- Email: [email protected]
Impact of non-small cell lung cancer histology on survival predicted from the graded prognostic assessment for patients with brain metastases
Lung Cancer 77 (2012) 389–393
Contents lists available at SciVerse ScienceDirect
Lung Cancer journal homepage: www.elsevier.com/locate/lungcan
Impact of non-small cell lung cancer histology on survival predicted from the graded prognostic assessment for patients with brain metastases Susan Guo a,∗ , Chandana A. Reddy a , Samuel T. Chao a,b , John H. Suh a,b , Gene H. Barnett b,c , Michael A. Vogelbaum b,c , Gregory M. Videtic a a
Department of Radiation Oncology, Cleveland Clinic, Cleveland, OH, United States Rose Ella Burkhardt Brain Tumor and Neuro-Oncology Center, United States c Department of Neurosurgery, Cleveland Clinic, Cleveland, OH, United States b
a r t i c l e
i n f o
Article history: Received 21 December 2011 Received in revised form 19 March 2012 Accepted 24 March 2012 Keywords: Non-small cell lung cancer Brain metastases Graded prognostic assessment Prognosis Histology
a b s t r a c t Introduction: The Graded Prognostic Assessment (GPA) provides prognostic classification for patients with brain metastases (BM), based on Radiation Therapy Oncology Group (RTOG) data. Recent evidence suggests differential response and outcomes to chemotherapy for different non-small cell lung cancer (NSCLC) histologies. Using a large BM patient database, we assessed the impact of histologic subtypes on survival stratified by the GPA. Methods: From an IRB-approved database, we analyzed 780 patients with NSCLC BM treated from 1982 to 2004. GPA classification variables included age, KPS, number of BM, and presence of extracranial disease. Histology was identified for each patient. Median survival time (MST) based on GPA class and histology were calculated using Kaplan–Meier analysis. The log rank test was used to determine statistical differences. Results: MST, in months, by histology were: adenocarcinoma (AC) 6.2 (n = 464), large cell (LC) 4.1 (n = 98), squamous (SQ) 4.2 (n = 108) (p = 0.0549). For GPA 3.5–4.0, MSTs did not differ significantly by histology. Differences in MST by histology were noted for GPA 3.0 (p = 0.04), GPA 1.5–2.5 (p = 0.01), and GPA 0–1.0 (p = 0.02). For all patients with brain metastases BM from NSCLC, MSTs by GPA score were: GPA 3.5–4.0, 12.6; GPA 3.0, 10.2; GPA 1.5–2.5, 5.8; and GPA 0–1.0, 2.7. Conclusions: Adenocarcinoma showed a statistically significant higher MST than other histologies of NSCLC for patients with GPA 0–3.0. Using histology as a prognostic factor for BM from NSCLC warrants further investigation. Our cohort of NSCLC BM patients validates the GPA, with MST comparable to that of published data. © 2012 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Brain metastases are diagnosed in approximately 200,000 patients per year, with rising incidence due to improving detection and treatment of systemic malignancy. Lung cancer is the most common primary tumor giving rise to brain metastases [1]. Approximately 50% of all brain metastases patients have a lung primary [2]. Given the variety of treatment options including whole-brain radiation therapy, stereotactic radiosurgery, and surgery, the management of brain metastases has become controversial [3]. Multiple prognostic indices have been published to use clinical factors to stratify brain metastasis patients into homogeneous
∗ Corresponding author at: Department of Radiation Oncology, Cleveland Clinic, 9500 Euclid Ave, Desk T28, Cleveland, OH 44113, United States. Tel.: +1 216 444 5570; fax: +1 216 445 1068. E-mail addresses: [email protected] (S. Guo), [email protected] (G.M. Videtic). 0169-5002/$ – see front matter © 2012 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.lungcan.2012.03.028
survival groups [4–7]. The graded prognostic assessment (GPA) stratifies patients into categories based on age, Karnofsky Performance Status (KPS), extracranial metastases (none and present), and number of metastases (one, two to three, and more than three) (Table 1) [7]. Recent studies have suggested that brain metastases from lung cancer respond to treatment differently than brain metastases from other primary sites [8]. Additional data suggests differential response and outcomes to selected chemotherapy for different NSCLC histologies [9]. Using a large single institutional database of patients with brain metastases, we assessed the impact of histologic subtypes on survival outcomes stratified by the GPA. A secondary endpoint was to validate the GPA with our data set. 2. Methods and materials Data was provided by the Cleveland Clinic brain tumor database, an institutional review board-approved registry of more than 1200
390
S. Guo et al. / Lung Cancer 77 (2012) 389–393
Table 1 Graded prognostic assessment.a
Table 2 Patient, tumor, and treatment characteristics for 780 NSCLC brain metastasis patients.
Score 0 Age KPS Number of brain metastases Extracranial metastases
>60 <70 >3 Present
0.5 50–59 70–80 2–3 –
1.0 <50 90–100 1 None
Abbreviations: KPS = Karnofsky Performance Status; CNS = central nervous system. a Adapted from Ref. [7].
patients with brain metastases. From the interval between January 1982 to September 2004, 780 consecutive patients with newly diagnosed brain metastases from NSCLC and known histologies were identified in this retrospective review (Table 2). Patients recorded in this database were either diagnosed and treated for primary and secondary tumors at our institution as well as at outside hospitals, or referred to the Cleveland Clinic for treatment of brain metastases [10]. Patient cohorts stratified either by GPA class or by histology were analyzed for survival differences. GPA was calculated for each patient based on the published criteria [7]. Tumor histologies were categorized into adenocarcinoma, large cell, squamous, mixed and NOS (not otherwise specified). Median survival times from time of diagnosis were calculated using Kaplan–Meier analysis, and comparisons were made to the published GPA results [7]. The log rank test was used to determine whether a statistically significant difference was present among patient groups. The statistical analyses were performed using StatView version 5.0 (SAS Institute, Cary, NC) and a p value of <0.05 was considered statistically significant.
3. Results For 780 patients, the median follow-up was 5.5 months (range, 0–122.9 months). The median age was 62.3 years (range, 25.3–90.8 years). The median KPS was 80 (range, 20–100). The median number of lesions at diagnosis was 2 (range, 1–50). Extracranial disease was present in 342 patients (43.8%) and absent in 438 patients (56.2%). The number of patients with known histologies was as follows: adenocarcinoma 468 (60.0%), large cell 98 (12.6%), mixed 15 (1.9%), NOS 91 (11.7%), and squamous 108 (13.8%). Patients with mixed and NOS histologies were excluded from GPA by histology analysis because of their relatively small numbers, and were included only in the validation of GPA analysis. Of all patients, 687 patients (88%) received WBRT, 583 patients (75%) received surgery, and 514 patients (65.9%) received stereotactic radiosurgery. Median survival times (MST, in months) by histology were: adenocarcinoma (AC) 6.2 (n = 464), large cell (LC) 4.1 (n = 98), squamous (SQ) 4.2 (n = 108) (p = 0.04549) (Fig. 1). GPA classification variables were identifiable in 751 patients. For GPA 3.5–4.0, MST by histology did not differ significantly (p = 0.8597): AC 11.2 (n = 39), LC 4.1 (n = 2), SQ 10.6 (n = 12) (p = 0.8597). Statistically significant differences in MST by histology were noted for GPA 3.0 (p = 0.0420), GPA 1.5–2.5 (p = 0.0052), and GPA 0–1.0 (p = 0.0205). For GPA 3.0, MST were: AC 12.1 (n = 54), LC 10.2 (n = 14), SQ 6.9 (n = 14) (p = 0.0420). For GPA 1.5–2.5, MST were: AC 6.1 (n = 275), LC 4.6 (n = 49), SQ 3.6 (n = 60) (p = 0.0052). For GPA 0–1.0, MST were: AC 3.2 (n = 83), LC 2.2 (n = 32), SQ 3.2 (n = 17) (p = 0.0205) (Fig. 2a–d; Table 3). For all patients with brain metastases from NSCLC, our MST by GPA score were: GPA 3.5–4.0 (Class I), 12.6; GPA 3.0 (Class II), 10.2; GPA 1.5–2.5 (Class III), 5.8; and GPA 0–1.0 (Class IV), 2.7 (p < 0.0001) (Fig. 3).
Tumor type Adenocarcinoma Large cell Mixed NOS Squamous Gender Female Male GPA score 0 0.5 1 1.5 2 2.5 3 3.5 GPA group 0–1 1.5–2.5 3 3.5–4 Age 50–60 <50 >60 Unknown Number of lesions 1 2 >2 Unknown KPS <70 70–80 90–100 Unknown Extracranial metastases N Y Received WBRT N Y WBRT fractionation 3000 cGy in 10 fx 3750 cGy in 15 fx 4000 cGy in 20 fx Other Unknown Received surgery N Y Received SRS N Y Treatment era 1982–1996 1997–2004 unknown
n
%
468 98 15 91 108
60.0 12.6 1.9 11.7 13.8
326 454
41.8 58.2
11 52 93 133 172 162 101 43
1.4 6.7 11.9 17.1 22.1 20.8 12.9 5.5
156 467 101 56
20.0 59.9 12.9 7.2
212 118 449 1
27.2 15.1 57.6 0.1
324 155 264 37
41.5 19.9 33.8 4.7
152 367 235 26
19.5 47.1 30.1 3.3
438 342
56.2 43.8
93 687
11.9 88.1
328 154 22 161 115
42.1 19.7 2.8 20.6 14.7
583 197
74.7 25.3
514 266
65.9 34.1
322 456 2
41.3 58.5 0.3
4. Discussion Our data is the first report suggesting that lung cancer histology may be prognostic for survival in patients with brain metastases from NSCLC. These improved survivals may reflect inherent biological differences among the various histologies. Recent evidence has shown that different histologies have varied responses to systemic therapy. Scagliotti and colleagues reported that adenocarcinomas treated with cisplatin and pemetrexed showed improved survival compared to treatment with cisplatin and gemcitabine in the firstline setting of advanced NSCLC [9]. Additional data suggests that
S. Guo et al. / Lung Cancer 77 (2012) 389–393
391
Table 3 Median survival (months) by GPA and by histology.
GPA 3.5–4.0 (Class I) (n = 53) GPA 3.0 (Class II) (n = 82) GPA 1.5–2.5 (Class III) (n = 384) GPA 0–1.0 (Class IV) (n = 132)
Adenocarcinoma
Large Cell
Squamous
p value
11.2 (n = 39) 12.1 (n = 54) 6.1 (n = 275) 3.2 (n = 83)
4.1 (n = 2) 10.2 (n = 14) 4.6 (n = 49) 2.2 (n = 32)
10.6 (n = 12) 6.9 (n = 14) 3.6 (n = 60) 3.2 (n = 17)
0.8597 0.0420 0.0052 0.0205
100
Overall Survival (%)
80 Adenocarcinoma (MST=6.2, N=464) Large Cell (MST=4.1, N=98) Squamous (MST=4.2, N=108)
60
40
20 p=0.0549 0 0
12
24
36
48
60
72
84
96
108
Fig. 3. Overall survival for non-small cell lung cancer patients with brain metastases by graded prognostic assessment group (n = 751).
120
Months Fig. 1. Overall survival from time of brain metastases diagnosis by histology for non-small cell lung cancer (n = 670).
brain metastases from lung primaries have different biology compared to other primary sites [11]. For example, subset analysis of RTOG (Radiation Therapy Oncology Group) 95–08 showed that patients with squamous or non-small cell tumors, mostly from lung primaries, had improved survival with whole brain radiation and GPA 3.5-4
100
Adenocarcinoma (n=275, MST= 6.1) Large Cell (n=49, MST= 4.6) Squamous (n=60, MST= 3.6)
Overall Survival (%)
80
60 40 20
60 40 20
p=0.8597
p=0.0052
0
0 0
12
24
36
48 60 Months
72
84
96
0
GPA 3
60 40
24
36
48 60 Months
84
96
108
.6
.4
.2
20
72
Adenocarcinoma (n=83, MST= 3.2) Large Cell (n=32, MST= 2.2) Squamous (n=17, MST= 3.2)
.8 Overall Survival (%)
Adenocarcinoma (n=54, MST= 12.1) Large Cell (n=14, MST= 10.2) Squamous (n=14, MST= 6.9)
80
12
GPA Group 0-1
1
100
Overall Survival (%)
GPA 1.5-2.5
100
Adenocarcinoma (n=39, MST= 11.2) Large Cell (n=2, MST= 4.1) Squamous (n=12, MST= 10.6)
80 Overall Survival (%)
stereotactic radiosurgery compared to whole brain radiation alone [11]. The exact mechanisms by which these biological differences relate to brain metastases remain to be elucidated. We emphasize that our data suggests lung cancer histology as a prognostic factor for improved survival in brain metastasis patients, but do not imply causation for improved survival. Adenocarcinoma is a cell type of NSCLC that has been recognized to metastasize commonly to the brain and to other distant
p=0.0205
p=0.0420 0
0 0
12
24
36
48
60 72 Months
84
96
108 120
0
12
24 Months
Fig. 2. Overall survival for non-small cell lung cancer patient with brain metastases by histology and by graded prognostic assessment group.
392
S. Guo et al. / Lung Cancer 77 (2012) 389–393
Table 4 Presence of extracranial disease by histology.
Adenocarcinoma Large cell Squamous
Number of patients
Extracranial metastases absent
Percentage without extracranial metastases by histology
Extracranial metastases present
Percentage with extracranial metastases by histology
451 97 103
252 53 62
55.9% 54.6% 60.2%
199 44 41
44.1 45.4 39.8
Chi square p-value 0.6786.
sites [12–14]. The incidence of brain metastases from adenocarcinomas is 43% of NSCLC in a retrospective review [15]. Our data suggests that brain metastases from this histology have improved survival compared to brain metastases from other histologies of non-small cell lung cancer. This improvement was seen across all four GPA groups, and was statistically significant in three out of four GPA groups. The outcomes may not have been statistically significant in GPA group 1 due to low patient numbers. Further evaluation of this association with larger patient numbers and multivariate analysis is warranted. Establishing the validity of histology for prognosis could permit improvements in clinical trial design by including this as a stratification factor in future brain metastasis trials. One limitation of our database is that no detailed patterns of failure information are available. Reflecting the relatively short life span of patients with brain metastases, whether these patients failed locally or systemically following brain treatment was rarely discernable or defined, and therefore was not readily captured in our database. To determine if our histology cohorts were balanced with respect to extracranial disease, we re-analyzed our data by these parameters and found no imbalance with respect to numbers of patients with systemic (extracranial) disease between any of the histologies (Table 4). Since the histologic groups were balanced with respect to percentage of patients with and without extracranial disease, and since randomized data had shown no significant advantages with respect to outcomes for any chemotherapy used during the era during which this data was collected (1982–2004), we therefore conclude that our observed improved outcomes based on histology were not confounded by improved control of extracranial disease in any histologic subset [16]. The GPA was based on 1960 analyzable patients from 5 randomized RTOG trials [7]. This initial report of the GPA reported MST of: 11.0 (Class I), 8.9 (Class II), 3.8 (Class III), and 2.6 (Class IV) months. Since the GPA was first published, it has been validated in a series of 140 patients treated at the University of Minnesota between April 2005 and December 2006. In this series, MST were: 21.7 (Class I), 17.5 (Class II), 5.9 (Class III), and 3.0 (Class IV). Longer survival seen in this series compared with our series and other series may be attributed to more recent dates of treatment and advances in systemic therapy [17]. The Diagnosis-Specific Graded Prognostic Assessment, (DSGPA) was later developed to account for inherent variations in prognosis among different primary tumors. This multi-institutional analysis reviewed 5067 patients from 11 institutions treated for BM between 1985 and 2007, of which included 1888 NSCLC patients [18]. All four GPA variables were demonstrated to be prognostic in NSCLC patients, and values are comparable to our data as well as the Minnesota series. Patients who were undergoing repeat treatment for BMs at the time of inclusion in the database were not counted. Our institution was one of many institutions that participated in this larger analysis. Some overlap is present between the patients in our series and this larger series, although our inclusion and exclusion criteria were different. Nieder and colleagues has also validated the GPA in brain metastasis patients in multiple series consisting of a variety of treatment modalities including whole brain radiation alone, with surgery,
and/or with SRS [19,20,21]. One series consisting of 232 patients, the majority of whom were treated with whole brain radiation alone, demonstrated that these patients had a significantly worse prognosis compared to the general population of brain metastasis patients. MST were: 10.3 (Class I), 5.6 (Class II), 3.5 (Class III), and 1.9 (Class IV) [19]. The same group also investigated the GPA in NSCLC patients in particular and also found the GPA to be prognostic in this cohort [19]. MST were: 9.5 (Class I), 4.3 (Class II), 2.5 (Class III), and 3.0 (Class IV). Because the GPA was found to be prognostic in their multiple series and across various treatment modalities for brain metastases, the authors proposed that the GPA was applicable in unselected patients with brain metastases from NSCLC. Patients with brain metastases comprise a heterogenous patient population. More accurate prognostic information may support the decision making process and treatment recommendations for these patients. Furthermore, a select number of these patients have a very limited survival, and more accurate survival prediction models could help to avoid overtreatment. The limitations of our study are the single institutional data, retrospective study design, and long time period for analysis. Although our treatment period encompassed a span of twenty years, there was no difference in overall survival by treatment era of 1982–1996 vs. 1997–2004. Our groups were not balanced with regard to treatment with WBRT, SRS, or surgery; however, data from Nieder and colleagues suggested that the GPA was valid regardless of treatment modality [21]. DS-GPA analysis also showed overlapping confidence intervals between the risk of death and median survival by treatment on multivariate analysis [18]. The current established brain metastases prognostic indices also do not control for the use of systemic treatment. 5. Conclusion Adenocarcinoma showed a higher median survival than other histologies in patients with brain metastases from NSCLC. This difference was seen across multiple GPA groups. The use of histology as a prognostic factor for brain metastases from NSCLC warrants further investigation. Our cohort of patients with brain metastases from NSCLC validated the use of GPA classification for prognosis in this population. Our reported median survivals were consistent with other series that also validated the GPA. Conflict of interest statement None declared. References [1] Nussbaum ES, Djalilian HR, Cho KH, Hall WA. Brain metastases, histology, multiplicity, surgery, and survival. Cancer 1996;78:1781–8. [2] Chang JE, Robins HI, Mehta MP. Therapeutic advances in the treatment of brain metastases. Clin Adv Hematol Oncol 2007;5:54–64. [3] Suh JH. Stereotactic radiosurgery for the management of brain metastases. N Engl J Med 2010;362:1119–27. [4] Lorenzoni J, Devriendt D, Massager N, David P, Ruiz S, Vanderlinden B, et al. Radiosurgery for treatment of brain metastases: estimation of patient eligibility using three stratification systems. Int J Radiat Oncol Biol Phys 2004;60:218–24.
S. Guo et al. / Lung Cancer 77 (2012) 389–393 [5] Gaspar L, Scott C, Rotman M, Asbell S, Phillips T, Wasserman T, 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;37:745–51. [6] Weltman E, Salvajoli JV, Brandt RA, de Morais Hanriot R, Prisco FE, Cruz JC, et al. Radiosurgery for brain metastases: a score index for predicting prognosis. Int J Radiat Oncol Biol Phys 2000;46:1155–61. [7] Sperduto PW, Berkey B, Gaspar LE, Mehta M, Curran W. A new prognostic index and comparison to three other indices for patients with brain metastases: an analysis of 1,960 patients in the RTOG database. Int J Radiat Oncol Biol Phys 2008;70:510–4. [8] Mehta MP, Rodrigus P, Terhaard CH, Rao A, Suh J, Roa W, et al. Survival and neurologic outcomes in a randomized trial of motexafin gadolinium and wholebrain radiation therapy in brain metastases. J Clin Oncol 2003;21:2529–36. [9] Scagliotti GV, Parikh P, von Pawel J, Biesma B, Vansteenkiste J, Manegold C, et al. Phase III study comparing cisplatin plus gemcitabine with cisplatin plus pemetrexed in chemotherapy-naive patients with advanced-stage non-smallcell lung cancer. J Clin Oncol 2008;26:3543–51. [10] Cannady SB, Cavanaugh KA, Lee SY, Bukowski RM, Olencki TE, Stevens GH, et al. Results of whole brain radiotherapy and recursive partitioning analysis in patients with brain metastases from renal cell carcinoma: a retrospective study. Int J Radiat Oncol Biol Phys 2004;58:253–8. [11] Andrews DW, Scott CB, Sperduto PW, Flanders AE, Gaspar LE, Schell MC, et al. Whole brain radiation therapy with or without stereotactic radiosurgery boost for patients with one to three brain metastases: phase III results of the RTOG 9508 randomised trial. Lancet 2004;363:1665–72. [12] Bajard A, Westeel V, Dubiez A, Jacoulet P, Pernet D, Dalphin JC, et al. Multivariate analysis of factors predictive of brain metastases in localised non-small cell lung carcinoma. Lung Cancer 2004;45:317–23.
393
[13] Strauss B, Weller CV. Bronchogenic carcinoma; a statistical analysis of two hundred ninety-six cases with necropsy as to relationships between cell types and age, sex, and metastasis. AMA Arch Pathol 1957;63:602–11. [14] Figlin RA, Piantadosi S, Feld R. Intracranial recurrence of carcinoma after complete surgical resection of stage I, II, and III non-small-cell lung cancer. N Engl J Med 1988;318:1300–5. [15] Mujoomdar A, Austin JH, Malhotra R, Powell CA, Pearson GD, Shiau MC, et al. Clinical predictors of metastatic disease to the brain from non-small cell lung carcinoma: primary tumor size, cell type, and lymph node metastases. Radiology 2007;242:882–8. [16] Schiller JH, Harrington D, Belani CP, Langer C, Sandler A, Krook J, et al. Comparison of four chemotherapy regimens for advanced non-small-cell lung cancer. N Engl J Med 2002;346:92–8. [17] Sperduto CM, Watanabe Y, Mullan J, Hood T, Dyste G, Watts C, et al. A validation study of a new prognostic index for patients with brain metastases: the Graded Prognostic Assessment. J Neurosurg 2008;109(Suppl.):87–9. [18] Sperduto PW, Chao ST, Sneed PK, Luo X, Suh J, Roberge D, et al. Diagnosisspecific prognostic factors, indexes, and treatment outcomes for patients with newly diagnosed brain metastases: a multi-institutional analysis of 4,259 patients. Int J Radiat Oncol Biol Phys 2010;77:655–61. [19] Nieder C, Marienhagen K, Geinitz H, Molls M. Validation of the graded prognostic assessment index for patients with brain metastases. Acta Oncol 2009;48:457–9. [20] Nieder C, Geinitz H, Molls M. Validation of the graded prognostic assessment index for surgically treated patients with brain metastases. Anticancer Res 2008;28:3015–7. [21] Nieder C, Bremnes RM, Andratschke NH. Prognostic scores in patients with brain metastases from non-small cell lung cancer. J Thorac Oncol 2009;4:1337–41.
Contents lists available at SciVerse ScienceDirect
Lung Cancer journal homepage: www.elsevier.com/locate/lungcan
Impact of non-small cell lung cancer histology on survival predicted from the graded prognostic assessment for patients with brain metastases Susan Guo a,∗ , Chandana A. Reddy a , Samuel T. Chao a,b , John H. Suh a,b , Gene H. Barnett b,c , Michael A. Vogelbaum b,c , Gregory M. Videtic a a
Department of Radiation Oncology, Cleveland Clinic, Cleveland, OH, United States Rose Ella Burkhardt Brain Tumor and Neuro-Oncology Center, United States c Department of Neurosurgery, Cleveland Clinic, Cleveland, OH, United States b
a r t i c l e
i n f o
Article history: Received 21 December 2011 Received in revised form 19 March 2012 Accepted 24 March 2012 Keywords: Non-small cell lung cancer Brain metastases Graded prognostic assessment Prognosis Histology
a b s t r a c t Introduction: The Graded Prognostic Assessment (GPA) provides prognostic classification for patients with brain metastases (BM), based on Radiation Therapy Oncology Group (RTOG) data. Recent evidence suggests differential response and outcomes to chemotherapy for different non-small cell lung cancer (NSCLC) histologies. Using a large BM patient database, we assessed the impact of histologic subtypes on survival stratified by the GPA. Methods: From an IRB-approved database, we analyzed 780 patients with NSCLC BM treated from 1982 to 2004. GPA classification variables included age, KPS, number of BM, and presence of extracranial disease. Histology was identified for each patient. Median survival time (MST) based on GPA class and histology were calculated using Kaplan–Meier analysis. The log rank test was used to determine statistical differences. Results: MST, in months, by histology were: adenocarcinoma (AC) 6.2 (n = 464), large cell (LC) 4.1 (n = 98), squamous (SQ) 4.2 (n = 108) (p = 0.0549). For GPA 3.5–4.0, MSTs did not differ significantly by histology. Differences in MST by histology were noted for GPA 3.0 (p = 0.04), GPA 1.5–2.5 (p = 0.01), and GPA 0–1.0 (p = 0.02). For all patients with brain metastases BM from NSCLC, MSTs by GPA score were: GPA 3.5–4.0, 12.6; GPA 3.0, 10.2; GPA 1.5–2.5, 5.8; and GPA 0–1.0, 2.7. Conclusions: Adenocarcinoma showed a statistically significant higher MST than other histologies of NSCLC for patients with GPA 0–3.0. Using histology as a prognostic factor for BM from NSCLC warrants further investigation. Our cohort of NSCLC BM patients validates the GPA, with MST comparable to that of published data. © 2012 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Brain metastases are diagnosed in approximately 200,000 patients per year, with rising incidence due to improving detection and treatment of systemic malignancy. Lung cancer is the most common primary tumor giving rise to brain metastases [1]. Approximately 50% of all brain metastases patients have a lung primary [2]. Given the variety of treatment options including whole-brain radiation therapy, stereotactic radiosurgery, and surgery, the management of brain metastases has become controversial [3]. Multiple prognostic indices have been published to use clinical factors to stratify brain metastasis patients into homogeneous
∗ Corresponding author at: Department of Radiation Oncology, Cleveland Clinic, 9500 Euclid Ave, Desk T28, Cleveland, OH 44113, United States. Tel.: +1 216 444 5570; fax: +1 216 445 1068. E-mail addresses: [email protected] (S. Guo), [email protected] (G.M. Videtic). 0169-5002/$ – see front matter © 2012 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.lungcan.2012.03.028
survival groups [4–7]. The graded prognostic assessment (GPA) stratifies patients into categories based on age, Karnofsky Performance Status (KPS), extracranial metastases (none and present), and number of metastases (one, two to three, and more than three) (Table 1) [7]. Recent studies have suggested that brain metastases from lung cancer respond to treatment differently than brain metastases from other primary sites [8]. Additional data suggests differential response and outcomes to selected chemotherapy for different NSCLC histologies [9]. Using a large single institutional database of patients with brain metastases, we assessed the impact of histologic subtypes on survival outcomes stratified by the GPA. A secondary endpoint was to validate the GPA with our data set. 2. Methods and materials Data was provided by the Cleveland Clinic brain tumor database, an institutional review board-approved registry of more than 1200
390
S. Guo et al. / Lung Cancer 77 (2012) 389–393
Table 1 Graded prognostic assessment.a
Table 2 Patient, tumor, and treatment characteristics for 780 NSCLC brain metastasis patients.
Score 0 Age KPS Number of brain metastases Extracranial metastases
>60 <70 >3 Present
0.5 50–59 70–80 2–3 –
1.0 <50 90–100 1 None
Abbreviations: KPS = Karnofsky Performance Status; CNS = central nervous system. a Adapted from Ref. [7].
patients with brain metastases. From the interval between January 1982 to September 2004, 780 consecutive patients with newly diagnosed brain metastases from NSCLC and known histologies were identified in this retrospective review (Table 2). Patients recorded in this database were either diagnosed and treated for primary and secondary tumors at our institution as well as at outside hospitals, or referred to the Cleveland Clinic for treatment of brain metastases [10]. Patient cohorts stratified either by GPA class or by histology were analyzed for survival differences. GPA was calculated for each patient based on the published criteria [7]. Tumor histologies were categorized into adenocarcinoma, large cell, squamous, mixed and NOS (not otherwise specified). Median survival times from time of diagnosis were calculated using Kaplan–Meier analysis, and comparisons were made to the published GPA results [7]. The log rank test was used to determine whether a statistically significant difference was present among patient groups. The statistical analyses were performed using StatView version 5.0 (SAS Institute, Cary, NC) and a p value of <0.05 was considered statistically significant.
3. Results For 780 patients, the median follow-up was 5.5 months (range, 0–122.9 months). The median age was 62.3 years (range, 25.3–90.8 years). The median KPS was 80 (range, 20–100). The median number of lesions at diagnosis was 2 (range, 1–50). Extracranial disease was present in 342 patients (43.8%) and absent in 438 patients (56.2%). The number of patients with known histologies was as follows: adenocarcinoma 468 (60.0%), large cell 98 (12.6%), mixed 15 (1.9%), NOS 91 (11.7%), and squamous 108 (13.8%). Patients with mixed and NOS histologies were excluded from GPA by histology analysis because of their relatively small numbers, and were included only in the validation of GPA analysis. Of all patients, 687 patients (88%) received WBRT, 583 patients (75%) received surgery, and 514 patients (65.9%) received stereotactic radiosurgery. Median survival times (MST, in months) by histology were: adenocarcinoma (AC) 6.2 (n = 464), large cell (LC) 4.1 (n = 98), squamous (SQ) 4.2 (n = 108) (p = 0.04549) (Fig. 1). GPA classification variables were identifiable in 751 patients. For GPA 3.5–4.0, MST by histology did not differ significantly (p = 0.8597): AC 11.2 (n = 39), LC 4.1 (n = 2), SQ 10.6 (n = 12) (p = 0.8597). Statistically significant differences in MST by histology were noted for GPA 3.0 (p = 0.0420), GPA 1.5–2.5 (p = 0.0052), and GPA 0–1.0 (p = 0.0205). For GPA 3.0, MST were: AC 12.1 (n = 54), LC 10.2 (n = 14), SQ 6.9 (n = 14) (p = 0.0420). For GPA 1.5–2.5, MST were: AC 6.1 (n = 275), LC 4.6 (n = 49), SQ 3.6 (n = 60) (p = 0.0052). For GPA 0–1.0, MST were: AC 3.2 (n = 83), LC 2.2 (n = 32), SQ 3.2 (n = 17) (p = 0.0205) (Fig. 2a–d; Table 3). For all patients with brain metastases from NSCLC, our MST by GPA score were: GPA 3.5–4.0 (Class I), 12.6; GPA 3.0 (Class II), 10.2; GPA 1.5–2.5 (Class III), 5.8; and GPA 0–1.0 (Class IV), 2.7 (p < 0.0001) (Fig. 3).
Tumor type Adenocarcinoma Large cell Mixed NOS Squamous Gender Female Male GPA score 0 0.5 1 1.5 2 2.5 3 3.5 GPA group 0–1 1.5–2.5 3 3.5–4 Age 50–60 <50 >60 Unknown Number of lesions 1 2 >2 Unknown KPS <70 70–80 90–100 Unknown Extracranial metastases N Y Received WBRT N Y WBRT fractionation 3000 cGy in 10 fx 3750 cGy in 15 fx 4000 cGy in 20 fx Other Unknown Received surgery N Y Received SRS N Y Treatment era 1982–1996 1997–2004 unknown
n
%
468 98 15 91 108
60.0 12.6 1.9 11.7 13.8
326 454
41.8 58.2
11 52 93 133 172 162 101 43
1.4 6.7 11.9 17.1 22.1 20.8 12.9 5.5
156 467 101 56
20.0 59.9 12.9 7.2
212 118 449 1
27.2 15.1 57.6 0.1
324 155 264 37
41.5 19.9 33.8 4.7
152 367 235 26
19.5 47.1 30.1 3.3
438 342
56.2 43.8
93 687
11.9 88.1
328 154 22 161 115
42.1 19.7 2.8 20.6 14.7
583 197
74.7 25.3
514 266
65.9 34.1
322 456 2
41.3 58.5 0.3
4. Discussion Our data is the first report suggesting that lung cancer histology may be prognostic for survival in patients with brain metastases from NSCLC. These improved survivals may reflect inherent biological differences among the various histologies. Recent evidence has shown that different histologies have varied responses to systemic therapy. Scagliotti and colleagues reported that adenocarcinomas treated with cisplatin and pemetrexed showed improved survival compared to treatment with cisplatin and gemcitabine in the firstline setting of advanced NSCLC [9]. Additional data suggests that
S. Guo et al. / Lung Cancer 77 (2012) 389–393
391
Table 3 Median survival (months) by GPA and by histology.
GPA 3.5–4.0 (Class I) (n = 53) GPA 3.0 (Class II) (n = 82) GPA 1.5–2.5 (Class III) (n = 384) GPA 0–1.0 (Class IV) (n = 132)
Adenocarcinoma
Large Cell
Squamous
p value
11.2 (n = 39) 12.1 (n = 54) 6.1 (n = 275) 3.2 (n = 83)
4.1 (n = 2) 10.2 (n = 14) 4.6 (n = 49) 2.2 (n = 32)
10.6 (n = 12) 6.9 (n = 14) 3.6 (n = 60) 3.2 (n = 17)
0.8597 0.0420 0.0052 0.0205
100
Overall Survival (%)
80 Adenocarcinoma (MST=6.2, N=464) Large Cell (MST=4.1, N=98) Squamous (MST=4.2, N=108)
60
40
20 p=0.0549 0 0
12
24
36
48
60
72
84
96
108
Fig. 3. Overall survival for non-small cell lung cancer patients with brain metastases by graded prognostic assessment group (n = 751).
120
Months Fig. 1. Overall survival from time of brain metastases diagnosis by histology for non-small cell lung cancer (n = 670).
brain metastases from lung primaries have different biology compared to other primary sites [11]. For example, subset analysis of RTOG (Radiation Therapy Oncology Group) 95–08 showed that patients with squamous or non-small cell tumors, mostly from lung primaries, had improved survival with whole brain radiation and GPA 3.5-4
100
Adenocarcinoma (n=275, MST= 6.1) Large Cell (n=49, MST= 4.6) Squamous (n=60, MST= 3.6)
Overall Survival (%)
80
60 40 20
60 40 20
p=0.8597
p=0.0052
0
0 0
12
24
36
48 60 Months
72
84
96
0
GPA 3
60 40
24
36
48 60 Months
84
96
108
.6
.4
.2
20
72
Adenocarcinoma (n=83, MST= 3.2) Large Cell (n=32, MST= 2.2) Squamous (n=17, MST= 3.2)
.8 Overall Survival (%)
Adenocarcinoma (n=54, MST= 12.1) Large Cell (n=14, MST= 10.2) Squamous (n=14, MST= 6.9)
80
12
GPA Group 0-1
1
100
Overall Survival (%)
GPA 1.5-2.5
100
Adenocarcinoma (n=39, MST= 11.2) Large Cell (n=2, MST= 4.1) Squamous (n=12, MST= 10.6)
80 Overall Survival (%)
stereotactic radiosurgery compared to whole brain radiation alone [11]. The exact mechanisms by which these biological differences relate to brain metastases remain to be elucidated. We emphasize that our data suggests lung cancer histology as a prognostic factor for improved survival in brain metastasis patients, but do not imply causation for improved survival. Adenocarcinoma is a cell type of NSCLC that has been recognized to metastasize commonly to the brain and to other distant
p=0.0205
p=0.0420 0
0 0
12
24
36
48
60 72 Months
84
96
108 120
0
12
24 Months
Fig. 2. Overall survival for non-small cell lung cancer patient with brain metastases by histology and by graded prognostic assessment group.
392
S. Guo et al. / Lung Cancer 77 (2012) 389–393
Table 4 Presence of extracranial disease by histology.
Adenocarcinoma Large cell Squamous
Number of patients
Extracranial metastases absent
Percentage without extracranial metastases by histology
Extracranial metastases present
Percentage with extracranial metastases by histology
451 97 103
252 53 62
55.9% 54.6% 60.2%
199 44 41
44.1 45.4 39.8
Chi square p-value 0.6786.
sites [12–14]. The incidence of brain metastases from adenocarcinomas is 43% of NSCLC in a retrospective review [15]. Our data suggests that brain metastases from this histology have improved survival compared to brain metastases from other histologies of non-small cell lung cancer. This improvement was seen across all four GPA groups, and was statistically significant in three out of four GPA groups. The outcomes may not have been statistically significant in GPA group 1 due to low patient numbers. Further evaluation of this association with larger patient numbers and multivariate analysis is warranted. Establishing the validity of histology for prognosis could permit improvements in clinical trial design by including this as a stratification factor in future brain metastasis trials. One limitation of our database is that no detailed patterns of failure information are available. Reflecting the relatively short life span of patients with brain metastases, whether these patients failed locally or systemically following brain treatment was rarely discernable or defined, and therefore was not readily captured in our database. To determine if our histology cohorts were balanced with respect to extracranial disease, we re-analyzed our data by these parameters and found no imbalance with respect to numbers of patients with systemic (extracranial) disease between any of the histologies (Table 4). Since the histologic groups were balanced with respect to percentage of patients with and without extracranial disease, and since randomized data had shown no significant advantages with respect to outcomes for any chemotherapy used during the era during which this data was collected (1982–2004), we therefore conclude that our observed improved outcomes based on histology were not confounded by improved control of extracranial disease in any histologic subset [16]. The GPA was based on 1960 analyzable patients from 5 randomized RTOG trials [7]. This initial report of the GPA reported MST of: 11.0 (Class I), 8.9 (Class II), 3.8 (Class III), and 2.6 (Class IV) months. Since the GPA was first published, it has been validated in a series of 140 patients treated at the University of Minnesota between April 2005 and December 2006. In this series, MST were: 21.7 (Class I), 17.5 (Class II), 5.9 (Class III), and 3.0 (Class IV). Longer survival seen in this series compared with our series and other series may be attributed to more recent dates of treatment and advances in systemic therapy [17]. The Diagnosis-Specific Graded Prognostic Assessment, (DSGPA) was later developed to account for inherent variations in prognosis among different primary tumors. This multi-institutional analysis reviewed 5067 patients from 11 institutions treated for BM between 1985 and 2007, of which included 1888 NSCLC patients [18]. All four GPA variables were demonstrated to be prognostic in NSCLC patients, and values are comparable to our data as well as the Minnesota series. Patients who were undergoing repeat treatment for BMs at the time of inclusion in the database were not counted. Our institution was one of many institutions that participated in this larger analysis. Some overlap is present between the patients in our series and this larger series, although our inclusion and exclusion criteria were different. Nieder and colleagues has also validated the GPA in brain metastasis patients in multiple series consisting of a variety of treatment modalities including whole brain radiation alone, with surgery,
and/or with SRS [19,20,21]. One series consisting of 232 patients, the majority of whom were treated with whole brain radiation alone, demonstrated that these patients had a significantly worse prognosis compared to the general population of brain metastasis patients. MST were: 10.3 (Class I), 5.6 (Class II), 3.5 (Class III), and 1.9 (Class IV) [19]. The same group also investigated the GPA in NSCLC patients in particular and also found the GPA to be prognostic in this cohort [19]. MST were: 9.5 (Class I), 4.3 (Class II), 2.5 (Class III), and 3.0 (Class IV). Because the GPA was found to be prognostic in their multiple series and across various treatment modalities for brain metastases, the authors proposed that the GPA was applicable in unselected patients with brain metastases from NSCLC. Patients with brain metastases comprise a heterogenous patient population. More accurate prognostic information may support the decision making process and treatment recommendations for these patients. Furthermore, a select number of these patients have a very limited survival, and more accurate survival prediction models could help to avoid overtreatment. The limitations of our study are the single institutional data, retrospective study design, and long time period for analysis. Although our treatment period encompassed a span of twenty years, there was no difference in overall survival by treatment era of 1982–1996 vs. 1997–2004. Our groups were not balanced with regard to treatment with WBRT, SRS, or surgery; however, data from Nieder and colleagues suggested that the GPA was valid regardless of treatment modality [21]. DS-GPA analysis also showed overlapping confidence intervals between the risk of death and median survival by treatment on multivariate analysis [18]. The current established brain metastases prognostic indices also do not control for the use of systemic treatment. 5. Conclusion Adenocarcinoma showed a higher median survival than other histologies in patients with brain metastases from NSCLC. This difference was seen across multiple GPA groups. The use of histology as a prognostic factor for brain metastases from NSCLC warrants further investigation. Our cohort of patients with brain metastases from NSCLC validated the use of GPA classification for prognosis in this population. Our reported median survivals were consistent with other series that also validated the GPA. Conflict of interest statement None declared. References [1] Nussbaum ES, Djalilian HR, Cho KH, Hall WA. Brain metastases, histology, multiplicity, surgery, and survival. Cancer 1996;78:1781–8. [2] Chang JE, Robins HI, Mehta MP. Therapeutic advances in the treatment of brain metastases. Clin Adv Hematol Oncol 2007;5:54–64. [3] Suh JH. Stereotactic radiosurgery for the management of brain metastases. N Engl J Med 2010;362:1119–27. [4] Lorenzoni J, Devriendt D, Massager N, David P, Ruiz S, Vanderlinden B, et al. Radiosurgery for treatment of brain metastases: estimation of patient eligibility using three stratification systems. Int J Radiat Oncol Biol Phys 2004;60:218–24.
S. Guo et al. / Lung Cancer 77 (2012) 389–393 [5] Gaspar L, Scott C, Rotman M, Asbell S, Phillips T, Wasserman T, 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;37:745–51. [6] Weltman E, Salvajoli JV, Brandt RA, de Morais Hanriot R, Prisco FE, Cruz JC, et al. Radiosurgery for brain metastases: a score index for predicting prognosis. Int J Radiat Oncol Biol Phys 2000;46:1155–61. [7] Sperduto PW, Berkey B, Gaspar LE, Mehta M, Curran W. A new prognostic index and comparison to three other indices for patients with brain metastases: an analysis of 1,960 patients in the RTOG database. Int J Radiat Oncol Biol Phys 2008;70:510–4. [8] Mehta MP, Rodrigus P, Terhaard CH, Rao A, Suh J, Roa W, et al. Survival and neurologic outcomes in a randomized trial of motexafin gadolinium and wholebrain radiation therapy in brain metastases. J Clin Oncol 2003;21:2529–36. [9] Scagliotti GV, Parikh P, von Pawel J, Biesma B, Vansteenkiste J, Manegold C, et al. Phase III study comparing cisplatin plus gemcitabine with cisplatin plus pemetrexed in chemotherapy-naive patients with advanced-stage non-smallcell lung cancer. J Clin Oncol 2008;26:3543–51. [10] Cannady SB, Cavanaugh KA, Lee SY, Bukowski RM, Olencki TE, Stevens GH, et al. Results of whole brain radiotherapy and recursive partitioning analysis in patients with brain metastases from renal cell carcinoma: a retrospective study. Int J Radiat Oncol Biol Phys 2004;58:253–8. [11] Andrews DW, Scott CB, Sperduto PW, Flanders AE, Gaspar LE, Schell MC, et al. Whole brain radiation therapy with or without stereotactic radiosurgery boost for patients with one to three brain metastases: phase III results of the RTOG 9508 randomised trial. Lancet 2004;363:1665–72. [12] Bajard A, Westeel V, Dubiez A, Jacoulet P, Pernet D, Dalphin JC, et al. Multivariate analysis of factors predictive of brain metastases in localised non-small cell lung carcinoma. Lung Cancer 2004;45:317–23.
393
[13] Strauss B, Weller CV. Bronchogenic carcinoma; a statistical analysis of two hundred ninety-six cases with necropsy as to relationships between cell types and age, sex, and metastasis. AMA Arch Pathol 1957;63:602–11. [14] Figlin RA, Piantadosi S, Feld R. Intracranial recurrence of carcinoma after complete surgical resection of stage I, II, and III non-small-cell lung cancer. N Engl J Med 1988;318:1300–5. [15] Mujoomdar A, Austin JH, Malhotra R, Powell CA, Pearson GD, Shiau MC, et al. Clinical predictors of metastatic disease to the brain from non-small cell lung carcinoma: primary tumor size, cell type, and lymph node metastases. Radiology 2007;242:882–8. [16] Schiller JH, Harrington D, Belani CP, Langer C, Sandler A, Krook J, et al. Comparison of four chemotherapy regimens for advanced non-small-cell lung cancer. N Engl J Med 2002;346:92–8. [17] Sperduto CM, Watanabe Y, Mullan J, Hood T, Dyste G, Watts C, et al. A validation study of a new prognostic index for patients with brain metastases: the Graded Prognostic Assessment. J Neurosurg 2008;109(Suppl.):87–9. [18] Sperduto PW, Chao ST, Sneed PK, Luo X, Suh J, Roberge D, et al. Diagnosisspecific prognostic factors, indexes, and treatment outcomes for patients with newly diagnosed brain metastases: a multi-institutional analysis of 4,259 patients. Int J Radiat Oncol Biol Phys 2010;77:655–61. [19] Nieder C, Marienhagen K, Geinitz H, Molls M. Validation of the graded prognostic assessment index for patients with brain metastases. Acta Oncol 2009;48:457–9. [20] Nieder C, Geinitz H, Molls M. Validation of the graded prognostic assessment index for surgically treated patients with brain metastases. Anticancer Res 2008;28:3015–7. [21] Nieder C, Bremnes RM, Andratschke NH. Prognostic scores in patients with brain metastases from non-small cell lung cancer. J Thorac Oncol 2009;4:1337–41.