All-cause mortality in patients with acromegaly treated with pegvisomant: an ACROSTUDY analysis

in European Journal of Endocrinology
View More View Less
  • 1 Neuroendocrine Unit, Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts, USA
  • | 2 Pfizer Endocrine Care, Pfizer Health AB, Sollentuna, Sweden
  • | 3 Division of Endocrinology and Metabolism, Department of Internal Medicine III, Medical University of Vienna, Vienna, Austria
  • | 4 Global Biometrics and Data Management, Pfizer Inc, New York, New York, USA
  • | 5 Pfizer Endocrine Care, Inc, New York, New York, USA
  • | 6 Department of Endocrine Neoplasia and Hormonal Disorders, University of Texas MD Anderson Cancer Center, Houston, Texas, USA

Correspondence should be addressed to N A Tritos; Email: ntritos@mgh.harvard.edu
Free access

Objective

To examine all-cause mortality rates in patients with acromegaly on pegvisomant and identify pertinent risk factors, including insulin-like growth factor I (IGF-I).

Design

Retrospective cohort analysis of data from ACROSTUDY (global surveillance study of patients with acromegaly treated with pegvisomant).

Methods

Kaplan–Meier analyses and Cox regression techniques were used to examine survival rates. Standardized mortality ratios (SMR) with reference to general population (WHO GBD 2016) were estimated. Multiplicative multiple Poisson regression models were used to characterize the association between SMR, IGF-I, and other risk factors associated with mortality risk.

Results

The study consisted of 2077 subjects who were followed for a median interval of 4.1 years, contributing to 8957 patient-years. Higher on-treatment IGF-I (P = 0.0035), older attained age (P < 0.0001), and longer duration of acromegaly (>10 years) before starting pegvisomant (P = 0.05) were associated with higher mortality rates. In reference to general population rates, higher SMR (1.10, 1.42, and 2.62, at attained age 55 years) were observed with higher serum IGF-I category (SMR trend: 1.44 (44%)/per fold level of IGF-I/ULN (95% CI: 1.10, 1.87), P = 0.0075). SMR increased per year of younger attained age (1.04 (1.02–1.04), P < 0.0001) and were higher for longer disease duration (>10 years) before starting pegvisomant (1.57 (1.02, 2.43), P = 0.042). Serum IGF-I levels within the normal range during pegvisomant therapy were associated with all-cause mortality rates that were indistinguishable from the general population.

Conclusions

Higher on-treatment IGF-I, older attained age, and longer duration of acromegaly before starting pegvisomant are associated with higher all-cause mortality rates. Younger patients with uncontrolled acromegaly have higher excess all-cause mortality rates in comparison with older patients.

Abstract

Objective

To examine all-cause mortality rates in patients with acromegaly on pegvisomant and identify pertinent risk factors, including insulin-like growth factor I (IGF-I).

Design

Retrospective cohort analysis of data from ACROSTUDY (global surveillance study of patients with acromegaly treated with pegvisomant).

Methods

Kaplan–Meier analyses and Cox regression techniques were used to examine survival rates. Standardized mortality ratios (SMR) with reference to general population (WHO GBD 2016) were estimated. Multiplicative multiple Poisson regression models were used to characterize the association between SMR, IGF-I, and other risk factors associated with mortality risk.

Results

The study consisted of 2077 subjects who were followed for a median interval of 4.1 years, contributing to 8957 patient-years. Higher on-treatment IGF-I (P = 0.0035), older attained age (P < 0.0001), and longer duration of acromegaly (>10 years) before starting pegvisomant (P = 0.05) were associated with higher mortality rates. In reference to general population rates, higher SMR (1.10, 1.42, and 2.62, at attained age 55 years) were observed with higher serum IGF-I category (SMR trend: 1.44 (44%)/per fold level of IGF-I/ULN (95% CI: 1.10, 1.87), P = 0.0075). SMR increased per year of younger attained age (1.04 (1.02–1.04), P < 0.0001) and were higher for longer disease duration (>10 years) before starting pegvisomant (1.57 (1.02, 2.43), P = 0.042). Serum IGF-I levels within the normal range during pegvisomant therapy were associated with all-cause mortality rates that were indistinguishable from the general population.

Conclusions

Higher on-treatment IGF-I, older attained age, and longer duration of acromegaly before starting pegvisomant are associated with higher all-cause mortality rates. Younger patients with uncontrolled acromegaly have higher excess all-cause mortality rates in comparison with older patients.

Introduction

Acromegaly has been associated with increased all-cause mortality, attributed to higher cardiovascular and cerebrovascular mortality (1, 2, 3, 4, 5). Whether malignancy-related morbidity and mortality is increased in patients with acromegaly is less certain (6, 7). Achieving biochemical control of excess growth hormone (GH) secretion in patients with acromegaly mitigates the excess mortality risk (8).

Pegvisomant is a GH receptor antagonist that effectively normalizes serum insulin-like growth factor I (IGF-I) levels in the majority of patients with acromegaly (9, 10, 11, 12, 13, 14). However, published data on all-cause mortality rates in patients with acromegaly on pegvisomant therapy in comparison with the general population are not available. Furthermore, it is not known whether IGF-I normalization in patients receiving pegvisomant therapy is sufficient to blunt excessive mortality. ACROSTUDY is a global pharmaco-epidemiological surveillance study that was open to patients with acromegaly who were treated with pegvisomant in routine clinical care (11, 15, 16).

The aim of the present study was to characterize all-cause mortality rates in patients with acromegaly treated with pegvisomant in comparison with the general population, using data extracted from ACROSTUDY, and examine risk factors for mortality in this patient population, including on-treatment serum IGF-I levels, attained age (as discussed subsequently), and patient comorbidities. The study hypothesis was that normal IGF-I levels, achieved during therapy with pegvisomant, are associated with mortality rates that are not higher than those in the general population.

Patients and methods

Patients who were participating in ACROSTUDY were managed by clinicians according to local medical practice. All study subjects provided written informed consent before entering ACROSTUDY. The study was approved by local ethics committees at each participating center (please see list under ‘Supplementary data’, see section on supplementary materials given at the end of this article) and conducted based on the principles of the Declaration of Helsinki (17).

All information used in the present study was extracted from database records, including age at diagnosis of acromegaly, age at entry into ACROSTUDY, attained age (defined subsequently), gender, serum IGF-I levels (measured locally at each participating site and expressed as ‘fold levels’ above the upper limit of the respective normal range (ULN)), presence of pituitary hormone deficiencies, comorbidities at study entry (hypertension, diabetes mellitus, cardiovascular or cerebrovascular disease, and malignancies), interval between the diagnosis of acromegaly and pegvisomant start, pegvisomant dose, additional treatments for acromegaly (surgery, radiation therapy, and medical therapies besides pegvisomant), birth date, first visit date in ACROSTUDY, last visit date in ACROSTUDY, and if patient died, date of death. Attained age is the age a patient ran through during follow-up from the first visit date to the last date in ACROSTUDY. It is calculated based on date of birth and all dates during follow-up, taking into consideration the number of patient-years accrued during follow-up. For example, if a patient entered follow-up at 48 years of age and exited the day before (s)he turned 55 years of age, the attained ages during follow-up were 48, 49, 50, 51, 52, 53, through 54 years. This patient accrued 7 patient-years during follow-up. Attained age can be classified, for instance, in 30–49 and 50–59 year classes, which means that this patient, in such a case, accrued 2 patient-years in 30–49 year classes and 5 patient-years in the 50–59 year attained age class.

All patients in ACROSTUDY who had available serum IGF-I data at study entry and during follow-up were included in the present report. Data on 142 patients and 2 deaths had to be excluded because of missing IGF-I values. The continuous (numerical) IGF-I/ULN was classified time-dependently, at blood sample dates, and a patient could contribute with patient-years and be classified in up to six different IGF-I/ULN classes during follow-up, depending on the measurement values at the different sample dates during follow-up. Classes are 1: <0.68, 2: 0.68 – <1.0, 3: 1.0 – <1.25, 4: 1.25 – <2.0, 5: 2.0 – <2.57, and 6: ≥2.57 (of IGF-I/ULN). In trend analyses, class values were represented by their mean value (i.e. from class 1 to 6: 0.52, 0.84, 1.11, 1.53, 2.24, and 3.59). In table presentations, the number of classes were reduced to three and given as <1 (within the reference range), 1 – <2, and ≥2 IGF-I/ULN.

Kaplan–Meier analyses and Cox regression techniques were used to analyze data on survival and present age and gender adjusted all-cause mortality hazards within the study population. Standardized mortality ratios (SMR) were calculated using the World Health Organization (WHO) Global Burden of Disease 2016 database (18) as general population mortality reference rates, and ACROSTUDY data were stratified in alignment to these WHO data by (attained) age, gender, and country (where the patient was treated and followed-up). Attained age were categorized as: 15–29, 30–49, 50–59, 60–69, and 70+ years. When attained age was represented as a numerical variable in the analyses, a patient-year mean weighted class midpoint was used. Fifteen countries contributed data in ACROSTUDY; Germany (2331 patient-years), Italy (2084), France (1400), Spain (870), USA (748), and Netherlands (671) were the largest contributing countries with over 90% of total patient-years. Other countries that contributed data were Austria, Belgium, Denmark, UK, Greece, Hungary, Portugal, Sweden, and Slovakia. Multiplicative multiple Poisson regression analyses were conducted to characterize the association between on-therapy IGF-I levels and SMR while controlling for potential confounders. The percentage change in trend estimates per fold level of IGF-I/ULN was calculated both for the crude estimates as well as for adjusted trend estimates.

Statistical multiple regression analyses were performed using SAS procedures PROC PHREG, PROC GENMOD (SAS institute, Inc, Cary, NC, USA). Unless otherwise noted, data are presented as median (10th percentile, 90th percentile) or percentages as appropriate. P values below 0.05 were considered as statistically significant.

Results

Demographic and clinical characteristics of the study population are shown in Table 1. There were 2077 subjects with acromegaly, who were followed for a median interval of 4.1 years (range: <1 to 13.1 years), contributing 8957 patient-years (PY) of follow-up during ACROSTUDY. Pegvisomant was initiated before ACROSTUDY entry in 95% of patients at a median interval of 0.9 years before ACROSTUDY enrollment (5th percentile: 5.2 years before ACROSTUDY start, 95th percentile: 0.01 year after ACROSTUDY start). Pegvisomant was administered at a median dose of 10 mg daily at ACROSTUDY entry and was titrated toward normal serum IGF-I levels at the discretion of treating clinicians at each participating site. There were 85 deaths (37 in women, 48 in men) during follow-up, yielding a crude mortality rate of 878.1/100,000 PY in women and 1011.9/100,000 PY in men. All-cause mortality rates were higher among patients who were older at entry into ACROSTUDY (Fig. 1).

Figure 1
Figure 1

Survival probability in patients with acromegaly, stratified by age at ACROSTUDY entry.

Citation: European Journal of Endocrinology 182, 3; 10.1530/EJE-19-0794

Table 1

Demographic and clinical characteristics of the study population.

Variablesn (%) of patients
Total n2077
Demographics and clinical characteristics
 Age at diagnosis of acromegaly
  <35 years683 (32.9%)
  35–50 years801 (38.6%)
  >50 years587 (28.3%)
  Missing6 (0.3%)
 Interval between diagnosis of acromegaly and initiation of pegvisomant therapy
  <2.5 years710 (34.2%)
  2.5–10 years820 (39.5%)
  >10 years541 (26.0%)
  Missing6 (0.3%)
 Age at ACROSTUDY entry (years) 51.6 (32.0, 69.6)*
  Gender
  Males1068 (51.4%)
  Females1009 (48.6%)
 IGF-I/ULN at ACROSTUDY entry
  ≤1*ULN1052 (50.7%)
  >1-2)*ULN767 (36.9%)
  >2*ULN258 (12.4%)
Comorbidities at ACROSTUDY entry
 Cardiovascular disease332 (15.9%)
 Cerebrovascular disease36 (1.7%)
 Diabetes mellitus600 (28.9%)
 Hypertension820 (39.5%)
 Hypopituitarism526 (25.3%)
 Malignancy 69 (3.3%)
Treatment
 Pituitary surgery2005 (96.5%)
 Radiation therapy648 (31.2%)
 Medical therapy before ACROSTUDY entry1723 (82.9%)
 Pegvisomant dose (mg) at ACROSTUDY entry 10 (5, 20)*
 Combination medical therapy during ACROSTUDY894 (43.0%)

*Values indicate median (10th, 90th percentile).

IGF-I: insulin-like growth factor I; ULN: upper limit of normal range.

In proportional hazards regression, all-cause mortality risk (hazard ratios) increased with higher serum IGF-I levels during follow-up (hazard ratio: 1 (referent), 1.19, 1.94, for IGF-I/ULN classes: <1, 1 – <2, ≥2, respectively, Table 2). In addition, older age at start of follow-up (P < 0.0001) and longer duration of acromegaly (>10 years) before starting pegvisomant therapy (P = 0.05) were associated with higher mortality risk. Patients with diabetes mellitus, cardiovascular, or cerebrovascular disease at study entry showed higher mortality risks, albeit not statistically significant.

Table 2

All-cause mortality hazard ratios as estimated in proportional hazards regression, classified by serum IGF-I levels, in patients with acromegaly followed on pegvisomant therapy during ACROSTUDY follow-up.

Ratio: IGF-I/ULNCrude estimatesAge at ACROSTUDY entry and gender – adjusted estimates
Crude mortality/100,000 PYHR95% CIP valueHR95% CIP value
Time-dependent IGF-I level/ULN in ACROSTUDY
 0. <1860.01 (ref)1 (ref)
 1. 1–21003.71.190.74–1.900.481.270.79–2.030.33
 2. >21580.11.940.95–3.980.072.631.28–5.400.009

CI, confidence intervals; HR, hazard ratio; IGF-I, insulin-like growth factor I; PY, person-years; ref, referent; ULN, upper limit of normal range.

Standardized mortality ratios (SMR) were calculated for the entire cohort (SMR, 95% confidence intervals (CI): 0.86 (0.69, 1.07)) and were stratified by gender (women: 0.78 (0.55, 1.08); men: 0.93 (0.69, 1.24)) and classified for serum IGF-I levels in patients on pegvisomant therapy as presented in Table 3.

Table 3

Standardized mortality ratios (SMR) in patients with acromegaly on pegvisomant therapy, stratified by serum IGF-I levels during follow-up.

IGF-I (time-dependent) Number of observed deathsNumber of PYCrude mortality/100 000 PYExpected deaths vs WHO 2016Crude SMR (Observed/expected deaths)95% CI
<1*ULN495698860.066.200.740.55–0.98
(>1 to 2)*ULN2726901003.727.740.970.64–1.42
>2*ULN95701580.14.721.910.87–3.62

* indicates multiplication (times above ULN) CI, confidence intervals; IGF-I, insulin-like growth factor I; PY, person-years; ULN, upper limit of normal range; WHO, World Health Organization.

The Poisson regression model with external reference and time-dependent IGF-I/ULN, attained age, and gender is presented in Table 4. The crude SMR was higher among patients with elevated serum IGF-I levels on therapy and even so after controlling for attained age and gender in the multiplicative multiple Poisson regression model (Table 4). SMR changes in a statistically significant manner with attained age and a common overall SMR is not a good representation of the results. SMRs must be specified for a certain attained age as presented in Table 4. For example, at attained age 55 years (median age for patients at the end of ACROSTUDY follow-up), the SMR was estimated to 1.10 in the lowest category <1*ULN, 1.42 in (1-2)*ULN, and 2.62 for IGF-I/ULN >2. (Table 4) and the attained age and gender-adjusted trend was estimated to 1.48 per IGF-I category (or 48% increase per fold-level of IGF-I/ULN).

Table 4

Standardized mortality ratios (SMR) in patients on pegvisomant therapy, stratified by serum IGF-I, attained age and gender.

IGF-I (time-dependent)Crude SMRPoisson regression: Attained age and gender adjusted SMR
(95% CI)Relative to <1*ULN40 years50 years55 years65 years70 yearsRelative to <*ULN
<1*ULN0.74 (0.55–0.98) 1 (ref)1.72 (0.96–3.08)1.28 (0.83–1.97)1.10 (0.77–1.59)0.82 (0.62–1.45)0.71 (0.53–0.94)1 (ref)
(0.55–0.98)(0.96–3.08)(0.83–1.97)(0.77–1.59)(0.62–1.45)(0.53–0.94)
(1 to 2)*ULN0.97 (0.64–1.42)1.32 (0.82–2.10)2.21 (1.20–4.07)1.64 (1.02–2.65)1.42 (0.92–2.17)1.05 (0.72–1.54)0.91 (0.62–1.33)1.28 (0.80–2.05)
(0.64–1.42)(0.82–2.10)(1.20–4.07)(1.02–2.65)(0.92–2.17)(0.72–1.54)(0.62–1.33)(0.80–2.05)
>2*ULN1.91 (0.87–3.62)2.57 (1.26–5.24)4.09 (1.86–9.00)3.04 (1.51–6.14)2.62 (1.33–5.14)1.94 (1.01–3.75)1.68 (0.86–3.26) 2.37 (1.16–4.84)
(0.87–3.62)(1.26–5.24)(1.86–9.00)(1.51–6.14)(1.33–5.14)(1.01–3.75)(0.86–3.26)(1.16–4.84)
Trend*: SMR per IGF-I/ULN fold-level (CI)1.52 (1.17–1.98)1.48 (1.13–1.93)
P value for trend*0.0020.004

*Trend estimates are based on six IGF-I/ULN classes (<0.7, 07 to <1.0, 1.0 to <1.25, 1.25 to 2.0, 2.0 to 2.57, and ≥2.57), where class values are represented by respective mean value (0.52, 0.84, 1.11, 1.53, 2.24, and 3.59).

CI: confidence intervals; IGF-I: insulin-like growth factor I; ref: referent; ULN: upper limit of normal range.

Additional potential confounders were subsequently incorporated and tested one by one: diabetes mellitus (1.54 (1.00, 2.42), P = 0.048), cardiovascular disease (1.65 (1.04, 2.60), P = 0.033) and cerebrovascular disease (2.34 (1.07, 5.11), P = 0.034), all at ACROSTUDY entry, and longer duration of acromegaly (>10 years vs <10 years) before starting pegvisomant therapy (1.57 (1.01, 2.43), P = 0.043) were associated with higher SMR. In contrast, age at diagnosis of acromegaly, hypopituitarism, hypertension, history of malignancy, radiation therapy, medical therapy before pegvisomant start or combination medical therapy during follow-up were not significantly associated with different levels of SMR. Adjustments by these potential confounders (both the significant and not significant) did not attenuate the association between serum IGF-I levels and SMR, which means that the IGF-I/ULN trend estimate was relatively unaffected by the 11 potential confounders. Indeed, the change in trend estimate varied between -2.7% and 0.7%, when potential confounders were tested one by one in the “reference model” (Table 4) that includes IGF-I/ULN, attained age and gender.

The multiple regression model presented in Table 5 is an extension of the “reference model” shown in Table 4 and incorporates 4 potential confounders/risk indicators (duration of acromegaly, diabetes mellitus, cardiovascular disease, cerebrovascular disease). In this model, the change in IGF-I/ULN trend estimate was only -2.7% (from 1.48 to 1.44). In this multiple (multivariable) Poisson regression model with external reference, incorporating the 4 potential confounders mentioned above, higher on-treatment serum IGF-I levels, younger attained age, and longer duration of acromegaly before starting pegvisomant therapy were associated with higher SMR (Table 5). The associations between diabetes mellitus, cardiovascular disease and cerebrovascular disease at ACROSTUDY entry with higher SMR were of borderline statistical significance in this larger multiple regression model, although the effect estimates were similar.

Table 5

Risk factors in the external reference multiple Poisson regression model estimating standardized mortality ratios (SMR) in patients on pegvisomant therapy.

Predictor variableSMR ratio95% CIChi square (χ2 )P value
IGF-I/ULN (time-dependent) (SMR, per fold-level)* 1.441.101.877.150.0075
Attained age (year) (SMR, relative change per year) 0.960.940.9817.61<0.0001
Gender (SMR ratio, F vs M) 0.960.621.500.030.87
Diabetes mellitus at study entry (SMR ratio, Y vs N)1.490.962.333.110.078
Duration of acromegaly before pegvisomant start (SMR ratio, >10 years vs <10 years)1.571.022.434.120.042
Cardiovascular disease at study entry (SMR ratio, Y vs N)1.510.952.403.090.079
Cerebrovascular disease at study entry (SMR ratio, Y vs N)2.070.944.563.250.071

*Trend estimates were based on 6 IGF-I/ULN classes (<0.7, 07 to <1.0, 1.0 to <1.25, 1.25 to 2.0, 2.0 to 2.57, and ≥2.57), where class values are represented by respective mean value (0.52, 0.84, 1.11, 1.53, 2.24, and 3.59).

CI: confidence intervals; IGF-I: insulin-like growth factor I; RR: relative risk; ULN: upper limit of normal range.

The corresponding Poisson regression model with internal reference (Table 6) showed similar results as the model with external reference, except for attained age, which showed that mortality rates increased by attained age (Fig. 2 and Table 6) in contrast to the external reference model (Table 5), where SMR (i.e. the relative rates in relation to the general population rates) decreased by age.

Figure 2
Figure 2

Mortality rates shown in relation to attained age in patients with acromegaly participating in ACROSTUDY (continuous line) and individuals in the general population in ACROSTUDY countries (dashed line).

Citation: European Journal of Endocrinology 182, 3; 10.1530/EJE-19-0794

Table 6

All-cause mortality rate ratios (RR) in patients with acromegaly followed on pegvisomant therapy based on a multiplicative multiple Poisson regression model with internal reference.

Risk indicatorRR95% CIChi square (χ2)P value
Trend over time-dependent IGF-I/ULN* (RR per fold-level)1.491.141.948.520.0035
Attained age (RR change per year i.e. increase by 7%)1.0751.051.1040.69<0.0001
Gender (RR female vs male)0.710.461.102.320.13
Interval between diagnosis of acromegaly and pegvisomant start (RR for >10 years relative to <10 years)1.541.002.373.760.053
Diabetes mellitus at study entry (RR for Y vs N)1.450.932.262.750.097
Cardiovascular disease at ACROSTUDY entry (RR Y vs N)1.510.952.393.020.082
Cerebrovascular disease at ACROSTUDY entry (RR Y vs N)2.261.034.984.100.043

*Trend estimates were based on 6 IGF-I/ULN classes (<0.7, 07 to <1.0, 1.0 to <1.25, 1.25 to 2.0, 2.0 to 2.57, and ≥2.57), where class values are represented by respective mean value (0.52, 0.84, 1.11, 1.53, 2.24, and 3.59).

CI: confidence intervals; RR: rate ratio; IGF-I: insulin-like growth factor I; ULN: upper limit of normal range.

Discussion

The present study examined all-cause mortality rates in patients with acromegaly on pegvisomant therapy in relation to on-treatment IGF-I levels, duration of acromegaly, age, and several comorbidities. Acromegaly has been associated with higher all-cause mortality rates in comparison with the general population (1). Recently reported data have suggested a decrease in mortality rates among patients with acromegaly over time, likely as a consequence of improved patient care (2, 3, 19, 20). Our findings are consistent with these observations. Crude SMR data in our cohort indicate that the all-cause mortality rate was not different from that in the general population.

Previous studies have reported that biochemical control of acromegaly, primarily achieved after pituitary surgery, mitigates the excess mortality risk in this population (8). However, there is debate whether GH or IGF-I levels (or both) are important predictors of mortality risk in acromegaly. Indeed, some data suggest that patients who achieve IGF-I normalization have survival rates that are comparable to those in the general population (8). On the other hand, other studies have reported that low GH levels, rather than IGF-I levels, are important predictors of survival in this population (21, 22, 23).

The findings of the present study suggest that on-therapy serum IGF-I levels are important risk factors for all-cause mortality in patients with acromegaly on pegvisomant therapy. In this cohort, IGF-I normalization is associated with all-cause mortality rates that are indistinguishable from those in the general population. These observations suggest that IGF-I normalization on pegvisomant therapy may be sufficient to optimize patient survival in acromegaly. Even though 43% of patients received combination therapy, this was not an independent predictor of SMR. Of note, patients with acromegaly of longer duration (>10 years) at ACROSTUDY entry had higher mortality rates in the present cohort. Cumulative exposure to GH excess may indeed predict morbidity and mortality in acromegaly (24, 25). In aggregate, these findings suggest the importance of achieving prompt disease control in order to mitigate the risk of excess mortality in acromegaly.

The present study also identified an excess mortality risk among younger patients with uncontrolled acromegaly (over the general population) in comparison with older patients, even though the latter subgroup had higher absolute mortality rates (as anticipated). Younger patients may have fewer competing health risks and are therefore more vulnerable to the excess mortality risk associated with GH excess. Alternatively, it is possible that younger patients may have more severe disease. In any case, the present observations further underscore the urgency to diagnose and tightly control acromegaly earlier in life.

The present study has several strengths, including the large size of the study population and substantial follow-up interval. The median follow-up (4.1 years) and maximal follow-up (13.1 years) intervals were substantial. However, it must be acknowledged that some patients were followed for a short time (<1 year) during the study. There are also study limitations as a consequence of its retrospective cohort design. Specifically, patients were managed at each participating site according to local clinical practice. It is therefore likely that pegvisomant therapy was not sufficiently uptitrated in some patients. It is not clear from the present data on how long serum IGF-I levels need to remain normal in order to mitigate a higher risk of all-cause mortality. As combination therapies varied among patients that received pegvisomant together with other agents, it was not possible to analyze data in this subgroup separately. It is possible that medications used to treat hypertension and diabetes mellitus might influence SMR in this population; however, detailed data on the use of these medications were not available and could not be incorporated in the present analysis.

In summary, higher on-treatment IGF-I levels, older chronologic age, and longer duration of GH excess before starting pegvisomant therapy are associated with higher all-cause mortality rates in patients with acromegaly treated with pegvisomant. Serum IGF-I levels within the normal range during pegvisomant therapy are associated with all-cause mortality rates that are indistinguishable from those in the general population. Younger patients with uncontrolled acromegaly have a higher risk of excess all-cause mortality (vs the general population) in comparison with older patients.

Supplementary materials

This is linked to the online version of the paper at https://doi.org/10.1530/EJE-19-0794.

Declaration of interest

G V is a member of the KIMS and BMKB is a member of the ACROSTUDY steering committee, respectively. A F M, S V, and J H-H are full-time employees of Pfizer Inc; N K was a full-time employee of Pfizer Inc. at the time of the manuscript development. ACROSTUDY is sponsored by Pfizer Inc. N A T, G V, B M K B, A K, and C J were not compensated for their participation in the present study. The investigators express their gratitude to all patients, site investigators, nurses, and other staff that participated in ACROSTUDY. Upon request, and subject to certain criteria, conditions, and exceptions (see https://www.pfizer.com/science/clinical-trials/trial-data-and-results for more information), Pfizer will provide access to individual de-identified participant data from Pfizer-sponsored global interventional clinical studies conducted for medicines, vaccines, and medical devices (1.) for indications that have been approved in the US and/or EU or (2.) in programs that have been terminated (i.e. development for all indications has been discontinued). Pfizer will also consider requests for the protocol, data dictionary, and statistical analysis plan. Data may be requested from Pfizer trials 24 months after study completion. The de-identified participant data will be made available to researchers whose proposals meet the research criteria and other conditions, and for which an exception does not apply, via a secure portal. To gain access, data requestors must enter into a data access agreement with Pfizer.

Funding

This work did not receive any specific grant from any funding agency in the public, commercial, or not-for-profit sector.

References

  • 1

    Katznelson L, Laws ER Jr, Melmed S, Molitch ME, Murad MH, Utz A, Wass JA & Endocrine Society. Acromegaly: an endocrine society clinical practice guideline. Journal of Clinical Endocrinology and Metabolism 2014 39333951. (https://doi.org/10.1210/jc.2014-2700)

    • Search Google Scholar
    • Export Citation
  • 2

    Bolfi F, Neves AF, Boguszewski CL, Nunes-Nogueira VS. Mortality in acromegaly decreased in the last decade: a systematic review and meta-analysis. European Journal of Endocrinology 2018 5971. (https://doi.org/10.1530/EJE-18-0255)

    • Search Google Scholar
    • Export Citation
  • 3

    Esposito D, Ragnarsson O, Granfeldt D, Marlow T, Johannsson G, Olsson DS. Decreasing mortality and changes in treatment patterns in patients with acromegaly from a nationwide study. European Journal of Endocrinology 2018 459469. (https://doi.org/10.1530/EJE-18-0015)

    • Search Google Scholar
    • Export Citation
  • 4

    Colao A, Grasso LFS, Giustina A, Melmed S, Chanson P, Pereira AM, Pivonello R. Acromegaly. Nature Reviews: Disease Primers 2019 20. (https://doi.org/10.1038/s41572-019-0071-6)

    • Search Google Scholar
    • Export Citation
  • 5

    Melmed S, Bronstein MD, Chanson P, Klibanski A, Casanueva FF, Wass JAH, Strasburger CJ, Luger A, Clemmons DR, Giustina A. A Consensus Statement on acromegaly therapeutic outcomes. Nature Reviews: Endocrinology 2018 552561. (https://doi.org/10.1038/s41574-018-0058-5)

    • Search Google Scholar
    • Export Citation
  • 6

    Boguszewski CL, Ayuk J. MANAGEMENT OF ENDOCRINE DISEASE: Acromegaly and cancer: an old debate revisited. European Journal of Endocrinology 2016 R147R156. (https://doi.org/10.1530/EJE-16-0178)

    • Search Google Scholar
    • Export Citation
  • 7

    Dal J, Leisner MZ, Hermansen K, Farkas DK, Bengtsen M, Kistorp C, Nielsen EH, Andersen M, Feldt-Rasmussen U & Dekkers OM et al. Cancer incidence in patients with acromegaly: a cohort study and meta-analysis of the literature. Journal of Clinical Endocrinology and Metabolism 2018 21822188. (https://doi.org/10.1210/jc.2017-02457)

    • Search Google Scholar
    • Export Citation
  • 8

    Swearingen B, Barker FG 2nd, Katznelson L, Biller BM, Grinspoon S, Klibanski A, Moayeri N, Black PM, Zervas NT. Long-term mortality after transsphenoidal surgery and adjunctive therapy for acromegaly. Journal of Clinical Endocrinology and Metabolism 1998 34193426. (https://doi.org/10.1210/jcem.83.10.5222)

    • Search Google Scholar
    • Export Citation
  • 9

    Trainer PJ, Drake WM, Katznelson L, Freda PU, Herman-Bonert V, van der Lely AJ, Dimaraki EV, Stewart PM, Friend KE & Vance ML et al. Treatment of acromegaly with the growth hormone-receptor antagonist pegvisomant. New England Journal of Medicine 2000 11711177. (https://doi.org/10.1056/NEJM200004203421604)

    • Search Google Scholar
    • Export Citation
  • 10

    van der Lely AJ, Hutson RK, Trainer PJ, Besser GM, Barkan AL, Katznelson L, Klibanski A, Herman-Bonert V, Melmed S & Vance ML et al. Long-term treatment of acromegaly with pegvisomant, a growth hormone receptor antagonist. Lancet 2001 17541759. (https://doi.org/10.1016/s0140-6736(01)06844-1)

    • Search Google Scholar
    • Export Citation
  • 11

    van der Lely AJ, Biller BM, Brue T, Buchfelder M, Ghigo E, Gomez R, Hey-Hadavi J, Lundgren F, Rajicic N & Strasburger CJ et al. Long-term safety of pegvisomant in patients with acromegaly: comprehensive review of 1288 subjects in ACROSTUDY. Journal of Clinical Endocrinology and Metabolism 2012 15891597. (https://doi.org/10.1210/jc.2011-2508)

    • Search Google Scholar
    • Export Citation
  • 12

    Buchfelder M, Schlaffer S, Droste M, Mann K, Saller B, Brubach K, Stalla GK, Strasburger CJ & German Pegvisomant Observational Study. The German ACROSTUDY: past and present. European Journal of Endocrinology 2009 (Supplement 1) S3S10. (https://doi.org/10.1530/EJE-09-0350)

    • Search Google Scholar
    • Export Citation
  • 13

    Jimenez C, Burman P, Abs R, Clemmons DR, Drake WM, Hutson KR, Messig M, Thorner MO, Trainer PJ, Gagel RF. Follow-up of pituitary tumor volume in patients with acromegaly treated with pegvisomant in clinical trials. European Journal of Endocrinology 2008 517523. (https://doi.org/10.1530/EJE-08-0205)

    • Search Google Scholar
    • Export Citation
  • 14

    Tritos NA, Biller BM. Pegvisomant: a growth hormone receptor antagonist used in the treatment of acromegaly. Pituitary 2017 129135. (https://doi.org/10.1007/s11102-016-0753-y)

    • Search Google Scholar
    • Export Citation
  • 15

    Trainer PJ. ACROSTUDY: an overview. Hormone Research 2007 (Supplement 5) 6869. (https://doi.org/10.1159/000110480)

  • 16

    Buchfelder M, van der Lely AJ, Biller BMK, Webb SM, Brue T, Strasburger CJ, Ghigo E, Camacho-Hubner C, Pan K & Lavenberg J et al. Long-term treatment with pegvisomant: observations from 2090 acromegaly patients in ACROSTUDY. European Journal of Endocrinology 2018 419427. (https://doi.org/10.1530/EJE-18-0616)

    • Search Google Scholar
    • Export Citation
  • 17

    Riis P. Thirty years of bioethics: the Helsinki Declaration 1964–2003. New Review of Bioethics 2003 1525. (https://doi.org/10.1080/1740028032000131396)

    • Search Google Scholar
    • Export Citation
  • 18

    Global Health Estimates 2016: Disease Burden by Cause, Sex, by country and by Region, 2000–2016. Geneva: World Health Organization, 2018.

    • Search Google Scholar
    • Export Citation
  • 19

    Mercado M, Gonzalez B, Vargas G, Ramirez C, de los Monteros AL, Sosa E, Jervis P, Roldan P, Mendoza V & Lopez-Felix B et al. Successful mortality reduction and control of comorbidities in patients with acromegaly followed at a highly specialized multidisciplinary clinic. Journal of Clinical Endocrinology and Metabolism 2014 44384446. (https://doi.org/10.1210/jc.2014-2670)

    • Search Google Scholar
    • Export Citation
  • 20

    Arosio M, Reimondo G, Malchiodi E, Berchialla P, Borraccino A, De Marinis L, Pivonello R, Grottoli S, Losa M & Cannavo S et al. Predictors of morbidity and mortality in acromegaly: an Italian survey. European Journal of Endocrinology 2012 189198. (https://doi.org/10.1530/EJE-12-0084)

    • Search Google Scholar
    • Export Citation
  • 21

    Ayuk J, Clayton RN, Holder G, Sheppard MC, Stewart PM, Bates AS. Growth hormone and pituitary radiotherapy, but not serum insulin-like growth factor-I concentrations, predict excess mortality in patients with acromegaly. Journal of Clinical Endocrinology and Metabolism 2004 16131617. (https://doi.org/10.1210/jc.2003-031584)

    • Search Google Scholar
    • Export Citation
  • 22

    Ayuk J, Sheppard MC, Clayton RN, Bates AS, Stewart PM. Evidence for the use of igf-I as a predictor of mortality in acromegaly is lacking. Journal of Clinical Endocrinology and Metabolism 2004 58675868; author reply 5868. (https://doi.org/10.1210/jc.2004-1475)

    • Search Google Scholar
    • Export Citation
  • 23

    Orme SM, McNally RJ, Cartwright RA, Belchetz PE. Mortality and cancer incidence in acromegaly: a retrospective cohort study. United Kingdom Acromegaly Study Group. Journal of Clinical Endocrinology and Metabolism 1998 27302734. (https://doi.org/10.1210/jcem.83.8.5007)

    • Search Google Scholar
    • Export Citation
  • 24

    Jayasena CN, Comninos AN, Clarke H, Donaldson M, Meeran K, Dhillo WS. The effects of long-term growth hormone and insulin-like growth factor-1 exposure on the development of cardiovascular, cerebrovascular and metabolic co-morbidities in treated patients with acromegaly. Clinical Endocrinology 2011 220225. (https://doi.org/10.1111/j.1365-2265.2011.04019.x)

    • Search Google Scholar
    • Export Citation
  • 25

    Varadhan L, Reulen RC, Brown M, Clayton RN. The role of cumulative growth hormone exposure in determining mortality and morbidity in acromegaly: a single centre study. Pituitary 2016 251261. (https://doi.org/10.1007/s11102-015-0700-3)

    • Search Google Scholar
    • Export Citation

 

     European Society of Endocrinology

Sept 2018 onwards Past Year Past 30 Days
Abstract Views 1149 75 0
Full Text Views 845 624 63
PDF Downloads 484 326 44
  • View in gallery

    Survival probability in patients with acromegaly, stratified by age at ACROSTUDY entry.

  • View in gallery

    Mortality rates shown in relation to attained age in patients with acromegaly participating in ACROSTUDY (continuous line) and individuals in the general population in ACROSTUDY countries (dashed line).