2019 Volume 66 Issue 2 Pages 187-192
Glucose intolerance is often observed in patients with pheochromocytoma. However, it remains controversial issue that glucose intolerance on pheochromocytoma is caused by impaired insulin secretion and/or by increased insulin resistance. We aimed to reveal the mechanism of glucose intolerance on pheochromocytoma with regard to the type and amount of catecholamines released. We evaluated 12 individuals diagnosed with pheochromocytoma and who underwent surgery to remove it. We examined glycemic parameters before and after surgery and investigated the association between the change of parameters of insulin secretion (homeostasis model assessment of β-cell function (HOMA-β)), insulin resistance (homeostasis model assessment of insulin resistance (HOMA-IR)) and that of urinary levels of metanephrine/normetanephrine before and after surgery. Overall, fasting plasma glucose, glycated hemoglobin (HbA1c), HOMA-β, and HOMA-IR were improved significantly after surgery. Regression analysis showed that the improvement in HOMA-β from before to after surgery was significantly positively associated with an improvement in urinary levels of metanephrine from before to after surgery and showed a significantly negative association with improvement in urinary levels of normetanephrine from before to after surgery. The improvement in HOMA-IR from before to after surgery was significantly positively associated with an improvement in urinary levels of normetanephrine from before to after surgery. Our results showed that pheochromocytoma extirpation improved glycemic parameters. Furthermore, the different effects elicited by excess amounts of adrenaline and noradrenaline on glucose intolerance were demonstrated.
PHEOCHROMOCYTOMA is a neuroendocrine tumor that produces catecholamines [1], and its clinical complications are variable [2]. Hypertension is most common, but glucose intolerance is also found in 25–75% of patients with pheochromocytoma [3, 4]. The causes of glucose intolerance in patients with pheochromocytoma have been reported to be impaired secretion of insulin and/or increased insulin resistance [5-10]. With regard to the effects of catecholamines on glucose intolerance, studies have shown that impaired secretion of insulin is mediated through the α2 receptors of β-cells in pancreatic islets [11], whereas increased insulin resistance is mediated through α1 and β receptors [7, 9].
Studies on differences in the mechanism of action between adrenaline and noradrenaline with regard to glucose intolerance in patients with pheochromocytoma are lacking. In the present study, we demonstrated that different types of catecholamines induce different effects on glucose intolerance in patients with pheochromocytoma.
All participants provided written informed consent for study inclusion. The study protocol was approved by the Ethics Review Committee of Fukuoka University (UMIN no. 000034090).
We recruited 12 individuals with pheochromocytoma who were admitted, diagnosed and underwent surgery at Fukuoka University Chikushi Hospital (Fukuoka, Japan) between April 2014 and March 2018. Pheochromocytoma was diagnosed because patients had increased plasma/urinary levels of catecholamines, increased urinary concentrations of metanephrine/normetanephrine and catecholamine metabolites, and their tumor had been identified using 123I-metaiodobenzylguanidine scintigraphy, as described previously [12]. After the diagnosis, patients commenced treatment with an oral alpha-blocker, which stopped on the day of surgery. The pathology of extirpated tumors was evaluated.
MethodsWe collected data on age, sex, diameter/laterality of the tumor, body mass index (BMI), glycemic parameters, systolic/diastolic blood pressure, and urinary levels of catecholamines and metanephrine/normetanephrine from all patients. With respect to glycemic parameters, we measured concentrations of fasting plasma glucose (FPG), glycated hemoglobin (HbA1c), homeostasis model assessment of β-cell function (HOMA-β), and homeostasis model assessment of insulin resistance (HOMA-IR) before and after surgery.
HOMA-IR was calculated using the following formula:
HOMA-IR = FPG × fasting insulin/405
HOMA-β was calculated using the following formula:
HOMA-β = 360 × fasting insulin/(FPG – 63)
Blood samples were obtained after an overnight fast. We investigated glycemic parameters except HbA1c were measured 1 day after hospital admission and 1 week after surgery. HbA1c levels were measured upon hospital admission and 1 month after surgery. Furthermore, to assess the glycemic parameters of patients taking hypoglycemic agents, we stopped treatment before surgery and restarted the same after hospital admission and 1 week after surgery. All patients with medical treatment for hypertension stopped on the day of surgery. Urinary levels of catecholamines and metanephrine/normetanephrine over 24 h were measured 1 day after hospital admission and 1 week after surgery. We analysed changes in glycemic parameters, urinary catecholamine levels, urinary levels of metanephrine/normetanephrine, and systolic/diastolic blood pressure before and after pheochromocytoma extirpation. Furthermore, we investigated relationship between improvement in HOMA-β/IR and improvement in urinary levels of metanephrine/normetanephrine with regression analysis after Spearman’s signed rank correlation test which revealed whether there was enough correlation to be analysed.
Statistical analysesData were shown as the median and internal quartile range (IQR). The significance of differences between median values was estimated by the Wilcoxon signed-rank test. Relationship was examined using Spearman’s signed rank correlation test regression analysis. Statistical analyses were performed using STATA® SE version 13.1 (Stata Corporation, College Station, TX, USA). P < 0.05 (*) was considered significant.
Table 1 shows the clinical characteristics of the patients in our study. Of the twelve patients, the median age was 67.0 (internal quartile range (IQR), 53–75) years. A tumor in the right adrenal glands was found in four patients. A tumor in the left adrenal glands was found in eight patients. The BMI and tumor diameter were 22.2 (IQR, 18.3–24.0) kg/m2 and 21.5 (IQR, 9–34) mm, respectively. Eight patients had hypertension and six patients were medicated with antihypertensive agents. Calcium blockers were only administered of all patients medicated to hypertension. Seven patients had dyslipidemia and six were medicated with lipid-lowering agents. Seven patients had diabetes mellitus and four patients were medicated with oral anti-diabetic agents, but no patients were medicated with insulin, which suggested that insulin secretion was preserved in all patients.
| Age (years) | 67.0 (53–75) |
| Male/female | 7/5 |
| Body mass index (kg/m2) | 22.2 (18.3–24.0) |
| Tumor laterality (right/left) | 4/8 |
| Tumor diameter (mm) | 21.5 (9–34) |
| Systolic blood pressure (mmHg) | 136.5 (122–139) |
| Diastolic blood pressure (mmHg) | 78.0 (66–82) |
| Fasting plasma glucose (mg/dL) | 106.5 (92–123) |
| Glycated hemoglobin (%) | 6.10 (5.6–7.1) |
| Low-density lipoprotein-cholesterol (mg/dL) | 99.0 (82–127) |
| High-density lipoprotein-cholesterol (mg/dL) | 57.0 (40–73) |
| Triglyceride (mg/dL) | 77.0 (56–122) |
| Aspartate transaminase (U/L) | 22.0 (17–25) |
| Alanine aminotransferase (U/L) | 14.0 (11–25) |
| γ-glutamyl transferase (U/L) | 23.5 (14–68) |
| Estimated glomerular filtration rate (mL/min/1.73 m2) | 78.5 (54.6–83.5) |
| Morbidity due to diabetes mellitus | 7 |
| Medical treatment for diabetes mellitus | 4 |
| Morbidity due to hypertension | 8 |
| Medical treatment for hypertension | 6 |
| Morbidity due to dyslipidemia | 7 |
| Medical treatment for dyslipidemia | 6 |
| Detail of medication for diabetes mellitus | |
| Only dipeptidyl peptidase-4 inhibitor | 2 |
| Dipeptidyl peptidase-4 inhibitor and biguanide | 1 |
| Sulfonylurea + dipeptidyl peptidase-4 inhibitor | 1 |
| Detail of medication for hypertension | |
| Only calcium blocker | 6 |
| Detail of medication for dyslipidemia | |
| Only statin | 6 |
Data are shown as the median (internal quartile range (IQR)).
Table 2 shows the glycemic parameters, urinary levels of catecholamines, and urinary levels of metanephrine/normetanephrine before and after surgery. The FPG level was reduced significantly by surgery (106.5 (IQR, 92–123) vs. 93.0 (IQR, 80–99) mg/dL) (p = 0.0022), as was the HbA1c level (6.10 (IQR, 5.6–7.1) vs. 5.65 (IQR, 5.2–6.0) %) (p = 0.0033). HOMA-β was increased significantly by surgery (45.7 (IQR, 22.3–78.0) vs. 76.2 (IQR, 39.5–87.1)) (p = 0.0499). HOMA-IR was reduced significantly by surgery (1.52 (IQR, 0.97–1.98) vs. 1.33 (IQR, 0.79–1.74)) (p = 0.0342). Urinary levels of metanephrine and normetanephrine were decreased significantly by surgery (0.19 (IQR, 0.09–43) vs. 0.05 (IQR, 0.05–0.08), p = 0.0022; 0.40 (IQR, 0.30–53) vs. 0.22 (IQR, 0.15–0.23), p = 0.0047, respectively). Table 3 shows systolic/diastolic blood pressure before and 1 week after surgery. Overall, as for systolic/diastolic blood pressure measured before and 1 week after surgery, there were not significant differences.
| Pre-tumor extirpation | Post-tumor extirpation | p | |
|---|---|---|---|
| FPG (mg/dL) | 106.5 (92–123) | 93.0 (80–99) | 0.0022* |
| Glycated hemoglobin (%) | 6.10 (5.6–7.1) | 5.65 (5.2–6.0) | 0.0033* |
| HOMA-β | 45.7 (22.3–78.0) | 76.2 (39.5–87.1) | 0.0499* |
| HOMA-IR | 1.52 (0.97–1.98) | 1.33 (0.79–1.74) | 0.0342* |
| Urinary adrenaline (μg/day) | 22.3 (6.4–35.9) | 5.1 (3.6–7.4) | 0.0022* |
| Urinary noradrenaline (μg/day) | 187.6 (119.3–276.0) | 91.8 (64.6–107.7) | 0.0029* |
| Urinary dopamine (μg/day) | 715.8 (493.3–798.7) | 602.1 (507.5–652.3) | 0.0342* |
| Urinary metanephrine (mg/day) | 0.19 (0.09–0.43) | 0.05 (0.05–0.08) | 0.0022* |
| Urinary normetanephrine (mg/day) | 0.40 (0.30–0.53) | 0.22 (0.15–0.23) | 0.0047* |
Data are shown as the median (internal quartile range (IQR)). The significance of differences between mean values was estimated by the Wilcoxon signed rank test. P < 0.05 (*) was considered significant. FPG: fasting plasma glucose, HOMA-β, homeostasis model assessment of β-cell function; HOMA-IR, homeostasis model assessment of insulin resistance.
| Pre-tumor extirpation | Post-tumor extirpation | p | |
|---|---|---|---|
| All patients (n = 12) | |||
| Systolic blood pressure (mmHg) | 136.5 (122–139) | 130.0 (125–132) | 0.4556 |
| Diastolic blood pressure (mmHg) | 78.0 (66–82) | 77.0 (70–80) | 0.6092 |
| Normotensive patients (n = 4) | |||
| Systolic blood pressure (mmHg) | 123.0 (92–133) | 120.0 (102–126) | 0.7150 |
| Diastolic blood pressure (mmHg) | 75.0 (55–78) | 75.0 (68–80) | 0.7150 |
| Hypertensive patients without medication before tumor extirpation (n = 2) | |||
| Systolic blood pressure (mmHg) | 136.5 (136–137) | 131.0 (130–132) | 0.1797 |
| Diastolic blood pressure (mmHg) | 85.5 (85–86) | 83.0 (76–90) | 0.6547 |
| Hypertensive patients with medication before tumor extirpation (n = 6) | |||
| Systolic blood pressure (mmHg) | 139.5 (124–142) | 135.0 (130–146) | 0.9165 |
| Diastolic blood pressure (mmHg) | 76.5 (66–82) | 75.0 (70–78) | 0.5971 |
Data were shown as the median (internal quartile range (IQR)). The significance of differences between mean values was estimated by the Wilcoxon signed rank test. P < 0.05 was considered significant. All hypertensive patients with medical treatment for hypertension were medicated with calcium blocker and stopped on the day of surgery.
Spearman’s signed rank correlation test revealed the significant or permissible correlation to investigate associations between the improvement in HOMA-β/IR and the improvement in urinary metanephrine/normetanephrine (the improvement in HOMA-β and the improvement in urinary metanephrine: r = 0.8386, p = 0.0007, the improvement in HOMA-β and the improvement in urinary normetanephrine: r = 0.2958, p = 0.3506, the improvement in HOMA-IR and the improvement in urinary metanephrine: r = –0.4526, p = 0.1395, the improvement in HOMA-IR and the improvement in urinary normetanephrine: r = 0.5071, p = 0.0925, respectively). Regression analysis revealed that the improvement in HOMA-β from before to after surgery had a significant positive association with the improvement in urinary levels of metanephrine after surgery (p = 0.0286), and a significant negative association with the improvement in urinary levels of normetanephrine after surgery (p = 0.0248). The improvement in HOMA-IR did not show a positive association with the improvement in urinary levels of metanephrine but showed a significant positive association with the improvement in urinary levels of normetanephrine (p = 0.0001) (Fig. 1 and Table 4).
A. Association between improvement in HOMA-β and improvement in urinary levels of metanephrine, B. Association between improvement in HOMA-β and improvement in urinary levels of normetanephrine, C. Association between improvement in HOMA-IR and improvement in urinary levels of metanephrine, D. Association between improvement in HOMA-IR and improvement in urinary levels of normetanephrine.
| Spearman’s rank correlation coefficient | ||
|---|---|---|
| r | p | |
| Improvement in HOMA-β and Improvement in urinary metanephrine | 0.8386 | 0.0007* |
| Improvement in HOMA-β and Improvement in urinary normetanephrine | –0.2958 | 0.3506 |
| Improvement in HOMA-IR and Improvement in urinary metanephrine | –0.4526 | 0.1395 |
| Improvement in HOMA-IR and Improvement in urinary normetanephrine | 0.5071 | 0.0925 |
| Regression analysis | |||
|---|---|---|---|
| Improvement in HOMA-β | |||
| β (SE) | 95% CI | p | |
| Improvement in urinary metanephrine | 46.69 (18.28) | 5.97–87.42 | 0.0286* |
| Improvement in urinary normetanephrine | –61.26 (23.21) | –112.97– –9.54 | 0.0248* |
| Improvement in HOMA-IR | |||
|---|---|---|---|
| β (SE) | 95% CI | p | |
| Improvement in urinary metanephrine | –0.45 (0.59) | –1.76–0.86 | 0.4625 |
| Improvement in urinary normetanephrine | 2.18 (0.36) | 1.37–2.99 | 0.0001* |
P < 0.05 (*) was considered significant. r, correlation coefficient; β, regression coefficient; SE, standard error; CI, confidence interval; HOMA-β, homeostasis model assessment of β-cell function; HOMA-IR, homeostasis model assessment of insulin resistance.
Glucose intolerance is observed in patients with pheochromocytoma. Several studies have focused on the effects of catecholamine excess on glucose intolerance, such as impaired secretion of insulin and/or increased insulin resistance. Two reports looked at the mechanism of glucose intolerance in patients with pheochromocytoma using clamp analyses. One study concluded that catecholamine excess induces glucose intolerance through increased insulin resistance [7] and the other study concluded that it was through impaired secretion of insulin [8]. Thus, the mechanism of glucose intolerance due to catecholamines has not been clarified.
We focused on the type of catecholamine in excess in patients with pheochromocytoma. It has been reported that impaired secretion of insulin by catecholamines is caused through the α2 receptors of β-cells in pancreatic islets [11, 13], and that increased insulin resistance is caused by excess of glucagon and free fatty acids through the α1 and β receptors of β-cells in pancreatic islets [7, 9, 14, 15]. Furthermore, adrenaline has been reported to have higher affinity to α2 receptors than noradrenaline [16, 17]. We hypothesized that excess amounts of different types of catecholamines induced different effects on glucose intolerance.
The metabolic product of adrenaline is metanephrine and that of noradrenaline is normetanephrine [1, 18]. Levels of metanephrine and normetanephrine are known to be stable [18]. Hence, we investigated the association of urinary levels of metanephrine/normetanephrine with HOMA-IR and HOMA-β, which are the parameters of insulin resistance and insulin secretion, respectively.
We found that extirpation of pheochromocytoma led to improvement of glycemic parameters but also elicited different effects between adrenaline and noradrenaline on glucose intolerance. Reduced urinary levels of metanephrine led to improved HOMA-β but had no effect on HOMA-IR. These findings suggested that adrenaline affects glucose intolerance mainly by impaired secretion of insulin, an observation that is in agreement with previous studies [11, 14]. Reduced levels of normetanephrine led to a decrease in HOMA-β and HOMA-IR. Values of HOMA-IR and HOMA-β are calculated using levels of insulin and glucose after an overnight fast, so these two values could show a positive association [19]. That is, reduced levels of insulin induced by an improvement in insulin resistance can cause a decrease in HOMA-β as judged from an eventual improvement of glycemic parameters. Hence, it is suggested that noradrenaline affects glucose intolerance mainly by increased insulin resistance. Furthermore, we investigated changes blood pressure before and after surgery because vascular defects caused insulin resistance [20, 21]. We found no significant differences between before and 1 week after surgery when we investigated HOMA-β/IR with respect to blood pressure. In addition, all patients with medical treatment for hypertension medicated by calcium blockers, not by angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, and diuretics which were reported to affect glucose metabolism. However, considering blood flow rarefaction in the skeletal muscle was reported to be one of the key in insulin resistance [21], reduced catecholamine by extirpation of pheochromocytoma could cause improvement in insulin resistance partly by increased blood flow in the skeletal muscle. We investigated HOMA-β/IR 1 day after hospital admission and 1 week after surgery, and surely HOMA-IR was improved overall as shown in Table 2. Nevertheless, to confirm the results of changes HOMA-β/IR, the investigation longer after surgery might be needed hereafter, considering above.
In summary, the present study provided insights into the mechanism of glucose intolerance in patients with pheochromocytoma. Our clinical study could aid development of a consensus on the effects of catecholamine excess on glucose intolerance in patients with pheochromocytoma. However, this study had two main limitations. First, our study cohort was small because pheochromocytoma is a rare tumor and pheochromocytoma is very heterogenous, which caused difficulty in our analysis. Second, we employed HOMA-IR and HOMA-β as surrogate markers of insulin resistance and insulin secretion as a substitute for an oral glucose tolerance test or hyperglycemic/hyperinsulinemic–euglycemic clamps. Thus, future studies are required to confirm our results.
We thank Mr. Hideaki Shimada and Ms. Yumi Iriguchi for assistance in undertaking our study.
This work was partly supported by grants from MSD and Mitsubishi-Tanabe. All authors declare that there is no conflict of interests regarding the publication of this paper. Toshihiko Yanase was supported financially in his research by MSD, Sanofi, Takeda, Daiichi Sankyo, Sumitomo Dainippon, Sanwa Chemistry, Eli Lilly Japan, Novo Nordisk, Novartis, Kowa, Boehringer Ingelheim, and Fujifilm. Kunihisa Kobayashi received honoraria from Mitsubishi-Tanabe, Ono, and MSD.
