Abstract
Background/Aim: The aim of this study was to analyze the survival predictions obtained from a recent graded prognostic model developed and validated in Japan. Patients and Methods: This was a retrospective single-institution analysis of 249 patients, managed with whole-brain radiotherapy for brain metastases. The sum of scores was calculated as in the Japanese study. The following parameters were included: number of brain metastases, volume of the largest lesion, sex, Karnofsky performance status, primary cancer type, control of primary cancer, and presence of extra-cerebral metastases. Results: Median overall survival was 3.0 months (95% CI= 2.6-3.4 months). The median sum of scores was 12, range=0-29. Statistically significant differences were observed between all prognostic strata. Conclusion: The graded prognostic model is also applicable to patients treated with whole-brain rather than stereotactic radiotherapy.
Parallel to changes in the treatment of brain metastases from solid tumors, more advanced prognostic models have been developed (1-6). In recent years, different models for different types of primary tumors have been proposed, e.g., melanoma- or renal cell carcinoma-specific prognostic scores (7-12). Nevertheless, a universally applicable score can still be considered advantageous due to its less complicated assignment in a busy working environment. Both three- and four-tiered scores have been validated and sometimes used for stratification in prospective clinical studies. Recently, Sato et al. have published a graded prognostic model for patients surviving 3 years or more after stereotactic radiosurgery (13). These authors assigned scores for seven statistically significant factors, i.e., number of brain metastases 1 vs. 2-4 vs. ≥5 (score; 6/1/0), volume of the largest lesion <10 ml vs. ≥10 ml (4/0), female/male sex (5/0), Karnofsky performance status (KPS) ≥80% vs. <80% (5/0), primary cancers of breast/lung/gastrointestinal tract/other (1/0/3/0), controlled primary cancer vs. uncontrolled primary cancer (8/0) and presence of extra-cerebral metastases vs. no extra-cerebral metastases (5/0). Patients were categorized into four strata according to the sum of scores, i.e., 0-9, 10-19, 20-29 and 30-36. The 3-year survival rates ranged between 0.7 and 45.1%. The median survival was 3.6, 6.8, 15.0 and 32.8 months, respectively. Relatively similar results were obtained in patients irradiated at a different institution in the same country (Japan), i.e. in a validation cohort (13). The aim of the present study was to provide additional data about the performance of this new prognostic model. In order to challenge its validity in several ways, patients managed in a different geographical region and with a different strategy (primary whole-brain radiotherapy, WBRT) were selected.
Patients and Methods
Analogous to previous validation studies (14-16), a single-institution database that includes all patients with unresected parenchymal brain metastases from histologically verified extracranial primary tumors managed with first-line WBRT was analyzed; both completed and interrupted treatment courses were included (according to the intention-to-treat principle; no previous prophylactic or other brain irradiation). The fractionation regimen was at the discretion of the treating physician. Further treatment for new or recurrent brain metastases was individualized. The strategies consisted of salvage surgery, stereotactic radiotherapy, systemic therapy or best supportive care (BSC). Systemic treatment was usually prescribed as judged appropriate by the patients’ medical oncologists. The patients were treated between January 01, 2007 and December 31, 2019. Extracranial staging consisted of computed tomography (CT). If clinically relevant, other modalities were added to clarify CT findings, e.g., isotope bone scan, ultrasound, positron emission tomography etc. The sum of scores was calculated as described by Sato et al. (outlined in the previous paragraph) (13). Overall survival (time to death) from the first day of radiotherapy was calculated employing the Kaplan–Meier method, and different groups were compared using the log-rank test (SPSS 25, IBM Corp., Armonk, NY, USA). Only three patients were censored after median 4 months of follow-up. Date of death was known in all other patients. Univariate Cox regression was employed to analyze the validity of the prognostic score.
Results
The study included 249 patients, whose baseline data are shown in Table I. Median overall survival was 3.0 months [95% confidence interval (CI)=2.6-3.4 months]. The median sum of scores was 12, range=0-29. The Cox regression analysis showed that the prognostic score (continuous variable) predicted overall survival [p=0.0001, Chi-square 26.4, Exp(B) 0.948]. As shown in Figure 1, the Kaplan-Meier survival curves were significantly different (p=0.0001 pooled over all strata). None of the study patients had a favorable sum of scores, i.e., 30-36 points. Patients with 0 points (n=8) had a median survival of 1.8 months (maximum 3.7 months). The respective figures were 2.1 months (5.0 months) for those with 1 point (n=7).
Patient characteristics.
Actuarial overall survival stratified by the sum score (p=0.0001 pooled over all strata). Eighty-one patients had 0-9 points (median survival 2.8 months, 3-year rate 0%). The largest group (n=134) had 10-19 points (median survival 3.0 months, 3-year rate 2%). Thirty-four patients had 20-29 points (median survival 7.0 months, 3-year rate 6%). Mean survival was 3.7, 7.3 and 14.3 months, respectively. The study did not include patients with 30-36 points. Pairwise comparison resulted in p-values of 0.025 (0 vs. 1), 0.00001 (0 vs. 2) and 0.001 (1 vs. 2), respectively.
Discussion
The validity of the new four-tiered score has already been demonstrated by Sato et al. (13). Their study included patients treated with stereotactic radiosurgery and was methodologically different from previous prognostic studies because the endpoint of interest was 3-year survival. The present results confirm the validity of the score in a very different population, both geographically and with regard to radiotherapy. Sato et al. reported a median survival time of 8.0 months and a median sum of scores of 15 (13). These figures illustrate that the WBRT study patients represent a negative selection (3.0 months, sum of scores 12). The study institution’s policy was to offer surgery or stereotactic radiosurgery to patients with favorable prognostic features and brain metastases eligible for such treatment. Interestingly, the poor-prognosis group (0-9 points) had a median survival of 2.8 months, whereas Sato et al. reported 3.6 and 3.8 months, respectively (development and validation cohort). This relatively small difference increased with increasing sum of scores (20-29 points: 7.0 months after WBRT vs. 15.0 and 15.7 months in the Sato et al. study). In other words, stereotactic radiosurgery is increasingly beneficial in patients with better prognostic features, both regarding the higher likelihood of local control and the lower likelihood of neurocognitive decline [endpoints that were not studied here, but in previous clinical trials reviewed in (17-19)].
In contrast to other prognostic models (20-22), the present one is not well suited to predict extremely short survival (even patients with 0-1 points sometimes survived for more than 3 months). This fact is not surprising, given that 3-year survival was the landmark chosen by the Japanese group. Limitations of the present study include the relatively small number of patients and its retrospective design. Strengths of our study include the completeness of data (both baseline and follow-up) and the real-world setting (all age groups, different disease burden, inclusion of patients who failed to complete treatment). It is reassuring that a validation study in a vastly different setting was able to confirm the applicability of the score published by Sato et al.
Footnotes
Authors’ Contributions
CN participated in the design of the study and performed the statistical analysis. CN, BM and RY conceived the study and drafted the article. All Authors read and approved the final article.
This article is freely accessible online.
Conflicts of Interest
The Authors declare that they have no conflicts of interest.
- Received January 15, 2021.
- Revision received February 4, 2021.
- Accepted February 10, 2021.
- Copyright© 2021, International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.