Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Prognostic indices for brain metastases – usefulness and challenges

Prognostic indices for brain metastases – usefulness and challenges Background: This review addresses the strengths and weaknesses of 6 different prognostic indices, published since the Radiation Therapy Oncology Group (RTOG) developed and validated the widely used 3-tiered prognostic index known as recursive partitioning analysis (RPA) classes, i.e. between 1997 and 2008. In addition, other analyses of prognostic factors in groups of patients, which typically are underrepresented in large trials or databases, published in the same time period are reviewed. Methods: Based on a systematic literature search, studies with more than 20 patients were included. The methods and results of prognostic factor analyses were extracted and compared. The authors discuss why current data suggest a need for a more refined index than RPA. Results: So far, none of the indices has been derived from analyses of all potential prognostic factors. The 3 most recently published indices, including the RTOG's graded prognostic assessment (GPA), all expanded from the primary 3-tiered RPA system to a 4-tiered system. The authors' own data confirm the results of the RTOG GPA analysis and support further evaluation of this tool. Conclusion: This review provides a basis for further refinement of the current prognostic indices by identifying open questions regarding, e.g., performance of the ideal index, evaluation of new candidate parameters, and separate analyses for different cancer types. Unusual primary tumors and their potential differences in biology or unique treatment approaches are not well represented in large pooled analyses. indices might also be used as inclusion/exclusion criteria Background Prognostic indices might represent a useful tool in pallia- for clinical trials and for comparison of results across dif- tive cancer treatment. Estimation of a patient's prognosis ferent studies in relatively homogeneous patient groups. in terms of overall survival might allow for tailored treat- ment, i.e. more aggressive approaches when these are Brain metastases continue to represent a formidable chal- likely to impact on survival and focus on disease stabilisa- lenge in oncology [1-3]. With increasing numbers of local tion, symptom control and toxicity minimization when and systemic treatment options, the issue of patient selec- the disease is more advanced, or comorbidity limits the tion gains importance. While surgery and stereotactic radi- tolerability of aggressive therapy. In addition, prognostic osurgery (SRS) provide long-term local control of Page 1 of 11 (page number not for citation purposes) Radiation Oncology 2009, 4:10 http://www.ro-journal.com/content/4/1/10 macroscopic disease and in combination with whole- tions. Their definition yielded class IIIa defined as age <65 brain radiotherapy (WBRT) the best available overall years, controlled primary tumor and single brain metasta- brain control for the remaining life time [4-10], they rep- sis, class IIIc defined as age ≥ 65 years, uncontrolled pri- resent overtreatment in patients with short survival, which mary tumor and multiple brain metastases, while other typically is caused by uncontrollable systemic disease. patients would make up class IIIb. The original RPA clas- This review will address the strengths and weaknesses of 6 sification has been validated by several authors, both in different prognostic indices, published since the Radia- selected and unselected patient groups, e.g., patients with tion Therapy Oncology Group (RTOG) developed and breast primary, lung primary (small cell and non-small validated the widely used 3-tiered prognostic index cell), malignant melanoma, unknown primary, or surgi- known as recursive partitioning analysis (RPA) classes cal resection and SRS as main local treatment modalities [11,12], i.e. between 1997 and 2008. In addition, other [14-35]. analyses of prognostic factors in groups of patients, which typically are underrepresented in large trials or databases, Probably, the surgically treated patients represent the published in the same time period are reviewed. These most homogeneous cohorts assessed with the RPA sys- include patients with primary tumors that do not com- tem, as these were patients with rather favourable progno- monly metastasize to the brain, and the elderly, who are sis, fit to undergo surgery and with limited brain disease. often either excluded or under-represented in clinical tri- Nevertheless, the differences in median survival between als. the individual studies were large. In RPA class I, median survival ranged from 15–29 months [31-35]. In class II, a survival range of 5.5–11 months has been reported. In Methods The present review compares different prognostic indices class III, these figures reached 1.4–9 months. As illustrated and analyses of prognostic factors based on a systematic here, survival within the same RPA class might vary by a literature search by use of Medline (Pub Med by the factor of 2 or more between different studies (identical National Library of Medicine, National Institutes of treatment approach). In series where the majority of Health, Bethesda, Maryland, USA). It is limited to adult patients were treated with WBRT, less variation between patients having received first-line treatment for parenchy- studies can be found (Figure 1). As shown in Table 1, both mal brain metastases in the absence of leptomeningeal RPA class II and III contain quite heterogeneous groups of disease. The key words used were "brain metastases", patients. The factor determining class III is KPS<70, which "metastatic brain tumor" and "cerebral metastases". The might result from many different causes including the final search was performed on June 30, 2008. It also brain metastases themselves, advanced and treatment- included the reference lists of all articles and the appropri- refractory extracranial metastases, severe pain or patho- ate chapters in textbooks on brain metastases, neuro- logical fracture in patients with bone metastases, atelecta- oncology and radiation oncology. Case reports and review sis or pneumonia from primary lung cancer, anemia articles were not assessed. Only studies with more than 20 induced by chemotherapy, recovery from recent surgery, patients were included. If several subsequent reports were and non-cancer-related comorbidity. In all the reports published from the same institution, the most recent pub- reviewed variable proportions of patients in the most lication was evaluated. The methods and results of prog- favourable RPA class I unexpectedly died within 2 nostic factor analyses were extracted and compared. months, while some patients in class III survived for more than 6 months. For these reasons, there obviously is a need for a more refined index than RPA. Results The search identified 6 different prognostic indices, which are shown in Table 1. Comparison of the patients' charac- The first attempt in 1999 resulted in the Rotterdam Score, teristics is shown in Table 2. Unfortunately, a considera- which did not gain wider acceptance [36]. Similar to RPA, ble amount of information can not be extracted from the performance status and extent of systemic disease were publications. The most widely used index over the last included, while the third parameter was response to ster- decade is the RPA index originally described by Gaspar et oids before WBRT. It can be assumed that the unavailabil- al. on behalf of the RTOG [11], which is based on 4 ity of this latter parameter in most databases or patient parameters (age, Karnofsky performance status (KPS), records prevented other groups from using the score. In presence or absence of extracranial metastases, and the addition, the definition of systemic tumor activity is not control status of the primary tumor), separating patients straight forward. The next attempt (Score Index for Radio- into 3 different classes. Lutterbach et al. suggested expan- surgery (SIR)) was derived from a limited number of sion of the classification by further dividing class III into patients treated with this particular focal approach, which 3 separate classes [13]. This was based on their multivari- might have resulted in overfitting of the data [37]. How- ate analysis of 916 patients from a single institution, but ever, several groups confirmed the performance of the SIR was not adopted by other authors in subsequent publica- in patients treated with SRS, surgery, and WBRT with or Page 2 of 11 (page number not for citation purposes) Table 1: Comparison of the prognostic scores published since 1997, empty fields indicate that a parameter is not used in the index Score Performanc Age Extracranial Controlled Steroid Number of Volume of Interval to Class I Class II Class III Class IV e status metastases primary treatment BM BM BM RPA KPS <65 years no vs yes no vs yes all 4 other KPS <70 none Derived ≥ 70 vs <70 favourable patients from 3 factors prospective RTOG studies, n = 1,200 Rotterdam ECOG limited systemic good, ECOG 0–1 other ECOG2-3 none Single 0–1 vs 2–3 activity vs extensive* moderate with no or patients with limited institution, or little limited or extensive n = 1,292 response systemic systemic tumor activity and activity and little good response to response to steroids steroids SIR KPS 80– ≤ 50: 2 no evidence disease: 1 1: 2 points largest 8–10 points 4–7 points 1–3 points none Single 100:2 points points of systemic point 2: 1 point lesion institution, KPS 60–70: 51–59: 1 disease or progressive ≥ 3: 0 points volume <5 n = 65 1 point point complete disease: 0 cc: 2 points KPS ≤ 50: 0 ≥ 60: 0 remission: 2 points 5–13 cc: 1 points points points point partial >13 cc: 0 remission or points stable BSBM KPS 80–100: no: 1 point yes: 1 point 3 points 2 points 1 point 0 points Single 1 point yes: 0 points no: 0 points institution, KPS ≤ 70: 0 n = 110 point GPA KPS 90–100: <50: 1 point none: 1 1: 1 point 3.5–4 points 3 points 1.5–2.5 0–1 points Derived 1 point 50–59: 0.5 point 2–3: 0.5 points from 5 KPS 70–80: points present: 0 points prospective 0.5 points >60: 0 points >3: 0 points RTOG KPS <70: 0 points studies, points n = 1,960 Rades et KPS ≥ 70: 5 ≤ 60: 4 none: 5 >8 mo: 4 17–18 14–16 11–13 9–10 points al. points points points points points points points Multi- KPS <70: 1 >60: 3 present: 2 ≤ 8 mo: institutional, point points points 3 points n = 1,085 BM: brain metastases, RPA: recursive partitioning analysis, RTOG: Radiation Therapy Oncology Group, KPS: Karnofsky performance score, SIR: score index for radiosurgery, BSBM: basic score for brain metastases, GPA: graded prognostic assessment, ECOG: Eastern Cooperative Oncology Group * limited systemic activity: no systemic metastases but progression of primary tumor or systemic metastases with primary tumor absent or controlled; extensive systemic activity: systemic metastases and progressive primary Radiation Oncology 2009, 4:10 http://www.ro-journal.com/content/4/1/10 Page 3 of 11 (page number not for citation purposes) Radiation Oncology 2009, 4:10 http://www.ro-journal.com/content/4/1/10 Table 2: Median values of reported patients' characteristics in each of the studies, empty fields indicate missing information Score Performance Age Extracranial Controlled Steroid Number of Volume of Interval to status metastases primary treatment BM BM BM RPA KPS 70 55–59 yrs. 38% 60% 2 n = 1,200 range Rotterdam ECOG 1 59 yrs. mean 15 mg 2 8.5 mo. n = 1,292 dexamethason e per day SIR KPS 80 61 yrs. 2 3.3 cc n = 65 BSBM 57 yrs. 2 9 cc n = 110 GPA KPS 80 60 yrs. 36% 67% 2 5–13 cc n = 1,960 Rades et al. KPS 70 60 yrs. 64% 8 mo. n = 1,085 BM: brain metastases, RPA: recursive partitioning analysis, KPS: Karnofsky performance score, SIR: score index for radiosurgery, BSBM: basic score for brain metastases, GPA: graded prognostic assessment, ECOG: Eastern Cooperative Oncology Group RTOG 1997 RTOG 2000 Nieder et al. 2000 Lutterbach et al. Months Saito et al. 2006 RTOG 2008 Nieder et al. 2008 Rades et al. 2008 RPA I RPA II RPA III Compar or without Figure 1 ison of median survival in 7 stud local measures, none of the studie ies uss ing the recursiv is limited to one pa e partit rtic ioning ular cancer ty analyses ( pe) RPA) classes (treatment was WBRT with Comparison of median survival in 7 studies using the recursive partitioning analyses (RPA) classes (treatment was WBRT with or without local measures, none of the studies is limited to one particular cancer type). without SRS, some of them with large numbers of patients patients managed with WBRT with or without SRS and (Figure 2) [35,38-44]. To accurately define systemic dis- surgery plus WBRT [35,42,44], however its performance is ease activity, comprehensive diagnostic work-up is not better than that of the other scores (Figure 3). needed. The RTOG has recently proposed a new index, which was When evaluating the SIR and RPA indices in their SRS compared to RPA, SIR, and BSBM (but not to the Rotter- database, the group from Brussels, Belgium, arrived at a dam score) [44]. The new score (Graded Prognostic new score, which they called Basic Score for Brain Metas- Assessment (GPA)) is different from RTOG's RPA, e.g., tases (BSBM) [43]. Based on its greater convenience and with regard to the number of prognostic classes, which simplicity, they advocated the use of this score, which uses increased from 3 to 4, and the larger number of patients. the same definition of extracranial disease activity as the The analysis also includes patients managed with WBRT RTOG. Recent data indicate that BSBM can be applied to plus SRS from RTOG study 9508 [5]. In the GPA system, Page 4 of 11 (page number not for citation purposes) Radiation Oncology 2009, 4:10 http://www.ro-journal.com/content/4/1/10 RTOG 2008 Months Nieder et al. SIR 8- SIR 4-7 SIR 1-3 C local Figure 2 omp mea arisso ures, studies n of median s not lim urvivali i te n 2 d to one particular studies using the s cancer type) core index for radiosurgery (SIR) (treatment was WBRT with or without Comparison of median survival in 2 studies using the score index for radiosurgery (SIR) (treatment was WBRT with or without local measures, studies not limited to one particular cancer type). RTOG 2008 Months Nieder et al. BSBM 3 BSBM 2 BSBM 1 BSBM 0 without Figure 3 Compar local measur ison of median survival in es, studies not 2 stud limited to ies using one particu the basic scor lar cancer typ e for brae) in metastases (BSBM) (treatment was WBRT with or Comparison of median survival in 2 studies using the basic score for brain metastases (BSBM) (treatment was WBRT with or without local measures, studies not limited to one particular cancer type). 3 different values (0, 0.5 or 1) are assigned for each of no longer mandated. It was concluded by the authors that these 4 parameters: age (≥ 60; 50–59; <50), KPS (<70; 70– "GPA is the least subjective, most quantitative and easiest 80; 90–100), number of brain metastases (>3; 2–3; 1), to use of the 4 indices" and that future trials should com- and extracranial metastases (present; not applicable; pare these scores and validate the GPA. One of the none). Assessment of primary tumor activity or control is authors' group has embarked on this comparison in 2 dif- Page 5 of 11 (page number not for citation purposes) Radiation Oncology 2009, 4:10 http://www.ro-journal.com/content/4/1/10 ferent patient populations, i.e. those managed with WBRT a patient, that can be used to estimate the chance of recov- with or without SRS (comparable to the RTOG study pop- ery from a disease or the chance of the disease recurring. Based on such prognostic factors, 6 different prognostic ulation) [42] and those managed with surgery and WBRT [35]. Both studies basically relied on the methods used by indices for adult patients with brain metastases from solid the RTOG in their analysis, though with patients treated in tumors have been developed over the last decade. As dem- clinical routine outside of randomized trials. Compared onstrated in Table 1, the 3 most recently published indices to RTOG's patients treated with WBRT with or without all expanded from the primary 3-tiered RPA system to a 4- SRS, the median age, KPS, number of lesions and lesion tiered system. The 6 indices are based on a different volume were similar. Obvious differences existed, how- number of prognostic factors, i.e. 3–6. Of course, increas- ever, regarding controlled primary tumor (47 vs. 67%) ing numbers of parameters will lead to less convenience and extracranial metastases (56 vs. 36%). Thus, the cohort and ease of administration. None of the groups that devel- is expected to have inferior survival. Figure 4 shows the oped these indices included all potential prognostic fac- survival results. tors in their analysis. This is most likely due to the unavailability of all the information in the databases and Last but not least, Rades et al. developed a new prognostic the difficulty in collecting missing data in 1,000 or more index based on 4 parameters (age, KPS, extracranial patients treated over many years. As can be seen in Figures metastases at the time of WBRT, interval from tumor diag- 3, 4, 5, the performance of the 4-tiered indices is not tre- nosis to WBRT) [45]. The major difference from the RPA mendously different, although further data are needed to classes is the replacement of primary tumor control by confirm this finding. interval from tumor diagnosis to WBRT (not by number of brain metastases as in the GPA). This index separated There is agreement in all indices on the importance of per- patients into 4 subgroups with significantly different formance status and extracranial disease activity. How- prognosis and was also validated in one of the authors' ever, whether both primary tumor and extracranial database (unpublished results, Figure 5). metastases should be considered is less clear (2 indices would not include primary tumor control). Assessment of Discussion extracranial disease status is not trivial. It might require As stated on the website of the National Cancer Institute considerable resources in patients with very limited life http://www.cancer.gov/templates/ expectancy and therapeutic options. When collecting data bd_alpha.aspx?CdrID=44246, a prognostic factor is over long time periods, one must expect a shift in diagnos- regarded as a situation or condition, or a characteristic of tic modalities, i.e. increasing use of magnetic resonance RTOG 2008 Months Nieder et al. GPA GPA 3 GPA GPA 0-1 3.5-4 1.5-2.5 Compar without Figure 4 local measur ison of median survival in es, studies not 2 st limited to udies using one particu the graded prognostic assess lar cancer type) ment (GPA) (treatment was WBRT with or Comparison of median survival in 2 studies using the graded prognostic assessment (GPA) (treatment was WBRT with or without local measures, studies not limited to one particular cancer type). Page 6 of 11 (page number not for citation purposes) Radiation Oncology 2009, 4:10 http://www.ro-journal.com/content/4/1/10 Rades et al. Months Nieder et al. 17-18 14-16 11-13 9-10 points points points points Compar Figure 5 ison of median survival in 2 studies using the index proposed by Rades et al Comparison of median survival in 2 studies using the index proposed by Rades et al. [45](treatment was WBRT with or without local measures, studies not limited to one particular cancer type, median survival estimated from the Kaplan-Meier curves in the publication). imaging of the brain as compared to computed tomogra- tumors in these models. Breast cancer poses an interesting phy (CT) or increasing use of chest CT or even positron dilemma here, because although tumor type and histology emission tomography (PET). Such a shift will likely result were not prognostically significant in the RPA, recent data, in no longer assigning patients to the most favourable especially since the advent of trastuzumab and lapatinib, prognostic class (stage migration). This might compro- suggest that, receptor status and her-2-neu expression mise the comparison of the different studies. might have prognostic impact, even if this issue is not with- out controversy (Table 3). The recently suggested prognos- Two of the 6 indices did not include age and the ones that tic factor lymphopenia falls into the same category [19,51]. did, used slightly different cut-off values. A minority of stud- Unusual primary tumors and their potential differences in ies (n = 2) included number of brain metastases and only biology or unique treatment approaches are not well repre- one each included response to steroids, volume of the largest sented in large pooled analyses. Table 4 provides examples lesion in the brain, and time interval to development of on analyses of prognostic factors in such groups. brain metastases, respectively. Other previous reports lend credence to the examination of each of these factors. In their Surrogate markers of disease activity that are easy to meas- multivariate analysis of 334 patients, DiLuna et al. reported ure and inexpensive, such as lactate dehydrogenase and significantly better survival in patients with 1–3 vs 4 or more other laboratory parameters have repeatedly been shown brain metastases and in those patients with both limited to be independent prognostic factors for survival [71-74]. number and volume of brain metastases (<5 cc total vol- Studies that were not limited to patients with brain metas- ume) [46]. Bhatnagar et al. also reported on the impact of tases suggest that the anorexia-cachexia syndrome, dysp- treatment volume as independent prognostic factor in nea, pain, and co-morbidity are further candidates for patients treated with SRS [47]. In a randomised trial with prospective evaluation [74]. The same holds true for neu- 544 patients, Priestman et al. found that dose of steroids was rofunction class [75,76] and mini mental status examina- independently associated with survival [48]. Interval to tion results, which was an independent prognostic factor development of brain metastases appears particularly impor- for survival in a multivariate model that also included KPS tant in patients with primary NSCLC and malignant [77]. The current prognostic indices unfortunately do not melanoma. The multivariate analyses of 3 studies with 292– incorporate these features. 686 patients support this observation [23,49,50]. One of the purposes of prognostic indices is to guide the The latter findings lead to the general question on the use- choice of treatment in individual patients. In this context, fulness of lumping together patients with different primary a prognostic index should be accurate enough to avoid Page 7 of 11 (page number not for citation purposes) Radiation Oncology 2009, 4:10 http://www.ro-journal.com/content/4/1/10 Table 3: Prognostic impact of hormone receptor and HER-2 status in patients with brain metastases from breast cancer n Prognostic impact of hormone receptor status Prognostic impact of HER-2 status Claude et al. 120 none not examined Bartsch et al. 174 none None Le Scodan et al. 117 receptor negative significantly worse None Nam et al. 126 receptor negative significantly worse HER-2 negative significantly worse Kirsch et al. 95 not examined HER-2 negative significantly worse* 55 ^ Eichler et al. 83 none HER-2 negative significantly worse Melisko et al. 112 receptor negative significantly worse none Harputluoglu et al. 144 none none Park et al. 125 none HER-2 positive significantly worse 86 not examined HER-2 negative significantly worse* Church et al. 80% of HER-2 overexpressing cases received trastuzumab after diagnosis of brain metastases * the difference in survival was limited to patients with HER-2 overexpressing cancer treated with trastuzumab after diagnosis of brain metastases overtreatment in patients that actually have very short sur- systems, the most favorable prognostic group is very small vival. Even more important, one should not withhold (e.g., GPA ≥ 3.5: 9% of RTOG and 7% of our own patients; treatment because the index erroneously predicts an unfa- RPA class I: 16% of RTOG and 11% of our own patients). vorable outcome. These aspects of the indices have not been thoroughly evaluated, even in the recent GPA analy- The open questions after publication of 6 prognostic indi- sis [44]. In our analysis of 239 patients, which confirms ces include: that RPA, SIR, BSBM and GPA each split the dataset into groups with significantly different prognosis, this issue - how should the ideal index perform? was addressed [42]. With regard to the outcome of patients with unfavorable survival, defined as ≤ 2 months - how many parameters should form the basis of the (n = 93), no significant difference between the indices was ideal index? observed. Regarding patients with favorable survival, defined as ≥ 6 months (n = 66), again no significant dif- - can we lump together patients with breast cancer, ference was observed, although RPA performed worse small-cell lung cancer, malignant melanoma etc. or do than the other indices. Overall, GPA misassigned 6% of we lose potentially important information? the patients (9 out of 159), compared to 11% with RPA. Therefore, the available validation data certainly do not - do we need candidate parameters beyond the ones discourage further evaluation of the new GPA. However, examined so far (lactate dehydrogenase, anemia, such evaluation should also include comparison with the weight loss, pain etc.)? 2 other scores (Rotterdam and Rades et al.). It is just the stark reality of the disease process that in all of the scoring Table 4: Prognostic factors in patients underrepresented in large studies (minimum number of patients n = 20) Author Population Significant prognostic factors Ogawa et al. esophageal cancer, n = 36 KPS, aggressive local treatment (multivariate) Weinberg et al. esophageal cancer, n = 27 no liver metastases, RPA class I (trend, p = 0.1, multivariate) Khuntia et al. esophageal cancer, n = 27 KPS, aggressive local treatment (multivariate) Cohen et al. ovarian cancer, n = 72 aggressive local treatment Cormio et al. ovarian cancer, n = 22 extracranial disease, time to development of brain metastases Growdon et al. gynaecological cancers, n = 30 extracranial disease, histology, use of chemotherapy (multivariate) Tremont-Lukats et al. prostate cancer, n = 103 adenocarcinoma vs other histology Rades et al. unknown primary, n = 101 KPS, extracranial metastases, RPA class Bartelt and Lutterbach unknown primary, n = 47 KPS, surgical resection status (multivariate) Ruda et al. unknown primary, n = 33 number of brain metastases (multivariate) Kim et al. patients ≥ 75 years, SRS treatment, n = 44 single brain metastasis, NSCLC vs other primary Noel et al. patients ≥ 65 years, SRS treatment, n = 117 KPS (multivariate) WBRT: whole-brain radiotherapy, KPS: Karnofsky performance status, RPA: recursive partitioning analysis, SRS: stereotactic radiosurgery, NSCLC: non-small cell lung cancer Page 8 of 11 (page number not for citation purposes) Radiation Oncology 2009, 4:10 http://www.ro-journal.com/content/4/1/10 6. Kondziolka D, Patel A, Lunsford LD, Kassam A, Flickinger JC: Stere- - is it justifiable to assign the same point value to dif- otactic radiosurgery plus whole brain radiotherapy versus ferent degrees of extracranial disease, e.g., 2 small radiotherapy alone for patients with multiple brain metas- asymptomatic lung metastases, 8 large liver metastases tases. Int J Radiat Oncol Biol Phys 1999, 45:427-434. 7. Patchell RA, Tibbs PA, Regine WF, Dempsey RJ, Mohiuddin M, Kry- with increased bilirubin, skin metastases already scio RJ, Markesbery WR, Foon KA, Young B: Postoperative radio- treated by radiotherapy etc.? therapy in the treatment of single metastases to the brain: a randomized trial. JAMA 1998, 80:1485-1489. 8. Noordijk EM, Vecht CJ, Haaxma-Reiche H, Padberg GW, Voormolen - can international groups collaborate to develop a JH, Hoekstra FH, Tans JT, Lambooij N, Metsaars JA, Wattendorf AR: consensus score, or maybe even an online tool? The choice of treatment of single brain metastasis should be based on extracranial tumor activity. Int J Radiat Oncol Biol Phys 1994, 29:711-717. Other aspects of predicting the outcome in patients with 9. Patchell RA, Tibbs PA, Walsh JW, Dempsey RJ, Maruyama Y, Kryscio brain metastases that many clinicians might appreciate, RJ, Markesbery WR, Macdonald JS, Young B: A randomized trial of surgery in the treatment of single metastases to the brain. N relate to the important issue of neurologic function and Engl J Med 1990, 322:494-500. quality of life. In many instances, radiotherapy aims more 10. Nieder C, Astner ST, Grosu AL, Andratschke NH, Molls M: The role of postoperative radiotherapy after resection of a single on improving deficits and preventing neurologic decline brain metastasis: combined analysis of 643 patients. Strahlen- than prolonging survival, but no attempts have been ther Onkol 2007, 183:576-580. made to develop scores that address endpoints other than 11. Gaspar L, Scott C, Rotman M, Asbell S, Phillips T, Wasserman T, McKenna WG, Byhardt R: Recursive partitioning analysis (RPA) overall survival. It appears therefore worthwhile to collect of prognostic factors in three Radiation Therapy Oncology data on such endpoints, as done, e.g., in the recently com- Group (RTOG) brain metastases trials. Int J Radiat Oncol Biol Phys 1997, 37:745-751. pleted randomized trial of radiotherapy with or without 12. Gaspar LE, Scott C, Murray K, Curran W: Validation of the RTOG motexafin gadolinium [78], which used time to neuro- recursive partitioning analysis (RPA) classification for brain logic progression as primary endpoint. Other opportuni- metastases. Int J Radiat Oncol Biol Phys 2000, 47:1001-1006. 13. Lutterbach J, Bartelt S, Stancu E, Guttenberger R: Patients with ties for future research include examination of prognostic brain metastases: hope for recursive partitioning analysis models that provide estimates on both risk of systemic (RPA) class 3. Radiother Oncol 2002, 63:339-345. cancer progression with death from non-neurologic 14. Nieder C, Nestle U, Motaref B, Walter K, Niewald M, Schnabel K: Prognostic factors in brain metastases: should patients be causes and risk of death from uncontrolled brain metas- selected for aggressive treatment according to recursive tases. partitioning analysis (RPA) classes? Int J Radiat Oncol Biol Phys 2000, 46:297-302. 15. Fleckenstein K, Hof H, Lohr F, Wenz F, Wannenmacher M: Prognos- Competing interests tic factors for brain metastases after whole brain radiother- The authors declare that they have no competing interests. apy. Data from a single institution. Strahlenther Onkol 2004, 180:268-273. 16. Saito EY, Viani GA, Ferrigno R, Nakamura RA, Novaes PE, Pellizzon Authors' contributions CA, Fogaroli RC, Conte MA, Salvajoli JV: Whole brain radiation CN and MM drafted the manuscript and participated in therapy in management of brain metastasis: results and prognostic factors. Radiat Oncol 2006, 1:20. the design of the study. Both authors read and approved 17. Mahmoud-Ahmed AS, Suh JH, Lee SY, Crownover RL, Barnett GH: the final manuscript. Results of whole brain radiotherapy in patients with brain metastases from breast cancer: a retrospective study. Int J Radiat Oncol Biol Phys 2002, 54:810-817. Acknowledgements 18. Viani GA, Castilho MS, Salvajoli JV, Pellizzon AC, Novaes PE, Guima- None rães FS, Conte MA, Fogaroli RC: Whole brain radiotherapy for brain metastases from breast cancer: estimation of survival using two stratification systems. BMC Cancer 2007, 7:53. References 19. Le Scodan R, Massard C, Mouret-Fourme E, Guinebretierre JM, 1. Kunthia D, Brown P, Li J, Mehta MP: Whole-brain radiotherapy in Cohen-Solal C, De Lalande B, Moisson P, Breton-Callu C, Gardner M, the management of brain metastasis. J Clin Oncol 2006, Goupil A, Renody N, Floiras JL, Labib A: Brain metastases from 24:1295-1304. breast carcinoma: validation of the Radiation Therapy 2. Langer CJ, Mehta MP: Current management of brain metas- Oncology Group recursive partitioning analysis classification tases, with a focus on systemic options. J Clin Oncol 2005, and proposition of a new prognostic score. Int J Radiat Oncol Biol 23:6207-6219. Phys 2007, 69:839-845. 3. Nieder C, Grosu AL, Astner ST, Thamm R, Molls M: Integration of 20. Kepka L, Cieslak E, Bujko K, Fijuth H, Wierzchowski M: Results of chemotherapy into current treatment strategies for brain the whole-brain radiotherapy for patients with brain metas- metastases from solid tumors. Radiat Oncol 2006, 1:19. tases from lung cancer: the RTOG RPA intra-classes analy- 4. Aoyama H, Shirato H, Tago M, Nakagawa K, Toyoda T, Hatano K, sis. Acta Oncol 2005, 44:389-398. Kenjyo M, Oya N, Hirota S, Shioura H, Kunieda E, Inomata T, Hay- 21. Videtic GM, Adelstein DJ, Mekhail TM, Rice TW, Stevens GH, Lee SY, akawa K, Katoh N, Kobashi G: Stereotactic radiosurgery plus Suh JH: Validation of the RTOG recursive partitioning analy- whole-brain radiation therapy vs stereotactic radiosurgery sis (RPA) classification for small-cell lung cancer-only brain alone for treatment of brain metastases. A randomized con- metastases. Int J Radiat Oncol Biol Phys 2007, 67:240-243. trolled trial. JAMA 2006, 295:2483-2491. 22. Gülbas H, Erkal HS, Serin M: The use of recursive partitioning 5. Andrews DW, Scott CB, Sperduto PW, Flanders AE, Gaspar LE, analysis grouping in patients with brain metastases from Schell MC, Werner-Wasik M, Demas W, Ryu J, Bahary JP, Souhami L, non-small-cell lung cancer. Jpn J Clin Oncol 2006, 36:193-196. Rotman M, Mehta MP, Curran WJ Jr: Whole brain radiation ther- 23. Rades D, Schild SE, Lohynska R, Veninga T, Stalpers LJ, Dunst J: Two apy with or without stereotactic radiosurgery boost for radiation regimens and prognostic factors for brain metas- patients with one to three brain metastases: phase III results tases in nonsmall cell lung cancer patients. Cancer 2007, of the RTOG 9508 randomised trial. Lancet 2004, 110:1077-1082. 363:1665-1672. Page 9 of 11 (page number not for citation purposes) Radiation Oncology 2009, 4:10 http://www.ro-journal.com/content/4/1/10 24. Rades D, Bohlen G, Lohynska R, Veninga T, Stalpers LJ, Schild SE, 42. Nieder C, Molls M: Validation of the graded prognostic assess- Dunst J: Whole-brain radiotherapy with 20 Gy in 5 fractions ment index for patients with brain metastases: in regards to for brain metastases in patients with cancer of unknown pri- Sperduto et al. (Int J Radiat Oncol Biol Phys 2008;70:510– mary (CUP). Strahlenther Onkol 2007, 183:631-636. 514). Int J Radiat Oncol Biol Phys 2008, 72:1619. 25. Kanner AA, Suh JH, Siomin VE, Lee SY, Barnett GH, Vogelbaum MA: 43. Lorenzoni J, Devriendt D, Massager N, David P, Ruiz S, Vanderlinden Posterior fossa metastases: aggressive treatment improves B, Van Houtte P, Brotchi J, Levivier M: Radiosurgery for treat- survival. Stereotact Funct Neurosurg 2003, 81:18-23. ment of brain metastases: Estimation of patient eligibility 26. Nam TK, Lee JI, Jung YJ, Im YS, An HY, Nam DH, Park K, Kim JH: using three stratification systems. Int J Radiat Oncol Biol Phys Gamma knife surgery for brain metastases in patients har- 2004, 60:218-224. boring four or more lesions: survival and prognostic factors. 44. Sperduto PW, Berkey B, Gaspar LE, Mehta M, Curran W: A new J Neurosurg 2005, 102(suppl):147-150. prognostic index and comparison to three other indices for 27. Buchsbaum JC, Suh JH, Lee SY, Chidel MA, Greskovich JF, Barnett patients with brain metastases: an analysis of 1,960 patients GH: Survival by Radiation Therapy Oncology Group recur- in the RTOG database. Int J Radiat Oncol Biol Phys 2008, sive partitioning analysis class and treatment modality in 70:510-514. patients with brain metastases from malignant melanoma. 45. Rades D, Dunst J, Schild SE: A new scoring system to predicting Cancer 2002, 94:2265-2272. the survival of patients treated with whole-brain radiother- 28. Harrison BE, Johnson JL, Clough RW, Halperin EC: Selection of apy for brain metastases. Strahlenther Onkol 2008, 184:251-255. patients with melanoma brain metastases for aggressive 46. DiLuna ML, King JT Jr, Knisely JP, Chiang VL: Prognostic factors for treatment. Am J Clin Oncol 2003, 26:354-357. survival after stereotactic radiosurgery vary with the 29. Morris SL, Low SH, A'Hern RP, Eisen TG, Gore ME, Nutting CM, Har- number of cerebral metastases. Cancer 2007, 109:135-145. rington KJ: A prognostic index that predicts outcome follow- 47. Bhatnagar AK, Kondziolka D, Lunsford LD, Flickinger JC: Recursive ing palliative whole brain radiotherapy for patients with partitioning analysis of prognostic factors for patients with metastatic malignant melanoma. Br J Cancer 2004, 91:829-833. four or more intracranial metastases treated with radiosur- 30. Chidel MA, Suh JH, Reddy CA, Chao ST, Lundbeck MF, Barnett GH: gery. Technol Cancer Res Treat 2007, 6:153-160. Application of recursive partitioning analysis and evaluation 48. Priestman TJ, Dunn J, Brada M, Rampling R, Baker PG: Final results of the use of whole brain radiation among patients treated of the Royal College of Radiologists' trial comparing two dif- with stereotactic radiosurgery for newly diagnosed brain ferent radiotherapy schedules in the treatment of cerebral metastases. Int J Radiat Oncol Biol Phys 2000, 47:993-999. metastases. Clin Oncol (R Coll Radiol) 1996, 8:308-315. 31. Agboola O, Benoit B, Cross P, Da Silva V, Esche B, Lesiuk H, Gon- 49. Tang SG, Tseng CK, Tsay PK, Chen CH, Chang JW, Pai PC, Hong JH: salves C: Prognostic factors derived from recursive partition- Predictors for patterns of brain relapse and overall survival ing analysis (RPA) of Radiation Therapy Oncology Group in patients with non-small cell lung cancer. J Neurooncol 2005, (RTOG) brain metastases trials applied to surgically 73:153-161. resected and irradiated brain metastatic cases. Int J Radiat 50. Fife KM, Colman MH, Stevens GN, Firth IC, Moon D, Shannon KF, Oncol Biol Phys 1998, 42:155-159. Harman R, Petersen-Schaefer K, Zacest AC, Besser M, Milton GW, 32. Paek SH, Audu PB, Sperling MR, Cho J, Andrews DW: Reevaluation McCarthy WH, Thompson JF: Determinants of outcome in of surgery for the treatment of brain metastases: review of melanoma patients with cerebral metastases. J Clin Oncol 208 patients with single or multiple brain metastases treated 2004, 22:1293-1300. at one institution with modern neurosurgical techniques. 51. Claude L, Perol D, Ray-Coquard I, Petit T, Blay JY, Carrie C, Bachelot Neurosurgery 2005, 56:1021-1034. T: Lymphopenia: A new independent prognostic factor for 33. Tendulkar RD, Liu SW, Barnett GH, Vogelbaum MA, Toms SA, Jin T, survival in patients treated with whole brain radiotherapy Suh JH: RPA classification has prognostic significance for sur- for brain metastases from breast carcinoma. Radiother Oncol gically resected single brain metastasis. Int J Radiat Oncol Biol 2005, 76:334-339. Phys 2006, 66:810-817. 52. Bartsch R, Fromm S, Rudas M, Wenzel C, Harbauer S, Roessler K, 34. Rades D, Pluemer A, Veninga T, Dunst J, Schild SE: A boost in addi- Kitz K, Steger GG, Weitmann HD, Poetter R, Zielinski CC, Dieck- tion to whole-brain radiotherapy improves patient outcome mann K: Intensified local treatment and systemic therapy sig- after resection of 1 or 2 brain metastases in recursive parti- nificantly increase survival in patients with brain metastases tioning analysis class 1 and 2 patients. Cancer 2007, from advanced breast cancer – A retrospective analysis. Radi- 110:1551-1559. other Oncol 2006, 80:313-317. 35. Nieder C, Geinitz H, Molls M: Validation of the graded prognos- 53. Nam BH, Kim SY, Han HS, Kwon Y, Lee KS, Kim TH, Ro J: Breast tic assessment index for surgically treated patients with cancer subtypes and survival in patients with brain metas- brain metastases. Anticancer Res 2008, 28:3015-3017. tases. Breast Cancer Res 2008, 10:R20. 36. Lagerwaard FJ, Levendag PC, Nowak PJ, Eijkenboom WM, Hanssens 54. Kirsch DG, Ledezma CJ, Mathews CS, Bhan AK, Ancukiewicz M, PE, Schmitz PI: Identification of prognostic factors in patients Hochberg FH, Loeffler JS: Survival after brain metastases from with brain metastases: a review of 1292 patients. Int J Radiat breast cancer in the trastuzumab era. J Clin Oncol 2005, Oncol Biol Phys 1999, 43:795-803. 23:2114-2116. 37. Weltman E, Salvajoli JV, Brandt RA, de Morais Hanriot R, Prisco FE, 55. Eichler AF, Kuter I, Ryan P, Schapira L, Younger J, Henson JW: Sur- Cruz JC, de Oliveira Borges SR, Wajsbrot DB: Radiosurgery for vival in patients with brain metastases from breast cancer. brain metastases: A score index for predicting prognosis. Int Cancer 2008, 112:2359-2367. J Radiat Oncol Biol Phys 2000, 46:1155-1161. 56. Melisko ME, Moore DH, Sneed PK, De Franco J, Rugo HS: Brain 38. Selek U, Chang EL, Hassenbusch SJ 3rd, Shiu AS, Lang FF, Allen P, metastases in breast cancer: clinical and pathologic charac- Weinberg J, Sawaya R, Maor MH: Stereotactic radiosurgical teristics associated with improvements in survival. J Neuroon- treatment in 103 patients for 153 cerebral melanoma metas- col 2008, 88:359-365. tases. Int J Radiat Oncol Biol Phys 2004, 59:1097-1106. 57. Harputluoglu H, Dizdar O, Aksoy S, Kilickap S, Dede DS, Ozisik Y, 39. Goyal S, Prasad D, Harrell F Jr, Matsumoto J, Rich T, Steiner L: Guler N, Barista I, Gullu I, Hayran M, Selek U, Cengiz M, Zorlu F, Gamma knife surgery for the treatment of intracranial Tekuzman G, Altundag K: Characteristics of breast cancer metastases from breast cancer. J Neurosurg 2005, 103:218-223. patients with central nervous system metastases: a single- 40. Gaudy-Marqueste C, Regis JM, Muracciole X, Laurans R, Richard MA, center experience. J Natl Med Assoc 2008, 100:521-526. Bonerandi JJ, Grob JJ: Gamma-Knife radiosurgery in the man- 58. Park BB, Uhm JE, Cho EY, et al.: Prognostic factor analysis in agement of melanoma patients with brain metastases: a patients with brain metastases from breast cancer: how can series of 106 patients without whole-brain radiotherapy. Int we improve the treatment outcomes? Cancer Chemother Phar- J Radiat Oncol Biol Phys 2006, 65:809-816. macol 2008. epub. 41. Akyurek S, Chang EL, Mahajan A, Hassenbusch SJ, Allen PK, Mathews 59. Church DN, Modgil R, Guglani S, Bahl A, Hopkins K, Braybrooke JP, LA, Shiu AS, Maor MH, Woo SY: Stereotactic radiosurgical Blair P, Price CG: Extended survival in women with brain treatment of cerebral metastases arising from breast can- metastases from HER2 overexpressing breast cancer. Am J cer. Am J Clin Oncol 2007, 30:310-314. Clin Oncol 2008, 31:250-254. Page 10 of 11 (page number not for citation purposes) Radiation Oncology 2009, 4:10 http://www.ro-journal.com/content/4/1/10 60. Ogawa K, Toita T, Sueyama H, Fuwa N, Kakinohana Y, Kamata M, brain radiotherapy prolongs time to neurologic progression Adachi G, Saito A, Yoshii Y, Murayama S: Brain metastases from in non-small-cell lung cancer patients with brain metastases: esophageal carcinoma: natural history, prognostic factors, results of a phase III trial. Int J Radiat Oncol Biol Phys 2009, and outcome. Cancer 2002, 94:759-764. 73:1069-1076. 61. Weinberg JS, Suki D, Hanbali F, Cohen ZR, Lenzi R, Sawaya R: Metas- tasis of esophageal carcinoma to the brain. Cancer 2003, 98:1925-1933. 62. Khuntia D, Sajja R, Chidel MA, Lee SY, Rice TW, Adelstein DJ, Carl- son TP, Saxton JP, Barnett GH, Suh JH: Factors associated with improved survival in patients with brain metastases from esophageal cancer: a retrospective review. Technol Cancer Res Treat 2003, 2:267-272. 63. Cohen ZR, Suki D, Weinberg JS, Marmor E, Lang FF, Gershenson DM, Sawaya R: Brain metastases in patients with ovarian carci- noma: prognostic factors and outcome. J Neurooncol 2004, 66:313-325. 64. Cormio G, Maneo A, Colamaria A, Loverro G, Lissoni A, Selvaggi L: Surgical resection of solitary brain metastasis from ovarian carcinoma: an analysis of 22 cases. Gynecol Oncol 2003, 89:116-119. 65. Growdon WB, Lopez-Varela E, Littell R, Oliva E, Seiden M, Krasner C, Lee H, Fuller A: Extent of extracranial disease is a powerful predictor of survival in patients with brain metastases from gynecological cancer. Int J Gynecol Cancer 2008, 18:262-268. 66. Tremont-Lukats IW, Bobustuc G, Lagos GK, Lolas K, Kyritsis AP, Puduvalli VK: Brain metastasis from prostate carcinoma: The M. D. Anderson Cancer Center experience. Cancer 2003, 98:363-368. 67. Bartelt S, Lutterbach J: Brain metastases in patients with cancer of unknown primary. J Neurooncol 2003, 64:249-253. 68. Rudà R, Borgognone M, Benech F, Vasario E, Soffietti R: Brain metastases from unknown primary tumour: a prospective study. J Neurol 2001, 248:394-398. 69. Kim SH, Weil RJ, Chao ST, Toms SA, Angelov L, Vogelbaum MA, Suh JH, Barnett GH: Stereotactic radiosurgical treatment of brain metastases in older patients. Cancer 2008, 113:834-840. 70. Noel G, Bollet MA, Noel S, Feuvret L, Boisserie G, Tep B, Delattre JY, Baillet F, Ambroise Valery C, Cornu P, Mazeron JJ: Linac stere- otactic radiosurgery: an effective and safe treatment for eld- erly patients with brain metastases. Int J Radiat Oncol Biol Phys 2005, 63:1555-1561. 71. Chatani M, Matayoshi Y, Masaki N, Inoue T: Radiation therapy for brain metastases from lung carcinoma. Prospective rand- omized trial according to the level of lactate dehydrogenase. Strahlenther Onkol 1994, 170:155-161. 72. Jacot W, Quantin X, Boher JM, Andre F, Moreau L, Gainet M, Depierre A, Quoix E, Chevalier TL, Pujol JL: Brain metastases at the time of presentation of non-small cell lung cancer: a multi-centric AERIO analysis of prognostic factors. Br J Cancer 2001, 84:903-909. 73. Gripp S, Moeller S, Bölke E, Schmitt G, Matuschek C, Asgari S, Asgharzadeh F, Roth S, Budach W, Franz M, Willers R: Survival pre- diction in terminally ill cancer patients by clinical estimates, laboratory tests, and self-rated anxiety and depression. J Clin Oncol 2007, 25:3313-3320. 74. Hauser CA, Stockler MR, Tattersall MH: Prognostic factors in patients with recently diagnosed incurable cancer: a system- atic review. Support Care Cancer 2006, 14:999-1011. 75. Rodrigus P, de Brouwer P, Raaymakers E: Brain metastases and non-small cell lung cancer. Prognostic factors and correla- tion with survival after irradiation. Lung Cancer 2001, Publish with Bio Med Central and every 32:129-136. scientist can read your work free of charge 76. Raizer JJ, Hwu WJ, Panageas KS, Wilton A, Baldwin DE, Bailey E, von Althann C, Lamb LA, Alvarado G, Bilsky MH, Gutin PH: Brain and "BioMed Central will be the most significant development for leptomeningeal metastases from cutaneous melanoma: sur- disseminating the results of biomedical researc h in our lifetime." vival outcomes based on clinical features. Neuro Oncol 2008, Sir Paul Nurse, Cancer Research UK 10:199-207. 77. Murray KJ, Scott C, Zachariah B, Michalski JM, Demas W, Vora NL, Your research papers will be: Whitton A, Movsas B: Importance of the mini-mental status available free of charge to the entire biomedical community examination in the treatment of patients with brain metas- tases: a report from the Radiation Therapy Oncology Group peer reviewed and published immediately upon acceptance protocol 91-04. Int J Radiat Oncol Biol Phys 2000, 48:59-64. cited in PubMed and archived on PubMed Central 78. Mehta MP, Shapiro WR, Phan SC, Gervais R, Carrie C, Chabot P, Patchell RA, Glantz MJ, Recht L, Langer C, Sur RK, Roa WH, Mahe yours — you keep the copyright MA, Fortin A, Nieder C, Meyers CA, Smith JA, Miller RA, Renschler BioMedcentral MF: Motexafin gadolinium combined with prompt whole Submit your manuscript here: http://www.biomedcentral.com/info/publishing_adv.asp Page 11 of 11 (page number not for citation purposes) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Radiation Oncology Springer Journals

Prognostic indices for brain metastases – usefulness and challenges

Radiation Oncology , Volume 4 (1) – Mar 4, 2009

Loading next page...
 
/lp/springer-journals/prognostic-indices-for-brain-metastases-usefulness-and-challenges-EtAVbJ6FZL
Publisher
Springer Journals
Copyright
Copyright © 2009 by Nieder and Mehta; licensee BioMed Central Ltd.
Subject
Medicine & Public Health; Oncology; Radiotherapy
eISSN
1748-717X
DOI
10.1186/1748-717X-4-10
pmid
19261187
Publisher site
See Article on Publisher Site

Abstract

Background: This review addresses the strengths and weaknesses of 6 different prognostic indices, published since the Radiation Therapy Oncology Group (RTOG) developed and validated the widely used 3-tiered prognostic index known as recursive partitioning analysis (RPA) classes, i.e. between 1997 and 2008. In addition, other analyses of prognostic factors in groups of patients, which typically are underrepresented in large trials or databases, published in the same time period are reviewed. Methods: Based on a systematic literature search, studies with more than 20 patients were included. The methods and results of prognostic factor analyses were extracted and compared. The authors discuss why current data suggest a need for a more refined index than RPA. Results: So far, none of the indices has been derived from analyses of all potential prognostic factors. The 3 most recently published indices, including the RTOG's graded prognostic assessment (GPA), all expanded from the primary 3-tiered RPA system to a 4-tiered system. The authors' own data confirm the results of the RTOG GPA analysis and support further evaluation of this tool. Conclusion: This review provides a basis for further refinement of the current prognostic indices by identifying open questions regarding, e.g., performance of the ideal index, evaluation of new candidate parameters, and separate analyses for different cancer types. Unusual primary tumors and their potential differences in biology or unique treatment approaches are not well represented in large pooled analyses. indices might also be used as inclusion/exclusion criteria Background Prognostic indices might represent a useful tool in pallia- for clinical trials and for comparison of results across dif- tive cancer treatment. Estimation of a patient's prognosis ferent studies in relatively homogeneous patient groups. in terms of overall survival might allow for tailored treat- ment, i.e. more aggressive approaches when these are Brain metastases continue to represent a formidable chal- likely to impact on survival and focus on disease stabilisa- lenge in oncology [1-3]. With increasing numbers of local tion, symptom control and toxicity minimization when and systemic treatment options, the issue of patient selec- the disease is more advanced, or comorbidity limits the tion gains importance. While surgery and stereotactic radi- tolerability of aggressive therapy. In addition, prognostic osurgery (SRS) provide long-term local control of Page 1 of 11 (page number not for citation purposes) Radiation Oncology 2009, 4:10 http://www.ro-journal.com/content/4/1/10 macroscopic disease and in combination with whole- tions. Their definition yielded class IIIa defined as age <65 brain radiotherapy (WBRT) the best available overall years, controlled primary tumor and single brain metasta- brain control for the remaining life time [4-10], they rep- sis, class IIIc defined as age ≥ 65 years, uncontrolled pri- resent overtreatment in patients with short survival, which mary tumor and multiple brain metastases, while other typically is caused by uncontrollable systemic disease. patients would make up class IIIb. The original RPA clas- This review will address the strengths and weaknesses of 6 sification has been validated by several authors, both in different prognostic indices, published since the Radia- selected and unselected patient groups, e.g., patients with tion Therapy Oncology Group (RTOG) developed and breast primary, lung primary (small cell and non-small validated the widely used 3-tiered prognostic index cell), malignant melanoma, unknown primary, or surgi- known as recursive partitioning analysis (RPA) classes cal resection and SRS as main local treatment modalities [11,12], i.e. between 1997 and 2008. In addition, other [14-35]. analyses of prognostic factors in groups of patients, which typically are underrepresented in large trials or databases, Probably, the surgically treated patients represent the published in the same time period are reviewed. These most homogeneous cohorts assessed with the RPA sys- include patients with primary tumors that do not com- tem, as these were patients with rather favourable progno- monly metastasize to the brain, and the elderly, who are sis, fit to undergo surgery and with limited brain disease. often either excluded or under-represented in clinical tri- Nevertheless, the differences in median survival between als. the individual studies were large. In RPA class I, median survival ranged from 15–29 months [31-35]. In class II, a survival range of 5.5–11 months has been reported. In Methods The present review compares different prognostic indices class III, these figures reached 1.4–9 months. As illustrated and analyses of prognostic factors based on a systematic here, survival within the same RPA class might vary by a literature search by use of Medline (Pub Med by the factor of 2 or more between different studies (identical National Library of Medicine, National Institutes of treatment approach). In series where the majority of Health, Bethesda, Maryland, USA). It is limited to adult patients were treated with WBRT, less variation between patients having received first-line treatment for parenchy- studies can be found (Figure 1). As shown in Table 1, both mal brain metastases in the absence of leptomeningeal RPA class II and III contain quite heterogeneous groups of disease. The key words used were "brain metastases", patients. The factor determining class III is KPS<70, which "metastatic brain tumor" and "cerebral metastases". The might result from many different causes including the final search was performed on June 30, 2008. It also brain metastases themselves, advanced and treatment- included the reference lists of all articles and the appropri- refractory extracranial metastases, severe pain or patho- ate chapters in textbooks on brain metastases, neuro- logical fracture in patients with bone metastases, atelecta- oncology and radiation oncology. Case reports and review sis or pneumonia from primary lung cancer, anemia articles were not assessed. Only studies with more than 20 induced by chemotherapy, recovery from recent surgery, patients were included. If several subsequent reports were and non-cancer-related comorbidity. In all the reports published from the same institution, the most recent pub- reviewed variable proportions of patients in the most lication was evaluated. The methods and results of prog- favourable RPA class I unexpectedly died within 2 nostic factor analyses were extracted and compared. months, while some patients in class III survived for more than 6 months. For these reasons, there obviously is a need for a more refined index than RPA. Results The search identified 6 different prognostic indices, which are shown in Table 1. Comparison of the patients' charac- The first attempt in 1999 resulted in the Rotterdam Score, teristics is shown in Table 2. Unfortunately, a considera- which did not gain wider acceptance [36]. Similar to RPA, ble amount of information can not be extracted from the performance status and extent of systemic disease were publications. The most widely used index over the last included, while the third parameter was response to ster- decade is the RPA index originally described by Gaspar et oids before WBRT. It can be assumed that the unavailabil- al. on behalf of the RTOG [11], which is based on 4 ity of this latter parameter in most databases or patient parameters (age, Karnofsky performance status (KPS), records prevented other groups from using the score. In presence or absence of extracranial metastases, and the addition, the definition of systemic tumor activity is not control status of the primary tumor), separating patients straight forward. The next attempt (Score Index for Radio- into 3 different classes. Lutterbach et al. suggested expan- surgery (SIR)) was derived from a limited number of sion of the classification by further dividing class III into patients treated with this particular focal approach, which 3 separate classes [13]. This was based on their multivari- might have resulted in overfitting of the data [37]. How- ate analysis of 916 patients from a single institution, but ever, several groups confirmed the performance of the SIR was not adopted by other authors in subsequent publica- in patients treated with SRS, surgery, and WBRT with or Page 2 of 11 (page number not for citation purposes) Table 1: Comparison of the prognostic scores published since 1997, empty fields indicate that a parameter is not used in the index Score Performanc Age Extracranial Controlled Steroid Number of Volume of Interval to Class I Class II Class III Class IV e status metastases primary treatment BM BM BM RPA KPS <65 years no vs yes no vs yes all 4 other KPS <70 none Derived ≥ 70 vs <70 favourable patients from 3 factors prospective RTOG studies, n = 1,200 Rotterdam ECOG limited systemic good, ECOG 0–1 other ECOG2-3 none Single 0–1 vs 2–3 activity vs extensive* moderate with no or patients with limited institution, or little limited or extensive n = 1,292 response systemic systemic tumor activity and activity and little good response to response to steroids steroids SIR KPS 80– ≤ 50: 2 no evidence disease: 1 1: 2 points largest 8–10 points 4–7 points 1–3 points none Single 100:2 points points of systemic point 2: 1 point lesion institution, KPS 60–70: 51–59: 1 disease or progressive ≥ 3: 0 points volume <5 n = 65 1 point point complete disease: 0 cc: 2 points KPS ≤ 50: 0 ≥ 60: 0 remission: 2 points 5–13 cc: 1 points points points point partial >13 cc: 0 remission or points stable BSBM KPS 80–100: no: 1 point yes: 1 point 3 points 2 points 1 point 0 points Single 1 point yes: 0 points no: 0 points institution, KPS ≤ 70: 0 n = 110 point GPA KPS 90–100: <50: 1 point none: 1 1: 1 point 3.5–4 points 3 points 1.5–2.5 0–1 points Derived 1 point 50–59: 0.5 point 2–3: 0.5 points from 5 KPS 70–80: points present: 0 points prospective 0.5 points >60: 0 points >3: 0 points RTOG KPS <70: 0 points studies, points n = 1,960 Rades et KPS ≥ 70: 5 ≤ 60: 4 none: 5 >8 mo: 4 17–18 14–16 11–13 9–10 points al. points points points points points points points Multi- KPS <70: 1 >60: 3 present: 2 ≤ 8 mo: institutional, point points points 3 points n = 1,085 BM: brain metastases, RPA: recursive partitioning analysis, RTOG: Radiation Therapy Oncology Group, KPS: Karnofsky performance score, SIR: score index for radiosurgery, BSBM: basic score for brain metastases, GPA: graded prognostic assessment, ECOG: Eastern Cooperative Oncology Group * limited systemic activity: no systemic metastases but progression of primary tumor or systemic metastases with primary tumor absent or controlled; extensive systemic activity: systemic metastases and progressive primary Radiation Oncology 2009, 4:10 http://www.ro-journal.com/content/4/1/10 Page 3 of 11 (page number not for citation purposes) Radiation Oncology 2009, 4:10 http://www.ro-journal.com/content/4/1/10 Table 2: Median values of reported patients' characteristics in each of the studies, empty fields indicate missing information Score Performance Age Extracranial Controlled Steroid Number of Volume of Interval to status metastases primary treatment BM BM BM RPA KPS 70 55–59 yrs. 38% 60% 2 n = 1,200 range Rotterdam ECOG 1 59 yrs. mean 15 mg 2 8.5 mo. n = 1,292 dexamethason e per day SIR KPS 80 61 yrs. 2 3.3 cc n = 65 BSBM 57 yrs. 2 9 cc n = 110 GPA KPS 80 60 yrs. 36% 67% 2 5–13 cc n = 1,960 Rades et al. KPS 70 60 yrs. 64% 8 mo. n = 1,085 BM: brain metastases, RPA: recursive partitioning analysis, KPS: Karnofsky performance score, SIR: score index for radiosurgery, BSBM: basic score for brain metastases, GPA: graded prognostic assessment, ECOG: Eastern Cooperative Oncology Group RTOG 1997 RTOG 2000 Nieder et al. 2000 Lutterbach et al. Months Saito et al. 2006 RTOG 2008 Nieder et al. 2008 Rades et al. 2008 RPA I RPA II RPA III Compar or without Figure 1 ison of median survival in 7 stud local measures, none of the studie ies uss ing the recursiv is limited to one pa e partit rtic ioning ular cancer ty analyses ( pe) RPA) classes (treatment was WBRT with Comparison of median survival in 7 studies using the recursive partitioning analyses (RPA) classes (treatment was WBRT with or without local measures, none of the studies is limited to one particular cancer type). without SRS, some of them with large numbers of patients patients managed with WBRT with or without SRS and (Figure 2) [35,38-44]. To accurately define systemic dis- surgery plus WBRT [35,42,44], however its performance is ease activity, comprehensive diagnostic work-up is not better than that of the other scores (Figure 3). needed. The RTOG has recently proposed a new index, which was When evaluating the SIR and RPA indices in their SRS compared to RPA, SIR, and BSBM (but not to the Rotter- database, the group from Brussels, Belgium, arrived at a dam score) [44]. The new score (Graded Prognostic new score, which they called Basic Score for Brain Metas- Assessment (GPA)) is different from RTOG's RPA, e.g., tases (BSBM) [43]. Based on its greater convenience and with regard to the number of prognostic classes, which simplicity, they advocated the use of this score, which uses increased from 3 to 4, and the larger number of patients. the same definition of extracranial disease activity as the The analysis also includes patients managed with WBRT RTOG. Recent data indicate that BSBM can be applied to plus SRS from RTOG study 9508 [5]. In the GPA system, Page 4 of 11 (page number not for citation purposes) Radiation Oncology 2009, 4:10 http://www.ro-journal.com/content/4/1/10 RTOG 2008 Months Nieder et al. SIR 8- SIR 4-7 SIR 1-3 C local Figure 2 omp mea arisso ures, studies n of median s not lim urvivali i te n 2 d to one particular studies using the s cancer type) core index for radiosurgery (SIR) (treatment was WBRT with or without Comparison of median survival in 2 studies using the score index for radiosurgery (SIR) (treatment was WBRT with or without local measures, studies not limited to one particular cancer type). RTOG 2008 Months Nieder et al. BSBM 3 BSBM 2 BSBM 1 BSBM 0 without Figure 3 Compar local measur ison of median survival in es, studies not 2 stud limited to ies using one particu the basic scor lar cancer typ e for brae) in metastases (BSBM) (treatment was WBRT with or Comparison of median survival in 2 studies using the basic score for brain metastases (BSBM) (treatment was WBRT with or without local measures, studies not limited to one particular cancer type). 3 different values (0, 0.5 or 1) are assigned for each of no longer mandated. It was concluded by the authors that these 4 parameters: age (≥ 60; 50–59; <50), KPS (<70; 70– "GPA is the least subjective, most quantitative and easiest 80; 90–100), number of brain metastases (>3; 2–3; 1), to use of the 4 indices" and that future trials should com- and extracranial metastases (present; not applicable; pare these scores and validate the GPA. One of the none). Assessment of primary tumor activity or control is authors' group has embarked on this comparison in 2 dif- Page 5 of 11 (page number not for citation purposes) Radiation Oncology 2009, 4:10 http://www.ro-journal.com/content/4/1/10 ferent patient populations, i.e. those managed with WBRT a patient, that can be used to estimate the chance of recov- with or without SRS (comparable to the RTOG study pop- ery from a disease or the chance of the disease recurring. Based on such prognostic factors, 6 different prognostic ulation) [42] and those managed with surgery and WBRT [35]. Both studies basically relied on the methods used by indices for adult patients with brain metastases from solid the RTOG in their analysis, though with patients treated in tumors have been developed over the last decade. As dem- clinical routine outside of randomized trials. Compared onstrated in Table 1, the 3 most recently published indices to RTOG's patients treated with WBRT with or without all expanded from the primary 3-tiered RPA system to a 4- SRS, the median age, KPS, number of lesions and lesion tiered system. The 6 indices are based on a different volume were similar. Obvious differences existed, how- number of prognostic factors, i.e. 3–6. Of course, increas- ever, regarding controlled primary tumor (47 vs. 67%) ing numbers of parameters will lead to less convenience and extracranial metastases (56 vs. 36%). Thus, the cohort and ease of administration. None of the groups that devel- is expected to have inferior survival. Figure 4 shows the oped these indices included all potential prognostic fac- survival results. tors in their analysis. This is most likely due to the unavailability of all the information in the databases and Last but not least, Rades et al. developed a new prognostic the difficulty in collecting missing data in 1,000 or more index based on 4 parameters (age, KPS, extracranial patients treated over many years. As can be seen in Figures metastases at the time of WBRT, interval from tumor diag- 3, 4, 5, the performance of the 4-tiered indices is not tre- nosis to WBRT) [45]. The major difference from the RPA mendously different, although further data are needed to classes is the replacement of primary tumor control by confirm this finding. interval from tumor diagnosis to WBRT (not by number of brain metastases as in the GPA). This index separated There is agreement in all indices on the importance of per- patients into 4 subgroups with significantly different formance status and extracranial disease activity. How- prognosis and was also validated in one of the authors' ever, whether both primary tumor and extracranial database (unpublished results, Figure 5). metastases should be considered is less clear (2 indices would not include primary tumor control). Assessment of Discussion extracranial disease status is not trivial. It might require As stated on the website of the National Cancer Institute considerable resources in patients with very limited life http://www.cancer.gov/templates/ expectancy and therapeutic options. When collecting data bd_alpha.aspx?CdrID=44246, a prognostic factor is over long time periods, one must expect a shift in diagnos- regarded as a situation or condition, or a characteristic of tic modalities, i.e. increasing use of magnetic resonance RTOG 2008 Months Nieder et al. GPA GPA 3 GPA GPA 0-1 3.5-4 1.5-2.5 Compar without Figure 4 local measur ison of median survival in es, studies not 2 st limited to udies using one particu the graded prognostic assess lar cancer type) ment (GPA) (treatment was WBRT with or Comparison of median survival in 2 studies using the graded prognostic assessment (GPA) (treatment was WBRT with or without local measures, studies not limited to one particular cancer type). Page 6 of 11 (page number not for citation purposes) Radiation Oncology 2009, 4:10 http://www.ro-journal.com/content/4/1/10 Rades et al. Months Nieder et al. 17-18 14-16 11-13 9-10 points points points points Compar Figure 5 ison of median survival in 2 studies using the index proposed by Rades et al Comparison of median survival in 2 studies using the index proposed by Rades et al. [45](treatment was WBRT with or without local measures, studies not limited to one particular cancer type, median survival estimated from the Kaplan-Meier curves in the publication). imaging of the brain as compared to computed tomogra- tumors in these models. Breast cancer poses an interesting phy (CT) or increasing use of chest CT or even positron dilemma here, because although tumor type and histology emission tomography (PET). Such a shift will likely result were not prognostically significant in the RPA, recent data, in no longer assigning patients to the most favourable especially since the advent of trastuzumab and lapatinib, prognostic class (stage migration). This might compro- suggest that, receptor status and her-2-neu expression mise the comparison of the different studies. might have prognostic impact, even if this issue is not with- out controversy (Table 3). The recently suggested prognos- Two of the 6 indices did not include age and the ones that tic factor lymphopenia falls into the same category [19,51]. did, used slightly different cut-off values. A minority of stud- Unusual primary tumors and their potential differences in ies (n = 2) included number of brain metastases and only biology or unique treatment approaches are not well repre- one each included response to steroids, volume of the largest sented in large pooled analyses. Table 4 provides examples lesion in the brain, and time interval to development of on analyses of prognostic factors in such groups. brain metastases, respectively. Other previous reports lend credence to the examination of each of these factors. In their Surrogate markers of disease activity that are easy to meas- multivariate analysis of 334 patients, DiLuna et al. reported ure and inexpensive, such as lactate dehydrogenase and significantly better survival in patients with 1–3 vs 4 or more other laboratory parameters have repeatedly been shown brain metastases and in those patients with both limited to be independent prognostic factors for survival [71-74]. number and volume of brain metastases (<5 cc total vol- Studies that were not limited to patients with brain metas- ume) [46]. Bhatnagar et al. also reported on the impact of tases suggest that the anorexia-cachexia syndrome, dysp- treatment volume as independent prognostic factor in nea, pain, and co-morbidity are further candidates for patients treated with SRS [47]. In a randomised trial with prospective evaluation [74]. The same holds true for neu- 544 patients, Priestman et al. found that dose of steroids was rofunction class [75,76] and mini mental status examina- independently associated with survival [48]. Interval to tion results, which was an independent prognostic factor development of brain metastases appears particularly impor- for survival in a multivariate model that also included KPS tant in patients with primary NSCLC and malignant [77]. The current prognostic indices unfortunately do not melanoma. The multivariate analyses of 3 studies with 292– incorporate these features. 686 patients support this observation [23,49,50]. One of the purposes of prognostic indices is to guide the The latter findings lead to the general question on the use- choice of treatment in individual patients. In this context, fulness of lumping together patients with different primary a prognostic index should be accurate enough to avoid Page 7 of 11 (page number not for citation purposes) Radiation Oncology 2009, 4:10 http://www.ro-journal.com/content/4/1/10 Table 3: Prognostic impact of hormone receptor and HER-2 status in patients with brain metastases from breast cancer n Prognostic impact of hormone receptor status Prognostic impact of HER-2 status Claude et al. 120 none not examined Bartsch et al. 174 none None Le Scodan et al. 117 receptor negative significantly worse None Nam et al. 126 receptor negative significantly worse HER-2 negative significantly worse Kirsch et al. 95 not examined HER-2 negative significantly worse* 55 ^ Eichler et al. 83 none HER-2 negative significantly worse Melisko et al. 112 receptor negative significantly worse none Harputluoglu et al. 144 none none Park et al. 125 none HER-2 positive significantly worse 86 not examined HER-2 negative significantly worse* Church et al. 80% of HER-2 overexpressing cases received trastuzumab after diagnosis of brain metastases * the difference in survival was limited to patients with HER-2 overexpressing cancer treated with trastuzumab after diagnosis of brain metastases overtreatment in patients that actually have very short sur- systems, the most favorable prognostic group is very small vival. Even more important, one should not withhold (e.g., GPA ≥ 3.5: 9% of RTOG and 7% of our own patients; treatment because the index erroneously predicts an unfa- RPA class I: 16% of RTOG and 11% of our own patients). vorable outcome. These aspects of the indices have not been thoroughly evaluated, even in the recent GPA analy- The open questions after publication of 6 prognostic indi- sis [44]. In our analysis of 239 patients, which confirms ces include: that RPA, SIR, BSBM and GPA each split the dataset into groups with significantly different prognosis, this issue - how should the ideal index perform? was addressed [42]. With regard to the outcome of patients with unfavorable survival, defined as ≤ 2 months - how many parameters should form the basis of the (n = 93), no significant difference between the indices was ideal index? observed. Regarding patients with favorable survival, defined as ≥ 6 months (n = 66), again no significant dif- - can we lump together patients with breast cancer, ference was observed, although RPA performed worse small-cell lung cancer, malignant melanoma etc. or do than the other indices. Overall, GPA misassigned 6% of we lose potentially important information? the patients (9 out of 159), compared to 11% with RPA. Therefore, the available validation data certainly do not - do we need candidate parameters beyond the ones discourage further evaluation of the new GPA. However, examined so far (lactate dehydrogenase, anemia, such evaluation should also include comparison with the weight loss, pain etc.)? 2 other scores (Rotterdam and Rades et al.). It is just the stark reality of the disease process that in all of the scoring Table 4: Prognostic factors in patients underrepresented in large studies (minimum number of patients n = 20) Author Population Significant prognostic factors Ogawa et al. esophageal cancer, n = 36 KPS, aggressive local treatment (multivariate) Weinberg et al. esophageal cancer, n = 27 no liver metastases, RPA class I (trend, p = 0.1, multivariate) Khuntia et al. esophageal cancer, n = 27 KPS, aggressive local treatment (multivariate) Cohen et al. ovarian cancer, n = 72 aggressive local treatment Cormio et al. ovarian cancer, n = 22 extracranial disease, time to development of brain metastases Growdon et al. gynaecological cancers, n = 30 extracranial disease, histology, use of chemotherapy (multivariate) Tremont-Lukats et al. prostate cancer, n = 103 adenocarcinoma vs other histology Rades et al. unknown primary, n = 101 KPS, extracranial metastases, RPA class Bartelt and Lutterbach unknown primary, n = 47 KPS, surgical resection status (multivariate) Ruda et al. unknown primary, n = 33 number of brain metastases (multivariate) Kim et al. patients ≥ 75 years, SRS treatment, n = 44 single brain metastasis, NSCLC vs other primary Noel et al. patients ≥ 65 years, SRS treatment, n = 117 KPS (multivariate) WBRT: whole-brain radiotherapy, KPS: Karnofsky performance status, RPA: recursive partitioning analysis, SRS: stereotactic radiosurgery, NSCLC: non-small cell lung cancer Page 8 of 11 (page number not for citation purposes) Radiation Oncology 2009, 4:10 http://www.ro-journal.com/content/4/1/10 6. Kondziolka D, Patel A, Lunsford LD, Kassam A, Flickinger JC: Stere- - is it justifiable to assign the same point value to dif- otactic radiosurgery plus whole brain radiotherapy versus ferent degrees of extracranial disease, e.g., 2 small radiotherapy alone for patients with multiple brain metas- asymptomatic lung metastases, 8 large liver metastases tases. Int J Radiat Oncol Biol Phys 1999, 45:427-434. 7. Patchell RA, Tibbs PA, Regine WF, Dempsey RJ, Mohiuddin M, Kry- with increased bilirubin, skin metastases already scio RJ, Markesbery WR, Foon KA, Young B: Postoperative radio- treated by radiotherapy etc.? therapy in the treatment of single metastases to the brain: a randomized trial. JAMA 1998, 80:1485-1489. 8. Noordijk EM, Vecht CJ, Haaxma-Reiche H, Padberg GW, Voormolen - can international groups collaborate to develop a JH, Hoekstra FH, Tans JT, Lambooij N, Metsaars JA, Wattendorf AR: consensus score, or maybe even an online tool? The choice of treatment of single brain metastasis should be based on extracranial tumor activity. Int J Radiat Oncol Biol Phys 1994, 29:711-717. Other aspects of predicting the outcome in patients with 9. Patchell RA, Tibbs PA, Walsh JW, Dempsey RJ, Maruyama Y, Kryscio brain metastases that many clinicians might appreciate, RJ, Markesbery WR, Macdonald JS, Young B: A randomized trial of surgery in the treatment of single metastases to the brain. N relate to the important issue of neurologic function and Engl J Med 1990, 322:494-500. quality of life. In many instances, radiotherapy aims more 10. Nieder C, Astner ST, Grosu AL, Andratschke NH, Molls M: The role of postoperative radiotherapy after resection of a single on improving deficits and preventing neurologic decline brain metastasis: combined analysis of 643 patients. Strahlen- than prolonging survival, but no attempts have been ther Onkol 2007, 183:576-580. made to develop scores that address endpoints other than 11. Gaspar L, Scott C, Rotman M, Asbell S, Phillips T, Wasserman T, McKenna WG, Byhardt R: Recursive partitioning analysis (RPA) overall survival. It appears therefore worthwhile to collect of prognostic factors in three Radiation Therapy Oncology data on such endpoints, as done, e.g., in the recently com- Group (RTOG) brain metastases trials. Int J Radiat Oncol Biol Phys 1997, 37:745-751. pleted randomized trial of radiotherapy with or without 12. Gaspar LE, Scott C, Murray K, Curran W: Validation of the RTOG motexafin gadolinium [78], which used time to neuro- recursive partitioning analysis (RPA) classification for brain logic progression as primary endpoint. Other opportuni- metastases. Int J Radiat Oncol Biol Phys 2000, 47:1001-1006. 13. Lutterbach J, Bartelt S, Stancu E, Guttenberger R: Patients with ties for future research include examination of prognostic brain metastases: hope for recursive partitioning analysis models that provide estimates on both risk of systemic (RPA) class 3. Radiother Oncol 2002, 63:339-345. cancer progression with death from non-neurologic 14. Nieder C, Nestle U, Motaref B, Walter K, Niewald M, Schnabel K: Prognostic factors in brain metastases: should patients be causes and risk of death from uncontrolled brain metas- selected for aggressive treatment according to recursive tases. partitioning analysis (RPA) classes? Int J Radiat Oncol Biol Phys 2000, 46:297-302. 15. Fleckenstein K, Hof H, Lohr F, Wenz F, Wannenmacher M: Prognos- Competing interests tic factors for brain metastases after whole brain radiother- The authors declare that they have no competing interests. apy. Data from a single institution. Strahlenther Onkol 2004, 180:268-273. 16. Saito EY, Viani GA, Ferrigno R, Nakamura RA, Novaes PE, Pellizzon Authors' contributions CA, Fogaroli RC, Conte MA, Salvajoli JV: Whole brain radiation CN and MM drafted the manuscript and participated in therapy in management of brain metastasis: results and prognostic factors. Radiat Oncol 2006, 1:20. the design of the study. Both authors read and approved 17. Mahmoud-Ahmed AS, Suh JH, Lee SY, Crownover RL, Barnett GH: the final manuscript. Results of whole brain radiotherapy in patients with brain metastases from breast cancer: a retrospective study. Int J Radiat Oncol Biol Phys 2002, 54:810-817. Acknowledgements 18. Viani GA, Castilho MS, Salvajoli JV, Pellizzon AC, Novaes PE, Guima- None rães FS, Conte MA, Fogaroli RC: Whole brain radiotherapy for brain metastases from breast cancer: estimation of survival using two stratification systems. BMC Cancer 2007, 7:53. References 19. Le Scodan R, Massard C, Mouret-Fourme E, Guinebretierre JM, 1. Kunthia D, Brown P, Li J, Mehta MP: Whole-brain radiotherapy in Cohen-Solal C, De Lalande B, Moisson P, Breton-Callu C, Gardner M, the management of brain metastasis. J Clin Oncol 2006, Goupil A, Renody N, Floiras JL, Labib A: Brain metastases from 24:1295-1304. breast carcinoma: validation of the Radiation Therapy 2. Langer CJ, Mehta MP: Current management of brain metas- Oncology Group recursive partitioning analysis classification tases, with a focus on systemic options. J Clin Oncol 2005, and proposition of a new prognostic score. Int J Radiat Oncol Biol 23:6207-6219. Phys 2007, 69:839-845. 3. Nieder C, Grosu AL, Astner ST, Thamm R, Molls M: Integration of 20. Kepka L, Cieslak E, Bujko K, Fijuth H, Wierzchowski M: Results of chemotherapy into current treatment strategies for brain the whole-brain radiotherapy for patients with brain metas- metastases from solid tumors. Radiat Oncol 2006, 1:19. tases from lung cancer: the RTOG RPA intra-classes analy- 4. Aoyama H, Shirato H, Tago M, Nakagawa K, Toyoda T, Hatano K, sis. Acta Oncol 2005, 44:389-398. Kenjyo M, Oya N, Hirota S, Shioura H, Kunieda E, Inomata T, Hay- 21. Videtic GM, Adelstein DJ, Mekhail TM, Rice TW, Stevens GH, Lee SY, akawa K, Katoh N, Kobashi G: Stereotactic radiosurgery plus Suh JH: Validation of the RTOG recursive partitioning analy- whole-brain radiation therapy vs stereotactic radiosurgery sis (RPA) classification for small-cell lung cancer-only brain alone for treatment of brain metastases. A randomized con- metastases. Int J Radiat Oncol Biol Phys 2007, 67:240-243. trolled trial. JAMA 2006, 295:2483-2491. 22. Gülbas H, Erkal HS, Serin M: The use of recursive partitioning 5. Andrews DW, Scott CB, Sperduto PW, Flanders AE, Gaspar LE, analysis grouping in patients with brain metastases from Schell MC, Werner-Wasik M, Demas W, Ryu J, Bahary JP, Souhami L, non-small-cell lung cancer. Jpn J Clin Oncol 2006, 36:193-196. Rotman M, Mehta MP, Curran WJ Jr: Whole brain radiation ther- 23. Rades D, Schild SE, Lohynska R, Veninga T, Stalpers LJ, Dunst J: Two apy with or without stereotactic radiosurgery boost for radiation regimens and prognostic factors for brain metas- patients with one to three brain metastases: phase III results tases in nonsmall cell lung cancer patients. Cancer 2007, of the RTOG 9508 randomised trial. Lancet 2004, 110:1077-1082. 363:1665-1672. Page 9 of 11 (page number not for citation purposes) Radiation Oncology 2009, 4:10 http://www.ro-journal.com/content/4/1/10 24. Rades D, Bohlen G, Lohynska R, Veninga T, Stalpers LJ, Schild SE, 42. Nieder C, Molls M: Validation of the graded prognostic assess- Dunst J: Whole-brain radiotherapy with 20 Gy in 5 fractions ment index for patients with brain metastases: in regards to for brain metastases in patients with cancer of unknown pri- Sperduto et al. (Int J Radiat Oncol Biol Phys 2008;70:510– mary (CUP). Strahlenther Onkol 2007, 183:631-636. 514). Int J Radiat Oncol Biol Phys 2008, 72:1619. 25. Kanner AA, Suh JH, Siomin VE, Lee SY, Barnett GH, Vogelbaum MA: 43. Lorenzoni J, Devriendt D, Massager N, David P, Ruiz S, Vanderlinden Posterior fossa metastases: aggressive treatment improves B, Van Houtte P, Brotchi J, Levivier M: Radiosurgery for treat- survival. Stereotact Funct Neurosurg 2003, 81:18-23. ment of brain metastases: Estimation of patient eligibility 26. Nam TK, Lee JI, Jung YJ, Im YS, An HY, Nam DH, Park K, Kim JH: using three stratification systems. Int J Radiat Oncol Biol Phys Gamma knife surgery for brain metastases in patients har- 2004, 60:218-224. boring four or more lesions: survival and prognostic factors. 44. Sperduto PW, Berkey B, Gaspar LE, Mehta M, Curran W: A new J Neurosurg 2005, 102(suppl):147-150. prognostic index and comparison to three other indices for 27. Buchsbaum JC, Suh JH, Lee SY, Chidel MA, Greskovich JF, Barnett patients with brain metastases: an analysis of 1,960 patients GH: Survival by Radiation Therapy Oncology Group recur- in the RTOG database. Int J Radiat Oncol Biol Phys 2008, sive partitioning analysis class and treatment modality in 70:510-514. patients with brain metastases from malignant melanoma. 45. Rades D, Dunst J, Schild SE: A new scoring system to predicting Cancer 2002, 94:2265-2272. the survival of patients treated with whole-brain radiother- 28. Harrison BE, Johnson JL, Clough RW, Halperin EC: Selection of apy for brain metastases. Strahlenther Onkol 2008, 184:251-255. patients with melanoma brain metastases for aggressive 46. DiLuna ML, King JT Jr, Knisely JP, Chiang VL: Prognostic factors for treatment. Am J Clin Oncol 2003, 26:354-357. survival after stereotactic radiosurgery vary with the 29. Morris SL, Low SH, A'Hern RP, Eisen TG, Gore ME, Nutting CM, Har- number of cerebral metastases. Cancer 2007, 109:135-145. rington KJ: A prognostic index that predicts outcome follow- 47. Bhatnagar AK, Kondziolka D, Lunsford LD, Flickinger JC: Recursive ing palliative whole brain radiotherapy for patients with partitioning analysis of prognostic factors for patients with metastatic malignant melanoma. Br J Cancer 2004, 91:829-833. four or more intracranial metastases treated with radiosur- 30. Chidel MA, Suh JH, Reddy CA, Chao ST, Lundbeck MF, Barnett GH: gery. Technol Cancer Res Treat 2007, 6:153-160. Application of recursive partitioning analysis and evaluation 48. Priestman TJ, Dunn J, Brada M, Rampling R, Baker PG: Final results of the use of whole brain radiation among patients treated of the Royal College of Radiologists' trial comparing two dif- with stereotactic radiosurgery for newly diagnosed brain ferent radiotherapy schedules in the treatment of cerebral metastases. Int J Radiat Oncol Biol Phys 2000, 47:993-999. metastases. Clin Oncol (R Coll Radiol) 1996, 8:308-315. 31. Agboola O, Benoit B, Cross P, Da Silva V, Esche B, Lesiuk H, Gon- 49. Tang SG, Tseng CK, Tsay PK, Chen CH, Chang JW, Pai PC, Hong JH: salves C: Prognostic factors derived from recursive partition- Predictors for patterns of brain relapse and overall survival ing analysis (RPA) of Radiation Therapy Oncology Group in patients with non-small cell lung cancer. J Neurooncol 2005, (RTOG) brain metastases trials applied to surgically 73:153-161. resected and irradiated brain metastatic cases. Int J Radiat 50. Fife KM, Colman MH, Stevens GN, Firth IC, Moon D, Shannon KF, Oncol Biol Phys 1998, 42:155-159. Harman R, Petersen-Schaefer K, Zacest AC, Besser M, Milton GW, 32. Paek SH, Audu PB, Sperling MR, Cho J, Andrews DW: Reevaluation McCarthy WH, Thompson JF: Determinants of outcome in of surgery for the treatment of brain metastases: review of melanoma patients with cerebral metastases. J Clin Oncol 208 patients with single or multiple brain metastases treated 2004, 22:1293-1300. at one institution with modern neurosurgical techniques. 51. Claude L, Perol D, Ray-Coquard I, Petit T, Blay JY, Carrie C, Bachelot Neurosurgery 2005, 56:1021-1034. T: Lymphopenia: A new independent prognostic factor for 33. Tendulkar RD, Liu SW, Barnett GH, Vogelbaum MA, Toms SA, Jin T, survival in patients treated with whole brain radiotherapy Suh JH: RPA classification has prognostic significance for sur- for brain metastases from breast carcinoma. Radiother Oncol gically resected single brain metastasis. Int J Radiat Oncol Biol 2005, 76:334-339. Phys 2006, 66:810-817. 52. Bartsch R, Fromm S, Rudas M, Wenzel C, Harbauer S, Roessler K, 34. Rades D, Pluemer A, Veninga T, Dunst J, Schild SE: A boost in addi- Kitz K, Steger GG, Weitmann HD, Poetter R, Zielinski CC, Dieck- tion to whole-brain radiotherapy improves patient outcome mann K: Intensified local treatment and systemic therapy sig- after resection of 1 or 2 brain metastases in recursive parti- nificantly increase survival in patients with brain metastases tioning analysis class 1 and 2 patients. Cancer 2007, from advanced breast cancer – A retrospective analysis. Radi- 110:1551-1559. other Oncol 2006, 80:313-317. 35. Nieder C, Geinitz H, Molls M: Validation of the graded prognos- 53. Nam BH, Kim SY, Han HS, Kwon Y, Lee KS, Kim TH, Ro J: Breast tic assessment index for surgically treated patients with cancer subtypes and survival in patients with brain metas- brain metastases. Anticancer Res 2008, 28:3015-3017. tases. Breast Cancer Res 2008, 10:R20. 36. Lagerwaard FJ, Levendag PC, Nowak PJ, Eijkenboom WM, Hanssens 54. Kirsch DG, Ledezma CJ, Mathews CS, Bhan AK, Ancukiewicz M, PE, Schmitz PI: Identification of prognostic factors in patients Hochberg FH, Loeffler JS: Survival after brain metastases from with brain metastases: a review of 1292 patients. Int J Radiat breast cancer in the trastuzumab era. J Clin Oncol 2005, Oncol Biol Phys 1999, 43:795-803. 23:2114-2116. 37. Weltman E, Salvajoli JV, Brandt RA, de Morais Hanriot R, Prisco FE, 55. Eichler AF, Kuter I, Ryan P, Schapira L, Younger J, Henson JW: Sur- Cruz JC, de Oliveira Borges SR, Wajsbrot DB: Radiosurgery for vival in patients with brain metastases from breast cancer. brain metastases: A score index for predicting prognosis. Int Cancer 2008, 112:2359-2367. J Radiat Oncol Biol Phys 2000, 46:1155-1161. 56. Melisko ME, Moore DH, Sneed PK, De Franco J, Rugo HS: Brain 38. Selek U, Chang EL, Hassenbusch SJ 3rd, Shiu AS, Lang FF, Allen P, metastases in breast cancer: clinical and pathologic charac- Weinberg J, Sawaya R, Maor MH: Stereotactic radiosurgical teristics associated with improvements in survival. J Neuroon- treatment in 103 patients for 153 cerebral melanoma metas- col 2008, 88:359-365. tases. Int J Radiat Oncol Biol Phys 2004, 59:1097-1106. 57. Harputluoglu H, Dizdar O, Aksoy S, Kilickap S, Dede DS, Ozisik Y, 39. Goyal S, Prasad D, Harrell F Jr, Matsumoto J, Rich T, Steiner L: Guler N, Barista I, Gullu I, Hayran M, Selek U, Cengiz M, Zorlu F, Gamma knife surgery for the treatment of intracranial Tekuzman G, Altundag K: Characteristics of breast cancer metastases from breast cancer. J Neurosurg 2005, 103:218-223. patients with central nervous system metastases: a single- 40. Gaudy-Marqueste C, Regis JM, Muracciole X, Laurans R, Richard MA, center experience. J Natl Med Assoc 2008, 100:521-526. Bonerandi JJ, Grob JJ: Gamma-Knife radiosurgery in the man- 58. Park BB, Uhm JE, Cho EY, et al.: Prognostic factor analysis in agement of melanoma patients with brain metastases: a patients with brain metastases from breast cancer: how can series of 106 patients without whole-brain radiotherapy. Int we improve the treatment outcomes? Cancer Chemother Phar- J Radiat Oncol Biol Phys 2006, 65:809-816. macol 2008. epub. 41. Akyurek S, Chang EL, Mahajan A, Hassenbusch SJ, Allen PK, Mathews 59. Church DN, Modgil R, Guglani S, Bahl A, Hopkins K, Braybrooke JP, LA, Shiu AS, Maor MH, Woo SY: Stereotactic radiosurgical Blair P, Price CG: Extended survival in women with brain treatment of cerebral metastases arising from breast can- metastases from HER2 overexpressing breast cancer. Am J cer. Am J Clin Oncol 2007, 30:310-314. Clin Oncol 2008, 31:250-254. Page 10 of 11 (page number not for citation purposes) Radiation Oncology 2009, 4:10 http://www.ro-journal.com/content/4/1/10 60. Ogawa K, Toita T, Sueyama H, Fuwa N, Kakinohana Y, Kamata M, brain radiotherapy prolongs time to neurologic progression Adachi G, Saito A, Yoshii Y, Murayama S: Brain metastases from in non-small-cell lung cancer patients with brain metastases: esophageal carcinoma: natural history, prognostic factors, results of a phase III trial. Int J Radiat Oncol Biol Phys 2009, and outcome. Cancer 2002, 94:759-764. 73:1069-1076. 61. Weinberg JS, Suki D, Hanbali F, Cohen ZR, Lenzi R, Sawaya R: Metas- tasis of esophageal carcinoma to the brain. Cancer 2003, 98:1925-1933. 62. Khuntia D, Sajja R, Chidel MA, Lee SY, Rice TW, Adelstein DJ, Carl- son TP, Saxton JP, Barnett GH, Suh JH: Factors associated with improved survival in patients with brain metastases from esophageal cancer: a retrospective review. Technol Cancer Res Treat 2003, 2:267-272. 63. Cohen ZR, Suki D, Weinberg JS, Marmor E, Lang FF, Gershenson DM, Sawaya R: Brain metastases in patients with ovarian carci- noma: prognostic factors and outcome. J Neurooncol 2004, 66:313-325. 64. Cormio G, Maneo A, Colamaria A, Loverro G, Lissoni A, Selvaggi L: Surgical resection of solitary brain metastasis from ovarian carcinoma: an analysis of 22 cases. Gynecol Oncol 2003, 89:116-119. 65. Growdon WB, Lopez-Varela E, Littell R, Oliva E, Seiden M, Krasner C, Lee H, Fuller A: Extent of extracranial disease is a powerful predictor of survival in patients with brain metastases from gynecological cancer. Int J Gynecol Cancer 2008, 18:262-268. 66. Tremont-Lukats IW, Bobustuc G, Lagos GK, Lolas K, Kyritsis AP, Puduvalli VK: Brain metastasis from prostate carcinoma: The M. D. Anderson Cancer Center experience. Cancer 2003, 98:363-368. 67. Bartelt S, Lutterbach J: Brain metastases in patients with cancer of unknown primary. J Neurooncol 2003, 64:249-253. 68. Rudà R, Borgognone M, Benech F, Vasario E, Soffietti R: Brain metastases from unknown primary tumour: a prospective study. J Neurol 2001, 248:394-398. 69. Kim SH, Weil RJ, Chao ST, Toms SA, Angelov L, Vogelbaum MA, Suh JH, Barnett GH: Stereotactic radiosurgical treatment of brain metastases in older patients. Cancer 2008, 113:834-840. 70. Noel G, Bollet MA, Noel S, Feuvret L, Boisserie G, Tep B, Delattre JY, Baillet F, Ambroise Valery C, Cornu P, Mazeron JJ: Linac stere- otactic radiosurgery: an effective and safe treatment for eld- erly patients with brain metastases. Int J Radiat Oncol Biol Phys 2005, 63:1555-1561. 71. Chatani M, Matayoshi Y, Masaki N, Inoue T: Radiation therapy for brain metastases from lung carcinoma. Prospective rand- omized trial according to the level of lactate dehydrogenase. Strahlenther Onkol 1994, 170:155-161. 72. Jacot W, Quantin X, Boher JM, Andre F, Moreau L, Gainet M, Depierre A, Quoix E, Chevalier TL, Pujol JL: Brain metastases at the time of presentation of non-small cell lung cancer: a multi-centric AERIO analysis of prognostic factors. Br J Cancer 2001, 84:903-909. 73. Gripp S, Moeller S, Bölke E, Schmitt G, Matuschek C, Asgari S, Asgharzadeh F, Roth S, Budach W, Franz M, Willers R: Survival pre- diction in terminally ill cancer patients by clinical estimates, laboratory tests, and self-rated anxiety and depression. J Clin Oncol 2007, 25:3313-3320. 74. Hauser CA, Stockler MR, Tattersall MH: Prognostic factors in patients with recently diagnosed incurable cancer: a system- atic review. Support Care Cancer 2006, 14:999-1011. 75. Rodrigus P, de Brouwer P, Raaymakers E: Brain metastases and non-small cell lung cancer. Prognostic factors and correla- tion with survival after irradiation. Lung Cancer 2001, Publish with Bio Med Central and every 32:129-136. scientist can read your work free of charge 76. Raizer JJ, Hwu WJ, Panageas KS, Wilton A, Baldwin DE, Bailey E, von Althann C, Lamb LA, Alvarado G, Bilsky MH, Gutin PH: Brain and "BioMed Central will be the most significant development for leptomeningeal metastases from cutaneous melanoma: sur- disseminating the results of biomedical researc h in our lifetime." vival outcomes based on clinical features. Neuro Oncol 2008, Sir Paul Nurse, Cancer Research UK 10:199-207. 77. Murray KJ, Scott C, Zachariah B, Michalski JM, Demas W, Vora NL, Your research papers will be: Whitton A, Movsas B: Importance of the mini-mental status available free of charge to the entire biomedical community examination in the treatment of patients with brain metas- tases: a report from the Radiation Therapy Oncology Group peer reviewed and published immediately upon acceptance protocol 91-04. Int J Radiat Oncol Biol Phys 2000, 48:59-64. cited in PubMed and archived on PubMed Central 78. Mehta MP, Shapiro WR, Phan SC, Gervais R, Carrie C, Chabot P, Patchell RA, Glantz MJ, Recht L, Langer C, Sur RK, Roa WH, Mahe yours — you keep the copyright MA, Fortin A, Nieder C, Meyers CA, Smith JA, Miller RA, Renschler BioMedcentral MF: Motexafin gadolinium combined with prompt whole Submit your manuscript here: http://www.biomedcentral.com/info/publishing_adv.asp Page 11 of 11 (page number not for citation purposes)

Journal

Radiation OncologySpringer Journals

Published: Mar 4, 2009

References