Medizinische Universität Graz - Research portal
Selected Publication:
Elmer, P.
Value of prognostic scoring models in patients with spinal and extremity bone metastases.
Humanmedizin; [ Diplomarbeit ] Medizinische Universität Graz; 2024. pp.
- Authors Med Uni Graz:
- Advisor:
- Leitner Lukas
- Smolle Maria Anna
- Altmetrics:
- Abstract:
- Background: Cancer treatment has rapidly advanced over the past few decades, improving prognosis and life expectancy. Consequently, the number of patients developing bone metastases has increased, with a subsequent need for adequate therapy goals and treatment. In the decision-making progress regarding best treatment for patients with metastatic bone disease, their prognosis plays an integral role. As prognosis depends on various factors, prognostic scoring models have been developed in patients with spine and extremity metastases. To implement such a score in everyday clinical practice, it should be repeatedly validated, easy to use, and reliably and accurately estimate the patient's prognosis. In this study, we set out to assess nine different prognostic scoring models regarding their prognostic accuracy. We investigated whether the differentiation between spinal and extremity metastases commonly employed in the scores investigated is adequate, or whether a uniform applicability of the scores regardless of the development cohort is possible. Methods: Patients (n=862; median age at diagnosis 67 years; 47.6% females) treated surgically and conservatively for spinal and extremity bone metastases at a single university hospital were retrospectively included. Assessed parameters included demographic variables (e.g., age, sex), tumour-specific factors (e.g., histology of the primary tumour, visceral metastases), treatment-related variables (e.g., previous chemotherapy, surgical philosophy), and outcome-specific parameters (e.g., date of last follow-up/death). The prognostic scoring models were tested on the entire cohort (n=89), as well as patients with spine (n=445) and extremity metastases (n=417) separately. Results: In all three separate test cohorts (entire, spine, extremity) the Sorenssen model showed the best performance, followed by the Katagiri and Modified Bauer models, respectively. Our findings suggested that models based on more recent data performed better than older models. This implies the need for continuous re-evaluation of the models using up-to-date datasets. Furthermore, our study indicates superior performance of models based on a moderate number of variables (i.e., 6–8) compared with fewer (e.g., 4) variables. In addition, we discovered that the differentiation between spinal and extremity metastases may be obsolete. We identified the presence of visceral metastases, a low Karnofski performance score, and rapid tumour growth to be the most important negative predictors for overall survival. Conclusion: This study provides thorough insight into the performance of nine prognostic scoring models for patients with bone metastases. We discovered that the differentiation between spinal and extremity bone metastases in prognostic scoring models may be obsolete, given the comparable performance regardless of the underlying development cohort. This may grant the possibility to create a unified prognostic scoring model for both patients with spinal and extremity bone metastases in the future.