Session Item

In most major cancer sites a low socioeconomic status (SES) is consistently associated with poorer survival. However, the generalizability of these results may be limited due to differences in local healthcare systems. Ideally, a patient’s SES should not affect survival. Therefore, we studied the effect of SES on survival in a cohort of brain metastasis patients.

We used a consecutive retrospective cohort of 401 patients treated with stereotactic radiosurgery for brain metastases (BM) in our institute between 2012 and 2017. Patients had one of the five most common primary tumour types: non-small cell lung carcinoma (NSCLC), breast cancer, renal cell carcinoma (RCC), melanoma and gastro-intestinal (GI) tumours. Baseline tumour-specific prognostic factors, taken from the Graded Prognostic Assessment (GPA), were collected. These were age, Karnofsky performance status, number of BMs, presence of extracranial metastases and tumour genotype.

The SES is represented by a score, based on education level, income and employment, and is determined periodically by the Netherlands Institute for Social Research. The scores are published as averages per postal code. The individual SES scores used in analysis were those most current at baseline.

A log-rank test was used to compare SES quartiles. Cox proportional hazard models were made for each tumour group, as well as for the cohort as a whole. Included were the SES score, together with the prognostic factors from the GPA.

In total, 401 patients were included. Median follow-up was 9.6 months, and 345 deaths were observed. The SES score in this cohort ranged from -2.69 to 2.62, with a mean and standard deviation comparable to that of the general population (0.41 and 0.91, respectively). Baseline characteristics per SES quartile are shown in Table 1. Individual adjusted hazard ratios for SES per primary tumour are shown in Figure 1, along with the HR for the entire cohort. The HR for SES in all tumour types is 1.00 (95% CI 0.89-1.13). Similar HRs are seen in NSCLC and GI, with broader confidence intervals for the smallest subgroups, with the one in the RCC group not including 0. However, in the full model an interaction between SES and tumour type was not significant, suggesting the prognostic effect of SES is not dependent on the primary tumour.

Unadjusted Kaplan-Meier survival curves are shown in Figure 2, showing largely comparable survival across the SES quartiles during the first months of follow-up. After 20 months the curves start to deviate. However, the log-rank test found no significant difference.

After adjusting for prognostic factors from the GPA, SES has no independent prognostic value for survival in the full cohort of patients with brain metastases undergoing radiation therapy. No robust prognostic value of SES was seen in any of the tumour-based subgroups either, although an effect in patients with a RCC cannot be excluded. This leads us to conclude that SES does not impact survival in this cohort.