Shorter-Course Whole-Brain Radiotherapy for Brain Metastases in Elderly Patients

Int. J. Radiation Oncology Biol. Phys., Vol. 81, No. 4, pp. e469–e473, 2011 Copyright Ó 2011 Elsevier Inc. Printed in the USA. All rights reserved 0360-3016/$ - see front matter

doi:10.1016/j.ijrobp.2011.01.058

CLINICAL INVESTIGATION

Brain

SHORTER-COURSE WHOLE-BRAIN RADIOTHERAPY FOR BRAIN METASTASES IN ELDERLY PATIENTS DIRK RADES, M.D.,*y JASMIN N. EVERS, M.D.,* THEO VENINGA, M.D.,z LUKAS J. A. STALPERS, M.D.,x RADKA LOHYNSKA, M.D.,k AND STEVEN E. SCHILD, M.D.{ *Department of Radiation Oncology, University Hospital Schleswig-Holstein, Lubeck, Germany; yDepartment of Radiation Oncology, University Hospital Hamburg-Eppendorf, Hamburg, Germany; zDepartment of Radiotherapy, Dr. Bernard Verbeeten Institute, Tilburg, Netherlands; xDepartment of Radiotherapy, Academic Medical Center, Amsterdam, Netherlands; kDepartment of Radiation Oncology, University Hospital, Prague, Czech Republic; and {Department of Radiation Oncology, Mayo Clinic Scottsdale, Arizona Purpose: Many patients with brain metastases receive whole-brain radiotherapy (WBRT) alone. Using 10  3 Gy in 2 weeks is the standard regimen in most centers. Regarding the extraordinarily poor survival prognosis of elderly patients with multiple brain metastases, a shorter WBRT regimen would be preferable. This study compared 10  3 Gy with 5  4 Gy in elderly patients ($65 years). Methods and Materials: Data from 455 elderly patients who received WBRT alone for brain metastases were retrospectively analyzed. Survival and local (= intracerebral) control of 293 patients receiving 10  3 Gy were compared with 162 patients receiving 5  4 Gy. Eight additional potential prognostic factors were investigated including age, gender, Karnofsky performance score (KPS), primary tumor, number of brain metastases, interval from tumor diagnosis to WBRT, extracerebral metastases, and recursive partitioning analysis (RPA) class. Results: The 6-month overall survival rates were 29% after 5  4 Gy and 21% after 10  3 Gy (p = 0.020). The 6-month local control rates were 12% and 10%, respectively (p = 0.32). On multivariate analysis, improved overall survival was associated with KPS $ 70 (p < 0.001), only one to three brain metastases (p = 0.029), no extracerebral metastasis (p = 0.012), and lower RPA class (p < 0.001). Improved local control was associated with KPS $ 70 (p < 0.001), breast cancer (p = 0.029), and lower RPA class (p < 0.001). Conclusions: Shorter-course WBRT with 5  4 Gy was not inferior to 10  3 Gy with respect to overall survival or local control in elderly patients. 5  4 Gy appears preferable for the majority of these patients. Ó 2011 Elsevier Inc. Elderly patients, Brain metastases, Shorter-course WBRT, Survival, Local control.

ule such as 5  4 Gy given in 1 week. However, a shortercourse program can only be recommended if it provides similar outcomes when compared to 10  3 Gy. Only a few studies compared ‘‘standard’’ WBRT with 10  3 Gy with shorter-course WBRT programs such as 1  10 Gy, 2  6 Gy, and 5  4 Gy with respect to overall survival (4–8). One study suggested a slightly better overall survival with 10  3 Gy compared to 2  6 Gy (4), whereas the other studies did not observe a significant difference (5–8). Only one study compared 10  3 Gy to 5  4 Gy for local (= intracerebral) control. No study has compared 10  3 Gy with shorter-course WBRT programs focusing specifically on elderly patients. The present study compares 10  3 Gy with 5  4 Gy for brain metastases in a series of elderly patients ($65 years) with respect to overall survival and local (= intracerebral) control.

INTRODUCTION Elderly patients are generally defined as patients age $65 years. About 70% of all cancer deaths occur in this age group and the proportion of elderly patients receiving cancer treatment is constantly increasing. By 2030, the number of elderly patients dying of cancer will have more than doubled (1, 2). Most patients with brain metastases are treated with whole-brain radiotherapy (WBRT) alone. The optimal dose–fractionation regimen of WBRT is controversial. Using 10  3 Gy given in 2 weeks is considered the standard regimen in most countries. Many patients with brain metastases have an extraordinarily poor survival prognosis. The median survival of untreated patients with multiple lesions is only about 1 month (3). In particular, these patients would benefit most from a shorter treatment sched-

Conflict of interest: none. Received Nov 23, 2010, and in revised form Jan 4, 2011. Accepted for publication Jan 18, 2011.

Reprint requests to: Dirk Rades, M.D., Department of Radiation Oncology, University Hospital Schleswig-Holstein, Campus Lubeck, Ratzeburger Allee 160, D-23538 Lubeck, Germany. Tel.: ++49-451-500-6661; Fax: ++49-451-500-3324; E-mail: [email protected] e469

e470

I. J. Radiation Oncology d Biology d Physics

METHODS AND MATERIALS This retrospective analysis included 455 elderly patients, treated between 2001 and 2010 with WBRT alone for brain metastases. All patients included in this study had brain metastases treated with WBRT alone (lateral opposed fields), no prior radiotherapy to the brain, confirmation of metastases by magnetic resonance imaging, and received dexamethasone during RT. The data were obtained from the patients, their general practitioners, treating oncologists, and patient files. The patient characteristics are summarized in Table 1. The survival and local (= intracerebral) control of 162 patients who received 5  4 Gy in 1 week were compared with 293 patients who received 10  3 Gy in 2 weeks. Time to death and time to local failure were measured from the completion of radiotherapy. Local failure was confirmed by magnetic resonance imaging. All patients treated in The Netherlands received 5  4 Gy and those patients treated in Germany or in the Czech Republic received 10  3 Gy. Thus, each series from a contributing center represented an unselected series of patients. In addition to the radiation schedule, the following potential prognostic factors were evaluated: age (65–70 years vs. >70 years), gender, Karnofsky performance score (KPS <70 versus KPS $70), primary tumor type (breast cancer vs. lung cancer vs. other tumors), number of brain metastases (1–3 vs. $4), extracerebral metastases at the time of WBRT, interval between tumor diagnosis and WBRT (#12 months vs. >12 months), and recursive partitioning analysis (RPA) class (RPA 2 vs. RPA 3). The survival of patients with brain

Volume 81, Number 4, 2011

metastases is strongly associated with the RPA class, which was derived from an analysis of three Radiation Therapy Oncology Group brain metastases trials (9). Three RPA classes were defined. RPA class 1 patients have a KPS $70 plus age <65 years, no extracerebral metastases, and a controlled primary tumor. RPA class 2 patients have a KPS $70, and at least one unfavorable prognostic factor (age $65 years, extracerebral metastases, or uncontrolled primary tumor). RPA class 3 includes all patients with a KPS <70. Overall survival and local control rates were calculated with the Kaplan-Meier method (10). The differences between the KaplanMeier curves were determined with the log–rank test (univariate analysis). The prognostic factors found to be significant (p < 0.05) or nearly significant (p < 0.06) were included in a multivariate analysis, which was performed with the Cox proportional hazards model. The comparisons of the two dose groups (5  4 Gy vs. 10  3 Gy) regarding the other potential prognostic factors (Table 1) were performed with the chi-square test. Additional subgroup analyses were performed for both RPA class 2 and RPA class 3 patients.

RESULTS Patients were followed until death or from 3 to 73 months (median, 8 months) in those patients who were alive at last follow-up. The median and 1-year survival rates were 2.5 Table 2. Results of the univariate analysis of overall survival At 6 At 12 Months (%) Months (%)

Table 1. Patient characteristics of the two treatment groups 5  4 Gy (n = 162) n (%) Age 65–70 years (n = 212) > 70 years (n = 243) Gender Female (n = 196) Male (n = 259) Karnofsky performance scale < 70 (n = 222) $ 70 (n = 233) Type of primary tumor Breast cancer (n = 76) Lung cancer (n = 223) Other tumors (n = 156) Number of brain metastases 1–3 (n = 137) $4 (n = 318) Extracerebral metastases No (n = 152) Yes (n = 303) Interval from tumor diagnosis to radiotherapy # 12 months (n = 272) > 12 months (n = 183) RPA class RPA 2 (n = 233) RPA 3 (n = 222)

10  3 Gy (n = 293) n (%)

p

75 (46) 87 (54)

137 (47) 156 (53)

0.95

65 (40) 97 (60)

131 (45) 162 (55)

0.58

77 (48) 85 (52)

145 (49) 148 (51)

0.82

23 (14) 77 (48) 62 (38)

53 (18) 146 (50) 94 (32)

0.67

43 (27) 119 (73)

94 (32) 199 (68)

0.53

58 (36) 104 (64)

94 (32) 199 (68)

0.68

93 (57) 69 (43)

179 (61) 114 (39)

0.68

85 (52) 77 (48)

148 (51) 145 (49)

0.82

Abbreviation: RPA = recursive partitioning analysis. The p value was obtained from the chi-square test of the comparison of the two treatment groups regarding the distribution of the other potential prognostic factors for treatment outcome.

Radiation schedule 5  4 Gy (n = 162) 10  3 Gy (n = 293) Age 65–70 years (n = 212) > 70 years (n = 243) Gender Female (n = 196) Male (n = 259) Karnofsky performance scale < 70 (n = 222) $ 70 (n = 233) Type of primary tumor Breast cancer (n = 76) Lung cancer (n = 223) Other tumors (n = 156) Number of brain metastases 1-3 (n = 137) $4 (n = 318) Extracerebral metastases No (n = 152) Yes (n = 303) Interval from tumor diagnosis to radiotherapy # 12 months (n = 272) > 12 months (n = 183) RPA class RPA 2 (n = 233) RPA 3 (n = 222) All patients (n = 455)

p

29 21

15 11

0.020

30 18

14 14

0.021

27 22

16 10

0.17

5 43

2 23

32 23 21

13 12 13

0.35

39 18

26 7

<0.001

36 18

21 8

<0.001

22 27

12 14

0.18

43 5 15

23 2 11

<0.001

<0.001

Abbreviation: RPA = recursive partitioning analysis. Bold indicates that the results were significant in the univariate analysis.

Whole brain radiotherapy for elderly patients d D. RADES et al.

Table 3. Results of the univariate analysis of local (= intracerebral) control

Table 5. Subgroup analysis of recursive partitioning analysis Class 3 patients

At 6 At 12 Months (%) Months (%) Radiation schedule 5  4 Gy (n = 162) 10  3 Gy (n = 293) Age 65–70 years (n = 212) > 70 years (n = 243) Gender Female (n = 196) Male (n = 259) Karnofsky performance scale < 70 (n = 222) $ 70 (n = 233) Type of primary tumor Breast cancer (n = 76) Lung cancer (n = 223) Other tumors (n = 156) Number of brain metastases 1-3 (n = 137) $4 (n = 318) Extracerebral metastases No (n = 152) Yes (n = 303) Interval from tumor diagnosis to radiotherapy # 12 months (n = 272) > 12 months (n = 183) RPA class RPA 2 (n = 233) RPA 3 (n = 222) All patients (n = 455)

p

12 10

0 1

0.32

18 5

1 1

0.001

15 8

1 1

0.18

11 21

0 2

<0.001

26 11 5

3 1 0

0.054

16 9

0 1

0.001

11 11

1 1

0.49

10 13

2 0

0.39

21 11 11

2 0 1

<0.001

Abbreviation: RPA = recursive partitioning analysis.

months and 13% respectively. On univariate analysis (Table 2), improved overall survival was significantly associated with WBRT regimen 5  4 Gy (vs. 10  3 Gy, p = 0.020, Fig. 1), age of 65–70 years (vs. >70 years, p = 0.021), KPS $70 (vs. <70, p < 0.001), one to three brain metastases (vs. $4, p < 0.001), no extracerebral metastases (vs. yes, p < 0.001), and RPA class 2 (vs. RPA class 3, p < 0.001). On multivariate analysis, KPS (p < 0.001), number of brain metastases (p = 0.029), extracerebral metastasis (p = 0.012), and RPA class (p < 0.001) maintained significant, whereas the WBRT regimen (p = 0.13) and age (p = 0.66) did not achieve significance. On univariate analysis (Table 3), improved local control was significantly associated with age of 65–70 Gy (vs. >70

Overall survival 5  4 Gy (n = 85) 10  3 Gy (n = 148) Local control 5  4 Gy (n = 85) 10  3 Gy (n = 148)

At 12 Months (%)

At 6 Months (%)

At 12 Months (%)

p

4 5

0 5

0.78

0 1

0 0

0.11

Overall survival 5  4 Gy (n = 77) 10  3 Gy (n = 145) Local control 5  4 Gy (n = 77) 10  3 Gy (n = 145)

years, p = 0.001), KPS $70 (vs. <70, p < 0.001), one to three brain metastases (vs. $4, p = 0.001), and RPA class 2 (vs. RPA Class 3, p < 0.001). Breast cancer (vs. lung cancer vs. other tumors) was almost significantly associated with better local control (p = 0.054). The WBRT regimen was not associated with local control (p = 0.32, Fig. 2). On multivariate analysis, KPS (p < 0.001), type of primary tumor (p = 0.029), and RPA class (p < 0.001) maintained significant, whereas age (p = 0.11) and number of brain metastases (p = 0.12) did not achieve significance. Acute WBRT related toxicity was mild in both groups. However, the patients treated with 5  4 Gy received higher doses of dexamethasone than those treated with 10  3 Gy (median 24 mg per day vs. median 20 mg per day). In the subgroup analysis of RPA class 2 patients, overall survival was associated with the WBRT regimen favoring 5  4 Gy (p = 0.004), and a trend toward better local control was observed with 5  4 Gy (p = 0.11). In the subgroup analysis of RPA class 3 patients, the WBRT regimen was not associated with overall survival (p = 0.78) and local control (p = 0.60). The results of both subgroup analyses are summarized in Table 4 (RPA class 2 patients) and Table 5 (RPA class 3 patients). DISCUSSION This is the first study in elderly patients that compared ‘‘standard’’ WBRT (10  3 Gy given in 2 weeks) to a shorter-course WBRT program (5  4 Gy given in 1 1.0 0.9 0.8

P=0.020

0.7 0.6 0.5 0.4

Table 4. Subgroup analysis of recursive partitioning analysis Class 2 patients At 6 Months (%)

e471

p

54 37

30 19

0.004

33 19

0 2

0.11

0.3 0.2 0.1 0.0

5x4 Gy 10x3 Gy 0

10

20

30

40

50

60

70

time to death (months)

Fig. 1. Comparison of 5  4 Gy and 10  3 Gy with respect to overall survival. (p = log–rank test, univariate analysis).

I. J. Radiation Oncology d Biology d Physics

e472

1.0 0.9 0.8

P=0.32

0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0

5x4 Gy 10x3 Gy 0

10

20

time to recurrence (months)

Fig. 2. Comparison of 5  4 Gy and 10  3 Gy with respect to local (= intracerebral) control. (p = log–rank test, univariate analysis).

week). The prognosis of elderly patients with brain metastases is extraordinarily poor with a median survival time of only 2.5 months. A shorter course of WBRT allows patients to spend the smallest proportion of their limited lifespans receiving therapy. One week of treatment includes about 10% of the remaining median lifetime of elderly patients with brain metastases. However, a shorter course of WBRT is only optimal if it resulted in similar outcomes in terms of overall survival and local (= intracerebral) control as the longer program. One might expect a better treatment effect with higher total radiation doses. However, the biological effectiveness of radiotherapy depends on both the total dose and the dose per fraction. Different radiation schedules can be compared with the equivalent dose in 2 Gy fractions (EQD2), which takes into account both the total dose and the dose per fraction (11). The EQD2 is calculated with the equation EQD2 = D x [(d + a/b) / (2 Gy + a/b)], as derived from the linearquadratic model; D = total dose, d = dose per fraction, a = linear (first-order dose-dependent) component of cell killing, b = quadratic (second-order dose dependent) component of cell killing, a/b ratio = the dose at which both components of cell killing are equal. Assuming an a/b ratio of 10 Gy for tumor cell kill, the EQD2 of the radiation schedules are 23.3 Gy (5  4 Gy) and 32.5 Gy (10  3 Gy). In contrast to these expectations, the data of the present study did not demonstrate a significantly worse overall survival or local control with 5  4 Gy when compared to 10  3 Gy. On univariate analysis of overall survival, 5  4 Gy appeared even more effective than 10  3 Gy in the entire cohort and in the subset of RPA class 2 patients. However, on multivariate analysis, the difference was no longer significant. Because 5  4 Gy was not inferior to 10  3 Gy, the shorter-course WBRT regimen appears preferable for the elderly. The retrospective design of the present study should be taken into account

Volume 81, Number 4, 2011

when interpreting the results. Retrospective studies always bear the risk of a hidden selection bias even if both treatment groups were well balanced with respect to the potential prognostic factors as in this study. Our survival results are in accordance with most other studies that compared 10  3 Gy to shorter-course programs (5–8). In 1977, Harwood et al. compared 10  3 Gy with 1  10 Gy in 101 patients and found no significant difference in median survival (4.4 vs. 4.0 months, p = 0.08) (5). Three studies were identified that compared 10  3 Gy with 5  4 Gy as done in the present study (6–8). Borgelt et al. reported median survival times ranging from 2.3 to 3.5 months in patients treated with either 10  3 Gy, 5  4 Gy, or 15  2.67 Gy (p > 0.05) (6). Chatani et al. observed 1-year survival rates of 6% after 10  3 Gy and 4% after 5  4 Gy (p = 0.94) (7). In our own previous study of 442 patients at any age with more than three brain metastases, both overall survival and local control rates were not significantly different (8). In that previous study, the 1-year overall survival rates were 12% after 10  3 Gy and 11% after 5  4 Gy (p = 0.29). The 1-year local control rates within the brain were 22% and 26%, respectively (p = 0.07). These figures when compared with the 1-year overall survival (13%) and local control rates (1%) of the present study in elderly patients demonstrate the poor prognosis of elderly patients. In the present study, overall survival was significantly associated with KPS, number of brain metastases, extracerebral metastasis, and RPA class. These findings are in accordance with overall survival data of previous studies not focusing on elderly patients. In the study of Gaspar et al., age, KPS, and lack of extracerebral metastases were significant in the RPA (9). In a previous study of our group performed to develop a new prognostic score for patients irradiated for brain metastases, overall survival was significantly associated with age, KPS, extracerebral metastases, interval tumor diagnosis to WBRT, and number of brain metastases (12). In the present study, improved local control was significantly associated with a better performance status, breast cancer, and a lower RPA class. These findings are in accordance with previous studies of our group, which is still the only group that compared 10  3 Gy to shortercourse WBRT regimens for local control within the brain (8, 12). That breast cancer is associated with better local control than other primary tumor types has been reported also for other palliative situations such as bone metastases and metastatic spinal cord compression (13, 14). In conclusion, WBRT with 5  4 Gy given in 1 week was not inferior to ‘‘standard’’ WBRT with 10  3 Gy given in 2 weeks with respect to overall survival and local (= intracerebral) control in elderly patients. Thus, 5  4 Gy appears preferable for the majority of these patients.

REFERENCES 1. Hayman JA, Langa KM, Kabeto MU, et al. Estimating the cost of informal caregiving for elderly patients with cancer. J Clin Oncol 2001;19:3219–3225.

2. Yancik R, Ries LA. Aging and cancer in America: Demographic and epidemiologic perspectives. Hematol Oncol Clin North Am 2000;14:17–23.

Whole brain radiotherapy for elderly patients d D. RADES et al.

3. Zimm S, Wampler GL, Stablein D, et al. Intracerebral metastases in solid tumor patients: Natural history and results of treatment. Cancer 1981;48:384–394. 4. Priestman TJ, Dunn J, Brada M, et al. Final results of the Royal College of Radiologists trial comparing two different radiotherapy schedules in the treatment of cerebral metastases. Clin Oncol 1996;8:308–315. 5. Harwood AR, Simson WJ. Radiation therapy of cerebral metastases: A randomized prospective clinical trial. Int J Radiat Oncol Biol Phys 1977;1091–1094. 6. Borgelt B, Gelber R, Kramer S, et al. The palliation of brain metastases: Final results of the first two studies by the Radiation Therapy Oncology Group. Int J Radiat Oncol Biol Phys 1980;6:1–9. 7. Chatani M, Matayoshi Y, Masaki N, et al. Radiation therapy for brain metastases from lung carcinoma. Prospective randomized trial according to the level of lactate dehydrogenase. Strahlenther Onkol 1994;170:155–161. 8. Rades D, Kieckebusch S, Lohynska R, et al. Reduction of the overall treatment time in patients irradiated for more than three brain metastases. Int J Radiat Oncol Biol Phys 2007;69:1509–1513.

e473

9. Gaspar L, Scott C, Rotman M, et al. Recursive partitioning analysis (RPA) of prognostic factors in three Radiation Therapy Oncology Group (RTOG) brain metastases trials. Int J Radiat Oncol Biol Phys 1997;37:745–751. 10. Kaplan EL, Meier P. Non parametric estimation from incomplete observations. J Am Stat Assoc 1958;53:457–481. 11. Joiner MC, Van der Kogel AJ. The linear-quadratic approach to fractionation and calculation of isoeffect relationships. In: Steel GG, editor. Basic clinical radiobiology. New York: Oxford University Press; 1997. p. 106–112. 12. Rades D, Dziggel L, Haatanen T, et al. Scoring systems to estimate intracerebral control and survival rates of patients irradiated for brain metastases. Int J Radiat Oncol Biol Phys, in press. 13. Steenland E, Leer J, van Houwelingen H, et al. The effect of a single fraction compared to multiple fraction on painful bone metastases: A global analysis of the Dutch Bone Metastasis Study. Radiother Oncol 1999;52:101–109. 14. Rades D, Fehlauer F, Schulte R, et al. Prognostic factors for local control and survival after radiotherapy of metastatic spinal cord compression. J Clin Oncol 2006;24:3388–3393.