| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Editorial |
Division of Endocrinology & Metabolism Department of Medicine Mayo Clinic Rochester, Minnesota 55905
Address all correspondence and requests for reprints to: F. John Service, Division of Endocrinology & Metabolism, Department of Medicine, Mayo Clinic, 200 First Street SW, Rochester, Minnesota 55905. E-mail: service.john{at}mayo.edu.
The hypothesized existence of endogenous hyperinsulinemic hypoglycemia as a clinical entity was confirmed 80 yr ago after an extract from the hepatic metastases of an islet cell carcinoma in a patient with recurrent spontaneous hypoglycemia produced hypoglycemia in rabbits (1). Subsequent to the identification of insulinoma as the source of excessive insulin secretion and the realization that food deprivation provoked hypoglycemia, the prolonged fast evolved as the chief diagnostic test for what was then a solely pancreaticocentric view of hypoglycemia.
With the passage of time, reports of patients having mild autonomic symptoms in the postprandial period consistent with those generated by hypoglycemia culminated in the creation of the poorly defined diagnostic entity of reactive hypoglycemia (2) for which the oral glucose tolerance test (OGTT) became established as the diagnostic test. Reactive hypoglycemia was considered to be a functional disorder arising from a mismatch between insulin secretion and glucose entry into the circulation after meals and not due to a structural lesion in the pancreas.
The ensuing classification of hypoglycemic disorders was based on the differentiation between disorders that cause fasting as opposed to postprandial hypoglycemia. The limitations of this classification became apparent with the broadening experience with patients with hypoglycemic disorders predicated on the not infrequent difficulty in determining whether symptoms occur during the fasting state or in the postprandial period and the recognition that some hypoglycemic disorders, e.g. insulinoma and insulin autoimmune hypoglycemia, can cause both fasting and postprandial hypoglycemia.
In addition, the use of the OGTT to document postprandial hypoglycemia ultimately was discredited because at least 10% of healthy persons have a plasma glucose nadir less than 50 mg/dl during this test (3). Furthermore, in persons considered to have reactive hypoglycemia (nadir plasma glucose < 50 mg/dl in an OGTT), there was no electroencephalographic evidence for hypoglycemia (4). Hence, the concept of reactive hypoglycemia as a diagnostic entity was at least in serious doubt if not actually abandoned. These observations led to the proposal of an alternate classification of hypoglycemic disorders (5).
To attribute symptoms of hypoglycemia (which by their nature are nonspecific) correctly to a hypoglycemic disorder, a low plasma glucose concentration must be documented at the time that spontaneous symptoms occur. These symptoms must be relieved through the correction of the low glucose level (<50 mg/dl) (Whipple’s triad) (6). Through the discipline of applying Whipple’s triad to the evaluation of hypoglycemia regardless of whether it occurs in the fasting or postprandial period, recognition of additional disorders causing endogenous hyperinsulinemic hypoglycemia has been possible.
Noninsulinoma pancreatogenous hypoglycemia syndrome (NIPHS), first described in 1999, is characterized by neuroglycopenic symptoms 2 to 4 h postprandially and not in the fasting state (7). Imaging studies of the pancreas (transabdominal, endoscopic, intraoperative ultrasound, and CT imaging) are uniformly negative. However, insulin secretory responses to the selective arterial calcium stimulation test in multiple vascular territories of the arterial supply of the pancreas are positive (8). After gradient-guided partial pancreatectomy, amelioration of symptoms can be expected. Pancreatic tissue from these patients shows evidence of islet hypertrophy and nesidioblastosis (when compared with surgical specimens from age-matched patients with no endocrine disease). In addition, no disease-causing mutation was detected in the Kir6.2 and SUR1 genes, which has been associated with familial persistent hyperinsulinemic hypoglycemia of infancy.
Subsequently, similar clinical presentations and histological findings were noted by us in patients who had undergone Roux-en-Y gastric bypass for medically complicated obesity months or years before (9). This raised the possibility that the bypass procedure contributed to the functional and anatomical abnormalities seen in the pancreata of these patients. Concurrent with the evaluation of the cohort of six patients in the original report (9), the diagnosis of post-gastric bypass hypoglycemia was confirmed in another patient with identical clinical features (not reported in the original series) referred for evaluation. This patient and two others with post-gastric bypass hypoglycemia were the subjects of a later publication by Patti et al. (10). As a result of these reports, this condition has been recognized with increasing frequency nationwide.
In the 18 months following the original description of this condition (9), we have evaluated 37 additional patients with documented endogenous hyperinsulinemic hypoglycemia in the postprandial period who had previously undergone Roux-en-Y gastric bypass surgery. Because of the severity of their symptoms, 23 of these underwent gradient-guided pancreatic resection. About 60% of pancreatic tissue was removed. Most of the resected specimen exhibited islet hypertrophy with nesidioblastosis. However, some also had one or more functioning insulinomas. To date, 21 patients remain symptom-free.
The editorial accompanying the initial report of post-gastric bypass hypoglycemia pointed out that elevated glucagon-like peptide-1 (GLP-1) concentrations may contribute to the abnormal growth and development of the islets (11). High concentrations of GLP-1 are commonly encountered in the immediate postprandial period in patients with upper gastrointestinal surgery including gastric bypass surgery. Although in vitro GLP-1 and GLP-1 receptor agonists have been shown to prevent islet apoptosis and promote islet growth in rodents (12), high dose therapy with a GLP-1 receptor agonist (exenatide) in primates over a 9-month period caused minimal islet hypercellularity and no hypoglycemic symptoms (13).
At this juncture, it is important to note that there has been controversy regarding the pathophysiology of this disorder. Proponents of the view that hypoglycemia is due to the unopposed action (via reduced caloric intake) of premorbid hypertrophic islets characteristic of obesity fail to account for the interval of months to a few years between the gastric bypass surgery and the development of hypoglycemia. This is especially noteworthy given the demonstration that hyperinsulinemia and defective insulin action evident preoperatively are restored to normal shortly after bariatric surgery (14) and by the observation that, in the rare case of a preexisting insulinoma, hypoglycemia is evident immediately after the gastric bypass.
Meier et al. (15) reported no evidence of islet hypertrophy or increased β-cell turnover in the pancreatic tissue of our post-gastric bypass hypoglycemia patients when compared with autopsy specimens from weight-matched controls. However, increased nuclear size was taken to imply β-cell hyperfunction. It has been argued that our use of surgical tissue as a control is inappropriate because of the possibility that the underlying disease, e.g. pancreatic cancer, disturbs islet morphology. On the other hand, the use of autopsy specimens as controls is equally problematic given the rapidity of postmortem changes in the islets leading to swelling and autolysis. This could mask a real difference in islet size.
Uncertainty about the existence of an anatomical cause of post-gastric bypass hypoglycemia (if one can discount the islet cell tumors encountered in some patients with this condition) has given voice to the argument that post-gastric bypass hypoglycemia may represent a functional problem arising in part from excessive GLP-1 secretion after meal ingestion (without necessarily leading to islet hypertrophy). In this issue of the Journal, Goldfine et al. (16) provide evidence that bears on this hypothesis. The investigators measured incretin responses over 2 h to a mixed meal (albeit in liquid form) in patients with and without hypoglycemia after gastric bypass. They report modestly increased GLP-1 in patients with post-gastric bypass surgery hypoglycemia compared with those who have undergone gastric bypass surgery but have not experienced hypoglycemia.
The incretin effect has been demonstrated to potentiate both the static and dynamic β-cell responses to glucose (17). The effect of GLP-1 on insulin secretion has been considered to be glucose-dependent and therefore unlikely to cause hypoglycemia. However, it has been reported that high concentrations of GLP-1 may override this glucose dependency and cause reactive hypoglycemia (18). It is difficult to be sure that the study design which used simultaneous infusion of glucose and GLP-1 is relevant to patients who have undergone gastric bypass. A case report of tumoral secretion of GLP-1 presenting with hypoglycemia failed to account for the probable contribution of somatostatin cosecretion to the pathogenesis of hypoglycemia (19). Another study using an oral glucose load suggested that nadir glucose was associated with the integrated, early (0–30 min) GLP-1 concentrations as well as postmeal suppression of glucagon (20). Although intriguing, these reports are far from definite proof of a direct role of GLP-1 in the pathogenesis of postprandial hypoglycemia.
The caveats regarding the limitations of the OGTT for the diagnosis of hypoglycemia are also likely to apply to the use of a liquid "meal" in patients who have undergone gastric bypass, considering the rapid rate of absorption of liquid calories. Furthermore, measurements of venous concentrations of insulin and c-peptide used in the absence of modeling techniques are qualitative and not quantitative measures of insulin secretion (which is presumed to be abnormal) and action. Although the elevated GLP-1 concentrations reported in this study may be a reasonable explanation for some of the abnormalities seen in affected patients, many questions remain unanswered. Precise measurement of postprandial active and intact GLP-1 concentrations is necessary, as is accurate matching (presurgical weight, postsurgical weight loss, roux en-Y limb length, etc.) of affected and unaffected subjects who participate in such experiments. Elevated GLP-1 may merely be a marker of the rapid rate of meal appearance in the distal intestine. Other considerations include the potential role of vagal efferents in stimulating L cell secretion (21) and therefore contributing to observed differences in GLP-1 secretion depending on the preservation of the vagus at the time of surgery.
In our experience, somatostatin analogs have been an imperfect treatment for this condition, despite their ability to inhibit endogenous secretion of GLP-1 and insulin. Their inhibition of glucagon secretion may be one possible explanation for treatment failure, perhaps illustrating the importance of counterregulation in this setting. Indeed, differences in the secretion of counterregulatory hormones may be another explanation as to why some patients and not others develop hypoglycemia after Roux-en-Y surgery. Acarbose delays absorption of simple carbohydrates, and yet it too has been relatively ineffective in preventing hypoglycemia in most patients. This suggests that rapid caloric absorption may not be the sole explanation for disordered insulin secretion in these situations.
In conclusion, it remains to be ascertained why hyperinsulinemic hypoglycemia develops only in a subset of patients who have undergone gastric bypass surgery. This condition is not simply another manifestation of NIPHS because, unlike NIPHS, it is more common in women and some patients develop insulinomas after gastric bypass surgery. The abnormal islet morphology encountered in affected patients suggests a structural as well as a functional basis for this disorder. Whether these phenomena are mediated primarily via the insulinotropic action of GLP-1 or whether hypersecretion of this peptide is an epiphenomenon has yet to be determined.
Furthermore, it should be emphasized that patients who have undergone gastric bypass surgery may have a myriad of postprandial symptoms, many of which could be incorrectly attributed to hypoglycemia. Symptoms attributable to the "dumping syndrome" usually occur in the early postoperative period, respond to dietary modification, and subsequently resolve. The criteria of Whipple’s triad need to be rigorously applied if a correct diagnosis is to be made and mischaracterization of the defects present in post-bypass hypoglycemia are to be avoided.
Footnotes
A.V. is supported by National Institutes of Health Grant DK078646-01.
Abbreviations: GLP-1, Glucagon-like peptide-1; NIPHS, noninsulinoma pancreatogenous hypoglycemia syndrome; OGTT, oral glucose tolerance test.
Received October 9, 2007.
Accepted October 17, 2007.
References
This article has been cited by other articles:
 |
|
 |
|
  M. D. Marcus, M. A. Kalarchian, and A. P. Courcoulas Psychiatric Evaluation and Follow-Up of Bariatric Surgery Patients Am J Psychiatry, March 1, 2009; 166(3): 285 - 291. [Full Text] [PDF]
|
|
|
|
 |
|
 P. E. Cryer, L. Axelrod, A. B. Grossman, S. R. Heller, V. M. Montori, E. R. Seaquist, and F. J. Service Evaluation and Management of Adult Hypoglycemic Disorders: An Endocrine Society Clinical Practice Guideline J. Clin. Endocrinol. Metab., March 1, 2009; 94(3): 709 - 728. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Vidal, J. Nicolau, F. Romero, R. Casamitjana, D. Momblan, I. Conget, R. Morinigo, and A. M. Lacy Long-Term Effects of Roux-en-Y Gastric Bypass Surgery on Plasma Glucagon-Like Peptide-1 and Islet Function in Morbidly Obese Subjects J. Clin. Endocrinol. Metab., March 1, 2009; 94(3): 884 - 891. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
