Cushing's Disease in Older Patients

Presentation and Outcome

Nidan Qiao; Brooke Swearingen; Nicholas A. Tritos


Clin Endocrinol. 2018;89(4):444-453. 

In This Article


Between 2000 and 2017, 545 patients were admitted to Massachusetts General Hospital for surgery with a diagnosis of CD. Fifty-one patients were older than 60 years of age. Six patients were excluded: one patient had Nelson's disease and five patients had recurrent disease after an initial presentation at age less than 60 years old. Forty-five patients were therefore included in the older CD group. Ninety patients less than 60 years old with CD matched by year of surgery were included in control group I, and an additional 45 patients with NFPA over 60 years old matched by year of surgery were included in control group II.

Clinical Characteristics

Demographic, clinical and endocrine characteristics in the three groups of the study population are summarized in Table 1. The female to male ratio tended to be higher among younger than older patients with CD (P = 0.063), and was lower among patients with NFPA than both younger and older patients with CD. In patients with CD, weight gain (of any magnitude) was more common in younger patients (60.0% in older patients vs 72.2% in younger patients), while central obesity was more common in older patients (40.0% vs 28.9%,), but neither difference was statistically significant. The BMI was significantly lower in older patients (30.2 ± 8.1, P = 0.031), and was comparable with the BMI in older patients with NFPA (29.2 ± 5.2, P = NS). Muscle wasting was more common in older patients with CD (44.4% vs 20.0%, P = 0.006) and female hirsutism was more frequently observed in younger patients (24.2% vs 48.1%, P = 0.033). Other symptoms including skin abnormalities, alopecia, fatigue, multiple fractures, and hypokalemia showed no significant difference between older and younger patients with CD. Patients with NFPA had a greater prevalence of visual loss (P < 0.001).

Hypertension (80.0% vs 60.0%, P = 0.033), diabetes mellitus (44.4% vs 25.6%, P = 0.043), cardiovascular disease (20.0% vs 1.1%, P < 0.001) and venous thromboembolism (17.8% vs 1.1%, P < 0.001) were more prevalent in older patients compared to younger patients with CD. The American Society of Anesthesiologists (ASA) grade was also higher in older patients (48.9% in ASA3) compared to younger patients (15.6% in ASA3, P < 0.001). Compared to older patients with NFPA, older CD patients had a higher prevalence of hypertension (P = 0.004), diabetes mellitus (P = 0.001) and an ASA score of 3 (P < 0.001). Mood disorders were equally common in both groups of patients with CD (28.9%).

Imaging and Hormonal Assessments

Over 80% of both older and younger patients with CD had either microadenomas or negative MRIs (Table 2, P = NS). Macroadenomas were seen in 13.3% of older patients versus 16.7% of younger patients. After IPSS, CD patients had central:peripheral ratios predictive of a central source at baseline in 65.8% of older and 77.9% of younger patients, which increased after CRH stimulation (86.8% and 94.9%, respectively) in both groups. No differences in ACTH, UFC or LNSC levels were observed between the two (older versus younger) groups of patients with CD. There was no difference in free T4, TSH, IGF-1 and testosterone between two groups with CD. Patients with NFPA had an increased prevalence of hypopituitarism at presentation, including lower IGF-I level (0.65 ± 0.32 in older CD patients vs 0.31 ± 0.20 in older NFPA patients, P < 0.001).

We compared the magnitude of hypercortisolemia in patients with and without multiple fractures, hypertension, diabetes mellitus, cardiovascular disease and venous thromboembolism (Table S3, found in the Supporting Information). None of these comorbidities were statistically associated with the severity of hypercortisolism based on LNSC and UFC data. There was a trend towards higher LNSC values in patients with hypertension or diabetes mellitus.

Surgical Outcome

Remission was achieved in 34/45 (75.6%) of older patients and 73/90 (81.1%) of younger patients after the initial operation. An additional 4/5 patients achieved remission after early reoperation in the older group compared to 5/11 patients in the younger group. The overall remission rate was therefore 38/45 (84.4%) patients in the older group and 78/90 (86.7%) patients in the younger group, including both macro and microadenomas, and those patients (8) explored despite a noncentralizing IPSS (Table S2, found in the Supporting Information). A higher remission rate was observed in younger CD patients with visualized microadenomas (100% younger as opposed to 78.9% older; P = 0.011) but not in those with macroadenomas (80.0% younger as opposed to 83.3% older; P = NS) nor with MRI-negative tumours (P = NS) compared to older patients. If we exclude those (8) patients with noncentralizing IPSS, the remission rate was 34/40(85.0%) in older patients and 75/87(86.2%) in younger patients.

An ACTH-staining tumour was found on immunohistochemistry in comparable proportions of both patient groups (80.0% vs 75.6%, P = NS, Table 3). The Ki-67 index tended to be higher in younger patients (2.8 ± 3.5) compared to older patients (1.6 ± 1.6), but the difference did not reach statistical significance (P = 0.082).

No difference in postoperative adverse events was observed between the two groups. Four patients in the older group (8.9%) and three patients in the younger group (3.3%) developed transient DI after surgery; none of them developed permanent DI. Nine patients had SIADH with 4 in the older group (8.9%) and 5 in the younger group (5.6%). One patient in the younger group developed haemorrhage that requires re-exploration. CSF leakage developed in one patient in the younger group. One patient in the older group with CD died possible of pulmonary embolism after surgery.

The mean follow-up was 38.5 months in older patients and 46.9 months in younger patients (P = NS). Of those patients in initial remission, one patient (2.6%) in the older group and fifteen patients (19.2%) in the younger group recurred during follow-up. Survival analysis showed that the recurrence risk was significantly different between the two groups (P = 0.030, Figure 1). In subgroup analyses, no difference in recurrence rates was observed between younger versus older patients with CD who had MRI-negative tumours, microadenomas or macroadenomas (Figure 2).

Figure 1.

Kaplan-Meier Recurrence-free Survival Plot in Patients with CD

Figure 2.

Kaplan-Meier Recurrence-free Survival Plot in MRI Subgroups

Seven older and 12 younger patients had persistent disease despite surgical treatment. Figure 3 shows adjunctive treatments, which included radiation therapy, medical management or bilateral adrenalectomy. Younger patients were more likely to receive radiotherapy with adjuvant medical treatment (10/12 surgical failures with adequate follow-up) than older patients (1/5 surgical failures with adequate follow-up, P = 0.03). Medical treatment alone was used only in some of the older patients.

Figure 3.

Flowchart of Patients with Persistent Disease