Bone Mineral Density in Childhood Survivors of Acute Lymphoblastic Leukemia Treated without Cranial Irradiation

Adult survivors of childhood acute lymphoblastic leukemia (ALL) whose treatment included cranial irradiation (XRT) have reduced bone mineral density (BMD). Fifty-three survivors of ALL (aged 6–17 yr; 22 males and 31 females), who had completed their treatment without XRT, at least 1 yr previously, and 187 (5–19 yr; 86 males and 101 females) healthy controls were examined with dual energy x-ray absorptiometry of the total body and L1–L4 vertebrae and peripheral quantitative computer tomography at the distal and midradial sites. The total body and lumbar spine BMDs did not differ between the ALL survivors and controls. Distal radial trabecular BMD (difference, −0.080 mg/cm³; 95% confidence interval, −0.139 to −0.020; P = 0.009), but not total BMD (difference, −0.006 mg/cm³; confidence interval, −0.051 to 0.039; P = 0.80), was lower in ALL survivors compared with controls. At the midradial site, both endosteal (11% larger; P = 0.0001) and periosteal (4% larger; P = 0.001) circumferences were greater, and cortical thickness was thinner by 6% (P = 0.006) in the ALL subjects, leading to an increase in the axial moment of inertia in the ALL subjects (difference, 13%; P = 0.008). In conclusion, BMD, except at the radius, is normal in childhood survivors of ALL treated without XRT. At the midradial site, we speculate that ALL or its treatment resulted in endosteal bone loss and cortical bone thinning, but the axial moment of inertia and, hence, strength was maintained as a result of bone gain at the periosteal surface.

AS A RESULT of modern treatment protocols, the majority of children with acute lymphoblastic leukemia (ALL) will now survive to adulthood; hence, the long-term complications of their treatment have become increasingly important (1). Furthermore, because of the ability to treat children with ALL successfully from the 1970s onward, a significant proportion of this cohort is now reaching their third and fourth decades, so the specific issues of bone health need to be addressed.

We have previously shown that adults treated for ALL in childhood have a highly significant reduction in bone mineral density (BMD) (2). Their treatment included multiagent chemotherapy, corticosteroids, and cranial irradiation (XRT), a known risk factor for impaired GH secretion (3), but we were unable to show a relationship between the cohort’s GH status and BMD. GH status is relevant to bone health, because reduced GH secretion is a causative factor for osteoporosis (4). Other factors implicated in such children to explain the reduced BMD include chemotherapy, specifically methotrexate (5), corticosteroids (6), the leukemia itself (7, 8), poor nutrition and decreased physical activity (9) during treatment, and ultimately a combination of all these factors. A reduction in BMD may limit bone strength and increase the risk of fragility fractures in adult survivors (10) and even in childhood.

The majority of children recently treated for ALL will not have received XRT and hence should have normal GH status. Little is known about their bone health (11), and few investigators have used volumetric methods for assessing BMD (12). A reduction in BMD in children treated for ALL, including XRT, has been documented (12, 13). The majority of previous studies in ALL patients (11, 13–15) have used dual energy x-ray absorptiometry (DXA), which bases its measurements on a two-dimensional projection of a three-dimensional structure. It does not measure the depth of the site that is scanned and thus does not fully account for bone size; it underestimates the areal BMD (aBMD: grams per square centimeter) of small bones and exaggerates the aBMD of large bones.

In this study we used DXA to determine the lumbar spine (LS) BMD and bone mineral apparent density (LSBMAD) in 53 childhood survivors of ALL, who had completed their treatment without XRT at least 1 yr previously, and in 187 healthy controls. We also used peripheral quantitative computed tomography (pQCT) to measure total and trabecular volumetric BMD at a distal radius site. The geometric properties of the midradial cortical shell were determined, including the endosteal circumference, periosteal circumference, cortical thickness, and axial moment of inertia, which is a measure of the distribution of cortical bone away from the central axis and is related to bending strength of bone (16). Finally, we used DXA to measure the whole body lean body mass and fat mass.

Subjects and Methods

Subjects

Fifty-three Caucasian children (22 males and 31 females) were assessed at a median age of 4.6 yr (range, 1.2–8.3 yr) after treatment for ALL. The median age at the assessment of BMD was 11.2 yr (range, 6.4–17.5 yr). They had all received multiagent chemotherapy for 2 yr without XRT. Chemotherapy had been completed at least 1 yr previously in all cases; all were in first remission, and none had received any hormone replacement therapy. Two patients had fractured their wrists 1 yr into treatment, and four patients had fractured bones at varying sites (wrist, arm, finger, and patella) after treatment had been completed. A review of the patients’ medical records was performed to obtain details of their chemotherapy. Standing height (meters) was measured for each participant using a stadiometer (Holtain Ltd., Crymych, UK), and weight (kilograms) was measured by a digital scale [Autoweigh, Ltd., Elland, UK]. Height and weight values were converted to sd scores (SDS) using the standards of Freeman et al. (17). Body mass index (BMI) was calculated using the formula kilograms per meter squared. Tanner stages of puberty (18) were determined by one investigator (B.B.). The study was approved by the Salford and Trafford local ethical committee, and written consent was obtained from parents and older subjects.

Treatment

Treatment for ALL according to the Medical Research Council UKALL XI (92) protocol was carried out at one oncology center, Royal Manchester Children’s Hospital, over a 2-yr period (19). In summary, this treatment consisted of 1 month of induction treatment, including vincristine, asparaginase, and prednisolone, followed by two intensification blocks at wk 5 and 20. These blocks included iv vincristine, daunorubicin, etoposide, cytarabine, and oral thioguanine plus prednisolone. In between these blocks, all patients received central nervous system-directed therapy in which they were randomized to either high dose methotrexate iv with intrathecal methotrexate or intrathecal methotrexate alone. In between the intensification blocks and after wk 20 they received maintenance therapy, consisting of oral weekly methotrexate and daily mercaptopurine and monthly injections of vincristine with a 5-d course of prednisolone. All patients were also randomized to receive, or not, a third intensification block of oral dexamethasone, vincristine iv, l-asparaginase, intrathecal methotrexate, iv cyclophosphamide, cytarabine, and oral thioguanine. Cumulative doses of drugs were 5,400–10,800 IU/m² asparaginase, 43.5–46.5 mg/m² vincristine, 6,120–6,520 mg/m² prednisolone, 1,000 mg/m² etoposide, 180 mg/m² daunorubicin, 2,000–3,200 mg/m² cytarabine, 800–2,480 mg/m² thioguanine, 1,520–1,740 mg/m² methotrexate orally, 18–24 g/m² methotrexate iv, and 120–212.5 mg/m² methotrexate intrathecally.

Eleven patients received intrathecal methotrexate and no third intensification block; 18 patients received high dose iv methotrexate, intrathecal methotrexate, and no third intensification; 15 patients received intrathecal methotrexate and the third intensification block; and 10 patients received high dose methotrexate, intrathecal methotrexate, and the third intensification block.

Controls

The controls consisted of 187 (86 male) healthy Caucasian Manchester school children, aged 5–19 yr, who had been recruited to obtain normative data on BMD and body composition. Ethical approval was granted by the central Manchester local research ethics committee. Informed written consent was obtained from the parents of children and from subjects over 16 yr of age. Children with a personal or a family history of medical problems known to affect bone mineralization were excluded. Standing height (meters) was measured for each participant using a stadiometer (Leicester Height Meter, Child Growth Foundation, London, UK), and weight (kilograms) was measured using a Secca electronic digital scale. BMI was calculated using the formula kilograms/meters squared. Weight, height, and BMI SDS were calculated using the standards reported by Freeman et al. (17). Their pubertal status was assessed by direct assessment (18) or by using the previously validated self-assessment method (20).

Bone densitometry

A QDR-4500A (Hologic, Waltham, MA) DXA scanner was used to measure total body bone mineral content (TBBMC; grams), total body bone area (TBBA; square centimeters), and total aBMD (TBaBMD; grams per square centimeter). The total body lean body mass (TBLBM) and fat mass (TBFM) values were derived from total body DXA scans. The lumbar spine (LS: L1–L4) bone mineral content (LSBMC; grams), bone area (LSBA; square centimeters), and aBMD (LSaBMD; grams per square centimeter) were also measured. The bone mineral apparent density of LS (LSBMAD; grams per cubic centimeter) was calculated by dividing LS bone mineral content (LSBMC; grams) by LS bone area (LSBA; square centimeters) to power of 1.5 (LSBA^1.5) (21). The total body scans were performed with subjects lying supine and wearing T-shirts and shorts, and analyses were performed in the adult mode, using the software version V8.26a:3. The in vivo coefficient of variation (CV) for TBBMC for adults is 0.92%, and the CV for LSaBMD IS 1.09%. The in vivo CVs for TBLBM and TBFM for adults are 1.75% and 0.56%, respectively.

Peripheral quantitative computed tomography (pQCT) measurements were made using a Stratec XCT-2000 scanner (Stratec, Pforzheim, Germany). All measurements were made using the nondominant arm in accordance with the manufacturer’s recommendations, with software version 5.40b. Measurements were made at 4% and 50% of the forearm length. A 1.2-mm-thick single tomographic slice was sampled using the scan speed of 25 mm/sec and a voxel size of 0.4 mm. The volumetric BMD (milligrams per cubic centimeter) were determined at the distal radius (4% site; proximal to the growth plate) for the total cross-section (cortical and trabecular) and for the trabecular (<400 mg/cm³) compartment. At the 50% midshaft, the cross-sectional area of cortical bone (square millimeters), cortical thickness (millimeters), and the outer and inner cortical bone contours (periosteal and endosteal circumferences; millimeters) were determined at a threshold of 710 mg/cm³. At this site, the volumetric cortical BMD (milligrams per cubic millimeter) was determined at a threshold of 710 mg/cm³. The axial moment of inertia (1 mm⁴), which is a measure of the distribution of cortical bone mass about the center of the cross-section of a tubular bone and is related to bending strength of bone, was also measured. In our unit, the in vivo CVs for total and trabecular volumetric BMD at the distal radius in children are 0.78% and cortical 0.86% (Fig. 1).

Fig. 1.

A schematic view of the cross-section of the midradius, indicating the parameters measured by pQCT: cortical thickness, endosteal circumference, and periosteal circumference.

Statistical analysis

Population parameters were summarized by their medians and interquartile ranges. Analysis of covariance was used to compare the ALL and control groups. Because the relationship between the bone parameters and age was nonlinear, a cubic spline function of age with 4 degrees of freedom was used as the covariate, with separate spline functions being fitted for each combination of gender and pubertal stage (the latter being represented as three groups, Tanner stages 1 and 2, 3 and 4, and 5). The effect sizes were insensitive to the degrees of freedom employed. Height, weight, and BMI were analyzed as SDS, and all other variables were logarithmically transformed for analysis; the effect sizes then were exponentiated to give percent changes. Additional terms were considered involving interactions of pubertal stage and gender with participant group, thus providing a test for nonhomogeneity in response across age/gender groups. These interactions were generally not significant, and where they were, additional exploration was undertaken, but the primary results are based on the age/gender-averaged effects.

For presentation the model was used to produce age/puberty/gender-adjusted results, centered on a midpubertal male of median age. To explore the potential transience of the ALL effects, an additional interaction term (time since treatment by participant group) was added to the model.

Analyses were performed using the R statistical package, version 1.8.1 (22).

Results

The median values and interquartile ranges for age, height, weight, and BMI, expressed as SDS, in children treated for ALL and controls are presented in Table 1. There were no statistically significant differences between ALL subjects and controls for height and weight. BMI SDS, however, was significantly higher in the ALL subjects compared with the controls (P = 0.0002).

TABLE 1.

The female and male age, height, weight, and BMI SDS for ALL cases and controls

	ALL cases	Controls
Female age (yr)	10.4 (8.6–12.7)	10.1 (8.4–14.4)
Male age (yr)	12.8 (11.0–14.1)	10.5 (8.7–13.5)
Female height SDS	0.29 (−0.49–1.22)	−0.05 (−0.59–0.58)
Male height SDS	−0.03 (−0.37–0.51)	0.26 (−0.24–0.81)
Female weight SDS	0.56 (−0.12–1.25)	0.09 (−0.45–0.82)
Male weight SDS	0.08 (−0.08–0.51)	0.40 (−0.21–0.95)
Female BMI SDS	1.09 (0.54–1.84)	0.16 (−0.44–0.90)
Male BMI SDS	0.49 (−0.20–1.06)	0.24 (−0.26–0.82)

Based on 53 ALL patients and 187 controls. Values are medians (interquartile range).

The differences in TBBA, TBBC, TBaBMD, TBLBM, and TBFM between the ALL subjects and controls are presented in Table 2. There was a small (4%) difference in TBBA, higher in the ALL subjects compared with the controls, which was of borderline statistical significance. There were no detectable differences, however, for the rest of the bone parameters. TBFM was substantially greater (by 25%) in the ALL subjects compared with the controls, and there was a smaller difference (5%) in TBLBM.

TABLE 2.

DXA total body scans for ALL cases and controls

	ALL effect
	Effect size (95% CI)	% Change	P
TBBA (cm2)	0.035 (−0.002–−0.072)1	3.6 (0.2–6.9)	0.066
TBBMC (g)	0.024 (−0.025–0.073)	2.4 (−2.5–7.6)	0.34
TBaBMD (g/cm2)	−0.011 (−0.033–0.011)	−1.1 (−3.2–1.1)	0.32
TBLBM (g)	0.051 (0.008–0.093)1	5.2 (0.8–9.7)	0.019
TBFM (g)	0.22 (0.09–0.35)1	24.6 (9.3–41.3)	0.0010

Based on 53 ALL patients and 187 controls. The effect size is the difference in the log-transformed measures between ALL patients and controls after adjustment for age, gender, and pubertal stage. This is converted to a percent change between the groups. Values are medians (interquartile range). CI, Confidence interval.

¹

Significant heterogeneity in response across puberty/gender groups.

Table 3 demonstrates the differences in LS DXA data. Bone area at L1–L4 was higher in the ALL subjects by a modest amount (4%; P = 0.043), but there was no difference for LSBMC and LSBMAD.

TABLE 3.

DXA LS scan results for ALL cases and controls

	ALL effect
	Effect size (95% CI)	% Change	P
LSBMC (g)	0.037 (−0.025–0.098)	3.8 (−2.5–10.3)	0.24
LSBA (cm2)	0.038 (0.001–0.075)	3.9 (0.1–7.8)	0.043
LSBMAD (g/cm3)	−0.020 (−0.058–0.018)	−2.0 (−5.6–1.8)	0.31

Based on 53 ALL patients and 187 controls. The effect size is the difference in the log-transformed measures between ALL patients and controls after adjustment for age, gender, and pubertal stage. This is converted to a percent change between the groups. CI, Confidence interval.

The differences in pQCT-measured parameters at the distal radial (4%) and midradial site (50%) are shown in Table 4. The median distal radial trabecular density for ALL subjects was significantly lower (8%) compared with that in the controls (P = 0.009; Fig. 2B). However, the median values for the total bone area or total (cortical and trabecular) BMD (Fig. 2A) at this radial site did not differ between ALL subjects and controls. At the midradial site, the median cortical thickness was reduced by 6% (P = 0.006) in the ALL subjects, and both endosteal (11% larger; P = 0.0001) and periosteal (4% larger; P = 0.001) circumferences were greater in the ALL subjects compared with controls. This displacement of the bone cortex outward led to a significant increase in bending strength of bone, as reflected by the axial moment of inertia in the ALL subjects compared with the controls (difference, 13%; P = 0.008).

Fig. 2.

A and B, Box and whisker plots of the distal radial volumetric total (A) and trabecular (B) BMD in ALL cases and controls. The horizontal line represents the median, the box indicates the interquartile range, and the whiskers show the data range, excluding the indicated extreme observations. The data have been adjusted for age, gender, and pubertal stage.

TABLE 4.

The results from pQCT scans of the distal radial (4%) and midradial (50%) sites for ALL cases and controls

	ALL effect
	Effect size (95% CI)	% Change	P
Distal radial volumetric total BMD (mg/cm3)	−0.006 (−0.051–0.039)	−0.6 (−5.0–−4.0)	0.80
Distal radial volumetric trabecular BMD (mg/cm3)	−0.080 (−0.139–−0.020)	−7.7 (−13.0–−2.0)	0.009
Midradial periosteal circumference (mm)	0.043 (0.017–0.068)1	4.4 (1.7–7.0)	0.001
Midradial endosteal circumference (mm)	0.104 (0.052–0.157)1	11.0 (5.3–17.0)	0.0001
Midradial cortical thickness (mm)	−0.061 (−0.1085–−0.018)	−5.9 (−10.0–−1.8)	0.006
Midradial axial moment of inertia (1 mm4)	0.123 (−0.008–0.161)1	12.7 (3.4–23.6)	0.0075
Midradial cortical density (mg/cm3)	−0.061 (−0.105–−0.018)	−6.0 (−10.0–−1.7)	0.0063

Based on 53 ALL patients and 187 controls. The effect size is the difference in the log-transformed measures between ALL patients and controls after adjustment for age, gender, and pubertal stage. This is converted to a percent change between the groups. CI, Confidence interval.

¹

Significant heterogeneity in response across puberty/gender groups.

Figure 3 shows the age/gender/puberty-adjusted midradius trabecular BMD for the ALL patients plotted against the interval between the end of treatment and the scans. Patients scanned close to the end of their treatment showed a larger difference from the controls than patients scanned after a longer recovery period. This plot indicates that there is a trend for the difference between the groups to improve with time and that these effects may be transitory. A formal test of this decline or improvement in effect over time is provided by the inclusion of an appropriate time by effect interaction factor in the analysis of covariance, that is, after allowance for age, gender, and pubertal stage. Within this framework we can demonstrate a significant decline in effect over time for BMI (P = 0.019), midradial periosteal circumference (P = 0.0004), endosteal circumference (P = 0.015), and moment of inertia (0.0007), whereas TBFM (P = 0.051) is of borderline significance. For distal radial trabecular BMD, the improvement with time is not statistically significant (P = 0.63)

Fig. 3.

Plot of distal radial volumetric trabecular BMD against time since end of treatment for ALL patients. Data are adjusted for age, gender, and pubertal stage. The line represents a simple linear regression against time. For comparison, the data for the controls are represented by a box plot on the right of the plot.

Discussion

To the best of our knowledge, this is the first study to measure the TBBMC, LSBMAD, distal radial volumetric total and trabecular BMD, and geometric properties of the midradial cortical shell in childhood survivors of ALL who did not receive XRT as part of their ALL treatment.

The ultimate interest in the bone health of ALL survivors is to determine whether the disease or its treatment has had a detrimental effect on bone strength and, hence, might lead to an increased risk of fractures. All previous pediatric studies of the effects of ALL on bone health have focused on the assessment of aBMD, which is an important determinant of the risk of fracture in both adults (10) and children (23). However, bone strength depends not only on the BMC or BMD, which reflect material properties, but also on the size, shape, and three-dimensional architecture of the bone (24).

The majority of previous studies in ALL (11, 13, 14, 25) have used DXA, which provides a measurement of the total amount of BMC contained within the scanned skeletal region and the two-dimensional projected bone area. The ratio of BMC and bone area, expressed in units of grams per square centimeter, is referred to as the aBMD. aBMD is a function of a bone’s size and its BMC; aBMD increases with bone size due to the greater thickness of larger bones. Thus, the interpretation of aBMD poses major challenges in childhood survivors of ALL, because the disease or the treatment may have affected their growth and pubertal development and thus the bone size. This problem can be overcome by either measuring volumetric spinal BMD by QCT (12) or by estimating BMAD from two-dimensional DXA-measured variables (21, 26).

We found that after adjustment for age, gender, and pubertal stage, the median TBBMC, TBaBMD, and LSBMAD of ALL subjects were not different from those of controls. The findings of normal LSBMAD supports those reported by Kadan-Lottick et al. (11), although the population of ALL subjects they studied varied from ours in that 31% of their cohort was still undergoing maintenance chemotherapy. This, however, could only reduce the cohort’s LSBMAD and not falsely increase it, as confirmed in their study. Individuals still receiving maintenance chemotherapy had a lower LSBMAD SDS than those who had completed treatment (11). Our findings of normal LSBMAD is in contrast to other studies that have employed DXA to assess LSBMAD in ALL survivors treated without cranial XRT (9, 14, 15). Tillmann et al. (9) assessed LSBMAD in 28 children treated for ALL in a contemporary time period to our study with subjects treated with the same ALL protocol, UKALL XI. Although the mean LSBMAD was reduced in the ALL subjects compared with the controls, there was significant overlap between the two groups. Furthermore, because their controls were not matched for pubertal stage, sex, or age and were recruited from the children of hospital staff, this might explain the higher LSBMAD in their control population compared with their ALL subjects. Finally, the difference might be due in part to the fact that Tillmann et al. (9) used an approach for estimating LSBMAD that assumes that the vertebral body is a cylinder (26), whereas we used the approach described by Carter et al. (21), which assumes the vertebral body to be a cube. Nysom et al. (14) examined aBMD in a large cohort of ALL survivors (n = 95) and had sufficient numbers to analyze separately patients less than 19 yr who did not receive XRT (n = 33). The ALL subjects had significantly lower LSaBMD compared with controls even in the nonirradiated cohort, partially explained by reduced bone size (14). Arikoski et al. (15) found LSBMAD, but not aBMD, reduced in a cohort of 22 survivors of ALL at the end of chemotherapy, of whom two had received cranial XRT. This contrasts with the findings of an earlier study by the same group, where both LSBMAD and aBMD were normal in long-term survivors of ALL treated without cranial XRT (25).

Two studies have reported the QCT-measured volumetric trabecular LSBMD in childhood ALL survivors, including a population treated without cranial XRT (12, 27). Gilsanz et al. (12) demonstrated that in the cohort treated without cranial XRT, their BMD was no different from controls who were individually matched with each subject for both sex and age. In the study by Kaste et al. (27), 46 of the 141 subjects studied did not receive cranial XRT and had a significantly better trabecular BMD than those who had received 24 Gy cranial XRT (n = 15), but had similar trabecular BMD as those who had received 18 Gy. Furthermore, the trabecular BMD SDS for those who did not receive cranial XRT was not significantly reduced compared with that of the reference population. Another possible factor that might have had a detrimental effect on the trabecular BMD of subjects in the study by Kaste et al. (27) who did not receive cranial XRT was their treatment with antimetabolite therapy, mercaptopurine and methotrexate. The trabecular BMD SDS of patients who received higher doses of antimetabolites were lower than those of the other patients (27).

We found that the distal radial trabecular BMD, but not total BMD (trabecular and cortical compartments), was reduced in ALL survivors compared with controls. Because BMD is related to bone strength in both adults (10) and children (23), it is possible that children treated for ALL might be at an increased risk of fracture of the distal forearm bones. However, as shown in Fig. 3, any increased risk of fracture is likely to be transient, because the difference between the distal radial trabecular BMD of ALL survivors and that of controls shows an improvement with time after completion of chemotherapy. The use of pQCT to measure the geometrical properties of midradius has allowed us to study how the skeleton adapts to illness. The results of this study show that at midradius, survivors of ALL had larger endosteal circumference and thinner cortical shell, probably due to endosteal bone loss during the acute and treatment phases of the illness. The net effect of these changes would be to decrease the BMC and BMD of the radial diaphysis, and thus bone strength. However, we also found that ALL survivors had significantly greater midradial periosteal diameter compared with the controls, suggesting that there was a compensatory periosteal apposition, In other words, the new bone is placed further away from the long axis of the cortical shell, thus improving the axial moment of inertia (16, 28, 29), which is related to bone’s ability to withstand bending and torsional forces despite the cortical thinning. For trabecular BMD at the distal radius, we observed a decline in the difference in midradial endosteal diameter, periosteal diameter, and axial moment of inertia with passage of time after the end of chemotherapy, suggesting that these effects may be transitory. However, this conclusion is based on exploratory analyses of cross-sectional data, and given the correlation between participant age and time since treatment, this can only be a tentative conclusion and needs to be confirmed in a longitudinal study.

Not unexpectedly, the total body fat mass was greater by 25% in the ALL subjects compared with the controls. This finding is supported by previous body composition studies in children previously treated for childhood ALL without cranial XRT (9, 30). As for the radial bone parameters, the trend for a decline in fat mass with time from the end of chemotherapy needs to be explored in a longitudinal study.

In conclusion, our study confirms the growing evidence that modern ALL treatment, without cranial XRT, may not have long-term detrimental effects on BMD of the total body nor at axial and appendicular skeletal sites in children whose bone development is not complete. The use of pQCT has allowed us to study how the skeleton adapts to illness; the midradius undergoes endosteal resorption and periosteal apposition, which help to restore diaphyseal bone strength. A follow-up study of this cohort is required to confirm that the observed changes in distal radial volumetric trabecular BMD and midradial bone geometry improve or normalize with longer duration of survival of ALL children.

First Published Online November 23, 2004

Deceased.

This work was supported by a grant from the Friends of Rosie Children’s Cancer and Leukemia Research Fund.

Abbreviations:

 
• aBMD,

  Areal BMD;

 
• ALL,

  acute lymphoblastic leukemia;

 
• BA,

  bone area;

 
• BMC,

  bone mineral content;

 
• BMD,

  bone mineral density;

 
• BMI,

  body mass index;

 
• CV,

  coefficient of variation;

 
• DXA,

  dual energy x-ray absorptiometry;

 
• LS,

  lumbar spine;

 
• LSaBMD,

  LS areal BMD;

 
• LSBMAD,

  LS bone mineral apparent density;

 
• pQCT,

  peripheral quantitative computed tomography;

 
• SDS,

  sd score;

 
• TBaBMD,

  total body aBMD;

 
• TBBA,

  total body BA;

 
• TBBMC,

  total body BMC;

 
• TBFM,

  total body fat mass;

 
• TBLBM,

  total body lean body mass.