Relative Risks of Contributing Factors to Morbidity and Mortality in Adults With Craniopharyngioma on Growth Hormone Replacement

Relative Risks of Contributing Factors to Morbidity and Mortality in Adults With... Abstract Context In adults, craniopharyngioma (CP) of either childhood-onset (CO-CP) or adult-onset (AO-CP) is associated with increased morbidity and mortality, but data on the relative risks (RRs) of contributing factors are lacking. Objective To assess the RRs of factors contributing to morbidity and mortality in adults with CO-CP and AO-CP. Methods Data on 1669 patients with CP from KIMS (Pfizer International Metabolic Database) were analyzed using univariate and multiple Poisson and Cox regression methods. Results When CO-CP and AO-CP groups were combined, history of stroke and hyperlipidemia increased cardiovascular risk, higher body mass index (BMI) and radiotherapy increased cerebrovascular risk, and increased waist circumference increased the risk of developing diabetes mellitus (DM). Compared with patients with CO-CP, patients with AO-CP had a threefold higher risk of tumor recurrence, whereas being female and previous radiotherapy exposure conferred lower risks. Radiotherapy and older age with every 10 years from disease onset conferred a 2.3- to 3.5-fold risk for developing new intracranial tumors, whereas older age, greater and/or increasing BMI, history of stroke, and lower insulinlike growth factor I (IGF-I) standard deviation score measured at last sampling before death were related to increased all-cause mortality. Compared with the general population, adults with CP had 9.3-, 8.1-, and 2.2-fold risks of developing DM, new intracranial tumors, and early death, respectively. Conclusion Conventional factors that increase the risks of cardio- and cerebrovascular diseases and DM and risks for developing new intracranial tumors contributed to excess morbidity and mortality. In addition, lower serum IGF-I level measured from the last sample before death was inversely associated with mortality risk in patients with CP. Craniopharyngiomas (CPs) are embryonic malformations that arise within the sellar and suprasellar regions and often infiltrate the pituitary gland and hypothalamus. These tumors may occur at any age but typically demonstrate a bimodal distribution, with peak childhood-onset (CO-CP) and adult-onset (AO-CP) incidence rates occurring between the ages of 5 and 14 years and the ages of 50 and 74 years, respectively (1). Although CPs are histologically benign, these tumors tend to invade locally, frequently resulting in recurrence even after aggressive surgery (2) and/or radiotherapy (3). Previous studies demonstrated that combined populations of pediatric and adult patients with CP but without growth hormone (GH) replacement have morbidity and mortality rates that are four- to ninefold higher than those of the general population (4, 5) and nearly 10-fold higher than those of individuals with other causes of hypopituitarism (6). The excess morbidity is associated with early disease onset, primary treatment regimen, severity of hypopituitarism, visual disturbances, hypothalamic damage, and initial symptoms of intracranial hypertension (7–11), whereas the increased mortality has been attributed mainly to cardio- and cerebrovascular diseases (4–6, 12). In patients who had received more contemporary hormone replacement therapy regimens, including GH in those with GH deficiency (GHD) and lower glucocorticoid replacement doses, the overall mortality rate was still at least fourfold higher than expected, whereas in patients treated for nonfunctioning pituitary adenomas and prolactinomas, mortality rates were comparable to those of the reference population (13). Accordingly, several other studies (14–16), including a recent meta-analysis by Pappachan et al. (17), reported beneficial effects favoring GH replacement in adults with hypopituitarism, including substantial reductions in mortality. In the current study, the main objective was to evaluate the relative risks (RRs) of contributing factors to excess morbidity and mortality in a large cohort of patients with CP receiving long-term GH replacement. Another objective was to examine the overall and cause-specific mortality rates and comorbidities, specifically cardio- and cerebrovascular diseases, diabetes mellitus (DM), and second brain tumors, of these patients. Patients and Methods Patients Data on 1669 adults with CO-CP (CP diagnosed at <18 years of age; n = 744) and AO-CP (CP diagnosed at ≥18 years of age; n = 925) were extracted from KIMS (Pfizer International Metabolic Database), the largest pharmaco-epidemiological surveillance study on GH replacement in adults with GHD. The 16,484 adult patients in KIMS were enrolled between 1994 and 2012 from 28 countries, and data were collected on case report forms by treating physicians at routine clinic visits. All patients included in the current study had confirmed diagnosis of severe GHD by GH stimulation testing or had at least three pituitary hormone deficiencies and a low serum insulinlike growth factor I (IGF-I) standard deviation score (SDS) < −2, in accordance with current clinical practice guidelines (18, 19). The median time from KIMS entry to the last follow-up clinic visit was 5.3 years (range, 0 to 18.3 years), representing a total of 10,284 patient-years. Data collection into KIMS was approved by the institutional review boards/ethical committees as required by local regulations in each participating center. Written informed consent was obtained from all patients before any data were entered into KIMS. Methods All data extracted from the KIMS database were provided from all the participating centers. Follow-up data were analyzed and consisted of average GH dose used, the development of DM, and the reporting of adverse events (AEs) by the participating investigators. DM was defined by physician documentation or use of antidiabetes medication(s) at baseline and/or biochemical data in accordance with the American Diabetes Association guidelines (20), as previously described (21). This study was divided into four parts. In the first part, a cross-sectional analysis was performed by comparing background clinical characteristics of patients with CO-CP and patients with AO-CP to determine whether demographics, phenotypic features, and biochemical parameters differed between the two groups. In the second part, differences in rates of serious AEs (SAEs) between patients with CO-CP and AO-CP were analyzed. In the third part, Cox regression analyses with internal reference were performed with adjustment for age and sex to assess the RRs of the following outcomes: cardio- and cerebrovascular diseases, DM, recurrence of CP, development of new intracranial and extracranial tumors, and all-cause mortality. These outcomes were assessed in relation to covariates hypothesized to systematically influence their rates of occurrence, such as medical history (whether CO-CP or AO-CP, number and type of surgeries, use of radiotherapy, history of stroke, smoking, and naivety to GH replacement), clinical presentation at baseline [age, sex, body mass index (BMI), hypertension, waist circumference, hyperlipidemia, diabetes insipidus, prediabetes defined as hemoglobin A1c value between 5.7% and 6.4%, DM, and metabolic syndrome at baseline based on the International Diabetes Federation definition] (22), IGF-I SDS, and follow-up data during GH replacement (time from baseline, age at KIMS baseline or attained age during follow-up, GH dose, and IGF-I SDS during GH replacement). In the fourth part, rates were compared with those in the general population for available end points (i.e., incidence rates of DM, incidence rates of intracranial tumors and extracranial tumors, respectively, and mortality from all causes, and from cardio- and cerebrovascular causes). The dose-titration regimens for GH replacement against serum IGF-I levels were prescribed by treating physicians on the basis of local practice. Serum IGF-I and lipid levels were measured centrally, as described previously (23), and plasma glucose and hemoglobin A1c levels were analyzed locally at each participating center. To make meaningful comparisons between IGF-I measurements and age- and sex-specific reference ranges, IGF-I SDS was calculated according to the following formulae: between 1994 and 1997, IGF-I SDS = [ln (IGF-I) − (5.95 − 0.0197 × age)]; between 1997 and 2002, IGF-I SDS = [ln (IGF-1) − (15.92 − 0.0146 × age)/0.272]; and from 2002 onward, based on data by Brabant et al. (24). Statistical methods For descriptive statistics, means, standard deviations, medians with fifth and 95th percentiles, and proportions (percentages) were calculated depending on the type of variable. For comparisons between patients with CO-CP and AO-CP, heterogeneity tests for nominal or categorical variables were performed with χ2 tests, whereas for numerical variables, Student t tests were performed. A crude presentation was performed to assess the overall incidence rates of reported SAEs presented according to the Medical Dictionary for Regulatory Activities System Organ Class (SOC) level comparing patients with CO-CP and those with AO-CP. The calculation was based on the first SAE in a patient on preferred term level but was presented on Medical Dictionary for Regulatory Activities SOC level. Crude comparisons with incidence rate ratios between patients with CO-CP and those with AO-CP were performed with score tests (25), whereas 95% confidence intervals (CIs) were calculated with the Byar equation (26). Significant rate differences on the crude level were further analyzed with adjustment for age and sex by proportional hazard regression. For the seven specific end points, regression analyses were performed with proportional hazards multiple regression. Time-to-event was defined as years from KIMS baseline to the date of first AE report or, if no event, to last visit date or date of death. Unless defined differently, the end points with hypothesized risk factors measured at baseline were the following: sex (for all risk factors), age at baseline (all), attained age (all), obesity by waist circumference or by BMI as time-dependent (TD) covariates (a, b, c, f, g), DM (a, b, g), and specifically: For occurrence of cardiovascular disease: hyperlipidemia (yes vs. no), total cholesterol, hypertension (yes vs. no), systolic and diastolic blood pressure (TD), metabolic syndrome, cardio/cerebrovascular events (TD) in medical history, smoking (yes vs. no), IGF-I SDS (TD), GH dose (TD), and hydrocortisone equivalent dose For occurrence of cerebrovascular disease: hyperlipidemia (yes vs. no), total cholesterol, hypertension (yes vs. no), systolic and diastolic blood pressure (TD), cardio/cerebrovascular events (TD) in medical history, radiotherapy (yes vs. no), diabetes insipidus (yes vs. no), smoking (yes vs. no), IGF-I SDS (TD), and GH dose (TD) For occurrence of new DM: hemoglobin A1c, fasting blood glucose, follow-up year after baseline, hydrocortisone equivalent dose, and TD sex and age effects For recurrence of CP: surgery (yes vs. no), number and type of surgeries, radiotherapy (yes vs. no), and IGF-I SDS (TD) For occurrence of new intracranial malignant tumors: surgery (yes vs. no), number and type of surgeries, radiotherapy (yes vs. no), previous cancer in medical history, smoking (yes vs. no), and IGF-I SDS (TD) For occurrence of new extracranial malignant tumors: radiotherapy (yes vs. no), previous cancer in medical history, smoking (yes vs. no), and IGF-I SDS (TD) For all-cause mortality: myocardial infarction and cerebrovascular disease in medical history (yes vs. no), diabetes insipidus (yes vs. no), adrenocorticotropic hormone deficiency (yes vs. no), radiotherapy, blood pressure, hypertension, smoking (yes vs. no), and IGF-I SDS (TD) Standardized incidence ratios and standardized mortality ratios were calculated using stratification for attained age, and/or sex, and/or country/region, depending on the level of stratification in the specific external reference. These ratios compare the observed number of cases in the patient group (O) and the expected number of cases (E). The latter number (E) quantifies the expected cases in the patient-group when the patient-group had the same specific rates as the chosen external reference population. Patient-years were calculated from the date of KIMS entry or GH start date (if later than the KIMS entry date) to the date of the studied event or, if no event, the date of the last visit or date of death. General population (reference) rates for cancer incidence were from the International Agency for Research on Cancer, with cancer incidence from five continents (27) stratified for age, sex, and country. Reference rates for DM were age- and sex-specific from a study on incidence of DM in southern Sweden (28). The age-, sex-, and region-specific reference rates for mortality were taken from the World Health Organization project Global Burden of Disease (29). The 95% CIs were calculated with the Byar approximation formula. In the regression models, CIs and P values were based on likelihood methods (26). P values <0.05 were considered statistically significant. Higher risk estimates in combination with a P value <0.10 were also considered. Results Baseline clinical characteristics, medical history, biochemical parameters, and hypothesized risk factors There were no differences in sex ratio between the patients with CO-CP and those with AO-CP. Compared with patients with AO-CP, patients with CO-CP more often underwent combined surgery and radiotherapy; had less than three pituitary hormone deficiencies; more often had thyroid-stimulating hormone, gonadotropin, and arginine vasopressin deficiencies; underwent longer duration of KIMS follow-up; had lower BMI, waist circumference, systolic and diastolic blood pressure values, fasting glucose level, and hemoglobin A1c value; had more favorable lipid profiles; had lower IGF-I SDS at baseline; and had lower IGF-I SDS at 1 year despite being treated with higher GH doses at 1 year (Table 1). Table 1. Background, GH Replacement Chronology, Clinical Characteristics, and Biochemical Characteristics of Adults With CO-CP and AO-CP   CO-CP  CO-CP vs. AO-CP, Crude P Valuea  AO-CP  Combined CO-CP + AO-CP  Sex, males/females, %  53.1/46.9  NS  53.0/47.0  53.0/47.0  Age at diagnosis of CP, y  11.1 (4.5)  <0.0001  36.7 (13.3)  25.3 (16.4)  Treatment of CP, %           Surgery (transsphenoidal/transcranial) only  65.5  NS  72.4  69.3   Surgery (transsphenoidal/transcranial) and radiotherapy  22.9  0.02  18.0  20.1   Radiotherapy only  4.4  NS  2.8  3.5  Pituitary hormone deficiencies excluding GHD, %           No hormonal deficit  1.2  NS  1.5  1.4   1 Hormonal deficit  3.1  NS  4.4  3.8   2 Hormonal deficits  5.1  NS  9.4  7.5   3 Hormonal deficits  24.6  <0.0001  30.0  27.6   4 Hormonal deficits  66.0  NS  54.6  59.7  Types of pituitary hormone deficiencies, %           ACTH  90.4  NS  89.1  89.7   TSH  95.4  0.004  91.9  93.4   LH/FSH  94.9  <0.0001  89.1  91.6   AVP  71.9  <0.0001  62.8  66.8  GH replacement chronology, y           Age at GH start in childhood  11.8 (4.2)  —  —  —   Age at GH stop in childhood  17.9 (2.8)  —  —  —   Age at GH start in adulthood  26.9 (9.7)  <0.0001  44.2 (13.1)  36.5 (14.5)   Age at KIMS entry  27.7 (9.7)  <0.01  44.7 (13.2)  37.1 (14.4)   Duration of KIMS follow-up since GH start  6.4 (4.8)  0.03  5.9 (4.6)  6.2 (4.8)  Clinical characteristics           BMI, kg/m2  29.5 (7.6)  0.026  30.3 (6.5)  29.9 (7.0)   Waist circumference, cm  96.8 (16.7)  0.0004  100.1 (15.1)  98.6 (15.9)   Systolic blood pressure, mm Hg  118.0 (14.6)  <0.0001  125.0 (16.4)  121.9 (16.0)   Diastolic blood pressure, mm Hg  74.5 (9.8)  <0.0001  78.5 (10.5)  76.8 (10.4)  Biochemical characteristics           Fasting glucose, mmol/L  4.7 (1.0)  0.043  4.9 (0.9)  4.8 (1.0)   Hemoglobin A1c, %  5.2 (2.0)  0.002  5.6 (3.1)  5.4 (2.7)   Total cholesterol, mmol/L  5.5 (1.3)  0.0002  5.9 (1.2)  5.7 (1.3)   HDL cholesterol, mmol/L  1.3 (0.4)  0.04  1.2 (0.4)  1.2 (0.4)   LDL cholesterol mmol/L  3.4 (1.1)  0.007  3.6 (1.1)  3.5 (1.1)   Triglycerides, mmol/L  2.2 (1.8)  0.001  2.6 (1.9)  2.4 (1.9)   IGF-I SDS at baseline  −2.7 (2.6)  <0.0001  −1.1 (2.0)  −1.8 (2.4)   GH dose, mg/d, at 1 y  0.6 (0.4)  <0.0001  0.4 (0.3)  0.5 (0.3)   IGF-I SDS at 1 y  −0.9 (2.0)  <0.0001  0.4 (1.7)  −0.2 (1.9)    CO-CP  CO-CP vs. AO-CP, Crude P Valuea  AO-CP  Combined CO-CP + AO-CP  Sex, males/females, %  53.1/46.9  NS  53.0/47.0  53.0/47.0  Age at diagnosis of CP, y  11.1 (4.5)  <0.0001  36.7 (13.3)  25.3 (16.4)  Treatment of CP, %           Surgery (transsphenoidal/transcranial) only  65.5  NS  72.4  69.3   Surgery (transsphenoidal/transcranial) and radiotherapy  22.9  0.02  18.0  20.1   Radiotherapy only  4.4  NS  2.8  3.5  Pituitary hormone deficiencies excluding GHD, %           No hormonal deficit  1.2  NS  1.5  1.4   1 Hormonal deficit  3.1  NS  4.4  3.8   2 Hormonal deficits  5.1  NS  9.4  7.5   3 Hormonal deficits  24.6  <0.0001  30.0  27.6   4 Hormonal deficits  66.0  NS  54.6  59.7  Types of pituitary hormone deficiencies, %           ACTH  90.4  NS  89.1  89.7   TSH  95.4  0.004  91.9  93.4   LH/FSH  94.9  <0.0001  89.1  91.6   AVP  71.9  <0.0001  62.8  66.8  GH replacement chronology, y           Age at GH start in childhood  11.8 (4.2)  —  —  —   Age at GH stop in childhood  17.9 (2.8)  —  —  —   Age at GH start in adulthood  26.9 (9.7)  <0.0001  44.2 (13.1)  36.5 (14.5)   Age at KIMS entry  27.7 (9.7)  <0.01  44.7 (13.2)  37.1 (14.4)   Duration of KIMS follow-up since GH start  6.4 (4.8)  0.03  5.9 (4.6)  6.2 (4.8)  Clinical characteristics           BMI, kg/m2  29.5 (7.6)  0.026  30.3 (6.5)  29.9 (7.0)   Waist circumference, cm  96.8 (16.7)  0.0004  100.1 (15.1)  98.6 (15.9)   Systolic blood pressure, mm Hg  118.0 (14.6)  <0.0001  125.0 (16.4)  121.9 (16.0)   Diastolic blood pressure, mm Hg  74.5 (9.8)  <0.0001  78.5 (10.5)  76.8 (10.4)  Biochemical characteristics           Fasting glucose, mmol/L  4.7 (1.0)  0.043  4.9 (0.9)  4.8 (1.0)   Hemoglobin A1c, %  5.2 (2.0)  0.002  5.6 (3.1)  5.4 (2.7)   Total cholesterol, mmol/L  5.5 (1.3)  0.0002  5.9 (1.2)  5.7 (1.3)   HDL cholesterol, mmol/L  1.3 (0.4)  0.04  1.2 (0.4)  1.2 (0.4)   LDL cholesterol mmol/L  3.4 (1.1)  0.007  3.6 (1.1)  3.5 (1.1)   Triglycerides, mmol/L  2.2 (1.8)  0.001  2.6 (1.9)  2.4 (1.9)   IGF-I SDS at baseline  −2.7 (2.6)  <0.0001  −1.1 (2.0)  −1.8 (2.4)   GH dose, mg/d, at 1 y  0.6 (0.4)  <0.0001  0.4 (0.3)  0.5 (0.3)   IGF-I SDS at 1 y  −0.9 (2.0)  <0.0001  0.4 (1.7)  −0.2 (1.9)  Data are presented as mean (standard deviation) unless otherwise stated. Significant P values are shown in bold. Abbreviations: ACTH, adrenocorticotropic hormone; AVP, arginine vasopressin; FSH, follicle-stimulating hormone; HDL, high-density lipoprotein; LDL, low-density lipoprotein; LH, luteinizing hormone; NS, not significant; TSH, thyroid-stimulating hormone. a P values for crude comparison between CO-CP vs. AO-CP groups. Heterogeneity tests for nominal or categorical variables were performed with χ2. Unpaired Student t tests were performed for numerical variables. After adjustment for age at baseline and sex, diastolic blood pressure, HDL cholesterol, triglycerides, IGF-I SDS, and GH dose remained significantly different between onset groups. View Large At baseline before KIMS entry, more patients with AO-CP were smokers and more often had DM, cardiovascular and cerebrovascular disease, hyperlipidemia, and cancer than those with CO-CP. By contrast, more patients with CO-CP underwent transcranial surgery and combined radiotherapy as well as transcranial and transsphenoidal surgeries, whereas more patients with AO-CP underwent transsphenoidal surgeries (Table 2). Table 2. Risk Factors of Adults With CO-CP and AO-CP at Baseline   CO-CP  AO-CP  Combined CO-CP + AO-CP  Smokers  89 (12.0)  206 (22.3)  295 (17.7)  DM  45 (6.0)  100 (10.8)  145 (8.7)  Metabolic syndrome (per IDF definition)a  102 (35.2)  236 (58.3)  338 (48.6)  Cardiovascular disease in medical history  20 (2.7)  58 (6.3)  78 (4.7)  Cerebrovascular disease in medical history  10 (1.4)  20 (2.2)  30 (1.8)  Hyperlipidemia  51 (6.8)  121 (13.1)  172 (10.3)  Cancer diagnosis before KIMS  4 (0.5)  8 (0.9)  12 (0.7)  Number of surgeries before KIMSb         0  21 (3.0)  20 (2.3)  41 (2.6)   1  507 (72.0)  677 (77.4)  1184 (75.0)   ≥2  176 (25.0)  178 (20.3)  354 (22.4)  Type of primary treatmentb         Transcranial surgery  376 (50.5)  419 (45.3)  795 (47.6)   Transsphenoidal surgery  104 (14.0)  235 (25.4)  339 (20.3)   Radiotherapy ± transcranial ± transsphenoidal surgeries  210 (28.2)  208 (22.5)  418 (25.0)    CO-CP  AO-CP  Combined CO-CP + AO-CP  Smokers  89 (12.0)  206 (22.3)  295 (17.7)  DM  45 (6.0)  100 (10.8)  145 (8.7)  Metabolic syndrome (per IDF definition)a  102 (35.2)  236 (58.3)  338 (48.6)  Cardiovascular disease in medical history  20 (2.7)  58 (6.3)  78 (4.7)  Cerebrovascular disease in medical history  10 (1.4)  20 (2.2)  30 (1.8)  Hyperlipidemia  51 (6.8)  121 (13.1)  172 (10.3)  Cancer diagnosis before KIMS  4 (0.5)  8 (0.9)  12 (0.7)  Number of surgeries before KIMSb         0  21 (3.0)  20 (2.3)  41 (2.6)   1  507 (72.0)  677 (77.4)  1184 (75.0)   ≥2  176 (25.0)  178 (20.3)  354 (22.4)  Type of primary treatmentb         Transcranial surgery  376 (50.5)  419 (45.3)  795 (47.6)   Transsphenoidal surgery  104 (14.0)  235 (25.4)  339 (20.3)   Radiotherapy ± transcranial ± transsphenoidal surgeries  210 (28.2)  208 (22.5)  418 (25.0)  Data are presented as number of patients (%). Abbreviation: IDF, International Diabetes Federation. a Among patients with nonmissing values (CO-CP: 290; AO-CP: 405). b Among patients with nonmissing values (CO-CP: 704; AO-CP: 875). View Large Factors affecting morbidity and mortality rates Older age at baseline, history of stroke, and hyperlipidemia were associated with increased risk of cardiovascular disease, whereas higher BMI (P = 0.055) showed a tendency toward increasing that risk. Older age at baseline, higher BMI, and a history of stroke (P = 0.10) increased the risk of cerebrovascular disease, whereas previous radiotherapy exposure (P = 0.058) tended to increase that risk. The risk of developing new DM increased with older age at baseline and with higher waist circumference and baseline hemoglobin A1c level. Female sex conferred a 2.1-fold increased risk at baseline following KIMS entry; thereafter, this risk became TD, with a decreasing rate of 44% per year of follow-up compared with males. This indicated that after 1.5 to 2.5 years of follow-up, rates were similar between sexes, and after 5 to 6 years of follow-up, males showed higher rates. Risk of recurrence of CP was lower in females and in those with previous surgery (vs. no surgery) and previous exposure to radiotherapy and was higher in adult-onset disease. Radiotherapy was associated with a 3.5-fold risk (95% CI: 0.9 to 13; P = 0.067) for developing new intracranial tumors (five de novo malignant brain tumors consisting of high-grade glioblastoma, glioblastoma multiforme, and anaplastic oligoastrocytoma; four meningiomas; and three unspecified primary brain tumors), whereas older attained age with every 10 years from disease onset conferred a 2.3-fold risk for developing new extracranial tumors (Fig. 1). The risk of extracranial tumors was similar compared with that of the reference population (standardized incidence ratio = 0.93; Fig. 2). Greater and/or increasing BMI, history of stroke, and older age with every 10 years from disease onset were associated with a 1.5- to 3.5-fold risk, respectively, of increased mortality. In addition, IGF-I SDS measured at the last sampling before death or at the corresponding time from baseline for patients who were still alive was inversely associated with all-cause mortality (27% lower rate per unit higher IGF-1 SDS; 95% CI: −13% to −38%; P = 0.0003) (Fig. 1). Other variables, as listed in the “Statistical methods” section, were not statistically significant in explaining the variation in risk(s). Figure 1. View largeDownload slide RRs (hazard ratios) of individual contributing factors (cardio- and cerebrovascular diseases and DM as well as CP recurrence, new intracranial and extracranial tumors, and all-cause mortality) to morbidity and mortality in adults with CO-CP and AO-CP. Patients with CO-CP and AO-CP had similar risks for all end points except recurrence (data not shown). Significant P values are shown in bold. *BMI TD variable; unit = kg/m2. **Females had hazard ratio = 2.13 vs. males at KIMS baseline, but after 1 to 2 years, the hazard ratio was lower in females (44% decrease per year). ***Surgery per se decreases the risk, but more surgeries indicate increasing risk. HbA1c, hemoglobin A1c; NV, numerical variable (no specific reference range); PY, per year. Figure 1. View largeDownload slide RRs (hazard ratios) of individual contributing factors (cardio- and cerebrovascular diseases and DM as well as CP recurrence, new intracranial and extracranial tumors, and all-cause mortality) to morbidity and mortality in adults with CO-CP and AO-CP. Patients with CO-CP and AO-CP had similar risks for all end points except recurrence (data not shown). Significant P values are shown in bold. *BMI TD variable; unit = kg/m2. **Females had hazard ratio = 2.13 vs. males at KIMS baseline, but after 1 to 2 years, the hazard ratio was lower in females (44% decrease per year). ***Surgery per se decreases the risk, but more surgeries indicate increasing risk. HbA1c, hemoglobin A1c; NV, numerical variable (no specific reference range); PY, per year. Figure 2. View largeDownload slide Observed/expected (O/E) ratios with 95% CIs for new intracranial malignant tumors, new extracranial malignant tumors, new DM, and mortality in adults with CO-CP and AO-CP. Significant P values are shown in bold. External incidence rates of intracranial (ICDO: C70-C72) and extracranial (ICDO: C00-C69, C73-C97) malignant tumors are from the International Agency for Research on Cancer (27). External incidence rates for DM are from Kronoberg County of southern Sweden (28). External mortality rates are from the World Health Organization project Global Burden of Disease (29). ICDO, International Classification of Diseases for Oncology. Figure 2. View largeDownload slide Observed/expected (O/E) ratios with 95% CIs for new intracranial malignant tumors, new extracranial malignant tumors, new DM, and mortality in adults with CO-CP and AO-CP. Significant P values are shown in bold. External incidence rates of intracranial (ICDO: C70-C72) and extracranial (ICDO: C00-C69, C73-C97) malignant tumors are from the International Agency for Research on Cancer (27). External incidence rates for DM are from Kronoberg County of southern Sweden (28). External mortality rates are from the World Health Organization project Global Burden of Disease (29). ICDO, International Classification of Diseases for Oncology. Both patients with CO-CP and those with AO-CP showed increased and similar O/E ratios for developing DM (~9.3-fold) and new intracranial malignant tumors (~8.1-fold) and for all-cause mortality (∼2.2-fold in the observed time-window of follow-up, stratified by attained age, sex, and region) (Fig. 2). For cardio- and cerebrovascular mortalities, patients with CO-CP and those with AO-CP showed increased O/E ratios of 2.0-fold (95% CI: 1.20 to 3.20; 18 cases) and 3.2-fold (95% CI: 1.30 to 6.30; 8 cases), respectively, when compared with the general population. Incidence rates of reported SAEs There were no significant differences in crude rates of SAEs, except for the SOCs “surgical and medical procedures,” “neoplasms benign, malignant, unspecified (including cysts and polyps),” and “infections and infestations,”’ respectively (Table 3). After adjustment for age and sex, the incidence rate ratios were similar [0.40 (95% CI: 0.27 to 0.58) and 0.51 (95% CI: 0.32 to 0.80)] for the first two SOCs above and still significant (P < 0.005) but changed for infections and infestations, from 1.43 (95% CI: 1.02 to 2.00) to 1.16 (95% CI: 0.75 to 1.80). Table 3. Number of SAEsa and Crude Incidence Rate Differences per 1000 Patient-Years and Incidence Rate Ratios for Patients With CO-CP vs AO-CP SOC  CO-CP Crude Rate per 
1000 PY (n)  AO-CP Crude Rate per 
1000 PY (n)  Crude Rate Difference per 
1000 PY  Crude IRR (Approximate 95% CI)  P Value
 (Score-Test)  Blood and lymphatic system disorders  0.0 (0)  0.4 (2)  −0.37  0  0.190  Cardiac disorders  2.6 (12)  3.2 (17)  −0.56  0.82 (0.39–1.72)  0.602  Congenital, familial, and genetic disorders  0.2 (1)  0.2 (1)  0.03  1.16 (0.07–18.62)  0.914  Ear and labyrinth disorders  1.1 (5)  0.6 (3)  0.52  1.94 (0.46–8.12)  0.355  Endocrine disorders  2.4 (11)  3.2 (17)  −0.78  0.75 (0.35–1.61)  0.463  Eye disorders  0.9 (4)  1.5 (8)  −0.62  0.58 (0.18–1.93)  0.371  Gastrointestinal disorders  2.8 (13)  5.2 (28)  −2.39  0.54 (0.28–1.04)  0.063  General disorders and administration site conditions  6.1 (28)  6.5 (35)  −0.45  0.93 (0.57–1.53)  0.779  Hepatobiliary disorders  1.9 (9)  2.6 (14)  −0.65  0.75 (0.32–1.73)  0.496  Infections and infestationsb  16.5 (76)  11.5 (62)  4.93  1.43 (1.02–2.00)  0.037  Injury, poisoning, and procedural complications  4.5 (21)  5.2 (28)  −0.66  0.87 (0.50–1.54)  0.639  Investigations  1.7 (8)  2.4 (13)  −0.69  0.72 (0.30–1.73)  0.456  Metabolic and nutritional disorders  4.3 (20)  4.1 (22)  0.24  1.06 (0.58–1.94)  0.854  Musculoskeletal and connective tissue disorders  3.2 (15)  3.7 (20)  −0.47  0.87 (0.45–1.71)  0.691  Neoplasms benign, malignant, and unspecified (including cysts and polyps)c  7.8 (36)  19.7 (106)  −11.92  0.40 (0.27–0.58)  0.001  Nervous system disorders  12.1 (56)  9.5 (51)  2.64  1.28 (0.87–1.87)  0.203  Pregnancy, puerperal and perinatal conditions  0.9 (4)  0.6 (3)  0.31  1.55 (0.35–6.94)  0.562  Psychiatric disorders  2.2 (10)  2.6 (14)  −0.44  0.83 (0.37–1.87)  0.656  Renal and urinary disorders  1.5 (7)  1.9 (10)  0.15  0.82 (0.31–2.14)  0.678  Reproductive system and breast disorders  1.1 (5)  0.9 (5)  −1.62  1.16 (0.34–4.02)  0.810  Respiratory, thoracic, and mediastinal disorders  1.7 (8)  3.3 (18)  0.31  0.52 (0.23–1.19)  0.114  Skin and subcutaneous tissue disorders  0.9 (4)  0.6 (3)  0.03  1.55 (0.35–6.94)  0.562  Social circumstances  0.2 (1)  0.2 (1)  −5.87  1.16 (0.07–18.62)  0.914  Surgical and medical procedures  4.5 (21)  10.4 (56)  0.06  0.44 (0.26–0.72)  0.001  Vascular disorders  1.7 (17)  1.7 (9)  0.15  1.03 (0.40–2.68)  0.944  Total  (383)  (546)        SOC  CO-CP Crude Rate per 
1000 PY (n)  AO-CP Crude Rate per 
1000 PY (n)  Crude Rate Difference per 
1000 PY  Crude IRR (Approximate 95% CI)  P Value
 (Score-Test)  Blood and lymphatic system disorders  0.0 (0)  0.4 (2)  −0.37  0  0.190  Cardiac disorders  2.6 (12)  3.2 (17)  −0.56  0.82 (0.39–1.72)  0.602  Congenital, familial, and genetic disorders  0.2 (1)  0.2 (1)  0.03  1.16 (0.07–18.62)  0.914  Ear and labyrinth disorders  1.1 (5)  0.6 (3)  0.52  1.94 (0.46–8.12)  0.355  Endocrine disorders  2.4 (11)  3.2 (17)  −0.78  0.75 (0.35–1.61)  0.463  Eye disorders  0.9 (4)  1.5 (8)  −0.62  0.58 (0.18–1.93)  0.371  Gastrointestinal disorders  2.8 (13)  5.2 (28)  −2.39  0.54 (0.28–1.04)  0.063  General disorders and administration site conditions  6.1 (28)  6.5 (35)  −0.45  0.93 (0.57–1.53)  0.779  Hepatobiliary disorders  1.9 (9)  2.6 (14)  −0.65  0.75 (0.32–1.73)  0.496  Infections and infestationsb  16.5 (76)  11.5 (62)  4.93  1.43 (1.02–2.00)  0.037  Injury, poisoning, and procedural complications  4.5 (21)  5.2 (28)  −0.66  0.87 (0.50–1.54)  0.639  Investigations  1.7 (8)  2.4 (13)  −0.69  0.72 (0.30–1.73)  0.456  Metabolic and nutritional disorders  4.3 (20)  4.1 (22)  0.24  1.06 (0.58–1.94)  0.854  Musculoskeletal and connective tissue disorders  3.2 (15)  3.7 (20)  −0.47  0.87 (0.45–1.71)  0.691  Neoplasms benign, malignant, and unspecified (including cysts and polyps)c  7.8 (36)  19.7 (106)  −11.92  0.40 (0.27–0.58)  0.001  Nervous system disorders  12.1 (56)  9.5 (51)  2.64  1.28 (0.87–1.87)  0.203  Pregnancy, puerperal and perinatal conditions  0.9 (4)  0.6 (3)  0.31  1.55 (0.35–6.94)  0.562  Psychiatric disorders  2.2 (10)  2.6 (14)  −0.44  0.83 (0.37–1.87)  0.656  Renal and urinary disorders  1.5 (7)  1.9 (10)  0.15  0.82 (0.31–2.14)  0.678  Reproductive system and breast disorders  1.1 (5)  0.9 (5)  −1.62  1.16 (0.34–4.02)  0.810  Respiratory, thoracic, and mediastinal disorders  1.7 (8)  3.3 (18)  0.31  0.52 (0.23–1.19)  0.114  Skin and subcutaneous tissue disorders  0.9 (4)  0.6 (3)  0.03  1.55 (0.35–6.94)  0.562  Social circumstances  0.2 (1)  0.2 (1)  −5.87  1.16 (0.07–18.62)  0.914  Surgical and medical procedures  4.5 (21)  10.4 (56)  0.06  0.44 (0.26–0.72)  0.001  Vascular disorders  1.7 (17)  1.7 (9)  0.15  1.03 (0.40–2.68)  0.944  Total  (383)  (546)        System Organ Class data are derived from the Medical Dictionary for Regulatory Activities. Significant P values are shown in bold. Abbreviations: IRR, incidence rate ratio; PY, patient-year. a First reported SAE in a patient on preferred term level (Medical Dictionary for Regulatory Activities preferred term) was counted. b With adjustment for age and sex (and to first event per patient on preferred term level), the IRR was 1.16 (95% CI: 0.75–1.80; P = 0.51). c With adjustment for age and sex (and to first event per patient on preferred term level), the IRR was 0.51 (95% CI: 0.32–0.80; P = 0.004). View Large Discussion CPs have consistently been associated with higher morbidity and mortality than have other pituitary disorders, which cannot be attributed solely to endocrine abnormalities (7). Despite advances in surgical technique and adjuvant radiotherapy (3), adults with CP continue to have excess morbidity and mortality, for which the RRs of contributing factors remain unclear. In this study, we assessed the RRs of a range of hypothesized factors that may contribute to increased morbidity and mortality of patients with CP by examining data from a large cohort of such patients enrolled in KIMS. Our results demonstrated that conventional factors that increase the risks of cardio- and cerebrovascular diseases and DM in the general population (e.g., history of stroke, hyperlipidemia, increasing BMI, and higher baseline hemoglobin A1c levels) conferred greater risks in patients with CP. In addition, the increased risk of recurrence of CP was associated with a higher number of surgeries and adult-onset of the disease, whereas previous radiotherapy and increasing age for every 10 years from disease onset conferred increased risks for developing new intracranial and extracranial tumors, respectively. Furthermore, CP and/or its treatment regimen increased the risks of developing DM and intracranial tumors and early death, whereas lower serum IGF-I levels measured at last sampling before death or at the corresponding time from baseline in patients who were still alive was associated with increased all-cause mortality. When adults with CO-CP were compared with those with AO-CP at KIMS entry, there were more CO-CP patients who underwent both surgery (and more transcranial surgeries) and radiotherapy, implying that larger and more aggressive CP tumors tended to present during childhood. These patients also had lower BMIs and waist circumference measurements and better blood pressure levels, hemoglobin A1c levels, and lipid profiles, suggesting more favorable body composition and cardiometabolic status. In line with previous findings (30), patients with CO-CP had lower baseline IGF-I SDS than patients with AO-CP, implying that patients with CO-CP may be more GH deficient. In addition, patients with CO-CP had lower IGF-I SDS values at 1 year despite being treated with higher GH doses. The reasons for this observation are unclear and may include a lack of adherence to daily GH injections by patients with CO-CP or the inability of the younger cohort with CO-CP to achieve optimal GH replacement dose after 1 year. By contrast, the less favorable cardiometabolic status in patients with AO-CP may be exacerbated by the fact that there were more smokers and patients who had a history of DM, metabolic syndrome, cardio- and cerebrovascular disease, hyperlipidemia, and cancer. These findings are also in keeping with those reported by Kendall-Taylor et al. (30) in a smaller cohort of patients with CO-CP and AO-CP who were also from the KIMS database and with findings of two older studies comparing adult GHD of mixed etiologies (31, 32). In this study, history of stroke and hyperlipidemia conferred approximately sixfold and threefold higher risks of cardiovascular disease, respectively, whereas increasing BMI increased the risk of cerebrovascular disease by approximately twofold. The infrequent prescription of antihypertensive and statin therapy offers a plausible explanation and may be reflected by the relatively young age of the CP population, particularly patients with CO-CP. Thus, our data suggest that history of stroke, hyperlipidemia, and higher and/or increasing BMI in adults with either CO-CP or AO-CP are important contributors to increased risk of cardio- and cerebrovascular diseases. In this study, incidence rates for DM in Sweden were used as a reference because the majority of patients in KIMS were from Europe and these rates have been reported to be comparable to those of other European populations (33, 34). Similar to patients without hypothalamic-pituitary diseases, adults with CO-CP and AO-CP with higher baseline hemoglobin A1c levels and greater and/or increasing waist circumference had an increased risk of developing new DM. Interestingly, females had higher rates of developing DM at the start of KIMS entry; however, this risk decreased over time: after ~1.5 to 2.5 years, to rates seen in males, and after 5 to 6 years, to lower rates than males. These data are consistent with those previously reported from KIMS that showed adults with GHD from a variety of etiologies apart from CP demonstrated a greater incidence of DM, with a female preponderance that was thought to be due to abnormal body composition (33) but an overall incidence of DM after GH replacement that was similar between sexes (21). The reason for the female propensity to develop DM before GH is initiated is unclear and may be attributable to a different expression of GHD with higher visceral adiposity and lower physical activity, in keeping with data previously reported in GH-treated patients with mixed etiologies of hypopituitarism (35). It is well known that GH has both mitogenic and antiapoptotic effects (36) and that these actions have led to concerns that long-term GH replacement may increase recurrence rates of CP, although previous studies have indicated that GH replacement is safe (37, 38). Our findings provide further evidence that long-term GH replacement in adults with CP neither facilitated tumor recurrence nor induced any SAEs. Our study also found that the risk of recurrence of CP was decreased in females and in those with previous radiotherapy exposure and prior surgery but increased in the setting of a greater number of surgeries and the onset of CP in adulthood. Knowledge of these factors is important for the physician when treatment regimens are being formulated for these patients. In their recent study involving a large cohort of patients in KIMS with hypopituitarism due to pituitary/sellar lesions, including CP, who were followed up for a median of 5.3 years, Burman et al. (39) found that radiotherapy increased the risk of malignant brain tumors and meningiomas, particularly when treated at younger ages. That same trend of radiotherapy conferring an increased risk of new intracranial tumors was also observed in the current study, although it was not as statistically significant as in the aforementioned study, which may be attributed to a smaller cohort of patients with CP studied. Older age with every 10 years from disease onset might also be a risk for developing new extracranial tumors, but this risk did not exceed the risk of extracranial tumors in the reference population, implying the need for longer-term studies. Taken together, our findings and those of Burman et al. (39) further reinforce the notion that radiotherapy should be used more cautiously. In three previous studies comprising 485 patients with CP (4, 6, 10), the mortality rate was increased between fourfold and ninefold. This observation was initially thought to be mainly due to cardio- and cerebrovascular diseases (4, 6), although hypothalamic damage and its treatment regimen(s) (11), hypopituitarism with inadequately treated hypocortisolism during stress (13), and insufficient GH replacement (14) have more recently been proposed as important additional contributing factors. By contrast, our study demonstrated a lower all-cause mortality rate of 2.2-fold compared with rates of older studies (4, 6, 10) (with cardiovascular and cerebrovascular mortality observed mainly in patients with CO-CP and AO-CP, respectively). This difference may be attributable to recent refinements in surgical techniques, more physiological utilization of glucocorticoid and estrogen replacement doses, and observational bias with closer surveillance of patients on GH replacement in KIMS. Our analysis also found that greater and/or increasing BMI and history of stroke further increased the risk of early death. In addition, because patients with CP had a significantly higher risk of developing new DM and new intracranial tumors after radiotherapy compared with the general population (39), these risk factors may also play a key role in exacerbating the excess mortality of these patients. Furthermore, the risk of developing DM, intracranial tumors, and early death compared with the general population is increased because of having CP and its treatment-related complications. These patients should therefore be considered as having a chronic disease, and ongoing follow-up addressing these comorbidities should be performed beginning at disease onset. We analyzed IGF-I SDS as a TD covariate, where IGF-I SDS measured at last sampling before death or at the corresponding time from baseline for those still alive was used in the risk calculation of all-cause mortality, and we found that lower IGF-I SDS was related to increased all-cause mortality. However, it is noteworthy that the time between the event and the last IGF-I measurement may vary between patients because of the observational nature of how KIMS data were collected. The reasons for these findings are unclear. It is possible that lack of adherence to daily GH injections, reduction in GH responsiveness, or patients discontinuing GH nearing their terminal phase resulted in lower IGF-I SDS before death. Because this was a retrospective study, several limitations are acknowledged. First, the KIMS database was not designed to provide detailed individual information about the degree of tumor invasiveness and the extent of surgical resection, details on the total and fractionated radiotherapy doses administered, and the severity of hypopituitarism. Thus, the role and adequacy of glucocorticoid, sex hormone, and thyroid hormone replacement therapies on the comorbidities cannot be elucidated in this study. Second, we cannot exclude the possibility of patient infrequent use of and/or nonadherence to daily GH injections. Third, given that the patients were enrolled between 1994 and 2012, serum IGF-I levels were measured by several different assays, which could have produced variable results because of the use of different populations to establish reference values. Therefore, for the comparison of values obtained with different assays, IGF-I SDS was used with reference to the normative data for the assay in question, after appropriate transformation for nonnormality data. Although these limitations are inherent to any registry-based study, the large number of adults with CO-CP and AO-CP in our series followed up on long-term GH replacement strengthens our findings. In conclusion, we demonstrated that conventional factors that increase the risks of cardio- and cerebrovascular diseases and DM, risks for developing new intracranial tumors after radiotherapy, and lower IGF-I SDS measured at last sampling before death or at the corresponding time point from baseline for those still alive contributed to the increased morbidity and mortality of patients with CP. Close follow-up is therefore imperative for early detection of tumor recurrence in irradiated patients, and early preventive management of modifiable cardiovascular risks is important to potentially decrease the mortality of these patients. Abbreviations: AE adverse event AO-CP adult-onset craniopharyngioma BMI body mass index CI confidence interval CO-CP childhood-onset craniopharyngioma CP craniopharyngioma DM diabetes mellitus E expected number of cases GH growth hormone GHD growth hormone deficiency IGF-I insulinlike growth factor I O observed number of cases RR relative risk SAE serious adverse event SOC System Organ Class TD time-dependent. Acknowledgments The authors thank all the KIMS investigators and patients who consented to providing data. Financial Support: This research was not supported by any specific grant from any funding agency in the public, commercial, or not-for-profit sector. The KIMS database is sponsored by Pfizer Inc. Current Affiliation: K.C.J. Yuen’s current affiliation is Barrow Pituitary Center, Departments of Neuroendocrinology and Neurosurgery, Barrow Neurological Institute, Phoenix, Arizona 85013. Disclosure Summary: K.C.J.Y., P.B., and R.A. are current members of the KIMS Steering Committee. K.C.J.Y. has received research grants from Pfizer, Novo Nordisk, Teva Pharmaceuticals, OPKO Biologics, and Versartis and has served on advisory boards for Pfizer, Novo Nordisk, Sandoz, and Versartis. A.F.M., J.L.F., and C.C.-H. are employed by Pfizer, Inc. All statistical analyses were performed by a statistician (A.F.M.). P.B. has received research grants from Novartis and speaker fees from Ipsen and Novo Nordisk and has served on advisory boards for Pfizer. J.V. has received honoraria for lectures and/or advisory boards from Pfizer, Ipsen, and Novartis. M.E.G. was a KIGS investigator and a member of the KIGS International Board, is currently on advisory boards for Abbvie, Diurnal, Endo, Pfizer, Millendo Therapeutics, Novo Nordisk, Ipsen, and Sandoz, and has a research contract from Versartis. E.-M.E. has nothing to disclose. References 1. Bunin GR, Surawicz TS, Witman PA, Preston-Martin S, Davis F, Bruner JM. The descriptive epidemiology of craniopharyngioma. J Neurosurg . 1998; 89( 4): 547– 551. Google Scholar CrossRef Search ADS PubMed  2. Müller HL. Consequences of craniopharyngioma surgery in children. J Clin Endocrinol Metab . 2011; 96( 7): 1981– 1991. Google Scholar CrossRef Search ADS PubMed  3. Clark AJ, Cage TA, Aranda D, Parsa AT, Auguste KI, Gupta N. Treatment-related morbidity and the management of pediatric craniopharyngioma: a systematic review. J Neurosurg Pediatr . 2012; 10( 4): 293– 301. Google Scholar CrossRef Search ADS PubMed  4. Bülow B, Attewell R, Hagmar L, Malmström P, Nordström CH, Erfurth EM. Postoperative prognosis in craniopharyngioma with respect to cardiovascular mortality, survival, and tumor recurrence. J Clin Endocrinol Metab . 1998; 83( 11): 3897– 3904. Google Scholar PubMed  5. Olsson DS, Andersson E, Bryngelsson IL, Nilsson AG, Johannsson G. Excess mortality and morbidity in patients with craniopharyngioma, especially in patients with childhood onset: a population-based study in Sweden. J Clin Endocrinol Metab . 2015; 100( 2): 467– 474. Google Scholar CrossRef Search ADS PubMed  6. Tomlinson JW, Holden N, Hills RK, Wheatley K, Clayton RN, Bates AS, Sheppard MC, Stewart PM; West Midlands Prospective Hypopituitary Study Group. Association between premature mortality and hypopituitarism. Lancet . 2001; 357( 9254): 425– 431. Google Scholar CrossRef Search ADS PubMed  7. Crowley RK, Hamnvik OP, O’Sullivan EP, Behan LA, Smith D, Agha A, Thompson CJ. Morbidity and mortality in patients with craniopharyngioma after surgery. Clin Endocrinol (Oxf) . 2010; 73( 4): 516– 521. Google Scholar PubMed  8. Fjalldal S, Holmer H, Rylander L, Elfving M, Ekman B, Osterberg K, Erfurth EM. Hypothalamic involvement predicts cognitive performance and psychosocial health in long-term survivors of childhood craniopharyngioma. J Clin Endocrinol Metab . 2013; 98( 8): 3253– 3262. Google Scholar CrossRef Search ADS PubMed  9. Gautier A, Godbout A, Grosheny C, Tejedor I, Coudert M, Courtillot C, Jublanc C, De Kerdanet M, Poirier JY, Riffaud L, Sainte-Rose C, Van Effenterre R, Brassier G, Bonnet F, Touraine P; Craniopharyngioma Study Group. Markers of recurrence and long-term morbidity in craniopharyngioma: a systematic analysis of 171 patients. J Clin Endocrinol Metab . 2012; 97( 4): 1258– 1267. Google Scholar CrossRef Search ADS PubMed  10. Pereira AM, Schmid EM, Schutte PJ, Voormolen JH, Biermasz NR, van Thiel SW, Corssmit EP, Smit JW, Roelfsema F, Romijn JA. High prevalence of long-term cardiovascular, neurological and psychosocial morbidity after treatment for craniopharyngioma. Clin Endocrinol (Oxf) . 2005; 62( 2): 197– 204. Google Scholar CrossRef Search ADS PubMed  11. Yuen KC, Kołtowska-Häggström M, Cook DM, Fox JL, Jönsson PJ, Geffner ME, Abs R. Primary treatment regimen and diabetes insipidus as predictors of health outcomes in adults with childhood-onset craniopharyngioma. J Clin Endocrinol Metab . 2014; 99( 4): 1227– 1235. Google Scholar CrossRef Search ADS PubMed  12. Bülow B, Hagmar L, Eskilsson J, Erfurth EM. Hypopituitary females have a high incidence of cardiovascular morbidity and an increased prevalence of cardiovascular risk factors. J Clin Endocrinol Metab . 2000; 85( 2): 574– 584. Google Scholar PubMed  13. Burman P, Mattsson AF, Johannsson G, Höybye C, Holmer H, Dahlqvist P, Berinder K, Engström BE, Ekman B, Erfurth EM, Svensson J, Wahlberg J, Karlsson FA. Deaths among adult patients with hypopituitarism: hypocortisolism during acute stress, and de novo malignant brain tumors contribute to an increased mortality. J Clin Endocrinol Metab . 2013; 98( 4): 1466– 1475. Google Scholar CrossRef Search ADS PubMed  14. Gaillard RC, Mattsson AF, Akerblad AC, Bengtsson BA, Cara J, Feldt-Rasmussen U, Koltowska-Häggström M, Monson JP, Saller B, Wilton P, Abs R. Overall and cause-specific mortality in GH-deficient adults on GH replacement. Eur J Endocrinol . 2012; 166( 6): 1069– 1077. Google Scholar CrossRef Search ADS PubMed  15. Olsson DS, Trimpou P, Hallén T, Bryngelsson IL, Andersson E, Skoglund T, Bengtsson BA, Johannsson G, Nilsson AG. Life expectancy in patients with pituitary adenoma receiving growth hormone replacement. Eur J Endocrinol . 2017; 176( 1): 67– 75. Google Scholar CrossRef Search ADS PubMed  16. van Bunderen CC, van Nieuwpoort IC, Arwert LI, Heymans MW, Franken AA, Koppeschaar HP, van der Lely AJ, Drent ML. Does growth hormone replacement therapy reduce mortality in adults with growth hormone deficiency? Data from the Dutch National Registry of Growth Hormone Treatment in adults. J Clin Endocrinol Metab . 2011; 96( 10): 3151– 3159. Google Scholar CrossRef Search ADS PubMed  17. Pappachan JM, Raskauskiene D, Kutty VR, Clayton RN. Excess mortality associated with hypopituitarism in adults: a meta-analysis of observational studies. J Clin Endocrinol Metab . 2015; 100( 4): 1405– 1411. Google Scholar CrossRef Search ADS PubMed  18. Cook DM, Yuen KC, Biller BM, Kemp SF, Vance ML; American Association of Clinical Endocrinologists. American Association of Clinical Endocrinologists medical guidelines for clinical practice for growth hormone use in growth hormone-deficient adults and transition patients: 2009 update. Endocr Pract . 2009; 15( Suppl 2): 1– 29. Google Scholar CrossRef Search ADS PubMed  19. Molitch ME, Clemmons DR, Malozowski S, Merriam GR, Vance ML; Endocrine Society. Evaluation and treatment of adult growth hormone deficiency: an Endocrine Society clinical practice guideline. J Clin Endocrinol Metab . 2011; 96( 6): 1587– 1609. Google Scholar CrossRef Search ADS PubMed  20. American Diabetes Association. 2. Classification and diagnosis of diabetes. Diabetes Care . 2016; 39( Suppl 1): S13– S22. CrossRef Search ADS PubMed  21. Luger A, Mattsson AF, Koltowska-Häggström M, Thunander M, Góth M, Verhelst J, Abs R. Incidence of diabetes mellitus and evolution of glucose parameters in growth hormone-deficient subjects during growth hormone replacement therapy: a long-term observational study. Diabetes Care . 2012; 35( 1): 57– 62. Google Scholar CrossRef Search ADS PubMed  22. Alberti KG, Zimmet P, Shaw J. Metabolic syndrome--a new world-wide definition: a consensus statement from the International Diabetes Federation. Diabet Med . 2006; 23( 5): 469– 480. Google Scholar CrossRef Search ADS PubMed  23. Verhelst J, Mattsson AF, Luger A, Thunander M, Góth MI, Koltowska-Häggström M, Abs R. Prevalence and characteristics of the metabolic syndrome in 2479 hypopituitary patients with adult-onset GH deficiency before GH replacement: a KIMS analysis. Eur J Endocrinol . 2011; 165( 6): 881– 889. Google Scholar CrossRef Search ADS PubMed  24. Brabant G, von zur Mühlen A, Wüster C, Ranke MB, Kratzsch J, Kiess W, Ketelslegers JM, Wilhelmsen L, Hulthén L, Saller B, Mattsson A, Wilde J, Schemer R, Kann P; German KIMS Board. Serum insulin-like growth factor I reference values for an automated chemiluminescence immunoassay system: results from a multicenter study. Horm Res . 2003; 60( 2): 53– 60. Google Scholar PubMed  25. Hughes AO, Hills M. Statistical Models in Epidemiology . Oxford, UK: Oxford University Press; 1993. 26. Breslow NE, Day NE. Statistical Methods in Cancer Research. Vol. 2. The Design and Analysis of Cohort Studies.  IARC Scientific Publications No. 82. Lyon, France: International Agency for Research on Cancer; 1987. 27. Curado MP, Edwards B, Shin HR, Storm H, Ferlay J, Heanue M, Boyle P. Cancer Incidence in Five Continents. Vol. 9. No. 160 . Lyon, France: International Agency for Research on Cancer; 2007. 28. Thunander M, Petersson C, Jonzon K, Fornander J, Ossiansson B, Torn C, Edvardsson S, Landin-Olsson M. Incidence of type 1 and type 2 diabetes in adults and children in Kronoberg, Sweden. Diabetes Res Clin Pract . 2008; 82( 2): 247– 255. Google Scholar CrossRef Search ADS PubMed  29. World Health Organization. World health statistics 2008. Available at: http://www.who.int/gho/publications/world_health_statistics/EN_WHS08_Full.pdf. Accessed 31 October 2017. 30. Kendall-Taylor P, Jönsson PJ, Abs R, Erfurth EM, Koltowska-Häggström M, Price DA, Verhelst J. The clinical, metabolic and endocrine features and the quality of life in adults with childhood-onset craniopharyngioma compared with adult-onset craniopharyngioma. Eur J Endocrinol . 2005; 152( 4): 557– 567. Google Scholar CrossRef Search ADS PubMed  31. Attanasio AF, Bates PC, Ho KK, Webb SM, Ross RJ, Strasburger CJ, Bouillon R, Crowe B, Selander K, Valle D, Lamberts SW; Hypopituitary Control and Complications Study International Advisory Board. Human growth hormone replacement in adult hypopituitary patients: long-term effects on body composition and lipid status--3-year results from the HypoCCS Database. J Clin Endocrinol Metab . 2002; 87( 4): 1600– 1606. Google Scholar PubMed  32. Koranyi J, Svensson J, Götherström G, Sunnerhagen KS, Bengtsson B, Johannsson G. Baseline characteristics and the effects of five years of GH replacement therapy in adults with GH deficiency of childhood or adulthood onset: a comparative, prospective study. J Clin Endocrinol Metab . 2001; 86( 10): 4693– 4699. Google Scholar CrossRef Search ADS PubMed  33. Abs R, Mattsson AF, Thunander M, Verhelst J, Góth MI, Wilton P, Kołtowska-Häggström M, Luger A. Prevalence of diabetes mellitus in 6050 hypopituitary patients with adult-onset GH deficiency before GH replacement: a KIMS analysis. Eur J Endocrinol . 2013; 168( 3): 297– 305. Google Scholar CrossRef Search ADS PubMed  34. Andersson T, Ahlbom A, Carlsson S. Diabetes prevalence in Sweden at present and projections for year 2050. PLoS One . 2015; 10( 11): e0143084. Google Scholar CrossRef Search ADS PubMed  35. Holmer H, Svensson J, Rylander L, Johannsson G, Rosén T, Bengtsson BA, Thorén M, Höybye C, Degerblad M, Bramnert M, Hägg E, Edén Engström B, Ekman B, Norrving B, Hagmar L, Erfurth EM. Nonfatal stroke, cardiac disease, and diabetes mellitus in hypopituitary patients on hormone replacement including growth hormone. J Clin Endocrinol Metab . 2007; 92( 9): 3560– 3567. Google Scholar CrossRef Search ADS PubMed  36. Bogazzi F, Ultimieri F, Raggi F, Russo D, Vanacore R, Guida C, Brogioni S, Cosci C, Gasperi M, Bartalena L, Martino E. Growth hormone inhibits apoptosis in human colonic cancer cell lines: antagonistic effects of peroxisome proliferator activated receptor-gamma ligands. Endocrinology . 2004; 145( 7): 3353– 3362. Google Scholar CrossRef Search ADS PubMed  37. Karavitaki N, Warner JT, Marland A, Shine B, Ryan F, Arnold J, Turner HE, Wass JA. GH replacement does not increase the risk of recurrence in patients with craniopharyngioma. Clin Endocrinol (Oxf) . 2006; 64( 5): 556– 560. Google Scholar CrossRef Search ADS PubMed  38. Olsson DS, Buchfelder M, Wiendieck K, Kremenevskaja N, Bengtsson BA, Jakobsson KE, Jarfelt M, Johannsson G, Nilsson AG. Tumour recurrence and enlargement in patients with craniopharyngioma with and without GH replacement therapy during more than 10 years of follow-up. Eur J Endocrinol . 2012; 166( 6): 1061– 1068. Google Scholar CrossRef Search ADS PubMed  39. Burman P, van Beek AP, Biller BM, Camacho-Hübner C, Mattsson AF. Radiotherapy, especially at young age, increases the risk for de novo brain tumors in patients treated for pituitary/sellar lesions. J Clin Endocrinol Metab . 2017; 102( 3): 1051– 1058. Google Scholar PubMed  Copyright © 2018 Endocrine Society http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Clinical Endocrinology and Metabolism Oxford University Press

Relative Risks of Contributing Factors to Morbidity and Mortality in Adults With Craniopharyngioma on Growth Hormone Replacement

Loading next page...
 
/lp/ou_press/relative-risks-of-contributing-factors-to-morbidity-and-mortality-in-LOVjrFW5WJ
Publisher
Endocrine Society
Copyright
Copyright © 2018 Endocrine Society
ISSN
0021-972X
eISSN
1945-7197
D.O.I.
10.1210/jc.2017-01542
Publisher site
See Article on Publisher Site

Abstract

Abstract Context In adults, craniopharyngioma (CP) of either childhood-onset (CO-CP) or adult-onset (AO-CP) is associated with increased morbidity and mortality, but data on the relative risks (RRs) of contributing factors are lacking. Objective To assess the RRs of factors contributing to morbidity and mortality in adults with CO-CP and AO-CP. Methods Data on 1669 patients with CP from KIMS (Pfizer International Metabolic Database) were analyzed using univariate and multiple Poisson and Cox regression methods. Results When CO-CP and AO-CP groups were combined, history of stroke and hyperlipidemia increased cardiovascular risk, higher body mass index (BMI) and radiotherapy increased cerebrovascular risk, and increased waist circumference increased the risk of developing diabetes mellitus (DM). Compared with patients with CO-CP, patients with AO-CP had a threefold higher risk of tumor recurrence, whereas being female and previous radiotherapy exposure conferred lower risks. Radiotherapy and older age with every 10 years from disease onset conferred a 2.3- to 3.5-fold risk for developing new intracranial tumors, whereas older age, greater and/or increasing BMI, history of stroke, and lower insulinlike growth factor I (IGF-I) standard deviation score measured at last sampling before death were related to increased all-cause mortality. Compared with the general population, adults with CP had 9.3-, 8.1-, and 2.2-fold risks of developing DM, new intracranial tumors, and early death, respectively. Conclusion Conventional factors that increase the risks of cardio- and cerebrovascular diseases and DM and risks for developing new intracranial tumors contributed to excess morbidity and mortality. In addition, lower serum IGF-I level measured from the last sample before death was inversely associated with mortality risk in patients with CP. Craniopharyngiomas (CPs) are embryonic malformations that arise within the sellar and suprasellar regions and often infiltrate the pituitary gland and hypothalamus. These tumors may occur at any age but typically demonstrate a bimodal distribution, with peak childhood-onset (CO-CP) and adult-onset (AO-CP) incidence rates occurring between the ages of 5 and 14 years and the ages of 50 and 74 years, respectively (1). Although CPs are histologically benign, these tumors tend to invade locally, frequently resulting in recurrence even after aggressive surgery (2) and/or radiotherapy (3). Previous studies demonstrated that combined populations of pediatric and adult patients with CP but without growth hormone (GH) replacement have morbidity and mortality rates that are four- to ninefold higher than those of the general population (4, 5) and nearly 10-fold higher than those of individuals with other causes of hypopituitarism (6). The excess morbidity is associated with early disease onset, primary treatment regimen, severity of hypopituitarism, visual disturbances, hypothalamic damage, and initial symptoms of intracranial hypertension (7–11), whereas the increased mortality has been attributed mainly to cardio- and cerebrovascular diseases (4–6, 12). In patients who had received more contemporary hormone replacement therapy regimens, including GH in those with GH deficiency (GHD) and lower glucocorticoid replacement doses, the overall mortality rate was still at least fourfold higher than expected, whereas in patients treated for nonfunctioning pituitary adenomas and prolactinomas, mortality rates were comparable to those of the reference population (13). Accordingly, several other studies (14–16), including a recent meta-analysis by Pappachan et al. (17), reported beneficial effects favoring GH replacement in adults with hypopituitarism, including substantial reductions in mortality. In the current study, the main objective was to evaluate the relative risks (RRs) of contributing factors to excess morbidity and mortality in a large cohort of patients with CP receiving long-term GH replacement. Another objective was to examine the overall and cause-specific mortality rates and comorbidities, specifically cardio- and cerebrovascular diseases, diabetes mellitus (DM), and second brain tumors, of these patients. Patients and Methods Patients Data on 1669 adults with CO-CP (CP diagnosed at <18 years of age; n = 744) and AO-CP (CP diagnosed at ≥18 years of age; n = 925) were extracted from KIMS (Pfizer International Metabolic Database), the largest pharmaco-epidemiological surveillance study on GH replacement in adults with GHD. The 16,484 adult patients in KIMS were enrolled between 1994 and 2012 from 28 countries, and data were collected on case report forms by treating physicians at routine clinic visits. All patients included in the current study had confirmed diagnosis of severe GHD by GH stimulation testing or had at least three pituitary hormone deficiencies and a low serum insulinlike growth factor I (IGF-I) standard deviation score (SDS) < −2, in accordance with current clinical practice guidelines (18, 19). The median time from KIMS entry to the last follow-up clinic visit was 5.3 years (range, 0 to 18.3 years), representing a total of 10,284 patient-years. Data collection into KIMS was approved by the institutional review boards/ethical committees as required by local regulations in each participating center. Written informed consent was obtained from all patients before any data were entered into KIMS. Methods All data extracted from the KIMS database were provided from all the participating centers. Follow-up data were analyzed and consisted of average GH dose used, the development of DM, and the reporting of adverse events (AEs) by the participating investigators. DM was defined by physician documentation or use of antidiabetes medication(s) at baseline and/or biochemical data in accordance with the American Diabetes Association guidelines (20), as previously described (21). This study was divided into four parts. In the first part, a cross-sectional analysis was performed by comparing background clinical characteristics of patients with CO-CP and patients with AO-CP to determine whether demographics, phenotypic features, and biochemical parameters differed between the two groups. In the second part, differences in rates of serious AEs (SAEs) between patients with CO-CP and AO-CP were analyzed. In the third part, Cox regression analyses with internal reference were performed with adjustment for age and sex to assess the RRs of the following outcomes: cardio- and cerebrovascular diseases, DM, recurrence of CP, development of new intracranial and extracranial tumors, and all-cause mortality. These outcomes were assessed in relation to covariates hypothesized to systematically influence their rates of occurrence, such as medical history (whether CO-CP or AO-CP, number and type of surgeries, use of radiotherapy, history of stroke, smoking, and naivety to GH replacement), clinical presentation at baseline [age, sex, body mass index (BMI), hypertension, waist circumference, hyperlipidemia, diabetes insipidus, prediabetes defined as hemoglobin A1c value between 5.7% and 6.4%, DM, and metabolic syndrome at baseline based on the International Diabetes Federation definition] (22), IGF-I SDS, and follow-up data during GH replacement (time from baseline, age at KIMS baseline or attained age during follow-up, GH dose, and IGF-I SDS during GH replacement). In the fourth part, rates were compared with those in the general population for available end points (i.e., incidence rates of DM, incidence rates of intracranial tumors and extracranial tumors, respectively, and mortality from all causes, and from cardio- and cerebrovascular causes). The dose-titration regimens for GH replacement against serum IGF-I levels were prescribed by treating physicians on the basis of local practice. Serum IGF-I and lipid levels were measured centrally, as described previously (23), and plasma glucose and hemoglobin A1c levels were analyzed locally at each participating center. To make meaningful comparisons between IGF-I measurements and age- and sex-specific reference ranges, IGF-I SDS was calculated according to the following formulae: between 1994 and 1997, IGF-I SDS = [ln (IGF-I) − (5.95 − 0.0197 × age)]; between 1997 and 2002, IGF-I SDS = [ln (IGF-1) − (15.92 − 0.0146 × age)/0.272]; and from 2002 onward, based on data by Brabant et al. (24). Statistical methods For descriptive statistics, means, standard deviations, medians with fifth and 95th percentiles, and proportions (percentages) were calculated depending on the type of variable. For comparisons between patients with CO-CP and AO-CP, heterogeneity tests for nominal or categorical variables were performed with χ2 tests, whereas for numerical variables, Student t tests were performed. A crude presentation was performed to assess the overall incidence rates of reported SAEs presented according to the Medical Dictionary for Regulatory Activities System Organ Class (SOC) level comparing patients with CO-CP and those with AO-CP. The calculation was based on the first SAE in a patient on preferred term level but was presented on Medical Dictionary for Regulatory Activities SOC level. Crude comparisons with incidence rate ratios between patients with CO-CP and those with AO-CP were performed with score tests (25), whereas 95% confidence intervals (CIs) were calculated with the Byar equation (26). Significant rate differences on the crude level were further analyzed with adjustment for age and sex by proportional hazard regression. For the seven specific end points, regression analyses were performed with proportional hazards multiple regression. Time-to-event was defined as years from KIMS baseline to the date of first AE report or, if no event, to last visit date or date of death. Unless defined differently, the end points with hypothesized risk factors measured at baseline were the following: sex (for all risk factors), age at baseline (all), attained age (all), obesity by waist circumference or by BMI as time-dependent (TD) covariates (a, b, c, f, g), DM (a, b, g), and specifically: For occurrence of cardiovascular disease: hyperlipidemia (yes vs. no), total cholesterol, hypertension (yes vs. no), systolic and diastolic blood pressure (TD), metabolic syndrome, cardio/cerebrovascular events (TD) in medical history, smoking (yes vs. no), IGF-I SDS (TD), GH dose (TD), and hydrocortisone equivalent dose For occurrence of cerebrovascular disease: hyperlipidemia (yes vs. no), total cholesterol, hypertension (yes vs. no), systolic and diastolic blood pressure (TD), cardio/cerebrovascular events (TD) in medical history, radiotherapy (yes vs. no), diabetes insipidus (yes vs. no), smoking (yes vs. no), IGF-I SDS (TD), and GH dose (TD) For occurrence of new DM: hemoglobin A1c, fasting blood glucose, follow-up year after baseline, hydrocortisone equivalent dose, and TD sex and age effects For recurrence of CP: surgery (yes vs. no), number and type of surgeries, radiotherapy (yes vs. no), and IGF-I SDS (TD) For occurrence of new intracranial malignant tumors: surgery (yes vs. no), number and type of surgeries, radiotherapy (yes vs. no), previous cancer in medical history, smoking (yes vs. no), and IGF-I SDS (TD) For occurrence of new extracranial malignant tumors: radiotherapy (yes vs. no), previous cancer in medical history, smoking (yes vs. no), and IGF-I SDS (TD) For all-cause mortality: myocardial infarction and cerebrovascular disease in medical history (yes vs. no), diabetes insipidus (yes vs. no), adrenocorticotropic hormone deficiency (yes vs. no), radiotherapy, blood pressure, hypertension, smoking (yes vs. no), and IGF-I SDS (TD) Standardized incidence ratios and standardized mortality ratios were calculated using stratification for attained age, and/or sex, and/or country/region, depending on the level of stratification in the specific external reference. These ratios compare the observed number of cases in the patient group (O) and the expected number of cases (E). The latter number (E) quantifies the expected cases in the patient-group when the patient-group had the same specific rates as the chosen external reference population. Patient-years were calculated from the date of KIMS entry or GH start date (if later than the KIMS entry date) to the date of the studied event or, if no event, the date of the last visit or date of death. General population (reference) rates for cancer incidence were from the International Agency for Research on Cancer, with cancer incidence from five continents (27) stratified for age, sex, and country. Reference rates for DM were age- and sex-specific from a study on incidence of DM in southern Sweden (28). The age-, sex-, and region-specific reference rates for mortality were taken from the World Health Organization project Global Burden of Disease (29). The 95% CIs were calculated with the Byar approximation formula. In the regression models, CIs and P values were based on likelihood methods (26). P values <0.05 were considered statistically significant. Higher risk estimates in combination with a P value <0.10 were also considered. Results Baseline clinical characteristics, medical history, biochemical parameters, and hypothesized risk factors There were no differences in sex ratio between the patients with CO-CP and those with AO-CP. Compared with patients with AO-CP, patients with CO-CP more often underwent combined surgery and radiotherapy; had less than three pituitary hormone deficiencies; more often had thyroid-stimulating hormone, gonadotropin, and arginine vasopressin deficiencies; underwent longer duration of KIMS follow-up; had lower BMI, waist circumference, systolic and diastolic blood pressure values, fasting glucose level, and hemoglobin A1c value; had more favorable lipid profiles; had lower IGF-I SDS at baseline; and had lower IGF-I SDS at 1 year despite being treated with higher GH doses at 1 year (Table 1). Table 1. Background, GH Replacement Chronology, Clinical Characteristics, and Biochemical Characteristics of Adults With CO-CP and AO-CP   CO-CP  CO-CP vs. AO-CP, Crude P Valuea  AO-CP  Combined CO-CP + AO-CP  Sex, males/females, %  53.1/46.9  NS  53.0/47.0  53.0/47.0  Age at diagnosis of CP, y  11.1 (4.5)  <0.0001  36.7 (13.3)  25.3 (16.4)  Treatment of CP, %           Surgery (transsphenoidal/transcranial) only  65.5  NS  72.4  69.3   Surgery (transsphenoidal/transcranial) and radiotherapy  22.9  0.02  18.0  20.1   Radiotherapy only  4.4  NS  2.8  3.5  Pituitary hormone deficiencies excluding GHD, %           No hormonal deficit  1.2  NS  1.5  1.4   1 Hormonal deficit  3.1  NS  4.4  3.8   2 Hormonal deficits  5.1  NS  9.4  7.5   3 Hormonal deficits  24.6  <0.0001  30.0  27.6   4 Hormonal deficits  66.0  NS  54.6  59.7  Types of pituitary hormone deficiencies, %           ACTH  90.4  NS  89.1  89.7   TSH  95.4  0.004  91.9  93.4   LH/FSH  94.9  <0.0001  89.1  91.6   AVP  71.9  <0.0001  62.8  66.8  GH replacement chronology, y           Age at GH start in childhood  11.8 (4.2)  —  —  —   Age at GH stop in childhood  17.9 (2.8)  —  —  —   Age at GH start in adulthood  26.9 (9.7)  <0.0001  44.2 (13.1)  36.5 (14.5)   Age at KIMS entry  27.7 (9.7)  <0.01  44.7 (13.2)  37.1 (14.4)   Duration of KIMS follow-up since GH start  6.4 (4.8)  0.03  5.9 (4.6)  6.2 (4.8)  Clinical characteristics           BMI, kg/m2  29.5 (7.6)  0.026  30.3 (6.5)  29.9 (7.0)   Waist circumference, cm  96.8 (16.7)  0.0004  100.1 (15.1)  98.6 (15.9)   Systolic blood pressure, mm Hg  118.0 (14.6)  <0.0001  125.0 (16.4)  121.9 (16.0)   Diastolic blood pressure, mm Hg  74.5 (9.8)  <0.0001  78.5 (10.5)  76.8 (10.4)  Biochemical characteristics           Fasting glucose, mmol/L  4.7 (1.0)  0.043  4.9 (0.9)  4.8 (1.0)   Hemoglobin A1c, %  5.2 (2.0)  0.002  5.6 (3.1)  5.4 (2.7)   Total cholesterol, mmol/L  5.5 (1.3)  0.0002  5.9 (1.2)  5.7 (1.3)   HDL cholesterol, mmol/L  1.3 (0.4)  0.04  1.2 (0.4)  1.2 (0.4)   LDL cholesterol mmol/L  3.4 (1.1)  0.007  3.6 (1.1)  3.5 (1.1)   Triglycerides, mmol/L  2.2 (1.8)  0.001  2.6 (1.9)  2.4 (1.9)   IGF-I SDS at baseline  −2.7 (2.6)  <0.0001  −1.1 (2.0)  −1.8 (2.4)   GH dose, mg/d, at 1 y  0.6 (0.4)  <0.0001  0.4 (0.3)  0.5 (0.3)   IGF-I SDS at 1 y  −0.9 (2.0)  <0.0001  0.4 (1.7)  −0.2 (1.9)    CO-CP  CO-CP vs. AO-CP, Crude P Valuea  AO-CP  Combined CO-CP + AO-CP  Sex, males/females, %  53.1/46.9  NS  53.0/47.0  53.0/47.0  Age at diagnosis of CP, y  11.1 (4.5)  <0.0001  36.7 (13.3)  25.3 (16.4)  Treatment of CP, %           Surgery (transsphenoidal/transcranial) only  65.5  NS  72.4  69.3   Surgery (transsphenoidal/transcranial) and radiotherapy  22.9  0.02  18.0  20.1   Radiotherapy only  4.4  NS  2.8  3.5  Pituitary hormone deficiencies excluding GHD, %           No hormonal deficit  1.2  NS  1.5  1.4   1 Hormonal deficit  3.1  NS  4.4  3.8   2 Hormonal deficits  5.1  NS  9.4  7.5   3 Hormonal deficits  24.6  <0.0001  30.0  27.6   4 Hormonal deficits  66.0  NS  54.6  59.7  Types of pituitary hormone deficiencies, %           ACTH  90.4  NS  89.1  89.7   TSH  95.4  0.004  91.9  93.4   LH/FSH  94.9  <0.0001  89.1  91.6   AVP  71.9  <0.0001  62.8  66.8  GH replacement chronology, y           Age at GH start in childhood  11.8 (4.2)  —  —  —   Age at GH stop in childhood  17.9 (2.8)  —  —  —   Age at GH start in adulthood  26.9 (9.7)  <0.0001  44.2 (13.1)  36.5 (14.5)   Age at KIMS entry  27.7 (9.7)  <0.01  44.7 (13.2)  37.1 (14.4)   Duration of KIMS follow-up since GH start  6.4 (4.8)  0.03  5.9 (4.6)  6.2 (4.8)  Clinical characteristics           BMI, kg/m2  29.5 (7.6)  0.026  30.3 (6.5)  29.9 (7.0)   Waist circumference, cm  96.8 (16.7)  0.0004  100.1 (15.1)  98.6 (15.9)   Systolic blood pressure, mm Hg  118.0 (14.6)  <0.0001  125.0 (16.4)  121.9 (16.0)   Diastolic blood pressure, mm Hg  74.5 (9.8)  <0.0001  78.5 (10.5)  76.8 (10.4)  Biochemical characteristics           Fasting glucose, mmol/L  4.7 (1.0)  0.043  4.9 (0.9)  4.8 (1.0)   Hemoglobin A1c, %  5.2 (2.0)  0.002  5.6 (3.1)  5.4 (2.7)   Total cholesterol, mmol/L  5.5 (1.3)  0.0002  5.9 (1.2)  5.7 (1.3)   HDL cholesterol, mmol/L  1.3 (0.4)  0.04  1.2 (0.4)  1.2 (0.4)   LDL cholesterol mmol/L  3.4 (1.1)  0.007  3.6 (1.1)  3.5 (1.1)   Triglycerides, mmol/L  2.2 (1.8)  0.001  2.6 (1.9)  2.4 (1.9)   IGF-I SDS at baseline  −2.7 (2.6)  <0.0001  −1.1 (2.0)  −1.8 (2.4)   GH dose, mg/d, at 1 y  0.6 (0.4)  <0.0001  0.4 (0.3)  0.5 (0.3)   IGF-I SDS at 1 y  −0.9 (2.0)  <0.0001  0.4 (1.7)  −0.2 (1.9)  Data are presented as mean (standard deviation) unless otherwise stated. Significant P values are shown in bold. Abbreviations: ACTH, adrenocorticotropic hormone; AVP, arginine vasopressin; FSH, follicle-stimulating hormone; HDL, high-density lipoprotein; LDL, low-density lipoprotein; LH, luteinizing hormone; NS, not significant; TSH, thyroid-stimulating hormone. a P values for crude comparison between CO-CP vs. AO-CP groups. Heterogeneity tests for nominal or categorical variables were performed with χ2. Unpaired Student t tests were performed for numerical variables. After adjustment for age at baseline and sex, diastolic blood pressure, HDL cholesterol, triglycerides, IGF-I SDS, and GH dose remained significantly different between onset groups. View Large At baseline before KIMS entry, more patients with AO-CP were smokers and more often had DM, cardiovascular and cerebrovascular disease, hyperlipidemia, and cancer than those with CO-CP. By contrast, more patients with CO-CP underwent transcranial surgery and combined radiotherapy as well as transcranial and transsphenoidal surgeries, whereas more patients with AO-CP underwent transsphenoidal surgeries (Table 2). Table 2. Risk Factors of Adults With CO-CP and AO-CP at Baseline   CO-CP  AO-CP  Combined CO-CP + AO-CP  Smokers  89 (12.0)  206 (22.3)  295 (17.7)  DM  45 (6.0)  100 (10.8)  145 (8.7)  Metabolic syndrome (per IDF definition)a  102 (35.2)  236 (58.3)  338 (48.6)  Cardiovascular disease in medical history  20 (2.7)  58 (6.3)  78 (4.7)  Cerebrovascular disease in medical history  10 (1.4)  20 (2.2)  30 (1.8)  Hyperlipidemia  51 (6.8)  121 (13.1)  172 (10.3)  Cancer diagnosis before KIMS  4 (0.5)  8 (0.9)  12 (0.7)  Number of surgeries before KIMSb         0  21 (3.0)  20 (2.3)  41 (2.6)   1  507 (72.0)  677 (77.4)  1184 (75.0)   ≥2  176 (25.0)  178 (20.3)  354 (22.4)  Type of primary treatmentb         Transcranial surgery  376 (50.5)  419 (45.3)  795 (47.6)   Transsphenoidal surgery  104 (14.0)  235 (25.4)  339 (20.3)   Radiotherapy ± transcranial ± transsphenoidal surgeries  210 (28.2)  208 (22.5)  418 (25.0)    CO-CP  AO-CP  Combined CO-CP + AO-CP  Smokers  89 (12.0)  206 (22.3)  295 (17.7)  DM  45 (6.0)  100 (10.8)  145 (8.7)  Metabolic syndrome (per IDF definition)a  102 (35.2)  236 (58.3)  338 (48.6)  Cardiovascular disease in medical history  20 (2.7)  58 (6.3)  78 (4.7)  Cerebrovascular disease in medical history  10 (1.4)  20 (2.2)  30 (1.8)  Hyperlipidemia  51 (6.8)  121 (13.1)  172 (10.3)  Cancer diagnosis before KIMS  4 (0.5)  8 (0.9)  12 (0.7)  Number of surgeries before KIMSb         0  21 (3.0)  20 (2.3)  41 (2.6)   1  507 (72.0)  677 (77.4)  1184 (75.0)   ≥2  176 (25.0)  178 (20.3)  354 (22.4)  Type of primary treatmentb         Transcranial surgery  376 (50.5)  419 (45.3)  795 (47.6)   Transsphenoidal surgery  104 (14.0)  235 (25.4)  339 (20.3)   Radiotherapy ± transcranial ± transsphenoidal surgeries  210 (28.2)  208 (22.5)  418 (25.0)  Data are presented as number of patients (%). Abbreviation: IDF, International Diabetes Federation. a Among patients with nonmissing values (CO-CP: 290; AO-CP: 405). b Among patients with nonmissing values (CO-CP: 704; AO-CP: 875). View Large Factors affecting morbidity and mortality rates Older age at baseline, history of stroke, and hyperlipidemia were associated with increased risk of cardiovascular disease, whereas higher BMI (P = 0.055) showed a tendency toward increasing that risk. Older age at baseline, higher BMI, and a history of stroke (P = 0.10) increased the risk of cerebrovascular disease, whereas previous radiotherapy exposure (P = 0.058) tended to increase that risk. The risk of developing new DM increased with older age at baseline and with higher waist circumference and baseline hemoglobin A1c level. Female sex conferred a 2.1-fold increased risk at baseline following KIMS entry; thereafter, this risk became TD, with a decreasing rate of 44% per year of follow-up compared with males. This indicated that after 1.5 to 2.5 years of follow-up, rates were similar between sexes, and after 5 to 6 years of follow-up, males showed higher rates. Risk of recurrence of CP was lower in females and in those with previous surgery (vs. no surgery) and previous exposure to radiotherapy and was higher in adult-onset disease. Radiotherapy was associated with a 3.5-fold risk (95% CI: 0.9 to 13; P = 0.067) for developing new intracranial tumors (five de novo malignant brain tumors consisting of high-grade glioblastoma, glioblastoma multiforme, and anaplastic oligoastrocytoma; four meningiomas; and three unspecified primary brain tumors), whereas older attained age with every 10 years from disease onset conferred a 2.3-fold risk for developing new extracranial tumors (Fig. 1). The risk of extracranial tumors was similar compared with that of the reference population (standardized incidence ratio = 0.93; Fig. 2). Greater and/or increasing BMI, history of stroke, and older age with every 10 years from disease onset were associated with a 1.5- to 3.5-fold risk, respectively, of increased mortality. In addition, IGF-I SDS measured at the last sampling before death or at the corresponding time from baseline for patients who were still alive was inversely associated with all-cause mortality (27% lower rate per unit higher IGF-1 SDS; 95% CI: −13% to −38%; P = 0.0003) (Fig. 1). Other variables, as listed in the “Statistical methods” section, were not statistically significant in explaining the variation in risk(s). Figure 1. View largeDownload slide RRs (hazard ratios) of individual contributing factors (cardio- and cerebrovascular diseases and DM as well as CP recurrence, new intracranial and extracranial tumors, and all-cause mortality) to morbidity and mortality in adults with CO-CP and AO-CP. Patients with CO-CP and AO-CP had similar risks for all end points except recurrence (data not shown). Significant P values are shown in bold. *BMI TD variable; unit = kg/m2. **Females had hazard ratio = 2.13 vs. males at KIMS baseline, but after 1 to 2 years, the hazard ratio was lower in females (44% decrease per year). ***Surgery per se decreases the risk, but more surgeries indicate increasing risk. HbA1c, hemoglobin A1c; NV, numerical variable (no specific reference range); PY, per year. Figure 1. View largeDownload slide RRs (hazard ratios) of individual contributing factors (cardio- and cerebrovascular diseases and DM as well as CP recurrence, new intracranial and extracranial tumors, and all-cause mortality) to morbidity and mortality in adults with CO-CP and AO-CP. Patients with CO-CP and AO-CP had similar risks for all end points except recurrence (data not shown). Significant P values are shown in bold. *BMI TD variable; unit = kg/m2. **Females had hazard ratio = 2.13 vs. males at KIMS baseline, but after 1 to 2 years, the hazard ratio was lower in females (44% decrease per year). ***Surgery per se decreases the risk, but more surgeries indicate increasing risk. HbA1c, hemoglobin A1c; NV, numerical variable (no specific reference range); PY, per year. Figure 2. View largeDownload slide Observed/expected (O/E) ratios with 95% CIs for new intracranial malignant tumors, new extracranial malignant tumors, new DM, and mortality in adults with CO-CP and AO-CP. Significant P values are shown in bold. External incidence rates of intracranial (ICDO: C70-C72) and extracranial (ICDO: C00-C69, C73-C97) malignant tumors are from the International Agency for Research on Cancer (27). External incidence rates for DM are from Kronoberg County of southern Sweden (28). External mortality rates are from the World Health Organization project Global Burden of Disease (29). ICDO, International Classification of Diseases for Oncology. Figure 2. View largeDownload slide Observed/expected (O/E) ratios with 95% CIs for new intracranial malignant tumors, new extracranial malignant tumors, new DM, and mortality in adults with CO-CP and AO-CP. Significant P values are shown in bold. External incidence rates of intracranial (ICDO: C70-C72) and extracranial (ICDO: C00-C69, C73-C97) malignant tumors are from the International Agency for Research on Cancer (27). External incidence rates for DM are from Kronoberg County of southern Sweden (28). External mortality rates are from the World Health Organization project Global Burden of Disease (29). ICDO, International Classification of Diseases for Oncology. Both patients with CO-CP and those with AO-CP showed increased and similar O/E ratios for developing DM (~9.3-fold) and new intracranial malignant tumors (~8.1-fold) and for all-cause mortality (∼2.2-fold in the observed time-window of follow-up, stratified by attained age, sex, and region) (Fig. 2). For cardio- and cerebrovascular mortalities, patients with CO-CP and those with AO-CP showed increased O/E ratios of 2.0-fold (95% CI: 1.20 to 3.20; 18 cases) and 3.2-fold (95% CI: 1.30 to 6.30; 8 cases), respectively, when compared with the general population. Incidence rates of reported SAEs There were no significant differences in crude rates of SAEs, except for the SOCs “surgical and medical procedures,” “neoplasms benign, malignant, unspecified (including cysts and polyps),” and “infections and infestations,”’ respectively (Table 3). After adjustment for age and sex, the incidence rate ratios were similar [0.40 (95% CI: 0.27 to 0.58) and 0.51 (95% CI: 0.32 to 0.80)] for the first two SOCs above and still significant (P < 0.005) but changed for infections and infestations, from 1.43 (95% CI: 1.02 to 2.00) to 1.16 (95% CI: 0.75 to 1.80). Table 3. Number of SAEsa and Crude Incidence Rate Differences per 1000 Patient-Years and Incidence Rate Ratios for Patients With CO-CP vs AO-CP SOC  CO-CP Crude Rate per 
1000 PY (n)  AO-CP Crude Rate per 
1000 PY (n)  Crude Rate Difference per 
1000 PY  Crude IRR (Approximate 95% CI)  P Value
 (Score-Test)  Blood and lymphatic system disorders  0.0 (0)  0.4 (2)  −0.37  0  0.190  Cardiac disorders  2.6 (12)  3.2 (17)  −0.56  0.82 (0.39–1.72)  0.602  Congenital, familial, and genetic disorders  0.2 (1)  0.2 (1)  0.03  1.16 (0.07–18.62)  0.914  Ear and labyrinth disorders  1.1 (5)  0.6 (3)  0.52  1.94 (0.46–8.12)  0.355  Endocrine disorders  2.4 (11)  3.2 (17)  −0.78  0.75 (0.35–1.61)  0.463  Eye disorders  0.9 (4)  1.5 (8)  −0.62  0.58 (0.18–1.93)  0.371  Gastrointestinal disorders  2.8 (13)  5.2 (28)  −2.39  0.54 (0.28–1.04)  0.063  General disorders and administration site conditions  6.1 (28)  6.5 (35)  −0.45  0.93 (0.57–1.53)  0.779  Hepatobiliary disorders  1.9 (9)  2.6 (14)  −0.65  0.75 (0.32–1.73)  0.496  Infections and infestationsb  16.5 (76)  11.5 (62)  4.93  1.43 (1.02–2.00)  0.037  Injury, poisoning, and procedural complications  4.5 (21)  5.2 (28)  −0.66  0.87 (0.50–1.54)  0.639  Investigations  1.7 (8)  2.4 (13)  −0.69  0.72 (0.30–1.73)  0.456  Metabolic and nutritional disorders  4.3 (20)  4.1 (22)  0.24  1.06 (0.58–1.94)  0.854  Musculoskeletal and connective tissue disorders  3.2 (15)  3.7 (20)  −0.47  0.87 (0.45–1.71)  0.691  Neoplasms benign, malignant, and unspecified (including cysts and polyps)c  7.8 (36)  19.7 (106)  −11.92  0.40 (0.27–0.58)  0.001  Nervous system disorders  12.1 (56)  9.5 (51)  2.64  1.28 (0.87–1.87)  0.203  Pregnancy, puerperal and perinatal conditions  0.9 (4)  0.6 (3)  0.31  1.55 (0.35–6.94)  0.562  Psychiatric disorders  2.2 (10)  2.6 (14)  −0.44  0.83 (0.37–1.87)  0.656  Renal and urinary disorders  1.5 (7)  1.9 (10)  0.15  0.82 (0.31–2.14)  0.678  Reproductive system and breast disorders  1.1 (5)  0.9 (5)  −1.62  1.16 (0.34–4.02)  0.810  Respiratory, thoracic, and mediastinal disorders  1.7 (8)  3.3 (18)  0.31  0.52 (0.23–1.19)  0.114  Skin and subcutaneous tissue disorders  0.9 (4)  0.6 (3)  0.03  1.55 (0.35–6.94)  0.562  Social circumstances  0.2 (1)  0.2 (1)  −5.87  1.16 (0.07–18.62)  0.914  Surgical and medical procedures  4.5 (21)  10.4 (56)  0.06  0.44 (0.26–0.72)  0.001  Vascular disorders  1.7 (17)  1.7 (9)  0.15  1.03 (0.40–2.68)  0.944  Total  (383)  (546)        SOC  CO-CP Crude Rate per 
1000 PY (n)  AO-CP Crude Rate per 
1000 PY (n)  Crude Rate Difference per 
1000 PY  Crude IRR (Approximate 95% CI)  P Value
 (Score-Test)  Blood and lymphatic system disorders  0.0 (0)  0.4 (2)  −0.37  0  0.190  Cardiac disorders  2.6 (12)  3.2 (17)  −0.56  0.82 (0.39–1.72)  0.602  Congenital, familial, and genetic disorders  0.2 (1)  0.2 (1)  0.03  1.16 (0.07–18.62)  0.914  Ear and labyrinth disorders  1.1 (5)  0.6 (3)  0.52  1.94 (0.46–8.12)  0.355  Endocrine disorders  2.4 (11)  3.2 (17)  −0.78  0.75 (0.35–1.61)  0.463  Eye disorders  0.9 (4)  1.5 (8)  −0.62  0.58 (0.18–1.93)  0.371  Gastrointestinal disorders  2.8 (13)  5.2 (28)  −2.39  0.54 (0.28–1.04)  0.063  General disorders and administration site conditions  6.1 (28)  6.5 (35)  −0.45  0.93 (0.57–1.53)  0.779  Hepatobiliary disorders  1.9 (9)  2.6 (14)  −0.65  0.75 (0.32–1.73)  0.496  Infections and infestationsb  16.5 (76)  11.5 (62)  4.93  1.43 (1.02–2.00)  0.037  Injury, poisoning, and procedural complications  4.5 (21)  5.2 (28)  −0.66  0.87 (0.50–1.54)  0.639  Investigations  1.7 (8)  2.4 (13)  −0.69  0.72 (0.30–1.73)  0.456  Metabolic and nutritional disorders  4.3 (20)  4.1 (22)  0.24  1.06 (0.58–1.94)  0.854  Musculoskeletal and connective tissue disorders  3.2 (15)  3.7 (20)  −0.47  0.87 (0.45–1.71)  0.691  Neoplasms benign, malignant, and unspecified (including cysts and polyps)c  7.8 (36)  19.7 (106)  −11.92  0.40 (0.27–0.58)  0.001  Nervous system disorders  12.1 (56)  9.5 (51)  2.64  1.28 (0.87–1.87)  0.203  Pregnancy, puerperal and perinatal conditions  0.9 (4)  0.6 (3)  0.31  1.55 (0.35–6.94)  0.562  Psychiatric disorders  2.2 (10)  2.6 (14)  −0.44  0.83 (0.37–1.87)  0.656  Renal and urinary disorders  1.5 (7)  1.9 (10)  0.15  0.82 (0.31–2.14)  0.678  Reproductive system and breast disorders  1.1 (5)  0.9 (5)  −1.62  1.16 (0.34–4.02)  0.810  Respiratory, thoracic, and mediastinal disorders  1.7 (8)  3.3 (18)  0.31  0.52 (0.23–1.19)  0.114  Skin and subcutaneous tissue disorders  0.9 (4)  0.6 (3)  0.03  1.55 (0.35–6.94)  0.562  Social circumstances  0.2 (1)  0.2 (1)  −5.87  1.16 (0.07–18.62)  0.914  Surgical and medical procedures  4.5 (21)  10.4 (56)  0.06  0.44 (0.26–0.72)  0.001  Vascular disorders  1.7 (17)  1.7 (9)  0.15  1.03 (0.40–2.68)  0.944  Total  (383)  (546)        System Organ Class data are derived from the Medical Dictionary for Regulatory Activities. Significant P values are shown in bold. Abbreviations: IRR, incidence rate ratio; PY, patient-year. a First reported SAE in a patient on preferred term level (Medical Dictionary for Regulatory Activities preferred term) was counted. b With adjustment for age and sex (and to first event per patient on preferred term level), the IRR was 1.16 (95% CI: 0.75–1.80; P = 0.51). c With adjustment for age and sex (and to first event per patient on preferred term level), the IRR was 0.51 (95% CI: 0.32–0.80; P = 0.004). View Large Discussion CPs have consistently been associated with higher morbidity and mortality than have other pituitary disorders, which cannot be attributed solely to endocrine abnormalities (7). Despite advances in surgical technique and adjuvant radiotherapy (3), adults with CP continue to have excess morbidity and mortality, for which the RRs of contributing factors remain unclear. In this study, we assessed the RRs of a range of hypothesized factors that may contribute to increased morbidity and mortality of patients with CP by examining data from a large cohort of such patients enrolled in KIMS. Our results demonstrated that conventional factors that increase the risks of cardio- and cerebrovascular diseases and DM in the general population (e.g., history of stroke, hyperlipidemia, increasing BMI, and higher baseline hemoglobin A1c levels) conferred greater risks in patients with CP. In addition, the increased risk of recurrence of CP was associated with a higher number of surgeries and adult-onset of the disease, whereas previous radiotherapy and increasing age for every 10 years from disease onset conferred increased risks for developing new intracranial and extracranial tumors, respectively. Furthermore, CP and/or its treatment regimen increased the risks of developing DM and intracranial tumors and early death, whereas lower serum IGF-I levels measured at last sampling before death or at the corresponding time from baseline in patients who were still alive was associated with increased all-cause mortality. When adults with CO-CP were compared with those with AO-CP at KIMS entry, there were more CO-CP patients who underwent both surgery (and more transcranial surgeries) and radiotherapy, implying that larger and more aggressive CP tumors tended to present during childhood. These patients also had lower BMIs and waist circumference measurements and better blood pressure levels, hemoglobin A1c levels, and lipid profiles, suggesting more favorable body composition and cardiometabolic status. In line with previous findings (30), patients with CO-CP had lower baseline IGF-I SDS than patients with AO-CP, implying that patients with CO-CP may be more GH deficient. In addition, patients with CO-CP had lower IGF-I SDS values at 1 year despite being treated with higher GH doses. The reasons for this observation are unclear and may include a lack of adherence to daily GH injections by patients with CO-CP or the inability of the younger cohort with CO-CP to achieve optimal GH replacement dose after 1 year. By contrast, the less favorable cardiometabolic status in patients with AO-CP may be exacerbated by the fact that there were more smokers and patients who had a history of DM, metabolic syndrome, cardio- and cerebrovascular disease, hyperlipidemia, and cancer. These findings are also in keeping with those reported by Kendall-Taylor et al. (30) in a smaller cohort of patients with CO-CP and AO-CP who were also from the KIMS database and with findings of two older studies comparing adult GHD of mixed etiologies (31, 32). In this study, history of stroke and hyperlipidemia conferred approximately sixfold and threefold higher risks of cardiovascular disease, respectively, whereas increasing BMI increased the risk of cerebrovascular disease by approximately twofold. The infrequent prescription of antihypertensive and statin therapy offers a plausible explanation and may be reflected by the relatively young age of the CP population, particularly patients with CO-CP. Thus, our data suggest that history of stroke, hyperlipidemia, and higher and/or increasing BMI in adults with either CO-CP or AO-CP are important contributors to increased risk of cardio- and cerebrovascular diseases. In this study, incidence rates for DM in Sweden were used as a reference because the majority of patients in KIMS were from Europe and these rates have been reported to be comparable to those of other European populations (33, 34). Similar to patients without hypothalamic-pituitary diseases, adults with CO-CP and AO-CP with higher baseline hemoglobin A1c levels and greater and/or increasing waist circumference had an increased risk of developing new DM. Interestingly, females had higher rates of developing DM at the start of KIMS entry; however, this risk decreased over time: after ~1.5 to 2.5 years, to rates seen in males, and after 5 to 6 years, to lower rates than males. These data are consistent with those previously reported from KIMS that showed adults with GHD from a variety of etiologies apart from CP demonstrated a greater incidence of DM, with a female preponderance that was thought to be due to abnormal body composition (33) but an overall incidence of DM after GH replacement that was similar between sexes (21). The reason for the female propensity to develop DM before GH is initiated is unclear and may be attributable to a different expression of GHD with higher visceral adiposity and lower physical activity, in keeping with data previously reported in GH-treated patients with mixed etiologies of hypopituitarism (35). It is well known that GH has both mitogenic and antiapoptotic effects (36) and that these actions have led to concerns that long-term GH replacement may increase recurrence rates of CP, although previous studies have indicated that GH replacement is safe (37, 38). Our findings provide further evidence that long-term GH replacement in adults with CP neither facilitated tumor recurrence nor induced any SAEs. Our study also found that the risk of recurrence of CP was decreased in females and in those with previous radiotherapy exposure and prior surgery but increased in the setting of a greater number of surgeries and the onset of CP in adulthood. Knowledge of these factors is important for the physician when treatment regimens are being formulated for these patients. In their recent study involving a large cohort of patients in KIMS with hypopituitarism due to pituitary/sellar lesions, including CP, who were followed up for a median of 5.3 years, Burman et al. (39) found that radiotherapy increased the risk of malignant brain tumors and meningiomas, particularly when treated at younger ages. That same trend of radiotherapy conferring an increased risk of new intracranial tumors was also observed in the current study, although it was not as statistically significant as in the aforementioned study, which may be attributed to a smaller cohort of patients with CP studied. Older age with every 10 years from disease onset might also be a risk for developing new extracranial tumors, but this risk did not exceed the risk of extracranial tumors in the reference population, implying the need for longer-term studies. Taken together, our findings and those of Burman et al. (39) further reinforce the notion that radiotherapy should be used more cautiously. In three previous studies comprising 485 patients with CP (4, 6, 10), the mortality rate was increased between fourfold and ninefold. This observation was initially thought to be mainly due to cardio- and cerebrovascular diseases (4, 6), although hypothalamic damage and its treatment regimen(s) (11), hypopituitarism with inadequately treated hypocortisolism during stress (13), and insufficient GH replacement (14) have more recently been proposed as important additional contributing factors. By contrast, our study demonstrated a lower all-cause mortality rate of 2.2-fold compared with rates of older studies (4, 6, 10) (with cardiovascular and cerebrovascular mortality observed mainly in patients with CO-CP and AO-CP, respectively). This difference may be attributable to recent refinements in surgical techniques, more physiological utilization of glucocorticoid and estrogen replacement doses, and observational bias with closer surveillance of patients on GH replacement in KIMS. Our analysis also found that greater and/or increasing BMI and history of stroke further increased the risk of early death. In addition, because patients with CP had a significantly higher risk of developing new DM and new intracranial tumors after radiotherapy compared with the general population (39), these risk factors may also play a key role in exacerbating the excess mortality of these patients. Furthermore, the risk of developing DM, intracranial tumors, and early death compared with the general population is increased because of having CP and its treatment-related complications. These patients should therefore be considered as having a chronic disease, and ongoing follow-up addressing these comorbidities should be performed beginning at disease onset. We analyzed IGF-I SDS as a TD covariate, where IGF-I SDS measured at last sampling before death or at the corresponding time from baseline for those still alive was used in the risk calculation of all-cause mortality, and we found that lower IGF-I SDS was related to increased all-cause mortality. However, it is noteworthy that the time between the event and the last IGF-I measurement may vary between patients because of the observational nature of how KIMS data were collected. The reasons for these findings are unclear. It is possible that lack of adherence to daily GH injections, reduction in GH responsiveness, or patients discontinuing GH nearing their terminal phase resulted in lower IGF-I SDS before death. Because this was a retrospective study, several limitations are acknowledged. First, the KIMS database was not designed to provide detailed individual information about the degree of tumor invasiveness and the extent of surgical resection, details on the total and fractionated radiotherapy doses administered, and the severity of hypopituitarism. Thus, the role and adequacy of glucocorticoid, sex hormone, and thyroid hormone replacement therapies on the comorbidities cannot be elucidated in this study. Second, we cannot exclude the possibility of patient infrequent use of and/or nonadherence to daily GH injections. Third, given that the patients were enrolled between 1994 and 2012, serum IGF-I levels were measured by several different assays, which could have produced variable results because of the use of different populations to establish reference values. Therefore, for the comparison of values obtained with different assays, IGF-I SDS was used with reference to the normative data for the assay in question, after appropriate transformation for nonnormality data. Although these limitations are inherent to any registry-based study, the large number of adults with CO-CP and AO-CP in our series followed up on long-term GH replacement strengthens our findings. In conclusion, we demonstrated that conventional factors that increase the risks of cardio- and cerebrovascular diseases and DM, risks for developing new intracranial tumors after radiotherapy, and lower IGF-I SDS measured at last sampling before death or at the corresponding time point from baseline for those still alive contributed to the increased morbidity and mortality of patients with CP. Close follow-up is therefore imperative for early detection of tumor recurrence in irradiated patients, and early preventive management of modifiable cardiovascular risks is important to potentially decrease the mortality of these patients. Abbreviations: AE adverse event AO-CP adult-onset craniopharyngioma BMI body mass index CI confidence interval CO-CP childhood-onset craniopharyngioma CP craniopharyngioma DM diabetes mellitus E expected number of cases GH growth hormone GHD growth hormone deficiency IGF-I insulinlike growth factor I O observed number of cases RR relative risk SAE serious adverse event SOC System Organ Class TD time-dependent. Acknowledgments The authors thank all the KIMS investigators and patients who consented to providing data. Financial Support: This research was not supported by any specific grant from any funding agency in the public, commercial, or not-for-profit sector. The KIMS database is sponsored by Pfizer Inc. Current Affiliation: K.C.J. Yuen’s current affiliation is Barrow Pituitary Center, Departments of Neuroendocrinology and Neurosurgery, Barrow Neurological Institute, Phoenix, Arizona 85013. Disclosure Summary: K.C.J.Y., P.B., and R.A. are current members of the KIMS Steering Committee. K.C.J.Y. has received research grants from Pfizer, Novo Nordisk, Teva Pharmaceuticals, OPKO Biologics, and Versartis and has served on advisory boards for Pfizer, Novo Nordisk, Sandoz, and Versartis. A.F.M., J.L.F., and C.C.-H. are employed by Pfizer, Inc. All statistical analyses were performed by a statistician (A.F.M.). P.B. has received research grants from Novartis and speaker fees from Ipsen and Novo Nordisk and has served on advisory boards for Pfizer. J.V. has received honoraria for lectures and/or advisory boards from Pfizer, Ipsen, and Novartis. M.E.G. was a KIGS investigator and a member of the KIGS International Board, is currently on advisory boards for Abbvie, Diurnal, Endo, Pfizer, Millendo Therapeutics, Novo Nordisk, Ipsen, and Sandoz, and has a research contract from Versartis. E.-M.E. has nothing to disclose. References 1. Bunin GR, Surawicz TS, Witman PA, Preston-Martin S, Davis F, Bruner JM. The descriptive epidemiology of craniopharyngioma. J Neurosurg . 1998; 89( 4): 547– 551. Google Scholar CrossRef Search ADS PubMed  2. Müller HL. Consequences of craniopharyngioma surgery in children. J Clin Endocrinol Metab . 2011; 96( 7): 1981– 1991. Google Scholar CrossRef Search ADS PubMed  3. Clark AJ, Cage TA, Aranda D, Parsa AT, Auguste KI, Gupta N. Treatment-related morbidity and the management of pediatric craniopharyngioma: a systematic review. J Neurosurg Pediatr . 2012; 10( 4): 293– 301. Google Scholar CrossRef Search ADS PubMed  4. Bülow B, Attewell R, Hagmar L, Malmström P, Nordström CH, Erfurth EM. Postoperative prognosis in craniopharyngioma with respect to cardiovascular mortality, survival, and tumor recurrence. J Clin Endocrinol Metab . 1998; 83( 11): 3897– 3904. Google Scholar PubMed  5. Olsson DS, Andersson E, Bryngelsson IL, Nilsson AG, Johannsson G. Excess mortality and morbidity in patients with craniopharyngioma, especially in patients with childhood onset: a population-based study in Sweden. J Clin Endocrinol Metab . 2015; 100( 2): 467– 474. Google Scholar CrossRef Search ADS PubMed  6. Tomlinson JW, Holden N, Hills RK, Wheatley K, Clayton RN, Bates AS, Sheppard MC, Stewart PM; West Midlands Prospective Hypopituitary Study Group. Association between premature mortality and hypopituitarism. Lancet . 2001; 357( 9254): 425– 431. Google Scholar CrossRef Search ADS PubMed  7. Crowley RK, Hamnvik OP, O’Sullivan EP, Behan LA, Smith D, Agha A, Thompson CJ. Morbidity and mortality in patients with craniopharyngioma after surgery. Clin Endocrinol (Oxf) . 2010; 73( 4): 516– 521. Google Scholar PubMed  8. Fjalldal S, Holmer H, Rylander L, Elfving M, Ekman B, Osterberg K, Erfurth EM. Hypothalamic involvement predicts cognitive performance and psychosocial health in long-term survivors of childhood craniopharyngioma. J Clin Endocrinol Metab . 2013; 98( 8): 3253– 3262. Google Scholar CrossRef Search ADS PubMed  9. Gautier A, Godbout A, Grosheny C, Tejedor I, Coudert M, Courtillot C, Jublanc C, De Kerdanet M, Poirier JY, Riffaud L, Sainte-Rose C, Van Effenterre R, Brassier G, Bonnet F, Touraine P; Craniopharyngioma Study Group. Markers of recurrence and long-term morbidity in craniopharyngioma: a systematic analysis of 171 patients. J Clin Endocrinol Metab . 2012; 97( 4): 1258– 1267. Google Scholar CrossRef Search ADS PubMed  10. Pereira AM, Schmid EM, Schutte PJ, Voormolen JH, Biermasz NR, van Thiel SW, Corssmit EP, Smit JW, Roelfsema F, Romijn JA. High prevalence of long-term cardiovascular, neurological and psychosocial morbidity after treatment for craniopharyngioma. Clin Endocrinol (Oxf) . 2005; 62( 2): 197– 204. Google Scholar CrossRef Search ADS PubMed  11. Yuen KC, Kołtowska-Häggström M, Cook DM, Fox JL, Jönsson PJ, Geffner ME, Abs R. Primary treatment regimen and diabetes insipidus as predictors of health outcomes in adults with childhood-onset craniopharyngioma. J Clin Endocrinol Metab . 2014; 99( 4): 1227– 1235. Google Scholar CrossRef Search ADS PubMed  12. Bülow B, Hagmar L, Eskilsson J, Erfurth EM. Hypopituitary females have a high incidence of cardiovascular morbidity and an increased prevalence of cardiovascular risk factors. J Clin Endocrinol Metab . 2000; 85( 2): 574– 584. Google Scholar PubMed  13. Burman P, Mattsson AF, Johannsson G, Höybye C, Holmer H, Dahlqvist P, Berinder K, Engström BE, Ekman B, Erfurth EM, Svensson J, Wahlberg J, Karlsson FA. Deaths among adult patients with hypopituitarism: hypocortisolism during acute stress, and de novo malignant brain tumors contribute to an increased mortality. J Clin Endocrinol Metab . 2013; 98( 4): 1466– 1475. Google Scholar CrossRef Search ADS PubMed  14. Gaillard RC, Mattsson AF, Akerblad AC, Bengtsson BA, Cara J, Feldt-Rasmussen U, Koltowska-Häggström M, Monson JP, Saller B, Wilton P, Abs R. Overall and cause-specific mortality in GH-deficient adults on GH replacement. Eur J Endocrinol . 2012; 166( 6): 1069– 1077. Google Scholar CrossRef Search ADS PubMed  15. Olsson DS, Trimpou P, Hallén T, Bryngelsson IL, Andersson E, Skoglund T, Bengtsson BA, Johannsson G, Nilsson AG. Life expectancy in patients with pituitary adenoma receiving growth hormone replacement. Eur J Endocrinol . 2017; 176( 1): 67– 75. Google Scholar CrossRef Search ADS PubMed  16. van Bunderen CC, van Nieuwpoort IC, Arwert LI, Heymans MW, Franken AA, Koppeschaar HP, van der Lely AJ, Drent ML. Does growth hormone replacement therapy reduce mortality in adults with growth hormone deficiency? Data from the Dutch National Registry of Growth Hormone Treatment in adults. J Clin Endocrinol Metab . 2011; 96( 10): 3151– 3159. Google Scholar CrossRef Search ADS PubMed  17. Pappachan JM, Raskauskiene D, Kutty VR, Clayton RN. Excess mortality associated with hypopituitarism in adults: a meta-analysis of observational studies. J Clin Endocrinol Metab . 2015; 100( 4): 1405– 1411. Google Scholar CrossRef Search ADS PubMed  18. Cook DM, Yuen KC, Biller BM, Kemp SF, Vance ML; American Association of Clinical Endocrinologists. American Association of Clinical Endocrinologists medical guidelines for clinical practice for growth hormone use in growth hormone-deficient adults and transition patients: 2009 update. Endocr Pract . 2009; 15( Suppl 2): 1– 29. Google Scholar CrossRef Search ADS PubMed  19. Molitch ME, Clemmons DR, Malozowski S, Merriam GR, Vance ML; Endocrine Society. Evaluation and treatment of adult growth hormone deficiency: an Endocrine Society clinical practice guideline. J Clin Endocrinol Metab . 2011; 96( 6): 1587– 1609. Google Scholar CrossRef Search ADS PubMed  20. American Diabetes Association. 2. Classification and diagnosis of diabetes. Diabetes Care . 2016; 39( Suppl 1): S13– S22. CrossRef Search ADS PubMed  21. Luger A, Mattsson AF, Koltowska-Häggström M, Thunander M, Góth M, Verhelst J, Abs R. Incidence of diabetes mellitus and evolution of glucose parameters in growth hormone-deficient subjects during growth hormone replacement therapy: a long-term observational study. Diabetes Care . 2012; 35( 1): 57– 62. Google Scholar CrossRef Search ADS PubMed  22. Alberti KG, Zimmet P, Shaw J. Metabolic syndrome--a new world-wide definition: a consensus statement from the International Diabetes Federation. Diabet Med . 2006; 23( 5): 469– 480. Google Scholar CrossRef Search ADS PubMed  23. Verhelst J, Mattsson AF, Luger A, Thunander M, Góth MI, Koltowska-Häggström M, Abs R. Prevalence and characteristics of the metabolic syndrome in 2479 hypopituitary patients with adult-onset GH deficiency before GH replacement: a KIMS analysis. Eur J Endocrinol . 2011; 165( 6): 881– 889. Google Scholar CrossRef Search ADS PubMed  24. Brabant G, von zur Mühlen A, Wüster C, Ranke MB, Kratzsch J, Kiess W, Ketelslegers JM, Wilhelmsen L, Hulthén L, Saller B, Mattsson A, Wilde J, Schemer R, Kann P; German KIMS Board. Serum insulin-like growth factor I reference values for an automated chemiluminescence immunoassay system: results from a multicenter study. Horm Res . 2003; 60( 2): 53– 60. Google Scholar PubMed  25. Hughes AO, Hills M. Statistical Models in Epidemiology . Oxford, UK: Oxford University Press; 1993. 26. Breslow NE, Day NE. Statistical Methods in Cancer Research. Vol. 2. The Design and Analysis of Cohort Studies.  IARC Scientific Publications No. 82. Lyon, France: International Agency for Research on Cancer; 1987. 27. Curado MP, Edwards B, Shin HR, Storm H, Ferlay J, Heanue M, Boyle P. Cancer Incidence in Five Continents. Vol. 9. No. 160 . Lyon, France: International Agency for Research on Cancer; 2007. 28. Thunander M, Petersson C, Jonzon K, Fornander J, Ossiansson B, Torn C, Edvardsson S, Landin-Olsson M. Incidence of type 1 and type 2 diabetes in adults and children in Kronoberg, Sweden. Diabetes Res Clin Pract . 2008; 82( 2): 247– 255. Google Scholar CrossRef Search ADS PubMed  29. World Health Organization. World health statistics 2008. Available at: http://www.who.int/gho/publications/world_health_statistics/EN_WHS08_Full.pdf. Accessed 31 October 2017. 30. Kendall-Taylor P, Jönsson PJ, Abs R, Erfurth EM, Koltowska-Häggström M, Price DA, Verhelst J. The clinical, metabolic and endocrine features and the quality of life in adults with childhood-onset craniopharyngioma compared with adult-onset craniopharyngioma. Eur J Endocrinol . 2005; 152( 4): 557– 567. Google Scholar CrossRef Search ADS PubMed  31. Attanasio AF, Bates PC, Ho KK, Webb SM, Ross RJ, Strasburger CJ, Bouillon R, Crowe B, Selander K, Valle D, Lamberts SW; Hypopituitary Control and Complications Study International Advisory Board. Human growth hormone replacement in adult hypopituitary patients: long-term effects on body composition and lipid status--3-year results from the HypoCCS Database. J Clin Endocrinol Metab . 2002; 87( 4): 1600– 1606. Google Scholar PubMed  32. Koranyi J, Svensson J, Götherström G, Sunnerhagen KS, Bengtsson B, Johannsson G. Baseline characteristics and the effects of five years of GH replacement therapy in adults with GH deficiency of childhood or adulthood onset: a comparative, prospective study. J Clin Endocrinol Metab . 2001; 86( 10): 4693– 4699. Google Scholar CrossRef Search ADS PubMed  33. Abs R, Mattsson AF, Thunander M, Verhelst J, Góth MI, Wilton P, Kołtowska-Häggström M, Luger A. Prevalence of diabetes mellitus in 6050 hypopituitary patients with adult-onset GH deficiency before GH replacement: a KIMS analysis. Eur J Endocrinol . 2013; 168( 3): 297– 305. Google Scholar CrossRef Search ADS PubMed  34. Andersson T, Ahlbom A, Carlsson S. Diabetes prevalence in Sweden at present and projections for year 2050. PLoS One . 2015; 10( 11): e0143084. Google Scholar CrossRef Search ADS PubMed  35. Holmer H, Svensson J, Rylander L, Johannsson G, Rosén T, Bengtsson BA, Thorén M, Höybye C, Degerblad M, Bramnert M, Hägg E, Edén Engström B, Ekman B, Norrving B, Hagmar L, Erfurth EM. Nonfatal stroke, cardiac disease, and diabetes mellitus in hypopituitary patients on hormone replacement including growth hormone. J Clin Endocrinol Metab . 2007; 92( 9): 3560– 3567. Google Scholar CrossRef Search ADS PubMed  36. Bogazzi F, Ultimieri F, Raggi F, Russo D, Vanacore R, Guida C, Brogioni S, Cosci C, Gasperi M, Bartalena L, Martino E. Growth hormone inhibits apoptosis in human colonic cancer cell lines: antagonistic effects of peroxisome proliferator activated receptor-gamma ligands. Endocrinology . 2004; 145( 7): 3353– 3362. Google Scholar CrossRef Search ADS PubMed  37. Karavitaki N, Warner JT, Marland A, Shine B, Ryan F, Arnold J, Turner HE, Wass JA. GH replacement does not increase the risk of recurrence in patients with craniopharyngioma. Clin Endocrinol (Oxf) . 2006; 64( 5): 556– 560. Google Scholar CrossRef Search ADS PubMed  38. Olsson DS, Buchfelder M, Wiendieck K, Kremenevskaja N, Bengtsson BA, Jakobsson KE, Jarfelt M, Johannsson G, Nilsson AG. Tumour recurrence and enlargement in patients with craniopharyngioma with and without GH replacement therapy during more than 10 years of follow-up. Eur J Endocrinol . 2012; 166( 6): 1061– 1068. Google Scholar CrossRef Search ADS PubMed  39. Burman P, van Beek AP, Biller BM, Camacho-Hübner C, Mattsson AF. Radiotherapy, especially at young age, increases the risk for de novo brain tumors in patients treated for pituitary/sellar lesions. J Clin Endocrinol Metab . 2017; 102( 3): 1051– 1058. Google Scholar PubMed  Copyright © 2018 Endocrine Society

Journal

Journal of Clinical Endocrinology and MetabolismOxford University Press

Published: Feb 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 12 million articles from more than
10,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Unlimited reading

Read as many articles as you need. Full articles with original layout, charts and figures. Read online, from anywhere.

Stay up to date

Keep up with your field with Personalized Recommendations and Follow Journals to get automatic updates.

Organize your research

It’s easy to organize your research with our built-in tools.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

Monthly Plan

  • Read unlimited articles
  • Personalized recommendations
  • No expiration
  • Print 20 pages per month
  • 20% off on PDF purchases
  • Organize your research
  • Get updates on your journals and topic searches

$49/month

Start Free Trial

14-day Free Trial

Best Deal — 39% off

Annual Plan

  • All the features of the Professional Plan, but for 39% off!
  • Billed annually
  • No expiration
  • For the normal price of 10 articles elsewhere, you get one full year of unlimited access to articles.

$588

$360/year

billed annually
Start Free Trial

14-day Free Trial