Abstract
Purpose: We report the role of relative lymphocyte count (RLC) as a potential biomarker with prognostic impact for catumaxomab efficacy and overall survival (OS) based on a post hoc analysis of the pivotal phase II/III study of intraperitoneal catumaxomab treatment of malignant ascites.
Experimental Design: The impact of treatment and RLC on OS was evaluated using multivariate Cox models. Kaplan–Meier and log-rank tests were used for group comparisons. Survival analyses were performed on the safety population [patients with paracentesis plus ≥1 dose of catumaxomab (n = 157) and paracentesis alone (n = 88)]. Determination of the optimal cutoff value for RLC was based on five optimality criteria.
Results: OS was significantly longer with catumaxomab versus paracentesis alone (P = 0.0219). The 6-month OS rate with catumaxomab was 28.9% versus 6.7% with paracentesis alone. RLC had a positive impact on OS and was an independent prognostic factor (P < 0.0001). In patients with RLC > 13% (n = 159: catumaxomab, 100 and control, 59), catumaxomab was associated with a favorable effect on OS versus paracentesis alone (P = 0.0072), with a median/mean OS benefit of 41/131 days and an increased 6-month survival rate of 37.0% versus 5.2%, respectively. In patients with RLC ≤ 13% at screening (n = 74: catumaxomab, 50 and control, 24), the median (mean) OS difference between the catumaxomab and the control group was 3 (16) days, respectively (P = 0.2561).
Conclusions: OS was significantly improved after catumaxomab treatment in patients with malignant ascites. An RLC > 13% at baseline was a significant prognostic biomarker. Clin Cancer Res; 20(12); 3348–57. ©2014 AACR.
Translational Relevance
The aim of this study was to investigate the role of relative lymphocyte count (RLC: defined as the percentage of lymphocytes within the peripheral white blood cell count) as a potential new biomarker for the efficacy of catumaxomab. Catumaxomab is a trifunctional monoclonal antibody that is approved in the European Union for the intraperitoneal treatment of malignant ascites in patients with epithelial cell adhesion molecule (EpCAM)-positive carcinomas.
The RLC before therapy was found to be a potential biomarker for catumaxomab treatment efficacy. Predicting the response to cancer therapy is an increasingly important area of clinical research. For targeted therapies such as catumaxomab, biomarkers could potentially lead to significant improvements in current oncology practice by providing the ability to define optimal time schedules and to predict efficacy and tailor treatment to individual patients.
Introduction
Catumaxomab is a trifunctional monoclonal antibody that is approved in the European Union for the intraperitoneal treatment of malignant ascites in patients with epithelial cell adhesion molecule (EpCAM)-positive carcinomas (1). The trifunctional mode of action of catumaxomab occurs by binding to epithelial tumor cells via EpCAM to T cells via CD3 and activation of Fcγ-receptor I-, IIa-, and III-positive accessory cells via its functional Fc domain (2–5). The activation of different immune effector cells at the tumor site by catumaxomab leads to improved tumor cell elimination by a variety of immunologic killing mechanisms (4, 6) and the presence of a functional immune system is key to its activity (7). In addition to malignant ascites (8), catumaxomab has also shown efficacy in the treatment of peritoneal carcinomatosis (9) and ovarian cancer (10).
The efficacy of intraperitoneal catumaxomab in the treatment of malignant ascites was demonstrated in a pivotal phase II/III study (8). In this randomized study of patients with malignant ascites due to different epithelial cancers, catumaxomab plus paracentesis significantly prolonged puncture-free survival (PuFS; median, 46 vs. 11 days; HR, 0.254; P < 0.0001) and time to next paracentesis (median, 77 vs. 13 days; HR, 0.169; P < 0.0001) compared with the control arm (paracentesis alone). In addition, in the original analysis of this study, there was a trend for improvement in overall survival (OS) with catumaxomab versus paracentesis alone (median, 72 vs. 68 days; HR, 0.723; P = 0.0846) in the overall population and a significant difference in patients with gastric cancer (n = 66; median, 71 vs. 44 days; HR, 0.469; P = 0.0313).
An OS benefit of catumaxomab was also demonstrated in an analysis of patients with peritoneal carcinomatosis due to colon, gastric, or pancreatic cancer (11). In this post hoc matched-pair analysis, 24 patients with peritoneal carcinomatosis treated with catumaxomab in a clinical study were compared with an equally sized control group of patients with peritoneal carcinomatosis treated with conventional intravenous chemotherapy. The median OS from the time of diagnosis of peritoneal carcinomatosis was 502 days with catumaxomab versus 180 days in the control group (HR, 0.421; P = 0.0083). The potential survival benefit of catumaxomab is being investigated in further clinical studies.
Predicting the response to cancer therapy is an increasingly important area of clinical research with numerous attempts to identify biomarkers that correlate with therapeutic effects to better select those patients who benefit most from the treatment (12–16). For targeted therapies, biomarkers could potentially lead to significant improvements in current oncology practice. By providing the ability to predict the efficacy of a specific therapy, biomarkers can guide treatment decision making for individual patients.
Besides a clear efficacy in ascites control in the pivotal study (8), prolonged OS could be observed in about 50% of the patients after catumaxomab treatment. Therefore it was of interest if those patients with benefit in OS could be identified by prognostic biomarkers. As the proposed mode of action of catumaxomab involves the activation of immune effector cells, these cells could be potential biomarkers for the efficacy of catumaxomab. An independently performed hypothesis-generating analysis was conducted to identify potential biomarkers predicting a positive effect of catumaxomab on OS in patients with peritoneal carcinomatosis. The impact of the following parameters was investigated: age, Karnofsky index (KI), relative lymphocyte count (RLC) and absolute lymphocyte count (ALC), relative and absolute granulocyte count, T-cell subsets, natural killer (NK) cells, and monocytes before catumaxomab therapy. Correlation analysis, Kaplan–Meier curves, receiver operator characteristic (ROC) calculation, and multivariate regression were used for statistical analysis. KI (P = 0.002) and RLC (defined as the percentage of lymphocytes within the peripheral white blood cell count) before treatment (P = 0.039) were found to be of prognostic value, being positively correlated with OS (17). Therefore, RLC and other potential biomarkers were investigated for their impact on response to catumaxomab treatment in this study with focus on OS. This analysis was performed post hoc in a subset of patients in the pivotal study who received at least one dose of catumaxomab (safety population).
Patients and Methods
Study design
This post hoc analysis of the impact of RLC on OS and PuFS used data from the 2-arm, randomized, open-label, pivotal phase II/III study (EudraCT number: 2004-000723-15; ClinicalTrials.gov identifier: NCT00836654) of catumaxomab in patients with symptomatic malignant ascites secondary to epithelial cancers requiring symptomatic therapeutic paracentesis (8). In this study, patients were randomized in a 2:1 ratio to either paracentesis plus catumaxomab (catumaxomab: n = 170) or paracentesis alone (control: n = 88) and stratified by cancer type (ovarian: n = 129 and non-ovarian: n = 129). Patients in the control group who fulfilled the eligibility criteria and had 2 therapeutic punctures after paracentesis on day 0 were permitted to receive catumaxomab in a subsequent, single-arm, crossover period. The primary endpoint for this study was PuFS. After the end of the study, patients were followed up until death. After the end of the study, patients were followed up until death. Despite the primary endpoint for this study was PuFS this post hoc analysis focused on the secondary endpoint OS, the most relevant clinical parameter in oncology studies.
Potential biomarkers for catumaxomab outcome
In the hypothesis-generating analysis, the RLC at screening was identified as a potential biomarker for a survival benefit of catumaxomab therapy. Among a number different parameters investigated, including ALC, CD4+ T cells, CD8+ T cells, CD19+ B cells, and NK cells, RLC and KI were the only prognostic parameters that significantly correlated with OS (17). On the basis of these findings, the data of the pivotal study were analyzed for the impact of RLC as a predictive biomarker for the effect of catumaxomab treatment. In addition to the impact of RLC, the impact of the absolute numbers of lymphocytes, granulocytes, and monocytes in the peripheral blood and the KI before treatment (screening) was evaluated.
Safety population for OS analysis
The intent-to-treat (ITT) population of the pivotal phase II/III study comprised 258 patients (8). The survival analyses reported in this article were performed primarily on the safety population (n = 245), that is, patients who received at least one dose of catumaxomab (n = 157) or were treated with paracentesis alone (control group: n = 88). Thirteen patients (7.6%) randomized to the catumaxomab group (5 ovarian and 8 non-ovarian cancer) did not receive catumaxomab and were excluded from the analysis of the safety population. Four patients died between randomization and first treatment with catumaxomab, 5 patients withdrew their consent, and 4 patients were excluded for other reasons, including a serious adverse event (ileus), failure to meet inclusion criteria, and problems with ascites drainage. Exclusion of these patients from the analysis of OS was considered as appropriate and consistent with statistical principles for clinical trials (18, 19).
Statistical analysis
Of the 245 patients in the safety population, 233 (catumaxomab, n = 150; control, n = 83) had evaluable screening data for the parameters of interest and were included in the statistical biomarker analysis. The parameters analyzed as potential biomarkers were derived from the peripheral blood cell counts before the start of treatment: RLC, ALC, absolute neutrophil count (ANC), and absolute monocyte count (AMC). Furthermore, KI at screening was analyzed for its prognostic impact. The impact of treatment and the potential biomarkers was evaluated using various Cox models, including separate Cox models for the 2 treatment groups. For assessing the heterogeneity of treatment effects among the levels of a potential biomarker and for assessing the predictive value, a statistical test for interaction was performed within the Cox model, which is consistent with recommendations for reporting subgroup analyses in clinical trials (20).
To identify the optimal RLC cutoff value, a cutoff point determination method for survival analysis was applied (21)—together with further optimality criteria, which were applied for various RLC cutoff values. These optimality criteria included the P value (for the log-rank test comparing the treatment effects above and below the cutoffs), the HR (for assessing the treatment effects above and below the cutoffs), the OS medians (for patients above the cutoffs—separately for each treatment group), and the sample size (of patients above the cutoffs). All statistical analyses were explorative in nature; therefore, P values should be interpreted descriptively.
Cutoff value–defined treatment groups were compared for OS and PuFS by Kaplan–Meier curves and associated estimates, by log-rank test and HR including 95% confidence interval (CI). Patients randomized to the control group who crossed over to catumaxomab treatment were censored at the date of crossover, that is, any OS component observed after crossover was not considered for the OS analysis, as it may have been influenced by catumaxomab treatment.
Results
Follow-up OS data
Ten patients were alive at the cutoff used for reporting the OS results of the phase II/III study (8). This cutoff was 8 months (ovarian cancer) and 7 months (non-ovarian cancer) after last patient out (LPO). Nine of those 10 patients had been randomized to catumaxomab (7 ovarian, 1 gastric, and 1 uterine cancer) and 1 patient (ovarian cancer) to paracentesis alone. The latter patient crossed over to catumaxomab after reaching the primary endpoint of the study. KI, ALC, AMC, ANC, and RLC at screening of these patients were comparable to those patients considered for the OS cutoff applied in the phase II/III study (data not shown; ref. 8). The median RLC at screening of these 10 patients was 19.9% (range, 14.3%–49.0%). At the follow-up analysis 35 months (ovarian cancer) and 34 months (non-ovarian cancer) after LPO, 1 catumaxomab-treated patient with non-ovarian cancer was alive. Follow-up survival results in the ITT and safety populations are shown in Table 1. In the safety population, there was a statistically significant benefit in OS for catumaxomab-treated patients compared with control patients (P = 0.0219; HR, 0.649; 95% CI, 0.446–0.943; Fig. 1A). The 6-month OS rate in catumaxomab-treated patients was 28.9%, compared with 6.7% in the control group.
A, OS of the safety population including long-term follow-up data. B and C, OS in relation to the RLC at screening in catumaxomab-treated (B) and control (C) patients.
Follow-up OS in the ITT and safety populations
The following results are based on the follow-up analysis of the safety population.
Biomarkers with OS impact
The Cox analysis showed that treatment with catumaxomab and RLC at screening had a strong positive impact on OS (Table 1). There was a strong correlation between RLC and OS in catumaxomab-treated patients but this was not seen in the control group (Fig. 1B and C). A significant positive impact on OS (P < 0.0001) was also observed for KI at screening (Table 1). The effect of catumaxomab treatment on OS remained significant (P = 0.0060) after adjustment for RLC. According to this model, catumaxomab treatment resulted in an HR of 0.582, corresponding to a catumaxomab-related risk reduction of 41.8%. An increase of 1% in RLC was associated with a risk reduction of 3.8% (HR, 0.962). In contrast, RLC had no significant effect on OS in control patients (P = 0.0974).
In a sensitivity analysis, we further investigated the impact of the RLC components [i.e., absolute values of white blood cells (WBC)] to OS (Table 1). It turned out that ALC had a positive impact on OS (P = 0.0260). The impact of ANC (P < 0.0001) and AMC (P = 0.0425) at screening on OS was also significant but negative, that is, OS was worse with increasing numbers of neutrophils and monocytes.
Plotting median OS against various RLC cutoffs (Fig. 2A) showed that the difference between the catumaxomab and control groups in median OS increased with an RLC > 13%. In addition, an RLC of 13% was the first cutoff at which a substantial difference between the medians was observed, associated with a low P value of 0.0072 (Fig. 2B) for catumaxomab versus the control group and a corresponding HR of 0.5180 (Fig. 2C). These observations were supported by the results of a cutoff point determination method for survival analysis (Table 2). Therefore, 13% was selected as the RLC cutoff for subgroup analysis.
A, median OS in patients treated with catumaxomab and control patients for various RLC cutoff values. B, HRs for comparing treatment effects above/below various RLC cutoff values. C, P values log-rank test comparing treatment effects above/below various RLC cutoff values. Median, P value, and HR above/below cutoff refers to the comparison between the treatment groups (catumaxomab vs. control) for the subgroup of patients above (>)/below or equal (≤) the corresponding RLC cutoff displayed on the x-axis.
Cutoff point determination for RLC by a survival tree analysis according to Mandrekar and colleagues (20)
In the subgroup of patients with an RLC > 13% at screening [n = 159 (64.9%): catumaxomab, n = 100 (63.7%) and control, n = 59 (67.0%)], catumaxomab treatment was associated with a prolonged OS compared with the control group (P = 0.0072; HR, 0.518; 95% CI, 0.318–0.844), with a median (mean) of 109 (209) versus 68 (78) days (Fig. 3A). The 6-month and 1-year survival rates were 37.0% versus 5.2% and 18.5% versus 0.0%, respectively. In patients with an RLC ≤ 13% at screening (n = 74: catumaxomab, n = 50 and control, n = 24), the median (mean) OS of the catumaxomab and the control groups was 52 (76) and 49 (60) days, respectively (P = 0.2561; HR, 0.695; 95% CI, 0.368–1.311; Fig. 3B).
OS in patients with an RLC > 13% (A) and ≤13% (B) at screening (only patients with valid lymphocyte results were included in the analysis). Crossover patients were censored at the time of first catumaxomab infusion.
Cutoff value definition at an RLC of 13% had also an impact on PuFS. Compared with control patients, catumaxomab treatment prolonged PuFS in patients with low RLC (≤13%) and high RLC (>13%) patients. However, catumaxomab patients with an RLC > 13% had a significantly prolonged PuFS compared with patients with RLC ≤ 13% (P = 0.0027, HR, 0.555; Table 3). In control patients, the difference between high and low RLC was not significant (P = 0.265, HR, 0.760; log-rank).
PuFS: comparison of RLC subgroups
KI had a significant positive impact on OS (Table 1). The study inclusion criteria included patients with KI ≥ 60. A total of 135 (86%) patients (86% catumaxomab; 86% control) had KI ≥ 70. In patients with KI ≥ 70, OS was statistically significantly longer in catumaxomab-treated patients than in controls (median, 84 vs. 62 days, respectively; P = 0.0053; HR, 0.567). In patients with KI ≥ 90 (23% of catumaxomab-treated patients, 20% of controls), median OS was 203 days in catumaxomab-treated patients versus 68 days in controls (P = 0.0162; HR, 0.372).
Discussion
RLC was found to be a biomarker with a positive prognostic impact on OS with catumaxomab treatment. In patients with RLC > 13% at screening, catumaxomab treatment was associated with a pronounced beneficial effect on OS versus the control group (P = 0.0072) and a 7-fold higher 6-month survival rate (37.0% vs. 5.2%). This statement is also true regarding PuFS, which represented the primary endpoint in the pivotal phase II/III trial. Here, the cutoff value of RLC > 13% allowed to select patients with prolonged PuFS (median, 49 vs. 31 days, P = 0.0027, log-rank) in the catumaxomab treatment group.
These results were achieved by investigating the relevance of RLC as a potential biomarker for catumaxomab treatment based on a retrospective analysis of an independent, smaller, descriptive study. The results of this hypothesis-generating analysis showed that RLC at screening was a prognostic biomarker, being positively correlated with OS, and that patients with a high RLC (>13%) showed an increased OS after catumaxomab treatment versus patients with an RLC ≤ 13% (mean, 14.5 vs. 7.1 months, respectively; P = 0.04; ref. 17). In addition, our analysis shows that catumaxomab was associated with a statistically significant OS benefit in the safety population of the pivotal phase II/III study versus paracentesis alone (P = 0.0219, HR, 0.649). It should be noted that all the patients who were alive at the first cutoff for OS analysis had an RLC > 13% (8).
The results are consistent with the proposed immunologic mode of action of catumaxomab (3–6), which requires a functional immune system for its antitumor activity. Ott and colleagues showed that there was a strong correlation between early humoral immunologic response to catumaxomab, as shown by the development of human antimouse antibodies (HAMA), and clinical outcome (7). Patients who developed HAMAs after catumaxomab treatment showed significant improvements in the 3 clinical outcome parameters investigated (OS, PuFS, time to next therapeutic paracentesis). This indicates that the ability of a patient to mount a rapid immunologic response, that is, the presence of a functional immune system, is key to catumaxomab's positive effects on OS. Different immunologic mechanisms are induced after intraperitoneal administration of catumaxomab, for example, immediate T-cell–mediated cytotoxicity is involved in the reduction of malignant ascites (22). Complex immunologic mechanisms in addition to the direct and local effects of catumaxomab may be involved in its mode of action (23). Cellular and humoral immune responses to antigens other than EpCAM [e.g. human epidermal growth factor receptor 2 (HER2) or cancer testis antigens] as well as the detection of catumaxomab-induced long-term activated T cells and the expansion of preexisting EpCAM-specific T cells after catumaxomab therapy indicate the initiation of persistent systemic immunologic anti-tumor activity (22, 24).
An increased lymphocyte count has also been shown to be a prognostic factor for OS in patients with gastric and colorectal cancer (25, 26). In addition, lymphocyte subsets have been shown to be predictive markers for immunotherapy in patients with cancer (14, 27). Characiejus and colleagues concluded that the pretreatment level of CD8+ T cells could be a predictive biomarker for the response to cancer immunotherapy (14).
The positive impact of RLC on OS was supported by the sensitivity analysis results for ALC, which also showed a significant positive impact on OS (P = 0.0260). The reason for the stronger signal of RLC versus ALC may be because the RLC includes all peripheral blood cells, including lymphocytes, neutrophils, and monocytes, as it represents the relation of the ALC to all peripheral blood cell populations (17). A number of studies on the neutrophil/lymphocyte ratio (NLR, analogous to the RLC) in different tumor types, including gastric, colorectal, and ovarian, showed that a high NLR (corresponding to a low RLC) is a poor prognostic parameter (28–30). This could result from impaired immunologic function due to a low number of lymphocytes and also from inflammatory processes caused by inhibitory neutrophils that can lead to stimulation of the tumor. Neutrophils secrete vascularization factors, such as VEGF, which promote tumor growth (31). In addition, neutrophils are also known to inhibit the specific immune system (32). An and colleagues therefore suggested that the NLR could represent the balance between protumor inflammatory status and antitumor immune status (33). The findings of the present study that the ANC and AMC have a negative impact on OS support this interpretation. The advantage of using RLC is that it is easy to assess using an established routine diagnostic method. The results of the analysis of the outcome of catumaxomab-treated patients according to their performance status can be regarded as consistent with the RLC results as a functional immune system is generally associated with an adequate performance status. Moreover, as RLC could be shown to be an independent prognostic marker, it can be concluded that both RLC and KI may complement one another with regard to treatment decision making.
In conclusion, an RLC > 13% could be used as a pretherapeutic biomarker for enhanced efficacy of catumaxomab treatment, regarding both PuFS and OS. These benefits in patients with an RLC > 13% indicate that treatment with catumaxomab in this subgroup may also result in effects on disease outcome, in addition to symptom control. Appropriate patient selection is therefore important for achieving a maximum treatment effect with catumaxomab, including a potential survival benefit. Moreover, the RLC may be used to define the optimal timepoint to start catumaxomab treatment in context of multimodal treatment. As RLC is a dynamic and easily accessible marker, not only it allows to select appropriate patients but also optimal time management in context with systemic chemotherapy. Therefore, early use of catumaxomab in selected patients in the course of disease could increase the OS benefit in addition to the symptom relief provided.
Disclosure of Potential Conflicts of Interest
M.A. Ströhlein is a consultant/advisory board member for Fresenius Biotech and TRION Pharma. C. Bokemeyer reports receiving speakers bureau honoraria from and is a consultant/advisory board member for Fresenius Biotech. D. Arnold is a consultant/advisory board member for Fresenius Biotech. D. Seimetz is an employee of Fresenius Biotech. H. Lindhofer is an employee of and has ownership interests (including patents) in TRION Pharma. No potential conflicts of interest were disclosed by the other authors.
Authors' Contributions
Conception and design: M.M. Heiss, M.A. Ströhlein, C. Bokemeyer, S.L. Parsons, H. Lindhofer, E. Schulze, M. Hennig
Development of methodology: M.M. Heiss, M.A. Ströhlein, S.L. Parsons, M. Hennig
Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): M.M. Heiss, M.A. Ströhlein, C. Bokemeyer, D. Arnold, S.L. Parsons, E. Schulze
Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): M.M. Heiss, M.A. Ströhlein, C. Bokemeyer, D. Arnold, S.L. Parsons, D. Seimetz, H. Lindhofer, E. Schulze, M. Hennig
Writing, review, and/or revision of the manuscript: M.M. Heiss, M.A. Ströhlein, C. Bokemeyer, D. Arnold, S.L. Parsons, D. Seimetz, H. Lindhofer, E. Schulze, M. Hennig
Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): D. Seimetz, E. Schulze, M. Hennig
Study supervision: M.A. Ströhlein, M. Hennig
Grant Support
This study was funded by Neovii (formerly Fresenius Biotech GmbH, Munich, Germany.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Acknowledgments
The authors thank their fellow investigators who participated in the pivotal phase II/III study: P. Murawa (Wielkoposka Cancer Center, Poznan, Poland); P. Koralewski (Rydygier Memorial Hospital, Krakow, Poland); E. Kutarska (Center of Oncology of Lublin, Lublin, Poland); O.O. Kolesnik (Institute of Oncology, Academy of Medical Science of Ukraine, Kiev, Ukraine); V.V. Ivanchenko (Regional Clinical Oncology Dispensary, Velikiy Novgorod, Russia); A.S. Dudnichenko (Kharkov Medical Academy of Postgraduate Education, Kharkov, Ukraine); B. Aleknaviciene (Vilnius University, Institute of Oncology, Vilnius, Lithuania); A. Razbadauskas (Klaipeda Seamen's Hospital, Klaipeda, Lithuania); M. Gore (Royal Marsden Hospital, London, UK); E. Ganea-Motan (Spitalul Judetean de Urgenta “Sf Ioan cel Nou,” Suceava, Romania); T. Ciuleanu (Cancer Institute Ion Chiricuta, Cluj-Napoca, Romania); P. Wimberger (University of Duisburg-Essen, Essen, Germany); A. Schmittel (Charité, University Hospital Berlin, Berlin, Germany); B. Schmalfeldt (Technical University Munich, Munich, Germany); A. Burges (University Hospital Grosshadern, Munich, Germany); A. Adenis (Centre Oscar Lambret, Lille, France); M. Bębenek (Lower Silesian Oncology Center, Wrocław, Poland); M. Bidziński (Memorial Cancer Center, Warsaw, Poland); M. Bitina (Daugavpils Oncological Hospital, Daugavpils, Latvia); M. Błasińska-Morawiec (N. Copernicus Memorial Hospital, Lodz, Poland); A. Blidaru (Oncology Institute “Prof. Dr Alexandru Trestioreanu,” Bucharest, Romania); B. Bolyukh (Pirogov Vinnytsya National Medical University, Vinnytsya, Ukraine); A. Croitoru (Clinical Institute Fundeni, Bucharest, Romania); H. Curé (Centre Jean Perrin, Clermont-Ferrand, France); S. Curescu (Emergency Municipal Clinical Hospital, Timisoara, Romania); G.C. de Gast (The Netherlands Cancer Institute, Anton van Leeuwenhoek Hospital, Amsterdam, The Netherlands); P. Dufour (Centre Paul Strauss, Strasbourg, France); V.A. Gorbounova (Russian Cancer Research Centre, Moscow, Russia); R. Greil (Paracelsius Medical University, State Hospital Salzburg, Salzburg, Austria); P. Harper (Guy's and St Thomas' Hospital, London, UK); Y. Hotko (Uzhgorod National University, Uzhgorod, Ukraine); E. Jaeger (Hospital Northwest, Medical Department II, Frankfurt, Germany); E. Juozaityte (Kaunas Medical University Hospital, Kaunas, Lithuania); E. Kettner (City Hospital, Magdeburg, Germany); P. Kier (Danube Hospital, Vienna, Austria); T.D.V. Komov (N.N. Blokhin Russian Cancer Research Center, Moscow, Russia); I.Y. Kostynskyy (Ivano-Frankivsk State Medical University, Ivano-Frankivsk, Ukraine); O. Kovalyov (Zaporizhzhya Oblast Clinical Oncology Dispensary, Zaporizhzhya, Ukraine); L.I. Krikunova (Medical Radiology Research Center RAMN, Obninsk, Russia); T. Kühn (General Hospital Gifhorn, Gifhorn, Germany); K. Kukk (NERH Surgery Clinic, Tallinn, Estonia); J-E. Kurtz (Hospital Hautepierre, Strasbourg, France); M. Kuta (Hospital Chomutov, Chomutov, Czech Republic); S.A. Lazarev (Altay Regional Oncology Dispensary, Barnaul, Russia); K. Lesniewski-Kmak (PCK Maritime Hospital, Gdynia, Poland); F. Lordick (Hospital of the Technical University Munich, Munich, Germany); A. Makhson (Moscow Municipal City Hospital #62, Moscow, Russia); C. Marth (Innsbruck Medical University, Innsbruck, Austria); E. Petru (University of Graz, Graz, Austria); L. Petruzelka (General University Hospital, Prague, Czech Republic); P. Piso (University Hospital Regensburg, Regensburg, Germany); C. Poole (Sandwell and West Birmingham NHS Trust, City Hospital, Birmingham, UK); G. Purkalne (P. Stradins University Hospital, Riga, Latvia); W. Schmidt (Saarland University Hospital, Homburg/Saar, Germany); L. Schulz (Hospital Garmisch-Partenkirchen, Garmisch-Partenkirchen, Germany); J. Sehouli (Charité Campus Virchow Klinikum, Berlin, Germany); I. Shchepotin (Kiev City Oncology Hospital, Kiev, Ukraine); T. Shchetinina (Lugansk Oblast Clinical Oncology Dispensary, Lugansk, Ukraine); S.V. Sidorov (City Hospital #1, Novosibirsk, Russia); H.P. Sleeboom (Leyenburg Hospital, Den Haag, The Netherlands); M. Spaczyński (Poznan University of Medical Sciences, Poznan, Poland); V. Stahalova (Na Bulovce Teaching Hospital, Prague, Czech Republic); W. Stummvoll (Hospital Barmherzige Schwestern Linz, Linz, Austria); J. Thaler (Klinikum Kreuzschwestern Wels, Wels, Austria); B. Utracka-Hutka (Maria Sklodowska-Curie Memorial Institute, Gilwice, Poland); I. Vaasna (Tartu University Clinics, Tartu, Estonia); C. Volovat (Center of Medical Oncology, Iasi, Romania); G. Wallner (Medical University of Lublin, Lublin, Poland); J. Waters (Kent Oncology Centre, Maidstone Hospital, Kent, UK); J. Wolf (University Hospital Cologne, Cologne, Germany); I. Zhulkevych (Ternopil Oblast Communal Clinical Oncology Dispensary, Ternopil, Ukraine); and Z. Zvirbule (Latvian Oncology Center, Riga, Latvia); and the study personnel at Neovii (formerly Fresenius) Biotech GmbH: A. Lahr and H. Friccius. The authors would also like to thank the freelance medical writer Kevin De-Voy for his writing support.
- Received September 11, 2013.
- Revision received February 8, 2014.
- Accepted February 17, 2014.
- ©2014 American Association for Cancer Research.
References
Volume 20, Issue 12
