Tailored one anastomosis gastric bypass - Subgroup analysis of a randomised control trial based on bilio-pancreatic limb length with long-term results of 101 patients.

Background: : Ideal bilio-pancreatic limb (BPL) length is a highly debatable issue in one anastomosis gastric bypass (OAGB). Whether to use a tailored BPL or a fixed-length BPL needs to be answered. Materials and
Methods: : One-hundred and one patients who have undergone tailored OAGB based on basal metabolic index (BMI) and type 2 diabetes mellitus (T2DM) were analysed. Sixty-three patients had BPL of 150 cm and 38 patients had BPL of 180 cm. Mean pre-operative BMI of BPL 150 and 180 cm groups were 39.73 and 51.92 kg/m2, respectively.
Results: There was a significant drop in mean total body weight, BMI and excess body weight of both the groups at 1 year which persisted for 5 years post-operatively. The mean BMI of BPL 150 and BPL 180 cm group at 5 years was 29.17 and 32.88 kg/m2, respectively. Although mean percentage excess weight loss (%EWL) and percentage of excess BMI loss in the two groups was similar, the mean percentage total weight loss (%TWL) was significantly higher for the BPL 180 cm group. There was no difference between the two groups in the number of patients who had >50% EWL and >20% TWL. At 5 years of follow-up, the mean serum iron level was significantly low in BPL 180 cm group. There was a significant drop in mean haemoglobin A1c values postoperatively, with no difference between the two groups. Conclusions: Tailored BPL of 150 and 180 cm do not show any difference in the number of patients achieving >50% EWL or >20% TWL and so increasing limb length may not increase the number of good responders for weight loss. Although the resolution of T2DM and improvement of QoL score do not change significantly with increase in BPL length, mean serum iron levels may be lower with longer BPL.

INTRODUCTION

One anastomosis gastric bypass (OAGB) is now a well-established procedure for morbid obesity.[12] It has shown long-term weight loss, co-morbidity resolution and improvement in quality of life (QoL). The procedure has however faced regular criticism in view of excessive weight loss, nutritional deficiencies and other long-term complications associated with bile reflux. Excessive or inadequate weight loss and nutritional deficiencies may be related to bilio-pancreatic limb (BPL) limb length used in OAGB. It has been hypothesised that higher BPL length is associated with more weight loss and better comorbidity resolution but at the cost of nutritional deficiencies.[3] There have been multiple attempts by various researchers to find an appropriate limb length which will strike balance between weight loss and nutritional deficiencies.[45] Whether this BPL length needs to be tailored as per the basal metabolic index (BMI) or the comorbidities of the patient is still not answered.[6]

Authors have previously published the series with early, mid-term and long-term results (1, 3 and 5-year follow-up, respectively) of a randomised control trial (RCT), comparing laparoscopic sleeve gastrectomy (LSG) with mini-gastric bypass (MGB) and analysed the outcome based on Bariatric analysis reporting and outcome system.[789] Various other studies have shown similar long-term results with OAGB.[101112] The aim of this study is to assess whether tailored limb length according to the BMI and presence of type 2 diabetes mellitus (T2DM) has any impact on weight loss, nutritional deficiency and T2DM resolution.

MATERIALS AND METHODS

This study is one arm of previously published RCT[789] comparing OAGB to LSG and presents retrospective subgroup analysis of the OAGB group. The different BPL length used in this study were not randomised but tailored according to the BMI and presence of T2DM. As already described in previous publications, all patients included in this series were in the age group of 18–60 years and BMI >35 and <60 kg/m2. Patients with a history of psychiatric illness, previous bariatric surgery, not willing to be a part of RCT or lost to follow-up before 1 year were excluded from the series. The operative technique has been described in a previous paper[7] where in a tailored BPL length was used for OAGB. Patients with BMI ≥50 and with BMI ≥45 with T2DM were operated with BPL of 180 cm, while others had a BPL of 150 cm. Subgroup analysis is done based on the BPL length and results are presented.

A total of 107 patients underwent OAGB after randomisation between 2013 and 2015 but only 101 patients followed at 1 year and were enrolled for the study. These patients were followed yearly for 5 years for various weight-loss parameters, nutritional deficiencies, resolution of T2DM and changes in QoL besides various other parameters which were part of RCT. Bodyweight (or total body weight [TBW]) was recorded annually and other parameters such as BMI, excess body weight (EBW), total weight loss (TWL), percentage of excess weight loss (%EWL) and percentage of excess BMI loss (%EBMIL) were calculated. Ideal BMI was taken as 22.5 kg/m2 and ideal body weight was taken as weight which corresponds to BMI of 22.5. EBW was calculated by (present weight − ideal body weight). TWL was calculated by (pre-operative weight − present weight). %EWL was calculated as [(pre-operative weight − present weight)/(pre-operative weight − ideal body weight)] ×100. %TWL was calculated as [(pre-operative weight − present weight)/pre-operative weight] ×100. %EBMIL was calculated as [(pre-operative BMI − present BMI)/(pre-operative BMI − ideal BMI)] ×100. The mean value of these parameters and number of patients who had >50%EWL and >20%TWL were noted along with the number of patients who could achieve BMI <30.

Comorbidity resolution and improvement were analysed for T2DM. Pre-opeartive diagnosis of T2DM was made if the patient was on oral hypoglycaemic drugs or Insulin or had haemoglobin A1c (HbA1c) >6.5. Postoperatively, resolution of T2DM was defined as HbA1c <6.5 without medicines and improvement as HbA1c <6.5 on medicines. HbA1c >6.5 even on medicines was defined as no improvement. The nutritional deficiency was assessed by comparing means values of serum albumin, serum Iron, serum ferritin and Hb levels (values <3.5 g/dl, <40 mcg, <20 mcg/dl and <9 g/dl units, respectively, were defined as deficiency). Number of patients with nutritional deficiencies of these parameters was identified and expressed as percentage. Since both pre and post-operative vitamin and mineral deficiency is common in obese patients,[13] all patients in this study received post-operative protein supplementation along with calcium, iron, vitamin D, Vitamin B12 and other vitamins and minerals in appropriate doses.[614] The importance of nutritional supplements was stressed at every post-operative visit and they were advised to continue their supplements. QoL score of the patients was calculated according to Moorehead Ardelt QoL questionnaire.[1516]

Data analysis

Statistical analysis of the data was performed using (IBM Corp., Armonk, NY) SPSS statistical package. Continuous variables were expressed as mean and standard deviation and categorical variables as frequency and percentage. Data analysis was performed with unpaired t-test for continuous variables such as BMI, EBW, TWL, %EWL, %EBMIL and QoL score. For categorical variables (number of patients with %EWL >50, %TWL >20, BMI <30, and patients with nutritional deficiencies) 2 × 2 contingency table and two-tailed Fisher's exact test was used and P value was calculated. Chi-square test was used for patients with resolution/improvement/no change of T2DM. A two-sided P < 0.05 was considered statistically significant.

RESULTS

Demographic details and pre-operative profile of the 101 patients are mentioned in Table 1. Patients were divided into two groups, one which underwent OAGB with 150 cm BPL limb and other which underwent OAGB with 180 cm BPL limb. The data for the two groups are comparative except that patients with BPL of 180 cm were younger, had higher TBW, BMI and EBW and had higher proportion of diabetics (68% in BPL 180 as compared to 37% in BPL 150). The total number of patients and diabetic patients who followed for the 5 years is shown in Table 2. Results on early and delayed complications and mortality are included in the previous study and since the numbers are very small, sub-group analysis of this data based on BPL length was not done.

Table 1

Pre-operative patient profile

	Mean±SD			P
	Total	BPL 150	BPL 180
Number of patients	101	63	38	-
Age	43.12±10.44	44.97±9.73	40.05±10.97	0.0210*
Male/female	39/62	25/38	14/24	0.8349
Weight	114.40±22.63	102.85±17.52	133.54±16.32	0.0001*
BMI	44.32±7.92	39.73±4.75	51.92±6.05	0.0001*
EBW	53.47±19.32	42.27±11.99	72.03±14.18	0.0001*
QoL score	−2.17±0.36	−2.03±0.34	−2.39±0.25	0.0001*
Serum albumin	4.34±0.19	4.27±0.29	4.22±0.27	0.3854
Serum iron	64.55±21.69	64.59±23.31	64.48±18.99	0.9804
Serum ferritin	83.60±55.87	87.03±63.21	77.91±41.08	0.4295
Hemoglobin	12.32±1.48	12.37±1.50	12.24±1.47	0.6713
HbA1c (diabetics) (%)	9.63±1.06 49 (49)	9.58±1.12 23 (37)	9.67±1.03 26 (68)	0.7711
Iron deficiency (serum iron <40) (%)	16 (16)	10 (16)	6 (16)	-

*P<0.05. SD: Standard deviation, BPL: Bilio-pancreatic limb length, BMI: Basal metabolic index, EBW: Excess body weight, QoL: Quality of life, HbA1c: Hemoglobin A1c

Table 2

Follow up of patients

	Total patients			Diabetic patients, n (%)
	BPL 150	BPL 180	Total	BPL 150	BPL 180	Total
Pre-operative	63	38	101	23 (36.5)	26 (68.4)	49 (48.51)
1 year	63	38	101	23 (36.5)	26 (68.4)	49 (48.51)
2 years	60	37	97	23 (38.33)	26 (70.27)	49 (50.52)
3 years	59	34	93	22 (37.29)	24 (70.59)	46 (49.46)
4 years	49	33	82	19 (38.78)	23 (69.70)	42 (51.22)
5 years	41	32	73	17 (41.46)	23 (71.88)	40 (54.79)

BPL: Bilio-pancreatic limb length

Weight reduction

The mean BMI, TBW and EBW of the two groups over 5 years are mentioned in Table 3a. It shows there was a significant drop in TBW, BMI and EBW at the end of 1 year which persisted over the next 4 years without any significant difference over these years. The two groups however continued to be significantly different from each other over these years. Mean %EWL, %EBMIL, TWL and %TWL of the two groups at 1, 3 and 5 years is mentioned in Table 3b and it shows that although there was no significant difference in their mean %EWL and %EBMIL TWL and %TWL were significantly more in the BPL 180 cm group. Mean %EWL in diabetic and non-diabetic patients at 5 years were 65.85 ± 14.53 and 64.58 ± 13.98, respectively.

Table 3a

Weight, body mass index and excess body weight at 1, 3 and 5 years

	Weight, mean±SD (n)			BMI, mean±SD (n)			EBW, mean±SD (n)
	BPL 150	BPL 180	P	BPL 150	BPL 180	P	BPL 150	BPL 180	P
Pre-operative	102.85±17.52 (63)	133.54±16.32 (38)	0.0001*	39.73±4.75 (63)	51.92±6.05 (38)	0.0001*	42.27±11.99 (63)	72.03±14.18 (38)	0.0001*
1 year	74.60±11.51 (63)	87.47±12.68 (38)	0.0001*	28.68±2.46 (63)	33.52±4.09 (38)	0.0001*	13.70±6.24 (63)	25.99±10.62 (38)	0.0001*
3 years	73.59±11.11 (59)	86.03±13.92 (34)	0.0001*	28.50±2.71 (59)	33.11±4.92 (34)	0.0001*	13.13±7.06 (59)	24.79±12.80 (34)	0.0001*
5 years	78.29±8.34 (41)	86.13±10.17 (32)	0.0006*	29.17±2.05 (41)	32.88±3.66 (32)	0.0001*	14.51±5.66 (41)	24.16±9.60 (32)	0.0001*

*P<0.05. SD: Standard deviation, BPL: Bilio-pancreatic limb length, BMI: Basal metabolic index, EBW: Excess body weight

Table 3b

Weight loss parameters at 1, 3 and 5 years

	1 year (n=101), mean±SD (n)			3 years (n=93), mean±SD (n)			5 years (n=73), mean±SD (n)
	BPL 150	BPL 180	P	BPL 150	BPL 180	P	BPL 150	BPL 180	P
Percentage EWL	67.24±11.38 (63)	64.47±10.03 (38)	0.2188	67.57±16.63 (59)	66.34±14.65 (34)	0.7208	64.20±15.72 (41)	66.65±11.48 (32)	0.4615
Percentage EBMIL	63.91±11.55 (63)	62.81±10.00 (38)	0.6274	63.85±15.62 (59)	64.88±13.35 (34)	0.7479	59.92±14.15 (41)	64.88±10.15 (32)	0.0985
TWL	28.25±10.46 (63)	46.08±9.61 (38)	0.0001*	28.59±13.02 (59)	48.05±12.87 (34)	0.0001*	28.40±12.00 (41)	48.80±14.29 (32)	0.0001*
Percentage TWL	27.01±6.75 (63)	34.49±5.41 (38)	<0.0001*	27.18±8.91 (59)	35.72±7.72 (34)	<0.0001*	25.87±8.29 (41)	35.75±6.86 (32)	<0.0001*
>50% EWL, n (%)	59 (94)	36 (95)	0.3597	53 (90)	30 (88)	1.0000	33 (80)	30 (94)	0.1697
>20% TWL, n (%)	55 (87)	36 (95)	0.3124	47 (80)	32 (94)	0.0749	31 (76)	30 (94)	0.0556
BMI <30, n (%)	47 (75)	6 (16)	0.0001*	42 (71)	9 (26)	0.0001*	28 (68)	8 (25)	0.0004*

*P<0.05. SD: Standard deviation, BPL: Bilio-pancreatic limb length, BMI: Basal metabolic index, EWL: Excess weight loss, TWL: Total weight loss, EBMIL: Excess basal metabolic index loss

The number of patients who had >50%EWL and >20%TWL were identified in the two groups and expressed in percentage Table 3b. There was no significant difference between the two groups on these parameters (P > 0.05). Number of patients who achieved BMI <30 at 1, 3 and 5 years were noted and expressed in percentage Table 3b. It was found that significantly higher proportion of patients in the BPL 150 cm group could reach at BMI <30. This is probably due to the lower pre-operative BMI of patients in this group.

Nutritional deficiencies

Assessment for nutritional deficiency was done for serum albumin, serum iron, serum ferritin and Hb since hypo-albuminemia and anaemia are the most common deficiency associated with OAGB. Mean pre-operative and post-operative values of all these four are mentioned in Table 4a. There was a significant drop in the mean value of these parameters after 1 year. Hb dropped by about 1.5 g/dl and persisted in the same range for the next 4 years. Similarly, there was a sudden drop in mean serum iron at 1 year despite regular supplementation. Although serum iron values improved over the next 4 years, the rise was more in BPL 150 cm group and so the mean serum iron was significantly less in BPL 180 cm at 3 and 5 years. At 1-year follow-up, mean serum albumin was significantly lower in BPL 180 cm group but the two groups were similar at 3 and 5 years. Mean serum ferritin levels were similar for the two groups postoperatively.

Table 4a

Nutritional parameters

	Haemoglobin, mean±SD (n)			Serum iron, mean±SD (n)			Serum ferritin, mean±SD (n)			Serum albumin, mean±SD (n)
	BPL 150	BPL 180	P	BPL 150	BPL 180	P	BPL 150	BPL 180	P	BPL 150	BPL 180	P
Pre-operative	12.37±1.50 (63)	12.24±1.47 (38)	0.6713	64.59±23.31 (63)	64.48±18.99 (38)	0.9804	87.03±63.21 (63)	77.91±41.08 (38)	0.4295	4.27±0.29 (63)	4.22±0.27 (38)	0.3854
1 year	10.99±1.20 (63)	10.7±1.32 (38)	0.2606	53.30±18.69 (63)	50.05±16.94 (38)	0.3830	62.87±43.69 (63)	60.97±35.30 (38)	0.8209	4.19±0.31 (63)	4.04±0.32 (38)	0.0231*
3 years	10.85±1.58 (59)	10.94±1.57 (34)	0.7915	61.47±17.80 (59)	53.75±18.39 (34)	0.0496*	65.28±35.17 (59)	62.07±37.13 (34)	0.6789	4.23±0.30 (59)	4.17±0.37 (34)	0.3986
5 years	10.81±1.43 (41)	10.81±1.09 (32)	1.0000	66.64±13.33 (41)	53.47±14.26 (32)	0.0001*	71.06±44.86 (41)	64.77±40.39 (32)	0.5368	4.10±0.31 (41)	4.12±0.33 (32)	0.7920

*P<0.05. SD: Standard deviation, BPL: Bilio-pancreatic limb length

The number of patients with nutritional deficiencies were identified in both the groups and expressed in percentage Table 4b. Although it is difficult to interpret this data since the numbers are small, at 5 years, anaemia (Hb <9 g/dl) was seen in approximately 10% of patients in both the groups while Iron deficiency (serum iron <40 mcg/dl) was seen in 10% and 19% of patients with BPL of 150 and 180 cm, respectively. Hypo-albuminemia (serum albumin <3.5 gm/dl) was seen in 7% and 12% of patients with BPL length of 150 and 180 cm, respectively, at 5 years. The differences between the two groups were however not significant.

Table 4b

Patients with nutritional deficiency

	1 year (n=101)			3 years (n=93)			5 years (n=73)
	BPL 150 (n=63), n (%)	BPL 180 (n=38), n (%)	P	BPL 150 (n=59), n (%)	BPL 180 (n=34), n (%)	P	BPL 150 (n=41), n (%)	BPL 180 (n=32), n (%)	P
Anemia (haemoglobin<9 g/dl)	2 (3.2)	5 (13.2)	0.1000	6 (10.2)	4 (11.2)	1.0000	4 (9.8)	3 (9.4)	1.0000
Iron deficiency (<40 mcg/dl)	14 (22)	10 (26)	0.6385	9 (15)	7 (21)	0.5736	4 (10)	6 (19)	0.3174
Ferritin deficiency (<20 mcg/dl)	4 (6)	7 (18)	0.0961	4 (7)	6 (18)	0.1623	2 (5)	5 (16)	0.2283
Hypoalbuminemia (<3.5 g/dl)	3 (4.7)	5 (13.2)	0.1486	2 (3.4)	4 (11.8)	0.1865	3 (7.3)	4 (12.5)	0.6917

BPL: Bilio-pancreatic limb length

Resolution of type 2 diabetes mellitus

There were a total of 49 diabetic patients in the whole series, 23 (37%) and 26 (68%) in BPL 150 and 180 cm, respectively. The mean pre-operative and post-operative HbA1c values of the two groups and the number of patients in whom T2DM resolved at 1, 3 and 5 years are mentioned in Table 5. Although there was a significant drop in mean HbA1c values at 1 year, lower mean HbA1c continued for the subsequent years without any significant difference. The two groups also performed similarly irrespective of BPL length both in terms of mean HbA1c values and percentage of patients with resolution or improvement of T2DM.

Table 5

Outcome of diabetes mellitus and quality of life

	HbA1c, mean±SD (n)			Outcome of DM–R/I/N, mean±SD (n)			QoL parameters, mean±SD (n)
	BPL 150	BPL 180	P	BPL 150	BPL 180	P	BPL 150	BPL 180	P
Pre-operative	9.58±1.12 (23)	9.67±1.03 (26)	0.7711	-	-	-	−2.03±0.34 (63)	−2.39±0.26 (38)	0.0001*
1 year	6.29±0.27 (23)	6.41±0.26 (26)	0.1213	16/4/3 (70/17/13)	19/4/3 (73/15/12)	0.8923	2.01±0.43 (63)	1.93±0.48 (38)	0.3888
3 years	6.10±0.27 (22)	6.16±0.26 (24)	0.4424	15/4/3 (68/18/14)	19/2/3 (79/8/13)	0.0951	2.81±0.28 (59)	2.84±0.25 (34)	0.6032
5 years	6.26±0.58 (17)	6.30±0.68 (23)	0.8451	14/2/1 (82/12/6)	20/1/2 (87/4/9)	0.0931	2.35±0.39 (41)	2.35±0.45 (32)	1.0000

*P<0.05. SD: Standard deviation, BPL: Bilio-pancreatic limb length, QoL: Quality of life, HbA1c: Haemoglobin A1c, DM: Diabetes mellitus

Quality of life improvement

Patients in BPL 180 cm group had a significantly poor QoL score as compared to BPL 150 cm group preoperatively. The score however improved after surgery and there was no significant difference between the two groups at 1 year [Table 5]. Higher QoL scores persisted over the next 4 years after surgery without any significant difference between the two groups.

DISCUSSION

RYGB is the gold standard procedure for morbid obesity. It is however technically demanding with a steep learning curve and is associated with life-threatening complications. OAGB was first described by Rutledge in 1997.[17] It is technically less challenging, involves a single anastomosis between a long gastric tube at its most dependant part and the jejunal loop and has an option of easier reversal if the need arise. The procedure however met serious initial criticism mostly due to bile reflux and associated morbidity/complications.[3] Various long-term studies later proved that incidence of bile reflux and associated complications were not significantly higher than the other procedures. Furthermore, long-term studies have also shown sustained weight loss and comorbidity resolution along with overall improvement in QoL with OAGB.[12911] Musella et al. in their large series of OAGB in 2678 patients with 62.6% 5 year follow-up concluded that OAGB is a reliable and favourable procedure in terms of complications when compared to the RYGB and sleeve gastrectomy.[11] Parmar et al. in their large systemic review of 12,807 patients also concluded that there is sufficient evidence to include OAGB as main-stream bariatric procedure.[1]

Unlike RYGB which has an alimentary limb and a BPL, OAGB has only a BPL which is bypassed in loop GJ. The length of BPL appears to be an important determinant of post-operative weight loss. It is prudent to believe that there is a direct relation between BPL length and post-operative weight loss, and so more is the BPL length, more will be the TWL and %EWL. As a result, various researchers have described different criteria to tailor the BPL in OAGB. Most of them have tailored the limb length based on pre-operative BMI of the patients.[181920] While Scavone et al.[18] tailored the limb length from 180 to 240 cm on the basis of pre-operative BMI, Hussain et al.[20] used only two lengths, 150 and 200 cm for patients with BMI <50 and BMI >50 respectively. Noun et al. used a 150 cm BPL for patients with BMI <40 and then increased the BPL length by 10 cm for each BMI.[19] There is also a concern regarding the length of the common channel left after BPL length is increased and so it was advised to measure the total bowel length and tailor the BPL length accordingly. Komaei et al. compared results of fixed length BPL with tailored BPL based on total bowel length. They used 40% of small bowel length as biliopancreatic limb, subject to a minimum of 250 cm as common limb length.[5]

Increasing the BPL length is not free of complications. Since longer limb lengths are associated with nutritional deficiencies,[21] it is essential to strike a balance between post-operative weight loss and prevention of deficiencies. In this series of 101 patients mean pre-operative TBW and BMI of the 101 patients were 114.40 and 44.32, respectively. The two groups (BPL 150 and BPL 180 cm) differ significantly in terms of mean pre-operative TBW, EBW and BMI. Both the groups had a significant fall in these parameters at 1 year with no significant change in subsequent years. This means weight change occurs by 1 year and then persists for long term. Although %EWL was similar in the two groups, %TWL was significantly higher in patients with 180 cm BPL. Since the two groups differ in their pre-operative BMI, we used >20%TWL as the standard definition for identifying good responders after bariatric surgery since this is least influenced by pre-operative BMI as described by Grover et al.[22] Number of patients with more that >50% EWL were also noted since this is probably the next best measure of successful weight loss. It was found that there was no difference between the two groups in the number of patients with >20% TWL and >50% EWL. This suggests that by increasing the limb length we may not be able to increase the number of good responders in terms of weight loss. When the number of patients who could achieve a target BMI of <30 were counted, it was found that there significantly more patients in the BPL 150 group could reach that threshold. However, it could be due to the lower pre-operative BMI of this group.

Various researchers have reported similar results in terms of post-operative weight loss with tailored BPL[1819202324] and fixed-length BPL.[2526] Soong et al.,[23] reported results of 134 OAGB patients with tailored BPL limb of 150–250 cm. Mean %TWL and %EWL at 5 years was 29.2 and 72.1, respectively. Scavone et al.[18] used a tailored BPL and found that patients had 68.80% of EWL at 5 years follow-up while Hussain et al.[20] found 89% and 77% EWL at 1 and 3 years with tailored BPL. Liagre et al.[25] reported 85.7% EWL and 35%TWL at 8 years in OAGB patients with fixed length BPL of 150 cm. Similar results were reported by Neuberg et al.[26] using BPL of 150 cm and 84.7% EWL at 5 years. Other authors have compared results of different fixed length BPL in OAGB and found that there is no significant difference in weight loss by changing BPL length.[42728] Boyle et al.[27] compared results of fixed length BPL of 200 and 150 cm at a mean follow-up of 24.7 months. They found that the mean %EWL in OAGB 200 and OAGB 150 groups was 75% and 74%, respectively, and 89.7% and 85.5% of patients achieved >50% EWL. Pizza et al.[4] compared results of fixed length BPL of 150, 180 and 200 cm and found that there was no significant difference in post-operative weight, BMI, %EWL and %TWL of the three groups at 1 and 2 years follow-up. Komaei et al.,[5] compared results of fixed length BPL (200 cm) with tailored BPL and found that at 1-year follow-up, two groups did not differ in %EWL and TWL. Various authors have reported significantly less weight loss in diabetics as compared to non-diabetic patients with OAGB. While Garciacaballero et al.[29] reported that diabetics had 5%–7% less %EWL than non-diabetics, Ansar et al.[30] found that diabetics were less likely to achieve >50% EWL at 1-year follow-up. This is in contrast to our study where we did not find any significant difference in mean %EWL between diabetic and non-diabetic patients (mean %EWL 65.85 ± 14.53 and 64.58 ± 13.98 respectively, P > 0.05).

With respect to nutritional deficiencies, none of the patients in our series had anemia or hypo-albuminemia pre-operatively but about 16% of patients had pre-operative iron deficiency. This is in line with previously reported results, where pre-operative vitamin and mineral deficiency is common in morbidly obese patients.[1314] Post-operatively, there was a significant drop in Hb, serum iron and serum ferritin levels in both groups. The two groups were however similar in mean values of all these three parameters at 1 year. The significant drop in serum albumin levels was seen in BPL 180 cm group only at 1 year, but the mean value improved over the next 4 years. At 3 and 5 years, serum iron levels were significantly low in BPL 180 cm group as compared to the BPL 150 group suggesting a higher risk of nutritional deficiencies in longer BPL OAGB. Soong et al.[23] in their study on tailored OAGB, found that there was significant drop in mean Hb and serum albumin levels at 5 years while there was no difference in serum Iron levels. In a large study by Kular et al.,[31] anaemia was the most common complication at 6 years of mean follow-up and was seen in 7.6% of patients. Neuberg et al.[26] reported that Hb <10 g/dl was seen in 10.4% of patients at 5-year follow-up. Jedamzik et al.[32] reported a higher nutritional deficiency trend as the limb length is increased. Boyle et al.[27] compared OAGB with 150 and 200 cm BPL and found that there was significant drop in Hb and serum albumin levels post-operatively in both the groups but the difference between the two groups was not significant. Study by Khalaj et al.[28] compared OAGB with 200 and 160 cm BPL and found that at 1-year follow-up although the anaemia and iron deficiency were similar, the incidence of hypo-albuminemia was significantly higher in patients with 200 cm BPL.

In this series, there was no difference in the two groups in terms of QoL score and resolution or improvement of T2DM. QoL score improved significantly for both the groups at 1 year without any difference between the two groups and the improved score continued for the subsequent 4 years. With regard to T2DM, pre-operative incidence was about 50% (37% and 68% in BPL 150 and 180 cm, respectively) and there was a significant drop in mean HbA1c levels at 1 year in both the groups. These low mean HbA1c values continued for the subsequent 4 years without any significant difference between the two groups. At 5 years, 82% and 87% of patients had complete resolution of T2DM in BPL 150 and BPL 180 group, respectively. This implies, BPL length of 150 cm may be sufficient for co-morbidity resolution and improvement in QoL score, and there may be no added advantage of increasing the limb length. Previous studies on tailored BPL OAGB have shown 70%–85% resolution of T2DM at 1–5 year follow-up.[4232733] While Soong et al.[23] found that 76.1% and 64.2% of patients had complete remission of T2DM with tailored BPL at 1 and 5 years, respectively, Rheinwalt et al.[33] reported that 82.8% and 79% of patients had full remission at 1 and 3 years respectively. Boyle et al.[27] compared results of fixed length BPL of 200 and 150 cm and found that at a mean follow-up of 24.7 months, >80% of patients had improvement in T2DM and about 46% of patients were able to stop medicines in either group. Similarly, Pizza et al.[4] found that in three groups of OAGB patients with different fixed lengths of BPL, percentage of patients with resolution of T2DM at 2 years was similar.

On the basis of our results, we suggest that BPL of 150 cm may be sufficient in all patients undergoing OAGB and there may not be any significant benefits of tailoring the limb lengths. Other researchers have made similar conclusions in the past.[4202634] Hussain et al.[20] suggested for not doing OAGB with >200 cm due to high risk of liver dysfunction. Pizza et al.[4] suggested that BPL limb should be shorter than 200 cm, and concluded that BPL of 150–180 is safe and effective in terms of %EWL, co-morbidity resolution and minimising malnutrition. While Mahawar[34] advised against doing OAGB with BPL >150 cm, Neuberg et al.[26] suggested that 150 cm BPL is adequate for patients with BMI 35–50, and provides good results with minimal nutritional complications. These results however need to be validated by randomised control studies of OAGB with fixed length BPL.

Strengths and limitations

The strength of this article is that the data has been extracted from a prospectively collected database of an RCT with a good long-term follow-up. Furthermore, since the patients were allocated the OAGB procedure randomly, there is no selection bias for the procedure. Limitation however is a small study size and tailored BPL. For true comparison of results with different BPL lengths, it would be ideal to design a randomised trial with different fixed length BPL.

CONCLUSIONS

Although patients with 180 cm BPL show higher mean %TWL than 150 cm BPL, the mean %EWL is similar for the two groups. Since there is no difference between the two groups in the number of patients achieving >50% EWL or >20% TWL, increasing the limb length may not increase the number of good responders for weight loss. While the resolution of T2DM and improvement of QoL score does not change significantly with increase in BPL length from 150 to 180 cm, there is evidence to suggest that mean serum iron levels may be lower with longer BPL. RCT's with fixed length BPL would further confirm these findings.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.