As mentioned in Tables 1 and 3, the patients who had a rise in Amylase values didn't differ from the patients who had normal Amylase values in preoperative variables including demographic data or medical background details, as mentioned in other studies [14, 24], nor in surgery complications, surgery duration, peri-operative complications nor length of stay in the hospital.
As for the laboratory tests, which were found to have significant correlation to the rise in the Amylase, we can make a few hypotheses.
First of all, a rise in Amylase has been found to be significantly associated with a concurrent rise in ALT, AST and GGT above the ULN, without any change in ALK nor conjugated bilirubin, what defines this change as hepatocellular change. All these may point to a liver injury during the procedure, which is able to increase the Amylase values [25–28]. During the operation, a retraction of the liver towards the abdominal wall by a retractor is necessary to allow good visibility of the stomach. Such a minimal liver injury, which expressed only in a little laboratory change, seems not to have any clinical significance.
A rise in Amylase has also been found to be significantly correlated to a rise in Creatinine Phospho-Kinase (CPK). The common reason to a CPK raise is muscle injury, either from the heart, which has been excluded since we didn't experience a concurrent rise in Lactate dehydrogenase (LDH), or from an injury to the abdominal wall muscles from its cutting during the surgery. Such rise may also occur due to acute renal failure, which has been ruled out since there was no concurrent change in Creatinine nor Blood-urea-nitrogen (BUN); It may also happen in an injury to the brain, testicles, retina or inner ear, which has no connection to LSG, and in pancreatitis [29–31]. Such mild pancreatitis may be created from the separation of the stomach from the pancreas during the surgery and it may also happen due to the medications were given during the procedure [32]. Bariatric surgery has already been connected to a higher tendency of the pancreas to produce pancreatitis [8, 33]. Furthermore, a rise in Amylase has been found to be significantly correlated to a rise in Glucose, and such concomitant rise can reflect the a greater body stress in the patients who experienced Amylase rise or also be created from pancreatitis since pancreatitis decreases the amount of secreted insulin [34, 35]. Moreover, a rise in Amylase above ULN has significant association to a decrease in the levels of Calcium and a significant correlation to a decrease in the levels of Phosphorus, which are both known to be correlated to pancreatitis [36–38]. It should be noted that during this study we didn’t measure the level of Lipase, even though its' better sensitivity and specificity of pancreatitis. Nevertheless, such pancreatitis seems to be minimal and with no clinical significance.
In addition, a rise in Amylase has been found to be significantly associated with a concurrent rise in white-blood-cells count (WBC) which expressed via a rise in Neutrophils count and percentage and a decline in the counts and percentage of Lymphocytes. This change may be attributed to a release of inflammatory mediators from either the damaged liver or pancreas as described above.
A rise in Amylase has also been found to be significantly correlated to a rise in Indirect Bilirubin, without a change in Direct Bilirubin. Increasing evidence that indirect bilirubin stimulates Amylase release from the pancreas [39, 40]. As for the origin of this indirect bilirubin raise, we didn't check Haptoglobin, glucose-6-phosphate-dehydrogenase (G6PD) deficiency, Reticulocyte count nor peripheral blood smear, but we didn't reveal any LDH rise, Red-Cell-Distribution-Width (%RDW) rise or Hemoglobin decline, what implies that the probability that the unconjugated bilirubin comes from hemolysis is not very high. We have no good reason to believe that in this study we had a high percentage of patients with syndrome such as Gilbert syndrome that can explain this phenomenon. A better option is that it may come from a hematoma which might happen in LSG to one of the intra-abdominal organs, such as the liver or the pancreas as already been hypothesized above, or an intra-mural hematoma inside the abdominal wall muscles due to their cutting.
A leak may be the most feared complication when observing a peri-operative Amylase rise, since in two thirds of cases it may lead to sepsis and even to patient's death [18, 41], and hence it may be critical to diagnose this complication and treat accordingly. This complication is contributed especially to a long staple line and high intraluminal pressure, and also to ischemia, hematoma formation, and staple misfiring. According to The American Society for Metabolic and Bariatric Surgery's recommendations [42, 43], intraoperative leak may be discovered using endoscopy and/or distention of the anastomosis with dye, air, or other gas. Post-operative suspicion may be proceeding using abdominal CT, with or without a chest CT, or reexploration [42–48]. Despite gathering information about factors that are associated with an increased risk of a leak, and different ways to reduce this risk that have been offered, due to the heterogeneity of the studies and small statistical power there are still no recommendations of ways to prevent leaks [18, 42, 43, 49].
During the study, one complicated event ended with a leak as mentioned above, but it was revealed not because of the high-Amylase the patient had, but due to Systemic inflammatory response syndrome (SIRS) and especially hyperpyrexia, as already noted by other studies [41, 44, 50]. The Amylase in the urine became high but not in the blood nor the drain. It should be mentioned that the value of the Amylase was 147% of ULN which is much less than other hyperamylasurias we observed during the study without a leak. The diagnosis was done finally by gastroscopy.
Hence, although the non-specific hyperamylasuria, other clinical signs may be more helpful in recognition of such a leak. Furthermore, blood Amylase wasn't sensitive at all in this case. Since we did not experience a leak into the peritoneal cavity, we cannot conclude about the useful of drain Amylase in such leaks, but even though a lot of organized fistulas start as an open leak, we didn't observe Amylase raise in the drain before the occurrence of the fistula, what may imply that Amylase in the drain is not sensitive enough for small leaks. Except of this patient, during the study we didn't experience any leak even when there was a high suspicion, for example due to an Amylase rise.
Moreover, the total amount of complications was lower in the high-Amylase group than in the normal-Amylase group. In consequence, Amylase rise is not suitable to be a non-specific sign for complications other than leak.
An increase in Amylase seems not to be specific enough for a leak, since in the vast majority of cases, a rise in Amylase may be associated with hepatic and/or pancreatic manipulation which has no clinical significance. Also, blood Amylase is not sensitive for a leak, and drain Amylase may not be sensitive for small leaks. Even in the appearance of a leak, other methods may be helpful enough for revealing leaks without the need for urine Amylase. Hence, an Amylase test may have a no clinical justification of being taken.
Limitations of this study include a relatively small number of patients for detection of a clear trend in those with leakage, due to the low incidence of leak. Regardless of the fact that our study shows clearly that even though amylase levels were high in more than one fifth of patients, it has no clinical significance in the absence of appropriate clinical context. In addition, since pancreatitis may explain the rise in Amylase, a repeat on such research should include Lipase test and/or urine Amylase-to-Creatinine Clearance-Ratio to confirm or to refute this option.