Clinical Investigations
Radiosurgery for brain metastases: a score index for predicting prognosis

https://doi.org/10.1016/S0360-3016(99)00549-0Get rights and content

Abstract

Purpose: To analyze a prognostic score index for patients with brain metastases submitted to stereotactic radiosurgery (the Score Index for Radiosurgery in Brain Metastases [SIR]).

Methods and Materials: Actuarial survival of 65 brain metastases patients treated with radiosurgery between July 1993 and December 1997 was retrospectively analyzed. Prognostic factors included age, Karnofsky performance status (KPS), extracranial disease status, number of brain lesions, largest brain lesion volume, lesions site, and receiving or not whole brain irradiation. The SIR was obtained through summation of the previously noted first five prognostic factors. Kaplan-Meier actuarial survival curves for all prognostic factors, SIR, and recursive partitioning analysis (RPA) (RTOG prognostic score) were calculated. Survival curves of subsets were compared by log-rank test. Application of the Cox model was utilized to identify any correlation between prognostic factors, prognostic scores, and survival.

Results: Median overall survival from radiosurgery was 6.8 months. Utilizing univariate analysis, extracranial disease status, KPS, number of brain lesions, largest brain lesion volume, RPA, and SIR were significantly correlated with prognosis. Median survival for the RPA classes 1, 2, and 3 was 20.19 months, 7.75 months, and 3.38 months respectively (p = 0.0131). Median survival for patients, grouped under SIR from 1 to 3, 4 to 7, and 8 to 10, was 2.91 months, 7.00 months, and 31.38 months respectively (p = 0.0001). Using the Cox model, extracranial disease status and KPS demonstrated significant correlation with prognosis (p = 0.0001 and 0.0004 respectively). Multivariate analysis also demonstrated significance for SIR and RPA when tested individually (p = 0.0001 and 0.0040 respectively). Applying the Cox Model to both SIR and RPA, only SIR reached independent significance (p = 0.0004).

Conclusions: Systemic disease status, KPS, SIR, and RPA are reliable prognostic factors for patients with brain metastases submitted to radiosurgery. Applying SIR and RPA classifications to our patients’ data, SIR demonstrated better accuracy in predicting prognosis. SIR should be further tested with larger patient accrual and for all patients with brain metastases subjected or not to stereotactic radiosurgery.

Introduction

Reliable prognostic factors for patients with cancer have a strong correlation with outcome, often having a greater impact on the survival than treatment itself. They are also useful in understanding the natural history of cancer, predicting the results of therapeutic interventions, identifying homogeneous patient populations, comparing treatment results, identifying subsets of patients with poor outcomes and planning follow-up strategies (1).

Brain metastases from systemic malignancies are frequent complications of malignant tumors of lung, breast, kidney, and malignant melanoma. In the United States, more than 80,000 patients are diagnosed with brain metastases every year 2, 3. Expanding tumor and surrounding peritumoral edema mass effect may be life-threatening, and its control is fundamental to the context of multidisciplinary cancer treatment 3, 4, 5, 6, 7, 8, 9.

The usual outcome of untreated patients is very poor, with a mean survival rate of only 1 month after diagnosis (6, 9) with central nervous system dysfunction as the most common cause of death 9, 10. The use of steroids to decrease the surrounding tumor edema improves neurological conditions and extends the median survival to approximately 2 months (11).

Since the publication of Chao et al. (12) in the 1950s, whole brain irradiation (WBI) has became the most common form of palliative treatment, resulting in a median survival rate of 3 to 6 months 10, 13. Retrospective studies have suggested that adjuvant WBI reduces local and distal brain failures after radiosurgery 14, 15, 16, 17. The risk of death secondary to systemic disease for patients with progressive extracranial disease may jeopardize the potential survival advantage of WBI (17).

Surgery has been used for brain metastases treatment for many years; its major advantage is the immediate relief of neurological signs and symptoms 18, 19. Patients with a solitary brain metastasis, controlled primary tumor, no evidence of systemic disease, nonradiosensitive primary tumor, and good performance status (Karnofsky ≥ 70) are the best candidates for surgical resection and WBI 13, 17, 20, 21, 22. These patients have demonstrated both longer survival and longer functional independence than patients treated with WBI alone (9).

Leksell first described stereotactic radiosurgery in 1951 (23). The purpose of this treatment was to deliver a single and focal high-dose radiotherapy. Using modern three-dimensional image processing, it has been largely applied for the treatment of small intracranial targets 2, 5, 18, 24, 25. The sharp peripheral dose falloff, resulting in minimal exposure to the normal surrounding brain tissue, is the cornerstone of radiosurgery therapy. It allows dose reduction to critical structures such as the brainstem and optic chiasm (26).

Theoretically, brain metastases are ideal targets for radiosurgery (2). The vast majority of these lesions are round or pseudospherical (4, 21), and it is not difficult to achieve a spherical isodose configuration in radiosurgery treatment planning (24, 25). Frequently brain metastases are located in noneloquent areas at the gray–white junction (4), allowing the delivery of a single large fraction dose with relatively low morbidity 2, 3, 10, 26, 27. Routine use of brain computed tomography (CT) and magnetic resonance imaging (MRI) for identification of brain metastases provides for the detection of metastatic lesions that are less than 3 cm in diameter (20). This limits the treatment volume while helping to reduce associated morbidity (27).

The utility of radiosurgery for the treatment of brain metastases was demonstrated by Sturm et al. (28) in the 1980s. Twelve inoperable and radioresistant lesions were treated using a modified linear accelerator, with a single fraction dose from 2000 cGy to 3000 cGy. These lesions demonstrated reduction of the enhancing volume and edema resulting in improvement of motor and sensory deficits. These data were confirmed by Loeffler et al. (29) treating recurrent brain metastases. They reported dramatic and rapid neurological and radiological responses in their series, reaching local control in all patients, with a median follow-up period of 9 months.

Radiosurgery has been used in many centers for the treatment of small brain metastases with positive results 14, 24, 27, 28, 29, 30. The local control rates are quite good: 82% for all lesions (92% for small, 75% for medium, and 62% for large lesions) at the Tokyo Women’s Medical Center in a 5.5-month mean follow-up period (31), and 94% for all lesions based on 16 years’ experience of the Karolinska Institute (32). Alexander et al. (2), in a large retrospective study, showed that radiosurgery was as effective as surgery in providing local control rates, with the added advantage of the same effectiveness for controlling surgically inaccessible lesions, multiple lesions, and tumors usually resistant to conventional radiation treatment.

Potential prognostic factors for brain metastases patients are: age, Karnofsky performance status (KPS), extracranial disease status, maximum volume and number of brain lesions 2, 15, 24, 31, 32, 33, 34. Composed scores with strong accuracy for prognosis may help us in defining the most adequate treatment strategy for each patient and clinical situation 1, 35, 36, 37.

In order to better test new treatment techniques for brain metastases on homogeneous groups of patients, Gaspar et al. (35), analyzing 1200 patients from three consecutive Radiation Therapy Oncology Group (RTOG) brain metastases studies, developed a recursive partitioning analysis (RPA) classification. They suggested three classes or stages, in descending prognostic expectancy from 1 to 3: Class 1: patients with KPS ≥ 70, < 65 years of age, with controlled primary and no evidence of extracranial metastases; Class 3: KPS < 70; and Class 2: all others.

The Score Index for Radiosurgery in Brain Metastases (SIR) was developed with the same intention. It is a prognostic score based upon retrospective survival analysis of our brain metastases patients treated with radiosurgery (36). It is a pure number that resulted from the association of five major prognostic factors: age, KPS, extracranial disease status, number of brain lesions, and largest brain lesion volume. These prognostic factors were classified in categories 0, 1, and 2, in ascending order, according to expected survival. The SIR was calculated by the summation of these marks, which may range from 0 to 10. The cutoff level for each factor is described in Table 1. Tests conducted with our brain metastases patients submitted to radiosurgery utilizing SIR demonstrated significant correlation with survival (36, 37).

In this study we evaluated retrospectively the predictive power of prognostic factors and SIR for our brain metastases patients submitted to radiosurgery. We also compared the accuracy of SIR to RPA, the RTOG prognostic index (35).

Section snippets

Methods and materials

From July 1993 to December 1997, 65 brain metastases patients were submitted to radiosurgery at Hospital Israelita Albert Einstein in Sao Paulo, Brazil. Patients and treatment characteristics are listed in Table 2. Exclusion criteria included: more than 5 lesions, any lesion larger than 30 cm3, KPS less than 50, clinical evidence requiring urgent neurosurgical intervention, or a very poor overall prognosis due to progressive systemic disease.

Fifty-eight patients (89%) received WBI during some

Results

Among the 65 patients analyzed, survival ranged from less than 1 month to more than 65 months. Overall actuarial median survival, obtained by Kaplan-Meier method, was 6.8 months, with two patients living 30 and 65 months after radiosurgery (Table 2 and Fig. 1). Survival curves for WBI, metastatic brain lesion site, age, primary tumor histology, and WBI did not demonstrate significant difference among subsets by log-rank test. However, the number of lesions and the largest brain lesion volume

Discussion

Considering expected survival rate to be the best selection parameter when evaluating the benefits of radiosurgery for patients with brain metastases, we divided the patients into one of two groups according to their prognosis: (a) surviving more than 6 months after radiosurgery (with a possible indication for radiosurgery), or (b) surviving less than 6 months after the radiosurgery (without a clear indication for radiosurgery) based on patient and disease-related data (36).

Patient-related

References (45)

  • L. Gaspar 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)
  • T.W. Griffin et al.

    Predicting the response of head and neck cancers to radiation therapy with a multivariate modeling systemAn analysis of the RTOG head and neck registry

    Int J Radiat Oncol Biol Phys

    (1984)
  • J.M. Buatti et al.

    Treatment selection factors for stereotactic radiosurgery of intracranial metastases

    Int J Radiat Oncol Biol Phys

    (1995)
  • E.M. Noordijk et al.

    The choice of treatment of single brain metastasis should be based on extracranial tumor activity and age

    Int J Radiat Oncol Biol Phys

    (1994)
  • M.K. Gospodarowicz et al.

    Introduction

  • E. Alexander et al.

    Stereotactic radiosurgery for the definitive, noninvasive treatment of brain metastases

    J Natl Cancer Inst

    (1995)
  • A.R. Kagan

    Palliation of brain and spinal cord metastases

  • P. Davey et al.

    Disposition of cerebral metastases from malignant melanomaImplications for radiosurgery

    Neurosurgery

    (1991)
  • L.D. Lunsford et al.

    General introductionhistory of radiosurgery

  • W.R. Markesbery et al.

    Treatment for patients with cerebral metastases

    Arch Neurol

    (1978)
  • E.S. Nussbaum et al.

    Brain metastasesHistology, multiplicity, surgery and survival

    Cancer

    (1996)
  • J.B. Posner

    Management of central nervous system metastases

    Semin Oncol

    (1977)
  • Cited by (0)

    View full text