Short-term survival prediction scale in patients with metastatic brain disease caused by lung and breast cancer

Abstract

To develop a prognostic scale suitable for distinguishing a group of poor prognosis with low survival prior to deciding on the appropriateness of radiotherapy. We analyzed only those patients with reliably known date of death after previous WBRT to determine objective criteria allowing WBRT abandonment. WBRT was carried out in 100 patients with non-small cell lung cancer (n=49) and breast cancer (n=51) and confirmed metastatic brain disease. All procedures have been conducted at the radiotherapy department of the Herzen Moscow Oncology Research Institute since January 2014. The prescribed dose of 3 Gy was ensured in all patients. Total focal dose of 30 Gy delivered in 10 fractions was achieved in 77 cases, 36 Gy delivered in 12 fractions - in 23 cases. Death date was recorded in all patients (n=100) by January 2020. In the electronic SPSS database, death information was digitized for each patient up to 2-24 months, respectively. We identified eight the most significant factors by using of correlation analysis: primary tumor (controlled (0), uncontrolled (1)), number of brain metastases (<17 (0), ≥17 (1)), volume of brain metastases (<48 cm3 (0) ≥48 cm3 (1)), extracranial control (no metastases (0), metastases with positive dynamics after chemotherapy (1), continued growth after chemotherapy (2)), metastatic lesion of liver and lungs, respectively (no (0), yes (1)), functional status (≥ 70% (0), ≤ 60% (1)), carcinomatosis of the meninges (no (0), yes (1)). A simple summation of digital variables for factors 1-8 in each patient resulted a prognostic scale. Low risk of early mortality after WBRT was determined by 0-3 scores, intermediate risk - 4-5 scores, high risk - 6-9 scores. According to univariate analysis (log-rank 0.000), median survival rate varied in these groups: low risk - 15.5 months (11.4-19.7), intermediate risk - 5.26 months (4.6-6.0), high risk - only 1.35 months (0.9-1.8). Only 1 out of 15 high-risk patients (6-9 scores) survived 3 months (3.25 months). Inclusion of all eight factors into multivariate analysis revealed significant impact of only risk group on short-term survival. A 3-month survival in the high-risk group was 20.6 times lower (p=0.002) compared to the low and intermediate risk groups. High significance of prognostic model and low informative value of each of the included factors emphasize the advisability of determining risk groups for short-term survival according to the suggested scale for each patient scheduled for WBRT. A simple assessment of separate predictors is pointless to decide whether WBRT is necessary.