Pretransplant predictors of one-year bone mineral density change after allogeneic stem cell transplantation in Colombian adults

Technical Summary: Pretransplant Predictors of One-Year Bone Mineral Density Change After Allogeneic Stem Cell Transplantation in Colombian Adults

Problem Statement
Allogeneic hematopoietic stem cell transplantation (a-HSCT) has significantly improved survival rates for malignant and non-malignant hematologic diseases. However, it is associated with long-term complications, specifically disturbances in bone metabolism leading to reduced bone mineral density (BMD), osteoporosis, and fractures. While the mechanisms involving inflammatory cytokines (TNF-α, IFN-γ) and the impact of graft-versus-host disease (GvHD) treatment are understood, there is a lack of local data evaluating the specific impact of a-HSCT on bone health within the Colombian population. Furthermore, while global studies suggest multifactorial causes for bone loss, identifying specific pre-transplant predictors in this demographic is crucial for early risk stratification and preventive strategies.

Methodology
This study employed a retrospective cohort design involving 48 adult patients (aged ≥18) who underwent a-HSCT between 2009 and 2022 at the Fundación Oftalmológica de Santander (FOSCAL) in Bucaramanga, Colombia. From an initial pool of 123 patients, 75 were excluded due to missing one-year BMD assessments, death during follow-up, or being minors.

• Data Collection: Sociodemographic data, clinical history (including GvHD, diabetes, relapse), and paraclinical variables (serum 25-hydroxyvitamin D, calcium, TSH) were collected. Imaging variables included T-scores and Z-scores for the lumbar spine (LS) and femoral neck (FN) measured pre-transplant and 9–15 months post-transplant.
• Outcome Measures: The primary outcome was the change in BMD at the LS and FN. A composite variable for "osteoporosis/low bone mass" was created (T-score < −2.5 or Z-score < −2.0) to account for variations in reporting based on patient age or radiologist decision. Secondary outcomes included the prevalence of fractures.
• Statistical Analysis: Descriptive statistics were used for univariate analysis. Simple linear regression was performed for each independent variable against BMD change. Variables with a significance level of p < 0.2 in univariate analysis were entered into a multiple linear regression model to identify independent predictors. Analyses were conducted using STATA V15.

Key Results

• Bone Density Changes: At one year post-transplant, BMD decreased by an average of 4.7% at the lumbar spine and 9.8% at the femoral neck (p < 0.05).
• Prevalence of Bone Loss: The proportion of patients with osteoporosis or low bone mass at the femoral neck increased significantly from 4.1% pre-transplant to 29.2% post-transplant (p < 0.005). The incidence of this composite outcome was 27.1% at the femoral neck and 16.7% at the lumbar spine. Three fractures were recorded during the follow-up period.
• Predictors of Bone Loss:
  • Univariate Analysis: Glucocorticoid exposure was significantly associated with BMD loss at the femoral neck (B = −0.0086; p = 0.007) but not at the lumbar spine.
  • Multivariate Analysis: No independent predictors were identified for lumbar spine BMD change. For the femoral neck, cumulative prednisolone dose remained the only independent predictor of BMD loss (B = −0.0087; p = 0.014).
  • Dose-Response Relationship: The final model, which included cumulative prednisolone dose, type 2 diabetes (T2DM), and pre-transplant relapse, explained 19% of the variance in femoral neck BMD change. Specifically, for every 1 gram increase in prednisolone administered, femoral neck BMD decreased by approximately 0.9%.
• Other Factors: While diabetes and pre-transplant relapse were included in the final model, prednisolone exposure was identified as the strongest predictor, exceeding the contribution of the other variables.

Significance and Claims
The authors conclude that bone loss during the first year after a-HSCT is frequent and clinically relevant in this Colombian cohort. The study identifies cumulative glucocorticoid exposure as a primary determinant of this process.

• Clinical Implication: The findings support the necessity of early risk identification and the implementation of preventive strategies to preserve bone health, particularly for patients exposed to high doses of glucocorticoids.
• Limitations: The authors acknowledge limitations inherent to the retrospective design, including potential information bias and the influence of unmeasured variables. The sample size was limited, potentially leading to Type II errors. Additionally, the study lacked detailed data on pre-transplant conditioning regimens (chemotherapy and total body irradiation), preventing a precise assessment of their independent impact on bone loss compared to post-transplant immunosuppression.
• Future Directions: The paper suggests that while glucocorticoid dose is a key factor, the 19% variance explained indicates that bone loss is a multifactorial process. Future research should investigate additional factors such as immune reconstitution, nutritional status, and physical inactivity, and systematically collect data on conditioning regimens.

The study emphasizes that while the model explains a portion of the variance, the dose-response relationship between prednisolone and BMD loss provides a quantifiable parameter for clinical practice, reinforcing the need for targeted interventions in high-risk populations.