EFFICACY OF DUODENAL SWITCH OPERATION IN THE MANAGEMENT OF DIABETES IN MORBIDLY OBESE PATIENTS

Atif Mohamed Abd El -Latif, Yasser Ali Elsayed, Tamer Youssef mohamed, Waleed ahmed gado and Mohamed Hafez. ...................................................................................................................... Manuscript Info Abstract ......................... ........................................................................ Manuscript History Received: 12 August 2019 Final Accepted: 14 September 2019 Published: October 2019


Statistical analysis:-
The study was performed at 95% level of significance and power of 80%. The collected data were coded, processed and analysed using the SPSS (Statistical Package for Social Sciences) version 22 for Windows® (SPSS Inc, Chicago, IL, USA). Qualitative data was presented as number (frequency) and Percent. Paired samples t-test was used to compare patients in the same groups at different time points. P < 0.05 was considered to be statistically significant.

Results:-
A total of 25 cases were included in this study. Regarding their demographics, the mean age of the included cases was 34.16 (range, 18-45 years). The majority of the included cases were females (21 cases -84%) whereas only 4 males were included (16%). The mean preoperative BMI was 61.15 kg/m 2 . When it comes the comorbidities, all cases (100%) were having diabetes mellitus while hypertension was present only in 5 cases (20%). Furthermore, obstructive sleep apnea was evident in 2 cases (8%), and arthritis was reported only in one case (4%). These data are shown at table (1). The mean preoperative patient weight was 159.65 kg. After 3 months, there was a marked decrease in the patient weight as it reached 138.85 kg (p < 0.0001). The same significance was kept constant when comparing the following follow up visits' readings to the preoperative recordings. Patients weight continued to decrease down to 124. 4, 112.75, and 103.63 kg at 8-month, 12-month, and 24-month visits (p < 0.0001). These data are shown in table (2) and figure (1).     Like serum bilirubin, serum calcium levels were not significantly different when comparing preoperative levels to the post-operative readings at the scheduled follow up visits. These data are illustrated in table (6). At the planned follow up visits, the HDL levels showed a significant increase when compared to the baseline levels preoperatively (p < 0.0001). The highest level of HDL measured was at 2-year follow up (62.83 mg/dl). Table (7) illustrates these data. Regarding triglycerides, the post-operative readings were all significantly lower than the preoperative levels. However, the 2-year follow up visit readings were relatively of smaller significance (p = 0.021) when compared to the previous readings at 8-month and 12-month follow up visits (p < 0.0001). Table (8) illustrates these data. When it comes to HbA1c, all the readings measured postoperatively at follow up visits were significantly lower than the preoperative levels (p < 0.0001). These data are illustrated in table (9) and figure (3). 7.03 ± 2.11 P value (in relation to pre-operative value) < 0.0001** At 2 years 6.58 ± 1.85 P value (in relation to pre-operative value) < 0.0001** The blood glucose levels showed a significant decrease in the follow up visit readings when compared to the preoperative levels (p < 0.0001). Table 10 and figure 4 illustrate these data.   Regarding the incidence of gall stones postoperatively, at 3-month follow up, there was no new cases. At 8-month and 12-month visits, one case and other 2 cases developed gall stones. Moreover, at 2-year follow up, 1 additional case developed gall stones. Table ( 11) illustrates these data. As illustrated in table (12) and figure (5), at 1-year follow up visit, twelve cases had full remission of diabetes (48%), whereas diabetes control and improvement were achieved in 4 and 2 cases respectively. On the other hand, failure was encountered in 7 cases (28%).
On 2-year follow up visit, eight cases were still showing full remission (32%) while four cases (16%) showed disease control. The number of cases who showed full improvement increased up to 9 cases (36%). The number of cases with failure decrease down to 4 cases (16%).  In the early postoperative period, leakage was encountered in 1 case. On the long term, adhesive intestinal obstruction was detected in 1 case while GERD was experienced in 2 cases. These data are shown in table (13) and figure (6).

Discussion:-
It has been reported that there is a global rise in the incidence of type II diabetes which leads to an increase in its related morbidity and mortality especially from cardiovascular complications 12 .
Type II diabetes is a chronic disease that could be managed by diet modifications, exercise, and medications. However, there is a weak chance of cure. Unfortunately, the results of the previously mentioned modalities are disappointing according to multiple reports in the literature 13, 14 .
Conversely, it has been reported that good diabetes remission and control rates could be achieved in obese diabetic patients who underwent bariatric surgery procedures. Although multiple bariatric procedures are existing, better diabetic control was achieved via gastric bypass and biliopancreatic diversion and duodenal switch procedures 15 .
This study was conducted at Mansoura University Hospitals aiming to evaluate the effect of loop duodenal switch on diabetes. A total of 25 diabetic patients with mean age of 34.16 years and mean BMI of 61.15 kg/m 2 were included in the study.
Regarding comorbidities other than diabetes expressed in our study patients, 4 cases (16%) was complaining of hypertension while 2 cases (8%) had OSA. Only one case (4%) expresses arthritis at the time of examination.
Obesity itself is considered to be a major cause of morbidity and mortality. Among the well-known complications of severe obesity are increased risks of developing diabetes, hypertension, hyperlipidemia and obstructive sleep apnea. Furthermore, overweight causes excess weight bearing exerted on weight bearing joints leading to degenerative osteoarthritis 16 .
Incidence of type II diabetes increases in obese individuals and it is positively correlated with their BMI. Besides, about 25% of hypertension in adults results from excess body weight. Impaired lipid profile is also present in obese individuals, especially persons having central fat distribution 16  It is thought that super obesity (BMI> 50kg/m2) is the main clinical indication for duodenal switch operation. Nevertheless, this concept cannot be generalized as many of these patients can not adhere to the postoperative instructions. It is also reported that Roux-en-Y gastric bypass is not as effective as DS in that patient group as about 20% of these cases fail to reach the expected weight loss 17 . However, two years after operation, other reports have published similar results between the two procedures (%EWL 72 % and 60 % respectively) 18 .
As Ds has a malabsorptive component, late weight regain is also uncommon. A more durable long term result can be obtained by DS 18 with 90 % EWL maintained between 2 -5 years after the procedure 19 .
BPD/DS procedure was found to be superior to all other bariatric procedures as reported by Buchwald's landmark meta-analysis as it achieved 70.1 % excess weight loss, as compared to 61.2 % and 45% for gastric bypass and adjustable gastric banding respectively 20 . Of note, this superiority is more noticed in superobese individuals 21 .
Prachand and his colleagues included 350 cases with BMI > 50kg/m 2 in their study conducted to compare between BPD/ DS and RYGB with 3-year follow-up period. A significant decrease in weight and BMI were noticed in BPD/DS when compared to the other group. Additionally, more patients had a BMI < 30 kg/m2 in BPD/DS group at 12 and 18 months. The two groups did not differ significantly regarding post-operative complications and morbidity rates. However, one mortality was reported in the BPD/DS group, but it was statistically insignificant. Hospital stay was significantly loner for DS group compared to the other group 22 .
Another prospective randomized study compared the outcomes of BPD/DS and RYGB in the super obese individuals at 1 year follow up. BPD/DS group was associated with significantly much excess BMI loss (75 %) compared to the RYGB group (54 %) (p< 0.001). Nevertheless, Ds was associated with longer operative time as well as longer hospital stay. No significant differences were detected regarding complication or morbidity rates. Besides, no mortality was reported in that study 23 .
Biertho et al. included 810 obese cases in their series who had DS operation. They had a mean initial BMI of 44.2 ± 3.6 kg/m 2 . After 8 year follow up, an EBWL of 76% was reported which denoted that Ds operation can be also performed for non-superobese individuals 24 .
In addition to marvelous results regarding weight loss, DS patients can also gain a great benefit regarding resolution of obesity related comorbidities like diabetes, hypertension, and sleep apnea. When it comes to diabetes, a 98% diabetes resolution rate was reported in the meta-analysis conducted by Buchwald, compared to 84 % and 48% in RYGB and gastric banding cases respectively 20 .
In this study, all HbA1c readings measured postoperatively at follow up visits were significantly lower than the preoperative levels (p < 0.0001). Moreover, blood glucose levels showed a significant decrease in the follow up visit readings when compared to the preoperative levels (p < 0.0001).
As a result, at 1-year follow up visit, twelve cases had full remission of diabetes (48%), whereas diabetes control and improvement were achieved in 4 and 2 cases respectively. On the other hand, failure was encountered in 7 cases (28%). On 2-year follow up visit, eight cases were still showing full remission (32%) while four cases (16%) showed disease control. The number of cases who showed full improvement increased up to 9 cases (36%). The number of cases with failure decrease down to 4 cases (16%).
The effect of DS operation on diabetes is more noticed in superobese patients. In a previous study, all DS patients were free of their diabetic medications compared to 60% only in the gastric bypass group although diabetic state was more severe in the DS patients 27 .

678
The metabolic benefits of DS operations should be clarified for patients planning to perform bariatric surgery as better diabetic control as well as better lipid profile could be achieved with this procedure when compared to other bariatric procedures 28,29 . In another practice, no fewer than 90 % of type 2 diabetics were rendered euglycaemic after surgery 19 , a finding echoed by others 29,30 .
Dorman and his associates have reported diabetes remission rates of 82% and 64% after DS and RYGB respectively. Moreover, DS cases showed better hypertension control (69% vs. 39 in RYGB), as well as resolution of dyslipidemia (81% vs. 51%) 30 . Another randomized trial reported significantly lower HbA1c levels beside better weight loss at 1 and 3 year follow up in the DS group 31 .
The positive effect of DS operation on diabetes starts early after surgery and persists for years. Scopinaro reported that 75% of patients had an FPG <110 mg/dl 1 -2 months after operation. At 1 year, more than 90% of patients had an FPG <90 mg/dL, and that level was maintained for 10 -20 years of follow-up 32 .
Given the low number of patients who still have diabetes after BPD/DS (<2%), limited conclusions have been made in the literature about predictors of failure with this particular procedure. In a series of patients undergoing GB by Pories, the factors that predicted poor resolution of diabetes were again duration of disease (4.6 vs. 1.6 years, p<0.04) and patient age (48.0 vs. 40.7 years, p<0.01) 33 . The reason for this phenomenon and the utility of these factors in choosing surgical and nonsurgical strategies remain areas for research 34 .
Mingrone et al. conducted a randomized study that included 60 morbidly obese patients with T2DM. They compared the effect of medical therapy (lifestyle modifications and hypoglycemic agents) to surgical intervention (RYGB or BPD). No diabetes remission was reported in the medical group whereas 75% and 95% of cases developed diabetic remission in the RYGB and BPD groups after 2 year follow-up respectively 35  Another systematic review and metaanalysis confirmed that best diabetes resolution rate can be achieved after Ds, followed by RYGB and gastric banding in order of speech 2 .
Astiarraga et al. recently assessed the effect of BPD/DS on T2DM in nonobese patients demonstrating marked amelioration (improved glycemia) of metabolic control and remission (HbA1C <6.5% and normal oral glucose tolerance test) in 1/3 of patients, suggesting a weight independent effect of the operation, as only modest weight loss (-12 kg at 2 months, -14 kg at 1 year) was observed in this nonobese patient population 38 .
Other cardiometabolic risk factors, including hypertension and dyslipidemia, have also shown marked improvement following BPD/DS. Additionally, obstructive sleep apnea was resolved in the majority of patients 26 .
In our study, the lipid profile of the included cases showed a marvelous improvement when compared to the preoperative values. The preoperative mean level of serum cholesterol was 202.21 mg/dl. At the first follow up visit, it decreased down to 171.71 mg/dl (p < 0.0001). On the following visits, there was significant reduction in the cholesterol levels reaching down to 140.05, 123.26, and 104.73 mg/dl at 8-month, 12-month, and 24-month visits respectively.
In addition, at the planned follow up visits, the HDL levels showed a significant increase when compared to the baseline levels preoperatively (p < 0.0001). The highest level of HDL measured was at 2-year follow up (62.83 mg/dl). Regarding triglycerides, the post-operative readings were all significantly lower than the preoperative levels. However, the 2-year follow up visit readings were relatively of smaller significance (p = 0.021) when compared to the previous readings at 8-month and 12-month follow up visits (p < 0.0001).
Regarding post-operative complications, leakage was encountered in 1 case (4%). On the long term, adhesive intestinal obstruction was detected in 1 case (4%) while GERD was experienced in 2 cases (8%).
There is undoubtedly a steep learning curve that surgeons undertaking DS have to negotiate which probably explains the high leak rates reported in some early studies of open and laparoscopic duodenal switches performed 10 or more years ago (0-6.6 % leak rate), and more recent low volume studies (6.8-8 % leak rate). Higher volume, recent studies consistently report leak rates of 0-3.5 % 39 .
In a study of 805 DS patients, Biertho reported a 2.4 % incidence of intestinal obstruction, with 1.6 % requiring further surgery 24 , findings that are similar to those reported after RYGB. Intestinal obstruction after laparoscopic DS can result from simple adhesions, port site hernia, incorrect anastomotic technique (twisting or narrowing) or ischemic stenosis. However, the most dangerous causes are internal herniation of the bowel and organo-axial rotation of the very long alimentary limb 40 .
In this study, hypocalcemia was not encountered in the patients during the scheduled follow up visits. This can be explained by the strict information given for the patients about the serious effects of missing multivitamin and mineral tablets commenced for them every day.
The main drawback of this study is that included a small sample size (n = 25 cases). Furthermore, the follow up of the cases included short and medium term only. As a result, more studies including larger number of cases with longer follow up periods should be conducted in the future.