Pituitary Evaluation in Patients With Low Prostate-Specific Antigen

Andjela Drincic, MD; Anh-Thu Nguyen, MD; Shilpi Singh, MBBS; Mohsen Zena, MD; Ryan Walters, PhD; Kathryn Friedman, RN; Robert J. Anderson, MD

Disclosures

Endocr Pract. 2018;24(12):1030-1037. 

In This Article

Results

Subjects

Low-PSA Group. Subjects were screened for the low-PSA group during in the period November 28, 2007, to March 9, 2011. The total number of screening PSA values obtained was 48,127. Upon elimination of all the patients with PSA >0.1 ng/mL and those with diagnosis of prostate cancer or history of prostate surgery, 329 subject files were prescreened through in-depth chart review, with further exclusion of subjects with prostate disease. Out of 95 subjects we contacted, 32 met exclusion criteria, 19 declined or could not participate in the evaluation, and 44 subjects who met inclusion and exclusion criteria were recruited and enrolled between December 1, 2007, and March 31, 2011. Listed in Table 1 are the baseline demographic characteristics. Subjects in the low-PSA group had a mean age (±SD) of 62.8 ± 6.9 years and were obese, with a mean body mass index (BMI) (±SD) of 33.9 ± 8.1 kg/m2. One subject was underweight, with BMI <18.5 kg/m2; 3 subjects (10%) had a healthy weight, with BMI 18.5 to 24.9 kg/m2; 12 subjects (27%) were overweight, with a BMI 25 to 29.9 kg/m2, and 28 (64%) subjects were obese, with BMI ≥30 kg/m2. Out of those, 14 had a BMI of ≥35 kg/m2. Nineteen (43.2%) subjects had type 2 diabetes mellitus (T2DM), and 11 (25%) were on various opiates for acute and chronic pain control.

Normal PSA Group. Screening and enrollment of the control group subjects occurred during the period of January 11, 2016, to August 11, 2016. Our enrollment goal was 20 subjects; thus, the enrollment period reflects the time needed to achieve that goal. Out of 74 subjects contacted, 26 agreed to a screening visit. Six subjects failed the screening visit due to meeting exclusion criteria (4 subjects had hypogonadism, 1 subject was on steroid therapy, and 1 subject had cancer). Twenty patients in this group had a mean age of 64.4 ± 5.8 years and a BMI of 31.8 ± 5.2 kg/m2. One subject had a healthy weight, with BMI 18.5 to 24.9 kg/m2; 8 subjects (40%) were overweight, with a BMI 25 to 29.9 kg/m2, and 11 (55%) subjects were obese, with BMI ≥30 kg/m2. Out of those, 4 subjects had a BMI ≥35 kg/m2. Six (30%) had T2DM, and 4 subjects (20%) were on narcotics for pain control. Table 1 outlines a summary of baseline demographic data for the normal-PSA group. There were no statistically significant differences between the two groups when compared for age, BMI, presence of diabetes, obstructive sleep apnea (OSA), or narcotic use.

Gonadal Status

At the VA-NWIHCS, hypogonadism is defined as a total testosterone value below the reference range for the assay (normal, 170 to 780 ng/dL). The low-PSA group had a mean total testosterone level that was significantly lower than the normal-PSA group (181.7 ng/dL vs. 263.7 ng/dL; P = .008).

Twenty subjects (45.5%) in the low-PSA group had a low total testosterone (Figure 1 and Figure 2). This group comprised of 4 subjects who had hypergonadotropic hypogonadism and 16 who had hypogonadotropic hypogonadism. In the latter group, we identified 13 out of 16 patients as having isolated hypogonadotropic hypogonadism, and 3 had associated hypopituitarism that was further evaluated with pituitary MRI (Figure 1). In contrast, only 3 out of 20 men in the control group had a low testosterone. All hypogonadal men in control group had isolated hypogonadotropic hypogonadism due to comorbid conditions. Overall, the rate of hypogonadism was significantly higher in the low-PSA group when compared with the control group (45.5% vs. 15.0%; OR, 4.7; 95% CI, 1.2 to 18.4; P = .027) (Figure 2). From a clinical standpoint, upon further questioning, 41 of 44 subjects (93.2%) in the low-PSA group had symptoms associated with hypogonadism, such as erectile dysfunction and/or low libido.

Figure 1.

Gonadal and growth hormone evaluation in the low–prostate-specific antigen (PSA) group. IGF-1 = insulin-like growth factor 1.

Figure 2.

Rate of hypogonadism for the normal- and low–prostate-specific antigen (PSA) groups. The rate was significantly higher in the low-PSA group compared with the normal-PSA control group (45.5% vs. 15.0%; odds ratio, 4.7; 95% confidence interval, 1.2 to 18.4; P = .027). Error bars represent 95% confidence intervals.

Growth Hormone Axis

In the initial data analysis of the low-PSA group, 9 patients had IGF-1 values below the age-dependent reference range (20%), while all subjects in the control group had normal IGF-1 for their age. Mean IGF-1 for the 44 men in the low-PSA group was 116.5 ± 39.1 ng/mL, as presented in Table 2. Upon thorough chart review, 1 patient from the low-PSA group was excluded from the analysis due to active and severe liver disease likely affecting IGF-1 synthesis. His testosterone was actually normal. Therefore, as presented in Figure 1, in the low-PSA group, 8 patients (18.6%) had IGF-1 values below the age-dependent reference range, whereas all subjects in the control group had normal IGF-1 for their age. Mean IGF-1 level in the low-PSA group (n = 43 after excluding the subject with liver failure) was statistically comparable to the control group (114.6 ± 38.7 ng/mL vs. 132.5 ± 43.1 ng/mL, respectively; P = .148). However, since IGF-1 levels are age dependent, direct comparison of mean IGF-1 values for subjects of different age is less meaningful. When comparing the deviation of the observed IGF-1 values from their age-specific lower bound results indicated that patients with low PSA were, on average, significantly closer to their age-specific lower bound of IGF-1 compared with patients with normal PSA (33.5 ± 40.3 ng/mL vs. 90.8 ± 46.1 ng/mL, respectively; P<.001; Figure 3).

Figure 3.

Average difference from insulin-like growth factor 1 (IGF-1) age-specific lower bound for the normal- and low–prostate-specific antigen (PSA) groups. IGF-1 was overly low in 18.6% of men with a low PSA (n = 8) compared with 0% in the normal-PSA group. The low-PSA patients were, on average, significantly closer to their age-specific lower limit of IGF-1 compared with patients with a normal PSA (33.5 ng/mL vs. 90.8 ng/mL, respectively; P<.001). Error bars represent 95% confidence intervals.

Among 8 patients with low IGF-1, 2 had panhypopituitarism associated with pituitary macroadenomas, 1 had hyperprolactinemia (MRI was normal), 1 had vague history of liver disease that was not affecting his liver enzymes or synthetic function (thus remained in the analysis pool), and the remaining 4 patients with an isolated low IGF-1 level did not have an MRI, since there were no other hormonal abnormalities.

As a whole, the low-PSA group had 5 (11.3%) of the 44 subjects who were diagnosed with partial or complete hypopituitarism. Table 2 summarizes a breakdown of the full pituitary panel in both groups.

Pituitary Imaging

Pituitary imaging was obtained in subjects with more than one hormonal abnormality (Table 3). Thus, it was not deemed necessary in the 13 out of 16 subjects with isolated hypogonadotropic hypogonadism in the low-PSA group. They all had well-recognized conditions explaining their low testosterone, such as diabetes, obesity, sleep apnea, and use of narcotics. Pituitary MRI was done in the remaining 3 subjects with hypogonadotropic hypogonadism and associated other hormonal abnormalities (Table 3). In subject 1 (with a history of macroadenoma), MRI showed a 3.4-cm tumor, and panhypopituitarism with profound adrenal insufficiency was diagnosed and treated because of study evaluation. Subject 2 had hypogonadism, hyperprolactinemia, and low IGF-1, and MRI results identified a pituitary macroadenoma (1.3 cm). Subject 3 had low FT4 in addition to profound hypogonadism, and the MRI identified a 6-mm microadenoma. Pituitary MRI was obtained on subject 4 with low IGF-1 and hyperprolactinemia and was unremarkable.

In the control group, all the patients diagnosed with hypogonadism had isolated hypogonadotropic hypogonadism and did not require MRI per clinical criteria.[14]

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....