Bone Marrow Transplantation (2004) 33, 435–441 & 2004 Nature Publishing Group All rights reserved 0268-3369/04 $25.00 www.nature.com/bmt Post-transplant complications Bone mineral density and osteonecrosis in survivors of childhood allogeneic bone marrow transplantation SC Kaste1,2, TJ Shidler1, X Tong3, DK Srivastava3, R Rochester4, MM Hudson4,5, PD Shearer4,5 and GA Hale4,5 1Department of Diagnostic Imaging, St Jude Children’s Research Hospital, Memphis, TN, USA; 2Department of Radiology, University of Tennessee College of Medicine, Memphis, TN, USA; 3Biostatistics, St Jude Children’s Research Hospital, Memphis, TN, USA; 4Department of Hematology-Oncology (Division of Stem Cell Transplantation), St Jude Children’s Research Hospital, Memphis, TN, USA; and 5Department of Pediatrics, University of Tennessee College of Medicine, Memphis, TN, USA Summary: are alive after undergoing allogeneic bone marrow transplantation (alloBMT) between 1968 and 2002 (M Our purpose was to evaluate frequency and severity of bone Horowitz, personal communication). Owing to the growing mineral decrements and frequency of osteonecrosis in success of pediatric alloBMT, an increasing number of long- survivors of pediatric allogeneic bone marrow transplanta- term survivors are at risk of disease- and therapy-induced tion (alloBMT). We retrospectively reviewed demographic adverse late effects, including diminished bone mineral density information, treatment, magnetic resonance (MR)imaging (BMD) and osteonecrosis. Limited guidelines are available studies (hips and knees), and bone mineral density (BMD) to direct management of survivors of pediatric alloBMT,1–3 studies of 48 patients as measured by quantitative computed of which few address potential skeletal complications.4 tomography (QCT). In all, 24 patients were male; 37 To assess the frequency and severity of bone mineral were Caucasian. Median age at alloBMT was 10.3 years deficits and of osteonecrosis in survivors of childhood (1.6–20.4 years). Of the 48 patients, 43 underwent QCT. alloBMT, we retrospectively reviewed the records of routine Median time between alloBMT and imaging was 5.1 years annual follow-up evaluations of survivors of pediatric (1.0–10.2 years). Median BMD Z-score was À0.89 (À4.06 alloBMT treated at a single pediatric transplant center. to 3.05). BMD Z-score tended to be associated with female sex (P ¼ 0.0559)but not with age at BMT, race, primary diagnosis, time from alloBMT, T-cell depletion of graft, Patients and methods total-body irradiation, or acute/chronic graft-versus-host disease (GVHD). MR showed osteonecrosis in 19 of 43 All patients who had follow-up evaluations for alloBMT (44%). We found no associations between osteonecrosis between March 1, 2001 and November 8, 2001, and who and sex, race, diagnosis, age at BMT, history of GVHD, underwent BMD assessment, and/or magnetic resonance time from BMT, or T-cell depletion. Seven patients (15%) (MR) of the hips and/or knees were eligible for inclusion. had MR changes of osteonecrosis and BMD Z-scores of Routine evaluation of BMD screening for osteonecrosis less than –1 s.d. We conclude that pediatric alloBMT using MR was initiated in March 2001. The reported survivors have decreased BMD and are at risk of patient cohort comprises a consecutive group of patients osteonecrosis. They should be monitored to assure early who returned for their annual post-BMT evaluation for intervention that may ameliorate adverse outcomes. whom these studies were part of routine evaluation. Bone Marrow Transplantation (2004) 33, 435–441. Information abstracted from medical records review doi:10.1038/sj.bmt.1704360 included demographic information, the date and type of Published online 12 January 2004 primary diagnosis, the alloBMT conditioning regimen Keywords: osteoporosis; childhood bone marrow trans- (including use of total body irradiation), T-cell depletion plantation; osteonecrosis of the graft, donor type, date of transplantation, and history of acute or chronic graft-versus-host disease (GVHD). We also retrospectively reviewed all quantitative computed tomography (QCT) and MR imaging studies The International Bone Marrow Transplant Registry performed during the annual evaluation of those patients. reports that 14 309 persons less than 21 years of age This study was approved by the St Jude Children’s Research Hospital Institutional Review Board (IRB); informed consent was waived for this retrospective study. Correspondence: Dr SC Kaste, Department of Diagnostic Imaging, St Jude Children’s Research Hospital, 332 North Lauderdale, Memphis, Transplantation TN 38105-2794, USA; E-mail: [email protected] Received 11 August 2003; accepted 24 September 2003 All patients underwent alloBMT on IRB-approved Published online 12 January 2004 protocols. In all, 26 patients received bone marrow from BMD and AVN after childhood allogeneic BMT SC Kaste et al 436 HLA-matched siblings, 25 from matched unrelated donors Hip and knee radiographs (matched at five or six HLA loci), and eight from We no longer routinely perform radiographic evaluation of mismatched family members. All patients who had the hips or knees for assessment of osteonecrosis until such hematologic malignancies (n ¼ 50) were conditioned with time as orthopedic surgery is contemplated. Thus, from cytarabine (3 g/m2/dose, six doses) and cyclophosphamide radiographic reports, we captured whether or not patients (45 mg/kg/dose, two doses), with mesna uroprotection with MR findings of osteonecrosis underwent radiographic (45 mg/kg divided into five doses) and received total-body evaluation during the study period and whether or not the irradiation (TBI) in eight fractions over 4 days in doses of radiographs identified the presence of osteonecrosis. 150 or 175 cGy per fraction. Recipients of matched sibling grafts received 12 Gy and recipients of unrelated or mismatched family member grafts received 14 Gy. Definitions Patients with nonmalignant diseases (n ¼ 9) received a The Z-score (number of s.d.’s from the mean for age- and preparatory regimen of busulfan (1 mg/kg/dose, 16 doses) gender-matched control subjects) was used as an end point and cyclophosphamide (50 mg/kg/dose, four doses) with to allow analysis of the results obtained. BMD values with mesna uroprotection (50 mg/kg, administered as described a Z-score above –1 were considered to be normal. We above). considered osteopenia as a Z-score between –1 and –2 s.d. Recipients of unrelated donor or mismatched family and osteoporosis as a Z-score more than 2 s.d. below the member grafts received antithymocyte globulin to enhance mean (as modified from the World Health Organization immunosuppression. Marrow from mismatched family and National Osteoporosis Foundation criterion for members and from unrelated donors was depleted of osteopenia).10–12 At our institution, we initiate nutritional T lymphocytes by a factor of approximately 1.5 logs using counseling and suggest calcium and vitamin D supplemen- anti-T-cell monoclonal antibodies CD6 and CD8 plus tation for all patients with a BMD Z-score below –1 s.d.; all 5 rabbit complement. All recipients began receiving cyclo- patients with a BMD Z-score below –2 s.d. are also sporin A 2 days before transplantation, at a dosage referred for endocrinologic consultation for more rigorous adjusted to maintain plasma levels of 200–350 ng/ml. evaluation of risk factors that may contribute to BMD Patients given matched sibling marrow received additional deficits. Patients who have undergone BMT are evaluated GVHD prophylaxis with pentoxifylline or methotrexate annually in the Endocrine Clinic. (15 mg/m2 at day þ 1 followed by 10 mg/m2 at days þ 3, þ 6, and þ 11). Statistical analysis Acute and chronic GVHD were graded according to standard criteria.6 Patients with mild/moderate GVHD We used univariate analysis and logistic regression to assess (grade 1 or 2) were treated with methylprednisolone (1– if the following factors: age at BMT, age at BMT by group 2 mg/kg daily). If GVHD remained at grade 2 for more (o10 year vs 410 years), time from completion of therapy, than 7 days or progressed, the steroid dosage was increased gender, race (black vs white vs others), diagnosis (malignant to 5–10 mg/kg daily. vs nonmalignant), T-cell depletion (yes vs no), TBI (yes vs no), BMT type (matched sibling vs matched unrelated vs QCT technique donor mismatched family donor), acute GVHD (grade 2–4 vs others), chronic GVHD (yes vs no), hypothyroidism Vertebral trabecular BMD was determined by QCT (yes vs no), hypogonadism (yes vs no), adrenal insufficiency imaging using a Siemens Somatom-Plus spiral CT (yes vs no), and growth hormone deficiency (yes vs no) were scanner (Siemens, Iselin, NY, USA) and Mindwaves associated with osteopenia/osteoporosis. QCT Calibration Phantoms and software (Mindwaves The generalized linear model (GLM) was used to further Software, Inc., South San Francisco, CA, USA). Bone investigate if the following factors were correlated with mineral content was determined through direct axial images QCT Z-scores, while ignoring the definition of osteopenia of the centers of the bodies of the first and second and osteoporosis: age at BMT, age at BMT by group (o10 lumbar vertebrae (L1 and L2, respectively), as determined year vs 410 years), time from completion of therapy, from a sagittal or coronal scout image as described gender, race (black vs white vs others), diagnosis (malignant previously.6–9 If either or both bodies showed evidence vs nonmalignant), T-cell depletion (yes vs no), TBI (yes vs of fracture or deformity, the affected vertebral bodies no), BMT type (matched sibling vs matched unrelated were excluded from analysis and an alternate verte- donor vs mismatched family donor), acute GVHD (grade bral body between T11 and L4 was chosen. BMD was 2–4 vs others), chronic GVHD (yes vs no), hypothyroidism calculated as the mean mg/cm3 value of the two vertebral (yes vs no), hypogonadism (yes vs no), adrenal insufficiency bodies. (yes vs no), growth hormone deficiency (yes vs no).