0021-972X/99/$03.00/0 The Journal of Clinical Endocrinology & Metabolism Copyright © 1999 by The Endocrine Society

Vol. 84, No. 8 Printed in U.S.A.

Plasma Chromogranin A in Patients with Multiple Endocrine Neoplasia Type 1 DAN GRANBERG, MATS STRIDSBERG, REIN SEENSALU, BARBRO ERIKSSON, ¨ BERG, AND BRITT SKOGSEID GUDMAR LUNDQVIST, KJELL O ¨ ., B.S.) and Clinical Chemistry (M.S., G.L.), Departments of Medical Sciences (D.G., B.E., K.O University Hospital, S-751 85 Uppsala, Sweden; and the Departments of Medicine (R.S.), Huddinge University Hospital and St. Go¨ran Hospital, S-11281 Stockholm, Sweden ABSTRACT Plasma chromogranin A (CgA) has been claimed to be a sensitive marker for neuroendocrine tumors, but its role in the early diagnosis of multiple endocrine neoplasia type 1 (MEN 1) pancreatic endocrine tumors has not been evaluated. We measured CgA in 36 patients with MEN 1, of whom 9 lacked pancreatic involvement, 20 had biochemical evidence of pancreatic endocrine tumors, and 7 displayed radiologically detectable pancreatic tumors. CgA was also analyzed in 25 patients with sporadic pancreatic endocrine tumors, 39 subjects with inflammatory bowel disease, 7 patients harboring nonendocrine pancreatic disease, and 19 healthy controls. Four of 9 of the MEN 1 patients without pancreatic involvement had elevated CgA. Furthermore, 60% with biochemically unequivocal tumors and all with a

P

ANCREATIC endocrine tumors occur in about 75% of multiple endocrine neoplasia type 1 (MEN 1) gene carriers, and about half of these patients develop malignant disease (1). It has been shown that a standardized meal stimulation test with measurement of pancreatic polypeptide (PP) and gastrin can detect the pancreatic lesions several years before they can be seen by radiological examinations (2). Several reports have pinpointed the value of plasma CgA as a marker for neuroendocrine tumors (3–7), but elevated CgA levels have also been found in patients with other diseases, such as impaired renal function (8 –10), deteriorating liver function (8), untreated benign essential hypertension (11), chronic atrophic gastritis (12), and prostatic carcinoma (13, 14). In primary hyperparathyroidism (HPT), conflicting results have been reported (3, 15). This study was designed to 1) evaluate the usefulness of basal CgA as a marker for early diagnosis of pancreatic endocrine tumors in MEN 1 patients, 2) investigate the value of adding measurements of CgA to PP and gastrin during the meal test, 3) recognize the degree of spontaneous variation in neuroendocrine tumor patients, and 4) identify the frequency of elevations of CgA in some nonendocrine diseases of the gut and pancreas. Subjects and Methods Informed consent was obtained from all patients participating in the study. All 36 patients with MEN 1 regularly attending our endocrine Received August 27, 1998. Revision received April 6, 1999. Accepted May 14, 1999. Address all correspondence and requests for reprints to: Britt Skogseid, M.D., Department of Endocrine Oncology, Division of Internal Medicine, University Hospital, S-751 85 Uppsala, Sweden. E-mail: [email protected].

radiologically visible tumor showed elevations. All 25 patients with sporadic pancreatic endocrine tumor had increased CgA, as had 28% of patients with inflammatory bowel disease and 57% with nonendocrine pancreatic disease. Mean day to day CgA variation was 29% (range, 0 –113%) in the neuroendocrine tumor patients and 21.0% (range, 0.0 – 47%, within reference range) among healthy controls. In summary, nonendocrine diseases may cause elevation of CgA, and its spontaneous variation can be considerable. Plasma chromogranin A is the most sensitive of the basal markers for neuroendocrine tumors, but cannot replace other established measures when screening for early pancreatic involvement in MEN 1. (J Clin Endocrinol Metab 84: 2712–2717, 1999)

oncology clinic, 27 with pancreatic endocrine tumors, were included in the study as soon as a meal stimulation test and measurement of CgA could be performed simultaneously. If the radiological examinations failed to demonstrate a pancreatic tumor, increasingly abnormal levels of at least 2 independent biochemical tumor markers were demanded to seal the diagnosis of pancreatic involvement (1).

MEN 1 subjects (groups A–C) Group A consisted of nine patients, five men and four women, who did not fulfill the biochemical criteria for pancreatic endocrine tumor diagnosis mentioned above and had no radiologically visible tumor in the pancreas (Table 1). However, eight patients in this group had intermittently abnormal meal stimulation tests, two patients had elevated serum proinsulin levels, and one had slightly elevated plasma glucagon levels. Five of the patients had increased serum calcium; the other four had previously been subjected to surgery for primary HPT. Three had been treated for pituitary tumors with surgery or bromocriptine, and another two showed elevated PRL or GH. Their mean age was 46 yr (range, 32–76 yr). The 20 group B patients (10 men and 10 women; mean age, 51 yr; range, 26 –74 yr) had biochemical evidence of endocrine pancreatic tumors by means of at least 2 independent and consistently elevated circulating tumor markers (Table 1) (1), but no radiological signs of an endocrine tumor in the pancreas. The tumors were subsequently extirpated in 16 of the patients. Five of the patients were hypercalcemic, and all but 1 of the other 15 had undergone surgery for primary HPT. Two patients had ongoing pituitary tumors disease, and another 8 had already been successfully treated. Seven patients (three men and four women; mean age, 59 yr; range, 39 –71 yr) fulfilled the criteria for belonging to group C, i.e. they all had demonstrable pancreatic endocrine tumors either by percutaneous ultrasonography or computed tomography scan (Table 1). All were normocalcemic, and all but one had previously been operated on for primary HPT. Altogether three of the seven patients harbored pituitary lesions at inclusion in this study.

2712

PLASMA CHROMOGRANIN A IN MEN 1

2713

TABLE 1. Hormone values and response to the meal test for all MEN 1 patients Patient no.

Group

CgA ,3.7 nmol/L

Calcium 2.2–2.6 mmol/L

PTH 12–55 ng/L

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36

A A A A A A A A A B B B B B B B B B B B B B B B B B B B B C C C C C C C

3.6 1.9 3.9 5.0 3.5 1.4 4.5 2.0 6.2 3.8 11.0 4.2 47.4 2.5 12.2 5.0 1.2 2.1 31.5 9.8 14.4 22.5 10.1 1.5 2.8 2.3 1.5 11.7 2.3 125.4 46.5 5.5 6.9 7.1 13.2 8.8

2.80 2.14 2.23 2.59 2.84 2.36 3.05 2.36 2.73 2.30 2.57 2.73 2.24 2.70 2.31 2.69 2.20 2.55 2.18 2.36 2.71 2.51 2.47 2.30 2.33 2.21 2.45 2.49 2.39 2.52 2.33 2.54 2.56 2.52 2.42 2.29

51 11 ,10 69 82 143 33 66 33 64 34 61 12 75 16 43 52 32 14 27 52 19 52 19 46 44 33 ,10

GH ,5 mg/L

PRL ,15 mg/L

PP ,70 pmol/L

Gastrin ,55 pmol/L

Glucagon ,140 ng/L

Insulin ,20 mU/L

5.40 1.49 0.33 0.69 0.43 0.63 9.20 0.83 0.48 2.10 0.44 0.70 2.00 0.66 0.78 0.79 7.80 1.08 4.30 0.97 0.59 0.31 1.41 0.82 0.68 0.96 1.27 0.87 0.72 0.59 3.10 0.54 1.90 16.70 2.60 3.40

5.2 2.2 8.0 13.2 2.8 11.0 16.3 17.0 7.7 9.0 5.7 14.6 ,0.42 7.2 7.5 0.14 3.6 3.8 9.3 7.0 3.5 4.2 5.0 1.3 5.2 5.4 5.2 1.4 3.4 11.4 3.1 11.0 3.4 196.0 7.5 6.3

19 16 6 27 24 30 10 8 9 25 24 45 725 92 14 46 7 53 60 115 207 30 38 117 25 31 60 58 20 115 ,3 30 320 98 45 92

14.4 39.4 18.7 15 35.7 16.1 18 ,6 14.9 14.5 58 13 250 39.2 27.3 26 12 30.5 83 59.4 740 209 100 81 19.2 ,8.7 31.7 28.4 46.4 182 40.3 47.7 17 123 86 28.8

38 49 159 85 75 84 35 33 28 26 72 118 120.2 69 66 79 199 45 84 14 90 50 82 34 81.6 43 44 37 22 233 97 39 80 90 204 170

11.1 4.7 9.6 17.7 18.1 11.1 4.9 8.8 28 9.1 13.8 10.8 18.8 10.2 30 11.2 14 16.8 8.4 6.6 6.8 4.8 31 11.2 13.3 10.8 23 9.3 11.5 11.3 10.3 14.4 13.5 13.7 12.3

C Peptide 0.2–1.9 nmol/L

Proinsulin ,9.8 pmol/L

Meal test

0.75 0.41 0.99 1.71 1.27 1.34 1.14 0.36 0.76 1.26 0.69 1.2 1 1.07 0.85 2 1.55 0.79 1.1 0.85 0.67 0.84 0.6 0.36 0.84 1.13 0.96 1.03 0.63 1.9 0.86 0.63 0.88 1.85 1.23 0.62

4.4 ,2.5 9.4 40.4 14 6.7 5.9 ,2.5 2.5 11.9 27.8 14 8.7 55.7 9.1 31 14.6 15.3 8.9 3 2.89 4 ,2.5 6.72 121.6 15.3 11.6 67 10.3 21.3 12.1 3.53 12a 21 7.5 38.8

2 2 2 2 2 1 1 2 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 1 1 1 1 1 1 1 2 1 1 1 1

Group A, No endocrine pancreatic tumor; B, biochemical endocrine pancreatic tumor with normal radiology; C, radiologically detectable endocrine pancreatic tumor. 2, Normal PP and gastrin response to the meal test; 1, abnormal PP or gastrin response to the meal test. a Reference range, below 13 pmol/L. Underlined figures indicate value above the upper reference limit.

Non-MEN 1 subjects Twenty-five patients (12 men and 13 women) with sporadic pancreatic endocrine tumor constituted group D. Their mean age was 55 yr (range, 20 –74 yr). One patient had an insulinoma, 3 had glucagonoma, 6 suffered from gastrinoma, 1 had a vasoactive intestinal polypeptidesecreting tumor, another suffered from combined insulinoma and glucagonoma, and the remaining 13 patients harbored nonfunctioning tumors. Twenty-two of the patients had proven metastases. Group E consisted of seven patients (five men and two women; mean age, 62 yr; range, 48 –75) with nonendocrine pancreatic disease. Five suffered from a pancreatic adenocarcinoma, one had a benign exocrine pancreatic tumor, and another had chronic pancreatitis. The inflammatory bowel disease (IBD) group consisted of 39 patients (33 men and 6 women) diagnosed to have inflammatory bowel disease (ulcerative colitis, n 5 23; Crohn’s disease, n 5 16) with up to 20 diarrhea episodes daily (mean, 3.3). The mean age in this group was 40 yr (range, 17– 81 yr). In the control group we included 19 healthy volunteers, 9 men and 10 women. Their mean age was 37 yr (range, 21– 67 yr). To assess the degree of spontaneous variation, CgA was measured on 2 consecutive days in 7 healthy individuals and in 40 neuroendocrine tumor patients. The patients constituted 21 with midgut carcinoids, 12 with sporadic pancreatic endocrine tumors, and 7 with MEN 1 syndrome with pancreatic endocrine lesions. The tumor patients had CgA values between 2.1–1060 nmol/L (mean, 108.1; median, 14.9). In some rare cases, we observed a possible influence of nonsteroid

anti-inflammatory drugs medication on circulating CgA levels. To assess this idea, CgA was analyzed in seven healthy controls (four men and three women; mean age, 43 yr) before and after daily oral ingestion of 200 mg ketoprofen during 1 week.

Hormone assays Serum insulin, proinsulin, C peptide, PP, and gastrin as well as plasma glucagon were measured after 12 h of fasting, together with serum PRL, GH, and PTH. These hormones were analyzed as described previously (2, 16 –23). Serum calcium, serum albumin, blood glucose, and other routine serum markers were analyzed at the routine clinical chemistry laboratory. A standardized meal stimulation test was performed in patients and the 19 healthy controls as follows. After an overnight fast a 563-Cal mixed meal composed of 66 g carbohydrate, 18 g protein, and 22 g fat was eaten during 20 min. Blood for analysis of gastrin, PP, and CgA was collected 5 min before starting, just before starting, and at 10, 20, 30, 45, and 60 min after starting the meal. An elevation of PP or gastrin of more than 2 sd above the mean in normal controls was considered an abnormal response to the meal according to earlier comprehensive evaluation of the meal test (2). The CgA response to the meal was assessed in 18 MEN 1 patients (6 in group A, 9 in group B, and 3 in group C) and compared to the response in the 19 control subjects of the present study. Blood samples for basal plasma CgA measurements were collected while subjects were fasting in the morning. Before 1995, the analysis was

2714

JCE & M • 1999 Vol 84 • No 8

GRANBERG ET AL.

accomplished by RIA, as previously described (4). From 1995, CgA was analyzed by a novel method with a detection limit of 10 pmol/L and an interassay variation of 6.4% (24). As there is a strong correlation between the two methods (r 5 0.96; P , 0.001; n 5 320), the older values could be transformed to the new ones by multiplication by a factor of 0.015. The upper reference limit in this study was defined as 2 sd above the mean level of our control subjects, i.e. 3.7 nmol/L.

Statistics Results are reported as the mean 6 sd unless otherwise stated. Intergroup comparison of mean values was performed according to the nonparametric test of Kruskal-Wallis and Mann-Whitney U tests, as well as parametric Student’s t test for unpaired samples. Correlation between CgA and other hormones was calculated by linear regression. P , 0.05 was considered significant. The spontaneous variation in CgA was expressed as a percentage and calculated according to the following formula: (CgA on day 1 2 CgA on day 2)/CgA on day 1.

Results

The results of the CgA analyses in the different patient groups are presented in Table 2 and Fig. 1. All of the MEN 1 patients with radiologically visible pancreatic endocrine tumor had elevated CgA (of whom the 2 patients with liver metastases had the highest values), as had all of the patients with sporadic pancreatic endocrine tumor. Both of these groups had mean CgA levels higher than those in the healthy controls (P 5 0.0001 and P , 0.0001, respectively). Twelve of the 20 patients (60%) with biochemical evidence of a pancreatic endocrine tumor (group B), and 4 of the 9 (44%) MEN 1 patients without established pancreatic endocrine tumor diagnosis (group A) had elevated CgA. None of these groups showed mean CgA levels significantly different from those in the control subjects (P 5 0.08 and P 5 0.18, respectively) when applying nonparametric statistics, but both groups showed significantly higher values than controls by Student’s t test (Table 2). When the MEN 1 groups were compared, a significant difference was found between groups A and C (P 5 0.001), but not between groups A and B (P 5 0.21) or between groups B and C (P 5 0.11). Four of the seven patients (57%) with nonendocrine pancreatic disease had increased CgA, and the mean level in this group was significantly higher than that in controls (P , 0.001). Eleven of the 39 patients with IBD (28%) had CgA above the upper reference limit. Mean CgA did not differ between control subjects and the IBD patients. Although patients with Crohn’s disease had a tendency toward higher CgA than those with ulcerative colitis, the difference did not reach

significance (P 5 0.07). There was no correlation between CgA levels and the number of daily diarrhea episodes. Six of 10 MEN 1 patients with hypercalcemia due to HPT had elevated CgA. All 3 group A patients with primary HPT and increased CgA had some sign of a pancreatic endocrine tumor (abnormal meal test in 2 cases and elevated proinsulin in 1). In group B there was no difference in CgA between the hyper- and normocalcemic patients, but the hypercalcemic group A patients had near-significantly higher CgA levels than those who were normocalcemic (nonparametric, P 5 0.05; parametric, P 5 0.02; Table 3). The highest CgA value among the hypercalcemic patients was recorded in the only patient with concomitant hypergastrinemia. There was no correlation between log plasma CgA and log serum calcium or PTH. The CgA levels were not significantly altered after surgery for primary HPT (but only 3 patients were evaluable

FIG. 1. CgA values for the different groups (nanomoles per L; log10 scale). The dotted line represents the upper reference limit (mean 6 2SD) calculated from our healthy controls. The short lines show the mean values. H, Healthy controls; A, MEN 1, no pancreatic involvement; B, MEN 1, biochemically detected pancreatic endocrine tumor; C, MEN 1, radiologically demonstrable pancreatic tumor; D, sporadic endocrine pancreatic tumor; E, nonendocrine pancreatic disease.

TABLE 2. Statistical comparison of CgA levels between healthy controls and the various patient groups Patient group

Healthy controls (n 5 19) MEN 1 groups No pancreatic tumor (n 5 9) Biochemical pancreatic tumor (n 5 20) Radiological pancreatic tumor (n 5 7) Sporadic pancreatic endocrine tumor (n 5 25) Nonendocrine pancreatic disease (n 5 7) Inflammatory bowel disease (n 5 39) Ulcerative colitis (n 5 23) Crohn’s disease (n 5 16)

Mean (nmol/L)

SD

2.7

0.5

3.5 10.0 30 136 4.6 2.9 2.7 3.2

1.6 11.8 44 271 2.4 1.0 1.0 1.0

Range

P value for Mann-Whitney U/Student’s t test

2.0–3.9 1.4– 6.2 1.2– 47 5.5–125 4.8–1305 1.4–10.0 1.3–5.6 1.3–5.6 1.7–5.6

NS/,0.05 0.08/,0.05 ,0.001/,0.01 ,0.001/,0.05 ,0.001/,0.01 NS/NS

PLASMA CHROMOGRANIN A IN MEN 1

2715

TABLE 3. Statistical comparison between hypercalcemic and normocalcemic patients of both groups A and B MEN 1 group

No pancreatic endocrine tumor Hypercalcemia (n 5 5) Normocalcemia (n 5 4) B. Biochemical pancreatic tumor Hypercalcemia (n 5 5) Normocalcemia (n 5 15)

Mean (nmol/L)

SD

Range

P value for Mann-Whitney U/Student’s t test

A.

4.5 2.3

1.1 1.1

3.5– 6.2 1.4–3.9

0.05/0.02

5.6 11.4

5.1 13.2

2.1–14.4 1.2– 47.4

NS/NS

FIG. 2. Spontaneous variation of CgA. a, Healthy individuals (n 5 7); b, neuroendocrine tumor patients with CgA below 40 nmol/L (n 5 27); c, neuroendocrine tumor patients with CgA between 40 –250 nmol/L (n 5 8); d, neuroendocrine tumor patients with CgA above 250 nmol/L (n 5 5).

in this respect). CgA was elevated in 10 of the 11 (90%) group B and C patients displaying high serum gastrin levels. The corresponding figure for patients with insulin/proinsulinproducing tumors was 9 of 17 (53%; Table 1). Three of five MEN 1 patients with pituitary tumors had CgA levels slightly above the upper reference limit (4.2–7.1 nmol/L). Two of these had verified pancreatic endocrine tumors, and the third (group A) had an abnormal response to the meal. During the meal test, the mean CgA increased 16% (within the reference range) in control subjects. In group A patients, CgA increased 25%, whereas in groups B and C, the responses were 31% and 20%, respectively. The maximum CgA values occurred between 30 – 60 min after the onset of the meal. In the patients with neuroendocrine tumors, the mean day to day variation in CgA was 29.3 6 22.4% (range, 0.0 –113.5%). The patients with normal CgA (n 5 3) had a mean variation of 27.0% compared to 29.5% among those with elevated CgA (n 5 37; P 5 NS). The variation did not differ significantly between the tumor types. In healthy subjects, mean day to day variation was 21.0 6 16.6% (range, 0.0 – 47%). This was not different from the variation in the neuroendocrine tumor patients, although all CgA values were normal in all healthy individuals. The day to day variation is shown in Fig. 2. Plasma CgA before and after 1 week

of ketoprofen ingestion in healthy individuals was not significantly changed (2.7 6 1.9 and 1.9 6 0.7 nmol/L, respectively; P 5 NS). Discussion

In this study we have focused on evaluating the specificity of CgA as a marker for early detection of small endocrine tumors of the pancreas in MEN 1 patients. The results show that CgA is elevated in a substantial proportion of patients with MEN 1. The levels were highest when a pancreatic endocrine tumor was radiologically visible, intermediate in patients with mere biochemical evidence of a pancreatic tumor, and lowest when no tumor was provable. Furthermore, the two MEN 1 patients with metastatic disease had the highest values. These results indicate that the CgA level in MEN 1 patients may correlate to tumor burden, as earlier shown for midgut carcinoids (5). Altogether 70% of the MEN 1 patients with established endocrine pancreatic tumor diagnosis had increased CgA. None of the other basal markers was as sensitive (33–59%). The meal test, however, reached 85% sensitivity. The analysis of CgA during the meal test added no further information to basal CgA. Of the MEN 1 patients without established pancreatic endocrine tumor diagnosis, 44% showed CgA elevations. This finding might constitute false positive elevations. On the

2716

JCE & M • 1999 Vol 84 • No 8

GRANBERG ET AL.

other hand, it could be due to CgA secretion from an otherwise undetectable tumor in the pancreas or represent secretion of CgA from other MEN 1 lesions. The majority of these patients will most likely develop pancreatic endocrine tumors and only two patients, both with normal CgA, lacked all signs of pancreatic endocrine tumor at inclusion in the study, although they later developed an abnormal PP response to the meal test. A long term prospective study will be necessary to actually pinpoint the efficacy of CgA as a marker for the very early onset of the pancreatic lesion in MEN 1, but in the meantime we recommend annual follow-up of known MEN 1 carriers with elevated CgA. It has previously been shown that CgA is elevated in hypercalcemic primary HPT patients only when they have a concomitant Zollinger-Ellison syndrome (15). In our study, 6 of 10 MEN 1 patients with hypercalcemia due to primary HPT had increased CgA. All 6 showed elevations of at least 1 marker for pancreatic involvement. Serum gastrin was increased in 1 of these MEN 1 patients. Thus, our results may support the suggestion that coexisting pancreatic endocrine tumors are responsible for perceivable elevations of CgA in MEN 1, but not necessarily in gastrinomas (15). Earlier studies have found elevated CgA in up to one third of patients with pituitary tumors. Whether these patients represented MEN 1 gene carriers was not determined (25, 26). We found elevated CgA in three of five MEN 1 patients with pituitary tumors. Since two of these three also had pancreatic endocrine tumors, no conclusions can be drawn about the impact of a pituitary tumor on the plasma CgA level in MEN 1 patients. Our results indicate that slightly increased CgA may also occur in ulcerative colitis and Crohn’s disease as well as in patients with a variety of nonendocrine pancreatic diseases. Previously, O’Connor and Deftos claimed that normal CgA levels were found in various nonpeptide-producing tumors (3). However, in their study, the cut-off limit was 6.4 sd above the mean CgA level in healthy controls. In our laboratory, upper reference limits are most commonly defined as 2 sd above the mean level in control subjects. When reanalyzing the data from the report by O’Connor and Deftos and applying our principles for reference intervals, as much as 40% of their patients with pancreatic adenocarcinoma would have been considered to have elevated CgA. Thus, in accordance with the results from our study, moderate elevation of CgA in a patient disclosing radiological findings of a pancreatic lesion does not exclude an adenocarcinoma, and histological diagnosis is mandatory. Conflicting data have been presented on the day to day variation of CgA in healthy subjects, with values between 0 –21% (3, 27). In this study we found a spontaneous variation of 21% in CgA in healthy individuals (range, 0.0 – 47%), although all measurements were within the reference range. The variation of CgA in patients with neuroendocrine tumors, never previously assessed, is pronounced, most likely due to tumoral instability and liability to release hormones. Repeated measurements and careful interpretation are thus warranted when using CgA for treatment monitoring. A recent comprehensive study evaluated the efficacy of CgA, neuron-specific enolase, and the a-subunit of glycoprotein hormones as circulating tumor markers in patients

with neuroendocrine tumors (28). CgA was found to be the most sensitive and specific of these three. It was, however, concluded that some patients with neuroendocrine tumors had normal levels of CgA, and some patients with nonneuroendocrine diseases displayed elevated levels of CgA. We conclude that CgA is the single most sensitive marker of pancreatic involvement in MEN 1. On the other hand, CgA elevations are less prominent in cases of limited disease and often within the same range as in patients with nonendocrine gastrointestinal or pancreatic diseases. Thus, mere CgA analysis cannot be recommended as a single marker for pancreatic tumor involvement in young MEN 1 patients. However, middle-aged MEN 1 patients harboring a substantial tumor burden frequently display high CgA levels. Finally, as the spontaneous day to day variation is not negligible, we emphasize that slight elevations of CgA should be interpreted with caution, and repeated measurements should be considered. References 1. Skogseid B, Eriksson B, Lundqvist G, et al. 1991 Multiple endocrine neoplasia type 1: a 10-year prospective study in four kindreds. J Clin Endocrinol Metab. 73:281–287. ¨ berg K, Benson L, et al. 1987 A standardized meal stimulation 2. Skogseid B, O test of the endocrine pancreas for early detection of pancreatic endocrine tumors in multiple endocrine neoplasia type I syndrome: five years experience. J Clin Endocrinol Metab. 64:1233–1240. 3. O’Connor DT, Deftos LJ. 1986 Secretion of chromogranin A by peptideproducing endocrine neoplasms. N Engl J Med. 314:1145–1151. ¨ berg K, et al. 1990 A polyclonal antiserum against 4. Eriksson B, Arnberg H, O chromogranin A and B: a new sensitive marker for neuroendocrine tumours. Acta Endocrinol (Copenh). 122:145–155. 5. Janson ET, Holmberg L, Stridsberg M, et al. 1997 Carcinoid tumors: analysis of prognostic factors and survival in 301 patients from a referral center. Ann Oncol. 8:685– 690. 6. Deftos LJ. 1991 Chromogranin A: its role in endocrine function and as an endocrine and neuroendocrine tumor marker. Endocr Rev. 12:181–186. ¨ berg K. 1991 Peptide hormones as tumor markers in neuroen7. Eriksson B, O docrine gastrointestinal tumors. Acta Oncol. 30:477– 483. 8. O’Connor DT, Pandian MR, Carlton E, Cervenka JH, Hsiao RJ. 1989 Rapid radioimmunoassay of circulating chromogranin A: in vitro stability, exploration of the neuroendocrine character of neoplasia, and assessment of the effects of organ failure. Clin Chem. 35:1631–1637. 9. Hsiao RJ, Mezger MS, O’Connor DT. 1990 Chromogranin A in uremia: progressive retention of immunoreactive fragments. Kidney Int. 37:955–964. 10. Stridsberg M, Husebye ES. 1997 Chromogranin A and chromogranin B are sensitive circulating markers for phaeochromocytoma. Eur J Endocrinol. 136:67–73. 11. Takiyyuddin MA, Cervenka JH, Hsiao RJ, Barbosa JH, Parmer RJ, O’Connor DT. 1990 Chromogranin A. Storage and release in hypertension. Hypertension. 15:237–246. 12. Borch K, Stridsberg M, Burman P, Rehfeld JF. 1997 Basal chromogranin A and gastrin concentrations in circulation correlate to endocrine cell proliferation in type-A gastritis. Scand J Gastroenterol. 32:198 –202. 13. Kadmon D, Thompson TC, Lynch GR, Scardino PT. 1991 Elevated plasma chromogranin-A concentrations in prostatic carcinoma. J Urol. 146:358 –361. 14. Angelsen A, Syversen U, Haugen OA, Stridsberg M, Mjølnerød OK, Waldum HL. 1997 Neuroendocrine differentiation in carcinomas of the prostate: do neuroendocrine serum markers reflect immunohistochemical findings? Prostate. 30:1– 6. 15. Nanes MS, O’Connor DT, Marx SJ. 1989 Plasma chromogranin A in primary hyperparathyroidism. J Clin Endocrinol Metab. 69:950 –955. 16. Heding LG. 1977 Specific and direct radioimmunoassay for human pro-insulin in serum. Diabetologica. 13:467– 474. 17. Heding LG, Larssen UD, Markussen J, Jo¨rgensen KH, Hallund O. 1974 Radioimmunoassay for human, pork and ox C-peptides and related substances. Horm Metab Res. 5:40 – 44. 18. Unger RH, Aquillar-Parada E, Muller WA, Eisentraut AM. 1970 Studies of pancreatic alpha cell function in normal and diabetic subjects. J Clin Invest. 49:837– 848. 19. Ha¨llgren R, Lundqvist G, Chance RE. 1977 Serum levels of human pancreatic polypeptide in renal disease. Scand J Gastroenterol. 12:923–927.

PLASMA CHROMOGRANIN A IN MEN 1 20. Lundqvist G, Wide L. 1977 Serum gastrin determination with a radioimmunosorbent technique. Clin Chim Acta. 79:357–362. 21. Nyberg F, Bergman P, Wide L, Roos P. 1985 Stability studies on human pituitary prolactin. Ups J Med Sci. 90:265–277. ¨ berg K, Norheim I, Wide L. 1985 Serum growth hormone in patients with 22. O carcinoid tumors. Basal levels and response to glucose and thyrotropin releasing hormone. Acta Endocrinol (Copenh). 109:13–18. 23. Benson L, Ljunghall S, Groth T, et al. 1987b Optimal discrimination of mild hyperparathyroidism with total serum calcium, ionized calcium, and parathyroid hormone measurements. Ups J Med Sci. 92:141–176. ¨ berg K. 24. Stridsberg M, Hellman U, Wilander E, Lundqvist G, Hellsing K, O 1993 Fragments of chromogranin A are present in the urine of patients with carcinoid tumors: development of a specific radioimmunoassay for chromogranin A and its fragments. J Endocrinol. 139:329 –337.

25. Deftos LJ, O’Connor DT, Wilson CB, Fitzgerald PA. 1989 Human pituitary tumors secrete chromogranin A. J Clin Endocrinol Metab. 68:869 – 872. 26. Nobels FRE, Kwekkeboom DJ, Coopmans W, et al. 1993 A comparison between the diagnostic value of gonadotropins, a-subunit, and chromogranin-A and their response to thyrotropin-releasing hormone in clinically nonfunctioning, a-subunit-secreting, and gonadotroph pituitary adenomas. J Clin Endocrinol Metab. 77:784 –789. 27. Takiyyuddin MA, Neumann HP, Cervenka JH, et al. 1991 Ultradian variations of chromogranin A in humans. Am J Physiol. 261:R939 –R944. 28. Nobels FRE, Kwekkeboom DJ, Coopmans W, et al. 1997 Chromogranin A as serum marker for neuroendocrine neoplasia: comparison with neuron-specific enolas and the a subunit of glycoprotein hormones. J Clin Endocrinol Metab. 82:2622–2628.

Seeking individuals with Idiopathic Hypogonadotropic Hypogonadism or Kallmann syndrome for participation in a genetic study Familial cases most desired, but all cases welcome. Mutational screening provided, free of charge. For more information, please contact: Dr. William Crowley or Dr. Stephanie Seminara, Reproductive Endocrine Unit, Massachusetts General Hospital, Boston, MA 02114. •E-mail: [email protected] •Voice Mail: 617-726-5387 •Fax: 617-726-5357

2717