Page 75 - Predicting survival in patients with spinal bone metastasesL

models were used to assess the effect of the molecular phenotypes on survival. Survival curves were compared using log-rank tests. Harrell’s C-statistic was used as a measure of the predictive accuracy of the model before and after adjustment. The C-statistic estimates the probability of concordance between predicted and observed responses15. A p-value of <0.05 was considered statistically significant. All analyses were performed using SPSS 20.0, Armonk NY, IBM Corp.
RESULTS
Table 1 summarizes the characteristics of the 111 patients studied, of whom 110
(99%) were female and one (1%) was male. The mean age at the start of treatment
for the spinal metastasis was 59.9 years (SD ± 13.9 years).The symptomatic SBM
were most commonly located in the thoracic (n=34, 31%) part of the spine.
Primary treatment of the spinal metastasis consisted either of radiotherapy (n=69,
62%), surgery (n=21, 19%) or systemic therapy only (n=21, 19%). Patients were V classified as luminal A (n=67, 60%), luminal B (n=9, 8%), HER2 (n=11, 10%) and
triple negative (n=24, 22%).
Median follow-up was 3.2 years (minimum 0.6, maximum 5.5 years) and there was no loss to follow-up. The overall median survival was 18.0 months (95%CI 12.7- 23.2) and there was no significant difference in median survival times between the participating centers (p=0.529) (table 2). In total 82 patients (74%) died during follow-up.
Survival analysis showed that median survival times in the luminal A, luminal B and HER2 categories were not significantly different, whereas the median survival time in the triple negative category was shorter (figure 1A and table 2). Due to the fact that there were few patients in the luminal B and HER2 categories, these were added to the luminal A group, creating a new category ‘receptor positive’ as opposed to the triple negative category.
BREAST CANCER
73