| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Departments of Endocrinology (K.L., E.M.E., B.A.), Pediatrics (C.M., S.G.), Oncology (E.C.-S.), and Cardiology (U.T.) and Competence Centre for Clinical Research (J.B.), Lund University Hospital, SE 221 85 Lund, Sweden
Address all correspondence and requests for reprints to: Eva Marie Erfurth, Department of Endocrinology, Lund University Hospital, SE-221 85 Lund, Sweden. E-mail: eva_marie.erfurth{at}med.lu.se.
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsDiscussionReferences |
|---|
The aim of the present study was to elucidate mechanisms of increased cardiovascular risk in former childhood ALL patients. A group of 44 ALL survivors (23 males, median age 25 yr, range 19–32 yr at the time of study) treated with CRT (median 24 Gy, 18–30 Gy) at a median age of 5 yr (1–18 yr) and chemotherapy were investigated for prevalence of GH deficiency and cardiovascular risk factors. Comparison was made with controls randomly selected from the general population and individually matched for sex, age, smoking habits, and residence. All patients and controls underwent a GHRH-arginine test, and patients with a peak GH 3.9 µg/liter or greater were further investigated with an additional insulin tolerance test.
Significantly higher plasma levels of insulin (P = 0.002), blood glucose (P = 0.01), and serum levels of low-density lipoprotein cholesterol, apolipoprotein (Apo) B, triglycerides, fibrinogen, and leptin (all P 
0.05) were recorded among the ALL patients, compared with controls. Furthermore, the serum levels of high-density lipoprotein cholesterol (P = 0.03) and Apo A1 (P = 0.005) were significantly lower among the patients. Compared with controls, the patients had higher body mass index and waist to hip ratio, and body composition measured with dual-energy x-ray absorptiometry showed significantly higher fat mass and lower lean mass (P < 0.001). Forty of 44 ALL patients (91%) were considered GH deficient according to the insulin tolerance test and/or the GHRH-arginine test, and the rest were considered GH insufficient.
In patients, peak GH during GHRH-arginine was significantly negatively correlated to total body fat mass measured with dual-energy x-ray absorptiometry (r = –0.48, P = 0.001), waist to hip ratio (r = –0.32, P = 0.03), plasma insulin (r = –0.49, P = 0.001), and leptin (r = –0.46, P = 0.002). Moreover, a significantly positive correlation was recorded with high-density lipoprotein cholesterol (r = 0.38, P = 0.012). Using Doppler echocardiography, a marked reduction in cardiac dimensions and performance (ejection fraction P < 0.001 and fractional shortening P = 0.01), compared with controls, was recorded.
In conclusion, at a median 17 yr after treatment with CRT and chemotherapy in former childhood ALL patients, a significant increase in cardiovascular risk factors was recorded. We suggest that GH deficiency, induced by CRT, is a primary cause for this because strong correlations between the stimulated GH peak and several of the cardiovascular risk factors were observed.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
The mechanisms behind the increase in cardiovascular risk in survivors of childhood cancer are not clarified, and several possibilities exist. Treatment-induced hypogonadism, eventually leading to cellular atrophy in the target organ, was associated with hyperinsulinemia in this patient group (7). Moreover, young adults surviving childhood cancer had significantly more manifestations of the metabolic syndrome, with reduced spontaneous GH secretion (5); patients with GH deficiency due to primary hypothalamic or pituitary disease are characterized by increased fat mass, dyslipidemia, and insulin resistance (8). Another possibility is an alteration in the regulation of leptin; elevated serum levels of leptin are found in GH-deficient patients treated with CRT for ALL (9, 10), which may be due to either radiation-induced hypothalamic damage causing leptin insensitivity or GH deficiency. A problem in previous studies of ALL patients has been the inclusion of heterogeneous groups with regard to age at treatment and treatment regimens (with and without CRT), which made interpretation of causal relations difficult due to low statistical power. Furthermore, the previous study of ALL survivors had a rather short follow-up and no control group, which limits interpretation of the results (6).
The aim of the present study was to further elucidate mechanisms of increased cardiovascular risk in former childhood ALL patients, using an appropriate control group. Therefore, a group of 44 ALL survivors, treated at a single department (Children’s Hospital Lund), who were at least 18 yr old at the time of study and treated with CRT and chemotherapy, were investigated for prevalence of GH deficiency and cardiovascular risk factors. The ALL patients were compared with controls randomly selected from the general population and individually matched for sex, age, smoking habits, and residence.
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
A consecutive series of 58 patients treated for childhood ALL with chemotherapy and CRT during 1971–1992 at the Children’s Hospital Lund and at least 18 yr of age were invited to participate. Fourteen patients were excluded for various reasons: seven patients declined to participate due to lack of time or fear of more hospital visits, two were treated for severe epilepsy with uncontrolled seizures, one was pregnant, one was breast-feeding, one had recently had surgery for a brain tumor, one was on treatment with GH, and one had emigrated. Patient characteristics, including doses of radiotherapy and anthracycline, and ages at diagnosis, treatment, and present investigation are shown in Table 1
. All patients had been off all chemotherapy for a median of 16.7 yr (range 6.3–23.9 yr).
|
). In the period 1971–1981, all patients diagnosed with ALL received CRT. In the latter period (1982–1992), children belonging to the lower-risk groups received no cranial radiation and were not eligible for the present study; remission was induced with prednisone, vincristine, and, in some patients, doxorubicin and asparaginase. When consolidation therapy was given, it consisted of intermediate-dose methotrexate with citrovorum factor rescue or cyclophosphamide and cytosine arabinoside. Central nervous system therapy included intrathecal methotrexate and prophylactic CRT with 18–24 Gy; in cases of central nervous system involvement, the dose was 30 Gy. Maintenance therapy included 6-mercaptopurine and methotrexate; in some patients cyclophosphamide was also given. In higher-risk groups reinductions with prednisolone and vincristine and/or doxorubicin and intermediate-dose methotrexate were applied. Five patients had received CRT in combination with spinal radiation (median 23.0 Gy, range 14–25 Gy) (Table 2). Twelve boys were given radiation to the testis (Table 1). Five patients had been treated for relapse and one patient for more than one relapse. One patient was treated with high-dose chemotherapy with autologous stem cell rescue.
|
) as injections (n = 7), tablets (n = 1), or patches (n = 2). Two patients subjected to testis radiation had no androgen substitution; one had not been properly irradiated due to an undescended testis and subsequently had normal testosterone levels, and the other was not sufficiently substituted. One patient suffered from heart failure and hypertension, treated with digoxin and enalapril maleate; another had hypertension treated with atenolol and enalapril maleate; and two female patients received sertraline hydrochloride for depression. Three men and two women were smokers. Control subjects and study design
The aim was to select one control subject for each patient enrolled in the study. To obtain this, 10 potential control subjects matched for age, gender, and residence (rural/nonrural) were selected randomly from a computerized register of the population in the catchment area of the patients (Southern Swedish Medical Region). Potential controls were contacted by telephone and then also matched for smoking. The first eligible control that agreed to participate in the study was chosen. If none of the 10 selected control subjects accepted, a new set of 10 controls were selected, and this process was repeated until appropriate controls for all patients were chosen. Thirty-six of 81 potential controls declined to participate when contacted, including six due to illness and two because of ongoing breast-feeding.
The present investigation was performed before any change in medication. For each patient and control, except for the insulin tolerance test, all investigations were performed on a single day. The Ethics Committee of Lund University approved the protocol, and all subjects gave written informed consent.
Test procedures for GH secretion
All patients and controls underwent a GHRH-arginine test in the morning according to the method of Ghigo et al. (14). Patients with a peak GH 3.9 µg/liter or greater were further investigated with an additional ITT, except in one patient, who refused this test. The two tests were performed on different occasions with at least 1 month between. The ITT was performed after an overnight fast with 0.1 U/kg iv soluble insulin (Actrapid, Novo Nordisk A/S, Bagsvaerd, Denmark). Hypoglycemia was obtained with blood glucose levels less than 2.2 mmol/liter in all but one patient, who had a GH response greater than 3.9 µg/liter but initial blood glucose level lowered by more than 50%. Previously defined cut-off levels, based on the GH assay used in the present study, were used to determine GH status (15). Severe GH deficiency was defined as a peak GH response less than 3 µg/liter to an ITT (15, 16) and less than 9 µg/liter during the GHRH-arginine test (14). GH insufficiency was defined as a peak GH response between 3 and 5 µg/liter during the ITT and between 9 and 16.5 µg/liter during the GHRH-arginine test (14). However, when the GH response to the GHRH-arginine test was greater than 9 µg/liter and the ITT in the same patient was less than 3 µg/liter, the ITT result determined the GH deficiency classification, consistent with Darzy et al. (17).
Anthropological measurements and blood pressure
Waist circumference was measured at the midpoint between the lower rib margin and the iliac crest and hip circumference at the level of the trochanters, enabling the calculation of the waist to hip ratio (WHR). Body mass index (BMI) was calculated as body weight (kilograms) divided by height (meters) squared. Body composition was assessed with dual-energy x-ray absorptiometry (DXA) (Expert XL, Lunar Corp., Madison, WI). Data are expressed as estimated fat tissue, lean tissue, and total body content (kilograms). Body composition was also measured in the supine position by bioelectric impedance analysis using the bioelectrical impedance assay (BIA) 101-S technique (RJL Systems, Detroit, MI), with a 50-KHz, 800-µA current. Data are expressed as percentages of fat and muscle, respectively. Blood pressure was measured in the right arm in the supine position after 10 min of rest. The mean value of two measurements was used.
Carotid artery ultrasound
A high-resolution four-duplex ultrasound system (128 XP and Sequoia 512, Acuson, Mountain View, CA), with a 7.5-MHz linear ultrasound probe, was used to measure intima-media thickness and plaque of the carotid arteries of the far wall. Measurements, among 29 patients with the longest follow-up and their matched controls were performed blinded by two skilled operators. Differences in measurements between the two operators were less than 0.02 mm. Details on the technique have been reported elsewhere (18).
Doppler echocardiography
Transthoracic echocardiography (Sonos 5500, Hewlett Packard, Portland, OR) was performed blinded by one of three experienced echocardiographers. All measurements were averaged on the basis of three consecutive heart beats (sinus rhythm). Heart rate was obtained from the R-R-interval (heart rate = 60 x 1000/R-R interval in milliseconds) measured from the screen. Cardiac dimensions, areas, volumes, and mass were indexed to body surface area (BSA), which was calculated from height (in centimeters) and weight (in kilograms) by the formula: BSA (in square meters) = (height + weight – 60)/100. M-mode dimensions were obtained from a parasternal view in a standard fashion according to the American Society of Echocardiography. Left ventricular end-diastolic (LVIDd) and end-systolic (LVIDs) cavity dimension as well as interventricular septal wall thickness (IVSd), left ventricular posterior wall thickness (LVPWd), and left atrial dimensions were measured. Left ventricular fractional shortening (FS) and ejection fraction (LVEF) were used as indices of left ventricular systolic function and calculated as: FS = [(LVIDd – LVIDs)/LVIDd] x 100 and LVEF = [(LVIDd3 – LVIDs3)/LVIDd3] x 100. Left ventricular mass (LV-mass) was calculated as: 1.04 (IVS + LVID + LVPW)3 – LVID3) – 13.6 (19). From an apical four-chamber view, the largest dimensions and areas of all heart chambers during the cardiac cycle were obtained (end diastole for the ventricles and end systole for the atria). Left ventricular systolic and diastolic functions were also assessed by measuring the movement of the atrioventricular plane from an apical acoustic window. The systolic amplitude and the maximal diastolic deflection slope were registered at four points (anterior, posterior, septal, lateral), and the average value was taken. The amplitude of the systolic atrioventricular movement was indexed to BSA.
Peak early (E-wave) and peak late diastolic filling (A-wave) velocities of mitral flow were measured, and the E/A ratio and the time required for the E velocity to decline from its peak to its baseline (deceleration time) were calculated. The systolic contribution to pulmonary venous flow was calculated as: systolic flow velocity/(systolic + diastolic flow velocity). A trivial tricuspid regurgitation was present in many but not all of the subjects, allowing the calculation of the systolic right ventricular/right atrial gradient. Apart from the parameters described above, the echocardiographic examination also comprised a complete morphological and functional evaluation of all the valves and heart chambers (data not shown).
Questionnaire of physical exercise
The degree of physical exercise during spare time and working time was assessed by a self-rating questionnaire in which patients and controls classified their physical activity according to a four-grade scale.
Biochemical assays
Blood samples were drawn in the morning, after fasting since midnight. Serum IGF-I was assayed by immunoradiometric assay (Nichols Institute of Diagnostics, San Juan Capistrano, CA); the normal range was 122–400 µg/liter in subjects aged 19–40 yr. The intraassay coefficient of variation (CV) was 16% at the level of 60 µg/liter and 11% at the level of 300 µg/liter. Serum GH was analyzed by an immunofluorometric method (DELFIA hGH, Wallac Oy, Turku, Finland). The detection level for serum GH was 0.01 µg/liter, and the intra- and interassay CVs were 5 and 3%, respectively, at a level of 1.5 µg/liter and 3 and 5%, respectively, at a level of 7.7 µg/liter. The kit standards were calibrated against the first IS 80/505 International Standard, and the assay is specific for GH 22 kDa. Serum insulin was measured with a competitive RIA, with intraassay CVs at low, medium, and high levels, of less than 5.2%. Serum TSH, free T4, free T3, and cortisol were analyzed with an immunofluorometric technique (Auto Delfia, Wallac). The intraassay CVs for serum free T4 and serum free T3 were less than 4.1 and 8.0%, respectively, and interassay CVs were 4.1 and 8%, respectively. The reference ranges for serum free T4 and serum free T3 were 9–22 and 3.1–6.9 pmol/liter, respectively, and for TSH 0.33–4.7 mUI/liter. Serum FSH and LH were analyzed with an electrochemiluminescence technique (Elecsys, Roche, Mannheim, Germany). Serum testosterone, estradiol, and SHBG levels were measured by commercially available immunoassays. Venous blood glucose was analyzed with a blood glucose analyzer (Hemocue AB, Ängelholm, Sweden). According to the manufacturer, the SD between the cuvettes is less than 0.3 mmol/liter. Serum leptin was analyzed with a double-antibody RIA using rabbit antihuman leptin antibodies, 125I-labeled human leptin tracer, and human leptin as standard (Linco Research, St. Charles, MO). Interassay CV was 1.9% at low levels (<5 ng/ml) and 3.2% at high levels (10–15 ng/ml), intraassay CV was 1.5% at low levels and 3% at high levels, and the limit of detection was 0.5 ng/ml. Fasting total, low-density lipoprotein (LDL), and high-density lipoprotein (HDL) cholesterol, Apo (apolipoprotein) A1, Apo B, triglycerides, and fibrinogen levels were measured by standard procedures.
Statistical analysis
Data are presented as median and range. Patients and controls were compared with the Wilcoxon signed rank test for matched pairs. Bivariate correlations were assessed using the Spearman rank correlation coefficient. We regarded P < 0.05 as statistically significant. To assess gender dissimilarities, we calculated the difference within in each patient-control pair and then tested these differences using the Mann-Whitney U test with two independent groups, i.e. male and female patient-control pairs. Data stratified by gender are presented only if the Mann-Whitney U test for gender differences resulted in P < 0.10. Ordinary linear regression was used to assess the association between peak GH response and BMI and patient/control-status. The natural logarithm of the GH response was used as dependent variable to obtain approximately normally distributed regression residuals. Ordinary linear regression was also used to assess the association between cardiovascular risk factors as (untransformed) dependent variables and GH response and BMI.
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Significantly higher plasma levels of insulin and blood glucose and serum levels of LDL cholesterol, Apo B, triglycerides, fibrinogen, and leptin were recorded among the ALL patients, compared with controls (Table 3). Furthermore, the serum levels of HDL cholesterol and Apo A1 were significantly lower among the patients, compared with controls. Higher leptin levels per kilogram fat mass among patients than controls was indicated (P = 0.06); however, when stratified for gender, both leptin/kilogram fat mass (1.04 vs. 0.25; P = 0.001) and serum leptin (25.2 vs. 3.9 ng/ml; P < 0.001) were significantly higher among the female patients, compared with female controls. No such difference was recorded in the male patients (0.29 vs. 0.33 leptin/kilogram fat mass; P > 0.3 and 7.0 vs. 4.8 ng/ml; P > 0.3, respectively). Compared with their controls, there was no difference in the serum leptin levels among the male ALL patients with (7.9 vs. 3.0 ng/ml; P = 0.11) or without (4.2 vs. 9.5 ng/ml; P = 0.14) testosterone substitution.
|
Peak GH responses to stimulation tests and other hormone assessments
Table 2 shows the individual responses during both tests. The median peak GH response to the GHRH-arginine test was 74% lower in the patients, compared with controls (6.2 vs. 23.9 µg/liter) (Table 4). Twenty-nine patients (67%) had peak GH less than 9 µg/liter during the GHRH-arginine test. In 13 patients with peak GH 3.9 µg/liter or greater and 9 µg/liter or less during the GHRH-arginine test, only one patient had peak GH greater than 3 µg/liter to the ITT (4.2 µg/liter). For the 15 patients with a peak GH greater than 9 µg/liter during the GHRH-arginine test, only three had a peak GH greater than 3 µg/liter (4.6, 5.0, and 6.9 µg/liter) during the ITT. Thus, 40 of 44 ALL patients (91%) were considered GH deficient according to the ITT and/or the GHRH-arginine test, and the rest were considered to be GH insufficient.
|
Serum IGF-I levels were significantly lower in the patients than controls (Table 4). No significant difference was recorded in serum TSH, free T4, free T3, LH, testosterone, estradiol, SHBG, and cortisol levels between patients and controls. A difference in serum FSH was seen among male patients only (9.2 vs. 4.1 IU/liter; P = 0.002).
Correlations between peak GH response during GHRH-arginine or years since radiotherapy and cardiovascular risk factors
In patients, peak GH during GHRH-arginine was significantly negatively correlated to the following parameters: total body fat mass (kilograms) measured with DXA (r = –0.48, P = 0.001; Fig. 1), WHR (r = –0.32, P = 0.03), plasma insulin (r = –0.49, P = 0.001; Fig. 2), and leptin (r = –0.46, P = 0.002). Moreover, significantly positive correlations were recorded with HDL cholesterol (r = 0.38, P = 0.012) and lean mass (percent) measured by BIA (r = 0.43, P < 0.004). However, in multiple regression analyses adjusting for BMI, the peak GH response was no longer significantly associated (all P > 0.3) with any of these markers, whereas BMI was significantly associated with total body fat mass, muscle mass, plasma insulin, and leptin (P < 0.001 for each) and HDL cholesterol (P = 0.005) but not WHR (P = 0.10).
|
|
Estimated physical exercise
There was no difference in the degree of physical exercise during spare time between patients and controls (P > 0.3). However, during work time the patients reported significantly more physical exercise, compared with controls (P = 0.04).
Carotid artery ultrasound
No differences between patients and controls were seen for the intima-media thickness in the common carotid artery or the internal carotid artery on the right and left side (Table 5). The intima-media thickness was, however, significantly more pronounced in the right bifurcation among the patients, but this was not seen for the left bifurcation. No plaques were detected in any of the patients.
|
Heart rate was significantly higher in the patients, compared with controls (Table 6). The patients’ BSAs were significantly smaller, compared with the controls. After correction for BSA, left and right ventricular areas and right atrial area were significantly smaller in patients, compared with controls. However, no significant difference was seen in left atrial area, LVIDd, and LV-mass index measurements. When stratified for gender, LV-mass index among the females was significantly smaller, compared with controls (65 vs. 78 g/m2, P = 0.03), but no such difference was recorded among the males (91 vs. 94 g/m2, P > 0.3). Left ventricular systolic function measured as fractional shortening, ejection fraction, and atrium-ventricular plane displacement was significantly lower in the patients, compared with controls. Also, for left ventricular diastolic function, a significant reduction was recorded in the patients, compared with controls, when measurements were made with E/A ratio and atrium-ventricular plane diastolic velocity. However, when the deceleration time or pulmonary venous velocity was analyzed, no difference in left ventricular diastolic function was recorded between patients and controls. There was no significant valvular leakage or stenosis in either patients or controls. No abnormalities were found in the calculation of the systolic right ventricular/right atrial gradient between patients and controls. Of the cardiac function parameters, E/A ratio correlated to the peak GH response to GHRH-arginine test in the ALL patients (r = 0.46, P = 0.002). No significant correlations were observed between the cardiac mass or area measurements and peak GH response to GHRH-arginine or between cardiac measurements and time since treatment.
|
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
The strength of this study is that it is the largest reported and with the longest follow-up of a homogenous group of adults treated with CRT and chemotherapy for childhood ALL. Moreover, it included proper testing for GH secretion together with measurements of cardiovascular risk factors. In addition, a control group that was randomly recruited from the general population and matched not only for age and gender but also for smoking habits and residence was used. We consider this control group to be appropriate despite the fact that, as with all control groups, it represents a slightly selected part of the study base. However, only 7% of the potential controls declined participation due to health reasons, which was almost the same percentage as the ALL patients (5%) who were excluded due to disease. Thus, it was unlikely that the recruitment of the control group was hampered by a selection bias.
We were not able to include a control group of nonirradiated ALL patients because no such group comparable with respect to risk group classification, number of patients, calendar year of diagnosis, length of follow-up, and chemotherapy schedule was available.
Obesity is an early and progressive complication in childhood ALL and, in accordance with previous studies (20, 21, 22), the patients in the present study were significantly more obese in comparison with matched controls, according to WHR, BMI, BIA, or DXA measurements. Furthermore, abdominal obesity was pronounced with significantly higher WHR in former ALL patients, which are profound risk factors for cardiovascular disease (23). Steroid treatment has been suggested as an explanation (24). However, in a very large multicenter study of ALL patients who had reached final height, CRT but not steroid treatment during childhood was related to obesity (20). The present study revealed that former ALL patients are insulin resistant, consistent with the finding of a high prevalence (52%) of impaired glucose tolerance reported in allogeneic bone marrow transplanted patients in whom the majority were subjected to total body irradiation and in some cases CRT (7). Treatment-induced hypogonadism and GH deficiency, eventually leading to cellular atrophy in the target organ, was also suggested, but exact information regarding GH secretion was missing (7). In a previous study of childhood cancer survivors, the high-risk group with metabolic syndrome, including significantly higher blood glucose levels, had lower nocturnal GH secretion (5). This is in accordance with the present study showing a significant negative correlation between GH peak during GHRH-arginine test and basal insulin levels.
Compared with controls, the former ALL patients in the present study suffered from significant dyslipidemia; of particular interest, the present study is the first to show a significantly negative correlation between stimulated GH secretion and HDL cholesterol levels in ALL patients. Previously a significant decrease in HDL cholesterol was recorded in childhood cancer survivors (5). Also, a significant increase in triglycerides and very low-density lipoprotein cholesterol was seen in children with ALL but only for patients subjected to CRT, compared with those treated with chemotherapy (6). This is in line with the hypertriglyceridemia found in patients surviving allogeneic bone marrow transplantation (7).
The vast majority (91%) of the former ALL patients in the present study were considered GH deficient from the insulin tolerance test or the GHRH-arginine test (14, 15, 16). Based on the background to hypothalamic-pituitary disease, different GH tests must be carefully considered (17). Thus, in accordance with Darzy et al. (17), we recorded that the ITT clearly reflected the presence of radiation-induced GH deficiency, but this was not always the case with the GHRH-arginine test. However, when the GH response to GHRH-arginine was low, we considered the patient to be GH deficient (17). Thus, it would appear that primarily the hypothalamic and then later direct pituitary damage from CRT was the cause of GH deficiency among the former ALL patients in the present study. There was probably a continuum from insufficiency to deficiency in GH secretion, illustrated by the correlation between fat mass, insulin, and WHR and the GH peak during the GHRH-arginine test. This is consistent with a recent study in adults with GH deficiency showing a significant correlation between the GH response during the GHRH-arginine test and the severity of lipid profile abnormalities (25).
In normal subjects, the most important predictors of integrated 24-h GH concentration are visceral fat mass and fasting insulin (26), and a bidirectional correlation was suggested between these entities. It has been shown that obesity may lower the GH response to the GHRH-arginine test, both in GH-deficient patients and controls (27), which was corroborated by the present results. We could, however, show a 67% lower GH response among ALL patients, even after adjustment for BMI and gender. Thus, obesity alone cannot explain the very low GH response to the GHRH-arginine test. The GH deficiency or insufficiency after CRT in the present study may explain the increase in abdominal fat mass; this in turn is closely associated with an increase in fasting plasma insulin (28) and increased levels of total cholesterol, LDL cholesterol, and triglycerides (28, 29). This is in line with the significant correlation between estimated fat mass and CRT or GH deficiency shown in a study on children with ALL (30). However, after adjustment for BMI in the present study, the associations between peak GH and the cardiovascular risk factors disappeared, which indicates that obesity is part of a strong chain of events from GH deficiency preceding other risk factors, i.e. insulin resistance and dyslipidemia.
The dyslipidemia among the former ALL patients in the present study accords with the lipid profiles in GH deficiency, primarily caused by hypothalamic-pituitary diseases (8). In addition, an increase in HDL (8) and a reduction in LDL levels have been recorded after GH treatment in GH-deficient patients (31).
Significantly increased fibrinogen levels were found in the former ALL patients in the present study. This is a new finding in ALL patients but is in agreement with the GH deficiency (32). An elevated fibrinogen level together with increased WHR and the lipid derangements link thrombogenesis with atherogenesis and is an independent risk factor for cardiovascular disease and at least as important as blood pressure and blood lipids (33). A strong association between smoking habits and fibrinogen concentration has been observed (33), which was, however, controlled for in the present study.
An increase in serum leptin levels has been recorded previously in former ALL patients (9, 10), and the highest serum leptin levels were seen in the most severely GH-deficient patients (9). In the present study, there was a highly significant increase in serum leptin in the female ALL patients, but this was not seen among the male ALL patients, which is a new and intriguing finding. Furthermore, the ratio leptin to kilogram fat mass, used as a measure of adipose secretory function, was also significantly higher in the female ALL patients. Thus, a difference in total fat mass between women and men was not an explanation for the increase in serum leptin levels in the female ALL patients. Why only the female patients, compared with their male counterparts, were hyperleptinemic is unclear; the relation between sc and visceral fat mass may be different, but this was not investigated in the present study. However, the gender difference in the serum leptin levels was not affected by testosterone substitution among the male ALL subjects. Finally, this difference could not be explained by gender differences in treatment regimens with CRT or chemotherapy.
Another explanation for the increased cardiovascular risk among the former ALL patients could be a reduction in physical exercise, which has previously been emphasized (34). However, contrary to this, the present study did not show a significant decrease in exercise during spare time; in fact, the ALL patients experienced significantly more physical exercise at work than their matched controls.
After controlling for body surface area, and in line with a previous study (35), several of the echocardiographic measurements, including functional parameters, were significantly abnormal in the ALL patients. In the female ALL patients, but not in the males, LV mass index was significantly reduced. This gender difference could not be explained by different doses of anthracycline. In comparison with other studies (35), the total dose of anthracycline in the present study was rather low (120 mg/m2 BSA); nevertheless, a considerable impairment of cardiac function was recorded. Controversy exists on the frequency, severity, and risk factors for late subclinical cardiotoxicity after anthracycline therapy for childhood cancer (35). Female gender, young age at therapy, and high cumulative dose of anthracycline have been suggested as contributing factors. In the present study, there were also significantly negative correlations between FS and ejection fraction and the dose of anthracycline. Thus, anthracycline may have contributed to the decrease in systolic cardiac function. However, hitherto, no study has proposed GH deficiency as a possible additive effect to the cardiac dysfunction in this patient group. In the present study, no correlation was recorded between time since anthracycline treatment (mean 17 yr) and any cardiac measures, but the GH peak during the GHRH-arginine test correlated positively to the diastolic function measured as the E/A ratio. Of interest, adults with childhood-onset GH deficiency, due to primary hypothalamic-pituitary diseases, have a reduction in LV-mass and impairment in cardiac systolic function, whereas GH treatment improved both parameters (36, 37). Furthermore, in adolescents with GH deficiency, withdrawal of GH treatment for 6 months significantly lowered LV-mass index, LVEF, and E/A ratio, and reintroduction of GH treatment for 6 months significantly improved LV-mass index (31). In addition, when these adolescents were off GH treatment, a significant increase in heart rate was recorded (31), which is in accordance with the results in the present study and seems to be a new finding in ALL patients. The cause may be related to the intense sympathetic nerve activity (38) or abnormal heart rate variability (39) shown in GH-deficient patients.
In the right bifurcation, but not the left, the intima-media thickness of the carotid artery was significantly increased, which is difficult to explain. Among childhood brain cancer survivors, a significantly increased intima-media thickness was recorded in the carotid bulb, compared with controls, but in no other part of the carotid arteries (4).
The significant increase in cardiovascular risk factors among the former ALL patients in the present study indicates an increased risk for cardiovascular morbidity and mortality (23, 28). Thus, intervention is highly warranted including professional advice concerning nutrition, physical activity, smoking cessation, and if necessary lipid-lowering therapy. The positive effects of GH substitution on cardiovascular risk factors among GH-deficient patients due to other causes has been demonstrated previously but remains to be confirmed in this patient group.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Abbreviations: ALL, Acute lymphoblastic leukemia; Apo, apolipoprotein; A-wave, peak late diastolic filling; BIA, bioelectrical impedance assay; BMI, body mass index; BSA, body surface area; CI, confidence interval; CRT, cranial radiotherapy; CV, coefficient of variation; DXA, dual-energy x-ray absorptiometry; E-wave, peak early diastolic filling; FS, fractional shortening; HDL, high-density lipoprotein; ITT, insulin tolerance test; LDL, low-density lipoprotein; LVEF, left ventricular ejection fraction; LVIDd, left ventricular end-diastolic dimension; LVIDs, left ventricular end-systolic dimension; LV-mass, left ventricular mass; LVPWd, left ventricular posterior wall thickness; WHR, waist to hip ratio.
Received January 23, 2004.
Accepted July 12, 2004.
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  L. A. Robinson, K. M. Emmons, E. T. Moolchan, and J. S. Ostroff Developing Smoking Cessation Programs for Chronically Ill Teens: Lessons Learned from Research with Healthy Adolescent Smokers J. Pediatr. Psychol., March 1, 2008; 33(2): 133 - 144. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. M. Shankar, N. Marina, M. M. Hudson, D. C. Hodgson, M. J. Adams, W. Landier, S. Bhatia, K. Meeske, M. H. Chen, K. E. Kinahan, et al. Monitoring for Cardiovascular Disease in Survivors of Childhood Cancer: Report From the Cardiovascular Disease Task Force of the Children's Oncology Group Pediatrics, February 1, 2008; 121(2): e387 - e396. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. T. Armstrong, C. A. Sklar, M. M. Hudson, and L. L. Robison Long-Term Health Status Among Survivors of Childhood Cancer: Does Sex Matter? J. Clin. Oncol., October 1, 2007; 25(28): 4477 - 4489. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. M. Janiszewski, K. C. Oeffinger, T. S. Church, A. L. Dunn, D. A. Eshelman, R. G. Victor, S. Brooks, A. J. Turoff, E. Sinclair, J. C. Murray, et al. Abdominal Obesity, Liver Fat, and Muscle Composition in Survivors of Childhood Acute Lymphoblastic Leukemia J. Clin. Endocrinol. Metab., October 1, 2007; 92(10): 3816 - 3821. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Bjork, K. Link, and E. M. Erfurth The Utility of the Growth Hormone (GH) Releasing Hormone-Arginine Test for Diagnosing GH Deficiency in Adults with Childhood Acute Lymphoblastic Leukemia Treated with Cranial Irradiation J. Clin. Endocrinol. Metab., November 1, 2005; 90(11): 6048 - 6054. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M Jarfelt, B Lannering, I Bosaeus, G Johannsson, and R Bjarnason Body composition in young adult survivors of childhood acute lymphoblastic leukaemia Eur. J. Endocrinol., July 1, 2005; 153(1): 81 - 89. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
