J. Endocrino/. /nvest. 13: 807 -815 , 1990
Evaluation of clonidine suppression and various provocation tests in the diagnosis of pheochromocytoma H. Koshida, I. Miyamori, R. Soma, T. Matsubara, S. Okamoto, M. Ikeda, and R. Takeda Second Department of lnternal Medieine, School of Medicine, Kanazawa University, Takara-machi 13-1, Kanazawa 920, Japan
ABSTRACT. Recent investigations have shown that the widely used clonidine suppression test is sometimes fallible for the diagnosis of pheochromocytoma. A comparative assessment was made of the following suppression and provocation tests, the clonidine suppression test, and the glucagon, metoclopramide, and naloxone provocation tests. The assessment was performed in 6 patients with pheochromocytoma and in 19 patients without pheochromocytoma who were initially suspected of harboring a tumor. BP response did not predict the presence of pheochromocytoma in any test. Plasma norepinephrine (NE) concentrations determined at 120 and 180 min after oral 150 JLg of clonidine gave false negative results in 2 of the 5 patients with pheochromocytoma tested. Both plasma NE and epinephrine (E) concer'ltrations were measured before and sequen-
tially after each provocative agent. Neither NE nor E responded to 1 mg of glucagon iv in 2 of the 4 patients with pheochromocytoma tested. Determination of the peak level, peak increment, and 0/0 peak increment of NE and E following 10 mg of naloxone iv did not distinguish the two groups. The % peak increments of both NE and E in all 4 patients with pheochromocytoma given 5 mg of metoclopramide iv exceeded the mean + 3 SD values of the patients without pheochromocytoma (25 + 28% for NE, and 25 + 42% for E). These results suggested that, when performed with judicious patient selection (ambiguous plasma or urinary catecholamine levels), the measurement of plasma catecholamines in response to metoclopramide can be a useful adjunctive tool in the diagnosis of pheochromocytoma.
INTRODUCTION
(1, 2). In reeent years, the clonidine suppression test has been widely used, since it avoids a pressor response and was initially thought to have a high specificity and sensitivity (1-3). However, recent experienees have indicated a number of pitfalls of this test. Both false-negative and false-positive results have oecurred in patients with small pheoehromoeytoma when the plasma eateeholamine levels were within the normal range (4, 5), and in patients without pheochromocytoma who were taking diuretics for a short term (6). Clinically silent, unsuspected adrenal masses are being increasingly detected incidentally due to the wider application of eomputerized tomography by unrelated reasons (7-10), and some of these prove to be pheochromoeytoma. It is critical, therefore, to determine whether these adrenal masses are pheochromoeytomas prior to needle aspiration biopsy (10) or surgical resection (11, 12). In this situation and when the clinical suspicion of pheochro-
Pheochromocytoma is an uncommon disease but the early diagnosis of this eatecholamine-produeing tumor is important, because it is usually eurable by surgieal reseetion and because of the risk of lethai eomplications when the diagnosis is delayed. The diagnosis is based on the elinical symptoms and a variety of biochemical procedures which indicate exeessive catecholamine production (1, 2). When the measurement of basal plasma and urinary catecholamines (metabolites) gives an equivocal result, further evaluation depends on pharmacologieal tests
Key-words: Pheochromocytoma, norepinephrine, epinephrine, clonidine, glucagon, metoclopramide, naloxone. Correspondence: Hideo Koshida, MD, 2nd Department of Internal Medicine, School of Medicine, Kanazawa University, Takara-machi 13-1, Kanazawa 920, Japan. Received December 20, 1989; accepted July 12, 1990.
807
H. Koshida, I. Miyamori, R. Soma, et al.
mocytoma persists, the only useful diagnostic rnaneuver is a provocation test (1, 2). To evaluate the reliability ot three provocation tests, the glucagon, metoclopramide, and naloxone tests, we performed these tests with meticulous care and compared the results with the clonidine test.
containing 200 meq of Na' and 60 meq of K+ per day. Antihypertensive medications were withdrawn at least 2 weeks before the study. No food, tea, coffee, or smoking were allowed for at least 12 h preceding the tests. The clonidine suppression test was performed first, followed by the glucagon, metoclopramide, and naloxone provocation tests, in a random order on separate days. Every test was begun between 08:00 hand 09:00 h. After insertion of two catheters into bilateral forearm veins and at least 30 min of rest in the supine position in a warm, quiet, and well-lighted room, two blood sam pies were drawn at 5-min intervals for plasma catecholamine measurement. Control measurements of BP and pulse rate were made three times at 5-min intervals. The patients then took 150 J.lg of clonidine orally with 100 ml of tap water, or received an iv bolus injection of 1 mg of glucagon, 5 mg of metoclopramide, or 10 mg of naloxone. Thereafter, in the clonidine test, BP and pulse rate were measured at 30-min intervals by an automated sphygmomanometer (BF-203, Nippon Colin Co. Ltd., Tokyo) and blood sampies were taken at 120 and 180 min. In the glucagon and metoclopramide tests, BP and pulse rate were measured at 1-min intervals for 15 min, and blood sampies were drawn at 1, 2, 3, 5, 10, and 15 min. In the naloxone test, BP and pulse rate were recorded at 5-min intervals for 30 min and blood sam pies were drawn at 5, 10, 15, 20, and 30 min. There was no pressor response to saline in any patient with pheochromocytoma and in the 8 selected patients without pheochromocytoma before the provocation tests were commenced. Patients remained supine throughout the study period. Phentolamine was always available at the bed side for immediate use. Blood sampies for plasma catecholamine determination were collected into pre-
MATERIALS AND METHODS We studied six patients with pheochromocytoma, in five of whom the diagnosis was confirmed by histopathological examination after surgery. Their clinical data are shown in Table 1. Case no. 1 refused an operation but the presence of pheochromocytoma was later confirmed by extremely elevated urinary and plasma catecholamine levels, computerized tomography, and 1311-metaiodobenzylguanidine (MIBG) scintigraphy. Nineteen additional patients (11 men and 8 women, mean age 53 yr, range 3569 yr), who were initially suspected of harboring a pheochromocytoma because of anxiety, headache, palpitations, diaphoresis, and vasomotor lability, were also enrolled in this study. In these patients, both pheochrömocytoma and secondary hypertension were ruled out by clinical examination and the following investigations: urinalysis, measurement of serum electrolytes, creatinine, BUN, creatinine clearance, urinary catecholamines, PRA, plasma aldosterone concentration, and intravenous pyelography. In selected cases, renal arteriography, computerized tomography of the adrenal glands, and 131 1_ MIBG scintigraphy were also performed. None of the patients without pheochromocytoma underwent surgical exploration. Informed consent was obtained from all the patients enrolled in this study according to the principles of the Declaration of Helsinki. All the patients were hospitalized and fed a diet
Table 1 - Profiles of the pheochromocytoma pa tients. Tests performed Patient (no.)
Age (yr)
Sex
Hypertension
C
1 2
58 59
3
51
4 5 6
sustained paroxysmal sustained paroxysmal paroxysmal sustained
+ + + +
71
male female female male female male
44 68
C, clonidine test; G, glucagon test; M, metoclopramide test: N, naloxone test. * patient no. 1 refused surgery.
808
+
Tumor
G
M N
Localization
+ + + +
+ + + +
right adrenal right adrenal right adrenal right adrenal left adrenal left adrenal
+ + + +
Size (em)
(3 X 2.5 X 2.5) (4 X 3 X 3) (6X 4 X 3) (7 X 5 X 4.2) (8 X 7 X 6)
Pharmacological test in pheochromocytoma
cooled tubes and kept on ice. The sampies were separated irnmedtately by centrifugation at 4 C and stored at -20 C until assay. Catecholamine levels were determined by high-performance liquid chromatography and the trihydroxy-indole fluorometric method, as previously described (13). The assay sensitivities were 20 pg/ml for both norepinephrine (NE) and epinephrine (E). Intra- and interassay coefficients of variation were 4.9°/0 and 7.2°/0 for NE, and 6.5°/0 and 9.2°/0 for E. Undetectable concentrations were assigned the value of the detection limit for the assay. The upper limit was defined as 3 SO above the mean (3) determined in 52 age-matched normal subjects; for NE it was 423 pg/ml while for E it was 100 pg/ml. All assays were performed within a week of sampling. Specimens from the
same patient were assayed together in the same run to avoid possible interassay variations. Statistical analysis was performed using the Kendall rankcorrelation test, the Wilcoxon signed rank test and the Kruskal-Wallis test with repeated measures followed by Dunnett's multiple range test. Results were expressed as the mean ± SO.
RESULTS Clonidine test (Table 2) Systolic and diastolic BP and pulse rate were significantly (p < 0.01) decreased after 150 IJg clonidine administration in the group of non-pheochromocytoma. In the "group of the patients without pheochrornocytorna, the plasma NE and E levels
Table2 - Cardiovascular, ptesme norepinephrine (NE), and plasma epinephrine(E) responses to oral clonidine in 19 patients without
pheochromocytoma (NPHE) and individualresponses in five patients with pheochromocytoma (PHE). Measurement Group
Before
2h
3h
SBP (mmHg) NPHE PHE
Mean ± SO Mean ± SO
156 ± 19 148 ± 35
144 ± 182 138 ± 47
142 ± 172 135 ± 20
OBP (mmHg) NPHE PHE
Mean ± SO Mean ± SO
91 ± 13 87 ± 21
78 ± 142 78 ± 27
80 ± 152 76 ± 23
PR (beats/min) NPHE PHE
Mean ± SO Mean ± SO
69± 9 64± 8
62 ± 82 63 ± 8
64 ± 92 65± 4
NE (pg/ml)
Mean ± so Mean + 3 SO
218 537
Patient (no.) 1 2 3 4 6 * Normal Mean + 3 SO
7160 491 1780 298 980 423
NPHE
± 77
120 294
± 582
125 316
± 642
Minimum at 2 or 3 h
107 271
± 552
%
-50 -108
ß
±
19
PHE
E (pg/ml) NPHE
Mean ± SO Mean + 3 SO
49 ± 21 111
4650 340 3430 250 1230
35 ± 172 86
3340 485 1600 308 1420
36 ± 162 85
3340 340 1600 250 1230
31 ± 142 72
-53 -31 -10 -16 + 26
-32 ± 26 -109
PHE Patient (no.) 1 2 3 4 6 * Normal Mean + 3 SO
2480 168 23 282 840 100
1710 115 38 263 680
SBP, systolic blood pressure; DBP, diastolic blood press ure; PR, pulse rate. *, normal subjects n = 52. 1
p< 0.05, 2 p< 0.01 vs before.
809
1340 167 38 338 830
1340 115 38 223 680
-46 -31 + 69 -21 -19
H. Koshida, I. Miyamori, R. Soma, et al.
were decreased significantly after both 2 and 3 h, with the mean ± SO of the % NE changes of the lower value of 2 or 3 h being -50 ± 19%.According to the criteria of Bravo et al. (1), a normal (negative) test result is when the lower plasma catecholamine (NE + E) level at 2 or 3 h after administration of 300
J..Lg of clonidine is less than 500 pg/ml. Cases 2 and 4 showed such negative results. The oral intake of 150 J..Lg of clonidine caused no side effects except drowsiness, in contrast to a previous report in which 300 J..Lg of clonidine was used (14). No symptomatic orthostatic hypothension was observed.
Table 3 - Cardiovascular, plasma norepinephrine (NE), and plasma epinephrine(E) responses to glucagon in 19 patients without pheochromocytoma (NPHE) and individualresponses in four patients with pheochromocytoma (PHE). Measurement Group SBP (mmHg) NPHE
Mean ± SO Mean + 3 SO
Before
Max
154 ± 19 209
180 ± 271 261
27 ± 17 77
17± 11 50
126 134 100 132
214 202 103 151
88 68 3 19
70 51 3 14
93 ± 13 131
109 ± 181 162
16 ± 11 49
18± 12 52
67 78 53 77
112 105 68 88
45 27 15 11
67 35 28 14
70 ± 7 90
82 ± 71 104
12 ±6 31
18± 10 49
75 78 59 61
103 116 65 69
28 38 6 8
37 49 10 13
0/0 ~
PHE Patient (no.) 2 3 4 5 OBP (mmHg) NPHE
Mean ± SO Mean + 3 SO
PHE Patient (no.) 2 3 4 5 PR (beats/min) NPHE
Mean ± SO Mean + 3 SO
PHE
Patient (no.) 2 3 4 5
NE (pg/ml) NPHE
Mean ± SO Mean + 3 SO
PHE Patient (no.) 2 3 4 5 * Normal Mean + 3 SO E (pg/ml) NPHE
Mean ± SO Mean + 3 SO
302 ± 1001 601
201 ± 85 456 416, 1440 538 370 423
73600 25700 692 372
96 ± 281 180
52 ± 17 102
101 ± 56 268 73200 24200 154 2
44 ± 22 110
57 ± 32 154 17600 1690 29 1
95 ± 63 282
PHE Patient (no.) 2 3
153 20
1
88700 160
58000 800
4
425
550
125
29
5 * Normal Mean + 3 SO
"440 100
470
30
7
SBP, systolic blood pressure; DBP,diastolic blood pressure; PR. pulse rate. *,
88900 180
normal subjects n = 52.
p< 0.01 vs before.
810
Pharmacological test in pheochromocytoma
Glucagon test (TabJe 3) The peak increment of BP fulfilled the criteria of Hull et al. (15) in three patients without pheochromocytoma, but not in two patients (cases 4 and 5) with pheochromocytoma. One patient (case 2) with
pheochromocytoma required an iv injection of 5 mg of phentolamine for the treatment of an abrupt rise in BP. The E levels increased significantly (p < 0.05) only at 3 min in the group of the patients without pheochromocytoma. Only two patients
Table 4 - Cardiovascular, plasma norepinephrine (NE), and plasma epinephrine(E) responses to metoclopramide in 14 patients
withoutpheochromocytoma (NPHE) and individualresponses in tour patients with pheochromocytoma (PHE). Measurement Group
Before
SBP (mmHg) NPHE
Mean ± SO Mean + 3 SO
Max
0/0
Ll
149± 19 207
156 ± 18 2 211
7±6 23
5±4 16
137 155 108 131
189 186 140 173
52 31 32 42
38 20 30 32
89 ± 18 144
92 ± 17 1 144
3±4 15
4±5 19
68 98 59 75
95 109 75 93
27 11 16 18
40 11 27 24
71 ± 9 97
78 ± 11 2 111
7±4 18
9±5 24
11 6 2
17 8 4 15
PHE Patient (no.) 2 3 4 5 OBP (mmHg) NPHE
Mean ± SO Mean + 3 SO
PHE Patient (no.) 2 3 4 5 PR (beats/ min) NPHE
Mean ± SO Mean + 3 SO
PHE Patient (no.) 2 3 4
5
65 73 56 62
Mean ± so Mean + 3 SO
192 399
Patient (no.) 2 3 4 5 * Normal Mean + 3 SO
315 1770 212 300 423
NE (pg/ml) NPHE
76 79 58 71
± 69
240 519
9
± 93 2
49 133
± 28
25 53
±9
PHE
E (pg/ml) NPHE
Mean ± SO Mean + 3 SO
6300 10400 1346 590
62 ± 29 2 149
49 ± 22 114
5990 8620 1130 290
13 ± 9 41
1900 487 535 97
25 ± 14 67
PHE Patient (no.) 2 3 4 5 * Normal Mean + 3 SO
185 25 238 210 100
5900 68 2050 1500
SBP, systolic blood pressure; OBP, diastolic blood pressure; PR, pulse rate. *, normal subjects n = 52. 1
p «; 0.05, 2 p «. 0.01 vs before.
811
5720 43 1810 1290
3090 173 761 614
H. Koshida, I. Miyamori, R. Soma, et al.
(cases no. 2 and 3) with pheochromocytoma exhibited % peak increments of NE greater than the mean + 3 SO of those of the patients without pheochromocytoma. Cases 4 and 5 fell within the mean ± 3 SO of % peak increments of E of the patients
without pheochromocytoma. In both groups, there was no significant positive correlation between the peak absolute BP and the peak absolute catecholamine levels, or between the pressor responses and the catecholamine responses. The glucagon
Table 5 - Cardiovascular, plasma norepinephrine (NE), and plasma epinephrine (E) responses to naloxone in 9 patients without pheochromocytoma (NPHE) and individualresponses in tour patients with pheochromocytoma (PHE). Measurement Group SBP (mmHg) NPHE
Mean ± SO Mean + 3 SO
Before
Max
147 ± 21 210
150 ± 21 212
3±6 22
2±4 16
127 137 97 166
133 151 105 199
6 14 8 33
5 10 8 20
90 ± 15 135
94 ± 141 136
4±5 20
5±7 25
1 7 -2 15
1 8 -3 18
5±4 17
8±6 24
3 10 0 15
15 0 28
~
0/0 ~
PHE Patient (no.) 2 3 4 6 OBP (mmHg) NPHE
Mean ± SO Mean + 3 SO
PHE Patient (no.) 2 3 4 6 PR (beats/min) NPHE
Mean ± SO Mean + 3 SO
PHE Patient (no.) 2
3 4 6 NE (pg/ml) NPHE
Mean ± SO Mean + 3 SO
PHE Patient (no.) 2 3 4 6 * Normal Mean + 3 SO E (pg/ml) NPHE
Mean ± SO Mean + 3 SO
71 98 63 97
70 91 65 82 70 ± 13 110
74 ± 142 115
71 66 58 53
74 76 58 68
208 ± 60 388
273 ± 722 490
444 ' 1730 385 1180 423
888 4250 500 2320
56 ± 17 108
82 ± 282 164
65 ± 33 165 444 2510 115 1130
26 ± 17 76
4
33 ± 16 80 100 145 30 96
48 ± 24 120
PHE Patient (no.) 2 3 4 6 * Normal Mean + 3 SO
140 20 400 749 100
330 27 625 1190
SBP, systolic blood pressure; OBP, diastolic blood pressure; PR, pulse rate. *, 1
normal subjects n
= 52.
p< 0.05, 2 p< 0.01 vs before.
812
190 7 225 441
136 35 56 59
Pharmacological test in pheochromocytoma
test was weil tolerated by all patients and caused no signiticant morbidity. Mild nausea and palpitations were experienced by tour patients without pheochromocytoma.
sponses. This test did not cause any side effects in either groups.
Metoclopramide test (Table 4) In the group of the patients without pheochromocytoma, metoclopramide caused no signiticant changes o{ systolic and diastolic BP and pulse rate at any point, although the peak values of these cardiovascular parametersdiffered signiticantlyfrom those of the baseline levels. In contrast, all patients with pheochromocytoma showed transient pressor responses which did not require iv phentolamine. While neither NE nor E levels altered signiticantly at any point in the group of the patients without pheochromocytoma, all patients with pheochromocytoma \ tested showed marked elevations of both plasma NE and E. The % oeak increments of NE and E of the patients with pheochromocytoma exceeded the mean + 3 SD of the non-pheochromocytoma group, and there was no overlap of individual % peak increment value for NE and E between the two groups. Both groups tailed to show significant positive correlations between the peak absolute BP and the peak catecholamine levels, or between the pressor responses and the catecholamine responses. No side effects were observed in the patients without pheochromocytoma, while slight headache and palpitations were experienced by two patients with pheochromocytoma. Na/axone test (Table 5) Naloxone did not induce significant changes of BP and pulse rate at any point in the group of the patients without pheochromocytoma. Only case 6 of the patients with pheochromocytoma exhiblted an obvious pressor response that was counteracted by the injection of 5 mg ot phentolamine. Naloxone tended to raise plasma catecholamine concentrations in the non-pheochromocytoma group, but these changes did not reach a statistically significant level at any point. Although three of four patients with pheochromocytoma tested showed a 0/0 peak increment of NE that was above the mean + 3 SD, that of plasma E levels did not differentiate between the two groups. In both groups, there was no significant correlation between the peak absolute BP and the peak absolute catecholamine levels. Also, there was no significant correlation between the pressor responses and the catecholamine re-
813
DISCUSSION The clonidine suppression test was originally introduced by Bravo et al. (3) as a safe and specific maneuver to differentiate the patients with pheochromocytoma, whose plasma catecholamine levels were slightly elevated, from the patients without pheochromocytoma (1 ). However, it has since been shown that misleading results can occur in the patients with pheochromocytoma whose plasma catecholamine levels are initially within the normal range (4, 5), and in the patients without pheochromocytoma who are receiving diuretics for a short term (6). We performed this test using half the recommended dose, since 300 J.Lg of clonidine produced profound and prolonged symptomatic hypotension in 2 of the 5 Japanese patients without pheochromocytoma in our pilot study. When the criteria tor the 300 J.Lg of clonidine test were applied (1), two of the 5 patients with pheochromocytoma yielded false negative results. Taking account of the different dose used and the reports of problems with this test, we performed further investigations in five patients with pheochromocytoma whose plasma catecholamine values were not pathognomonic (not over 2,000 pg/ml) (1) and whose BP were not so high as to contraindicate a provocation test (1, 2). tn provocation tests, we compared the results of the BP and the plasma catecholamine levels using the peak values, since the peak point varied from 2 to 5 min in the glucagon and metoclopramide tests and trom 5 to 20 min in the naloxone test. Although a previous report has stated that the individual measurement of plasma NE concentrations in response to glucagon predicted the presence or absence of a tumor in 27 ot 28 patients under suspicion of having pheochromocytoma (16), our results failed to demonstrate such a discriminative capacity of the glucagon test. This discrepancy may result trom the small number of patients with pheochromocytoma in both studies, or trom an inappropriate sampling time in the previous study where blood was taken at only 3 min after the administration of glucagon. Mannelli et al. (17) observed for the first time an abrupt rise in both BP and plasma catecholamines after the administration of naloxone in a patient with pheochromocytoma. Naloxone has also been
H. Koshida, I. Miyamori, R. Soma, et al.
shown to increase the area under the curve (AUC) for plasma NE and E levels in the patients with Esecreting tumors (18). Among our patients with pheochromocytoma, only case 6 showed a marked pressor response. We assessed % peak increments of NE and E levels instead of the AUC, since case 6 required treatment with phentolamine which is known to elevate the plasma catecholamine concentrations (19). The % peak increment of either catecholamine did not distinguish the patients with pheochromocytoma from the patients without pheochromocytoma. In contrast to the above two provocation tests, the metoclopramide test successtully differentiated the patients with pheochromocytoma from the patients without pheochromocytoma,as shown first by Agabiti-Rosei et al. (20). In the patients without pheochromocytoma, metoclopramide induced only a minimal change of BP and a slight increase of NE and E levels. In comparison, all the patients with pheochromocytoma showed marked pressor responses, although none of them required phentolamine. No patient tested was within 3 SO above the mean of the 0/0 peak incrernents of. the non-pheochromocytoma group. This did not coincide with the report of Kawabe et al. (21), who stated that E concentrations remained unchanged after metoclopramide in the patients with pheochromocytoma, although % peak increment of NE in the patients with pheochromocytoma differed significantly from that of the patients without pheochromocytoma. This discrepancy might be attributed to differences of the sampling times (15 and 30 min after metoclopramide in that study) or of the major catecholamines secreted from the tumors. In order to reduce the potential danger ot these provocation tests, it has been proposed that pretreatment with an a-blocking agent (22) or a calcium-channel blocker (23) could allow pressor responses to be minimized without changing plasma catecholamine responsiveness. However, further studies are necessary to assess the validity of provocation tests under such conditions. In some cases, it has been shown that pressor responses were not altered despite there being no response of the plasma catecholamine concentrations after these pretreatments (16), and a-blocking agents (24) may affect plasma catecholamine levels. In conclusion, this study suggests that the metoclopramide test with the measurement of plasma NE and E concentrations is a useful adjunctive tool
for the diagnosis of pheochromocytoma. However, this test should be performed by experienced personnel taking proper precautions to counteract pressor responses and arrhythmias, and should only be used in cases where the basal levels of catecholamines and their metabolites in the plasma and urine are normal or equivocal but a clinical suspicion of pheochromocytoma persists.
REFERENCES 1. Bravo E.L., Gifford R.W. Jr. Pheochromocytoma: diagnosis, localization and management. N. Engl. J. Med. 311: 1298, 1984.
2. Sheps S.G., Jiang N-.S., Klee G.G. Oiagnostic evaluation of pheochromocytoma. Endocrinol. Metab. Clin. North Am. 17: 397, 1988.
3. Bravo E.L., Tarazi R.C., Fouad F.M., Vidt O.G., Gifford R.W. Jr. Clonidine-suppression test. A useful aid in the diagnosis of pheochromocytoma. N. Engl. J. Med. 305: 623, 1981. 4. Taylor H.C., Mayes 0., Anton A.H. Clonidine suppression test for pheochromocytoma: examples of misleading results. J. Clin. Endocrinol. Metab. 63: 238, 1986.
5. Plewe G., Klause U., Cordes U., Beyer J. Evaluation of the clonidine-suppression test in the diagnosis of pheochromocytoma. Klin. Wochenschr. 66: 764, 1988.
6. Hui T.P., Krakoff L.R., Felton K., Yeager K. Oiuretic treatment alters clonidine suppression of plasma norepinephrine. Hypertension 8: 272, 1986.
7. Abecassis M., McLoughlin M.J., Langer B., Kudlow J.E. Serendipitous adrenal masses: prevalence, significance, and management. Am. J. Surg. 149: 783,1985. 8. Krane N.K. Clinically unsuspected pheochromocytomas. Arch. Intern. Med. 146: 54, 1986. 9. Penn 1., Moulton J., Bracken B. Diagnosis and management of adrenal masses: 1987 Du Pont lecture. Can. J. Surg. 31: 105, 1988. 10. McCorkell S.J., Niles N.L. Fine-needle aspiration of catecholamine-producing adrenal masses: a possibly fatal mistake. AJR 145: 113, 1985.
814
Pharmacological test in pheochromocytoma
11 . Cross D.A., Meyer J.S. Postoperative deaths due to unsuspected pheochromocytoma. South. Med. J. 70: 1320, 1977.
18. Mannelli M., Maggi M., Oe Feo M.L., Boscaro M., Opocher G., Mantero F., Baldi E., Giusti G. Opioid modulation of normal and pathological human chromaffin tissue. J. Clin. Endocrinol. Metab. 62: 577, 1986.
12. Sellevold O.F.M., Raeder J., Stenseth R. Undiagnosed pheochromocytoma in the perioperative period. Case reports. Acta Anaesthesiol. Scand. 29: 474, 1985.
19. Galbo H., Christensen N.J., Holst J.J. Catecholamines and pancreatic hormones during autonomic blockade in exercising man. Acta Physiol. Scand. 101: 428, 1977.
13. Hifumi S. The responses of norepinephrine, renin and aldosterone to standing in diabetic patients with orthostatic hypotension. Folia Endocrinol. Jpn. 58: 1040, 1982.
20. Agabiti-Rosei E., Alicandri C.L., Corea L. Hypertensive crisis in patients with pheochromocytoma given metoclopramide. Lancet 1: 600, 1977.
14. Burris J.F., D'Angelo L.J. Complications of clonidine suppression test for pheochromocytoma. N. Engl. J. Med. 307: 756, 1982.
21. Kawabe H., Itaya Y., Suzuki H., Kondo K., Saruta T. Metoclopramide in the diagnosis of pheochromocytoma. Jap. Heart J., 26: 557, 1985.
15. Hull R.D., Sebel E., Stokes G.S. Results of screening for pheochromocytoma in 17 hypertensive patients with episodic symptoms. Med. J. Aust. 1: 381, 1974.
22. Elliott W.J., Murphy M.B., Straus 11 F.H., Jarabak J. Improved safety of glucagon testing for pheochromocytoma by prior o-receptor blockade. Arch. Intern. Med. 149: 214, 1989.
16. Levinson P.D., Hamilton B.P., Mersey J.H., Kowarski A.A. Plasma norepinephrine and epinephrine responses to glucagon in patients with suspected pheochromocytomas. Metabolism 32: 998, 1983.
23. Favre ~., Forster A., Fathi M., Vallotton M.B. Calcium-channel inhibition in pheochromocytoma. Acta Endocrinol. (Copenh.) 113: 385, 1986. 24. Mulvihill-Wilson J., Graham R.M., PettingerW., Muckleroy C., Anderson S., Gaffney F.A., Blomqvist C.G. Comparative effects of prazosin and phenoxybenzamine on arterial blood pressure, heart rate, and plasma catecholamines in essential hypertension. J. Cardiovasc. Pharmacol. 1 (Suppl.): S1, 1979.
17. Mannelli M., Maggi M., Oe Feo M.L., Cuomo S., Forti G., Moroni F., Giusti G. Naloxone administration releases catecholamines. N. Engl. J. Med. 308: 654, 1983.
815
Evaluation of clonidine suppression and various provocation tests in the diagnosis of pheochromocytoma H. Koshida, I. Miyamori, R. Soma, T. Matsubara, S. Okamoto, M. Ikeda, and R. Takeda Second Department of lnternal Medieine, School of Medicine, Kanazawa University, Takara-machi 13-1, Kanazawa 920, Japan
ABSTRACT. Recent investigations have shown that the widely used clonidine suppression test is sometimes fallible for the diagnosis of pheochromocytoma. A comparative assessment was made of the following suppression and provocation tests, the clonidine suppression test, and the glucagon, metoclopramide, and naloxone provocation tests. The assessment was performed in 6 patients with pheochromocytoma and in 19 patients without pheochromocytoma who were initially suspected of harboring a tumor. BP response did not predict the presence of pheochromocytoma in any test. Plasma norepinephrine (NE) concentrations determined at 120 and 180 min after oral 150 JLg of clonidine gave false negative results in 2 of the 5 patients with pheochromocytoma tested. Both plasma NE and epinephrine (E) concer'ltrations were measured before and sequen-
tially after each provocative agent. Neither NE nor E responded to 1 mg of glucagon iv in 2 of the 4 patients with pheochromocytoma tested. Determination of the peak level, peak increment, and 0/0 peak increment of NE and E following 10 mg of naloxone iv did not distinguish the two groups. The % peak increments of both NE and E in all 4 patients with pheochromocytoma given 5 mg of metoclopramide iv exceeded the mean + 3 SD values of the patients without pheochromocytoma (25 + 28% for NE, and 25 + 42% for E). These results suggested that, when performed with judicious patient selection (ambiguous plasma or urinary catecholamine levels), the measurement of plasma catecholamines in response to metoclopramide can be a useful adjunctive tool in the diagnosis of pheochromocytoma.
INTRODUCTION
(1, 2). In reeent years, the clonidine suppression test has been widely used, since it avoids a pressor response and was initially thought to have a high specificity and sensitivity (1-3). However, recent experienees have indicated a number of pitfalls of this test. Both false-negative and false-positive results have oecurred in patients with small pheoehromoeytoma when the plasma eateeholamine levels were within the normal range (4, 5), and in patients without pheochromocytoma who were taking diuretics for a short term (6). Clinically silent, unsuspected adrenal masses are being increasingly detected incidentally due to the wider application of eomputerized tomography by unrelated reasons (7-10), and some of these prove to be pheochromoeytoma. It is critical, therefore, to determine whether these adrenal masses are pheochromoeytomas prior to needle aspiration biopsy (10) or surgical resection (11, 12). In this situation and when the clinical suspicion of pheochro-
Pheochromocytoma is an uncommon disease but the early diagnosis of this eatecholamine-produeing tumor is important, because it is usually eurable by surgieal reseetion and because of the risk of lethai eomplications when the diagnosis is delayed. The diagnosis is based on the elinical symptoms and a variety of biochemical procedures which indicate exeessive catecholamine production (1, 2). When the measurement of basal plasma and urinary catecholamines (metabolites) gives an equivocal result, further evaluation depends on pharmacologieal tests
Key-words: Pheochromocytoma, norepinephrine, epinephrine, clonidine, glucagon, metoclopramide, naloxone. Correspondence: Hideo Koshida, MD, 2nd Department of Internal Medicine, School of Medicine, Kanazawa University, Takara-machi 13-1, Kanazawa 920, Japan. Received December 20, 1989; accepted July 12, 1990.
807
H. Koshida, I. Miyamori, R. Soma, et al.
mocytoma persists, the only useful diagnostic rnaneuver is a provocation test (1, 2). To evaluate the reliability ot three provocation tests, the glucagon, metoclopramide, and naloxone tests, we performed these tests with meticulous care and compared the results with the clonidine test.
containing 200 meq of Na' and 60 meq of K+ per day. Antihypertensive medications were withdrawn at least 2 weeks before the study. No food, tea, coffee, or smoking were allowed for at least 12 h preceding the tests. The clonidine suppression test was performed first, followed by the glucagon, metoclopramide, and naloxone provocation tests, in a random order on separate days. Every test was begun between 08:00 hand 09:00 h. After insertion of two catheters into bilateral forearm veins and at least 30 min of rest in the supine position in a warm, quiet, and well-lighted room, two blood sam pies were drawn at 5-min intervals for plasma catecholamine measurement. Control measurements of BP and pulse rate were made three times at 5-min intervals. The patients then took 150 J.lg of clonidine orally with 100 ml of tap water, or received an iv bolus injection of 1 mg of glucagon, 5 mg of metoclopramide, or 10 mg of naloxone. Thereafter, in the clonidine test, BP and pulse rate were measured at 30-min intervals by an automated sphygmomanometer (BF-203, Nippon Colin Co. Ltd., Tokyo) and blood sampies were taken at 120 and 180 min. In the glucagon and metoclopramide tests, BP and pulse rate were measured at 1-min intervals for 15 min, and blood sampies were drawn at 1, 2, 3, 5, 10, and 15 min. In the naloxone test, BP and pulse rate were recorded at 5-min intervals for 30 min and blood sam pies were drawn at 5, 10, 15, 20, and 30 min. There was no pressor response to saline in any patient with pheochromocytoma and in the 8 selected patients without pheochromocytoma before the provocation tests were commenced. Patients remained supine throughout the study period. Phentolamine was always available at the bed side for immediate use. Blood sampies for plasma catecholamine determination were collected into pre-
MATERIALS AND METHODS We studied six patients with pheochromocytoma, in five of whom the diagnosis was confirmed by histopathological examination after surgery. Their clinical data are shown in Table 1. Case no. 1 refused an operation but the presence of pheochromocytoma was later confirmed by extremely elevated urinary and plasma catecholamine levels, computerized tomography, and 1311-metaiodobenzylguanidine (MIBG) scintigraphy. Nineteen additional patients (11 men and 8 women, mean age 53 yr, range 3569 yr), who were initially suspected of harboring a pheochromocytoma because of anxiety, headache, palpitations, diaphoresis, and vasomotor lability, were also enrolled in this study. In these patients, both pheochrömocytoma and secondary hypertension were ruled out by clinical examination and the following investigations: urinalysis, measurement of serum electrolytes, creatinine, BUN, creatinine clearance, urinary catecholamines, PRA, plasma aldosterone concentration, and intravenous pyelography. In selected cases, renal arteriography, computerized tomography of the adrenal glands, and 131 1_ MIBG scintigraphy were also performed. None of the patients without pheochromocytoma underwent surgical exploration. Informed consent was obtained from all the patients enrolled in this study according to the principles of the Declaration of Helsinki. All the patients were hospitalized and fed a diet
Table 1 - Profiles of the pheochromocytoma pa tients. Tests performed Patient (no.)
Age (yr)
Sex
Hypertension
C
1 2
58 59
3
51
4 5 6
sustained paroxysmal sustained paroxysmal paroxysmal sustained
+ + + +
71
male female female male female male
44 68
C, clonidine test; G, glucagon test; M, metoclopramide test: N, naloxone test. * patient no. 1 refused surgery.
808
+
Tumor
G
M N
Localization
+ + + +
+ + + +
right adrenal right adrenal right adrenal right adrenal left adrenal left adrenal
+ + + +
Size (em)
(3 X 2.5 X 2.5) (4 X 3 X 3) (6X 4 X 3) (7 X 5 X 4.2) (8 X 7 X 6)
Pharmacological test in pheochromocytoma
cooled tubes and kept on ice. The sampies were separated irnmedtately by centrifugation at 4 C and stored at -20 C until assay. Catecholamine levels were determined by high-performance liquid chromatography and the trihydroxy-indole fluorometric method, as previously described (13). The assay sensitivities were 20 pg/ml for both norepinephrine (NE) and epinephrine (E). Intra- and interassay coefficients of variation were 4.9°/0 and 7.2°/0 for NE, and 6.5°/0 and 9.2°/0 for E. Undetectable concentrations were assigned the value of the detection limit for the assay. The upper limit was defined as 3 SO above the mean (3) determined in 52 age-matched normal subjects; for NE it was 423 pg/ml while for E it was 100 pg/ml. All assays were performed within a week of sampling. Specimens from the
same patient were assayed together in the same run to avoid possible interassay variations. Statistical analysis was performed using the Kendall rankcorrelation test, the Wilcoxon signed rank test and the Kruskal-Wallis test with repeated measures followed by Dunnett's multiple range test. Results were expressed as the mean ± SO.
RESULTS Clonidine test (Table 2) Systolic and diastolic BP and pulse rate were significantly (p < 0.01) decreased after 150 IJg clonidine administration in the group of non-pheochromocytoma. In the "group of the patients without pheochrornocytorna, the plasma NE and E levels
Table2 - Cardiovascular, ptesme norepinephrine (NE), and plasma epinephrine(E) responses to oral clonidine in 19 patients without
pheochromocytoma (NPHE) and individualresponses in five patients with pheochromocytoma (PHE). Measurement Group
Before
2h
3h
SBP (mmHg) NPHE PHE
Mean ± SO Mean ± SO
156 ± 19 148 ± 35
144 ± 182 138 ± 47
142 ± 172 135 ± 20
OBP (mmHg) NPHE PHE
Mean ± SO Mean ± SO
91 ± 13 87 ± 21
78 ± 142 78 ± 27
80 ± 152 76 ± 23
PR (beats/min) NPHE PHE
Mean ± SO Mean ± SO
69± 9 64± 8
62 ± 82 63 ± 8
64 ± 92 65± 4
NE (pg/ml)
Mean ± so Mean + 3 SO
218 537
Patient (no.) 1 2 3 4 6 * Normal Mean + 3 SO
7160 491 1780 298 980 423
NPHE
± 77
120 294
± 582
125 316
± 642
Minimum at 2 or 3 h
107 271
± 552
%
-50 -108
ß
±
19
PHE
E (pg/ml) NPHE
Mean ± SO Mean + 3 SO
49 ± 21 111
4650 340 3430 250 1230
35 ± 172 86
3340 485 1600 308 1420
36 ± 162 85
3340 340 1600 250 1230
31 ± 142 72
-53 -31 -10 -16 + 26
-32 ± 26 -109
PHE Patient (no.) 1 2 3 4 6 * Normal Mean + 3 SO
2480 168 23 282 840 100
1710 115 38 263 680
SBP, systolic blood pressure; DBP, diastolic blood press ure; PR, pulse rate. *, normal subjects n = 52. 1
p< 0.05, 2 p< 0.01 vs before.
809
1340 167 38 338 830
1340 115 38 223 680
-46 -31 + 69 -21 -19
H. Koshida, I. Miyamori, R. Soma, et al.
were decreased significantly after both 2 and 3 h, with the mean ± SO of the % NE changes of the lower value of 2 or 3 h being -50 ± 19%.According to the criteria of Bravo et al. (1), a normal (negative) test result is when the lower plasma catecholamine (NE + E) level at 2 or 3 h after administration of 300
J..Lg of clonidine is less than 500 pg/ml. Cases 2 and 4 showed such negative results. The oral intake of 150 J..Lg of clonidine caused no side effects except drowsiness, in contrast to a previous report in which 300 J..Lg of clonidine was used (14). No symptomatic orthostatic hypothension was observed.
Table 3 - Cardiovascular, plasma norepinephrine (NE), and plasma epinephrine(E) responses to glucagon in 19 patients without pheochromocytoma (NPHE) and individualresponses in four patients with pheochromocytoma (PHE). Measurement Group SBP (mmHg) NPHE
Mean ± SO Mean + 3 SO
Before
Max
154 ± 19 209
180 ± 271 261
27 ± 17 77
17± 11 50
126 134 100 132
214 202 103 151
88 68 3 19
70 51 3 14
93 ± 13 131
109 ± 181 162
16 ± 11 49
18± 12 52
67 78 53 77
112 105 68 88
45 27 15 11
67 35 28 14
70 ± 7 90
82 ± 71 104
12 ±6 31
18± 10 49
75 78 59 61
103 116 65 69
28 38 6 8
37 49 10 13
0/0 ~
PHE Patient (no.) 2 3 4 5 OBP (mmHg) NPHE
Mean ± SO Mean + 3 SO
PHE Patient (no.) 2 3 4 5 PR (beats/min) NPHE
Mean ± SO Mean + 3 SO
PHE
Patient (no.) 2 3 4 5
NE (pg/ml) NPHE
Mean ± SO Mean + 3 SO
PHE Patient (no.) 2 3 4 5 * Normal Mean + 3 SO E (pg/ml) NPHE
Mean ± SO Mean + 3 SO
302 ± 1001 601
201 ± 85 456 416, 1440 538 370 423
73600 25700 692 372
96 ± 281 180
52 ± 17 102
101 ± 56 268 73200 24200 154 2
44 ± 22 110
57 ± 32 154 17600 1690 29 1
95 ± 63 282
PHE Patient (no.) 2 3
153 20
1
88700 160
58000 800
4
425
550
125
29
5 * Normal Mean + 3 SO
"440 100
470
30
7
SBP, systolic blood pressure; DBP,diastolic blood pressure; PR. pulse rate. *,
88900 180
normal subjects n = 52.
p< 0.01 vs before.
810
Pharmacological test in pheochromocytoma
Glucagon test (TabJe 3) The peak increment of BP fulfilled the criteria of Hull et al. (15) in three patients without pheochromocytoma, but not in two patients (cases 4 and 5) with pheochromocytoma. One patient (case 2) with
pheochromocytoma required an iv injection of 5 mg of phentolamine for the treatment of an abrupt rise in BP. The E levels increased significantly (p < 0.05) only at 3 min in the group of the patients without pheochromocytoma. Only two patients
Table 4 - Cardiovascular, plasma norepinephrine (NE), and plasma epinephrine(E) responses to metoclopramide in 14 patients
withoutpheochromocytoma (NPHE) and individualresponses in tour patients with pheochromocytoma (PHE). Measurement Group
Before
SBP (mmHg) NPHE
Mean ± SO Mean + 3 SO
Max
0/0
Ll
149± 19 207
156 ± 18 2 211
7±6 23
5±4 16
137 155 108 131
189 186 140 173
52 31 32 42
38 20 30 32
89 ± 18 144
92 ± 17 1 144
3±4 15
4±5 19
68 98 59 75
95 109 75 93
27 11 16 18
40 11 27 24
71 ± 9 97
78 ± 11 2 111
7±4 18
9±5 24
11 6 2
17 8 4 15
PHE Patient (no.) 2 3 4 5 OBP (mmHg) NPHE
Mean ± SO Mean + 3 SO
PHE Patient (no.) 2 3 4 5 PR (beats/ min) NPHE
Mean ± SO Mean + 3 SO
PHE Patient (no.) 2 3 4
5
65 73 56 62
Mean ± so Mean + 3 SO
192 399
Patient (no.) 2 3 4 5 * Normal Mean + 3 SO
315 1770 212 300 423
NE (pg/ml) NPHE
76 79 58 71
± 69
240 519
9
± 93 2
49 133
± 28
25 53
±9
PHE
E (pg/ml) NPHE
Mean ± SO Mean + 3 SO
6300 10400 1346 590
62 ± 29 2 149
49 ± 22 114
5990 8620 1130 290
13 ± 9 41
1900 487 535 97
25 ± 14 67
PHE Patient (no.) 2 3 4 5 * Normal Mean + 3 SO
185 25 238 210 100
5900 68 2050 1500
SBP, systolic blood pressure; OBP, diastolic blood pressure; PR, pulse rate. *, normal subjects n = 52. 1
p «; 0.05, 2 p «. 0.01 vs before.
811
5720 43 1810 1290
3090 173 761 614
H. Koshida, I. Miyamori, R. Soma, et al.
(cases no. 2 and 3) with pheochromocytoma exhibited % peak increments of NE greater than the mean + 3 SO of those of the patients without pheochromocytoma. Cases 4 and 5 fell within the mean ± 3 SO of % peak increments of E of the patients
without pheochromocytoma. In both groups, there was no significant positive correlation between the peak absolute BP and the peak absolute catecholamine levels, or between the pressor responses and the catecholamine responses. The glucagon
Table 5 - Cardiovascular, plasma norepinephrine (NE), and plasma epinephrine (E) responses to naloxone in 9 patients without pheochromocytoma (NPHE) and individualresponses in tour patients with pheochromocytoma (PHE). Measurement Group SBP (mmHg) NPHE
Mean ± SO Mean + 3 SO
Before
Max
147 ± 21 210
150 ± 21 212
3±6 22
2±4 16
127 137 97 166
133 151 105 199
6 14 8 33
5 10 8 20
90 ± 15 135
94 ± 141 136
4±5 20
5±7 25
1 7 -2 15
1 8 -3 18
5±4 17
8±6 24
3 10 0 15
15 0 28
~
0/0 ~
PHE Patient (no.) 2 3 4 6 OBP (mmHg) NPHE
Mean ± SO Mean + 3 SO
PHE Patient (no.) 2 3 4 6 PR (beats/min) NPHE
Mean ± SO Mean + 3 SO
PHE Patient (no.) 2
3 4 6 NE (pg/ml) NPHE
Mean ± SO Mean + 3 SO
PHE Patient (no.) 2 3 4 6 * Normal Mean + 3 SO E (pg/ml) NPHE
Mean ± SO Mean + 3 SO
71 98 63 97
70 91 65 82 70 ± 13 110
74 ± 142 115
71 66 58 53
74 76 58 68
208 ± 60 388
273 ± 722 490
444 ' 1730 385 1180 423
888 4250 500 2320
56 ± 17 108
82 ± 282 164
65 ± 33 165 444 2510 115 1130
26 ± 17 76
4
33 ± 16 80 100 145 30 96
48 ± 24 120
PHE Patient (no.) 2 3 4 6 * Normal Mean + 3 SO
140 20 400 749 100
330 27 625 1190
SBP, systolic blood pressure; OBP, diastolic blood pressure; PR, pulse rate. *, 1
normal subjects n
= 52.
p< 0.05, 2 p< 0.01 vs before.
812
190 7 225 441
136 35 56 59
Pharmacological test in pheochromocytoma
test was weil tolerated by all patients and caused no signiticant morbidity. Mild nausea and palpitations were experienced by tour patients without pheochromocytoma.
sponses. This test did not cause any side effects in either groups.
Metoclopramide test (Table 4) In the group of the patients without pheochromocytoma, metoclopramide caused no signiticant changes o{ systolic and diastolic BP and pulse rate at any point, although the peak values of these cardiovascular parametersdiffered signiticantlyfrom those of the baseline levels. In contrast, all patients with pheochromocytoma showed transient pressor responses which did not require iv phentolamine. While neither NE nor E levels altered signiticantly at any point in the group of the patients without pheochromocytoma, all patients with pheochromocytoma \ tested showed marked elevations of both plasma NE and E. The % oeak increments of NE and E of the patients with pheochromocytoma exceeded the mean + 3 SD of the non-pheochromocytoma group, and there was no overlap of individual % peak increment value for NE and E between the two groups. Both groups tailed to show significant positive correlations between the peak absolute BP and the peak catecholamine levels, or between the pressor responses and the catecholamine responses. No side effects were observed in the patients without pheochromocytoma, while slight headache and palpitations were experienced by two patients with pheochromocytoma. Na/axone test (Table 5) Naloxone did not induce significant changes of BP and pulse rate at any point in the group of the patients without pheochromocytoma. Only case 6 of the patients with pheochromocytoma exhiblted an obvious pressor response that was counteracted by the injection of 5 mg ot phentolamine. Naloxone tended to raise plasma catecholamine concentrations in the non-pheochromocytoma group, but these changes did not reach a statistically significant level at any point. Although three of four patients with pheochromocytoma tested showed a 0/0 peak increment of NE that was above the mean + 3 SD, that of plasma E levels did not differentiate between the two groups. In both groups, there was no significant correlation between the peak absolute BP and the peak absolute catecholamine levels. Also, there was no significant correlation between the pressor responses and the catecholamine re-
813
DISCUSSION The clonidine suppression test was originally introduced by Bravo et al. (3) as a safe and specific maneuver to differentiate the patients with pheochromocytoma, whose plasma catecholamine levels were slightly elevated, from the patients without pheochromocytoma (1 ). However, it has since been shown that misleading results can occur in the patients with pheochromocytoma whose plasma catecholamine levels are initially within the normal range (4, 5), and in the patients without pheochromocytoma who are receiving diuretics for a short term (6). We performed this test using half the recommended dose, since 300 J.Lg of clonidine produced profound and prolonged symptomatic hypotension in 2 of the 5 Japanese patients without pheochromocytoma in our pilot study. When the criteria tor the 300 J.Lg of clonidine test were applied (1), two of the 5 patients with pheochromocytoma yielded false negative results. Taking account of the different dose used and the reports of problems with this test, we performed further investigations in five patients with pheochromocytoma whose plasma catecholamine values were not pathognomonic (not over 2,000 pg/ml) (1) and whose BP were not so high as to contraindicate a provocation test (1, 2). tn provocation tests, we compared the results of the BP and the plasma catecholamine levels using the peak values, since the peak point varied from 2 to 5 min in the glucagon and metoclopramide tests and trom 5 to 20 min in the naloxone test. Although a previous report has stated that the individual measurement of plasma NE concentrations in response to glucagon predicted the presence or absence of a tumor in 27 ot 28 patients under suspicion of having pheochromocytoma (16), our results failed to demonstrate such a discriminative capacity of the glucagon test. This discrepancy may result trom the small number of patients with pheochromocytoma in both studies, or trom an inappropriate sampling time in the previous study where blood was taken at only 3 min after the administration of glucagon. Mannelli et al. (17) observed for the first time an abrupt rise in both BP and plasma catecholamines after the administration of naloxone in a patient with pheochromocytoma. Naloxone has also been
H. Koshida, I. Miyamori, R. Soma, et al.
shown to increase the area under the curve (AUC) for plasma NE and E levels in the patients with Esecreting tumors (18). Among our patients with pheochromocytoma, only case 6 showed a marked pressor response. We assessed % peak increments of NE and E levels instead of the AUC, since case 6 required treatment with phentolamine which is known to elevate the plasma catecholamine concentrations (19). The % peak increment of either catecholamine did not distinguish the patients with pheochromocytoma from the patients without pheochromocytoma. In contrast to the above two provocation tests, the metoclopramide test successtully differentiated the patients with pheochromocytoma from the patients without pheochromocytoma,as shown first by Agabiti-Rosei et al. (20). In the patients without pheochromocytoma, metoclopramide induced only a minimal change of BP and a slight increase of NE and E levels. In comparison, all the patients with pheochromocytoma showed marked pressor responses, although none of them required phentolamine. No patient tested was within 3 SO above the mean of the 0/0 peak incrernents of. the non-pheochromocytoma group. This did not coincide with the report of Kawabe et al. (21), who stated that E concentrations remained unchanged after metoclopramide in the patients with pheochromocytoma, although % peak increment of NE in the patients with pheochromocytoma differed significantly from that of the patients without pheochromocytoma. This discrepancy might be attributed to differences of the sampling times (15 and 30 min after metoclopramide in that study) or of the major catecholamines secreted from the tumors. In order to reduce the potential danger ot these provocation tests, it has been proposed that pretreatment with an a-blocking agent (22) or a calcium-channel blocker (23) could allow pressor responses to be minimized without changing plasma catecholamine responsiveness. However, further studies are necessary to assess the validity of provocation tests under such conditions. In some cases, it has been shown that pressor responses were not altered despite there being no response of the plasma catecholamine concentrations after these pretreatments (16), and a-blocking agents (24) may affect plasma catecholamine levels. In conclusion, this study suggests that the metoclopramide test with the measurement of plasma NE and E concentrations is a useful adjunctive tool
for the diagnosis of pheochromocytoma. However, this test should be performed by experienced personnel taking proper precautions to counteract pressor responses and arrhythmias, and should only be used in cases where the basal levels of catecholamines and their metabolites in the plasma and urine are normal or equivocal but a clinical suspicion of pheochromocytoma persists.
REFERENCES 1. Bravo E.L., Gifford R.W. Jr. Pheochromocytoma: diagnosis, localization and management. N. Engl. J. Med. 311: 1298, 1984.
2. Sheps S.G., Jiang N-.S., Klee G.G. Oiagnostic evaluation of pheochromocytoma. Endocrinol. Metab. Clin. North Am. 17: 397, 1988.
3. Bravo E.L., Tarazi R.C., Fouad F.M., Vidt O.G., Gifford R.W. Jr. Clonidine-suppression test. A useful aid in the diagnosis of pheochromocytoma. N. Engl. J. Med. 305: 623, 1981. 4. Taylor H.C., Mayes 0., Anton A.H. Clonidine suppression test for pheochromocytoma: examples of misleading results. J. Clin. Endocrinol. Metab. 63: 238, 1986.
5. Plewe G., Klause U., Cordes U., Beyer J. Evaluation of the clonidine-suppression test in the diagnosis of pheochromocytoma. Klin. Wochenschr. 66: 764, 1988.
6. Hui T.P., Krakoff L.R., Felton K., Yeager K. Oiuretic treatment alters clonidine suppression of plasma norepinephrine. Hypertension 8: 272, 1986.
7. Abecassis M., McLoughlin M.J., Langer B., Kudlow J.E. Serendipitous adrenal masses: prevalence, significance, and management. Am. J. Surg. 149: 783,1985. 8. Krane N.K. Clinically unsuspected pheochromocytomas. Arch. Intern. Med. 146: 54, 1986. 9. Penn 1., Moulton J., Bracken B. Diagnosis and management of adrenal masses: 1987 Du Pont lecture. Can. J. Surg. 31: 105, 1988. 10. McCorkell S.J., Niles N.L. Fine-needle aspiration of catecholamine-producing adrenal masses: a possibly fatal mistake. AJR 145: 113, 1985.
814
Pharmacological test in pheochromocytoma
11 . Cross D.A., Meyer J.S. Postoperative deaths due to unsuspected pheochromocytoma. South. Med. J. 70: 1320, 1977.
18. Mannelli M., Maggi M., Oe Feo M.L., Boscaro M., Opocher G., Mantero F., Baldi E., Giusti G. Opioid modulation of normal and pathological human chromaffin tissue. J. Clin. Endocrinol. Metab. 62: 577, 1986.
12. Sellevold O.F.M., Raeder J., Stenseth R. Undiagnosed pheochromocytoma in the perioperative period. Case reports. Acta Anaesthesiol. Scand. 29: 474, 1985.
19. Galbo H., Christensen N.J., Holst J.J. Catecholamines and pancreatic hormones during autonomic blockade in exercising man. Acta Physiol. Scand. 101: 428, 1977.
13. Hifumi S. The responses of norepinephrine, renin and aldosterone to standing in diabetic patients with orthostatic hypotension. Folia Endocrinol. Jpn. 58: 1040, 1982.
20. Agabiti-Rosei E., Alicandri C.L., Corea L. Hypertensive crisis in patients with pheochromocytoma given metoclopramide. Lancet 1: 600, 1977.
14. Burris J.F., D'Angelo L.J. Complications of clonidine suppression test for pheochromocytoma. N. Engl. J. Med. 307: 756, 1982.
21. Kawabe H., Itaya Y., Suzuki H., Kondo K., Saruta T. Metoclopramide in the diagnosis of pheochromocytoma. Jap. Heart J., 26: 557, 1985.
15. Hull R.D., Sebel E., Stokes G.S. Results of screening for pheochromocytoma in 17 hypertensive patients with episodic symptoms. Med. J. Aust. 1: 381, 1974.
22. Elliott W.J., Murphy M.B., Straus 11 F.H., Jarabak J. Improved safety of glucagon testing for pheochromocytoma by prior o-receptor blockade. Arch. Intern. Med. 149: 214, 1989.
16. Levinson P.D., Hamilton B.P., Mersey J.H., Kowarski A.A. Plasma norepinephrine and epinephrine responses to glucagon in patients with suspected pheochromocytomas. Metabolism 32: 998, 1983.
23. Favre ~., Forster A., Fathi M., Vallotton M.B. Calcium-channel inhibition in pheochromocytoma. Acta Endocrinol. (Copenh.) 113: 385, 1986. 24. Mulvihill-Wilson J., Graham R.M., PettingerW., Muckleroy C., Anderson S., Gaffney F.A., Blomqvist C.G. Comparative effects of prazosin and phenoxybenzamine on arterial blood pressure, heart rate, and plasma catecholamines in essential hypertension. J. Cardiovasc. Pharmacol. 1 (Suppl.): S1, 1979.
17. Mannelli M., Maggi M., Oe Feo M.L., Cuomo S., Forti G., Moroni F., Giusti G. Naloxone administration releases catecholamines. N. Engl. J. Med. 308: 654, 1983.
815
