Nonautonomous Function of a Pancreatic Insulinoma ELLIOT J. RAYFIELD, MARIE PULINI, AVRUM GOLUB, ARTHUR H. RUBENSTEIN, AND DAVID L. HORWITZ Departments of Medicine and Pathology, Mount Sinai School of Medicine of the City University of New York, Neio York, and the Department of Medicine, University of Chicago School of Medicine, Chicago, Illinois ABSTRACT. A 56-year-old woman with symptoms of weakness, visual blurring, and sweating underwent diagnostic studies to evaluate the etiology of her hypoglycemia. Fasting hypoglycemia was never documented; in diagnostic studies performed during her two hospitalizations and several outpatient glucose tolerance tests, the lowest fasting plasma glucose recorded was 56 mg/dl. The patient displayed exaggerated plasma insulin responses following oral glucose (peak response: 447 /xU/ml at 30 min) and following 1 gm of iv tolbutamide (peak response: 719 /xU/ml at 5 min) with symptomatic profound hypoglycemia during both tests. Basal per cent proinsulin was elevated at 49% (normal range 5-22%).
W
HILE provocative tests of insulin secretion, such as iv tolbutamide or glucagon, may aid in the diagnosis of an insulinoma, false positive and false negative results are not uncommon (1). The most reliable means of documenting inappropriate insulin release in a patient with an insulinoma has been during a supervised fast (2). Recently, Turner and Harris reported that, following the injection of fish insulin, suppression of endogenous insulin secretion is either completely or partially lost in insulinomas (3). This report describes a patient with a pancreatic beta cell tumor who did not become significantly hypogly-
Throughout a 72 h fast, values for plasma glucose, insulin, and glucose/insulin ratios were all within the normal range. During the infusion of exogenous insulin (0.1 U/kg for 60 min) serum C-peptide reactivity suppressed to less than 1.3 ng/ml when the plasma glucose fell below 40 mg/dl representing normal suppression. At surgery, a pancreatic beta cell adenoma was found and removed. This patient represents the uncommon circumstance in which stimulation tests with tolbutamide and glucose were more helpful in establishing a preoperative diagnosis than were the suppression tests. (J Clin Endocrinol Metab 43: 1307, 1976)
cemic during a 72 h fast and who demonstrated almost complete suppression of C-peptide reactivity (CPR) following an infusion of exogenous insulin. 1 Case Report
A 56-year-old woman was referred to the Mount Sinai Medical Center in March 1975 because of increasing frequency of symptoms of weakness, visual blurring, and sweating of 7 years duration. The attacks were not triggered by fasting, excercise or periods of emotional upset and occurred at any time of the day. The patient was non-obese (height 125 cm, weight 40.5 kg) and there was no weight gain during this period. Past medical history revealed no family history of diabetes mellitus; she had a one year history Received April 29, 1976. of hypertension which had been treated with Supported in part by research grants from the Eli hydrochlorthiazide 50 mg daily and diazepam Lilly Co., The American Diabetes Association and the 5 mg three times daily. Physical examination NIH (AM 18522), an NIH Research Career Developwas negative except for a right sided \xh x 1 ment Award KO 4-AM00089 (Dr. Rayfield); NIH Grant RR-71, Division of Research Resources, Clinical Re- cm thyroid nodule; there were no neuromuscular search Center, Mount Sinai School of Medicine; abnormalities. The admission laboratory data inAmerican Cancer Society Institutional Grant No. cluded a normal blood count, urinalysis, serum IN-410 and a Research and Development Award of the American Diabetes Association (Dr. Horwitz) and a 1 A preliminary report of this study has appeared PHS Diabetes Center Grant AM-17-46 (Dr. Rubenas an abstract. Rayfield, E. J., M. Pulini, A. Golub, A. H. stein). Rubenstein, and D. Horwitz, Nonautonomous funcReprint requests to: Dr. Elliot J. Rayfield, Mount tion of a pancreatic insulinoma (p.i.) Clin Res 25: Sinai School of Medicine, Fifth Avenue and 100th 593A, 1975. Street, New York, N.Y. 10029. 1307
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 22:09 For personal use only. No other uses without permission. . All rights reserved.
JCE & M • 1976 Vol 43 • No 6
RAYFIELD ET AL.
1308
TABLE 1. Carbohydrate tolerance tests Tolbutamide test
Glucose tolerance test
IRIt
Time (min.)
Glucose mg/dl
-15 0 5 15 30 45 60 90 120 180 240
66 63 — 78 52 32 34 46 63 49
192 447 110 49 20 8 99 13
300
65
10
51
/xU/ml
Proinsulin* ng/ml
ProinsulinJ %
0.40
49.1 — 1.5 — — 16.8
11 4
—
0.48 — — 1.59 0.59 — 0.81 —
36.7 30.1 —
Glucagon pg/ml
Glucose mg/dl
68 92
68 68 64 42 12 14 24 29 39 51
91 61 80 80 80 80 93
—
73
IRIt /iU/ml 5 6 719 588 184 42 27 12 8 9 — —
* Read from a human proinsulin standard after gel filtration. f IRI = Immunoreactive Insulin. \ Read from an insulin standard.
electrolytes, blood urea nitrogen, serum creatinine, calcium, phosphorus, alkaline phosphatase, uric acid, bilirubin, cholesterol, thyroxine and triiodothyronine. Chest x-ray and EKG were normal. The patient underwent the following diagnostic studies: oral glucose tolerance test, tolbutamide test, three day fast and insulin suppression test. Fasting hypoglycemia was never documented during her two hospitalizations nor on 3 glucose tolerance tests performed as an outpatient by her private physician. The lowest fasting plasma glucose recorded was 56 mg/dl. The patient had symptoms identical to those she described at home during the hypoglycemic phase of the tolbutamide test, the oral glucose tolerance test, and the insulin suppression test. A celiac and superior mesenteric angiogram was negative. At exploratory laparotomy, a 3 x 2 cm adenoma projecting from the inferior border of the pancreas at the junction of the body and tail was removed. On histologic examination the tumor was composed of cells arranged in isletlike structures containing many beta cell granules. There was no evidence of metastases. Six months post-operatively, the patient had no recurrence of hypoglycemic episodes. An oral glucose tolerance test was normal with respect to both insulin and glucose levels. Materials and Methods Plasma glucose was measured by the glucoseoxidase method with the Technicon Auto Analyzer. Serum immunoreactive insulin (IRI) (4,5),
proinsulin (6), percentage proinsulin, C-peptide reactivity (CPR) (7), and plasma glucagon (8) were measured by radioimmunoassay procedures. Acid-ethanol extracts of the pancreatic islet cell tumor were assayed for insulin by radioimmunoassay (5). Informed consent was obtained for all the studies. Results 1. Glucose tolerance test (Table 1) After 100 g oral glucose, plasma glucose fell from 63 mg/dl at 45 min, and returned to baseline levels by 180 min. Although the baseline serum IRI values at - 1 5 and 0 min fell within the normal range (5-20 ju-U/ml), a greatly exaggerated insulin concentration occurred following glucose administration with a peak value of 447 /uU/ml, at 30 min. Serum proinsulin levels rose from a fasting level of 0.40 ng/ml read against a human proinsulin standard to a peak of 1.59 ng/ml at 60 min. The contribution of proinsulin to the total immunoreactive insulin concentration in the basal state was clearly elevated at 49% (normal range 5-22%). The proinsulin per cent reached a nadir of 1.5% 15 min following glucose and then increased to 36% at 120 min (Table 1). Basal plasma glucagon levels were within the normal range (68-92 pg/
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 22:09 For personal use only. No other uses without permission. . All rights reserved.
NONAUTONOMOUS PANCREATIC INSULINOMA ml), diminishing to 61 pg/ml at 30 min and returning to baseline at 45 min.
3.0
1
1
1
1309
I
I
2.5 --
o
2. Tolbutamide test Following 1.0 g tolbutamide iv, plasma glucose levels decreased from 68 mg/dl (time 0) to 12 mg/dl at 30 min with sustained hypoglycemia (Table 1). Serum IRI values increased markedly from 6 /x,U/ml (time 0) to 719 /-lU/ml at 5 min and 588 /x.U/ml 15 min following tolbutamide. 3. Insulin suppression test Regular insulin was infused at a rate of 0.1 U/h for 60 min using a Harvard infusion pump. During the infusion of exogenous insulin, serum CPR was measured to monitor endogenous insulin secretion. C-peptide reactivity suppressed normally during the first 40 min of the test (Fig. 1), and thereafter increased slightly above this range to levels of 1.13 and 1.35 ng/ml at 50 and 60 min, respectively. 4. 72 hour fast Plasma glucose values remained consistently above 50 mg/dl. IRI values varied between 6.5 and 2.5 /aU/ml and glucose/ IRI ratios ranged between 9.1 to 23 (normal range 4-14) (9). At the end of the 72 h fast, plasma glucose was 56 mg/dl, IRI 8 /tU/ml and the percentage proinsulin was 35. The true insulin level (following column fractionation with separation of proinsulin) was 4.9 /uU/ml and the glucose/IRI ratio was 7.0. 5. Tumor insulin content The tumor was extracted for immmunoreactive insulin and showed 23.8 U/gram. Normal pancreas values ranged from 0.5 to 2 U/g, while tumors vary from very high to very low values (10). This level is compatible with an insulinoma. Discussion Insulin suppression tests (prolonged fast or infusion of exogenous insulin) are con-
/
£2.0 1.5 h 1.0 0.8
60
/
^
/
:
I
I
I
-
I
20 30 40 50 60 70 80 90 100 PLASMA GLUCOSE (mq/IOOml)
FIG. 1. Relationship between plasma glucose and Cpeptide immunoreactivity during the 60 min insulin infusion. The hatched area depicts the normal range at each blood sugar concentration in 7 healthy non diabetic volunteers as previously determined (15); the closed circles connected by broken lines indicate the values for the insulinoma patient and the time (in minutes) at which each sample was taken is given next to each circle.
sidered to be more reliable than stimulation tests in the diagnosis of insulinomas (2,11). In the largest available series of surgically documented beta cell tumors (12), only 2 of 108 patients failed to manifest a "typical hypoglycemic attack" by the end of a prolonged fast lasting as long as 72 h. In the 106 patients who became hypoglycemic, the blood sugar was lower than 50 mg/dl by the methods of Folin-Wu (normal range 80-120 mg/dl) or Somogyi (normal range 65-90 mg/dl) (12). In this regard, Merimee and Tyson (13) have documented a sex difference in plasma glucose homeostasis, in that plasma glucose concentrations of as low as 35 mg/dl occurred after one day of fasting in healthy women, in contrast to 55 mg/dl in men. Thus, during a fast, concomitant glucose and insulin levels are necessary to determine whether an abnormality is present. However, the 72 h values (at the completion of the fast) for plasma glucose, insulin, and glucose/insulin ratios were all within the normal range in the patient described in this report. An alternative procedure has recently been developed in which the suppression of endogenous insulin secretion is assessed by measuring CPR (14,15) levels during the
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 22:09 For personal use only. No other uses without permission. . All rights reserved.
1310
JCE & M • 1976 Vol 43 • No 6
RAYFIELD ET AL.
infusion of porcine insulin. During the infusion of exogenous insulin (Fig. 1), healthy subjects, as well as patients with hypoglycemia due to causes other than insulinoma, suppress CPR levels to less than 1.3 ng/ml when the plasma glucose falls below 40 mg/dl. Furthermore, CPR levels remain below 1.3 ng/dl during the remainder of the insulin infusion (15). The response of our patient is unusual in that, unlike 8 other patients with proven insulinomas (16), CPR levels suppressed below 1.3 ng/ml when her plasma glucose fell below 40 mg/dl. Unlike control subjects, however, CPR levels did not remain below 1.3 ng/ml for the entire insulin infusion. The patient described in this report thus represents an uncommon circumstance in which stimulation tests with tolbutamide and glucose were more helpful in establishing a preoperative diagnosis than were the suppression tests. There are a few cases in the literature of atypical patterns of insulin secretion in pancreatic beta cell tumors. Harano et al. (17) described a patient with a surgically documented pancreatic insulinoma who had blunted insulin responses following tolbutamide and glucagon, but an exaggerated insulin response following secretin administration. Our patient's exaggerated insulin response following oral glucose, while rare, has been previously noted by Power (18) in a patient who also had normal glucose values during a 72 h fast. While the mechanism for the marked insulin responses following glucose remains speculative, it is possible that this tumor retained the capacity to respond normally to both decreases and increases in plasma glucose. While the response to each of the tumor cells was quantitatively the same as "normal beta cells" the overall release of insulin was quantitatively larger because of the larger number of beta cells in the tumor. The test responses presented here illustrate that pancreatic beta cell tumors like other endocrine tumors may exhibit a spectrum from autonomous to nonautonomous function (19). Finally, it should
be noted that the elevated percentage of proinsulin (49%) in the basal state in the face of normal insulin levels (Table 1) was most helpful in establishing that a beta cell tumor was the cause of the patient's symptoms (20). References 1. Fajans, S. S., In Wintrobe, M. M, G. W. Thorn, R. D. Adams, E. Braunwald, K. Isselbacher, and R. G. Petersdorf (eds.), Harrison's Principles of Internal Medicine, ed. 7, McGraw-Hill Book Company, New York, 1974, p. 554. 2. Schein, P. S., R. A. DeLellis, C. R. Kahn, P. Gorden, and A. R. Kraft, Islet cell tumors: Current concepts and management, Ann Intern Med 79: 239, 1973. 3. Turner, R. C , and E. Harris, Diagnosis of insulinomas by suppression tests, Lancet 2: 188, 1974. 4. Juan, C , and T. W. AvRuskin, A combined immunoassay of human growth hormone and insulin: Cumulative assessment of assay performance,/ Clin Endocrinol Metab 33: 150, 1971. 5. Morgan, C. R., and A. Lazarow, Immunoassay of insulin: Two antibody system: Plasma insulin levels of normal, subdiabetic, and diabetic rats, Diabetes 12: 115, 1963. 6. Melani, F., A. H. Rubenstein, and D. F. Steiner, Human serum proinsulin, J Clin Invest 49: 497, 1970. 7. Block, M. B., M. E. Mako, D. F. Steiner, and A. H. Rubenstein, Circulating C-peptide immunoreactivity: Studies in normals and diabetic patients, Diabetes 21: 1913, 1972. 8. Aquliar-Parada, E., A. M. Einsentraut, and R. H. Unger, Pancreatic glucagon secretion in normal and diabetic subjects, Am J Med Sci 257: 415, 1969. 9. Grunt, J. R., J. A. Pallotta, and J. S. Soeldner, Blood sugar, serum insulin and free fatty acid relationships during intravenous tolbutamide testing in normal young adults, and in patients with insulinoma, Diabetes 19: 122, 1970. 10. Creutzfeldt, W., R. Arnold, C. Creutzfeldt, U. Deuticke, H. Frerichs, and N. S. Track, Biochemical and Morphological Investigation of 30 Human Insulinomas: Correlation Between the Tumor Content of Insulin and Proinsulin-like Components and the Histological and Ultrastructural Appearance, Diabetologia 9: 217, 1973. 11. Editorial, Diagnosis of insulinoma, Lancet 2: 385, 1974. 12. Laroche, G. P., D. O. Ferris, J. T. Priestly, D. A. Scholz, and M. B. Dockerty, Hyperinsulinism, Arch Sur 96: 763, 1968. 13. Merimee, T. J., and J. E. Tyson, Stabilization of
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 22:09 For personal use only. No other uses without permission. . All rights reserved.
NONAUTONOMOUS PANCREATIC INSULINOMA plasma glucose during fasting: Normal variations in two separate studies, N Engl J Med 291: 1275, 1974. 14. Horwitz, D. L., and A. H. Rubenstein, Insulin suppression, Lancet 2: 1021, 1974. 15. Horwitz, D. L., A. H. Rubenstein, C. Reynolds, G. D. Molnar, and N. Yanaihara, Prolonged suppression of insulin release by insulin-induced hypoglycemia: Demonstration by C-peptide assay, Honn Metab Res 7: 449, 1975. 16. Service, F. J., D. L. Horwitz, H. Kuzuya, M. Mako, C. Reynolds, G. D. Molnar, and A. H. Rubenstein, Diagnosis of insulinomas by C-peptide measurement during hypoglycemia induced by insulin infusion, The Endocrine Society 58th Annual Meeting Program and Abstracts: 66, 1976 (Abstract).
1311
17. Harano, Y., M. Kohama, N. Araki, A. Sakamoto, M. Hoshi, M. Shichiri, K. Shima, J. Okamura, and Y. Shigeta, Qualitative abnormality of insulin secretion in a case with insulinoma, Endocrinol Jap 22: 175, 1975. 18. Power, L., A. glucose-responsive insulinoma, JAMA 207: 893, 1969. 19. Rayfield, E. J., L. I. Rose, J. P. Cain, R. G. Dluhy, and G. H. Williams, ACTH-responsive, dexamethasone-suppressible adrenocortical carcinoma, N Engl J Med 284: 591, 1971. 20. Alsever, R. N., J. P. Roberts, J. G. Gerber, M. E. Mako, and A. H. Rubenstein, Insulinoma with low circulating insulin levels: The diagnostic value of proinsulin measurements, Ann Intern Med 82: 347, 1975.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 22:09 For personal use only. No other uses without permission. . All rights reserved.
W
HILE provocative tests of insulin secretion, such as iv tolbutamide or glucagon, may aid in the diagnosis of an insulinoma, false positive and false negative results are not uncommon (1). The most reliable means of documenting inappropriate insulin release in a patient with an insulinoma has been during a supervised fast (2). Recently, Turner and Harris reported that, following the injection of fish insulin, suppression of endogenous insulin secretion is either completely or partially lost in insulinomas (3). This report describes a patient with a pancreatic beta cell tumor who did not become significantly hypogly-
Throughout a 72 h fast, values for plasma glucose, insulin, and glucose/insulin ratios were all within the normal range. During the infusion of exogenous insulin (0.1 U/kg for 60 min) serum C-peptide reactivity suppressed to less than 1.3 ng/ml when the plasma glucose fell below 40 mg/dl representing normal suppression. At surgery, a pancreatic beta cell adenoma was found and removed. This patient represents the uncommon circumstance in which stimulation tests with tolbutamide and glucose were more helpful in establishing a preoperative diagnosis than were the suppression tests. (J Clin Endocrinol Metab 43: 1307, 1976)
cemic during a 72 h fast and who demonstrated almost complete suppression of C-peptide reactivity (CPR) following an infusion of exogenous insulin. 1 Case Report
A 56-year-old woman was referred to the Mount Sinai Medical Center in March 1975 because of increasing frequency of symptoms of weakness, visual blurring, and sweating of 7 years duration. The attacks were not triggered by fasting, excercise or periods of emotional upset and occurred at any time of the day. The patient was non-obese (height 125 cm, weight 40.5 kg) and there was no weight gain during this period. Past medical history revealed no family history of diabetes mellitus; she had a one year history Received April 29, 1976. of hypertension which had been treated with Supported in part by research grants from the Eli hydrochlorthiazide 50 mg daily and diazepam Lilly Co., The American Diabetes Association and the 5 mg three times daily. Physical examination NIH (AM 18522), an NIH Research Career Developwas negative except for a right sided \xh x 1 ment Award KO 4-AM00089 (Dr. Rayfield); NIH Grant RR-71, Division of Research Resources, Clinical Re- cm thyroid nodule; there were no neuromuscular search Center, Mount Sinai School of Medicine; abnormalities. The admission laboratory data inAmerican Cancer Society Institutional Grant No. cluded a normal blood count, urinalysis, serum IN-410 and a Research and Development Award of the American Diabetes Association (Dr. Horwitz) and a 1 A preliminary report of this study has appeared PHS Diabetes Center Grant AM-17-46 (Dr. Rubenas an abstract. Rayfield, E. J., M. Pulini, A. Golub, A. H. stein). Rubenstein, and D. Horwitz, Nonautonomous funcReprint requests to: Dr. Elliot J. Rayfield, Mount tion of a pancreatic insulinoma (p.i.) Clin Res 25: Sinai School of Medicine, Fifth Avenue and 100th 593A, 1975. Street, New York, N.Y. 10029. 1307
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 22:09 For personal use only. No other uses without permission. . All rights reserved.
JCE & M • 1976 Vol 43 • No 6
RAYFIELD ET AL.
1308
TABLE 1. Carbohydrate tolerance tests Tolbutamide test
Glucose tolerance test
IRIt
Time (min.)
Glucose mg/dl
-15 0 5 15 30 45 60 90 120 180 240
66 63 — 78 52 32 34 46 63 49
192 447 110 49 20 8 99 13
300
65
10
51
/xU/ml
Proinsulin* ng/ml
ProinsulinJ %
0.40
49.1 — 1.5 — — 16.8
11 4
—
0.48 — — 1.59 0.59 — 0.81 —
36.7 30.1 —
Glucagon pg/ml
Glucose mg/dl
68 92
68 68 64 42 12 14 24 29 39 51
91 61 80 80 80 80 93
—
73
IRIt /iU/ml 5 6 719 588 184 42 27 12 8 9 — —
* Read from a human proinsulin standard after gel filtration. f IRI = Immunoreactive Insulin. \ Read from an insulin standard.
electrolytes, blood urea nitrogen, serum creatinine, calcium, phosphorus, alkaline phosphatase, uric acid, bilirubin, cholesterol, thyroxine and triiodothyronine. Chest x-ray and EKG were normal. The patient underwent the following diagnostic studies: oral glucose tolerance test, tolbutamide test, three day fast and insulin suppression test. Fasting hypoglycemia was never documented during her two hospitalizations nor on 3 glucose tolerance tests performed as an outpatient by her private physician. The lowest fasting plasma glucose recorded was 56 mg/dl. The patient had symptoms identical to those she described at home during the hypoglycemic phase of the tolbutamide test, the oral glucose tolerance test, and the insulin suppression test. A celiac and superior mesenteric angiogram was negative. At exploratory laparotomy, a 3 x 2 cm adenoma projecting from the inferior border of the pancreas at the junction of the body and tail was removed. On histologic examination the tumor was composed of cells arranged in isletlike structures containing many beta cell granules. There was no evidence of metastases. Six months post-operatively, the patient had no recurrence of hypoglycemic episodes. An oral glucose tolerance test was normal with respect to both insulin and glucose levels. Materials and Methods Plasma glucose was measured by the glucoseoxidase method with the Technicon Auto Analyzer. Serum immunoreactive insulin (IRI) (4,5),
proinsulin (6), percentage proinsulin, C-peptide reactivity (CPR) (7), and plasma glucagon (8) were measured by radioimmunoassay procedures. Acid-ethanol extracts of the pancreatic islet cell tumor were assayed for insulin by radioimmunoassay (5). Informed consent was obtained for all the studies. Results 1. Glucose tolerance test (Table 1) After 100 g oral glucose, plasma glucose fell from 63 mg/dl at 45 min, and returned to baseline levels by 180 min. Although the baseline serum IRI values at - 1 5 and 0 min fell within the normal range (5-20 ju-U/ml), a greatly exaggerated insulin concentration occurred following glucose administration with a peak value of 447 /uU/ml, at 30 min. Serum proinsulin levels rose from a fasting level of 0.40 ng/ml read against a human proinsulin standard to a peak of 1.59 ng/ml at 60 min. The contribution of proinsulin to the total immunoreactive insulin concentration in the basal state was clearly elevated at 49% (normal range 5-22%). The proinsulin per cent reached a nadir of 1.5% 15 min following glucose and then increased to 36% at 120 min (Table 1). Basal plasma glucagon levels were within the normal range (68-92 pg/
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 22:09 For personal use only. No other uses without permission. . All rights reserved.
NONAUTONOMOUS PANCREATIC INSULINOMA ml), diminishing to 61 pg/ml at 30 min and returning to baseline at 45 min.
3.0
1
1
1
1309
I
I
2.5 --
o
2. Tolbutamide test Following 1.0 g tolbutamide iv, plasma glucose levels decreased from 68 mg/dl (time 0) to 12 mg/dl at 30 min with sustained hypoglycemia (Table 1). Serum IRI values increased markedly from 6 /x,U/ml (time 0) to 719 /-lU/ml at 5 min and 588 /x.U/ml 15 min following tolbutamide. 3. Insulin suppression test Regular insulin was infused at a rate of 0.1 U/h for 60 min using a Harvard infusion pump. During the infusion of exogenous insulin, serum CPR was measured to monitor endogenous insulin secretion. C-peptide reactivity suppressed normally during the first 40 min of the test (Fig. 1), and thereafter increased slightly above this range to levels of 1.13 and 1.35 ng/ml at 50 and 60 min, respectively. 4. 72 hour fast Plasma glucose values remained consistently above 50 mg/dl. IRI values varied between 6.5 and 2.5 /aU/ml and glucose/ IRI ratios ranged between 9.1 to 23 (normal range 4-14) (9). At the end of the 72 h fast, plasma glucose was 56 mg/dl, IRI 8 /tU/ml and the percentage proinsulin was 35. The true insulin level (following column fractionation with separation of proinsulin) was 4.9 /uU/ml and the glucose/IRI ratio was 7.0. 5. Tumor insulin content The tumor was extracted for immmunoreactive insulin and showed 23.8 U/gram. Normal pancreas values ranged from 0.5 to 2 U/g, while tumors vary from very high to very low values (10). This level is compatible with an insulinoma. Discussion Insulin suppression tests (prolonged fast or infusion of exogenous insulin) are con-
/
£2.0 1.5 h 1.0 0.8
60
/
^
/
:
I
I
I
-
I
20 30 40 50 60 70 80 90 100 PLASMA GLUCOSE (mq/IOOml)
FIG. 1. Relationship between plasma glucose and Cpeptide immunoreactivity during the 60 min insulin infusion. The hatched area depicts the normal range at each blood sugar concentration in 7 healthy non diabetic volunteers as previously determined (15); the closed circles connected by broken lines indicate the values for the insulinoma patient and the time (in minutes) at which each sample was taken is given next to each circle.
sidered to be more reliable than stimulation tests in the diagnosis of insulinomas (2,11). In the largest available series of surgically documented beta cell tumors (12), only 2 of 108 patients failed to manifest a "typical hypoglycemic attack" by the end of a prolonged fast lasting as long as 72 h. In the 106 patients who became hypoglycemic, the blood sugar was lower than 50 mg/dl by the methods of Folin-Wu (normal range 80-120 mg/dl) or Somogyi (normal range 65-90 mg/dl) (12). In this regard, Merimee and Tyson (13) have documented a sex difference in plasma glucose homeostasis, in that plasma glucose concentrations of as low as 35 mg/dl occurred after one day of fasting in healthy women, in contrast to 55 mg/dl in men. Thus, during a fast, concomitant glucose and insulin levels are necessary to determine whether an abnormality is present. However, the 72 h values (at the completion of the fast) for plasma glucose, insulin, and glucose/insulin ratios were all within the normal range in the patient described in this report. An alternative procedure has recently been developed in which the suppression of endogenous insulin secretion is assessed by measuring CPR (14,15) levels during the
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 22:09 For personal use only. No other uses without permission. . All rights reserved.
1310
JCE & M • 1976 Vol 43 • No 6
RAYFIELD ET AL.
infusion of porcine insulin. During the infusion of exogenous insulin (Fig. 1), healthy subjects, as well as patients with hypoglycemia due to causes other than insulinoma, suppress CPR levels to less than 1.3 ng/ml when the plasma glucose falls below 40 mg/dl. Furthermore, CPR levels remain below 1.3 ng/dl during the remainder of the insulin infusion (15). The response of our patient is unusual in that, unlike 8 other patients with proven insulinomas (16), CPR levels suppressed below 1.3 ng/ml when her plasma glucose fell below 40 mg/dl. Unlike control subjects, however, CPR levels did not remain below 1.3 ng/ml for the entire insulin infusion. The patient described in this report thus represents an uncommon circumstance in which stimulation tests with tolbutamide and glucose were more helpful in establishing a preoperative diagnosis than were the suppression tests. There are a few cases in the literature of atypical patterns of insulin secretion in pancreatic beta cell tumors. Harano et al. (17) described a patient with a surgically documented pancreatic insulinoma who had blunted insulin responses following tolbutamide and glucagon, but an exaggerated insulin response following secretin administration. Our patient's exaggerated insulin response following oral glucose, while rare, has been previously noted by Power (18) in a patient who also had normal glucose values during a 72 h fast. While the mechanism for the marked insulin responses following glucose remains speculative, it is possible that this tumor retained the capacity to respond normally to both decreases and increases in plasma glucose. While the response to each of the tumor cells was quantitatively the same as "normal beta cells" the overall release of insulin was quantitatively larger because of the larger number of beta cells in the tumor. The test responses presented here illustrate that pancreatic beta cell tumors like other endocrine tumors may exhibit a spectrum from autonomous to nonautonomous function (19). Finally, it should
be noted that the elevated percentage of proinsulin (49%) in the basal state in the face of normal insulin levels (Table 1) was most helpful in establishing that a beta cell tumor was the cause of the patient's symptoms (20). References 1. Fajans, S. S., In Wintrobe, M. M, G. W. Thorn, R. D. Adams, E. Braunwald, K. Isselbacher, and R. G. Petersdorf (eds.), Harrison's Principles of Internal Medicine, ed. 7, McGraw-Hill Book Company, New York, 1974, p. 554. 2. Schein, P. S., R. A. DeLellis, C. R. Kahn, P. Gorden, and A. R. Kraft, Islet cell tumors: Current concepts and management, Ann Intern Med 79: 239, 1973. 3. Turner, R. C , and E. Harris, Diagnosis of insulinomas by suppression tests, Lancet 2: 188, 1974. 4. Juan, C , and T. W. AvRuskin, A combined immunoassay of human growth hormone and insulin: Cumulative assessment of assay performance,/ Clin Endocrinol Metab 33: 150, 1971. 5. Morgan, C. R., and A. Lazarow, Immunoassay of insulin: Two antibody system: Plasma insulin levels of normal, subdiabetic, and diabetic rats, Diabetes 12: 115, 1963. 6. Melani, F., A. H. Rubenstein, and D. F. Steiner, Human serum proinsulin, J Clin Invest 49: 497, 1970. 7. Block, M. B., M. E. Mako, D. F. Steiner, and A. H. Rubenstein, Circulating C-peptide immunoreactivity: Studies in normals and diabetic patients, Diabetes 21: 1913, 1972. 8. Aquliar-Parada, E., A. M. Einsentraut, and R. H. Unger, Pancreatic glucagon secretion in normal and diabetic subjects, Am J Med Sci 257: 415, 1969. 9. Grunt, J. R., J. A. Pallotta, and J. S. Soeldner, Blood sugar, serum insulin and free fatty acid relationships during intravenous tolbutamide testing in normal young adults, and in patients with insulinoma, Diabetes 19: 122, 1970. 10. Creutzfeldt, W., R. Arnold, C. Creutzfeldt, U. Deuticke, H. Frerichs, and N. S. Track, Biochemical and Morphological Investigation of 30 Human Insulinomas: Correlation Between the Tumor Content of Insulin and Proinsulin-like Components and the Histological and Ultrastructural Appearance, Diabetologia 9: 217, 1973. 11. Editorial, Diagnosis of insulinoma, Lancet 2: 385, 1974. 12. Laroche, G. P., D. O. Ferris, J. T. Priestly, D. A. Scholz, and M. B. Dockerty, Hyperinsulinism, Arch Sur 96: 763, 1968. 13. Merimee, T. J., and J. E. Tyson, Stabilization of
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 22:09 For personal use only. No other uses without permission. . All rights reserved.
NONAUTONOMOUS PANCREATIC INSULINOMA plasma glucose during fasting: Normal variations in two separate studies, N Engl J Med 291: 1275, 1974. 14. Horwitz, D. L., and A. H. Rubenstein, Insulin suppression, Lancet 2: 1021, 1974. 15. Horwitz, D. L., A. H. Rubenstein, C. Reynolds, G. D. Molnar, and N. Yanaihara, Prolonged suppression of insulin release by insulin-induced hypoglycemia: Demonstration by C-peptide assay, Honn Metab Res 7: 449, 1975. 16. Service, F. J., D. L. Horwitz, H. Kuzuya, M. Mako, C. Reynolds, G. D. Molnar, and A. H. Rubenstein, Diagnosis of insulinomas by C-peptide measurement during hypoglycemia induced by insulin infusion, The Endocrine Society 58th Annual Meeting Program and Abstracts: 66, 1976 (Abstract).
1311
17. Harano, Y., M. Kohama, N. Araki, A. Sakamoto, M. Hoshi, M. Shichiri, K. Shima, J. Okamura, and Y. Shigeta, Qualitative abnormality of insulin secretion in a case with insulinoma, Endocrinol Jap 22: 175, 1975. 18. Power, L., A. glucose-responsive insulinoma, JAMA 207: 893, 1969. 19. Rayfield, E. J., L. I. Rose, J. P. Cain, R. G. Dluhy, and G. H. Williams, ACTH-responsive, dexamethasone-suppressible adrenocortical carcinoma, N Engl J Med 284: 591, 1971. 20. Alsever, R. N., J. P. Roberts, J. G. Gerber, M. E. Mako, and A. H. Rubenstein, Insulinoma with low circulating insulin levels: The diagnostic value of proinsulin measurements, Ann Intern Med 82: 347, 1975.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 22:09 For personal use only. No other uses without permission. . All rights reserved.
