A validation study of the ProgMets score as a survival prediction tool in spinal metastasis

The ProgMets score is a tool developed to aid survival prediction in spinal metastasis. The ProgMets score encompasses primary tumour type, extent of metastasis, serum albumin and sarcopaenia, as measured by the total psoas / L3 vertebral body ratio. This was an external validation study to assess the utility of the ProgMets score and compare it to other modern prognostic systems.

This was a retrospective cohort study of 100 consecutive patients with known spinal metastasis. Demographic, clinical and pathological data were recorded. The total psoas / L3 vertebral body ratio was calculated from cross-sectional computer tomography (CT) using Image J (U.S. National Institutes of Health, Bethesda, Maryland, USA). Bloods were recorded from the date of CT. The primary outcome measure was 1-year survival. Survival was calculated in days. Model fitting was performed with logistic regression. Statistical analysis was performed with R, version 4.4.1.

Of 100 patients, 50 were female with a mean age of 58 years (SD: 14). A total of 48 patients underwent spinal surgery for metastasis, while 52 were treated non-operatively. The 90-day, 6-month and 1-year survival was 79%, 61% and 49% respectively. Of the six survival prediction models tested, the ProgMets Score had the lowest Brier Score (0.173) and highest C statistic (0.803), indicating the greatest predictive accuracy for one year survival. Furthermore, the ProgMets score had the most predictive Brier score for 90-day survival and 6-month survival.

This study of has validated the ProgMets score as a prediction tool for survival in spinal metastasis. Of six modern prognostic systems tested, the ProgMets score had the highest predictive accuracy for one year survival.