This case series looked at three middle-aged women (median age 42 years; range 39–51 years) suffering from super obesity (median body mass index (BMI) 50 kg/m
2; range 49–53 kg/m
2) (Table
1). Patients B and C had similar co-morbidity profiles, except for patient B having had previous abdominal surgery, and patient C had hypertension (HTN). Patient A, however, had obstructive sleep apnea (OSA), hypertension, recurrent incisional hernia, and previous abdominal surgery with a mesh. None of the patients suffered from diabetes or hiatus hernia (Table
1). The median operative time was 132 min (range 120–195 min) (Table
1). Post-operatively, all patients stayed for 2 days in the hospital and there was no leakage on the routine post-operative (30–48 h) CT (Table
1). Median excess weight loss (EWL) at 6 months was 39% (range 32–43%) and at 1 year 57.7% (range 50.4–63%). A delayed leakage (7–10 days post-operatively)—which could be attributed to heat-related sloughing of the gastric wall—was excluded on a clinical basis during the initial follow-up period that was extended to 30-days post-operatively.
Table 1
Demographics and peri-operative characteristics of three patients who underwent stapleless laparoscopic sleeve gastrectomy
Characteristic |
Age (years) | 39 | 51 | 42 |
Body mass index (BMI) | 52.67 | 50.31 | 49.31 |
Gender (f = 0; m = 1) | 0 | 0 | 0 |
Pre-op risk factor |
Diabetes (no = 0; yes = 1) | 0 | 0 | 0 |
Obstructive sleep apnea (no = 0; yes = 1) | 1 | 0 | 0 |
Hypertension (no = 0; yes = 1) | 1 | 0 | 1 |
Reflux (no = 0; yes = 1) | 0 | 0 | 0 |
Previous abdominal surgery (no = 0; yes = 1) | 1 | 1 | 0 |
Current hernia (no = 0; yes = 1) | 1 | 0 | 0 |
Hiatal hernia (no = 0; yes = 1) | 0 | 0 | 0 |
Peri-op technical consideration |
Instrument | Harmonic | Harmonic | Harmonic |
Number of layers | 2 | 2 | 2 |
Fashion | Stationed/continuous | Stationed/continuous | Stationed/continuous |
Post-op CT (no = 0; yes = 1) | 1 | 1 | 1 |
Main outcomes measured |
Operative time (minutes) | 195 | 120 | 132 |
Leak (no = 0; yes = 1) | 0 | 0 | 0 |
Hospital stay (days) | 2 | 2 | 2 |
BMI 6 months (EWL%) | 36.39 (51%) | 37.56 (36%) | 34.33 (48%) |
BMI 1 year (EWL%) | 34.2 (71%) | 29 (53%) | 33.9* (59%) |
All patients had some form of post-operative food intolerance. The commonest symptom among all patients was post-operative anorexia. All patients described nausea secondary to food odor; additionally, all patients experienced vomitus at least once post-operatively.
Discussion
Stapleless LSG is a potentially safe and efficient alternative to the standard stapled LSG. Stapleless LSG has not been widely discussed in literature. However, a recent comment on stapleless laparoscopic Roux-en-Y gastric bypass (LRYGBP), exemplifies some of the hypothetical critiques of stapleless LSG, stating that “
the potential advantage of avoiding leaks and fistulas may not be consistent overall,” describing it as “
cumbersome and time-consuming,” and finally that “
the reductions in cost may not be worth it if complication rates increase” [
9]. Given the dearth of literature looking specifically at stapleless LSG [
10,
11], we will look at LSG with oversewing as a proxy. We will place the hypothetical critiques in the context of the broader literature and address them one by one.
Some have doubted whether SLR by oversewing does carry any advantages in leak rates [
12], bleeding rates [
13], or both [
14], compared to LSG alone. They claim that some studies do not show statistically significant results, or that oversewing might increase bleeding rates for example. Firstly, this is contested by studies which point to a clear reduction in leak rates [
15], bleeding rates [
16], and both [
17,
18]. However, more importantly, the problem with many of the studies that cast doubt over SLR by oversewing is that they lack statistical power [
18]. This is in part because leak and bleeding rates are currently quite low to start with, especially once the learning curve has been overcome. Additionally, Chen et al. remarked that a large sample size (around 9346 procedures) would be needed to detect relatively significant differences in leak rate [
19]. None of the studies reviewed critiquing oversewing met this threshold. Whereas, the one meta-analysis that did meet that sample size threshold (Shikora and Mahoney 2015), showed a statistically significant link between oversewing and reduced leak and bleeding rates [
18].
Some might suggest that a stapleless method has increased risk of leaks and hemorrhage as the surgeon needs to open the stomach and then close it with sutures. This was alluded to by Póvoas and Vilas-Bôas (2006) when describing the complication that arose following their attempt at Stapleless RYGBP [
9]. Again, no literature could be found comparing a hand-sewn versus a stapled technique for closure of the stomach in LSG. Therefore, we used evidence surrounding gastrojejunostomy anastomosis creation as a proxy for this. Most of the studies reviewed compared hand-sewn anastomosis (HSA) with circular-stapled anastomosis (CSA) and linear-stapled anastomosis (LSA). CSA was ignored as these are not used in LSG.
Initial studies were inconclusive. Some are suggesting that HSA has a higher stenosis rate [
20], whereas others showed lower rates [
21]. Some these studies employed three different procedures early in the learning curve, which may have confounded the results. More recent evidence has demonstrated no increased risk from hand-sewn versus linear-stapled anastomosis with regards to leaks, bleeding, or stenosis [
20,
22]. Extrapolating from this, the evidence seems to suggest that hand-sewn gastric closure would be safe, though studies comparing this with LSA directly will be required.
Oversewing in SLR does increase the operative time when compared with LSG alone [
12]. Though this is trivially true, in so far as doing something will always take longer than not doing something, it is also true that oversewing does extend operative time, even when compared to other reinforcement methods [
23]. This was also true of our series, where the median operative time was 132 min (range 120–195 min), with patient A, the most obese and co-morbid, taking the longest, 195 min (Table
1). However, historically, surgery had often opted for more time-consuming techniques when benefits outweighed this [
24]. Moreover, it is unclear how big a problem this is in practice as one group reported comparable operative times with and without oversewing [
15,
18]. Finally, this factor can be easily reduced with experience, especially in high-volume centers [
18]. The same meta-analysis also showed that buttressing produced slightly better outcomes than oversewing. Unfortunately, these slight benefits need to be weighed against costs. Among different techniques for reinforcement, oversewing is seen to be the most affordable one [
25]. Some have reported that for large institutional hospitals, buttressing materials (e.g., absorbable polymer membranes, bovine pericardial strips, or fibrin sealants) are too expensive for use on a permanent basis; therefore, suturing is the best option regarding costs and benefits [
18]. This is particularly the case in countries with limited income per capita, as highlighted by Ettinger et al. using Brazil as an example. They explain that because materials are imported incurring taxes and currency exchange rates, the final cost is expensive compared with other methods. They illustrate this with the following example of the cost of materials for stapled RYGBP: the stapler for the laparoscopic RYGBP in Salvador, Brazil, is R$3220 (Brazilian Reals) = US$1340 (US dollars). A laparoscopic cartridge costs R$1260 = US$525. The total cost per operation using one stapler plus seven cartridges is R$12,040 = US$5016 [
26].
When these factors are taken together, a hand-sewn technique becomes not only better than LSG alone regarding outcomes but is also likely to be more cost-effective, when compared to LSG with no reinforcement and LSG with buttressing.
Unsurprisingly during the early post-operative period, most of the patients had symptoms related to food intolerance. However, the clinical impression is that this was not different from the standard sleeve gastrectomy [
27]. It is, of course, obvious that no reliable statistics could be generated from this small sample size. In future, research could be carried out using validated questionnaires [
28] to explore whether stapleless LSG has any effect on food tolerance.
A substantial modification of this method is the addition of invaginating sutures over underrunning continuous sutures. Reinforcement with invaginating sutures has been shown to have reduced the leak rate from around 7 to 0% in one our previous study [
8]. The study by Rogula et al. (2015) [
25] showed a decrease in leak rate by 70 to 0% in vitro, when comparing imbricating sutures and continuous through-and-through sutures [
24].
We propose that a single seromuscular (extra mucosal) layer could be sufficient; however, we have yet to test this hypothesis. The purpose of the two-layer suturing adopted here is to protect against energy leakage from energy induced sloughing of the gastric wall. The lateral thermal spread on the gastric wall from the harmonic is not known to the best of our knowledge. However, one of the advertising proclamations of the harmonic is the minimal lateral spread [
29]. It is worth remembering that the thermal lateral spread is theoretically larger than what should be on the stomach, such as porcine carotid arteries [
30] and bovine muscle fascia [
29]. Thus, given the hemostatic nature of the instrument, we think that the full thickness layer—taken for hemostasis mainly—can be dispensed. This is also an analogy to the bowel anastomosis in which single layer is proved to be as effective as two layers [
31].
In our limited case series, we routinely performed CT scan with oral contrast to detect any leak. We acknowledge that the cost of this extra imaging increases the cost of our stapleless technique; however, we felt that it was warranted in our initial series to be skeptical. Although our patients were closely monitored for clinical signs of early leak such as tachycardia, fever, and unequal inspiration [
7], we note that these signs are not specific to leakage and could indicate atelectasis, which is recognized as the most common complication in bariatric patients [
32]. Dehydration also causes low-grade pyrexia and tachycardia and could have confounded the diagnosis of an early leak. We found our patients were vulnerable to dehydration, as post-operative nausea and vomiting was a frequent event. An additional reason to perform CT scans in our study was to assess the geometry of the pouch.