Efficacy and safety of once-monthly pasireotide in Cushing's disease


Between Dec 28, 2011, and Dec 9, 2014, we randomly allocated 150 patients to pasireotide 10 mg (74 [49%] patients) or 30 mg (76 [51%] patients; figure 1). At baseline, patient demographics and disease history were balanced between dose groups (table 1). The primary efficacy endpoint (mUFC concentration of ≤ULN at month 7) was met by 31 (41·9% [95% CI 30·5–53·9]) of 74 patients in the 10 mg group and 31 (40·8% [29·7–52·7]) of 76 in the 30 mg group. Most patients in each group had a decrease in mUFC concentration from baseline to month 7 (figure 2). Of the patients who met the primary efficacy endpoint, two (6%) of 31 in the 10 mg group and none in the 30 mg were classified as responders on the basis of imputation of their last available measurement between month 4 and month 7. We observed higher response rates in patients with lower screening mUFC concentrations (figure 3). Partial control was achieved in four (5%) of 74 patients in the 10 mg group at month 7 and ten (13%) of 76 in the 30 mg group. Overall, we observed similar response rates between men and women and patients with and without previous surgery; response rates were higher in patients with a macroadenoma at baseline (24 [49%] of 49) than with a microadenoma (24 [35%] of 68), although this finding was seen only in the 10 mg group (nine [26%] of 34 and 12 [60%] of 20) and not in the 30 mg group (15 [44%] of 34 and 12 [41%] of 29; appendix).

 Opens large image

Figure 1

Trial profile

*Patients could have several reasons for exclusion, so the sum of the reasons for exclusion is greater than the total number of exclusions.

Table 1 

Baseline characteristics

Table Thumbnail. Opens Table in new tab.

 Opens large image

Figure 2

Absolute change in individual mUFC concentrations from baseline to month 7

For patients without an mUFC measurement at month 7, only baseline values are shown. The dashed line represents the upper limit of normal (ULN) for urinary free cortisol. Five (7%) patients in the 10 mg group and six (8%) in the 30 mg group had an mUFC concentration of less than or equal to the ULN at baseline (each patient had an mUFC concentration of 1·5–5·0 × ULN [range 1·6–4·5 × ULN] at screening, thus meeting the eligibility criterion); of these patients, four (80%) met the primary efficacy endpoint in the 10 mg group compared with three (50%) in the 30 mg group. mUFC=mean urinary free cortisol.

 Opens large image

Figure 3

Proportion of patients with an mUFC concentration of less than or equal to the ULN at month 7, according to mUFC stratum

mUFC=mean urinary free cortisol. ULN=upper limit of normal.

The number of patients who achieved an mUFC concentration of less than or equal to the ULN at month 7 with no previous dose uptitration was 21 (28·4% [95% CI 18·5–40·1]) of 74 in the 10 mg group and 24 (31·6% [21·4–43·3]) of 76 in the 30 mg group. 31 (42%) patients received dose uptitration at month 4 in the 10 mg group compared with 28 (37%) in the 30 mg group; of these patients, ten (32%) of 31 were responders at month 7 compared with seven (25%) of 28. At month 12, 26 (35·1% [24·4–47·1]) of 74 patients had an mUFC concentration less than or equal to the ULN compared with 19 (25·0% [15·8–36·3]) of 76; partial control was achieved in eight (11%) of 74 compared with 13 (17%) of 76. Of patients who were classified as responders at month 7, 20 (65%) of 31 patients had an mUFC concentration less than or equal to the ULN at month 12 compared with 15 (48%) of 31. Six (19%) of 31 responders at month 7 had an mUFC concentration of more than the ULN at month 12 compared with 13 (42%) of 31 (appendix); two (6%) of these patients (both in the 30 mg group) received a lower dose of pasireotide at month 11 than at month 6. Median mUFC concentration decreased within 1 month of treatment in both groups and remained lower than baseline until month 12 for patients who continued to receive pasireotide (figure 4). The median percentage change from baseline in mUFC concentration was −47·9% (−198·1 nmol per 24 h) in the 10 mg group and −48·5% (−178·8 nmol per 24 h) in the 30 mg group at month 7 compared with −52·5% (−164·7 nmol per 24 h) and −51·9% (−182·2 nmol per 24 h) at month 12.

 Opens large image

Figure 4

Median mUFC concentration from baseline until month 12

Error bars represent IQRs. The dashed line represents the ULN for urinary free cortisol. mUFC=mean urinary free cortisol. ULN=upper limit of normal.

We observed sustained reductions in both groups for median morning plasma ACTH and late-night salivary cortisol concentration (appendix). The median percentage change from baseline to month 12 in plasma ACTH concentration was −22·5% (−2·0 pmol/L) in the 10 mg group and −17·4% (−2·0 pmol/L) in the 30 mg group and in late-night salivary cortisol concentration was −30·7% (−2·4 nmol/L) compared with −23·7% (−1·4 nmol/L). At baseline, nine (13%) of 68 patients in the 10 mg group and three (4%) of 69 in the 30 mg group had normal late-night salivary cortisol concentrations; at month 7, 16 (30%) of 53 did compared with nine (15%) of 60; at month 12, 11 (25%) of 44 did compared with ten (21%) of 48 at month 12. Changes in serum cortisol concentrations were generally consistent with those seen for morning plasma ACTH and late-night salivary cortisol concentrations (appendix). Median percentage changes from baseline to month 12 for serum cortisol concentration were −9·2% (−49·1 nmol/L) in the 10 mg group and 0·1% (0·3 nmol/L) in the 30 mg group; median percentage changes ranged from 1·7% (9·7 nmol/L) to −14·5% (−78·2 nmol/L) at all other timepoints between month 7 and month 11 in the 10 mg group and from −8·0% (−37·8 nmol/L) to −16·1% (−86·1 nmol/L) in the 30 mg group.

Clinical improvements were observed at month 7 and most improved further to month 12 (appendix). Mean change from baseline to month 12 in systolic BP was −4·6 mm Hg (95% CI −9·9 to 0·7) in the 10 mg group and −5·0 mm Hg (−8·8 to −1·3) in the 30 mg group, in diastolic BP was −3·4 mm Hg (−7·3 to 0·4) compared with −3·1 mm Hg (−5·7 to −0·5), in waist circumference was −4·5 cm (−7·2 to −1·8) compared with −6·2 cm (−8·7 to −3·6), in BMI was −1·3 kg/m2 (−1·8 to −0·8) compared with −2·6 kg/m2 (−3·3 to −1·9), in weight was −3·4 kg (−4·8 to −2·0; 4·6% decrease) compared with −6·5 kg (−8·3 to −4·7; 8·6% decrease); and in the Cushing’s Quality of Life questionnaire was 6·4 points (1·3–11·6) compared with 7·0 points (3·0–10·9). We also observed improvements in facial rubor, fat pads, and other clinical signs of Cushing’s disease from baseline to month 7 and month 12 (appendix).

For patients with evaluable measurements at both baseline and month 12, median tumour volume decreased by 17·8% (31·0 mm3) in the 10 mg group (n=35) and 16·3% (33·5 mm3) in the 30 mg group (n=38; appendix). For patients with a macroadenoma at baseline, median tumour volume decreased by 14·6% (48·0 mm3; n=15) at month 12 compared with 11·6% (86·0 mm3; n=13). The proportion of patients with a 20% or greater reduction in tumour volume, 20% or greater increase, and less than 20% change at month 12 is shown in table 2. Median mUFC concentrations decreased from baseline to month 12 regardless of tumour volume change (appendix).

Table 2 

Proportion of patients with a 20% or greater reduction, 20% or greater increase, and less than 20% change (stable) in tumour volume from baseline to month 12, by maximum tumour diameter at baseline

Table Thumbnail. Opens Table in new tab.

By data cutoff, the median treatment duration was 449 days in the 10 mg group (IQR 197–756) and 381 days in the 30 mg group (IQR 280–532). Overall, 73 (99%) of 74 patients in the 10 mg group and all 76 patients in the 30 mg group had an AE. The most commonly reported AEs were hyperglycaemia, diarrhoea, cholelithiasis, diabetes mellitus, and nausea (table 3); most were of mild-to-moderate severity (grade 1–2). The most frequent grade 3–4 AE was diabetes mellitus. Nine (12%) of 74 patients in the 10 mg group and ten (13%) of 76 in the 30 mg group discontinued treatment because of an AE (appendix). Two patients in the 30 mg group died during the study: one (1%) from pulmonary artery thrombosis (30 days after first injection) and one (1%) from cardiorespiratory failure (16 days after the 16th injection [during the extension phase]); we did not suspect that either death was related to the study drug. SAEs were reported in 21 (28%) of 74 patients compared with 17 (22%) of 76; the most commonly reported SAE was cholelithiasis (two [3%] of 74 and two [3%] of 76). Eight (11%) of 74 patients compared with four (5%) of 76 had an SAE that was suspected to be related to study drug. Hyperglycaemia-related AEs occurred in 53 (72%) of 74 patients compared with 62 (82%) of 76; 17 (23%) of 74 compared with 18 (24%) of 76 had grade 3–4 events. Four (5%) patients per group discontinued treatment because of hyperglycaemia-related AEs. No instances of diabetic ketoacidosis or hyperosmolar hyperglycaemic non-ketotic syndrome were reported.

Table 3 

Adverse events reported in at least 10% of patients overall, regardless of study drug relationship

Table Thumbnail. Opens Table in new tab.

Mean fasting plasma glucose concentration increased within 1 month of treatment: the concentration was 93·9 mg/dL (SD 16·7) in the 10 mg group and 95·8 mg/dL (20·4) in the 30 mg group at baseline compared with 113·5 mg/dL (33·9) and 138·5 mg/dL (50·0) at month 1. Mean HbA1c concentration increased from 5·7% (SD 0·6) at baseline in the 10 mg group and 5·7% (0·7) in the 30 mg group to 6·4% (1·1) and 6·8% (1·3) at month 2, and to 6·9% (1·4) and 7·0% (1·4) at month 12. At month 12, 28 (58%) of 48 patients in the 10 mg group and 32 (60%) of 53 in the 30 mg group had an HbA1c concentration of less than 7·0%. At baseline, 27 (36%) of 74 patients compared with 33 (43%) of 76 had diabetes mellitus, 12 (16%) of 74 compared with 12 (16%) of 76 had prediabetes, and 35 (47%) of 74 compared with 31 (41%) of 76 had normal glucose tolerance (the appendix contains definitions of diabetes categories). Of patients with normal glucose tolerance at baseline, 13 (37%) of 35 in the 10 mg group and seven (23%) of 31 in the 30 mg group had a shift into the prediabetes category and 19 (54%) of 35 compared with 21 (68%) of 31 had a shift into the diabetes category at least once during the study. Seven (58%) of 12 patients compared with 11 (92%) of 12 who had prediabetes at baseline had a worst-reported shift into the diabetes category. Of patients with normal glucose tolerance at baseline, eight (23%) of 35 compared with 13 (42%) of 31 had a last available HbA1c concentration of 6·5% or higher; nine (75%) of 12 compared with seven (58%) of 12 patients who had prediabetes at baseline had a last available HbA1c concentration of 6·5% or higher (the appendix shows the shift in HbA1c concentration from baseline to the highest reported value). Of patients with normal glucose tolerance or prediabetes at baseline, 18 (38%) of 47 compared with 22 (51%) of 43 were receiving antidiabetic medication at last assessment; eight (73%) of 11 compared with all 14 patients with diabetes who were not receiving antidiabetic medication at baseline were receiving it at last assessment. Nine (56%) of 16 patients compared with ten (53%) of 19 on antidiabetic medication at baseline were receiving at least one additional agent at last assessment (the appendix contains HbA1c concentration changes by antidiabetic medication status).

Gallbladder or biliary-related AEs occurred in 18 (24%) of 74 patients in the 10 mg group and 34 (45%) of 76 in the 30 mg group; two (3%) patients per group underwent cholecystectomy. Of patients with normal baseline ultrasound results, eight (13%) of 64 compared with 24 (36%) of 67 had detectable sludge or gallstones at last assessment. 15 (20%) patients per group had liver safety-related AEs; six (8%) of 74 compared with five (7%) of 76 had grade 3–4 AEs, most commonly increased γ-glutamyltransferase concentration, which was reported in four (5%) compared with three (4%). Two (3%) patients per group discontinued treatment because of liver safety-related AEs. At baseline, one of the patients in the 10 mg group had steatosis and elevated alanine aminotransferase concentration (53 units per L) and another in the 30 mg group had steatosis. Alanine aminotransferase concentrations remained elevated at last observation (42 days [82 units per L] and 54 days [166 units per L]) after last pasireotide dose in two (50%) of the four patients. No instances of jaundice or other symptoms of hepatic dysfunction were reported.

Two (3%) patients in the 10 mg group had a grade 1 AE of decreased insulin-like growth factor 1 (IGF1) concentration without any associated symptoms; one (1%) of these patients initiated growth hormone (GH) replacement 2·5 years after the onset of this AE. Another patient in the 30 mg group had an AE of GH deficiency (IGF1 and GH concentration less than the normal range); no associated symptoms were reported and treatment was not adjusted. Mean IGF1 concentration decreased during pasireotide treatment (appendix). The initial decrease, seen at month 1 and month 3 (before any dose uptitration), was greater in the 30 mg group than in the 10 mg group. By month 7 (3 months after the first permitted dose uptitration), mean IGF1 SD score was similar in both groups and remained stable to month 12. Mean IGF1 SD score was 0·3 (SD 1·4) in the 10 mg group at baseline and 0·3 (1·2) in the 30 mg group, was −0·6 (1·5) compared with −1·5 (1·4) at month 3, was −1·2 (1·6) compared with −1·6 (1·4) at month 7, and was −1·0 (1·5) compared with −1·4 (1·6) at month 12. Most (>80%) patients continued to have IGF1 concentrations within the normal range at month 12 (appendix) and no clinical symptoms were reported. Newly occurring QTcF extension of greater than 480 ms occurred in two (3%) patients in the 30 mg group; events were sporadic and resolved without treatment interruption. No patients with QTcF extension of greater than 480 ms had arrhythmias or clinical symptoms associated with episodes of QTcF extension. Two (3%) patients per group had injection site-related AEs; one event in the 10 mg group required temporary treatment interruption and medication for pain.



Read more

Leave A Comment

Your email address will not be published. Required fields are marked *