Fax: (713) 563-2331
Nonsmall cell lung cancer presenting with synchronous solitary brain metastasis
Version of Record online: 29 MAR 2006
Copyright © 2006 American Cancer Society
Volume 106, Issue 9, pages 1998–2004, 1 May 2006
How to Cite
Hu, C., Chang, E. L., Hassenbusch, S. J., Allen, P. K., Woo, S. Y., Mahajan, A., Komaki, R. and Liao, Z. (2006), Nonsmall cell lung cancer presenting with synchronous solitary brain metastasis. Cancer, 106: 1998–2004. doi: 10.1002/cncr.21818
- Issue online: 18 APR 2006
- Version of Record online: 29 MAR 2006
- Manuscript Accepted: 18 NOV 2005
- Manuscript Revised: 5 NOV 2005
- Manuscript Received: 26 AUG 2005
- Radiological Society of North America International Radiology Education Program Grant to “Teach the Teachers” from Emerging Nations
- nonsmall cell lung cancer;
- synchronous brain metastasis;
- whole brain radiation therapy
Solitary brain metastases occur in about 50% of patients with brain metastases from nonsmall cell lung cancer (NSCLC). The standard of care is surgical resection of solitary brain metastases, or stereotactic radiosurgery (SRS) plus whole brain radiation therapy (WBRT). However, the optimal treatment for the primary site of newly diagnosed NSCLC with a solitary brain metastasis is not well defined. The goal was to distinguish which patients might benefit from aggressive treatment of their lung primary in patients whose solitary brain metastasis was treated with surgery or SRS.
The cases of 84 newly diagnosed NSCLC patients presenting with a solitary brain metastasis and treated from December 1993 through June 2004 were retrospectively reviewed at The University of Texas M. D. Anderson Cancer Center. All patients had undergone either craniotomy (n = 53) or SRS (n = 31) for management of the solitary brain metastasis. Forty-four patients received treatment of their primary lung cancer using thoracic radiation therapy (median dose 45 Gy; n = 8), chemotherapy (n = 23), or both (n = 13).
The median Karnofsky performance status score was 80 (range, 60-100). Excluding the presence of the brain metastasis, 12 patients had AJCC Stage I primary cancer, 27 had Stage II disease, and 45 had Stage III disease. The median follow-up was 9.7 months (range, 1-86 months). The 1-, 2-, 3-, and 5-year overall survival rates from time of lung cancer diagnosis were 49.8%, 16.3%, 12.7%, and 7.6%, respectively. The median survival times for patients by thoracic stage (I, II, and III) were 25.6, 9.5, and 9.9 months, respectively (P = .006).
By applying American Joint Committee on Cancer staging to only the primary site, the thoracic Stage I patients in our study with solitary brain metastases had a more favorable outcome than would be expected and was comparable to Stage I NSCLC without brain metastases. Aggressive treatment to the lung may be justified for newly diagnosed thoracic Stage I NSCLC patients with a solitary brain metastasis, but not for locally advanced NSCLC patients with a solitary brain metastasis. Cancer 2006. © 2006 American Cancer Society.
Lung cancer is the leading cause of cancer mortality in both women and men in the US. In 2005 an estimated 172,570 new cases of lung cancer and 168,140 deaths from it will occur in the US.1 Brain metastases occur in 30% to 50% of patients with nonsmall cell lung cancer (NSCLC) and confer a worse prognosis and quality of life.2–4 Historically, whole brain radiation therapy (WBRT) alone was offered as first-line therapy for the management of brain metastases. However, the landmark randomized trial by Patchell et al.5 comparing WBRT plus biopsy with WBRT plus surgical resection of single brain metastasis established surgery as the standard of care for patients with a single brain metastasis because it prolonged survival from 15 weeks to 40 weeks. In addition, the Radiation Therapy Oncology Group 95-08 trial demonstrated longer survival time for patients with solitary brain metastasis treated with stereotactic radiosurgery (SRS) and WBRT compared with WBRT alone (median survival time: 6.5 vs. 4.9 months).6 Additional studies have been published that support the use of SRS in patients with solitary brain metastases.7–13 These trials, which have demonstrated longer survival times for patients with aggressively treated single brain metastases, raise the question as to what level of treatment is appropriate for the primary site in NSCLC.
The optimal treatment for the primary site in patients with newly diagnosed NSCLC who have solitary brain metastases is not well defined, although several studies have shown that some patients might benefit from aggressive therapy.14–18 We sought to distinguish which patients with solitary brain metastases treated with surgery or SRS might benefit from aggressive treatment of their primary lung cancer.
MATERIALS AND METHODS
We searched the database of the Department of Radiation Oncology and Neurosurgery to identify patients with brain metastases who were treated at the University of Texas M. D. Anderson Cancer Center from December 1993 through June 2004 (N = 2200). We screened these cases to identify patients with solitary synchronous brain metastases from NSCLC that were diagnosed within 1 month of the primary lung cancer diagnosis and treated with SRS or neurosurgery (N = 84). Patients with any other distant metastases (e.g., liver, bone, or adrenal metastases) were not included in the study cohort.
The study population consisted of 46 men and 38 women. The median age at the time of diagnosis was 63 years (range, 29-77 years) and the median Karnofsky performance status (KPS) score was 80 (range, 60-100). The histopathologic subtype was adenocarcinoma in 47 cases, NSCLC (not otherwise specified) in 16 cases, squamous cell carcinoma in 11 cases, large cell carcinoma in 6 cases, and other in 4 cases. Excluding the presence of the brain metastasis, the stage distribution according to the 1997 American Joint Committee on Cancer (AJCC) staging criteria was Stage I in 12 cases, Stage II in 27, and Stage III in 45. The Radiation Therapy Oncology Group recursive partitioning analysis (RPA) class was applied as follows: Class I, patients who had a KPS score ≥70 and were <65 years of age with a controlled primary tumor and no extracranial metastases (n = 0); Class III, a KPS score <70 (n = 4); and Class II, all other patients (n = 80).12 The characteristics of the patients are listed in Table 1.
|N = 31||N = 53|
Patients had had a complete workup before definitive treatment. Workup included computed tomography of the chest and abdomen, bone scan, complete blood counts, and serum chemistries. Most brain metastases had been diagnosed by magnetic resonance imaging (MRI), and serial MRI had been used for follow-up. Computed tomography of the brain had been used as a diagnostic tool only in 5 cases in which MRI was contraindicated.
The institutional review board at M. D. Anderson cancer center approved this retrospective review of the patients' medical records and waived the requirement for informed consent.
Eight patients had received thoracic radiotherapy at a median dose of 45 Gy (range, 12-65 Gy), 23 patients had received chemotherapy, and 13 patients had received both. Most of the chemotherapy regimens consisted of a combination of carboplatin and paclitaxel (n = 30), etoposide + cisplatin in 3, etoposide alone in 2, and gemcitabine alone in 1. The median number of cycles was 4 (range, 2-6). Forty patients had received no treatment at all for the primary cancer.
Patients had undergone either SRS (n = 31) or craniotomy (n = 53) for the management of the solitary brain metastasis. The SRS technique has been previously published in detail.11 Briefly, SRS was performed using a linear accelerator with 6 MV photons. The median SRS dose to the brain metastases was 18 Gy and was delivered in a single session. The median prescribed isodose line was 95%. Details of the SRS are listed in Table 2. Seven SRS patients and 24 craniotomy patients received planned postprocedural WBRT consisting of 30 Gy in 10 fractions.
|Primary treatment, no. of patients||44|
|Radiotherapy and chemotherapy||13|
|Median SRS dose, Gy (range)|
|Prescription dose||18 (13–20)|
|Isocenter dose||22.3 (18–23.5)|
|Median prescription isodose, % (range)||85 (80–90)|
|Median WBRT dose, Gy (range)||30 (30)|
|Median cone diameter, cm (range)||2.25 (1.5–4)|
MRI studies had been routinely obtained for each patient every 1-3 months for the first 2 years after treatment. Recurrence of brain metastases was classified as local failure, and any new brain metastases in a location distinct from that of the original tumor were considered distant brain metastases. If any of the following criteria were met, the patient was considered to have a local failure at the lesion site: increase in tumor diameter by ≥25%, progression of an SRS-treated lesion that required subsequent surgery, or SRS-related complications. Distant brain metastasis-free survival was defined as the absence of new brain metastases.
Survival time was calculated in months from the date of each patient's diagnosis. Actuarial local tumor control and distant brain metastases-free survival curves were estimated using the Kaplan-Meier method. Univariate comparisons of potential prognostic factors for tumor control and patient survival were performed with log-rank using Stata software (Release 9; College Station, TX, 2005). A P value of <.05 was considered statistically significant.
The median follow-up time was 9.7 months (range, 1-86 months). The 1-, 2-, 3-, and 5-year overall survival rates for all patients were 49.8%, 16.3%, 12.7%, and 7.6%, respectively (Fig. 1). The median survival times for patients with Stage I, II, or III thoracic disease (primary cancer) were 25.6, 9.5, and 9.9 months, respectively, and the 3-year of overall survival rates were 37.5%, 10.3%, and 7.6%, respectively (P = .006; Fig. 2). The 3-year overall survival rates for patients whose primary cancer had been treated (n = 44) were 50% for Stage I and 10% for non-Stage I patients (P = .018). The median survival times for the SRS and surgery groups were 7.4 and 12 months, respectively (P = .09).
Primary Cancer Treatment
The median survival durations for patients who had received treatment (including radiotherapy and chemotherapy) for their primary cancers and those who had not were 15.5 and 5.9 months, respectively (P = .046). The corresponding 2-year overall survival rates were 18.5% and 13.9%, respectively; however, both survival curves collapsed after 2 years (Fig. 3). The 2-year primary cancer recurrence-free survival rate for patients whose primary cancer has been treated was 53% for Stage I disease and 24% for non-Stage I patients (P = .058). The median survival time for patients whose primary cancer had been treated was 25.6 months for Stage I disease and 7.3 months for non-Stage I patients (P = .018).
Local Brain Tumor Control Rate
The 1- and 2-year local control rates of brain tumor for all patients were 83.9% and 77.9%, respectively (Fig. 1). The corresponding rates were 61% and 61% for the SRS group and 97% and 88% for the neurosurgery group. The rates were significantly different between the SRS and neurosurgery groups (P = .0007). The 3-year local brain tumor control rate for patients whose primary cancer had been treated was 100% for Stage I cancer and 77.5% for non-Stage I patients (P = .26).
The 2-year distant brain metastasis-free survival rates were not significantly different in the SRS and surgery groups (52% and 45.2%; P = .38). In addition, the 2-year rates were not significantly different by stage (Stage I, 64.3%; Stage II, 52.7%; Stage III, 46.4%; P = .15).
The median tumor volume of brain metastasis for all patients was 5.34 cm3 (range, 0.24-113.1 cm3). The mean tumor volume for the SRS and the neurosurgery groups were 3.7 and 16.3 cm3, respectively (P = .004). Tumor volume (categorized as <5, 5-10, and >10 cm3) did not differ significantly by primary cancer stage (Table 3). The categorization of tumor volume also revealed no significant differences in the 2-year local control rates (72.0%, 63.9%, and 100%; P = .085) or overall survival rates (12.9%, 12.2%, and 26.5%; P = .26). However, tumor volume >10 cm3 was significantly associated with the development of distant brain metastases (2 years, 53.0%, 0.0%, and 79.3%, P = .007).
|Tumor Volume (cm3)||Primary cancer stage, no. (%)||Total, no.|
|<5||6 (50)||13 (48.2)||22 (48.9)||41|
|5-10||3 (25)||7 (25.9)||12 (26.7)||22|
|>10||3 (25)||7 (25.9)||11 (24.4)||21|
Causes of Death
Sixty-eight patients died; 62 died from disease, 11 died with brain disease, and 42 with distant metastases including lung (n = 21), bone (n = 10), adrenal gland (n = 5), liver (n = 3), brain (n = 11), and others (n = 4). Most of the patients had multiple metastases; 9 patients died with disease but not specified. Five patients died of unknown causes. One patient died of intercurrent disease.
Univariate analysis revealed no significant associations between the KPS score, RPA, or histologic subtype with local brain control, brain distant metastases, or overall survival rates. These results are not surprising because there were no RPA I patients and only 4 RPA III patients; almost all were RPA II.
In our retrospective study, the median survival times and survival rates were better than what we expected for NSCLC patients with solitary synchronous brain metastases. The 5-year overall survival rates from time of lung cancer diagnosis for the entire group was 7.6%, compared with 2% with Stage IV disease.19
Treatment of Primary Cancer and Synchronous Solitary Brain Metastases
There is no agreed upon standard treatment of the primary site for patients with synchronous brain metastases from NSCLC. Table 4 summarizes the findings of four previously published studies and our own on the treatment of synchronous brain metastases from NSCLC. Billing et al.14 studied 28 patients with brain metastases from NSCLC who underwent craniotomy plus resection of the lung primary tumor. WBRT was given to 24 patients, and 9 of these also received postoperative systemic chemotherapy. Median survival was 24 months. Overall survival rates at 1, 2, and 5 years were 64.3%, 54.0%, and 21.4%, respectively. Chidel et al.15 conducted a retrospective analysis of 33 patients with newly diagnosed NSCLC with a single synchronous brain metastasis between 1982 and 1996. An aggressive approach to the primary lung cancer was undertaken in 13 patients, and most of the brain metastases were treated by surgical resection and WBRT. The median overall and disease-free survival durations were 6.9 months and 3.3 months, respectively. There were 2 long-term survivors up to 65.8 months. The authors concluded that an aggressive approach to selected patients was reasonable because it resulted in an occasional long-term survival duration. Bonnette et al.16 observed median survival durations of 12.4 months in 103 patients treated with surgery for both the primary tumor and brain metastases; the 5-year overall survival rate was 11%. Downey et al.17 conducted a Phase II trial of chemotherapy and surgery to brain and lung for NSCLC in 23 patients with a solitary synchronous brain metastasis (N0 or N1). Ten of 23 patients finished chemotherapy and 2 patients survived for 5 years. Harita et al.18 reported 2 NSCLC patients with isolated metastases to the brain who received concurrent chemotherapy and radiotherapy for the thoracic disease and the brain disease. Cisplatin and docetaxel were administrated for up to three cycles. One patient survived for 53 months and the other for 37 months. Nakayama et al.20 analyzed 15 patients who initially presented with brain metastases from NSCLC treated by SRS. Those treated for the primary and mediastinal lymph node (22 months) survived longer than those who did not (8 months; P<.001). In summary, patients treated for the primary tumor (with chemotherapy, radiotherapy, or both) and brain metastases had better outcome. In our cohort, the 44 patients who had received such treatment had a significantly longer median survival time, although the prolonged survival rate was not durable.
|AuthorRef. No.||No. of Patients||% Craniotomy||No. of SRS or WBRT||% Primary Cancer Treatment||Median Survival, mo||5-y Overall Survival Rate, %|
|Billing et al.14||28||100||0/15||100||24||21.4|
|Chidel et al.15||33||63.6||7/28||84.8||6.9||N/A|
|Bonnette et al.16||103||100||0/75||100||12.4||11|
|Downey et al.17||23||56.5||1/NA||56.5||11||8.7|
Who Will Benefit from Aggressive Treatment
Billing et al.14 studied 28 patients with brain metastases from NSCLC who underwent craniotomy and resection of the lung primary tumor. Seventeen patients had N0 disease, 5 had N1 disease, and 6 had N2 disease. The 5-year survival rate for patients with N0 disease was 35% compared with 0% in patients with N1 or N2 lymph node disease. Negative lymph nodes appear to be an important prognostic factor for patients even in the presence of a single brain metastasis. Bonnette et al.16 observed median survival durations of 12.4 months in 103 patients treated with surgery for both the primary tumor and brain metastases, with a 5-year survival overall rate of 11%. We compared the current study group of patients with our database of 1002 lung cancer patients treated in the Department of Radiation Oncology at M. D. Anderson. The overall survival rate for patients with synchronous solitary brain metastasis was worse than that of the whole database (P<.01). But for the Stage I patients with synchronous brain metastases, there were no significant differences between the 2 groups of patients with or without brain metastases, P = .64 (Fig. 4), and the result was similar to Mountain et al.'s.19 However, there were significant differences of survival rates for Stage II and III patients between patients with or without brain metastases (P < .01; Figs. 5, 6). Also, there were significant differences of overall survival rate and chest control for Stage I patients compared with Stage II and III (53% vs. 24%). The data suggest that aggressive treatment of both the brain metastases and primary NSCLC is indicated for thoracic Stage I patients.
We recognize that there are several limitations to our retrospective study. The cohort may be biased by patient selection. Most of the patients in the early 1990s were treated by neurosurgery and were generally considered more favorable patients, whereas patients who were not operative candidates were referred for radiosurgery. This may explain the more favorable outcome associated with resection of brain metastasis compared with SRS treatment. The number of patients is small, especially for Stage I patients.
Stage I patients in our study with solitary brain metastases had a more favorable outcome than would be expected and was comparable to Stage I NSCLC patients without brain metastases. These results, within the limits of a retrospective study, suggest that aggressive treatment of the lung may be justified for newly diagnosed thoracic Stage I NSCLC patients with a solitary brain metastasis, but not for locally advanced NSCLC patients with a solitary brain metastasis.
- 19Lung cancer handbook for staging and imaging, 3rd ed. Houston: Clifton F. Mountain Foundation, 1996., , .