(P 285) A NOVEL METHOD TO PREDICT HEIGHT AT SKELETAL MATURITY IN PEDIATRIC SARCOMA PATIENTS

Objective: Sarcoma and its treatment may impede potential height attainment in skeletally immature patients. Limited research has proven or quantified this effect, posing challenges for orthopaedic oncologists who must plan surgical reconstruction of lower extremity tumors in growing children. Our objective with the present study was to answer the following questions in the light of the aforementioned knowledge gap:
1. Does pediatric sarcoma and its treatment impede adult height attainment?
2. Using existing or novel height prediction methods, is there a way to accurately predict the adult height of pediatric patients diagnosed with and treated for sarcoma to aid with surgical planning?

Methods: Two-hundred twenty-three pediatric sarcoma patients treated at a single pediatric sarcoma center were identified via retrospective chart review. Inclusion criteria were completion of chemotherapy before skeletal maturity, survival to maturity, and complete height data. Fifty-six patients met inclusion criteria (24 female, 32 male). Data encompassed cancer type, age, height, and weight at diagnosis and maturity, and treatment characteristics. Skeletal maturity was determined via radiographs in 50% of patients, but due to missing radiographic data, age (16 for females, 18 for males) was used as a surrogate marker of skeletal maturity for the remaining 50% of patients. Center for Disease Control (CDC) Z-scores were compared at diagnosis and maturity using U-tests. Paley Multiplier Method (PMM) predictions based on height at diagnosis were compared to final attained heights using T-tests. A novel height prediction model accounting for sarcoma and its treatment was derived via multiple regression and cross-validation techniques.

Results: Among 56 included patients, 70% had primary bone tumors (62% of these involved lower extremity long bones). Osteosarcoma (41%) and Ewing’s sarcoma (36%) were the most common sarcoma types, with the femur (25%) being the most common tumor location. Doxorubicin (82%) and cyclophosphamide (61%) were the most common chemotherapeutics with treatment duration averaging 76 weeks. Females were diagnosed at a mean age of 11.4±3.9 years, reaching skeletal maturity at 16.2±1.0 years. Males were diagnosed at a mean age of 13.7±3.4 years, reaching skeletal maturity at 17.3±1.3 years.

CDC Z-scores at maturity were significantly lower than at diagnosis by an average of 0.26±0.7 (p = 0.008). Correlation tests revealed that differences in CDC Z-scores at maturity and diagnosis were greater for patients diagnosed at younger ages (R = 0.33, 95% CI 0.1-0.5, p = 0.02). The PMM overpredicted adult height in the majority of cases (75%). Heights were overpredicted by the PMM by an average of 4.3±5.9 cm in females (95% CI 1.8-6.8, p = 0.002) and 4.9±6.1 cm in males (95% CI 2.6-7.1, p = 0.0001). Correlation testing revealed that patients diagnosed at younger ages had greater discrepancies between PMM predicted heights and actual heights (R = 0.55, 95% CI 0.3-0.7, p < 0.0001).

Our novel height prediction model incorporated sex, height at time of sarcoma diagnosis, and age at first chemotherapy, which were all found to be statistically significant height predictor variables. Compared to the PMM, our model yielded adult height predictions that were most similar to actual heights attained.

Conclusion: Sarcoma and its treatment during skeletal immaturity limits adult height attainment. CDC Z-scores and the PMM appear unreliable for height predictions in this context. Our novel height prediction model, accounting for sarcoma and its treatment, provided more accurate predictions and may better assist orthopaedic oncologists during surgical planning.