Abstract
Aim: Two dose groups of patients treated with stereotactic radiosurgery (SRS) alone for 1-3 brain metastases from non-small cell lung cancer (NSCLC) were compared for outcomes. Patients and Methods: Based on the SRS dose administered to the margins of the brain lesions, 46 patients were assigned to groups treated with 15-18 Gy (n=13) or with 20 Gy (n=33). Seven additional factors were investigated: age (≤58 vs. ≥59 years), gender, Karnofsky performance score (KPS 70-80 vs. 90-100), number of brain metastases (1 vs. 2-3), histology (adenocarcinoma vs. other) extracerebral metastases and interval from NSCLC diagnosis to SRS (≤6 vs. >6 months). Results: Local control rates for 15-18-Gy and 20-Gy groups were 75% and 92% at one year (p=0.043). SRS dose was significant on multivariate analysis (p=0.030). SRS dose was not associated with freedom from new brain metastases (p=0.24) or survival (p=0.37). Conclusion: SRS with 20 Gy resulted in better control of the irradiated metastases than 15-18 Gy did.
Non-small cell lung cancer (NSCLC) is the most common primary tumor leading to the development of brain metastases in adult patients with cancer (7). Most of these patients have multiple lesions and receive whole-brain radiotherapy (WBRT) alone (2-4). However, a considerable proportion of patients with brain metastases from NSCLC have a very limited number of lesions and may benefit from more intensive local treatment such as stereotactic radiosurgery (SRS) or neurosurgical resection (5, 6). Resection is usually reserved for patients with one or two cerebral lesions (7, 8). Radiosurgery, either alone or in combination with WBRT, is widely used for one to three lesions (5). Several studies have shown that the addition of WBRT to SRS results in better intra-cerebral control but has no significant impact on survival (9, 10). According to one randomized trial of 58 patients, the cost of the positive effect on intracerebral control is worse neurocognitive function (11). In that trial, the mean probability of decline in learning and memory function at four months was 52% in patients receiving SRS-plus-WBRT compared to 24% in patients receiving SRS-alone. In contrast, a prospective study of 92 patients from a randomized trial of 132 patients showed that neurocognitive function was better after WBRT-plus-SRS than after SRS alone at one year and at two years following treatment (12). This finding could be explained by the significantly lower rate of intra-cerebral recurrences after WBRT-plus-SRS (9). Because of these contradictory findings, it is unclear whether patients with a very limited number of brain metastases benefit from the addition of WBRT to SRS. Therefore, a considerable number of these patients are treated with SRS alone. However, when SRS alone is administered, the optimal SRS dose is currently undefined. The optimal dose may vary with respect to the type of primary tumor responsible for the brain metastases. The present study compared two dose levels of SRS alone, 15-18 Gy and 20 Gy, with respect to treatment outcomes in patients with newly-diagnosed brain metastases from NSCLC.
Patients and Methods
Patients and treatment approaches. Forty-six patients treated with SRS alone for 1-3 newly-diagnosed brain metastases from NSCLC between 1999 and 2012 were included in this retrospective study. Forty-three patients received linear accelerator-based radiosurgery and two patients CyberKnife radiosurgery. Two groups were designated according to the dose given to the margin of the metastases (representing the 78% to 90% isodose level): equivalent (alpha/beta=10 Gy) to 15-18 Gy (n=13) versus 20 Gy (n=33). The difference regarding the SRS doses was due to the variation of dose concepts at the contributing centers during the time the patients were treated and the treating physician's opinions.
Both dose groups were compared for local control of the irradiated metastases, freedom from new brain metastases and survival. In addition to the SRS dose, seven further potential prognostic factors were investigated including age (≤58 vs. ≥59 years, median age=58 years), gender, Karnofsky performance score (KPS 70-60 vs. KPS 90-100), number of brain metastases (1 vs. 2-3), histology (adenocarcinoma vs. other), extracranial metastases (no vs. yes) and the interval from first diagnosis of NSCLC to SRS (≤6 vs. >6 months). Patients' characteristics of the dose groups are shown in Table I.
Both dose groups were well balanced with respect to the other seven potential prognostic factors. The median maximum diameters of the lesions were 14 mm (range=2-28 mm) in the 15-18 Gy group and 13 mm (range=3-30 mm) in the 20 Gy group, respectively. All but three lesions, one in the 15-18 Gy group and two in the 20 Gy group, were supratentorial.
Statistical considerations. The univariate analyses of local control, freedom from new brain metastases and survival were performed with the Kaplan–Meier method and the log-rank test (13). The prognostic factors that were significant or showed a trend in the univariate analysis (p<0.07) were included in a multivariate analysis performed with the Cox hazards proportional model. Patients were followed-up until death or for a median time of 13 months (range=6-30 months) in patients alive at the last follow-up.
Results
The local control rates of the entire cohort at six months and at one year following SRS were 87% and 79%, respectively. In the univariate analysis, improved local control of the irradiated metastases was significantly associated with a SRS dose of 20 Gy (p<0.001) (Figure 1). An interval between NSCLC diagnosis and SRS greater than six months was associated with a trend towards better survival when compared to an interval of six months or less (p=0.066). The results of the univariate analysis of local control are summarized in Table II. In the multivariate analysis of local control, the SRS dose [risk ratio (RR)=4.50; 95% confidence interval (CI)=1.16-21.74; p=0.030] and the interval between NSCLC diagnosis and SRS (RR=4.34; 95% CI=1.02-22.41; p=0.047) were significant.
For the entire cohort, the rates of freedom from new metastases at six months and one year following SRS were 73% and 51%, respectively. In the univariate analysis of freedom from new metastases, none of the investigated potential prognostic factors, including SRS dose (p=0.24) (Figure 1), was associated with outcome (Table III). Therefore, a multivariate analysis was not performed for this endpoint.
For the entire cohort, the survival rates at six months and one year following SRS were 70% and 58%, respectively. In the univariate analysis, improved survival was associated with absence of extracranial metastases (p=0.019) (Table IV). The SRS dose was not significantly associated with survival (p=0.37) (Figure 1). In the multivariate analysis of survival, extracranial metastases (RR=2.55; 95% CI=1.09-5.72; p=0.032) maintained significance.
Patient characteristics of the dose groups 15-18 Gy and 20 Gy.
Discussion
Patients with NSCLC represent about 40% of all patients with cancer developing brain metastases (1). Many patients with 1-3 brain metastases receive SRS either alone or in combination with WBRT. Some physicians are hesitant to add WBRT, since a small randomized trial has suggested that this approach is associated with a higher rate of neurocognitive dysfunction than SRS alone (11). When SRS alone had been administered in prior studies, generally, the median SRS dose plus the range of SRS doses were stated and similar to those of the present study. However, the optimal SRS dose is undefined.
In the present study, we compared SRS doses of 15-18 Gy to a dose of 20 Gy given to the margins of the metastatic lesions. Both dose groups were well-balanced with respect to the other seven potential prognostic factors. The median maximum diameters and the corresponding ranges were also similar in both groups. These facts likely reduce the risk of a hidden selection bias. However, due to the retrospective design of this study and the relatively small number of patients included, such a bias cannot be completely excluded.
Comparison of the two dose groups, 15-18 Gy and 20 Gy, with respect to local control of the irradiated metastases (A), freedom from new brain metastases (B) and survival (C).
According to the findings of this study, a SRS dose of 20 Gy results in significantly better local control of the irradiated metastases than doses of 15-18 Gy. The finding that a higher SRS dose is associated with better local control agrees with the results of a previous retrospective study of 86 patients (14). In 2002, a retrospective study compared <18 Gy/80%-isodose to ≥18 Gy/80%-isodose for local control of 1 to ≥4 brain metastases from NSCLC. The median follow-up in that study was six months. Local control at six months was 80% in the entire cohort. Local control rates were 57% after <18 Gy and 84% after ≥18 Gy (p<0.01). However, 20% of the patients in that study received WBRT in addition to SRS, which may have influenced the results.
Univariate analysis of local control of the irradiated metastases.
In the present study, freedom from new brain metastases and survival were not significantly different in the dose groups. This can be explained by the fact that more patients of the 20 Gy group died from new brain metastases than did those in the 15-18 Gy group. The rates of freedom from new brain metastases at one year were 72% in the 15-18-Gy group and only 41% in the 20-Gy group. The finding that the local control rate in the 20 Gy group is more than two-fold greater than the rate of freedom from new lesions supports the concept that the addition of WBRT to SRS is of value. This conclusion is strengthened by the fact that the higher rate of new brain metastases in the 20-Gy group when compared to the 15-18-Gy group led to a non-significantly worse survival at one year in the 20-Gy group (52% vs. 68% in the 15-18 Gy group).
Univariate analysis of freedom from new brain metastases (distant control).
In addition to the SRS dose, improved local control was significantly associated with a longer interval (>6 months) from the first diagnosis of NSCLC and SRS. Improved survival was significantly associated with an absence of extracranial metastases. Patients with adenocarcinoma showed a trend towards improved survival when compared to patients with other histologies in univariate analysis.
A positive association between treatment outcomes and a longer interval from the diagnosis of NSCLC and SRS has been previously reported (15). Extracranial metastases and histology have also been recognized before as prognostic factors for treatment outcomes for both NSCLC and small-cell lung cancer (15-19). The results of the present study agree with the literature.
In summary, this study suggests that a SRS dose of 20 Gy is associated with improved local control when compared to doses of 15-18 Gy. Freedom from new brain metastases and survival were not significantly different in the two groups. The fact that many patients of the 20 Gy group died from new brain metastases despite very good local control of the irradiated metastases supports the addition of WBRT to SRS in terms of achieving optimal control within the brain in spite of controversy.
Univariate analysis of survival.
Footnotes
-
Conflicts of Interest
On behalf of all Authors, the corresponding Author states that there is no conflict of interest related to this study.
- Received July 19, 2014.
- Revision received August 29, 2014.
- Accepted September 5, 2014.
- Copyright© 2014 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved
References
In this issue
Jump to section
Related Articles
Cited By...
- A Tool to Predict the Probability of Intracerebral Recurrence or New Cerebral Metastases After Whole-brain Irradiation in Patients with Head-and-Neck Cancer
- Defining the Optimal Dose of Stereotactic Radiosurgery for Treating Cerebral Metastases in Elderly Patients
- Impact of Stereotactic Radiosurgery Dose on Control of Cerebral Metastases from Renal Cell Carcinoma