Article Tools
Radiation Oncology
Article Tools
Regression After Whole-Brain Radiation Therapy for Brain Metastases Correlates With Survival and Improved Neurocognitive Function
Abstract
Brain metastasis (BM) is a major cause of suffering and health costs in cancer patients. Whole-brain radiation therapy (WBRT) offers tumor shrinking and palliation in many cases, but it has been speculated that these benefits may be outweighed by adverse effects on neurocognitive function (NCF).
Two hundred eight BM patients from the WBRT arm of phase III trial PCI-P120-9801 evaluating motexafin gadolinium were analyzed. NCF, assessed by tests of memory, executive function, and fine motor coordination, was correlated to magnetic resonance imaging–measured BM volume. NCF and survival were compared in 135 patients assessable at 2 months with tumor shrinkage below (poor responders) and above (good responders) the population median (45%). Mean NCF scores and BM volume at 4 and 15 months were compared.
Good responders experienced a significantly improved survival (unidirectional P = .03). For all tests, the median time to NCF deterioration was longer in good compared with poor responders, with statistical significance seen for Trailmaking B (executive function), and two Pegboard tests (fine motor). In long-term survivors, tumor shrinkage significantly correlated with preservation of executive function and fine motor coordination (r = 0.68 to 0.88). During the early follow-up period, the population mean NCF scores were dominated by patients with progressive disease. A small subset of 15-month survivors had stable or improving scores, and greater mean BM reduction.
Brain metastasis (BM), which occurs in 10% to 30% of adult cancer patients, is an important cause of morbidity and mortality.1 The annual incidence in the United States, currently estimated at 170,000, is presumed to be rising as a result of an aging population, better treatment of systemic disease, and utilization of improved imaging techniques earlier in the staging process.1 The prognosis of patients with BM is generally poor; with median survival time of 1 month in untreated patients2 and 4 to 6 months in treated patients.3 Newer and more aggressive treatment approaches have demonstrated survival benefit in subpopulations of BM patients.4 In this context, neurocognitive function (NCF) and quality-of-life (QOL) issues have become important concerns for survivors. Recently, a battery of validated, language-specific, and population-normalized NCF tests evaluating memory, fine motor coordination, and executive functions has been established. This new instrument has been shown to measure NCF changes in BM patients more accurately and comprehensively.5
Whole-brain radiation therapy (WBRT) is a major treatment modality for unresectable BM, especially in patients with multiple metastases in whom surgery or radiosurgery is of limited use. Many patients whose baseline NCF is already impaired6,7 are concerned about possible worsening of their NCF after WBRT. Few prospective trials have evaluated the neurocognitive impact of WBRT in BM patients. One retrospective report associating dementia with WBRT in long-term survivors is frequently quoted as justification for the avoidance of WBRT, even though it is a study without baseline testing and using nonconventional radiation dose fractionation.8 Recent studies have suggested potential benefit of WBRT on NCF preservation. One prospective Radiation Therapy Oncology Group trial, 91-04, demonstrated that the average reduction in mini-mental status examination (MMSE) scores for patients whose BMs were radiographically controlled, was less than for those with uncontrolled metastases, suggesting that progression of BM explains some of the NCF impairment after WBRT.9 However, the MMSE does not measure NCF in specific domains.10 Another recent publication from a phase III BM trial evaluating WBRT with or without motexafin gadolinium (MGd; PCI-P120-9801) showed that neurocognitive decline correlated with tumor growth.6 However, the authors did not evaluate NCF beyond 2 months post-WBRT, and NCF is of particular concern in long-term survivors, with selected patients potentially surviving for a long time. In summary, few prospective studies that specifically address the effects of WBRT on NCF, and aim to tease out whether disease progression or WBRT is the cause of NCF deterioration in these patients, have been published.
In this article, we present the data from 208 patients in the control arm of a phase III BM trial (PCI-P120-9801) in which all of the patients received WBRT and were prospectively evaluated using validated NCF tools on a regular basis from enrollment to death.6,7,11 We believe that this is one of the first few analyses of prospectively recorded data to correlate NCF directly with volume regression after WBRT in BM patients.
Two hundred eight patients in this analysis were from the control arm of a prospective, open-label, multicenter clinical trial (PCI-P120-9801) investigating the effects of adding MGd to WBRT in patients with unresectable BM. Detailed information on the primary study outcomes has been published previously.6,7,11 Adult patients were eligible to participate if they had radiologically demonstrated BM from histologically proven solid tumor, required WBRT, and had a Karnofsky performance score (KPS) ≥ 70. All patients received 10 daily fractions of 3 Gy WBRT.
A battery of standardized neurocognitive tests was administered monthly for 6 months and then every 3 months until death by trained and certified nurses or clinical research associates.6,7,11 This includes tests of memory (recall, delayed recall, recognition), verbal fluency (controlled oral word association [COWA]), fine motor coordination (Pegboard Nondominant Hand [NDH] and Pegboard Dominant Hand [DH]), and executive function (Trailmaking A and Trailmaking B).5 All tests were scored centrally by a blinded reviewer to avoid potential bias.
Magnetic resonance imaging scans were obtained at baseline; at 2, 4, and 6 months; and every 3 months thereafter. The scans were reviewed centrally in a blinded manner. Up to six brain metastases were selected at presentation as indicator lesions; the sum of the volumes of these is referred in the text and tables herein as “indicator lesion volume.”6,7
Patients were classified as good (n = 66) versus poor (n = 65) responders on the basis of whether their indicator lesion volume reduction at 2 months was above or below the population median reduction of 45%. For each specific NCF test, deterioration was defined as a reduction of the test score of more than two standard deviations from the patient's own baseline value on two consecutive measurements or on the last follow-up visit before death, using an age-adjusted normative distribution from an unimpaired population. The time to NCF deterioration was calculated from the Kaplan-Meier estimates. Patients were censored if they did not experience deterioration at their last visit or before death. In this case, the censoring time was defined as time to the last visit or time to death. Kaplan-Meier analysis was performed comparing survival and time to NCF deterioration. A two-tailed sign test was used to evaluate the overall significance of the observed difference in median time to NCF deterioration. A unidirectional P value of less than .05 is considered statistically different when two subgroups were compared.
In patients who survived to the 15th month, the Spearman correlations between percentage of tumor volume reduction from patients' own baseline and decline in NCF test scores were calculated, using measurements collected at the 15th month. A unidirectional P value less than .05 was considered statistically significant.
Mean tumor volume from indicator lesions at different time points was calculated on the basis of magnetic resonance imaging. The statistical significance of change of mean tumor volume over time was analyzed using Spearman's correlation. A two-sided P value less than .05 was considered significant. For patients who were available for follow-up at the fourth month (n = 84) and those at the 15th month (n = 9), tumor volume at different time points was normalized to patient's own baseline values before the mean was calculated and compared.
All of the statistical analyses were carried out using SPSS (SPSS Inc, Chicago, IL).
Among 208 patients enrolled onto the study, 135 were available for evaluation at 2 months. To test the hypothesis that patients with greater tumor shrinkage after WBRT survive longer and preserve their NCF better, we divided patients equally into good and poor responders as described herein. Figure 1A shows that good responders survived significantly longer than did poor responders (median survival, 300 ± 26 v 240 ± 19 days; P = .03). One-year survival (± 1 SE of the estimate) in these 135 patients was 28% ± 6% in the good responders compared with 16% ± 5% in the poor responders.
Time to NCF deterioration was compared in the good and poor responders using the Kaplan-Meier estimate. Figure 1B shows that good responders had significantly prolonged median time to deterioration of fine motor coordination test Pegboard NDH as compared with the poor responders (401 ± 38 v 291 ± 38 days; P = .02). Table 1 shows that the median time to NCF deterioration for all eight tests was longer in good responders than in poor responders (two-tailed P sign test = .008). The largest gains in delay of NCF deterioration in good responders were noted in an executive function test Trailmaking B (131 days' gain; P = .02), and the fine motor coordination tests Pegboard NDH (110 days' gain; P = .02) and Pegboard DH (93 days' gain; P = .05). These results demonstrate that patients with volume regression after radiation had a longer delay in NCF deterioration, especially for executive function and fine motor coordination. Memory function appears to have a weaker association with tumor reduction, but the trend is still in favor of longer preservation of memory with greater tumor reduction, with a net gain of 61, 52, and 59 days for recall, recognition, and delayed recall, respectively.
To further test the hypothesis that greater tumor volume reduction is associated with better preservation of NCF, we calculated Spearman's correlations between tumor shrinkage from patients' own baseline values and reduction of NCF test scores in the nine patients who were surviving at the 15th month (comparing the baseline v 15th-month NCF results). As shown in Table 2, despite the limited sample size (only nine patients, < 5% of the original 208 patients, were available for analysis), we discovered statistically significant, strong correlations in four of the eight NCF tests (r = 0.68 to 0.88), including executive function (Trailmaking A and Trailmaking B) and fine motor coordination (Pegboard DH and NDH). The three memory tests showed the weakest and statistically nonsignificant correlations with tumor shrinkage. These results again demonstrate that greater tumor shrinkage is associated with better preservation of specific NCF functions, especially executive function and fine motor coordination.
To further explore the relationships between survival, tumor volume, and NCF, we characterized the mean NCF scores and mean tumor volume in patients who were surviving at early (4 months) and late (15 months) time points. We found that, at early follow-up, the population mean NCF scores were dominated by patients with progressive disease, as reflected by a sharp drop in the mean NCF scores at 4 months (Figs 2 and 3). The decline in the mean scores over time is statistically significant for memory function (delayed recall and recall) and COWA. On the contrary, in the 15-months survivors the mean NCF test scores appeared to be stable in the first few months with gradual improvement later on (Figs 4 and 5). Statistically significant improvement in the mean NCF test scores was observed in verbal fluency (COWA), executive function (Trailmaking A), and fine motor coordination (Pegboard NDH).
When the mean tumor volume in patients surviving at 4 and 15 months was compared, greater reduction was observed at 15 (79% reduction from baseline) than at 4 months (22% reduction from baseline), although the baseline volume was indistinguishable in the two subpopulations (Fig 6). Thus, patients who became long-term survivors experienced larger tumor volume reduction after WBRT, and these patients also had the best NCF outcomes.
Although deterioration of QOL and memory in BM patients receiving WBRT has been reported, no prospective randomized studies have specifically teased out the roles of tumor progression compared with WBRT in terms of their effect on NCF. Multiple factors may contribute to NCF deterioration in BM patients: disease progression, radiation, surgery, chemotherapy, medications, or paraneoplastic effects. In this regard, our results show that better control of intracranial disease by WBRT is associated with stabilization and improvement of NCF, indicating that, at least in this population, studied for up to 21 months, the beneficial effect of tumor regression from WBRT on cognition outweighs its potential harm, and disease progression is the main contributor to neurologic decline. These results are consistent with a previous report from the same trial that showed that neurocognitive decline correlates with tumor growth.6 To our knowledge, our report is the first prospective study that has used a validated and comprehensive NCF tool to fully explore the correlation between preservation of NCF and WBRT-induced tumor volume reduction in BM patients.
An interesting finding from our studies is that patients who demonstrated good radiologic response to WBRT had improvement in their executive function and fine motor coordination, but not memory, although even for this, the trends were in the right direction. This suggests that WBRT, although improving certain aspects of cognition by reducing intracranial tumor burden, may not improve memory to the same extent, and one possible explanation for this is that WBRT might specifically impair hippocampus-related functions such as memory and learning. In rats, a dose as low as 2 Gy can reduce proliferation of neural precursor cells isolated from hippocampi, but not those from whole brain.12-14 Therefore, radiation techniques that spare hippocampus or agents that prevent hippocampal damage may help to preserve memory and to reduce the risk and severity of radiation-induced dementia. Our group is currently investigating specific conformal avoidance techniques to reduce radiation doses to the hippocampus during WBRT. With intensity-modulated radiation therapy, especially tomotherapy, it is possible to create dose distributions that deliver the full prescribed dose to the majority of the brain, while keeping the radiation dose to the hippocampus relatively low.15
QOL has become an increasingly important end point in addition to conventional measurement of survival in cancer trials. Although it is commonly speculated that deterioration of NCF in BM patients who receive WBRT will negatively influence QOL, few retrospective or prospective studies have explored the relationship between NCF and QOL in this population. In trial PCI-P120-9801, QOL was monitored on a regular basis, and our preliminary analysis has shown that NCF and QOL were correlated in this patient population, suggesting that any efforts to delay NCF decline may help to preserve QOL and, therefore, improve overall patient care.
In summary, we report that WBRT-induced reduction in total BM volume is correlated with improved survival and delay in NCF deterioration. Our study supports the use of techniques to maximize intracranial control, such as WBRT in BM patients. Future study directions include identifying patients who are most likely to respond to WBRT, further exploration of the relationship between NCF and QOL, investigating hippocampus-sparing WBRT, and testing of neuroprotective agents. Also, the relative weighting of WBRT and localized RT should be investigated as a function of the number and size of metastasis at presentation.
Although all authors completed the disclosure declaration, the following authors or their immediate family members indicated a financial interest. No conflict exists for drugs or devices used in a study if they are not being evaluated as part of the investigation. For a detailed description of the disclosure categories, or for more information about ASCO's conflict of interest policy, please refer to the Author Disclosure Declaration and the Disclosures of Potential Conflicts of Interest section in Information for Contributors.
Employment: Markus Renschler, Pharmacyclics Leadership: N/A Consultant: Minesh P. Mehta, Schering Plough Stock: N/A Honoraria: N/A Research Funds: N/A Testimony: N/A Other: N/A
Conception and design: Markus Renschler, Minesh P. Mehta
Administrative support: Markus Renschler
Collection and assembly of data: Markus Renschler, Minesh P. Mehta
Data analysis and interpretation: Jing Li, Soren M. Bentzen, Minesh P. Mehta
Manuscript writing: Jing Li, Soren M. Bentzen
Final approval of manuscript: Jing Li, Soren M. Bentzen, Markus Renschler, Minesh P. Mehta
Fig 1. (A) Kaplan-Meier survival curves in subgroups of patients with different extent of tumor reduction. Patients were classified as “good” v “poor” responders depending on their volume regression at 2 moths relative to the population median value. (B) Time to deterioration of Pegboard Nondominant Hand (NDH) in subgroups of patients with different extent of tumor reduction. Patients were classified as “good” v “poor” responders depending on their volume regression at 2 months relative to the population median value. Fraction of patients who had not experienced deterioration was plotted against time using Kaplan-Meier analysis. (+) Censored patients.
Fig 2. Change of mean normalized neurocognitive function (NCF) test scores in patients who were surviving at the fourth month. The scores were normalized to individual patient's own baseline. Figure represents NCF tests where higher scores reflect better function. (A) Recall, (B) delayed recall, (C) recognition, and (D) controlled oral word association (COWA). Each data point represents the mean NCF score ± SE. r represents Spearman's correction coefficient between mean test scores and time. (*) P value < .05 is considered statistically significant.
Fig 3. Change of mean normalized neurocognitive function (NCF) scores in patients who were surviving at the fourth month. Figure represents NCF tests where lower scores reflect better function. Each data point represents the mean normalized NCF scores ± SE. The trend of mean NCF changes is reflected in r (Spearman correlation coefficient between mean test scores and time). (*) P value < .05 is considered statistically significant. (A) Pegboard Dominant Hand, (B) Pegboard Nondominant Hand, (C) Trailmaking A, and (D) Trailmaking B.
Fig 4. Change of mean normalized neurocognitive function (NCF) test scores in patients who were surviving at the 15th month. The scores were normalized to individual patient's baseline. Figure represents NCF tests where higher scores reflect better function. (A) Recall, (B) delayed recall, (C) recognition, and (D) controlled oral word association (COWA). Each data point represents the mean normalized NCF scores ± SE. The trend of mean NCF changes is reflected in r (Spearman correlation coefficient between mean test scores and time). (*) P value < .05 is considered statistically significant.
Fig 5. Change of mean normalized neurocognitive function (NCF) test scores in patients who were surviving at the 15th month. Figure represents NCF tests where lower scores reflect better function. Each data point represents the mean normalized NCF scores ± SE. The trend of mean NCF changes is reflected in r (Spearman correlation coefficient between mean test scores and time). (*) P value < .05 is considered statistically significant. (A) Pegboard Dominant Hand, (B) Pegboard Nondominant Hand, (C) Trailmaking A, and (D) Trailmaking B.
Fig 6. Change of mean tumor volume over time. (A) Reduction in mean tumor volume after whole-brain radiation therapy in patients who were alive at the fourth month v those who were alive at the 15th month. The tumor volume in both groups was normalized to patients' own baseline values. Each data point represents the mean ± SE. (B) Mean tumor volume at the time of enrollment in patients who survived to the fourth month v those who survived to the 15th month. Each data point represents the mean ± SE.
|
| Group | Memory | Verbal Fluency (COWA) | Pegboard | Executive Function | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Recall | Recognition | Delayed Recall | DH | NDH | Trailmaking A | Trailmaking B | ||||||
| Good responders | 416 ± 37 | 374 ± 37 | 431 ± 37 | 512 ± 30 | 380 ± 38 | 401 ± 38 | 391 ± 37 | 462 ± 35 | ||||
| Poor responders | 355 ± 41 | 322 ± 38 | 372 ± 39 | 441 ± 39 | 287 ± 37 | 291 ± 38 | 386 ± 36 | 331 ± 38 | ||||
| Net gain (days) | 61 | 52 | 59 | 71 | 93 | 110 | 5 | 131 | ||||
| P | .205 | .478 | .315 | .243 | .049 | .021 | .237 | .017 | ||||
| No. of patients | 131 | 131 | 131 | 131 | 132 | 132 | 132 | 131 | ||||
NOTE. Patients were classified as good versus poor responders based on whether their indicator lesion volume reduction at 2 months was above or below the population median reduction of 45%.
Abbreviations: NCF, neurocognitive function; COWA, controlled oral word association; DH, Dominant Hand; NDH, Nondominant Hand.
|
| Correlation | Memory | Verbal Fluency (COWA) | Pegboard | Executive Function | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Recall | Recognition | Delayed Recall | DH | NDH | DH | NDH | ||||||
| r | 0.034 | 0.421 | 0.282 | −0.511 | 0.884 | 0.741 | 0.678 | 0.829 | ||||
| P | .465 | .129 | .249 | .080 | .002 | .029 | .025 | .011 | ||||
| No. of patients | 9 | 9 | 8 | 9 | 8 | 7 | 9 | 7 | ||||
NOTE. In patients who survived to the 15th month, the Spearman correlations between % tumor volume reduction from patients' own baseline and decline in NCF test scores were calculated, using measurements collected at the 15th month.
Abbreviations: NCF, neurocognitive function; COWA, controlled oral word association; DH, Dominant Hand; NDH, Nondominant Hand.
Authors' disclosures of potential conflicts of interest and author contributions are found at the end of this article.
| 1. | Wen PY, Black PM, Loeffler JS: Metastatic brain cancer, in Devita V, Hellman S, Rosenberg SA (eds): Cancer: Principles and Practice of Oncology (ed 6). Philadelphia, PA, Lippincott Williams & Wilkins, , pp 2665,2001-2670 Google Scholar |
| 2. | Zimm S, Wampler GL, Stablein D, et al: Intracerebral metastases in solid-tumor patients: Natural history and results of treatment. Cancer 48::384,1981-394, Crossref, Medline, Google Scholar |
| 3. | Sundstrom JT, Minn H, Lertola KK, et al: Prognosis of patients treated for intracranial metastases with whole-brain irradiation. Ann Med 30::296,1998-299, Crossref, Medline, Google Scholar |
| 4. | Stea B, Shaw E, Fortin A, et al: Efaproxiral (efaproxyn) as an adjunct to whole brain radiation therapy for the treatment of brain metastases originating from breast cancer: Updated survival results of the randomized REACH (RT-009) study. Proceedings of the San Antonio Breast Cancer Symposium , San Antonio, TX, December 8, 2004 Google Scholar |
| 5. | Herman MA, Tremont-Lukats I, Meyers CA, et al: Neurocognitive and functional assessment of patients with brain metastases: A pilot study. Am J Clin Oncol 26::273,2003-279, Medline, Google Scholar |
| 6. | Meyers CA, Smith JA, Bezjak A, et al: Neurocognitive function and progression in patients with brain metastases treated with whole-brain radiation and motexafin gadolinium: Results of a randomized phase III trial. J Clin Oncol 22::157,2004-165, Link, Google Scholar |
| 7. | Mehta MP, Rodrigus P, Terhaard CH, et al: Survival and neurologic outcomes in a randomized trial of motexafin gadolinium and whole-brain radiation therapy in brain metastases. J Clin Oncol 21::2529,2003-2536, Link, Google Scholar |
| 8. | DeAngelis LM, Delattre JY, Posner JB: Radiation-induced dementia in patients cured of brain metastases. Neurology 39::789,1989-796, Crossref, Medline, Google Scholar |
| 9. | Regine WF, Scott C, Murray K, et al: Neurocognitive outcome in brain metastases patients treated with accelerated-fractionation vs accelerated-hyperfractionated radiotherapy: An analysis from Radiation Therapy Oncology Group Study 91-04. Int J Radiat Oncol Biol Phys 51::711,2001-717, Crossref, Medline, Google Scholar |
| 10. | Meyers CA, Brown PD: Role and relevance of neurocognitive assessment in clinical trials of patients with CNS tumors. J Clin Oncol 24::1305,2006-1309, Link, Google Scholar |
| 11. | Mehta MP, Shapiro WR, Glantz MJ, et al: Lead-in phase to randomized trial of motexafin gadolinium and whole-brain radiation for patients with brain metastases: Centralized assessment of magnetic resonance imaging, neurocognitive, and neurologic end points. J Clin Oncol 20::3445,2002-3453, Link, Google Scholar |
| 12. | Mizumatsu S, Monje ML, Morhardt DR, et al: Extreme sensitivity of adult neurogenesis to low doses of X-irradiation. Cancer Res 63::4021,2003-4027, Medline, Google Scholar |
| 13. | Monje ML, Palmer T: Radiation injury and neurogenesis. Curr Opin Neurol 16::129,2003-134, Crossref, Medline, Google Scholar |
| 14. | Monje ML, Mizumatsu S, Fike JR, et al: Irradiation induces neural precursor-cell dysfunction. Nat Med 8::955,2002-962, Crossref, Medline, Google Scholar |
| 15. | Khuntia D, Brown P, Li J, et al: Whole-brain radiotherapy in the management of brain metastasis. J Clin Oncol 24::1295,2006-1304, Link, Google Scholar |
