related comorbidities [1]. The remission rate of T2D after BS is around 60–70% after 1 year of follow-up [2]. Therefore, there is a significant proportion of non-responders to BS in terms of diabetes remission. Additionally, after 5 years, there is about a 20–35% relapse of T2D after Y-de-Roux gastric by-pass (RYGB) [3–5]. A score based on clinical variables for the pre-operative prediction of T2D remission following RYGB surgery (DiaRem) was proposed [6]. However, this model has several limiting factors [7] and it has not been generally adopted in clinical practice. At present, there are no reliable predictors of T2D remission and relapse after BS. J. Clin. Med. 2019, 8, 964; doi:10.3390/jcm8070964 www.mdpi.com/journal/jcm J. Clin. Med. 2019, 8, 964 2 of 8 In recent years, interest in the genetic influence on the response of different treatments for obesity has increased. Two retrospective studies [8,9] showed that several single nucleotide polymorphisms (SNPs) were associated with a poor response to BS. However, in these studies the discrimination capacity of the GPS was not significant, and the role of T2D in the response to BS and the impact of these genetic factors on diabetes remission were not evaluated. On this basis, the aim of the present study was to evaluate whether genetic markers can be used for the prediction of adequate weight loss and diabetes remission after BS. 2. Material and Methods A single-center, retrospective observational pilot study in a third-level university hospital (Vall d´Hebron University Hospital, Barcelona, Spain) was conducted following the Strengthening the Reporting of Observational Studies in Epidemiology guidelines. The study comprised patients that underwent RYBG surgery between January 2010 and December 2012. The inclusion criteria were women, stable weight in the prior 6 months before BS, and minimum of 5 years of follow-up after BS. In order to avoid heterogeneity and given that the vast majority of the patients under bariatric surgery were women, we decided to rule out the inclusion of men in this pilot study. The patients were informed about the study and they all signed the written informed consent form. The exclusion criteria were male, marked mobility problems, a different BS technique apart from RYBG, and severe psychiatric or eating disorders. For the genetic study, a sample of saliva was collected. The characteristics of RYBG were food loop length: 150–180 cm, and bilio-pancreatic loop length: 120 cm, gastric pouch 30 cc3 . The technique was the same in all cases, performed by the same surgical team in our hospital. Excess body weight (EBW) was defined as the amount of weight that was in excess of the ideal body weight (IBW). The percentage of excess weight loss (EWL) was calculated according to the formula: %EWL = (weight before BS (kg) − weight after BS (kg)/EBW(kg)) × 100. The post-BMI weight regain was defined as a 10% regain of the minimal weight after BS. The minimal weight after BMI was achieved at 2 years follow-up for all of the patients. Diabetes remission was defined according to American Diabetes Association (ADA) criteria [10]. Relapse of T2D was defined as one or more of the following conditions: (a) restarting diabetes medication; (b) one or more HbA1c measures ≥ 6.5%; and/or (c) one or more fasting glucose measures ≥ 126 mg/dL [11]. The study was approved by the Local Ethics Committee and registered at Clinical.Trials.gov, NCT02405949. 2.1. Genotyping and Sequencing The DNA was extracted from saliva samples and processed by GoldenGate® Genotyping Assay for VeraCode. The genetic predisposition was assessed using Nutri inCode (NiC) (Ferrer inCode) and selecting the 57 SNPs associated with susceptibility to diabetes, obesity, appetite regulation, weight loss in response to hypocaloric diet, and the response to BS. The details about the SNPs are reflected in the Supplementary Materials. The selected SNPs were grouped into three genetic predisposition risk scores (GPS): diabetes remission, weight loss in non-diabetic subjects, and weight loss in subjects with diabetes. 2.2. Statistical Analysis In order to assess the best predictive GPS, patients were distributed into 4 subgroups according to the BS response (%EWL) and the presence of T2D: (1) %EWL < 40% without diabetes (n = 15); (2) %EWL < 40% with diabetes (n = 16); (3) %EWL > 75% without diabetes (n = 35); and (4) T2D and %EWL > 75% with diabetes (n = 31). Univariate and multivariate logistic regressions were used to establish associations. Akaike Information Criterion (AIC)-based backward selection was used to remove insignificant terms from an initial model containing all the candidate predictors. The calibration J. Clin. Med. 2019, 8, 964 3 of 8 of the model’s adequacy was determined by the Hosmer–Lemeshow test. The area under the ROC curve (AUROC) was used for evaluating the prediction performance of the models. The cut-offs for the developed algorithms were selected as the point which maximizes the Youden index. J. Clin. Med. 2019, 8, x FOR PEER REVIEW 3 of 8 3. Results the ROC curve (AUROC) was used for evaluating the prediction performance of the models. The The clinical cut-offs characteristics for the of the patients developed algorithms included were selected inpoint as the the study whichare shown the maximizes in Table Youden1. index. Apart from age, we did not find any significant differences between diabetic and non-diabetic subjects before 3. Results BS. The diabetic treatment received by subjects with diabetes is displayed in Figure 1. No other medicationThe clinical apart fromcharacteristics of the patients AINEs occasionally and included vitamin in the study areasshown supplements in Table after per protocol 1. Apart bariatric from age, we did not find any significant differences between diabetic and non-diabetic surgery (ciancobalamin 1000 mcg/month, colecalciferol 25.000–100.000 UI/month) were administered. subjects before BS. The diabetic treatment received by subjects with diabetes is displayed in Figure 1. No other medication apart Table from AINEs 1. Baseline occasionally characteristics andpatients of the vitaminincluded supplements in the as per protocol after study. bariatric surgery (ciancobalamin 1000 mcg/month, colecalciferol 25.000–100.000 UI/month) were administered. Non-Diabetic Patients Type 2 Diabetic Patients p N 50 47 Table 1. Baseline characteristics of the patients included in the study. Age (years) 48.0 (37.5; 55.0) 52.0 (46.0; 58.8) 0.0016 Initial BMI (Kg/m2 ) 45.2 (43.0; 48.5) Patients Type Non-Diabetic 42.5 (40.1; 46.4) 2 Diabetic Patients p 0.008 BMI (Kg/m2 ) 2 y post-BS N 31.8 (26.1; 35.6) 50 30.9 (26.8;47 35.7) n.s. 5 y post-BS Age (Kg/m2 ) BMI(years) 32.63 (21; 48.052.14) (37.5; 55.0) 33.68 52.0 (21; (46.0;46.43) 58.8) 0.0016 n.s. HypertensionInitial (%)BMI (Kg/m2) 45.2 (43.0; 48.5) 48.3 42.5 49.5 (40.1; 46.4) 0.008 n.s. Dyslipidemia post-BS BMI (Kg/m2) 2 y (%) 31.8 (26.1; 35.6) 43.2 30.9 45.7 (26.8; 35.7) n.s. n.s. 5 y Sleep apnea (%) post-BS BMI (Kg/m2) 32.63 (21; 52.14) 27.2 33.6829.7 (21; 46.43) n.s. n.s. Hypertension (%) 48.3 49.5 n.s. Dyslipidemia (%) 43.2 45.7 n.s. In the subgroup of the Sleep apnea (%) non-diabetic patients, 27.2 the multivariate 29.7logistic regression equation for n.s. predicting positive weight loss response (%EWL > 75%) after the BS (NiC-Bariatric-ND) includes In the subgroup SNPs associated of theloss with weight non-diabetic patients, in response the multivariate to a hypocaloric dietlogistic regression and SNPs equation associated to for appetite predicting positive weight loss response (%EWL > 75%) after the BS (NiC-Bariatric-ND) includes regulation. The model showed an AUROC of 0.763 (95% CI 0.605 to 0.920; p < 0.001), a sensitivity of SNPs associated with weight loss in response to a hypocaloric diet and SNPs associated to appetite 86.49%, and a specificity of 57.14%. The calibration of the adequacy of the model determined by the regulation. The model showed an AUROC of 0.763 (95% CI 0.605 to 0.920; p < 0.001), a sensitivity of Hosmer–Lemeshow test wasof0.679. 86.49%, and a specificity 57.14%. The calibration of the adequacy of the model determined by the The continuous variables Hosmer–Lemeshow test waswere 0.679.median (1st quartile; 3rd quartile) and the categorical data were percentages. TheBMI: body mass continuous index. variables wereEWL: excess median of weight (1st quartile; 3rdloss. BS: and quartile) bariatric surgery.data the categorical Hypertension were percentages. was defined BMI: body by increased mass index. systolic (≥140 EWL: mmHg) excess of weight loss. or increased BS: bariatric diastolic surgery. (≥90 mmHg) Hypertension blood pressure or by the usewas defined by increased of antihypertensive systolic drugs, (≥140 mmHg) according or increased to current diastolic guidelines. (≥90 mmHg)was Dyslipidemia blood pressure defined byorthe use by the use of antihypertensive drugs, according to current guidelines. Dyslipidemia was defined by of lipid-lowering drugs, decreased values of HDL cholesterol (men < 0.9 mmol/L, women < 1.0 mmol/L) the use of lipid-lowering drugs, decreased values of HDL cholesterol (men < 0.9 mmol/L, women < or by at least one increased value of total cholesterol (>5.2 mmol/L), LDL cholesterol or triglycerides 1.0 mmol/L) or by at least one increased value of total cholesterol (>5.2 mmol/L), LDL cholesterol or (>1.7 mmol/L). triglycerides (>1.7 mmol/L). 14 12 Diet 10 Diet+metformin 8 Diet+SU Diet+iDPPIV+Metformin 6 Diet+iDPPIV 4 Diet+GLP-1AR+Metformin 2 Diet+GLP-1AR 0 Number of patients FigureFigure 1. The1.diabetic The diabetic treatment treatment received received before before BS BS by by thethe subjects subjects with with diabetesincluded diabetes included in in the the study. study. SU: sulphonylurea, SU: sulphonylurea, iDPPIV: iDPPIV: DPPIV DPPIV enzyme enzyme inhibitor, inhibitor, GLP-1AR:GLP-1AR: GLP-1GLP-1 receptor receptor agonists. agonists. J. Clin. Med. 2019, 8, 964 4 of 8 J. Clin. Med. 2019, 8, x FOR PEER REVIEW 4 of 8 Weight regain Weight after regain 5 years’ after 5 years’follow-up follow-upwas was seen seen in 9.6% 9.6% ofofthe thepatients. patients.TheThe model model to identify to identify the the patients who patients had who hadpresented presentedweight weightregain regain after years’follow-up after 5 years’ follow-upshowed showed anan AUROC AUROC of 0.834 of 0.834 (95%(95% CI 0.705 CI 0.705 to 0.923; p

75%) after BS (NiC-Bariatric-D) included SNPs response (%EWL > 75%) after BS (NiC-Bariatric-D) included SNPs associated with weight loss associated with weight loss in in response to hypocaloric diet, SNPs associated to response to BS [9], and SNPs associated response to hypocaloric diet, SNPs associated to response to BS [9], and SNPs associated to response to to response to lifestyle interventions [11]. The model showed an AUROC of 0.929 (95% CI 0.850 to 0.99; p < 0.001), lifestyle interventions [11]. The model showed an AUROC of 0.929 (95% CI 0.850 to 0.99; p < 0.001), a sensitivity of 87.10% and a specificity of 93.33%. The calibration of the model’s adequacy a sensitivity of 87.10% and a specificity of 93.33%. The calibration of the model’s adequacy determined determined by the Hosmer–Lemeshow test was 0.291. Weight regain in subjects with diabetes was by the Hosmer–Lemeshow test was 0.291. Weight regain in subjects with diabetes was observed in observed in 17.5% of them. The model to identify patients with diabetes who will regain weight after 17.5% of them. of a follow-up The model 5 years to identify after bariatricpatients surgery with showed diabetes in thiswho case will regain weight an AUROC of 0.781after (95%aCI follow-up 0.623 to of 5 years after bariatric surgery showed in this case an AUROC of 0.781 (95% CI 0.623 0.896; p < 0.04), a sensitivity of 71.43%, and a specificity of 84.85%. The calibration of the model’s to 0.896; p < 0.04), a sensitivity of 71.43%, and a specificity of 84.85%. The calibration of adequacy determined by the Hosmer–Lemeshow test was 0.8664. Figure 2 shows the AUROC the model’s adequacy determined by the Hosmer–Lemeshow corresponding to weighttest wasin0.8664. regain Figure the whole 2 shows (Figure 2A)the AUROC corresponding population, to weight non-diabetic subjects regain (Figure in the 2B),whole (Figure and T2D 2A)(Figure patients population, 2C). non-diabetic subjects (Figure 2B), and T2D patients (Figure 2C). Figure 2. Cont. J. Clin. Med. 2019, 8, x FOR PEER REVIEW 5 of 8 J. Clin. Med. 2019, 8, 964 5 of 8 J. Clin. Med. 2019, 8, x FOR PEER REVIEW 5 of 8 Figure Figure 2. The 2. The predictive predictive capacity capacity of the of the genetic genetic score score for for weight weight regain regain after after 5 years’ 5 years’ follow-up follow-up in the in the Figure 2. The predictive capacity of the genetic score for weight regain after 5 years’ follow-up in the whole whole whole(A) population, (A) population, non-T2D subjects (B), and T2D patients non-T2D subjects (B), and T2D patients (C). (C). (A) population, non-T2D subjects (B), and T2D patients (C). Diabetes Diabetes remission remission waswas seen seen in in 73.91% 73.91% of of thethe typetype 2 diabetic patients 2 diabetic included in thethe study Diabetes remission was seen in 73.91% of the type 2 diabeticpatients patientsincluded included in in the study study (66.67% (66.67% in the group in theingroup of %EWL of %EWL < < 40%40% and andand 77.42% 77.42% in the in in group thethegroup of %EWL ofof%EWL > 75%). Diabetes relapse was (66.67% the group of %EWL < 40% 77.42% group %EWL>>75%). 75%). Diabetes relapse Diabetes relapse wasseen seen inin was 25%25% seen of inofthe 25% patients. theofpatients. the patients. The multivariate The multivariate The multivariate logistic logistic regression regression logistic regressionequation equation for equationforthetheprediction for thepredictionof diabetes prediction ofof diabetesremission diabetes remission remissionand relapse and and after BS relapserelapse(NiC-Bariatric-DR) after BSafter included BS (NiC-Bariatric-DR) (NiC-Bariatric-DR) SNPs associated included included SNPsSNPs with obesity, associatedwith associated SNPs withobesity, associated obesity,SNPs with SNPs associated associatedweight withloss with in response weight weight to loss inloss hypocaloric in response response diet, SNPs associated to hypocaloric to hypocaloric diet,diet, SNPs with SNPs appetite associated associated regulation, withappetite with appetite and SNPs associated regulation, regulation, and SNPs and SNPs with genetic predisposition associated with to genetic diabetes. This predisposition prediction to diabetes.model This showed prediction associated with genetic predisposition to diabetes. This prediction model showed an AUROC of an AUROC model showedof 0.868 an (95% AUROC CI 0.709 of 0.868 CI to 0.976; 0.868 (95% <0.709 p(95% CI 0.709 0.0001)to for todiabetes 0.976;0.976; p 75% (p = 0.674). In It isaddition, worth mentioning previous data that in diabetic showed patients, that about 30% oftheT2Dratepatients of remission that arewas ablenot significantly to discontinue thedifferent betweenmedication the group with after %EWL BS will < 40% present and within a relapse the group with the first %EWL 5 years [3–5].>Some (p = 0.674). 75% studies In addition, found weak previouscorrelation data showed between that aboutregain, weight 30% of T2D patients younger thatBMI age or lower arebefore able toBSdiscontinue as predictors of theT2D medication relapse after after BS [5,14], while other studies found no association [3]. Therefore, BS will present a relapse within the first 5 years [3–5]. Some studies found weak correlation between at present, there are no reliable predictors of T2D relapse after BS. In our study, the proposed score showed a high weight regain, younger age or lower BMI before BS as predictors of T2D relapse after BS [5,14], while predictive value of T2D relapse after BS, thus, underlying the potential key role of genetic testing in other studies found no association [3]. Therefore, at present, there are no reliable predictors of T2D precision medicine in order to assure better outcomes after BS. Interestingly, in our study, in the relapse subgroup after BS. of In T2D our patients, study, the theproposed inclusion of score SNPsshowed associated a high predictive to response to BS value did notofimprove T2D relapse the after BS, thus, underlying the potential key role of genetic testing in precision medicine in order to assure better outcomes after BS. Interestingly, in our study, in the subgroup of T2D patients, the inclusion of SNPs associated to response to BS did not improve the prediction scores, suggesting that these genes are not critical or do not intervene in the remission and relapse of T2D. This finding suggests that the J. Clin. Med. 2019, 8, 964 7 of 8 physiopathology of diabetes remission and relapse after RYGB might not be related with the %EWL in this population. Overall, these results are intriguing and point to a genuine genetic background in the mechanisms involved in diabetes remission and relapse after BS, perhaps related to insulin resistance. In conclusion, in this pilot study we have developed highly sensitive and specific genetic predictive scores of responses to BS in terms of weight loss and T2D remission and the long-term sustainability of these effects. These results would allow us not only to implement a more effective and personalized BS, but also to optimize healthcare resources. However, further studies with a larger sample size to confirm this pilot study are needed. Supplementary Materials: The supplementary materials are available online at http://www.mdpi.com/2077-0383/ 8/7/964/s1. Author Contributions: A.C. and E.F. contributed to the study design, obtained study data, assisted in data analysis, and drafted the initial manuscript. O.S. and A.O. obtained study data. S.P. and E.S. performed genetic and statistical analyses. J.M., C.H. and A.L. made substantial contributions to the acquisition, analysis and interpretation of data, and critically reviewed the manuscript. R.S. contributed to the study design, discussion, reviewed the manuscript, supervised the project, and is the guarantor of this work, and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. All authors approved the final version of the manuscript. Funding: This study was supported by grants from the “Pla Estratègic de Recerca i Innovació en Salut” (PERIS) 2016-2020 (SLT002/16/00497), the Instituto de Salud Carlos III (Fondo de Investigación sanitaria, PI 18/00964), and the European Union (European Regional Development Fund). O.S. is a recipient of a “Río Hortega” fellowship grant from the Instituto de Salud Carlos III. Conflicts of Interest: No potential conflict of interest relevant to this article were reported. References 1. Rubino, F.; Kaplan, L.M.; Schauer, P.R.; Cummings, D.E. Diabetes Surgery Summit Delegates. The Diabetes Surgery Summit consensus conference: Recommendations for the evaluation and use of gastrointestinal surgery to treat type 2 diabetes mellitus. Ann. Surg. 2010, 251, 399–405. [CrossRef] [PubMed] 2. Esposito, K.; Maiorino, M.I.; Petrizzo, M.; Bellastella, G.; Giugliano, D. Remission of type 2 diabetes: Is bariatric surgery ready for prime time? Endocrine 2015, 48, 417–421. [CrossRef] [PubMed] 3. Schauer, P.R.; Bhatt, D.L.; Kirwan, J.P.; Wolski, K.; Aminian, A.; Brethauer, S.A.; Navaneethan, S.D.; Singh, R.P.; Pothier, C.E.; Nissen, S.E.; et al. Bariatric Surgery versus Intensive Medical Therapy for Diabetes-5-Year Outcomes. N. Engl. J. Med. 2017, 376, 641–651. [CrossRef] [PubMed] 4. Arterburn, D.E.; Bogart, A.; Sherwood, N.E.; Sidney, S.; Coleman, K.J.; Haneuse, S.; O’Connor, P.J.; Theis, M.K.; Campos, G.M.; McCulloch, D.; et al. A multisite study of long-term remission and relapse of type 2 diabetes mellitus following gastric bypass. Obes. Surg. 2013, 23, 93–102. [CrossRef] [PubMed] 5. DiGiorgi, M.; Rosen, D.J.; Choi, J.J.; Milone, L.; Schrope, B.; Olivero-Rivera, L.; Restuccia, N.; Yuen, S.; Fisk, M.; Inabnet, W.B.; et al. Re-emergence of diabetes after gastric bypass in patients with mid- to long-term follow-up. Surg. Obes. Relat. Dis. 2010, 6, 249–253. [CrossRef] [PubMed] 6. Still, C.D.; Wood, G.C.; Benotti, P.; Petrick, A.T.; Gabrielsen, J.; Strodel, W.E.; Ibele, A.; Seiler, J.; Irving, B.A.; Celaya, M.P.; et al. Preoperative prediction of type 2 diabetes remission after Roux-en-Y gastric bypass surgery: A retrospective cohort study. Lancet Diabetes Endocrinol. 2014, 2, 38–45. [CrossRef] 7. Aminian, A.; Brethauer, S.A.; Kashyap, S.R.; Kirwan, J.P.; Schauer, P.R. DiaRem score: External validation. Lancet Diabetes Endocrinol. 2014, 2, 12–13. [CrossRef] 8. Still, C.D.; Wood, G.C.; Chu, X.; Erdman, R.; Manney, C.H.; Benotti, P.N.; Petrick, A.T.; Strodel, W.E.; Mirshahi, U.L.; Mirshahi, T.; et al. High allelic burden of four obesity SNPs is associated with poorer weight loss outcomes following gastric bypass surgery. Obesity 2011, 19, 1676–1683. [CrossRef] [PubMed] 9. Mirshahi, U.L.; Still, C.D.; Masker, K.K.; Gerhard, G.S.; Carey, D.J.; Mirshahi, T. The MC4R(I251L) allele is associated with better metabolic status and more weight loss after gastric bypass surgery. J. Clin. Endocrinol. Metab. 2011, 96, E2088–E2096. [CrossRef] [PubMed] 10. American Diabetes Association. Standards of Medical Care in Diabetes 2012. Diabetes Care 2012, 35, S11–S63. [CrossRef] [PubMed] J. Clin. Med. 2019, 8, 964 8 of 8 11. Buse, J.B.; Caprio, S.; Cefalu, W.T.; Ceriello, A.; del Prato, S.; Inzucchi, S.E.; McLaughlin, S.; Phillips, G.L., II; Robertson, R.P.; Rubino, F.; et al. How do we define cure of diabetes? Diabetes Care 2009, 32, 2133–2135. [CrossRef] [PubMed] 12. Maggard, M.A.; Shugarman, L.R.; Suttorp, M.; Maglione, M.; Sugerman, H.J.; Sugarman, H.J.; Livingston, E.H.; Nguyen, N.T.; Li, Z.; Mojica, W.A.; et al. Meta-analysis: Surgical treatment of obesity. Ann. Intern. Med. 2005, 142, 547–559. [CrossRef] [PubMed] 13. de Hollanda, A.; Ruiz, T.; Jiménez, A.; Flores, L.; Lacy, A.; Vidal, J. Patterns of Weight Loss Response Following Gastric Bypass and Sleeve Gastrectomy. Obes. Surg. 2015, 25, 1177–1183. [CrossRef] [PubMed] 14. Chikunguwo, S.M.; Wolfe, L.G.; Dodson, P.; Meador, J.G.; Baugh, N.; Clore, J.N.; Kellum, J.M.; Maher, J.W. Analysis of factors associated with durable remission of diabetes after Roux-en-Y gastric bypass. Surg. Obes. Relat. Dis. 2010, 6, 254–259. [CrossRef] [PubMed] © 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).