2099
Prognostic Factors for Survival in Patients with Advanced Renal Cell Carcinoma Undergoing Nonmyeloablative Allogeneic Stem Cell Transplantation Jacopo Peccatori, M.D.1 Lisbeth Barkholt, M.D.2 Tanner Demirer, M.D.3 Maria Pia Sormani, M.D.4 Paolo Bruzzi, M.D.4 Fabio Ciceri, M.D.1 Alberto Zambelli, M.D.5 Gian Antonio Da Prada, M.D.5 Paolo Pedrazzoli, M.D.6 Salvatore Siena, M.D.6 Gero Massenkeil, M.D.7 Rodrigo Martino, M.D.8 Stig Lenhoff, M.D.9 Paolo Corradini, M.D.10 Giovanni Rosti, M.D.11 Olle Ringden, M.D.2 Marco Bregni, M.D.1 Dietger Niederwieser, M.D.12 on behalf of The European Bone Marrow Transplantation Solid Tumor Working Party
BACKGROUND. The objective of this study was to identify prognostic factors for predicting survival in patients with advanced renal cell carcinoma (RCC) who had undergone an allogeneic stem cell transplantation after failure on immunotherapy. METHODS. The authors studied 70 patients with advanced RCC who underwent allogeneic transplantation with a fludarabine-based, reduced-intensity regimen. Ten parameters were analyzed at the time of transplantation for their power to predict survival. Clinical features were examined first univariately; then, variables that were correlated significantly with survival in the univariate analysis were included in a multivariate Cox regression model. RESULTS. Factors that were found to be associated significantly with limited survival were performance status, the number of metastatic sites, the presence of mediastinal metastasis, hemoglobin level, C-reactive protein (CRP) level, lactate dehydrogenase (LDH) level, and neutrophil counts. All these variables were included in a multivariate Cox regression model, and three were retained in the final model. Patients were classified according to the score estimated by the final Cox model in two groups (above or below the median value): The median survival was 3.5 months for patients who had a poor prognosis patients versus 23 months for patients who had a good prognosis. CONCLUSIONS. The current findings suggested that three easily available parameters (performance status, CRP level, and LDH level) could be used to stratify patients with advanced RCC who are candidates for allografting and to assist clinicians in decision-making and selection of an appropriate treatment program. Cancer 2005;104:2099 –103. © 2005 American Cancer Society.
1
Hematology-Bone Marrow Transplant Unit, Istituto Scientifico H. San Raffaele, Milano, Italy.
KEYWORDS: C-reactive protein, renal cell carcinoma, allogeneic stem cell transplantation, prognostic factors.
2
Department of Hematology, Huddinge University Hospital, Huddinge, Sweden.
3
Stem Cell Transplantation Unit, Ankara University Medical School, Ankara, Turkey.
4
Department of Epidemiology, National Institute for Cancer Research, Genova, Italy.
5
Department of Medical Oncology, Fondazione S. Maugeri, Pavia, Italy.
10 Department of Hematology, Istituto Nazionale Tumori, Milano, Italy.
The last two authors contributed equally to this article.
6
Department of Medical Oncology, Ospedale Niguarda-Ca’ Granda, Milano, Italy.
11
Department of Medical Oncology, Ospedale Civile, Ravenna, Italy.
7
Department of Hematology-Oncology, Charite`, Berlin, Germany. 8
12
Department of Hematology-Oncology, University of Leipzig, Leipzig, Germany.
Department of Hematology, Hospital Santa Creu i San Paul, Barcelona, Spain.
9
Department of Hematology, Lund University Hospital, Lund, Sweden.
Supported in part by a grant from the Associazione Italiana per la Ricerca sul Cancro to M.B.
© 2005 American Cancer Society DOI 10.1002/cncr.21477 Published online 11 October 2005 in Wiley InterScience (www.interscience.wiley.com).
Address for reprints: Marco Bregni, M.D., Departments of Oncology and Hematology and Bone Marrow Transplantation Unit, Istituto Scientifico H. San Raffaele, Milano, Italy; Fax: (011) 39 0226434760; E-mail:
[email protected] Received September 1, 2004; revision received July 21, 2005; accepted July 21, 2005.
2100
CANCER November 15, 2005 / Volume 104 / Number 10
P
atients with metastatic renal cell carcinoma (RCC) have a poor prognosis, with a median survival of 10 months.1 Different treatment strategies, including hormone therapy, chemotherapy, and immunotherapy, have little impact on global survival. Cytokine combinations (interferon-␣ with or without interleukin-2) are used widely as first-line therapy, but the response rate is low (10 –20%), and very few patients achieve a long-term response.2 Interest has focused recently on allogeneic nonmyeloablative stem cell transplantation, and an immune-mediated graft-versus-tumor (GVT) effect has been reported.3–5 However, the onset of a GVT effect typically is delayed by months from the time of the transplantation and usually occurs after cyclosporine withdrawal. Indeed, most patients develop tumor progression soon after they undergo transplantation, and it is questionable whether patients with rapidly progressive tumors should be included in allogeneic stem cell transplantation programs. Several reports have outlined different prognostic factors for survival in patients with RCC who are treated with cytokines as first-line therapy.6,7 Recently, the Memorial Sloan-Kettering Cancer Center developed a scoring system to predict survival in previously treated patients with RCC.1 In the last few years, approximately 200 patients with RCC who underwent allogeneic stem cell transplantation in Europe have been reported to The European Bone Marrow Transplantation (EBMT) Solid Tumor Working Party. We selected 9 EBMT centers and collected data on 70 patients, and we analyzed 10 variables that had been identified in previous studies as prognostic factors for the survival of patients with advanced RCC undergoing allogeneic transplantation after failure on immunotherapy. In the current article, we report the results from a multivariate analysis of this patient population.
MATERIALS AND METHODS Between June 1999 and 2003, 70 consecutive patients with metastatic RCC underwent allogeneic transplantation in 9 EBMT centers. All patients received a fludarabine-based, reduced-intensity preparative regimen. Graft-versus-host disease prophylaxis consisted mainly of cyclosporine and short-course methotrexate. The clinical characteristics of these patients will be described in detail elsewhere (unpublished results). The median interval from the diagnosis of metastatic disease to transplantation was 12 months (range, 0 –186 mos). Patients with treated and quiescent brain metastases were not excluded from transplantation procedure, provided they had a suitable human leukocyte antigen full-matched sibling. At a median fol-
low-up of 10 months, 48 patients died, and 22 remained alive. We analyzed the power of 10 parameters: performance status (PS), disease-free interval (DFI), the number of metastatic sites, the presence of liver metastasis, the presence of mediastinal metastasis, hemoglobin level, C-reactive protein (CRP) level, lactate dehydrogenase (LDH) level, calcemia, and neutrophil count. We decided to focus only on pretransplantation prognostic factors to provide clinicians with clear indications for treatment strategies, such as allogeneic stem cell transplantation. The effect of all prognostic factors on survival after allogeneic stem cell transplantation was evaluated in univariate analyses using a Cox regression model; variables were entered as binary (yes or no), categorical, or continuous, according to their distribution. P values for trend were estimated for categorical variables. The DFI between tumor diagnosis and the appearance of metastasis was log transformed to perform the test for trend to account for its asymmetric distribution. Biochemical parameters with different normality ranges (as measured in different centers), including CRP, LDH, and calcium levels, were recoded as normal or not normal, according to the criteria of each center. In tests of significance, no correction for multiple comparisons was done because of the exploratory nature of these analyses. All the variables that were found to be correlated significantly with survival in the univariate analysis (P ⬍ 0.05) were included in a multivariate Cox regression model with a backward selection procedure (P for exclusion ⫽ 0.10). A prognostic score was calculated using the variables that were included in the final model, and Kaplan–Meier survival curves were estimated for low-risk and high-risk patients, as classified according to their prognostic score (values greater than or less than the median score). Transplantation-related mortality and disease-related mortality were estimated using the methodology for competing risks.8 The main patient characteristics are reported in Table 1.
RESULTS The primary endpoint of the current study was survival, which was defined as the time from the day of graft infusion to the date of death or of last follow-up. Factors that were considered in the univariate analysis included the number and sites of metastasis, Karnofsky PS, DFI between tumor diagnosis and metastasis appearance, and baseline biochemical parameters, as reported in Table 2. Factors that were associated significantly with limited survival were Karnofsky PS (P ⫽ 0.005; test for trend), the number of metastatic sites (P ⫽ 0.001; test for trend), the presence of mediastinal metastasis (P ⫽ 0.008), hemoglobin level (P ⫽ 0.002),
Survival in Allografted Renal Ca/Peccatori et al. TABLE 1 Main Patient Characteristics Variable
TABLE 2 Cox Regression Analysis No. of patients
% Variable
Karnofsky performance status 60% 70% 80% 90% 100% Unknown No. of metastatic sites 0–1 2 3 ⱖ4 Liver metastasis No Yes Mediastinal metastasis No Yes Unknown C-reactive protein (g/L) Normal High Unknown Lactate dehydrogenase (IU/L) Normal High Calcium (mm/L) Normal High Neutrophil count (⫻ 109/L): mean ⫾ SD Disease-free interval (mos) Median Range Hemoglobin (g/dL): mean ⫾ SD
2101
6 15 10 18 20 1
8.7 21.7 14.5 26.1 29.0
16 23 20 11
22.9 32.9 28.6 15.7
54 15
78.3 21.7
39 30 1
56.5 43.5
39 20 11
55.7 28.9 15.7
54 16
77.1 22.9
62 8
88.6 11.4
12.0 ⫾ 1.9 7 0–258 11.9 ⫾ 1.9
SD: standard deviation.
CRP level (P ⬍ 0.001), LDH level (P ⫽ 0.008), and neutrophil count (P ⫽ 0.036). All of these variables were included in a multivariate Cox regression model, and 3 of these were retained in the final model (Table 2): Karnofsky PS (hazard ratio [HR] ⫽ 0.74 from 1 category to the next; P ⫽ 0.03), CRP level (HR ⫽ 3.40; P ⬍ 0.001), and LDH level (HR ⫽ 2.41; P ⫽ 0.02). The ability of this model to discriminate patients with a poor prognosis is illustrated in Figure 1. Patients were classified according to the score estimated by the final Cox model into two groups (above or below the median value): High-risk patients had a median survival of 3.5 months, whereas low-risk patients had a median survival of 23 months. High-risk and low-risk patients had statistically significant differences in both transplantation-related mortality and disease-related mortality, as shown in Table 3.
Karnofsky performance status 60% 70% 80% 90% 100% No. of metastatic sites 0–1 2 3 ⱖ4 Liver metastasis No Yes Mediastinal metastasis No Yes Hemoglobin (g/dL) C-reactive protein (g/L) Normal High Lactate dehydrogenase (IU/L) Normal High Neutrophil count (⫻ 109/L) Disease-free interval (mos) 0 mos 1–19 mos ⱖ 20 mos Calcium (mm/L) Normal High
HR
95% CI
9.29 2.27 1.33 2.40 1.00b
3.18–27.14 0.94–5.49 0.47–3.79 1.04–5.57
P value
Multivariate P value
0.005a
0.03a
1.00b 1.91 2.07 5.47
0.77–4.71 0.84–5.07 2.03–14.77
0.001a
1.00b 1.49
0.75–2.97
0.26
1.00b 2.22 0.76
1.23–4.00 0.64–0.90
0.008 0.002
1.00b 3.41
1.78–4.84
⬍ 0.001
⬍ 0.001
1.00b 2.43 1.16
1.26–4.69 1.01–1.33
0.008 0.036
0.02
1.00b 0.76 0.62
0.38–1.49 0.30–1.25
0.11a
1.00b 1.09
0.45–2.61
0.85
HR: hazard ratio; 95% CI: 95% confidence interval. a Test for trend. P values for the multivariate analysis are reported for variables that were retained in the final model. b Reference value.
DISCUSSION Patients with advanced RCC have a very poor prognosis. Some reports have outlined survival analyses for patients who were treated with cytokines as first-line therapy. Their results almost uniformly report a median survival for patients with advanced stage RCC ranging from 10 months to 13 months after diagnosis.6,7 Recently, 2 studies examined the survival of patients with previously treated RCC, and, surprisingly, they found a median survival from the first day of second-line therapy of 12.7 months and 16.0 months, respectively.1,2 The median survival of patients receiving second-line therapy was better than in the survival of treatment-naive patients, most likely because clinical trials of salvage therapy select patients with more slowly progressive disease. Second-
2102
CANCER November 15, 2005 / Volume 104 / Number 10
FIGURE 1. Overall survival in low-risk and high-risk patients. TABLE 3 Transplantation-Related and Disease-Related Mortality Day 100 (%)
Day 365 (%)
Variable
Low risk
High risk
Low risk
High risk
TRM Disease-related mortality Death for any cause
0.0 3.7 3.7
17.7 26.7 43.3
19.1 18.9 38.9
26.7 56.7 83.3
TRM: transplantation-related mortality.
line therapy consisted of an alternative cytokine regimen, chemotherapy, or new agents under investigation, but never included allogeneic stem cell transplantation. All of the reports mentioned above indicated the possibility of stratifying patients according to some prognostic factors that significantly separate overall survival curves between low-risk patients and high-risk patients. To our knowledge, the current study is the first to consider prognostic factors for survival in patients with RCC who undergo allogeneic stem cell transplantation. This procedure has emerged in the last few years and seems to have a response rate ⬎ 25% with a nonrecurrence-mortality rate of approximately 10%. Currently, there are few data on the impact of allogeneic transplantation on the overall survival of patients with RCC. The population in the current study was comprised of patients with disease in a very advanced phase: Patients had received a median of 2 previous lines of therapy (range, 0 – 4 lines of therapy), and the median interval from diagnosis of advanced disease to transplantation was approximately 12 months, which was very close to the expected median survival.
We analyzed the power of the following 10 parameters: PS, DFI, the number of metastatic sites, the presence of liver metastasis, the presence of mediastinal metastasis, hemoglobin level, CRP level, LDH level, calcemia, and neutrophil count. All the variables that were correlated significantly with survival in the univariate analysis were included in a multivariate Cox regression model. A prognostic score was calculated using three parameters: PS, CRP level, and LDH level. We were able to divide our population according to the prognostic score into two significantly different subsets: high-risk patients, with a median survival of 3.5 months, and low-risk patients, with a median survival of 23 months. Our prognostic score not only was able to recognize patients who were affected by more aggressive disease but also could identify patients who were at high risk for transplantation-related mortality (TRM). Low-risk patients had a 0% 100-day TRM and an estimated overall survival ⬎ 60% at 1 year after transplantation. Although the current data need to be validated in prospective studies, we believe that some important conclusions can be drawn. Prognostic factors can identify patients with very different survival, and this must be considered in clinical trial design and in clinical decision making. It most likely is not conceivable to offer allogeneic stem cell transplantation to a highrisk patient who has a median survival of approximately 100 days. Conversely, allogeneic transplantation seems to offer the best therapeutic chance to a low-risk patient. The 23-month median survival of low-risk patients undergoing transplantation is remarkable: Compared with the same subset of patients treated with different therapies (i.e., patients with no risk factors in a Motzer analysis), we find a slight improvement in survival (23 mos vs. 20 mos). Moreover, we must emphasize that our patients usually underwent transplantation as third-line therapy. Future strategies will show the impact on overall survival of allogeneic stem cell transplantation performed as a second-line therapy, when patients are more likely to fulfill the low-risk criteria. Prospective studies of allografting compared with new therapeutic strategies, including targeted agents, are warranted.
REFERENCES 1.
2.
Motzer RJ, Bacik J, Schwartz JH, et al. Prognostic factors for survival in previously treated patients with metastatic renal cell carcinoma. J Clin Oncol. 2004;22:454 – 463. Leibovich BC, Han K, Bui MHT, et al. Scoring algorithm to predict survival after nephrectomy and immunotherapy in patients with metastatic renal cell carcinoma. Cancer. 2003; 98:2566 –2575.
Survival in Allografted Renal Ca/Peccatori et al. 3.
4.
5.
Childs R, Chernoff A, Contentin N, et al. Regression of metastatic renal-cell carcinoma after nonmyeloablative allogeneic peripheral-blood stem-cell transplantation. N Engl J Med. 2000;343:750 –758. Bregni M, Dodero A, Peccatori J, et al. Nonmyeloablative conditioning followed by hematopoietic cell allografting and donor lymphocyte infusions for patients with metastatic renal and breast cancer. Blood. 2002;99:4234 – 4236. Pedrazzoli P, Da Prada GA, Giorgiani G, et al. Allogeneic blood stem cell transplantation after a reduced-intensity, preparative regimen: a pilot study in patients with refractory malignancies. Cancer. 2002;94:2409 –2015.
6.
7.
8.
2103
Motzer RJ, Mazumdar M, Bacik J, et al. Survival and prognostic stratification of 670 patients with advanced renal cell carcinoma. J Clin Oncol. 1999;17:2530 –2540. Negrier S, Escudier B, Gomez F, et al. Prognostic factors of survival and rapid progression in 782 patients with metastatic renal carcinomas treated by cytokines: a report from the Groupe Francais d’Immunotherapie. Ann Oncol. 2002; 13:1460 –1468. Gooley TA, Leisering W, Crowley J, Storer BE. Estimation of failure probabilities in the presence of competing risks: new representations of old estimators. Stat Med. 1999;18: 695–706.