European Journal of Endocrinology (2011) 164 325–333 ISSN 0804-4643

CLINICAL STUDY Coadministration of Autogel and pegvisomant normalizes IGF1 levels and is well tolerated in patients with partially controlled by analogs alone Aart-Jan van der Lely, Ignacio Bernabeu1, Jan Cap2, Philippe Caron3, Annamaria Colao4, Josef Marek5, Sebastian Neggers and Pascal Birman6 Department of Internal Medicine, Erasmus University MC, 40 Dr Molewaterplein, 3015 GD Rotterdam, The Netherlands, 1Santiago de Compostela University, Complejo Hospitalario Universitario de Santiago (SERGAS), Santiago de Compostela 15706, Spain, 2Second Department of Internal Medicine, Faculty of Medicine and Teaching Hospital, Charles University, Hradec Kralove 50001, Czech Republic, 3Department of Endocrinology and Metabolic Diseases, CHU Larrey, Toulouse 31059, France, 4Federico II, University of Naples, Naples 80138, Italy, 5First School of Medicine, Charles University, Prague 12108, Czech Republic and 6Ipsen Innovation, Les Ulis 91966, France (Correspondence should be addressed to A-J van der Lely; Email: [email protected])

Abstract Objective: To evaluate the efficacy and safety of coadministered lanreotide Autogel (LA; 120 mg/month) and pegvisomant (40–120 mg/week) in acromegaly. Design: This is a 28-week, multicenter, open-label, single-arm sequential study. Methods: Patients (nZ92) biochemically uncontrolled, on somatostatin analogs (SSAs) or using pegvisomant monotherapy entered a 4-month run-in taking LA (120 mg/month). Patients uncontrolled after the run-in period (nZ57) entered a 28-week coadministration period, receiving LA 120 mg/month plus pegvisomant (60 mg once weekly, adapted every 8 weeks based on IGF1 levels to 40–80 mg once weekly or 40 or 60 mg twice weekly). Results: In total, 33 (57.9%) patients had normalized IGF1 following coadministration (P!0.0001 versus 30% minimum clinically relevant); median pegvisomant dose in normalized patients was 60 mg/week. IGF1 normalized at any time during coadministration in 45 (78.9%) patients (P!0.0001) with median pegvisomant dose at 60 mg/week. Being nondiabetic (odds ratio (OR): 4.65) and older (OR, upper versus lower quartile: 3.40) showed increased likelihood of normalization. Symptom reduction was greatest for arthralgia (K0.6G1.6) and soft tissue swelling (K0.6G1.8). Five patients reported treatment-emergent adverse events causing treatment withdrawal: three serious (treatment related – thrombocytopenia, urticaria; not treatment related – abdominal pain/vomiting) and two nonserious (hepatotoxicity and cytolytic , both elevating alanine aminotransferase to O5!upper limit of normal with normalization after withdrawal). Conclusions: In patients partially controlled by SSAs, LA (120 mg/month) plus pegvisomant normalized IGF1 in 57.9% of patients after 7 months, at a median effective pegvisomant dose of 60 mg/week, and 78.9% at any time. In these patients, results suggest a pegvisomant-sparing effect versus daily pegvisomant monotherapy.

European Journal of Endocrinology 164 325–333

Introduction symptom burden (1, 2). Long-acting SSAs, adminis- tered by monthly injection, are generally well tolerated; Acromegaly is a chronic disorder caused by excess the most notable adverse effect, increased incidence of GH secretion, usually due to a benign GH-secreting small gallstones and gallbladder sludge, rarely causes pituitary somatotroph adenoma. Pharmaceutical clinically significant symptoms (1, 3). management of acromegaly includes somatostatin Pegvisomant, a recombinant human GH analog and analogs (SSAs), the GH (GHRA) highly selective GHRA (4, 5), is the only approved agent pegvisomant, or, less commonly, the dopamine agonist in its class, and it is indicated in Europe in patients cabergoline. There are currently two commercially who have responded incompletely, or are intolerant, to available long-acting SSAs, LAR (Novartis maximal dose SSA therapy. Although typically injected International AG) and lanreotide Autogel (LA, daily, emerging data now support once-weekly dosing of Beaufour-Ipsen Industrie, Dreux, France), both of pegvisomant (6). Pegvisomant monotherapy has been which normalize insulin-like growth factor 1 (IGF1) shown to normalize IGF1 levels in w90–100% of levels in 30–60% of unselected patients and reduce patients at doses of 20–40 mg/day (7, 8), but long-term

q 2011 European Society of Endocrinology DOI: 10.1530/EJE-10-0867 Online version via www.eje-online.org

Downloaded from Bioscientifica.com at 09/23/2021 12:42:47PM via free access 326 A-J van der Lely and others EUROPEAN JOURNAL OF ENDOCRINOLOGY (2011) 164 data from the observational, registry-based ACROS- As such, they had either IGF1 levels that were greater TUDY, representing 2625 patient-years of treatment, than the upper limit of normal (ULN) following indicated IGF1 normalization in only w60% of patients pegvisomant 30 mg/day for at least 3 months or IGF1 at mean weekly doses exceeding 120 mg (9). The levels OULN 28 days after the last SSA injection, relatively high cost and traditional daily injection following at least 6 months of treatment with LA or frequency of pegvisomant remain key limitations to its octreotide LAR, including 3 months at the highest wider use, as does concern that pegvisomant may not marketed dose. Unlike similar studies conducted inhibit tumor growth. The potential for hepatitis is also previously (12, 18), this study employed a two-step acknowledged in the product’s ‘Summary of Product process to confirm that patients entering the coadmi- Characteristics’, along with a recommendation that nistration period were incomplete responders to full- hepatic transaminases be monitored at 4–6-week dose SSA therapy. Eligible subjects (nZ92) entered a intervals for the first 6 months of treatment, and in all 4-month run-in period, during which they received LA cases with symptoms consistent with hepatitis (10). at the highest available dose (120 mg/month). Patients There is therefore increasing interest in combination with serum IGF1O1.2! ULN (nZ54) or between ULN therapy as a means of sparing pegvisomant. and 1.2! ULN with serum GH nadir R1 m/l on oral Patients with only a partial response to full-dose SSA glucose tolerance testing (nZ3) 28 days after the third therapy can achieve further reductions in IGF1 levels by injection of LA were considered to have remained combining weekly pegvisomant with monthly SSA uncontrolled and entered the coadministration period injections (11–13). This strategy capitalizes on the (nZ57 total). Exclusion criteria were pituitary surgery complementary modes of action of the two drug classes or radiotherapy within the previous 6 months or and mitigates concerns about tumor growth by planned during the study; dopamine agonists within incorporating SSAs, agents that can induce pituitary 6 weeks of study entry; previous SSA/GHRA therapy; tumor shrinkage (14–16). Weekly, rather than daily, and aspartate aminotransferase (AST) or alanine pegvisomant dosing offers the advantage of reduced aminotransferase (ALT) O2! ULN. None of the injection frequency, which patients may prefer, even in patients in this study had participated in any previous the absence of further reduction in IGF1 levels (6, 17). coadministration studies. All patients provided written Based on the findings from two small (26–32 informed consent prior to entering the study. patients) published trials of combination SSA/pegviso- mant therapy (12, 18), this study is the first to examine LA in combination with pegvisomant in a rigorously Treatments selected larger population of patients partially con- LA (120 mg) was administered by a single s.c. injection trolled by SSAs. in the buttock at the same time each day, every 28 days. Pegvisomant was administered by a single s.c. injection once or twice weekly in the abdomen, buttock, thigh, or upper arm. The starting dose of pegvisomant in the Patients and methods coadministration period, 60 mg/week, was the median Study objective dose that normalized IGF1 levels in the Feenstra et al. (12) study. The pegvisomant dose was evaluated every This is a multicenter, open-label, single-arm, sequential 8 weeks based on IGF1 levels taken 4 weeks after the study in which patients served as their own controls; previous dose adaptation, and according to the rules study design is presented in Fig. 1a. The primary study presented in Fig. 1b. If dosing was twice weekly, the objective is to evaluate the efficacy of coadministered LA second injection was given 4 days after the first. (120 mg/month) and pegvisomant (40–120 mg/week), Pegvisomant doses exceeding 40 mg were delivered as as determined by the change in IGF1 levels over 28 two separate injections at the same time. weeks, in patients with acromegaly. Secondary objec- tives are to assess the effects of combined therapy on acromegaly symptoms and quality of life (QoL) and to Efficacy and safety analyses assess treatment safety. The primary efficacy end point was the number and percentage of patients with normalized age- and sex- Patient selection adjusted IGF1 levels at the end of coadministration; this was compared with the estimated minimum clinically Patients were eligible for inclusion if aged 18–75 years relevant percentage of 30%, using a one-sided exact test and with a documented diagnosis of acromegaly. They based on the binomial distribution. Sensitivity analyses were also either responders to daily pegvisomant were performed to investigate the potential impact on monotherapy (and thus given the approved indication the primary end point of patients achieving a for pegvisomant, previously uncontrolled on SSA normalized IGF1 level during the coadministration therapy) or partial responders to the highest marketed phase that was not maintained until study end because doses of either pegvisomant or a long-acting SSA. of a subsequent reduction in pegvisomant dose.

www.eje-online.org

Downloaded from Bioscientifica.com at 09/23/2021 12:42:47PM via free access EUROPEAN JOURNAL OF ENDOCRINOLOGY (2011) 164 Coadministration of LA and pegvisomant 327

(a) Run-in period Coadministration period Lanreotide Autogel 120 mg Lanreotide Autogel 120 mg monthly injection + GH monthly injection for 4 months antagonist injection once or twice a week for 7 months

Pegvisomant Pegvisomant Pegvisomant Pegvisomant dose 1 dose 2 dose 3 dose 4

2–4 4 4 4 4 4 4 4 4 4 4 4 weeks weeks weeks weeks weeks weeks weeks weeks weeks weeks weeks weeks

Visit 1 Visit 2 Visit 3 Visit 4 Visit 5 Visit 6 Visit 7 Visit 8 Visit 9 Visit 10 Visit 11

IGF1 IGF1 IGF1 IGF1 IGF1 IGF1 IGF1

Enrollment Confirmation of Pegvisomant Pegvisomant Pegvisomant eligibility based dose dose dose on visit 3 adaptation adaptation adaptation assessments

(b) Pegvisomant IGF1 ≤ ULN 40 mg/week Pegvisomant 40 mg/week ≤ Pegvisomant IGF1 ULN IGF1 > ULN Pegvisomant 60 mg/week 40 mg/week Pegvisomant IGF1 ≤ ULN IGF1 > ULN 60 mg/week IGF1 ≤ ULN Pegvisomant 60 mg/week Pegvisomant IGF1 > ULN Pegvisomant 80 mg/week 60 mg/week Pegvisomant IGF1 ≤ ULN 80 mg/week Pegvisomant IGF1 > ULN 80 mg/week ≤ IGF1 ULN IGF1 > ULN Pegvisomant Pegvisomant 40 mg × 2/week 80 mg/week IGF1 ≤ ULN Pegvisomant IGF1 > ULN 40 mg × 2/week Pegvisomant 40 mg × 2/week IGF1 > ULN Pegvisomant 60 mg × 2/week

Figure 1 (a) Study design and (b) pegvisomant dose adaptation schedule. If pegvisomant dosing was twice weekly, the second injection was given 4 days after the first. Pegvisomant doses exceeding 40 mg were delivered as two separate injections at the same time.

The analyses determined i) the number and percentage separate factors), and time since diagnosis. These of patients with a normal IGF1 level at any time during variables were tested using backward elimination the coadministration period while taking the dose used (with a significance level of PZ0.15 for keeping a prior to the last study assessment and ii) the number term in the final model). All efficacy analyses were and percentage of patients with a normal IGF1 level at performed on the intention-to-treat (ITT) population, any time during coadministration. IGF1 z-score was a defined as all coadministered subjects with R1 baseline secondary efficacy end point and was calculated using and R1 postbaseline IGF1 assessment where values the following formula: z-scoreZ(ln(IGF1 value)K were missing at the end of coadministration and last mean)/S.D., with mean and S.D. being age- and sex- observation carried forward was used. specific and were determined on the logarithmic scale. Safety end points were adverse events (AEs), defined Additional secondary end points included acromegaly as serious if they were life threatening or fatal symptoms, rated on a 0–8 scale of increasing severity, or resulted in significant disability. In this study, the and ACROQoL questionnaire scores. Only summary safety endpoints chosen were glucose tolerance (in statistics were prepared for secondary end points. To subjects without diabetes only), gallbladder ultrasound, investigate whether various predisposing factors were electrocardiograms (ECGs), hepatic function tests associated with the primary efficacy variable, logistic (including ALT and AST), and standard hematological regression methods were applied in an exploratory and biochemical laboratory tests. The duration of the analysis with normalization at the end of the coadmi- study was inadequate for the assessment of the full nistration period as the response variable. The predis- impact of treatment on pituitary size, and thus this posing factors included IGF1 z-score at the end of the parameter was included as a safety end point only. run-in period, previous treatment, diabetic status, age, The safety population comprised all patients who sex, pituitary tumor size, whether the patient had received at least one dose of both trial products during received previous radiotherapy or surgery (as two coadministration.

www.eje-online.org

Downloaded from Bioscientifica.com at 09/23/2021 12:42:47PM via free access 328 A-J van der Lely and others EUROPEAN JOURNAL OF ENDOCRINOLOGY (2011) 164

The planned sample size of 60 subjects was based on 7 z-score at baseline having at least 95% power to detect the difference 6 z-score at LVA between the null hypothesis of 30% (estimated as the 5 minimum clinically relevant percentage of responders) 4 and the expected response rate (80%) at the 5% 3 significance level. 2 -score for IGF1 z 1 0 40 mg 60 mg 40 mg/week 60 mg/week 80 mg/week Results –1 × 2/week × 2/week Final pegvisomant dose Patients and conduct of study Figure 2 Serum IGF1 concentration in 57 patients with acromegaly All 57 subjects received at least one dose of both before and after 28 weeks of combination therapy with lanreotide study medications and had at least one baseline and Autogel (120 mg/month) plus pegvisomant (40–120 mg/week). at least one postbaseline IGF1 assessment; the ITT and Data are z-scores for log-transformed IGF1 levels. A z-score C K safety populations were therefore identical. between 2 and 2 indicates a normal IGF1 concentration. LVA, last (post baseline) value available. Baseline demographics and clinical characteristics for the ITT population are presented in Table 1. The most frequent comorbid conditions in study patients were Z other endocrine disorders (n 36; 63%), previous inter- theoretical 30% estimated as minimum clinically ventional procedures (nZ33; 58%), and metabolic/ Z relevant; Fig. 2), with a median effective weekly nutritional disorders (n 31; 54%). Acromegaly treat- pegvisomant dose of 60 mg for all (normalized) subjects ments taken prior to run-in by patients who entered and for nondiabetics, and 80 mg for those with diabetes. into the coadministration period were lanreotide The proportion of subjects with normalized IGF1 at (nZ24; 42%), octreotide (nZ20; 35%), and pegviso- the end of coadministration (or for a small number of mant (nZ13; 23%). patients at early withdrawal) were similar across final pegvisomant dose groups of 40, 60, and 80 mg/week Efficacy (Fig. 3), and were statistically significant in nondiabetic (24/38; 63.2%, 90% CI: 48.5, 76.2; P!0.0001) but A total of 33 patients (57.9%; 90% confidence interval not in diabetic patients (9/19; 47.4%, 90% CI: 27.4, (CI): 46.1, 69.0) achieved normal serum IGF1 levels by 68.0; PZ0.008); this finding was consistent with ! the end of coadministration (P 0.0001 versus higher median reductions in z-scores in nondiabetics. Similarly, in an exploratory multivariate logistic Table 1 Baseline demographics and characteristics of the ITT regression analysis, being nondiabetic (OR: 4.65) and population. being older (OR for upper versus lower quartile: 3.40) were both associated with increased odds of normal- All patients Z (n 57) ization. None of the other potential predisposing factors Mean age (yearsGS.D.) 51.6G12.7 tested, including tumor size, baseline IGF1 SDS, and Sex (n,%) previous radiotherapy, were significant predictors of a Male 29 (50.9) Female 28 (49.1) response. The biochemical response rate at study end Diabetic status at entry (n,%) was 76.9% in patients taking a final pegvisomant dose Diabetic 15 (26.3) of 40 mg once weekly, 61.5% in those taking 60 mg Nondiabetic 42 (73.7) a once weekly, 75% with 80 mg once weekly, and 60% Baseline IGF1 (z-score) with 40 mg twice weekly (Fig. 3), leaving w30% of MeanGS.D. 3.36G0.85 Median (min, max) 3.13 (2.04, 6.03) patients across these groups suboptimally controlled on Years since diagnosis (meanGS.D.) 7.7G7.0 their final pegvisomant dose. A notable lack of efficacy Pituitary tumor size (n,%) was apparent for pegvisomant 60 mg twice weekly, % Microadenoma 10 mm 6 (10.5) suggesting that these subjects may need more pegviso- Macroadenoma O10 mm 49 (86.0) N/A 2 (3.5) mant and/or a different dose regimen to normalize Pretreatment (n,%) IGF1. Biochemical response was significant in patients Pituitary surgery 38 (66.7) previously treated with octreotide LAR (nZ14 (70.0%); Radiotherapy 18 (31.6) Z Z Z b P 0.0003) and LA (n 13 (54.2%); P 0.01), but not Prior acromegaly medications (n,%) Z Z All somatostatin analogs 44 (77.2) pegvisomant (n 6(46.2%);P 0.17). Of the 13 Lanreotide 24 (42.1) patients previously treated with pegvisomant, 10 Octreotide 20 (35.1) (76.9%) had normal IGF1 levels at baseline; none of GHRA (pegvisomant) 13 (22.8) these patients remained biochemically controlled at the aAt end of run-in period. end of run-in. However, only one patient failed to bIncludes medications for acromegaly administered until visit 2. normalize during coadministration. Of the three

www.eje-online.org

Downloaded from Bioscientifica.com at 09/23/2021 12:42:47PM via free access EUROPEAN JOURNAL OF ENDOCRINOLOGY (2011) 164 Coadministration of LA and pegvisomant 329

100 Safety 90 Mean study duration was 47.6 weeks; durations of 80 76.9 75.0 lanreotide and pegvisomant therapy during coadminis- 70 tration were 26.8 and 25.8 weeks respectively. Treat- 61.5 ment-emergent AEs (TEAEs) related to treatment 57.9 60.0 60 occurred in more subjects during coadministration 50 (nZ24; 42%), when predominant complaints were 40 gastrointestinal, hepatobiliary, and injection-site disorders (nZ6; 11% for all), than during run-in 30 Z

% IGF1 normalization (n 13; 23%), when the most frequent symptoms 20 were gastrointestinal and hepatobiliary (nZ5; 9% for 10 both) and metabolic/nutritional disorders (nZ4; 7%). 0 0 The most common gastrointestinal complaint during coadministration was diarrhea (nZ3; 5%), which

Overall = 5) occurred with similar frequency during run-in with = 57) = 13) = 13) = 16) (n = 10) (n (n (n (n (n lanreotide alone (nZ4; 7%). 80 mg/week 40 mg/week60 mg/week Of the serious AEs reported during coadministration 40 mg × 2/week60 mg × 2/week Z Normalization at study end by final dose (n 8 patients; 14%), two (thrombocytopenia and urticaria) were considered treatment-related and led Figure 3 Serum IGF1 normalization by final pegvisomant dose. to withdrawal from the study, according to protocol; a ‘Final dose’ was the dose taken by patients at the end of the study or third, severe abdominal pain and vomiting in one at the time of withdrawal. patient, was not considered treatment-related, but led to treatment discontinuation nonetheless. The patient with thrombocytopenia had a known history of patients previously treated with pegvisomant, who idiopathic thrombocytopenia purpura; her lowest had abnormal IGF1 levels at baseline, none of whom recorded platelet count, 87 G/l, occurred 6 weeks were biochemically controlled at the end of the run-in after the final pegvisomant dose and the thrombocy- period, two (66.7%) remained uncontrolled at the end topenia resolved 6 weeks later. A further two patients of coadministration. A post hoc analysis involving eight reported nonserious treatment-related AEs that led to patients whose mean IGF1 levels were similar while on treatment withdrawal: transient hepatotoxicity and pegvisomant monotherapy and on at least one occasion cytolytic hepatitis, both leading to transient elevations during the coadministration period (202.4G51.6 and in transaminases (ALT O5!ULN), which returned 193.3G53.0 ng/ml respectively) showed that these patients were able to reduce their weekly pegvisomant dose by an average of 51% (PZ0.008) from 131.3 100

G36.2 to 62.5G16.7 mg. 90 86.8

Serum IGF1 normalized at least once in 38 (66.7%; 78.9

80 73.7 90% CI: 55.0, 77.0) patients at their final doses (median 70 66.7 63.2 final pegvisomant dose of 60 mg/week; Fig. 4). At any 63.2 time during coadministration, serum IGF1 normalized 60 57.9 52.6 at least once in 45 patients (78.9%; 90% CI: 68.1, 87.4) 50 47.4 (both P!0.0001; Fig. 4). There was a positive correlation between IGF1 z-score 40 at the end of the run-in period and final weekly 30 % IGF1 normalization pegvisomant dose at normalization (Fig. 5): patients 20 with lower IGF1 z-scores on entering the coadministra- 10 tion period required less pegvisomant to achieve normal IGF1 levels than those with higher entry scores. 0 Overall Diabetic Nondiabetic Coadministration of LA and pegvisomant decreased (n = 57) subjects subjects mean acromegaly symptom scores (GS.D.) with greatest (n = 19) (n = 38) reductions for arthralgia (K0.6G1.6) and soft tissue swelling (K0.6G1.8); small mean improvements in Normalization at study end global QoL score (mean increase Z2.2G8.8) and Normalization at least once while taking final dose multiple subscores were also apparent, although there Normalization at any time was considerable data variability. There was no correlation between change in IGF1 z-score and change Figure 4 Serum IGF1 normalization occurring at least once, by in QoL score. diabetic status and time of normalization.

www.eje-online.org

Downloaded from Bioscientifica.com at 09/23/2021 12:42:47PM via free access 330 A-J van der Lely and others EUROPEAN JOURNAL OF ENDOCRINOLOGY (2011) 164

80 clinically significant changes in mean ECG para- meters during the coadministration period and there were no clinical significant individual treatment- emergent changes. 60

Discussion 40 This study investigated the efficacy and safety of r = 0.52 P = 0.0022 coadministration of LA (120 mg/month) and pegviso- Pegvisomant dose (mg/week) 20 mant (40–120 mg/week), administered once or twice 0123456weekly. Combination therapy in 57 patients with IGF1 z-score at visit 3 acromegaly and partial response to SSA normalized IGF1 levels in 58% of patients at a median pegvisomant Figure 5 Correlation between IGF1 z-score at the end of run-in and dose of 60 mg once weekly. In all, 79% of patients had final weekly pegvisomant dose at normalization. All subjects normalized IGF1 levels at least once during coadminis- normalized at the end of the coadministration period (intent-to-treat tration. Patients with lower IGF1 levels at the end of population). z-scores are based on log-transformed IGF1 concentrations; a z-score between C2 and K2 indicates a run-in, including nondiabetic patients who had lower normal value; r represents Spearman’s rank correlation with P baseline IGF1 z-scores than their noninsulin-dependent value. The pegvisomant dose is based on the final weekly dosage diabetic counterparts, were more likely to achieve (i.e. 40 mg!2/weekZ80 mg/week). normal IGF1 levels during coadministration, with a lower pegvisomant dose. The higher pegvisomant dose required in hyperinsulinemic diabetic compared with to normal after treatment discontinuation. Gallbladder nondiabetic patients may reflect the higher number of ultrasound was normal in both cases. hepatic GH receptors in individuals with diabetes (19), O ! A transient increase in AST and/or ALT to 2 ULN necessitating higher concentrations of pegvisomant for occurred in 6 (11%) patients, including the two patients full receptor blockade (19, 20). The lower response rate ! in whom the increase exceeded 5 ULN and led in patients previously treated with pegvisomant may be to withdrawal. In the four patients who continued in confounded by the higher frequency of previous radio- the study, increases occurred during coadministration therapy and longer time since diagnosis in this patient and were either normalized within the coadministration subset, both indicators of more severe baseline disease. period (three patients) or shortly after (one patient; In the, albeit small, subset of patients whose disease was 3 months later while continuing to receive LA). similarly controlled on pegvisomant monotherapy and Gallbladder echography was unchanged in all four coadministration with LA, there was a significant fall patients. There was no relationship between diabetic in the required weekly dose of pegvisomant. status and elevated transaminases. No clinically The rates of IGF1 normalization observed in this significant changes were observed during coadministra- study are somewhat lower than those in similar studies tion in mean values for hematological or biochemical of combination therapy, which typically report bio- parameters, nor for liver function tests; individual chemical response in R95% of patients. In a study by abnormalities were consistent with TEAEs. A clinically Neggers et al. (18), 100% of patients achieved normal significant decrease in mean fasting insulin (K12.7 IGF1 after 138 weeks, with a median pegvisomant dose G41.5 pmol/l) was observed in nondiabetic patients of 60 mg/week. In a similar study by Feenstra et al. (12), after coadministration; mean blood glucose parameters normalization was reported in 95% of patients after remained otherwise normal. Gallbladder sludge and/or 42 weeks with the same median pegvisomant dose. lithiasis developed in four (7%) patients during There are several possible explanations for the lower coadministration, and preexisting sludge and/or response rates in this study. First, this study had a much lithiasis appeared to resolve in five (9%). The majority shorter duration, with a coadministration period of of these changes were seen in patients previously w26 weeks. Second, our two-step recruitment process treated with SSAs. excluded SSA responders, whose inclusion cannot be Magnetic resonance imaging revealed considerable entirely ruled out in studies that employ a single-step interindividual variability in tumor size. During the selection process. Third, our titration schedule aimed run-in period, tumor size decreased by O20% in 3/54 to determine the lowest effective pegvisomant dose, patients (5.6%) and increased by O20% in 9/54 whereas the aforementioned studies used an escalating patients (16.7%). Corresponding values for the titration schedule to achieve maximum efficacy. Conse- coadministration period were 7/53 patients (13.2%) quently, of the 31 patients in this study who responded and 13/53 patients (24.5%) respectively. There was at a dose of 60 mg/week, 13 no longer responded when no statistically significant correlation between change the dose was reduced to 40 mg/week, and only 8 in tumor size and IGF1 z-score. There were no responded again when the 60 mg dose was reinstated.

www.eje-online.org

Downloaded from Bioscientifica.com at 09/23/2021 12:42:47PM via free access EUROPEAN JOURNAL OF ENDOCRINOLOGY (2011) 164 Coadministration of LA and pegvisomant 331

IGF1 normalization rates in this study were, however, therapy appears to be well tolerated in long-term similar to those in the Trainer et al.’s (21) randomized therapy of up to 5 years (11), and some studies have controlled trial, which compared 5–30 mg/day shown reversal of elevated transaminases during (35–210 mg/week) pegvisomant monotherapy (56% continued pegvisomant treatment (11, 18), it was not response rate) with the same dose combined with possible in this study to evaluate the outcomes in octreotide LAR combination (62% response rate). Two patients with transaminases R5!ULN as the protocol points should be noted with respect to Trainer et al.’s mandated withdrawal under these circumstances; we study: first, it employed a single-step selection process remain cautious about continuing treatment in the face and may have included some SSA responders; second, of raised transaminases. the RIA for IGF1 was discontinued by the manufacturer Compared with monotherapy, coadministration with midway through the study, requiring a switch to a an SSA requires a lower dose of pegvisomant per week chemiluminescent assay, which was used to reanalyze to achieve a similar degree of efficacy (7, 23, 24), the previous IGF1 levels, and could have accounted through a variety of mechanisms. These may include for lower than expected IGF1 normalization rates. elevation of serum pegvisomant levels (13), reduced Key, short- (8) and long-term (7), studies of insulin concentration in the portal vein, which pegvisomant monotherapy document IGF1 normal- decreases the number of available hepatocyte GH ization rates of up to 90% using doses %20 mg/day receptors (25), and reduction in endogenous GH levels. (%140 mg/week) and R97% at doses up to 40 mg/day Pegvisomant is also a more expensive treatment than (%280 mg/week), although the highest normalization high-dose, long-acting SSA treatment (LA and long- rates were reported when the definition included acting octreotide) (26). It can therefore be presumed normalization at any time, not specifically at study that combination therapy will result in a cost benefit. end point (7). Interestingly, the combination of Limitations of this study include the open-label design pegvisomant with SSAs in this study appears to and almost exclusively Caucasian study population. have similar efficacy to pegvisomant monotherapy The final sample size of 57 coadministered patients, 15 mg/day (105 mg/week; w75% IGF1 normalization while lower than the planned cohort of 60, retained a after 12 weeks) and more effective than 10 mg/day power close to 95% to detect the difference between the (70 mg/week; w38% response) (8), a dose that exceeds null hypothesis of 30% minimum clinically relevant the median weekly dose in this study. Further to this, percentage of responders and the expected percentage although direct comparisons between our data and of responders of 80%; this difference was not anticipated those from uncontrolled studies should be made with to affect the validityof statistical testing. The study was not caution, coadministration of LA and pegvisomant in designed to determine the pegvisomant dose that would this study achieved similar efficacy to pegvisomant maximize efficacy,but rather the lowest dose necessary to monotherapy in the uncontrolled registry-based normalize IGF1 levels; it is possible that higher peg- ACROSTUDY (9). ACROSTUDY, in which 80% of visomant doses and/or a different therapeutic regimen patients were taking pegvisomant %20 mg/day might have achieved greater efficacy. There is also a (%140 mg/week) reported a 60–70% response rate margin of error relating to intraindividual variability in during 5 years of follow-up, while the median dose the IGF1 assay. Finally, the study was not designed to required to achieve a similar level of efficacy in this study examine changes in tumor size; indeed, most patients was considerably lower. Although the ACROSTUDY either had microadenomas or had undergone tumor results offer a perspective on outcomes that may be debulking surgery, and SSA-induced tumor shrinkage expected in the ‘real-life’ setting, the value of such was likely to be maximal during prior SSA treatment comparisons is limited by important differences in study rather than with SSA–pegvisomant combination therapy. design and patient selection. In conclusion, combination therapy using pegviso- In this study, combination therapy was well tolerated mant (40–120 mg/week) and LA (120 mg/month) in with only three serious adverse events leading to partial SSA responders normalized IGF1 levels at the treatment withdrawal, of which two were related to end of coadministration in 57.9% of patients at a treatment. Furthermore, the increase in transaminase median effective pegvisomant dose of 60 mg/week, and levels to R2!ULN was observed in six patients, of in 79% at any time during coadministration, at the whom only two had levels R5!ULN; as such, these same median weekly dose. Treatment was generally well rates were lower than those in other combination tolerated. Compared with pegvisomant monotherapy, therapy studies (11, 12, 18), and closer to the rates combination therapy with lanreotide offers the benefit observed during pegvisomant monotherapy (7, 8, 22) of reduced pegvisomant injection frequency. and were described in the ‘Summary of Product Characteristics’ for pegvisomant (10). It should also be Declaration of interest noted that the incidence of raised transaminases in this Prof. van der Lely has received financial support for investigator- study was no higher in diabetic than nondiabetic initiated research and unrestricted grants from Novartis, Pfizer, and patients, a finding that differs from certain other study Ipsen. Prof. Bernabeu has received lecture fees from Ipsen, Pfizer, and data (11). Although SSA/pegvisomant combination Novartis. Prof. Cap has received travel grants from Ipsen, Pfizer, and

www.eje-online.org

Downloaded from Bioscientifica.com at 09/23/2021 12:42:47PM via free access 332 A-J van der Lely and others EUROPEAN JOURNAL OF ENDOCRINOLOGY (2011) 164

Novartis. Prof. Caron had been a consultant for and received lecture 9 Trainer P. ACROSTUDY: the first 5 years. European Journal of fees from Ipsen, Pfizer, and Novartis. Prof. Colao is the recipient of Endocrinology 2009 161 S19–S24. (doi:10.1530/EJE-09-0322) research grants from Ipsen, Pfizer, and Novartis and is a member of 10 Pfizer Limited. Summary of Product Characteristics: SOMAVERT scientific advisory boards for Ipsen and Novartis. Prof. Marek has 10 mg, 15 mg, 20 mg powder and solvent for solution for received travel grants from Ipsen, Pfizer, and Novartis. Prof. Neggers injection. Available at http://emc.medicines.org.uk/medici- has no conflicts to declare. Dr Birman is an employee of Ipsen. ne/14353/SPC/SOMAVERT (accessed 25 October 2010). 11 Neggers SJ, de Herder WW, Janssen JA, Feelders RA & van der Lely AJ. Combined treatment for acromegaly with long-acting Funding somatostatin analogs and pegvisomant: long-term safety for up to 4.5 years (median 2.2 years) of follow-up in 86 patients. European This study was supported by Ipsen. Journal of Endocrinology 2009 160 529–533. (doi:10.1530/EJE- 08-0843) 12 Feenstra J, de Herder WW, ten Have SM, van den Beld AW, Acknowledgements Feelders RA, Janssen JA & van der Lely AJ. Combined therapy with The authors accept direct responsibility for this paper and are grateful somatostatin analogues and weekly pegvisomant in active for the contribution made by Watermeadow Medical (supported by acromegaly. Lancet 2005 365 1644–1646. (doi:10.1016/ Ipsen) in developing a first draft from an agreed outline and in S0140-6736(05)63011-5) collating comments. Thanks are due to Jean-Luc Beffy of Ipsen and 13 Jørgensen JO, Feldt-Rasmussen U, Frystyk J, Chen JW, Alison Langley for their invaluable statistical support throughout the Kristensen LØ, Hagen C & Ørskov H. Cotreatment of acromegaly study, analysis of the results, and preparation of the manuscript, and with a somatostatin analog and a receptor to clinical study manager Christine Seymour. antagonist. Journal of Clinical Endocrinology and Metabolism 2005 90 5627–5631. (doi:10.1210/jc.2005-0531) 14 Attanasio R, Lanzi R, Losa M, Valentini F, Grimaldi F, De Menis E, Davı` MV, Battista C, Castello R, Cremonini N, Razzore P, Rosato F, References Montini M & Cozzi R. Effects of lanreotide Autogel on growth hormone, insulin-like growth factor I, and tumor size in 1 Melmed S, Colao A, Barkan A, Molitch M, Grossman AB, acromegaly: a 1-year prospective multicenter study. Endocrine Kleinberg D, Clemmons D, Chanson P, Laws E, Schlechte J, Practice 2008 14 846–855. Vance ML, Ho K, Giustina A & Acromegaly Consensus Group. 15 Colao A, Auriemma RS, Rebora A, Galdiero M, Resmini E, Guidelines for acromegaly management: an update. Journal of Minuto F, Lombardi G, Pivonello R & Ferone D. Significant Clinical Endocrinology and Metabolism 2009 94 1509–1517. tumour shrinkage after 12 months of lanreotide Autogel- (doi:10.1210/jc.2008-2421) 120 mg treatment given first-line in acromegaly. Clinical 2 Murray RD & Melmed S. A critical analysis of clinically Endocrinology 2009 71 237–245. (doi:10.1111/j.1365-2265. available somatostatin analog formulations for therapy of 2008.03503.x) acromegaly. Journal of Clinical Endocrinology and Metabolism 2008 16 Mazziotti G & Giustina A. Effects of lanreotide SR and Autogel on 93 2957–2968. (doi:10.1210/jc.2008-0027) tumor mass in patients with acromegaly: a systematic review. 3 Cook DM, Ezzat S, Katznelson L, Kleinberg DL, Laws ER Jr, Pituitary 2010 13 60–67. (doi:10.1007/s11102-009-0169-z) Nippoldt TB, Swearingen B, Vance ML & AACE Acromegaly 17 Neggers SJ, van Aken MO, de Herder WW, Feelders RA, Guidelines Task Force . AACE medical guidelines for clinical Janssen JA, Badia X, Webb SM & van der Lely AJ. Quality of practice for the diagnosis and treatment of acromegaly. Endocrine life in acromegalic patients during long-term somatostatin Practice 2004 10 213–225. analog treatment with and without pegvisomant. Journal of 4 Goffin V, Bernichtein S, Carrie`re O, Bennett WF, Kopchick JJ & Clinical Endocrinology and Metabolism 2008 93 3853–3859. Kelly PA.The human growth hormone antagonist B2036 does not (doi:10.1210/jc.2008-0669) interact with the prolactin receptor. Endocrinology 1999 140 18 Neggers SJ, van Aken MO, Janssen JA, Feelders RA, de Herder WW 3853–3856. (doi:10.1210/en.140.8.3853) & van der Lely AJ. Long-term efficacy and safety of combined 5 Fuh G, Cunningham BC, Fukunaga R, Nagata S, Goeddel DV & treatment of somatostatin analogs and pegvisomant in acrome- Wells JA. Rational design of potent antagonists to the human galy. Journal of Clinical Endocrinology and Metabolism 2007 92 . Science 1992 256 1677–1780. (doi:10. 4598–4601. (doi:10.1210/jc.2007-1234) 1126/science.256.5064.1677) 19 Kratzch J, Keliner K, Zilkens T, Schmidt-Gayk H, Selisko T & 6 Higham CE, Thomas JD, Bidlingmaier M, Drake WM & Trainer PJ. Scholz GH. Growth hormone-binding protein related immuno- Successful use of weekly pegvisomant administration in patients reactivity is regulated by the degree of insulinopenia in diabetes with acromegaly. European Journal of Endocrinology 2009 161 mellitus. Clinical Endocrinology 1996 44 673–678. (doi:10.1046/ 21–25. (doi:10.1530/EJE-08-0990) j.1365-2265.1996.672494.x) 7 van der Lely AJ, Hutson RK, Trainer PJ, Besser GM, Barkan AL, 20 Wurzburger MI, Prelevic GM, So¨nksen PH, Wheeler M & Balint- Katznelson L, Klibanski A, Herman-Bonert V,Melmed S, Vance ML, Peric L. Effect of recombinant human growth hormone treatment Freda PU, Stewart PM, Friend KE, Clemmons DR, Johannsson G, on insulin-like growth factor (IGF-I) levels in insulin-dependent Stavrou S, Cook DM, Phillips LS, Strasburger CJ, Hackett S, Zib KA, diabetic patients. Acta Diabetologica 1995 32 131–134. (doi:10. Davis RJ, Scarlett JA & Thorner MO. Long-term treatment of 1007/BF00569572) acromegaly with pegvisomant, a growth hormone receptor 21 Trainer PJ, Ezzat S, D’Souza GA, Layton G & Strasburger CJ. A antagonist. Lancet 2001 358 1754–1759. (doi:10.1016/S0140- randomized, controlled, multicentre trial comparing pegvisomant 6736(01)06844-1) alone with combination therapy of pegvisomant and long-acting 8 Trainer PJ, Drake WM, Katznelson L, Freda PU, Herman-Bonert V, octreotide in patients with acromegaly. Clinical Endocrinology 2009 van der Lely AJ, Dimaraki EV, Stewart PM, Friend KE, Vance ML, 71 549–557. (doi:10.1111/j.1365-2265.2009.03620.x) Besser GM, Scarlett JA, Thorner MO, Parkinson C, Klibanski A, 22 Biering H, Saller B, Bauditz J, Pirlich M, Rudolph B, Johne A, Powell JS, Barkan AL, Sheppard MC, Malsonado M, Rose DR, Buchfelder M, Mann K, Droste M, Schreiber I, Lochs H, Clemmons DR, Johannsson G, Bengtsson BA, Stavrou S, Strasburger CJ & German pegvisomant investigators . Elevated Kleinberg DL, Cook DM, Phillips LS, Bidlingmaier M, transaminases during medical treatment of acromegaly: a review Strasburger CJ, Hackett S, Zib K, Bennett WF & Davis RJ. of the German pegvisomant surveillance experience and a report Treatment of acromegaly with the growth hormone-receptor of a patient with histologically proven chronic mild active antagonist pegvisomant. New England Journal of Medicine 2000 hepatitis. European Journal of Endocrinology 2006 154 213–220. 342 1171–1177. (doi:10.1056/NEJM200004203421604) (doi:10.1530/eje.1.02079)

www.eje-online.org

Downloaded from Bioscientifica.com at 09/23/2021 12:42:47PM via free access EUROPEAN JOURNAL OF ENDOCRINOLOGY (2011) 164 Coadministration of LA and pegvisomant 333

23 Colao A, Pivonello R, Auriemma RS, De Martino MC, 25 Leung KC, Doyle N, Ballesteros M, Waters MJ & Ho KK. Insulin Bidlingmaier M, Briganti F, Tortora F, Burman P, Kourides IA, regulation of human hepatic growth hormone receptors: Strasburger CJ & Lombardi G. Efficacy of 12-month treatment with divergent effects on biosynthesis and surface translocation. the GH receptor antagonist pegvisomant in patients with Journal of Clinical Endocrinology and Metabolism 2000 85 acromegaly resistant to long-term, high-dose somatostatin analog 4712–4720. (doi:10.1210/jc.85.12.4712) treatment: effect on IGF-I levels, tumor mass, hypertension and 26 Moore D, Adi Y, Connock M & Bayliss S. Clinical effectiveness and glucose tolerance. European Journal of Endocrinology 2006 154 cost-effectiveness of pegvisomant for the treatment of acromegaly: 467–477. (doi:10.1530/eje.1.02112) a systematic review and economic evaluation. BMC Endocrine 24 Schreiber I, Buchfelder M, Droste M, Forssmann K, Mann K, Saller B, Disorders 2009 9 20. (doi:10.1186/1472-6823-9-20) Strasburger CJ & German Pegvisomant Investigators. Treatment of acromegaly with the GH receptor antagonist pegvisomant in clinical practice: safety and efficacy evaluation from the German Pegvisomant Observational Study. European Journal of Endocrinology Received 8 December 2010 2007 156 75–82. (doi:10.1530/eje.1.02312) Accepted 10 December 2010

www.eje-online.org

Downloaded from Bioscientifica.com at 09/23/2021 12:42:47PM via free access