Insulinoma: Cost-effective Care in Patients with a Rare Disease
L. Axelrod
DOI: https://doi.org/10.7326/0003-4819-123-4-199508150-00011
IF: 39.2
1995-08-15
Annals of Internal Medicine
Abstract:An insulinoma is the quintessential example of the elusive but curable endocrine disorder. It is a rare condition, estimated to occur four times per million person-years [1], and has nonspecific neurologic and adrenergic manifestations. It frequently eludes diagnosis for years. The neurologic manifestations, especially those that affect behavior and consciousness, are often frightening and embarrassing to those who have this condition and to those around them. An insulinoma is potentially lethal. Once identified, it is curable by surgery [1]. When an isolated benign insulinoma is removed (and most are benign), the patient has a normal life expectancy [1]. A patient with an insulinoma presents a series of challenges. The physician must recognize that the patient's symptoms are attributable to hypoglycemia, exclude other exogenous and endogenous causes of hypoglycemia, obtain appropriate hormonal studies to establish the diagnosis beyond any doubt, and guide the patient toward appropriate surgical treatment. Our ability to diagnose and treat patients with insulinomas has improved markedly since the disorder was initially described in 1927 [2] and since Allen O. Whipple proposed the criteria now known as the Whipple triad: 1) the occurrence of attacks attributable to hypoglycemia, especially but not exclusively in the fasting state or after exertion; 2) measurement of a low circulating glucose level when the patient has symptoms [spontaneously or after a fast]; and 3) alleviation of the patient's symptoms and signs by the administration of glucose [3, 4]. The availability of methods to measure accurately circulating levels of insulin, C-peptide, and proinsulin has dramatically improved the ability of the physician to diagnose an insulinoma preoperatively with a high degree of confidence and to exclude the surreptitious administration of insulin as the cause of hypoglycemia [5-7]. In fact, the diagnosis should be established before surgical exploration is done; this is in contrast to the situation that prevailed before sensitive and specific radioimmunoassays for insulin were available. In an earlier era, the presence of the Whipple triad led directly to surgery; thus, confirmation of the diagnosis, localization, and treatment occurred together. Because insulinomas are now diagnosed preoperatively, localization of an insulinoma is now approached as a separate step after diagnosis. Perhaps because of the difficulties encountered earlier in this century, many physicians have sought ways to localize this characteristically small tumor preoperatively. However, the assumption that the preoperative localization of an insulinoma improves the outcome of surgery remains unproved [8]. This assumption may have been rendered irrelevant by improvements in surgical technique and by the development of intraoperative ultrasonography for localization, for which published reports provide evidence of remarkable sensitivity [9, 10]. Intraoperative ultrasonography by itself detects 84% to 86% of solitary insulinomas [11]. When combined with palpation by a surgeon skilled in this operation, the overall sensitivity is 95% to 100% [9, 10]. For any preoperative method of localization to be of value, it should improve on this 95% to 100% success rate [9, 10]. In this issue, Doppman and colleagues [12] describe a procedure for the preoperative localization of insulinomas that is based on the effects of the intra-arterial injection of calcium [12]. They injected calcium into the gastroduodenal, splenic, and superior mesenteric arteries of 25 patients. Blood samples were obtained from the right and left hepatic veins. Results were compared with those of ultrasonography, computed tomography, magnetic resonance imaging, arteriography, and portal venous sampling. The authors found that calcium stimulation localized the tumor in 22 of 25 insulinomas and concluded that this procedure is the most sensitive preoperative localizing test for insulinomas. Their work represents an intelligent application of a physiologic principle and a superb technical achievement. But should this test now become part of the preoperative management of every patient with clinical and hormonal evidence of an insulinoma? Based on all of the available information, the answer must be no. First of all, the methods used do not precisely localize the lesions. The results indicate only whether the insulinoma is located to the right (pancreatic head and neck) or to the left (pancreatic body and tail) of the superior mesenteric artery. The authors claim that this has surgical relevance because adenomas to the left of the superior mesenteric artery can be treated by enucleation and distal pancreatectomy, whereas those to the right must be localized for enucleation. However, it is not apparent that this localization served that purpose for any patient in the series. Actual localization appears to have been achieved by the use of intraoperative ultrasonography and exploration by a skilled surgeon. Additionally, the comparison between the present procedure and the five methods of localization studied is not relevant to the current clinical context. None of these methods has sufficient sensitivity to compete with intraoperative ultrasonography and surgical exploration [9-12]. In the recent experience of Doppman and colleagues [12], conventional ultrasonography detected 9% of insulinomas, computed tomography detected 17%, magnetic resonance imaging detected 43%, arteriography detected 36%, and percutaneous transhepatic portal venous sampling detected 67%. Endoscopic ultrasonography has a sensitivity of approximately 80% [13]. Even when an insulinoma is localized preoperatively, it is not clear that the patient benefits, because it is possible that preoperative studies identify those tumors that are readily found at surgery and miss those that elude the surgeon's fingers. The appropriate comparison would be with intraoperative ultrasonography and careful pancreatic exploration by a skilled surgeon. There is no evidence to show that the proposed approach improves on the results obtained in this way. The authors previously reported [10] that intraoperative ultrasonography is essential for the intraoperative detection of insulinomas. In their present article [12], Intraoperative ultrasonography was done in each patient to visualize the tumor, to identify major pancreatic and biliary ducts adjacent to the tumor, and to direct the pancreatic incision for enucleation. The authors also state that the availability of experienced intraoperative ultrasonography is critical in patients with tumors located to the right of the superior mesenteric artery because adenomas smaller than 15 mm may be difficult to palpate. The conclusion of Doppman and colleagues that calcium stimulation with hepatic venous sampling. should replace portal venous sampling as a technique to localize occult insulinomas is reasonable. The new procedure appears to be the most sensitive one available for the preoperative localization of insulinomas, subject to confirmation in other centers. But the development of this new procedure should not obscure the fact that the need for preoperative localization may be infrequent. One could easily argue that neither procedure is routinely indicated. One might argue that calcium stimulation with hepatic venous sampling should be used in the occasional patient (perhaps in 5% of patients) in whom an insulinoma cannot be localized by intraoperative ultrasonography and careful exploration by a surgeon skilled in the detection of such tumors. Because insulinomas are rare and because the outcome of surgery depends on the availability of intraoperative ultrasonography and of a surgeon with the appropriate training and experience, patients should be directed to an appropriate referral center after the diagnosis has been established. The recommendation of Doppman and colleagues that the use of intra-arterial calcium injection will facilitate the management of patients with an insulinoma in community hospitals is a worrisome extrapolation from experience at a referral center and fails to consider the cumulative learning curve that applies to the performance of invasive radiologic procedures and surgery. Ultimately, the authors have not shown that their intervention had any effect on outcome. The intervention certainly increases cost and morbidity. In an age when cost-effectiveness is always an issue and outcome analysis is deemed essential, it is no longer acceptable to advocate or use costly invasive procedures for minimal (if any) marginal advantage. Until now, the health care debate has focused largely on the cost-effectiveness of commonly used procedures, such as cardiac catheterization, coronary angioplasty, and endoscopic gastroenterologic procedures. Doppman and colleagues' study reminds us that management strategies for rare disorders also must be judged on the basis of cost-effectiveness. When we treat rare disorders, we must address the same concerns about cost-effectiveness and outcome analysis that we address when we treat common ones.