A boost in addition to whole-brain radiotherapy improves patient outcome after resection of 1 or 2 brain metastases in recursive partitioning analysis class 1 and 2 patients
Dirk Rades,  Cancer 2007;110:1551

The current study was conducted to compare 2 treatment regimens including surgical resection and whole-brain radiotherapy (WBRT) for patients with 1 to 2 brain metastases. A total of 201 patients with recursive partitioning analysis (RPA) class 1 to 2 disease with 1 to 2 resectable brain metastases were analyzed retrospectively. Patients underwent either resection of the metastases plus WBRT with 10 fractions of 3 grays (Gy) each or 20 fractions of 2 Gy each (99 patients; Group A) or the same treatment plus a WBRT boost to the metastatic site (10 fractions of 3 Gy each plus 5 fractions of 3 Gy each or 20 fractions of 2 Gy each plus 5 fractions of 2 Gy each) (102 patients; Group B). Eight other potential prognostic factors were evaluated with regard to overall survival (OS), brain control (BC), and local control of resected metastases (LC): age, gender, Karnofsky performance status, extent of surgical resection, tumor type, extracranial metastases, RPA class, and interval from tumor diagnosis to WBRT.

RESULTS. Group B patients had better 1-year OS (66% vs 41%). On multivariate analysis of OS, treatment regimen (relative risk [RR] of 1.94; P < .001), extent of surgical resection (RR of 1.80; P = .001), and interval from tumor diagnosis to WBRT (RR of 1.62; P = .010) were found to be statistically significant. On multivariate analysis of BC, treatment regimen (RR of 2.15; P = .002), extent of surgical resection (RR of 2.78; P < .001), and interval from tumor diagnosis to WBRT (RR of 1.52; P = .034) were found to be statistically significant. On multivariate analysis of LC, treatment regimen (RR of 2.31; P = .002) and extent of surgical resection (RR of 3.79; P < .001) were found to be statistically significant. On RPA class subgroup analyses, outcome was found to be significantly better with a WBRT boost in both RPA class 1 and class 2 patients. A WBRT boost resulted in better outcome after both complete and incomplete surgical resection. However, the results concerning BC and LC were not found to be statistically significant if surgical resection was incomplete.

CONCLUSIONS. After surgical resection of 1 to 2 brain metastases, a boost of 10 to 15 Gy in addition to WBRT was found to improve outcome. After incomplete surgical resection, further dose escalation to the metastatic site may be considered.

Brain metastases occur in up to 40% of all adult cancer patients and occur 10 times more often than primary brain tumors. The prognosis of the majority of patients with brain metastasis is poor, with most patients surviving only 3 to 6 months. The prognosis of patients with 1 or 2 brain metastases appears to be better than that of patients with more brain metastases. Therefore, the treatment of patients with 1 or 2 brain metastases is often more aggressive than that of patients with a greater number of lesions. Whole-brain radiotherapy (WBRT) alone is the most common treatment for patients with multiple brain metastases, and is generally administered as WBRT. Uncontrolled, retrospective studies have suggested that patients with multiple brain metastases do not benefit from undergoing surgical resection in addition to WBRT. In contrast to the treatment of patients with multiple metastases, surgical resection plays an important role in the treatment of patients with 1 or 2 lesions. Two randomized trials found that surgical resection of metastasis followed by WBRT resulted in better overall survival (OS) and local control than WBRT alone for the treatment of solitary metastasis. The optimal radiation dose-fractionation schedule after the surgical resection of brain metastases remains controversial. It is possible that the treatment outcome can be improved further by escalating the WBRT dose after surgery. However, an increased dose administered to the entire brain could increase late toxicity. The risk of relevant late toxicity would be less if the increased dose was administered to the metastatic site only (WBRT boost) rather than to the whole brain. In the current study, surgery followed by WBRT alone was compared with surgery followed by WBRT plus a WBRT boost to the metastatic site in patients with 1 or 2 brain metastases.

Although the number of brain metastases may impact patient prognosis, other prognostic factors appear to be more relevant. These factors include the Karnofsky performance score (KPS), age, controlled primary tumor, and extracranial metastases. Based on a recursive partitioning analysis (RPA) of 3 Radiation Therapy Oncology Group (RTOG) brain metastases trials, 3 prognostic classes (RPA classes) were defined. With regard to the different OS prognoses in the RPA classes, the appropriate treatment regimen may vary with RPA class. Therefore, the current study compares the 2 treatment regimens with respect to outcome (OS, brain control [BC], and local control of the resected metastases [LC]) for the entire cohort and for RPA classes 1 and 2 individually.

WBRT is the most common treatment modality used for brain metastases. Using data from 3 RTOG brain metastases trials, an RPA was performed to better define OS prognosis. Three prognostic classes were defined: RPA class 1 includes a KPS 70, age <65 years, controlled primary tumor, and no extracranial metastases. RPA class 2 indicates a KPS 70, age 65 years, and/or uncontrolled primary tumor and/or extracranial metastases. RPA class 3 includes all patients with a KPS <70. On the RTOG analysis, the median OS for RPA classes 1, 2, and 3 was 7.1 months, 4.2 months, and 2.3 months, respectively.

RPA class 1 and class 2 patients with a limited number of resectable metastases were reported to benefit from neurosurgery in addition to WBRT with respect to OS and LC.  Resection followed by WBRT also has been reported to result in better treatment outcomes than surgery alone in patients with single metastasis. In a retrospective study from the Mayo Clinic, the brain recurrence rates were 85% after surgical resection alone and 21% after surgical resection plus WBRT.  The median OS times were 11.5 months and 21 months, respectively. A prospective trial reported by Patchell et al. in patients who underwent a macroscopic surgical resection of single brain metastasis found an improvement in LC with postoperative WBRT (82% vs 30%).  OS was not found to be significantly different. However, the risk of death due to neurologic causes was found to decrease with postoperative WBRT (14% vs 37%). The maximal benefit from a combined approach including surgery and WBRT when compared with WBRT alone or surgery alone was observed for RPA class 1 and RPA class 2 patients. In contrast, RPA class 3 patients were found to have a poor OS regardless of treatment type. Therefore, neurosurgery is usually limited to RPA class 1 and class 2 patients, as in the current study.

To our knowledge, the appropriate WBRT regimen as a part of a combined approach including surgery has not yet been sufficiently defined. In 3 randomized studies, the median OS after surgery followed by 30 to 40 Gy of WBRT alone was 6 to 10 months, which needs to be improved. Such an improvement may possibly be achieved with escalation of the WBRT dose.

The current study compared surgical resection of the metastases followed by 30 to 40 Gy of WBRT alone versus surgical resection of the metastases followed by 30 to 40 Gy of WBRT plus a 10 to 15-Gy boost to the metastatic site. The additional WBRT boost was associated with improved OS, BC, and LC of the resected metastases. Acute and late toxicity rates were found to be relatively low in both treatment groups, which is in accordance with the findings from the available literature. To our knowledge, there is little evidence that WBRT after the surgical resection of brain metastases increases toxicities such as neurocognitive deficits. Regine et al. found that control of brain metastases has a significant impact on neurocognitive function as measured by the Mini-Mental Status Examination (MMSE). However, the risk of cognitive dysfunction resulting from late WBRT toxicity has been reported to increase with the dose per fraction. Therefore, doses per fraction >3 Gy were not used in the current study, in particular because the OS prognosis of the patients was expected to be reasonably good.A considerable proportion of the patients are likely to survive long enough to develop late WBRT toxicity and neurocognitive dysfunction.

In the current study OS was found on univariate analysis to be significantly associated with administration of the WBRT boost, complete surgical resection, KPS, extracranial metastases, RPA class, and the interval between tumor diagnosis and WBRT. On multivariate analysis, the treatment regimen, extent of surgical resection, and the interval between tumor diagnosis and WBRT maintained statistical significance. These findings are in accordance with data regarding the OS of patients with brain metastases presented by Gaspar In that study, age, KPS, and lack of extracranial metastases were found to be significant on both univariate analysis and RPA. Prior brain surgery and the interval from tumor diagnosis to the treatment of metastases were found to be significant on the univariate analysis. The negative prognostic impact of a shorter interval from tumor diagnosis to the diagnosis and treatment of brain metastasis likely reflects the faster growth of more aggressive tumors, which has already been demonstrated for other palliative situations such as metastatic spinal cord compression.

In the current study, both treatment groups were balanced with regard to the other potential prognostic factors investigated, which considerably reduces the risk of a selection bias. However, the retrospective design of this study must be taken into account when interpreting these results. In the RPA class subgroup analyses, patients in either of the RPA classes (1 or 2) appeared to benefit from the WBRT boost in terms of OS and control of disease within the brain.

The subgroup analyses performed for different extents of resection (complete vs incomplete) demonstrated that an additional WBRT boost resulted in significantly better outcome in terms of OS, BC, and LC if a complete surgical resection of the metastases was performed. These results suggest that a boost dose of 10 to 15 Gy in addition to 30 to 40 Gy of WBRT is sufficient after complete surgical resection of the metastases. After incomplete surgical resection, the administration of an additional WBRT boost was associated with significantly better OS. However, BC and LC were not found to be significantly improved with the additional WBRT boost. Thus, if only an incomplete surgical resection can be performed, a further escalation of the WBRT dose to the metastatic site may be considered to improve control of disease within the brain. These findings can be explained by the fact that a cumulative dose of 45 to 50 Gy to the metastatic site, as administered in the current study, can be considered sufficient to control subclinical disease but not to control residual tumor.

A further dose escalation to the metastatic site may be performed either by increasing the dose of the external beam WBRT boost or with the administration of a stereotactic radiosurgery (SRS) boost. However, one needs to be aware that SRS is available in only a minority of WBRT departments worldwide.

With regard to SRS, it may be questioned whether higher doses of SRS alone, which was not administered in the current study, may result in a better outcome than surgical resection followed by conventional WBRT (WBRT with or without a WBRT boost). SRS may be indicated for the treatment of up to 3 metastases of a suitable size (4 cm). Because SRS is not available in many centers, surgical resection followed by conventional WBRT appears to be a reasonable alternative to SRS with or without WBRT. Furthermore, it is unclear whether SRS alone results in better outcome than surgery or vice versa. To our knowledge, randomized studies comparing these 2 treatment modalities are missing. The results from the few available retrospective series are controversial. In the series of 75 patients reported by Bindal et al., the median OS was 16.4 months after surgical resection versus 7.5 months after SRS. However, underdosage may have confounded the results after SRS. Muacevic et al. compared surgery plus WBRT versus SRS for single lesions measuring 3.5 cm in greatest dimension. OS and LC rates at 1 year were not found to be significantly different in both groups. Schoggl et al. compared SRS plus WBRT with surgery plus WBRT in a retrospective case-control series of 133 patients and found no significant difference with regard to median OS, but significantly better LC was noted in the SRS group.[26] Similar findings were reported by O'Neill et al. from an uncontrolled retrospective study with 97 patients, which suggested better LC after SRS than after surgical resection of solitary brain metastasis but with similar OS. The median OS after SRS alone in these studies was reported to be 7 to 12 months, which is comparable to the 9.5-month OS reported in patients in the current study who were treated without a WBRT boost (Group A) and worse than the 18-month OS noted in those patients who received the WBRT boost (Group B). Thus, surgical resection followed by WBRT plus a WBRT boost to the metastatic site should be considered an effective option in the treatment of patients with 1 or 2 brain metastases. This approach appears to be associated with better outcomes than SRS alone. However, randomized trials comparing surgery plus conventional WBRT versus treatment approaches including SRS with or without WBRT are required to determine which option is superior.

In conclusion, the results of the current study suggest that after surgical resection of 1 or 2 brain metastases, a WBRT boost to the metastatic site in addition to WBRT may improve treatment outcome in terms of OS, BC, and LC of the resected metastases, if a complete surgical resection of the metastases has been performed. After incomplete surgical resection, an WBRT boost dose of 10 to 15 Gy in addition to 30 to 40 Gy of WBRT appears to improve OS, but not to significantly improve intracerebral control. In these patients, a higher boost dose of external beam WBRT or a boost of SRS may be considered.