Long-term follow-up of metastatic renal cancer patients ... - Nature

Bone Marrow Transplantation (2009) 44, 237–242 & 2009 Macmillan Publishers Limited All rights reserved 0268-3369/09 $32.00

www.nature.com/bmt

ORIGINAL ARTICLE

Long-term follow-up of metastatic renal cancer patients undergoing reduced-intensity allografting M Bregni1,2, M Bernardi1, P Servida1,3, A Pescarollo1,4, R Crocchiolo1, E Treppiedi1, P Corradini1,5, F Ciceri1 and J Peccatori1 1

Strategic Program in Oncology and Hematology Unit, Department of Oncology, Scientific Institute San Raffaele, Milan, Italy; Oncology Unit, Ospedale San Giuseppe, Milano, Italy; 3Istituto Oncologico Svizzera Italiana (IOSI), Bellinzona CH, Switzerland; 4 Ospedale Evangelico, Genova, Italy and 5Istituto Nazionale Tumori, Milano, Italy 2

SCT from an HLA-compatible sibling donor is an adoptive immunotherapy for cytokine-refractory, metastatic clear-cell renal cell cancer (RCC). However, the recent introduction of targeted therapy compounds has reduced the interest in this therapeutic strategy. We have reanalyzed our series with the aim to assess long-term benefit from allografting. Twenty-five RCC patients received a reduced-intensity allograft from an HLAidentical sibling donor. All patients received a thiotepa, fludarabine and CY conditioning regimen, and a cyclosporine-based GVHD prophylaxis. Best response to allograft was evaluable in 24 patients: 1 CR, 4 PR, 12 minor response/stable disease, 7 progressive disease. Oneyear survival was 48%, and five-year survival was 20%. At a median observation time of 65 months, five patients are alive, one in CR, one in PR and three with stable disease. By multivariate analysis, C-reactive protein value before transplant, the number of CD34 þ infused cells and disease status at day þ 90 significantly correlated with survival. Survival of patients at favorable/intermediate-risk according to the MSKCC score that underwent allografting was better in comparison to the survival predicted by historical controls. We conclude that 20% of cytokine-refractory RCC patients are alive long-term after allografting. Transplantation is able to induce longterm disease control in a fraction of relapsed RCC patients Bone Marrow Transplantation (2009) 44, 237–242; doi:10.1038/bmt.2009.9; published online 23 February 2009 Keywords: renal carcinoma; allografting; reduced-intensity conditioning; graft-versus-tumor

Correspondence: Dr M Bregni, Medical Oncology Unit, Ospedale San Giuseppe, Via San Vittore, 12, Milano I-20121, Italy. E-mail: [email protected] Received 27 May 2008; revised 11 November 2008; accepted 1 December 2008; published online 23 February 2009

Introduction In 2000, Childs et al.1 at NIH published the first series of patients with cytokine-refractory, clear-cell renal cancer treated with a nonmyeloablative allogeneic transplantation from an HLA-identical sibling donor: they observed a 53% response rate, including complete responses. They established the proof of concept that a graft-versus-malignancy effect exists also in solid tumors, and that allogeneic SCT is able to mediate this effect. After the seminal paper of Childs et al., other groups have reported variable results of reduced-intensity allografting in cytokine-resistant/refractory renal cancer (reviewed in reference2). Recently, the EBMT Solid Tumour Working Party (STWP) reported 124 patients, treated in 16 institutions with different conditioning regimens and GVHD prophylaxis,3 and confirmed the response rate and the transplant-related mortality already noted by Childs et al. We now know that allogeneic transplantation after reduced-intensity conditioning can induce responses in 30–40% patients with clear-cell renal cell carcinoma; most of these responses are partial, but in some cases complete. These responses are often delayed in onset, frequently associated with GVHD or cyclosporine withdrawal or with infusion of donor lymphocytes. In an effort to prospectively identify the patients who could benefit most from allogeneic transplantation, Peccatori et al.,4 on behalf of the EBMT STWP, examined the pretransplant characteristics of 70 patients who underwent allografting in selected EMBT centers in Europe. They found that three factors (performance status, C-reactive protein (CRP) and lactate dehydrogenase (LDH)) were significantly associated with survival in multivariate analysis. These factors were able to stratify patients in two groups with very different median survival (3.5 months for bad-prognosis patients vs 23 months for good-prognosis patients). They suggested that these easily available parameters can identify patients who are candidates to allografting, and assist in treatment decision making and selection of appropriate program. Thus, there is evidence that allografting can induce clinical responses in cytokine-refractory renal cancer. We do not know, however, if allografting induces a long-term benefit, nor if it influences the natural history of cytokinerefractory renal cancer.

Allografting for renal cancer M Bregni et al

238

Notably, the therapeutic scenario of renal cancer has dramatically changed in the past 5 years. After elucidation of the key role of hypoxia-inducible factor 1-alfa and associated genes (VHL, VEGF, mTOR, and others) in inducing neovascularization and tumor growth and metastasis, inhibitory molecules have been identified and introduced into the clinic. The novel targeted agents sunitinib, temsirolimus, sorafenib and bevacizumab are now standard treatment for metastatic renal cell cancer, either in first line or after cytokine failure.5 As a consequence, allografting as a salvage therapy has been questioned, and its use has dramatically decreased, even though it can induce a long-lasting control of the disease. Even if there is temporary benefit, the outcome of metastatic RCC has not been substantially changed by the introduction of antiangiogenic therapies. In particular, patients who fail a first-line antiangiogenic treatment and patients with poor prognostic factors according to the Memorial Sloan Kettering Cancer Centre (MSKCC) score have a median survival ranging from 5 months to 1 year, irrespective of second-line treatment. At our institution we have treated in a prospective clinical protocol (designated ‘Allorene’) 25 patients with metastatic renal cancer who have failed a cytokine therapy, with a reduced-intensity allograft regimen and SCT from a HLA-identical sibling donor. We here report the long-term follow-up (5 years) of our series of patients treated in this protocol, as a contribution to the assessment of clinical benefit of allografting.

Patients and methods Patients Twenty-five patients with cytokine-refractory RCC underwent reduced-intensity allografting from an HLA-compatible sibling donor from February 1999 to May 2005. Main patient characteristics are reported in Table 1. Nephrectomy was carried out in all cases. All patients had cytokinerefractory metastatic disease at the time of the transplant and many had received additional lines of treatment including chemotherapy; only one patient received allografting as first-line treatment for metastatic disease. Median days from diagnosis of primary tumor to allograft were 822 (range, 165–3758). The first seven patients in this series have already been included in a previous report.6 Allografting All patients received a thiotepa, fludarabine and CY conditioning regimen, as already described.6 PBSC from a HLA-identical sibling donor were given at the target dose of X5 106 CD34 þ cells/kg recipient body wt. GVHD prophylaxis was based on CSA plus either short-course MTX (21 patients) or sirolimus (four patients). Hematopoietic reconstitution was assessed by peripheral blood counts and BM examination. Hematopoietic reconstitution was defined as the first of 3 days with neutrophil counts 4500 106/l, and unsustained platelet counts 420 109/l. Donor–recipient chimerism analysis was carried out on whole BM mononuclear cells, and on Bone Marrow Transplantation

Table 1

Main patient characteristics Value

Median age (range) Sex (M/F) Histology Clear cell Papillary Previous treatments (range)

53 (40–65) 18/7 24 1 1 (0–3)

ECOG PS 0 1 2

0 19 3

Patients with LDH level 41.5 times ULN Hb level oLLN Corrected serum calcium level 410 mg/100 ml (2.5 mmol/l) Disease-free interval o1 year

1 21 2 17

MSKCC score 0 risk factor 1 risk factor 2–3 risk factors

1 19 5

Abbreviations: LDH ¼ lactate dehydrogenase; LLN ¼ lower limit of normal; MSKCC ¼ Memorial Sloan Kettering Cancer Centre; ULN ¼ upper limit of normal.

CD3 þ and CD13 þ cell subsets obtained by cell sorting from PBMC using PCR of informative minisatellite regions at monthly intervals for the first 6 months, at 3-month intervals for the first 2 years and every 6 months thereafter. A complete myeloid or lymphoid chimerism was defined as 490% donor origin.

Follow-up and immune manipulation Disease assessment was carried out by clinical exam and conventional imaging (CT scan, MRI) at bimonthly intervals for the first year, at 4-month intervals for the second and third year and at 6-month intervals thereafter. Disease response was assessed by standard WHO criteria. Cyclosporine was reduced or withdrawn after day þ 90 or earlier based on disease progression. Donor lymphocyte infusion (DLI) at escalating doses, starting from 106 CD3 þ cells/kg, was given to patients with mixed chimerism or with progressive disease after CSA withdrawal and no concurrent GVHD. Patients with disease progression received sorafenib or sunitinib as they became available in the clinic (that is, after 2006). Statistical analysis We analyzed the correlation of the following variables with survival: age at transplant (continuous variable), time from diagnosis to transplant (continuous), serum calcium corrected for albumin levels, LDH, CRP, Hb level before transplantation, Karnofsky performance status (continuous), number of CD34 þ /kg infused (X5 vs o5 106/kg), number of CD3 þ /kg infused (above vs below the median value), progressive disease at day þ 90 after transplant (yes vs no), occurrence of acute (any grade) or chronic GVHD. Biochemical parameters were categorized as normal or abnormal according to laboratory normality ranges.


239

Prognostic classification Three risk categories for predicting survival of patients refractory to a first-line cytokine treatment were identified at Memorial Sloan Kettering Cancer Centre (‘MSKCC score’) on the basis of three pretreatment clinical features, that is, low Karnofsky performance status (p70), serum Hb below the lower limit of normal and high corrected serum calcium (410 mg/100 ml). Patients can be assigned to one of the three risk categories based on the number of risk factors: those with zero risk factors (favorable risk), those with one (intermediate risk) and those with two or three (poor risk). Median survival significantly differed in these categories (favorable: 22 months; intermediate: 11.9 months; poor: 5.4 months).11 Patients of our series were assigned a prognostic score, and their actual outcome was compared to that predicted by the respective risk category.

Results Hematopoietic engraftment All evaluable patients (N ¼ 24) had a complete hematopoietic engraftment within 1 month from transplant. A full donor chimerism was achieved in all patients by day þ 60 after transplant. In one patient there was a late graft failure at day þ 150 after transplant (lymphoid engraftment 20–25% donor, myeloid engraftment o40% donor), that was rescued by reinfusion of selected CD34 þ cells from the original donor. Median CSA withdrawal day after BMT was 120 (21–351) days. Six patients received DLI at escalating doses for progressing or nonresponding disease at a median of 360 (72–1095) days from transplant. Six patients had two infusions of DLI, four patients had three infusions, one patient had four infusions. Twelve patients had acute GVHD (aGVHD), five grade I, five grade II, one grade III and one grade IV. Nine patients had chronic GVHD (cGVHD), seven limited and two extensive. Response and survival Best response to allograft was evaluable in 24 patients: 1 CR and 4 PR. Twelve patients reported minor response

Survival Function 1.0

Survival function Censored

0.8 Cum survival

Univariate Cox regression7 was performed to assess any association between pretransplant variables and outcome, whereas time-dependent Cox regression was used with post-transplant variables.8 Factors that correlated to survival with a Po0.20 entered in a multivariate Cox regression model with a backward selection procedure; only variables with a Po0.05 were retained in the final model. Survival analysis was performed using the Kaplan–Meier method.9 A semilandmark analysis10 was used to graphically represent patients who progressed within 90 days after transplantation vs the others, the landmark day being the date of progression for those who progressed by day þ 90, and the day þ 120 (median day of progression for all patients) for the others.

0.6

0.4

0.2

0.0 0

500

1000

1500

2000

2500

Survival (to DOLC) Figure 1 Overall survival.

(MR) or stable disease (SD), and seven patients reported progressive disease (PD). Overall, the response rate (CR þ PR) was 20%, whereas the clinical benefit rate (CR þ PR þ MR þ SD) was 71%. Six patients died because of transplant related mortality (TRM) at days þ 12, þ 71, þ 86, þ 151, þ 259, þ 478. Cause of death was infection in four cases, GVHD in one case and acute renal failure in one case. Fourteen patients died from progressive disease at a median 415 (36–958) days from transplant. Median survival was 336 (12–2332 þ ) days. One-year overall survival was 48% (95% CI: 28–68), and five-year overall survival was 20% (95% CI: 4–36; Figure 1). At a median observation time of 65 (40–72) months, five patients are alive, one in CR, one in PR and three with stable disease. These latter patients are receiving a tyrosine kinase inhibitor for their disease, two patients are receiving sunitinib and one is receiving sorafenib. None of these patients has long-term side effects from the transplant, specifically none has cGVHD.

Survival analysis Twenty-one patients had Hb level below the lower limit of normal. Two patients had corrected serum calcium levels higher than 10 mg/100 ml. One patient had high lactate dehydrogenase level. Seventeen patients had a disease free interval of less than 1 year. On univariate analysis, LDH, CRP, Karnofsky performance status, number of CD34 þ /kg infused cells, aGVHD and disease status at day þ 90 significantly correlated with survival and were included in multivariate analysis. Factors that were retained in the final multivariate model were CRP, number of CD34 þ infused cells and disease status by day þ 90 after transplantation. Results are shown in Table 2. To graphically represent the impact of early disease progression after transplant on overall survival, a semilandmark plot was performed on 24 evaluable patients Bone Marrow Transplantation


240 Table 2

Cox regression analysis

Variable Age at Tx Time from diagnosis to Tx (days) Corrected calcium LDH CRP Hb Karnofsky performance status Infused CD34+/kg Infused CD3+/kg PD within +90 after Tx aGVHD occurrence cGVHD occurrence

HR

95% CI

P value

0.99 1.00 1.33 10.99 6.29 1.82 0.95 0.42 0.99 27.74 2.05 0.75

0.93–1.06 0.99–1.00 0.30–5.80 0.99–121.24 1.98–19.97 0.42–7.94 0.92–0.99 0.17–1.07 0.37–2.66 2.77–277.49 0.75–5.60 0.24–2.31

0.87 0.67 0.71 0.05 0.002 0.42 0.01 0.07 0.98 0.005 0.16 0.62

Multivariate P value

0.012

0.015 0.010

Abbreviations: 95% CI ¼ 95% confidence interval; aGVHD ¼ acute GVHD; cGVHD ¼ chronic GVHD; CRP ¼ C-reactive protein; HR ¼ hazard ratio; LDH ¼ lactate dehydrogenase; PD ¼ progressive disease; Tx ¼ transplant. P values for the multivariate analysis were reported only for variables included in the final model.

Survival functions

0.8 Cum survival

0.8

0.6

0.4

0.6

0.4

0.2

0.2

0.0

0.0 0

500

1000 1500 FU_semilandmark

2000

Group 1 2 1-censored

1.0

PD_semilandmark+90 0 1 0-censored

1.0

Cum survival

Survival functions

0

2500

500

1000

1500

2000

2500

Survival (to DOLC)

Survival according to early progression (day þ 90 after transplant). PD ¼ progressive disease.

Figure 3 Survival according to Memorial Sloan Kettering Cancer Centre

(Figure 2). Patients who progressed early (N ¼ 8) had a median survival of 71 days and no long-term survival; patients whose disease did not progress by day þ 90 had a median survival of 700 days (Po0.0001) and a 30% longterm survival. Thus, early progression is strictly related to survival after allografting.

One patient in our series fell into the favorable-risk group according to MSKCC score, 20 patients were in the intermediate- and 5 in the poor-risk group (Table 1). To study the correlation of MSKCC score with survival, we divided our patients into two groups: favorable/intermediate-risk vs a poor-risk group (Figure 3). The median survival of favorable/intermediate prognosis patients was 415 days vs 71 days of the poor-prognosis patients (P ¼ 0.018) The long-term survival of the favorable/ intermediate-risk group was 25%, and it was 0% in the high-risk group.

Figure 2

Prediction of survival according to MSKCC prognostic score The median survival of our patients overall was similar to that of the intermediate group of the MSKCC series: 11 months vs 11.9 in the Motzer series, respectively. However, there was a clear advantage in the long-term survival of our series (20% at 5 years) compared to that predicted by the MSKCC prognostic score (11% in the intermediate group and 0% in the poor-prognosis group at 3 years, respectively). Bone Marrow Transplantation

(MSKCC) score. Risk factors are described in the Patients and methods section.

Discussion At 5 years after allografting, 25% of cytokine-refractory RCC patients are alive. All these patients had a CRP within


241

normal limits, received more than 5 10e6 CD34 þ cells/ kg, and had stable or responding disease at day þ 90 after transplant. At the time of this writing, three patients with stable disease are receiving antiangiogenic therapy, and two are in remission without further therapies. Our experience compares favorably with that of Childs et al., who recently updated the follow-up of their series of 74 patients12: they reported a median survival of 525 days, and a long-term survival of approximately 10% in patients with a clear-cell histology. Three variables were correlated to survival in multivariate analysis, that is, C-reactive protein, number of infused CD34 þ cells and disease status after transplant. The number of infused CD34 þ cells is strictly related to the speed and completeness of engrafting: the higher the cell dose infused, the faster is the engraftment, and the achievement of a complete chimeric status. Disease status after transplant has a particular meaning in this setting. We know that the disease response attributable to graft-versusrenal cancer effect occurs slowly (4–5 months after transplant),1 and disease often progresses in this frame period. In our series, one-third of the patients had a progression of the disease at 3 months from transplant. Disease progression induces early withdrawal of cyclosporine, with subsequent increase in the rate of aGVHD and TRM. Most toxicity deaths (4 out of 6) were attributable to infectious events in the first year after transplant, and even if not statistically significant, we can assume that steroid treatment of aGVHD may have had a role in this high infectious rate. Better patient selection and control of disease progression after transplant are key to the success of this strategy. Other reports have found cGVHD to impact survival in multivariate analysis3,13: Blaise et al. have reported that in a series of 57 patients with various solid tumors, cGVHD and status of disease at transplant were the major determinants of survival. However, RCC patients were a minority of the whole population in that report. Barkholt et al. reported in a retrospective analysis of allografting in renal cancer in Europe that performance status, cGVHD, DLI and extent of disease (less than three metastatic sites) as statistically significant for survival. This analysis included also some of the patients described in this series. We did not find such an effect of cGVHD in our series, probably due to the limited number of patients. It is still not understood why some patients respond to this treatment and others do not. On a clinical ground, we have identified three factors that correlate with survival; we presume that biologic factors may better define the responders. Graft-verus-tumor (GVT) effects against renal cancer may be associated with tumor-specific immune responses to Ag expressed on RCC cells, such as gene products derived from HERV-E, expressed in a minority of RCC patients.12 The long-term survival of our patients was actually higher than that predicted by the prognostic score based on the MSKCC retrospective analysis. The 20% OS of our intermediate prognosis group is overlapping with the good prognosis group of MSKCC series. In contrast, the survival of patients with an intermediate prognostic score is 11% in the MSKCC analysis. It means that overall survival of our patients with one-risk factor is similar to the OS of

MSKCC patients without risk factors. The similarity in the median survival is probably accounted for by the shape of the survival curve in our series, which reflects the early TRM in the first year. Moreover, our long-term survival is evaluated at 5 years, compared to 3 years of the MSKCC series. Notably, this survival has not been reported before in cytokine-refractory patients, for whom, until recently, there was no salvage therapy available. Now we have witnessed a revolution in the management of metastatic renal cancer, after the elucidation of pathogenetic mechanisms of the disease and the subsequent design of specific inhibitory molecules. The blockade of AKT/PI3K/mTOR pathway has resulted in a survival advantage in patients with poor prognostic factors.14 Inhibition of VEGF or VEGFinitiated signaling has also demonstrated considerable antitumor activity.15–17 The introduction in the clinical setting of molecularly targeted agents (bevacizumab, sorafenib, sunitinib and temsirolimus) has considerably improved progression-free survival in first and second lines, and has ameliorated overall survival in first line in poor prognosis patients. Data on long-term survival of patients who have failed a cytokine therapy and have been treated with a tyrosine kinase inhibitor are not available yet for a comparison. Targeted therapy has practical advantages over a complex treatment as allogeneic transplantation; indeed, patient referral for transplant has considerably decreased over the past years. However, the benefit of these new agents on the overall survival remains limited, particularly in the poor-prognosis subgroup and in patients relapsed after a first-line antiangiogenic treatment. In contrast, allogeneic transplantation has the potential of inducing long-term disease control. A prospective, randomized study of comparison between antiangiogenic therapy and allografting would ideally assess the best treatment. Unfortunately, such a trial would be very difficult to organize, because patients eligible to allografting are a small subset of all renal cancer patients. Allografting requires the availability of an HLA-compatible sibling donor, age less than 60 years (median age of RCC patients is 65) and absence of significant comorbidities. In addition, given its still high TRM, allografting would be better positioned as a second-line treatment after antiangiogenic drug failure. A rationale exists for combining the immune effect of allografting and the molecularly targeted effect of the new drugs. The antiangiogenic effect of some of these agents may be directly beneficial, as tumor-produced VEGF may contribute to immune suppression and limit the generation of an antitumor immune response by inducing inadequate DC differentiation.18 These data suggest that the inhibition of VEGF may augment the antitumor effect of adoptive immunotherapy, and be advantageously associated with allogeneic transplantation. Other lines of evidence suggest that antiangiogenictargeted therapies have effects on the immune system that can be particularly relevant for their combination with immune-targeted therapies. Finke et al. suggested that treatment with sunitinib can promote a type-1 cytokine response (IFN-g) and simultaneously decrease the type-2 response (IL-4) in patients with metastatic RCC.19 Zeiser Bone Marrow Transplantation


242

et al.20 have demonstrated that rapamycin (the parent compound of temsirolimus) and CD4 þ CD25 þ regulatory T cells synergize in preserving GVT activity whereas inhibiting GVHD after BMT. Taken together, these findings could have implications for furthering the clinical benefit of such therapies in RCC and other tumors, and particularly for combining these agents with allogeneic transplant. As the transplant conditioning regimen is largely ineffective in cytoreducing the tumor burden, and because GVT effects are typically delayed, patients often succumb to their tumor before an antitumor effect can occur. A treatment that controls the disease progression before the GVT effect may reduce the need to withdraw immune suppression, avoid grade 3–4 aGVHD, reduce the TRM and increase progression-free survival. Temsirolimus may have the characteristics for the combination. In addition, temsirolimus, like its parent compound sirolimus, may have favorable effects on the immune cells that regulate GVHD and GVT activity of the graft. Based on these assumptions, a phase I–II study of allografting followed by temsirolimus will be started in RCC patients relapsed after or refractory to a first-line antiangiogenic treatment. In conclusion, allografting is able to induce long-term disease control in a fraction of cytokine-resistant RCC patients at intermediate prognosis. The association of the immune effects of allografting with the effects of novel targeted therapies is a promising new field of investigation. Future strategies should consider incorporating new drugs in the transplant regimen, with the aim of decreasing progression and prolonging overall survival.

Acknowledgements We thank the doctors and the nurses of the Bone Marrow Transplant Unit for their dedicated patient care, and Ms Stefania Trinca for data management. This work has been supported by AIRC Regional Grant and by Ricerca Finalizzata 2005 to MBr.

References 1 Childs R, Chernoff A, Contentin N, Bahceci E, Schrump D, Leitman S et al. Regression of metastatic renal-cell carcinoma after nonmyeloablative allogeneic peripheral-blood stem-cell transplantation. N Engl J Med 2000; 343: 750–758. 2 Demirer T, Barkholt L, Blaise D, Pedrazzoli P, Aglietta M, Carella AM et al. Transplantation of allogeneic hematopoietic stem cells: an emerging treatment modality for solid tumors. Nat Clin Pract Oncol 2008; 5: 256–267. 3 Barkholt L, Bregni M, Remberger M, Blaise D, Peccatori J, Massenkeil G et al. Allogeneic haematopoietic stem cell transplantation for metastatic renal carcinoma in Europe. Ann Oncol 2006; 17: 1134–1140. 4 Peccatori J, Barkholt L, Demirer T, Sormani MP, Bruzzi P, Ciceri F et al. Prognostic factors for survival in patients with advanced renal cell carcinoma undergoing nonmyeloablative allogeneic stem cell transplantation. Cancer 2005; 104: 2099–2103.

Bone Marrow Transplantation

5 Motzer RJ, Bukowski RM. Targeted therapy for metastatic renal cell carcinoma. J Clin Oncol 2006; 24: 5601–5608. 6 Bregni M, Dodero A, Peccatori J, Pescarollo A, Bernardi M, Sassi I 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. 7 Cox DR. Regression models and life tables. J R Stat Soc 1972; 34 Series B: 187.220–226. 8 Klein JP, Rizzo JD, Zhang MJ, Keiding N. Statistical methods for the analysis and presentation of the results of bone marrow transplants. Part I: Unadjusted analysis. Bone Marrow Transplant 2001; 28: 909–915. Part II: Regression modelling. Bone Marrow Transplant 2001; 28: 1001–1011. 9 Kaplan EL, Meier P. Non-parametric estimation from incomplete observations. J Am Stat Assoc 1958; 53: 457–481. 10 Baron F, Maris MB, Sandmaier BM, Storer BE, Sorror M, Diaconescu R et al. Graft-versus-tumor effects after allogeneic hematopoietic cell transplantation with nonmyeloablative conditioning. J Clin Oncol 2005; 23: 1993–2003. 11 Motzer RJ, Bacik J, Schwartz LH, Reuter V, Russo P, Marion S et al. Prognostic factors for survival in previously treated patients with metastatic renal cell carcinoma. J Clin Oncol 2004; 22: 454–463. 12 Takahashi Y, Harashima N, Kajigaya SH, Yokoyama H, Cherkasova E, McCoy JP et al. Regression of human kidney cancer following allogeneic stem cell transplantation is associated with recognition of an HERV-E antigen by T cells. J Clin Invest 2008; 118: 1099–1109. 13 Blaise D, Bay JO, Faucher C, Michallet M, Boiron JM, Choufi B et al. Reduced-intensity preparative regimen and allogeneic stem cell transplantation for advanced solid tumors. Blood 2004; 103: 435–441. 14 Hudes G, Carducci M, Tomczac P, Dutcher J, Figlin R, Kapoor A et al. Temsirolimus, interferon alfa, or both for advanced renal-cell carcinoma. N Engl J Med 2007; 356: 2271–2281. 15 Escudier B, Pluzanska A, Koralewski P, Ravaud A, Bracarda S, Szczylik C et al. Bevacizumab plus interferon alfa-2a for treatment of metastatic renal cell carcinoma: a randomised, double-blind phase III trial. Lancet 2007; 370: 2103–2111. 16 Motzer RJ, Hutson TE, Tomczac P, Michaelson MD, Bukowski RM, Rixe O et al. Sunitinib versus interferon alfa in metastatic renal-cell carcinoma. N Engl J Med 2007; 356: 115–124. 17 Escudier B, Eisen T, Stadler WM, Szczylik C, Oudard S, Siebels M et al. Sorafenib in advanced clear-cell renal-cell carcinoma. N Engl J Med 2007; 356: 125–134. 18 Gabrilovich DI, Chen HL, Girgis KR, Cunningham HT, Meny GM, Nadaf S et al. Production of vascular endothelial growth factor by human tumors inhibits the functional maturation of dendritic cells. Nat Med 1996; 2: 1096–1103. 19 Suppiah R, Finke J, Rini BI, Richmond A, Wood L, Elson P et al. T regulatory cells (Treg) in patients with metastatic renal cell carcinoma (mRCC) decrease during Sunitinib treatment: correlation with clinical responses and T helper 1/T helper 2 (Th1/Th2) bias. J Clin Oncol 2006; 24: 1065a. 20 Zeiser R, Leveson-Gower DB, Zambricki EA, Kambham N, Beilhack A, Loh J et al. Differential impact of mammalian target of rapamycin inhibition on CD4+CD25+Foxp3+ regulatory T cells compared with conventional CD4+ T cells. Blood 2008; 111: 453–462.