A multi-institutional review of radiosurgery alone vs. radiosurgery with whole brain radiotherapy as the initial management of brain metastases

A multi-institutional review of radiosurgery alone vs. radiosurgery with whole brain radiotherapy as the initial management of brain metastases

Int. J. Radiation Oncology Biol. Phys., Vol. 53, No. 3, pp. 519 –526, 2002 Copyright © 2002 Elsevier Science Inc. Printed in the USA. All rights reser...

93KB Sizes 1 Downloads 120 Views

Int. J. Radiation Oncology Biol. Phys., Vol. 53, No. 3, pp. 519 –526, 2002 Copyright © 2002 Elsevier Science Inc. Printed in the USA. All rights reserved 0360-3016/02/$–see front matter

PII S0360-3016(02)02770-0

CLINICAL INVESTIGATION

Brain

A MULTI-INSTITUTIONAL REVIEW OF RADIOSURGERY ALONE VS. RADIOSURGERY WITH WHOLE BRAIN RADIOTHERAPY AS THE INITIAL MANAGEMENT OF BRAIN METASTASES PENNY K. SNEED, M.D.,* JOHN H. SUH, M.D.,† STEVEN J. GOETSCH, PH.D.,‡ SEEMA N. SANGHAVI, M.D.,§ RICHARD CHAPPELL, PH.D.,㛳 JOHN M. BUATTI, M.D.,¶ WILLIAM F. REGINE, M.D.,# EDUARDO WELTMAN, M.D.,** VERNON J. KING, M.D.,†† JOHN C. BRENEMAN, M.D.,‡‡ PAUL W. SPERDUTO, M.D.,§§ AND MINESH P. MEHTA, M.D.§ *Department of Radiation Oncology, University of California, San Francisco, CA; †Department of Radiation Oncology, Cleveland Clinic Foundation, Cleveland, OH; ‡San Diego Gamma Knife Center, San Diego, CA; Departments of §Human Oncology and 㛳 Biostatistics and Medical Informatics, University of Wisconsin, Madison, WI; ¶Department of Radiation Oncology, University of Iowa, Iowa City, IA; #Department of Radiation Medicine, University of Kentucky, Lexington, KY; **Department of Radiation Oncology, Hospital Israelita Albert Einstein, Sao Paulo, Brazil; ††Department of Radiation Oncology, Albany Medical Center, Albany, NY; ‡‡ Department of Radiology, University of Cincinnati, Cincinnati, OH; §§Department of Radiation Oncology, Methodist Hospital Cancer Center, Minneapolis, MN Purpose: Data collected from 10 institutions were reviewed to compare survival probabilities of patients with newly diagnosed brain metastases managed initially with radiosurgery (RS) alone vs. RS ⴙ whole brain radiotherapy (WBRT). Methods and Materials: A database was created from raw data submitted from 10 institutions on patients treated with RS for brain metastases. The major exclusion criteria were resection of a brain metastasis and interval from the end of WBRT until RS >1 month (to try to ensure that the up-front intent was to combine RS ⴙ WBRT and that RS was not given for recurrent brain metastases). Survival was estimated using the Kaplan–Meier method from the date of first treatment for brain metastases until death or last follow-up. Survival times were compared for patients managed initially with RS alone vs. RS ⴙ WBRT using the Cox proportional hazards model to adjust for known prognostic factors or Radiation Therapy Oncology Group recursive partitioning analysis (RPA) class. Results: Out of 983 patients, 31 were excluded because treatment began after 6/1/98; 159 were excluded because brain metastases were resected; 179 were excluded because there was an interval >1 month from WBRT until RS; and 45 were excluded for other reasons. Of the 569 evaluable patients, 268 had RS alone initially (24% of whom ultimately had salvage WBRT), and 301 had RS ⴙ up-front WBRT. The median survival times for patients treated with RS alone initially vs. RS ⴙ WBRT were 14.0 vs. 15.2 months for RPA Class 1 patients, 8.2 vs. 7.0 months for Class 2, and 5.3 vs. 5.5 months for Class 3, respectively. With adjustment by RPA class, there was no survival difference comparing RS alone initially to RS ⴙ up-front WBRT (p ⴝ 0.33, hazard ratio ⴝ 1.09). Conclusions: Omission of up-front WBRT does not seem to compromise length of survival in patients treated with RS for newly diagnosed brain metastases. © 2002 Elsevier Science Inc. Radiosurgery, Stereotactic radiosurgery, Brain metastases, Whole brain radiotherapy, Gamma Knife, Linac radiosurgery, Brain neoplasms.

INTRODUCTION Brain metastases are common, affecting up to 25% of all cancer patients (1, 2). For decades, the standard treatment for most patients with brain metastases has been whole brain radiotherapy (WBRT), resulting in symptomatic improvement in a majority of patients (3). However, long-term local control has been poor (0%–14% at 1 year) (4, 5), and long-term survivors experience a relatively poorly defined but real risk of neurocognitive sequelae (6, 7). Surgical

resection is generally reserved for a minority of patients with a single metastasis and good prognosis. Recently radiosurgery (RS) has emerged as another treatment modality for selected patients with single or multiple brain metastases ⬍4 cm in diameter. Generally, RS has been used in conjunction with WBRT, but limited data are becoming available on the results of RS alone. Whole brain radiotherapy complements RS by helping to control micrometastases, but it may not be necessary or effective in all patients. Single-institution retrospective reviews of patients

Reprint requests to: Penny K. Sneed, M.D., Dept. of Radiation Oncology, Room L-08 (Box 0226), University of California, 505 Parnassus Avenue, San Francisco, CA 94143-0226. Tel: (415) 353-

8900; Fax: (415) 353-8679; E-mail: [email protected] Received Mar 5, 2001, and in revised form Feb 8, 2002. Accepted for publication Feb 11, 2002. 519

520

I. J. Radiation Oncology

● Biology ● Physics

with newly diagnosed brain metastases have confirmed a significantly increased risk of intracranial failure among patients managed initially with RS alone vs. RS ⫹ WBRT (8, 9), although one study showed that brain control after allowing for salvage therapy was the same for the two approaches (9). None of the single-institution studies have shown a significant survival difference comparing RS alone initially to RS ⫹ up-front WBRT (8 –11), although one study showed a trend toward improved survival for RS alone initially (8), and another study showed a disturbing trend toward worse survival after RS alone among patients with no known extracranial disease (11). To further investigate the question of whether omission of up-front WBRT compromises survival of patients treated with RS for brain metastases, the present study was performed by analyzing data from 10 institutions. In the first analysis of these data, the focus was on evaluating the outcome of patients treated with RS ⫹ WBRT. In that study, 242 patients were excluded because WBRT was never given; furthermore, patients treated initially with RS alone who received “delayed” WBRT were included, along with patients who were treated initially with RS ⫹ WBRT (12). For the present study, different exclusion and inclusion criteria were used to compare survival times for patients with newly diagnosed brain metastases managed with RS alone initially vs. RS ⫹ up-front WBRT. Available data were too limited to permit analysis of quality of life, intracranial freedom from progression, or the success of salvage therapy. METHODS AND MATERIALS Data collection Data on patients treated with RS alone or RS ⫹ WBRT were gathered from 10 different institutions (Albany Medical Center, Hospital Israelita Albert Einstein in Brazil, University of Cincinnati, Cleveland Clinic Foundation, Methodist Hospital Cancer Center in Minneapolis, San Diego Gamma Knife Center, University of California San Francisco, University of Florida, University of Kentucky, and the University of Wisconsin). In the original analysis of these data, patients who never received WBRT were excluded, and patients who received either immediate or “delayed” WBRT were included (12). Data collected included the following: date of birth; Karnofsky performance status (KPS); gender; date of diagnosis of brain metastases; date of resection of brain metastasis (if any); WBRT dates, dose, and number of fractions; date and type of RS (gamma knife or linac); whether all visible lesions were treated with RS; primary site; status of the primary tumor; presence of known extracranial metastases; and date of death or last follow-up. Control of the primary site was defined as primary tumor totally resected (unless there was evidence of recurrence), or treatment of the primary tumor completed at least 1 month before the diagnosis of brain metastases, without evidence of progression. Unknown primaries were defined as uncontrolled. For the present study, investigators were recontacted and

Volume 53, Number 3, 2002

asked to report the number of brain metastases treated and the dates of any salvage brain therapy (RS, WBRT, or brain surgery); they were also asked to confirm whether patients treated with RS followed by WBRT ⬎1.0 month later had WBRT as salvage for new, progressive, or recurrent brain metastases as opposed to planned adjuvant WBRT. Study exclusion criteria The current analysis used inclusion and exclusion criteria different from the previous analysis to compare results of RS alone initially to RS ⫹ up-front WBRT. Reasons for exclusion from the present study (in order of precedence) included (1) Start date of treatment for brain metastases (RS or WBRT) after June 1, 1998; (2) Prior WBRT for another reason (such as prophylactic WBRT or, in one case, WBRT 22 years earlier for brain metastases from a different primary site); (3) Surgical resection of a brain metastasis before RS ⫾ WBRT; (4) Interval from the end of WBRT until RS ⬎1.0 month (in which case RS may have been given as salvage therapy for new, recurrent, or progressive brain metastases rather than as part of the initial management for newly diagnosed brain metastases); (5) Interval from RS until the start of “adjuvant” WBRT ⬎1.0 month (in which case WBRT may have been given as salvage therapy for progressive or new brain metastases rather than as part of the initial management for newly diagnosed brain metastases; (6) Failure to treat all visible brain metastases with RS; (7) Treatment with partial brain radiotherapy instead of WBRT; and (8) Inadequate data available from the treating institution. Only one reason for exclusion was recorded for each excluded patient. Patients included in the present analysis were categorized as having had RS alone initially if WBRT was never given or if WBRT began ⬎1.0 month after RS and if the institution verified that WBRT was given as salvage treatment for new, recurrent, or progressive brain metastases. Whole brain radiotherapy Whether or not up-front WBRT was given depended on institutional preference, physician preference, patient preference, and referral patterns. Among patients treated with up-front WBRT, the most common dose-fractionation schemes included 30 Gy in 10 fractions, 30 Gy in 12 fractions, 35 Gy in 14 fractions, 37.5 Gy in 15 fractions, 40 Gy in 20 fractions, and 50.4 Gy in 28 fractions. Radiosurgery technique Linac RS (13) and gamma knife RS techniques (14) and the gamma knife device (Elekta Instruments, Inc., Atlanta, GA) (15) have been described previously. No details of RS, such as dose, prescription isodose contour, number of arcs or isocenters, or size or volume of metastases treated, were collected from the treating institutions. End points and statistical methods Survival was measured from the date treatment for brain metastases began (the date of RS or the date WBRT started,

A multi-institutional review of RS ⫾ WBRT

● P. K. SNEED et al.

521

Table 1. Numbers of patients and reasons for exclusion by institution

Total number of patients Number excluded Rx began after 6/1/98 Prior WBRT Resection of brain mets RS ⬎1 month after WBRT* RS ⬎1 month before WBRT* Failure to treat all mets Partial brain RT Inadequate data Number included RS alone initially RS ⫹ upfront WBRT

Alb

AE

Cin

CCF

MM

SD

UCSF

UF

UK

UW

Total

36

79

29

128

15

116

183

166

91

140

983

1 0 0 23 1 0 0 0

8 1 15 25 0 1 0 0

1 0 5 14 0 1 0 1

0 0 15 12 5 0 0 0

2 0 1 0 0 1 0 0

0 0 15 17 0 1 0 2

7 1 31 15 1 13 2 0

0 0 53 38 2 0 0 0

5 0 17 13 0 0 0 1

7 1 7 22 2 3 0 5

31 3 159 179 11 20 2 9

0 11

16 13

3 4

66 30

0 11

73 8

81 32

7 66

18 37

4 89

268 301

* Exclusion criteria for the “RS ⫹ upfront WBRT” group. Any patients classified as having “RS alone initially” never had WBRT or had WBRT ⬎1 month after RS, with WBRT known to have been given as salvage therapy. Abbreviations: Alb ⫽ Albany Medical Center; AE ⫽ Hospital Israelita Albert Einstein; Cin ⫽ University of Cincinnati; CCF ⫽ Cleveland Clinic Foundation; MM ⫽ Methodist Hospital Cancer Center in Minneapolis; SD ⫽ San Diego Gamma Knife Center; UCSF ⫽ University of California San Francisco; UF ⫽ University of Florida; UK ⫽ University of Kentucky; UW ⫽ University of Wisconsin; Rx ⫽ treatment; WBRT ⫽ whole brain radiotherapy; mets ⫽ metastases; RS ⫽ radiosurgery; RT ⫽ radiotherapy. Table 2. Summary of patient parameters for RS alone initially vs. RS ⫹ upfront WBRT Parameter

RS alone (n ⫽ 268)

RS ⫹ WBRT (n ⫽ 301)

Age, range (median) (years) Age ⬍65 years Age ⱖ65 years KPS ⬍70 KPS ⫽ 70–80 KPS ⫽ 90–100 KPS unknown No known extracranial metastases Known extracranial metastases Status of extracranial mets unknown Primary controlled Primary uncontrolled Status of primary unknown No known extracranial disease Known extracranial disease Status of extracranial disease unknown RPA Class 1‡ RPA Class 2‡ RPA Class 3‡ RPA class unknown Primary site Breast Kidney Lung Melanoma Other/unknown Number of brain metastases 1 2 3 ⱖ4 Unknown Gamma knife radiosurgery Linac radiosurgery

15–96 (61) 150 (56%) 118 (44%) 29 (11%) 82 (31%) 155 (58%) 2 (1%) 141 (53%) 121 (45%) 6 (2%) 161 (60%) 106 (40%) 1 (⬍0.5%) 77 (29%) 188 (70%) 3 (1%) 39 (15%) 197 (74%) 29 (11%) 3 (1%)

28–79 (59) 215 (71%) 86 (29%) 9 (3%) 134 (45%) 157 (52%) 1 (⬍0.5%) 161 (53%) 126 (42%) 14 (5%) 182 (60%) 118 (39%) 1 (⬍0.5%) 97 (32%) 194 (64%) 10 (3%) 64 (21%) 222 (74%) 9 (3%) 6 (2%)

15 (6%) 34 (13%) 115 (43%) 64 (24%) 40 (15%)

35 (12%) 28 (9%) 167 (55%) 29 (10%) 42 (14%)

⬍0.001†

168 (63%) 57 (21%) 21 (8%) 21 (8%) 1 (⬍0.5%) 204 (76%) 64 (24%)

175 (58%) 70 (23%) 19 (6%) 31 (10%) 6 (2%) 89 (30%) 212 (70%)

0.37* 0.55†

p value 0.010* 0.0001* ⬍0.001† 0.64* – 0.59* – 0.93* – 0.28* – 0.0005* –



– ⬍0.0001*

* Wilcoxon rank-sum p value for ordered categories or continuous variables shown in the table. Fisher exact p value. ‡ Recursive partitioning analysis (RPA) classes defined by Gaspar et al. (17). Abbreviations: RS ⫽ radiosurgery; WBRT ⫽ whole brain radiotherapy; KPS ⫽ Karnofsky performance score; mets ⫽ metastases. †

522

I. J. Radiation Oncology

● Biology ● Physics

Volume 53, Number 3, 2002

Table 3. Survival results by subgroup Parameter (code) Age ⬍65 years (0) Age ⱖ65 years (1) KPS ⬍70 (0) KPS 70–80 (1) KPS 90–100 (2) No known extracranial mets (0) Known extracranial mets (1) Primary uncontrolled (0) Primary controlled (1) No known extracranial disease (0) Known extracranial disease (1) RPA Class 1* (1) RPA Class 2* (2) RPA Class 3* (3) Primary site Breast Kidney Lung Melanoma Other/unknown Number of brain metastases 1 2 3 ⱖ4 Gamma knife radiosurgery (0) Linac radiosurgery (1) RS alone initially (0) RS ⫹ upfront WBRT (1)

Number of patients

Median survival (months)

One-year survival (95% CI)

Univariate Cox p value

365 204 38 216 312 302 247 224 343 174 382 103 419 38

8.8 7.1 5.5 6.3 10.4 10.4 6.8 6.5 9.1 12.5 6.8 14.6 7.2 5.5

40% (34–45%) 31% (25–38%) 18% (8–32%) 26% (21–32%) 46% (40–51%) 46% (40–51%) 27% (21–32%) 30% (24–36%) 41% (35–46%) 51% (43–58%) 30% (25–35%) 56% (46–65%) 33% (29–38%) 18% (8–32%)

0.027

50 62 282 93 82

8.6 9.6 8.7 7.1 8.4

35% (22–48%) 40% (28–52%) 39% (33–44%) 32% (23–41%) 32% (22–42%)

0.55†



343 127 40 52 293 276 268 301

8.3 8.7 7.6 6.6 7.9 8.7 8.2 8.6

39% (34–44%) 39% (30–47%) 35% (21–50%) 21% (11–33%) 37% (31–42%) 36% (30–42%) 38% (33–44%) 35% (30–40%)

0.018

1.12 (1.02–1.22)

0.45

0.93 (0.79–1.11) 0.99 (0.84–1.18)

⬍0.001 ⬍0.001 0.004 ⬍0.001 ⬍0.001

0.93

Hazard ratio (95% CI) 1.22 (1.02–1.46) 0.65 (0.57–0.75) 1.57 (1.32–1.87) 0.77 (0.65–0.92) 1.61 (1.33–1.95) 1.65 (1.38–1.97)

* Recursive partitioning analysis (RPA) classes defined by Gaspar et al. (17). Log–rank p value excluding “Other/unknown” category. Abbreviations: CI ⫽ confidence interval; KPS ⫽ Karnofsky performance score; mets ⫽ metastases; RS ⫽ radiosurgery; WBRT ⫽ whole brain radiotherapy. †

whichever was earlier) until death. Patients who were alive at last follow-up were censored. Actuarial survival was calculated using the Kaplan–Meier method with 95% Greenwood confidence intervals (95% CI). To test and then adjust for the influence of prognostic factors on survival, univariate and multivariate Cox proportional hazards analyses (16) were performed for various patient subgroups, for all patients, and for patients with breast, kidney, lung, or melanoma primary sites with stratification by primary site. Parameters tested for influence on survival included age (⬍65 vs. ⱖ65 years), KPS (⬍70 vs. 70 – 80 vs. 90 –100), history of known extracranial metastases (no vs. yes), control of the primary site (no vs. yes), Radiation Therapy Oncology Group (RTOG) recursive partitioning analysis (RPA) Class 1 (age ⬍65 years and KPS ⱖ70 and no extracranial disease) vs. Class 2 (KPS ⱖ70 but age ⱖ65 years and/or extracranial disease present) vs. Class 3 (KPS ⬍70) (17), number of brain metastases (1 vs. 2 vs. 3 vs. ⱖ4), gamma knife vs. linac RS technique, and initial management (RS alone vs. RS ⫹ WBRT). Multivariate Cox proportional hazards models were used to estimate the influence of treatment approach (RS alone initially vs. RS ⫹ up-front WBRT) on outcome, adjusting for RPA class or for

age, KPS, extracranial metastases, and control of the primary site. Subgroup analyses were also performed in patients with no known extracranial metastases, those with no known extracranial disease (controlled primary site and no extracranial metastases), those with single metastases, and those with multiple metastases to test RS alone vs. RS ⫹ WBRT, adjusting for RPA class in these specific subgroups. RESULTS Patient exclusions A total of 983 raw data sheets were received. Table 1 shows the number of patients excluded and the reasons for exclusion by institution. The most common reasons for exclusion were treatment start date after June 1, 1998 (31 patients), surgical resection of a brain metastasis before RS or WBRT (159 patients), RS given ⬎1.0 month after the end of WBRT (179 patients), and failure to treat all visible metastases with RS (20 patients). Among the 179 patients excluded because of an interval between the end of WBRT and RS ⬎1.0 month, this interval ranged from 1.1 to 113.9 months (25th percentile, 1.7 months; 50th percentile, 3.4 months; 75th percentile, 7.9 months). A total of 569 patients

A multi-institutional review of RS ⫾ WBRT

● P. K. SNEED et al.

523

Table 4. Survival results by subgroups for RS alone initially vs. RS ⫹ upfront WBRT Prognostic subgroup RPA Class 1* RPA Class 2 RPA Class 3 No known extracranial mets No known extracranial disease Single brain met Multiple brain mets

Parameter

Results for RS alone initially

Results for RS ⫹ upfront WBRT

Number of patients Median survival time 1-year survival (95% CI) Number of patients Median survival time 1-year survival (95% CI) Number of patients Median survival time 1-year survival (95% CI) Number of patients Median survival time 1-year survival (95% CI) Number of patients Median survival time 1-year survival (95% CI) Number of patients Median survival time 1-year survival (95% CI) Number of patients Median survival time 1-year survival (95% CI)

39 14.0 months 56% (40–70%) 197 8.2 months 38% (31–44%) 29 5.3 months 21% (8–37%) 141 10.1 months 47% (38–55%) 77 12.5 months 50% (39–61%) 168 8.3 months 40% (32–47%) 99 6.8 months 35% (26–45%)

64 15.2 months 56% (43–67%) 222 7.0 months 29% (23–35%) 9 5.5 months 11% (1–39%) 161 10.6 months 45% (37–52%) 97 12.5 months 52% (41–61%) 175 8.4 months 38% (31–45%) 120 8.5 months 33% (24–41%)

p value 0.98 0.38 0.51 0.75 0.44 0.94 0.71

* Recursive partitioning analysis (RPA) classes defined by Gaspar et al. (17). Abbreviations: RS ⫽ radiosurgery; WBRT ⫽ whole brain radiotherapy; CI ⫽ confidence interval; mets ⫽ metastases.

were evaluable for this study, including 268 patients managed initially with RS alone and 301 patients managed initially with RS ⫹ WBRT. Patient characteristics and treatment parameters Patient characteristics are summarized in Table 2 by treatment group. Prognostic factors were worse in the RSalone group in terms of percentages of patients ⱖ65 years old and of patients with KPS ⬍70 (RPA Class 3); there were also more melanoma patients and fewer breast cancer patients than in the RS ⫹ up-front WBRT group. The percentages of patients with no known extracranial metastases and with control of the primary tumor were similar. Of the 301 patients treated with RS ⫹ up-front WBRT, 73 had RS 1–31 days (median: 7 days) before the first day of WBRT, 41 had RS during the course of WBRT, and 187 had RS 1–31 days (median: 11 days) after the last day of WBRT. The dose of WBRT ranged from 2.0 to 57.0 Gy with only 13 patients (4%) receiving less than 30.0 Gy and only 2 patients (⬍1%) receiving more than 50.4 Gy. The following fractionation schemes account for 79% of the 301 patients: 30 Gy in 10 fractions (71 patients), 30 Gy in 12 fractions (21 patients), 35 Gy in 14 fractions (15 patients), 37.5 Gy in 15 fractions (35 patients), 40 Gy in 16 fractions (14 patients), 40 Gy in 20 fractions (43 patients), 45 Gy in 25 fractions (13 patients), 46 Gy in 23 fractions (6 patients), and 50.4 Gy in 28 fractions (20 patients). Two patients had external beam boosts to total doses of 39 Gy in 13 fractions or 42.5 Gy in 15 fractions after WBRT to 30 Gy in 10 fractions. Of all 569 evaluable patients, 293 were treated with gamma knife RS and 276 with linac RS.

Survival At the time of analysis, there were 20 living “RS-alone” patients with a median follow-up of 42.5 months (range: 4.1–103.4 months) and 19 living “RS ⫹ WBRT” patients with a median follow-up of 47.4 months (range: 13.9 –105.0 months). The median survival times for the 268 patients treated with RS alone initially vs. 301 patients treated with RS ⫹ up-front WBRT were 8.2 vs. 8.6 months overall with 1-year survival probabilities of 38% (95% CI: 33%– 44%) vs. 35% (95% CI: 30%– 40%), respectively (univariate Cox proportional hazards p ⫽ 0.93, hazard ratio ⫽ 0.99) (Table 3). By RPA class, the median survival times for RS alone initially vs. RS ⫹ up-front WBRT were 14.0 vs. 15.2 months for RPA Class 1 patients (Table 4, Fig. 1a), 8.2 vs. 7.0 months for RPA Class 2 patients (Table 4, Fig. 1b), and 5.3 vs. 5.5 months for RPA Class 3 patients (Table 4, Fig. 1c), respectively. As expected, univariate analyses confirmed significant associations between improved survival and age ⬍65 years, higher KPS, no known extracranial metastases, control of the primary tumor, no known extracranial disease, and RPA class (Table 3). Survival probabilities were worse with increasing numbers of brain metastases; patients with 1, 2, 3, or ⱖ4 brain metastases had median survival times of 8.3, 8.7, 7.6, and 6.6 months, respectively (p ⫽ 0.018). There was no significant difference in survival comparing patients treated with gamma knife vs. linac RS by univariate analysis (Table 3) or multivariate analysis adjusting for other prognostic factors (data not shown). To investigate the influence of adjuvant WBRT on survival, a multivariate analysis was performed, adjusting for

524

I. J. Radiation Oncology

● Biology ● Physics

Volume 53, Number 3, 2002

Fig. 1. Kaplan–Meier survival curves for patients with newly diagnosed brain metastases by Radiation Therapy Oncology Group (RTOG) RPA class, showing no significant survival differences for RS alone initially vs. RS ⫹ up-front WBRT. The median survival times were 14.0 vs. 15.2 months for Class 1 (Fig. 1a), 8.2 vs. 7.0 months for Class 2 (Fig. 1b), and 5.3 vs. 5.5 months for Class 3 (Fig. 1c), respectively.

age (⬍65 vs. ⱖ65 years), KPS (⬍70 vs. 70 – 80 vs. 90 – 100), extracranial metastases (no vs. yes), control of the primary tumor (no vs. yes), and number of brain metastases (1 vs. 2 vs. 3 vs. ⱖ4); the analysis showed that treatment approach (RS alone initially vs. RS ⫹ up-front WBRT) did not influence survival (multivariate p ⫽ 0.49, hazard ratio ⫽ 1.07) (Table 5). Similar results were obtained adjusting for RPA class (1 vs. 2 vs. 3), yielding p ⫽ 0.33 and hazard ratio ⫽ 1.09 for RS alone initially vs. RS ⫹ up-front WBRT (Table 5). Results of multivariate analyses were similar for each major primary site and for all breast, kidney, lung, and melanoma patients with stratification by primary site (data not shown). The influence of treatment approach was also analyzed in two favorable subsets with adjustment for RPA class (predominantly to adjust for imbalances in age and KPS). Among 300 patients with no known extracranial metastases, the multivariate p value was 0.72 and hazard ratio 1.05 (95% CI: 0.82–1.33) for RS alone vs. RS ⫹ WBRT (Table 5). Among 174 patients with no known extracranial disease (controlled primary site and no extracranial metastases), the

multivariate p value was 0.73 and hazard ratio 0.94 (95% CI: 0.68 –1.30) for RS alone vs. RS ⫹ WBRT (Table 5). In addition, treatment approach (RS alone initially vs. RS ⫹ up-front WBRT) was tested with adjustment for RPA class in patients with a single brain metastasis and in those with multiple brain metastases, in case results were different for these two patient subsets. Among the 337 patients with a single brain metastasis and known RPA class, the multivariate p value was 0.53 and hazard ratio 1.07 (95% CI: 0.86 –1.34) for RS alone vs. RS ⫹ WBRT (Table 5). Among the 216 patients with multiple brain metastases and known RPA class, the multivariate p value was 0.55 and hazard ratio 1.09 (95% CI: 0.82–1.46) for RS alone vs. RS ⫹ WBRT (Table 5). Salvage therapy Dates and types of salvage therapies for brain metastases were gathered for all but 11 “RS ⫹ up-front WBRT” patients in the study; these 11 patients were dropped from the denominator for purposes of calculating the percentage of patients receiving salvage therapy. Table 6 summarizes

A multi-institutional review of RS ⫾ WBRT

● P. K. SNEED et al.

525

Table 5. Results of multivariate survival analyses Parameter 539 patients (with known KPS, number of mets, status of extracranial mets, and status of primary) Age (⬍65 vs. ⱖ65 years) KPS (⬍70 vs. 70–80 vs. 90–100) Extracranial metastases (no vs. yes) Control of the primary (no vs. yes) Number of mets (1 vs. 2 vs. 3 vs. ⱖ4) RS alone vs. RS ⫹ WBRT initially 560 patients with known RPA class RPA class (1 vs. 2 vs. 3) RS alone vs. RS ⫹ WBRT initially 300 patients with no known extracranial metastases and with known RPA class RPA class (1 vs. 2 vs. 3) RS alone vs. RS ⫹ WBRT initially 174 patients with no known extracranial disease and with known RPA class RPA class (1 vs. 2 vs. 3) RS alone vs. RS ⫹ WBRT initially 337 patients with single brain metastasis and with known RPA class RPA class (1 vs. 2 vs. 3) RS alone vs. RS ⫹ WBRT initially 216 patients with multiple brain metastases and with known RPA class RPA class (1 vs. 2 vs. 3) RS alone vs. RS ⫹ WBRT initially

Multivariate p value

Multivariate hazard ratio (95% CI)

Better prognosis association

0.12 ⬍0.001 ⬍0.001 0.004 0.078 0.49

1.16 (0.96–1.40) 0.66 (0.58–0.76) 1.62 (1.35–1.95) 0.77 (0.64–0.92) 1.09 (0.99–1.20) 1.07 (0.89–1.27)

(Not significant) Higher KPS No extracranial mets Primary controlled Smaller number (Not significant)

⬍0.001 0.33

1.68 (1.40–2.01) 1.09 (0.92–1.30)

Class 1 ⬎ 2 ⬎ 3 (Not significant)

⬍0.001 0.72

1.48 (1.20–1.83) 1.05 (0.82–1.33)

Class 1 ⬎ 2 ⬎ 3 (Not significant)

⬍0.001 0.73

1.50 (1.14–1.97) 0.94 (0.68–1.30)

Class 1 ⬎ 2 ⬎ 3 (Not significant)

⬍0.001 0.53

1.75 (1.39–2.21) 1.07 (0.86–1.34)

Class 1 ⬎ 2 ⬎ 3 (Not significant)

0.003 0.55

1.58 (1.17–2.12) 1.09 (0.82–1.46)

Class 1 ⬎ 2 ⬎ 3 (Not significant)

Abbreviations: CI ⫽ confidence interval; KPS ⫽ Karnofsky performance status; mets ⫽ metastases; RS ⫽ radiosurgery; WBRT ⫽ whole brain radiotherapy; RPA ⫽ recursive partitioning analysis.

the various salvage therapies given. Of the 268 patients managed initially with RS alone, 98 (37%) underwent one or more salvage therapies at a median of 5.7 months after RS (range: 1.0 – 40.5 months), including salvage WBRT in a total of 63 patients (24%) (Table 6). In comparison, 20 of

Table 6. Salvage therapies given in patients treated with RS alone initially vs. RS ⫹ upfront WBRT Salvage therapy or therapies WBRT alone RS alone Surgery alone WBRT ⫹ RS WBRT ⫹ surgery RS ⫹ surgery WBRT ⫹ RS ⫹ surgery Any salvage therapy Months from end of treatment until first salvage therapy, range (median)

RS alone initially (n ⫽ 268) 47 25 8 11 4 2 1 98 (37%) 1.0–40.5 (5.7)

RS ⫹ upfront WBRT (n ⫽ 290)* 2 16 1 1 0 0 0 20 (7%) 1.6–23.7 (8.0)

* Salvage therapy data missing on 11 patients treated with RS ⫹ upfront WBRT Abbreviations: RS ⫽ radiosurgery; WBRT ⫽ whole brain radiotherapy.

290 “RS ⫹ up-front WBRT” patients with available data (7%) underwent one or more salvage therapies at a median of 8.0 months after the end of initial treatment for brain metastases (range: 1.6 –23.7 months), including only 3 patients who underwent salvage repeat WBRT. DISCUSSION Many institutions that use RS in the treatment of newly diagnosed brain metastases routinely recommend up-front WBRT in conjunction with RS, although several institutions have recently reported results of RS alone initially in comparison with RS ⫹ up-front WBRT (8 –11). Joseph et al. reviewed 120 patients treated with RS for brain metastases, including 24 patients who had undergone prior WBRT, 24 treated with concurrent WBRT, 33 treated with WBRT after RS (interval not specified), and 19 who did not have WBRT. The use of WBRT did not affect survival, but it was noted that WBRT given before or concurrent with RS increased the risk of delayed radiation necrosis (10). Pirzkall et al. reviewed 236 patients with 1–3 brain metastases, KPS ⱖ50, and no previous WBRT and found no significant survival difference overall for RS ⫹ WBRT vs. RS alone (p ⫽ 0.75). However, among patients with no known extracranial disease, the median survival time was 15.4 months for the 26 patients treated with RS ⫹ WBRT

526

I. J. Radiation Oncology

● Biology ● Physics

compared to 8.3 months for the 51 patients treated with RS alone (p ⫽ 0.08) (11). A retrospective review by Sneed et al. of 105 patients with newly diagnosed brain metastases comparing RS alone initially vs. RS ⫹ up-front WBRT showed no survival difference (median survival from diagnosis of brain metastases, 11.3 vs. 11.1 months, respectively) (9). There was a twofold increased risk of developing new brain metastases in the 62 patients managed initially with RS alone, but brain control after allowing for successful salvage therapy was the same for patients managed with RS alone initially vs. RS ⫹ up-front WBRT. Another recent retrospective review of patients with newly diagnosed brain metastases by Chidel et al. showed a trend toward longer survival among the 78 patients managed initially with RS alone vs. the 57 patients treated with RS ⫹ up-front WBRT (median survival times 10.5 vs. 6.4 months, respectively; log–rank p ⫽ 0.07, multivariate p ⫽ 0.078 after adjustment for other prognostic factors) (8). Similar to the study by Sneed et al. (9), the study by Chidel et al. (8) indicated that intracranial disease-free survival was significantly worse for patients managed initially with RS alone vs. RS ⫹ WBRT (60% vs. 34% at 2 years, p ⫽ 0.027).

Volume 53, Number 3, 2002

The current study suggests that the usage of up-front WBRT does not affect the length of survival of patients with newly diagnosed brain metastases treated with RS, even in the subset of patients with no known extracranial disease, in patients with a single brain metastasis, and in patients with multiple brain metastases. The median survival times for patients treated initially with RS alone vs. RS ⫹ up-front WBRT were 14.0 vs. 15.2 months for RPA Class 1, 8.2 vs. 7.0 months for RPA Class 2, and 5.3 vs. 5.5 months for RPA Class 3 patients, respectively. Of note, length of survival was not influenced by RS technique (gamma knife vs. linac). The study is limited by the fact that selection biases may have influenced the results of this retrospective review. It is also limited by the lack of information on neurologic outcome and quality of life, particularly as they relate to sequelae of RS, sequelae and success of up-front vs. salvage WBRT, sequelae and success of other salvage therapies, and the impact of the increased risk of new brain metastases in patients managed initially with RS alone. Prospective trials following patients through salvage therapy and collecting data on intracranial disease control, survival, cost, neurologic function, and quality of life are needed to further investigate RS alone initially vs. RS ⫹ up-front WBRT.

REFERENCES 1. Hojo S, Hirano A. Pathology of metastases affecting the central nervous system. In: Takakura K, Sano K, Hojo S, et al, editors. Metastatic tumors of the central nervous system. Tokyo: Igaku-Shoin; 1982. p. 5–35. 2. Posner JB, Chernik NL. Intracranial metastases from systemic cancer. Adv Neurol 1978;19:579 –592. 3. 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– 8. 4. Kondziolka D, Patel A, Lunsford LD, et al. Stereotactic radiosurgery plus whole brain radiotherapy versus radiotherapy alone for patients with multiple brain metastases. Int J Radiat Oncol Biol Phys 1999;45:427– 434. 5. Patchell RA, Tibbs PA, Walsh JW, et al. A randomized trial of surgery in the treatment of single metastases to the brain. N Engl J Med 1990;322:494 –500. 6. Asai A, Matsutani M, Kohno T, et al. Subacute brain atrophy after radiation therapy for malignant brain tumor. Cancer 1989;63:1962–1974. 7. DeAngelis LM, Delattre JY, Posner JB. Radiation-induced dementia in patients cured of brain metastases. Neurology 1989;39:789 –796. 8. Chidel MA, Suh JH, Reddy CA, et al. Application of recursive partitioning analysis and evaluation of the use of whole brain radiation among patients treated with stereotactic radiosurgery for newly diagnosed brain metastases. Int J Radiat Oncol Biol Phys 2000;47:993–999. 9. Sneed PK, Lamborn KR, Forstner JM, et al. Radiosurgery for

10.

11.

12.

13.

14. 15.

16. 17.

brain metastases: Is whole brain radiotherapy necessary? Int J Radiat Oncol Biol Phys 1999;43:549 –558. Joseph J, Adler JR, Cox RS, et al. Linear accelerator-based stereotaxic radiosurgery for brain metastases: The influence of number of lesions on survival. J Clin Oncol 1996;14:1085– 1092. Pirzkall A, Debus J, Lohr F, et al. Radiosurgery alone or in combination with whole-brain radiotherapy for brain metastases. J Clin Oncol 1998;16:3563–3569. Sanghavi SN, Miranpuri SS, Chappell R, et al. Radiosurgery for patients with brain metastases: A multi-institutional analysis, stratified by the RTOG recursive partitioning analysis methodology. Submitted. Kooy HM. Linear accelerators in stereotactic radiosurgery. In: Alexander E, 3rd, Loeffler JS, Lunsford LD, editors. Stereotactic radiosurgery. New York: McGraw-Hill, Inc.; 1993. p. 67–76. Lindquist C. Gamma knife radiosurgery. Semin Radiat Oncol 1995;5:197–202. Wu A, Lindner G, Maitz AH, et al. Physics of gamma knife approach on convergent beams in stereotactic radiosurgery. Int J Radiat Oncol Biol Phys 1990;18:941–949. Cox DR. Regression models and life tables. J R Stat Soc 1972;34:187–220. 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.