|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Originally published as JCO Early Release 10.1200/JCO.2007.12.1152 on November 5 2007 © 2007 American Society of Clinical Oncology. Computed Tomography Scans Do Not Improve the Predictive Power of 1996 National Cancer Institute–Sponsored Working Group Chronic Lymphocytic Leukemia Response Criteria
From the Division of Hematology-Oncology, Department of Internal Medicine; Biostatistical Core, The Arthur G. James Comprehensive Cancer Center; and the Division of Medicinal Chemistry and Pharmacology, College of Pharmacy, The Ohio State University, Columbus, OH Address reprint requests to Kristie A. Blum, MD, The Ohio State University, Division of Hematology-Oncology, Starling Loving Hall, Room B324, Columbus, OH 43210; e-mail: kristie.blum{at}osumc.edu
Purpose National Cancer Institute–sponsored Working Group (NCI-WG) response criteria for chronic lymphocytic leukemia (CLL) rely on physical examination, blood, and bone marrow evaluations. The widespread use of computed tomography (CT) scans has prompted many to advocate for the incorporation of this test into CLL response criteria. Patients and Methods In a retrospective review of 82 CLL patients treated at the Ohio State University (Columbus, OH), we compared CT assessed response using non-Hodgkin's lymphoma (NHL) response definitions with NCI-WG response. Results Responses by NCI-WG criteria included five complete responses (CRs), 32 partial responses (PRs), 21 patients with stable disease (SD), 17 patients with progressive disease (PD), and seven patients not assessable (NA). Responses by NHL-CT criteria included three CRs, 12 unconfirmed CRs (CRus), 16 PRs, 26 with SD, four with PD, and 21 NA. Using NCI-WG criteria, progression-free survival (PFS) was 27.3 months for CR and 11.4 months for PR. With NHL-CT criteria, PFS was 18.4 months for CR, 11.7 months for CRu, and 14.5 months for PR. In multivariate analysis, both NCI-WG and NHL-CT response correlated with PFS (P = .009 and .001, respectively). Conclusion Current NCI-WG CLL response criteria are a significant predictor of PFS in previously treated CLL patients, with no additional benefit from the inclusion of CT scans. Although retrospective, these results highlight the importance of prospective validation of CT scans before routine inclusion in CLL response criteria.
Chronic lymphocytic leukemia (CLL) is the most common adult leukemia occurring in the Western hemisphere, with an annual age-adjusted incidence of 3.8 per 100,000 men and women per year.1 Several effective therapies exist including fludarabine, rituximab, alkylating agents, and alemtuzumab. Despite the wealth of therapeutic options, no curative therapies, with the potential exception of allogeneic stem-cell transplantation, exist. However, a variety of novel therapies including monoclonal antibodies, immunomodulatory agents, and cyclin-dependent kinase inhibitors are under evaluation.2-7 Critical assessment of response and survival outcomes is required for the success of rapid drug development in this disease. These assessments rely on internationally accepted response criteria, justified by their association with meaningful patient outcomes, namely progression-free survival (PFS) and overall survival (OS). Addition of new diagnostic tests, particularly if invasive or expensive, should undergo evaluation before routine inclusion in response criteria. In 1996, the National Cancer Institute–sponsored Working Group (NCI-WG) criteria for CLL were revised and updated to provide practice and research guidelines for diagnosis, treatment, and response evaluation in CLL.8 These criteria have served as accepted response criteria in CLL for 10 years, providing a standardized framework for international clinical trials in regard to eligibility, response, and toxicity assessments. In accordance with the 1996 NCI-WG criteria, response assessments require a CBC and physical examination, with a bone marrow biopsy at the time of complete remission.8 Computed tomography (CT) scans are recommended only when indicated by patient symptoms. However, with improvements in CLL therapy9-13 and radiographic imaging techniques, incorporation of CT scans into updated CLL response criteria has been suggested. Although CT scans are utilized in non-Hodgkin's lymphoma (NHL) response criteria,14 the benefit of CT scan–measured lymph node response in predicting patient outcomes in CLL is unclear. To address the potential importance of CT scans in CLL, we therefore performed a retrospective review of three recently completed therapeutic trials conducted at the Ohio State University (OSU; Columbus, OH) where response by NCI-WG criteria was compared with response using CT scan–confirmed lymph node dimensions. Because current NCI-WG criteria do not define changes in lymph node dimensions required for complete response (CR), partial response (PR), stable disease (SD), and progressive disease (PD), response based on lymph node size was assessed in pre- and post-treatment CT scans using international response criteria for NHL.14 The results described herein fail to justify the routine use of CT scans for assessment of response in CLL.
Response was assessed in 82 patients (eight previously untreated and 74 previously treated) with NCI-WG–defined CLL8 or small lymphocytic lymphoma (SLL)15 treated at OSU from April 2002 to November 2006 using both 1996 CLL NCI-WG8 and international NHL14 response criteria. From April 2002 to September 2002, 14 patients with relapsed CLL were enrolled onto OSU 0049, a phase I trial of combination rituximab and etanercept. CT scans were required before treatment and at the completion of study therapy. OSU 0055, a phase I trial of single-agent flavopiridol open from April 2003 to October 2006, enrolled 58 patients with previously treated CLL.7 Although not specifically required by the study, patients underwent CT scans pretreatment, every two cycles, and at the time of study withdrawal or completion. OSU 0211, a phase I study of flavopiridol, fludarabine, and rituximab in patients with indolent B-cell lymphoproliferative disorders, enrolled a total of 10 patients with CLL/SLL from April 2003 to November 2006. Eight patients had received no prior therapies, and two were treated previously. In OSU 0211, CT scans were required pretreatment, after cycles 3 and 6, and at the end of study therapy. OSU 0049, 0055, and 0211 were approved by the OSU Institutional Review Board, and all patients provided written informed consent. Because of the differing intervals for CT scans on these trials, only pretreatment and end-of-therapy CT scans were used to determine response. All patients were treated at OSU and had their CT scans performed and reviewed at OSU, but there was no formal central radiographic review. The 1996 NCI-WG response evaluations were based on peripheral blood counts and physical examination findings, with bone marrow assessments performed at least 2 months after clinical and laboratory findings demonstrated CR.8 Specifically, a CR was defined as the absence of lymphadenopathy and organomegaly by physical examination, absolute lymphocyte count (ALC) no higher than 5,000/µL, absolute neutrophil count (ANC) of at least 1,500/µL, platelets more than 100,000/µL, hemoglobin more than 11.0 g/dL, and bone marrow without lymphoid nodules and fewer than 30% lymphocytes. A PR required at least a 50% decrease in ALC, lymphadenopathy, and organomegaly from pretreatment evaluation, and an ANC of at least 1,500/µL, platelets more than 100,000/µL, and hemoglobin more than 11.0 g/dL, or at least a 50% improvement from baseline in these values. PD was defined as an increase of at least 50% in either the product of measured lymph node dimensions, organomegaly, or ALC; the presence of new palpable nodes; or transformation. Responses less than a PR, but without evidence of progression, were classified as SD. Because historical response criteria for CLL8 do not require CT scans and do not incorporate lymph node or tumor dimensions into response definitions, International Working Group NHL criteria14 were utilized to define response on CT scan. According to these criteria,14 all lymph nodes or tumor masses must have regressed to no larger than 1.5 cm if more than 1.5 cm before therapy, or no larger than 1 cm if 1.1 to 1.5 cm at the start of therapy, with resolution of organomegaly, to qualify for a CR. A residual nodal mass larger than 1.5 cm with a more than 75% decrease in the sum of the products was defined as unconfirmed CR (CRu). A decrease of 50% or more in the sum of the products of the diameters of the six largest nodal masses from different regions of the body was required for a PR. PD was defined as an increase of 50% or more from nadir in the sum of the products of the diameters of any previously identified node or the appearance of new lesions, and SD was defined as less than a PR with no evidence of progression.
Response was determined in all patients using NCI-WG CLL criteria8 and NHL-CT criteria.14 Patients whose response could not be assessed by NCI-WG CLL because of early treatment related toxicity or by NHL-CT criteria because of lack of a post-treatment CT scan were not assessable (NA) in this analysis. PFS and OS were estimated from the time of study entry until time of progression or death using the Kaplan-Meier method. The log-rank test was used to assess the correlation of 1996 NCI-WG CLL response designation (CR, PR, SD, or PD) and NHL-CT response designation (CR, CRu, PR, SD, and PD) with PFS and OS. Univariate (log-rank test) and multivariate analysis (Cox proportional hazards regression model) were used to calculate the association of patient age, sex, Rai stage, number of prior therapies, pretreatment β2-microglobulin level, the presence of bulky lymphadenopathy (nodal mass either
Patient Characteristics The median age of the 82 CLL/SLL patients receiving treatment on OSU 0049, OSU 0055, and OSU 0211 was 60 years (range, 38 to 84 years). Seventy-eight patients had CLL, four patients were classified as SLL, and 60 patients were male. The majority of patients had Rai stage 3 to 4 disease and had received a median number of four prior therapies (range, 0 to 14). In 73 patients with available data, bulky lymphadenopathy of at least 5 cm in size was present in 54.9%, and 18.3% had at least one node larger than 10 cm. Pretreatment β2-microglobulin levels were elevated in 82.9% of patients. Additional pretreatment characteristics are provided in Table 1.
Response to Therapy Post-treatment response according to both NCI-WG CLL and NHL-CT criteria could not be determined in 23 of 82 patients because of early treatment-related toxicities or lack of a post-treatment CT scan. Therefore, only 59 patients were assessable by both NCI-WG CLL and NHL-CT criteria in this analysis. Response according to NCI-WG CLL criteria was as follows: CR, n = 5 (6.1%); PR, n = 32 (39%); SD, n = 21 (25.6%); PD, n = 17 (20.7%); and NA, n = 7 (8.5%). Response according to NHL-CT criteria was as follows: CR, n = 3 (3.7%); CRu, n = 12 (14.6%); PR, n = 16 (19.5%); SD, n = 26 (31.7%); PD, n = 4 (4.9%); and NA, n = 21 (25.6%).
Table 2 depicts the discrepancies in response assessment using NCI-WG criteria and NHL-CT criteria in the 59 assessable patients. The majority of patients classified as CRu by NHL-CT criteria achieved a PR by NCI-WG criteria (eight of 12 responders). More patients were classified as a PR or PD using NCI-WG CLL criteria than by NHL-CT criteria, whereas fewer instances of SD were observed using conventional CLL criteria compared with NHL-CT criteria. However, by McNemar
PFS and OS In 82 treated CLL/SLL patients, the median PFS and OS were 9.0 months (95% CI, 7.7 to 10.3 months; Fig A1A, online only) and 25.2 months (95% CI, 13.2 to 37.2 months; Fig A1B), respectively, with 55 of 82 patients experiencing disease progression, and 17 deaths. In the 75 patients assessable by NCI-WG CLL criteria, CR, PR, SD, and PD response designations correlated with PFS by log-rank test (P < .001, Fig 1A), with a median PFS of 27.3 months (95% CI, 13.0 to 41.6 months) for CR, 11.4 months (95% CI, 9.9 to 12.9 months) for PR, 5.6 months (95% CI, 0 to 11.5 months) for SD, and 2.5 months (95% CI 1.7 to 3.4 months) for PD. NCI-WG CLL response also significantly correlated with OS (P = .012, Fig 1B). For the 61 patients assessable by NHL-CT criteria, response designation also correlated with PFS (P < .001, Fig 2A) and OS (P < .001, Fig 2B). The median PFS by NHL-CT response was 18.4 months (95% CI, not computed with only three events) for CR, 11.7 months (95% CI, 8.4 to 14.9 months) for CRu, 14.5 months (95% CI, 8.7 to 20.2 months) for PR, 5.2 months (95% CI, 4.5 to 5.9 months) for SD, and 2.0 months (95% CI, 1.1 to 2.9 months) for PD. Achievement of CRu did not appear to prolong PFS compared with PR.
Because there is no validated tool for measuring lymph nodes in CLL as exists for NHL, alternative response definitions based on lymph node dimensions were examined in addition to NHL-CT response. If the definition of PR is loosened to require a 25% or greater decrease in the sum of the products of the diameters of measured lymph nodes, with PD requiring an increase of at least 25% in the sum of the products of the diameters of the measured nodes, little change is observed in the ability of these response designations to predict PFS compared with NCI-WG or NHL-CT response (Fig A2, online only).
Predictive Power of Patient Response on Survival End Points By multivariate Cox regression analysis, age, sex, number of prior therapies, Rai stage, β2-microglobulin, bulky lymphadenopathy, NCI-WG CLL response, and NHL-CT response failed to correlate with OS. However, response by NCI-WG CLL (CR and PR) or NHL-CT (CR, CRu, and PR) criteria were significantly associated with PFS (P = .009 and P = .001, respectively, Table 3).
The development of standardized response criteria are critical for the evaluation of novel therapeutic approaches in CLL, providing a uniform method of reporting response and evaluating outcomes across national and international clinical trials. Radiographic and laboratory assessments of response should be evaluated rigorously in the context of older accepted response indicators to determine their added benefit in the prediction of patient outcomes, particularly because these new methods are often time-consuming and costly. This comparative analysis of 1996 NCI-WG CLL and CT scan response in previously treated CLL patients demonstrates no advantage of the addition of CT scans to conventional physical examination, peripheral blood, and bone marrow evaluations in regard to patient PFS. In this retrospective study of 82 CLL/SLL patients enrolled onto three different OSU treatment protocols, both NCI-WG and NHL-CT response designation (CR, CRu, PR, SD, or PD; Figs 1A and 2A) and the achievement of a response (CR, CRu, and PR; Tables 3 and A1) correlate with PFS in univariate and multivariate analysis. Because both NCI-WG and NHL-CT response are predictive of PFS, CT scan–measured lymph node size is not necessary to improve the predictive power of traditional CLL response criteria. Therefore, on the basis of these results, the routine incorporation of CT-defined response into standard CLL response criteria is not justified, particularly in light of the added expense, patient inconvenience, and potential risks (ie, renal impairment or allergic reaction) that these studies entail. Although this study primarily examined changes in lymph node dimensions according to response definitions used in NHL trials,14 it is unclear whether these definitions are valid in CLL. Perhaps, the use of other measurement criteria, such as Response Evaluation Criteria in Solid Tumors (RECIST) definitions,16 may be more effective in CLL. As part of this analysis, we did explore an alternative system permitting 25% or greater improvement in the sum of the products of the diameter for PR, and this failed to demonstrate any noticeable improvement in predicted PFS for patients achieving CR, PR, or SD over the currently accepted NCI-WG CLL criteria. Also, in this analysis, the median PFS for patients achieving a CRu failed to prolong PFS when compared with PFS for a PR by NCI-WG or NHL-CT definitions, suggesting that this response category is not meaningful in CLL patients. Several recent studies suggest that the use of CT scans reduces routine overestimation of CR in CLL by NCI-WG criteria, and that CT-confirmed CR correlates with a longer response duration than CR by NCI-WG criteria.17-19,21 In one study, intra-abdominal adenopathy on CT correlated with a shortened time to disease progression and time to first treatment.17 Because the study evaluated CT scans as a pretreatment prognostic factor in newly diagnosed patients, these results cannot be used to justify altering existing therapeutic response criteria. However, an analysis of CT scans in patients receiving fludarabine, cyclophosphamide, and mitoxantrone as initial therapy demonstrated that residual abdominal adenopathy on CT scan shortened response duration in patients with an NCI-WG CR.19 Unlike the results in our study, 64% patients achieved CR. With the limited CR rate in our study, it is difficult to directly compare NCI-WG CR with NHL-CT CR. However, both NCI-WG and NHL-CT adequately differentiate survival outcomes in patients achieving CR and PR, and CT does not appear to improve on this. Also, given that most novel agents are first evaluated in the relapsed setting, where CRs are infrequent,13,20 the large proportion of previously treated patients in our study adds to the worth of our retrospective findings. Two studies suggest that NCI-WG CLL criteria overestimate CR and nodular PR in untreated and relapsed patients receiving fludarabine-containing regimens.18,20,21 These studies, however, do not correlate response with PFS, and as demonstrated in our study, strict definitions of response based on CT findings may not ultimately influence PFS. Specifically, in our study, although NCI-WG criteria overestimated CR, PFS was similar and perhaps favored NCI-WG criteria (n = 5; median PFS, 27.3 months), compared with NHL-CT criteria (n = 3; median PFS, 18.4 months). Relative to patient benefit, PFS represents a more rigorous end point and should serve as the focus of future prospective evaluations of CT scans in patients with CLL. This study was designed to examine the added value of CT scans in CLL response assessment as a result of recent interest in modifying these guidelines, and is therefore limited by its retrospective design. Specifically, CT scans were not consistently performed at similar time points across all three trials and the post-treatment CT scan may not have occurred at the time of best response assessed by NCI-WG CLL criteria. However, the median response duration for the responding patients on OSU 0055 (70% of the patients evaluated in this study) exceeds 12 months,7 and in all likelihood, the post-treatment CT reflects this response. Also, although treatment was not uniform in this population, the widely differing therapies including immunotherapy (rituximab and etanercept), chemotherapy (flavopiridol), and combination chemoimmunotherapy (rituximab, flavopiridol, and fludarabine) strengthen the findings. Because response criteria must accurately assess response to treatments ranging from chemotherapy to immunomodulatory drugs, the value of CT scans should be examined across multiple therapeutic CLL trials. The results of this study highlight the importance of prospective validation of CT scans before routine inclusion into response criteria for CLL. Future evaluations should occur in a large cohort of untreated and relapsed patients from multiple institutions, with central radiographic review and cost analysis to determine whether the added cost of these tests is offset by improvements in measurements of clinical benefit, including PFS and OS. Until such studies are performed, the findings of this analysis do not demonstrate any benefit for the routine incorporation of CT scans into the care and response evaluation of relapsed CLL patients, without clinical indication, despite the impression that CT measurements provide a more objective method of response assessment than physical examination, blood, and bone marrow findings.
The author(s) indicated no potential conflicts of interest.
Conception and design: Kristie A. Blum, Michael R. Grever, John C. Byrd Administrative support: Margaret S. Lucas, Beth Fischer Provision of study materials or patients: Margaret S. Lucas, Beth Fischer, Thomas S. Lin, John C. Byrd Collection and assembly of data: Kristie A. Blum, Sarah Broering, Margaret S. Lucas, Beth Fischer Data analysis and interpretation: Kristie A. Blum, Donn Young Manuscript writing: Kristie A. Blum, Donn Young, John C. Byrd Final approval of manuscript: Kristie A. Blum, Donn Young, Sarah Broering, Margaret S. Lucas, Beth Fischer, Thomas S. Lin, Michael R. Grever, John C. Byrd
published online ahead of print at www.jco.org on November 5, 2007. Supported by Grants No. K23 CA109004-01A1, U01 CA 76576, R21 CA112947-01A1, National Cancer Institute P01 CA95426, Specialized Center for Research from the Leukemia and Lymphoma Society P30 CA16058, and the D. Warren Brown Foundation. Presented in part as a poster presentation at the 48th Annual Meeting of the American Society of Hematology, December 10, 2007, Orlando, FL. Authors' disclosures of potential conflicts of interest and author contributions are found at the end of this article.
1. Reis L, Harkins D, Krapcho M, et al: SEER Cancer Statistics Review, 1975-2003. Bethesda, MD, National Cancer Institute, 2005 2. Chanan-Khan A, Miller KC, Musial L, et al: Clinical efficacy of lenalidomide in patients with relapsed or refractory chronic lymphocytic leukemia: Results of a phase II study. J Clin Oncol 24:5343-5349, 2006 3. Cherukuri A, Kadel E, Lee SH, et al: Human antagonistic anti-CD40 antibody, CHIR-12.12, inhibits CD40-mediated survival of chronic lymphocytic leukemia B cells leading to cellular apoptosis. Proc Am Soc Hematol 106:832, 2005 (abstr 2965) 4. Kater AP, Dicker F, Fukuda T, et al: CD40-activation of chronic lymphocytic leukemia cells induces latent sensitivity to Fas/TRAIL-mediated apoptosis via a P53-independent, C-Abl-dependent pathway. Proc Am Soc Hematol 104:101, 2004 (abstr 342) 5. Tong X, Georgakis GV, Li L, et al: In vitro activity of a novel fully human anti-CD40 antibody CHIR-12.12 in chronic lymphocytic leukemia: Blockade of CD40 activation and induction of ADCC. Proc Am Soc Hematol 104:686, 2004 (abstr 2504) 6. Byrd JC, O'Brien S, Flinn I, et al: Safety and efficacy results from a phase I trial of single-agent lumiliximab (anti-CD23 antibody) for chronic lymphocytic leukemia. Proc Am Soc Hematol 104:686, 2004 (abstr 2503) 7. Byrd JC, Lin TS, Dalton JT, et al: Flavopiridol administered using a pharmacologically derived schedule is associated with marked clinical efficacy in refractory, genetically high-risk chronic lymphocytic leukemia. Blood 109:399-404, 2007 8. Cheson B, Bennett J, Grever M, et al: National Cancer Institute-sponsored working group guidelines for chronic lymphocytic leukemia: Revised guidelines for diagnosis and treatment. Blood 87:4990-4997, 1996 9. Byrd JC, Peterson BL, Morrison VA, et al: Randomized phase 2 study of fludarabine with concurrent versus sequential treatment with rituximab in symptomatic, untreated patients with B-cell chronic lymphocytic leukemia: Results from Cancer and Leukemia Group B 9712 (CALGB 9712). Blood 101:6-14, 2003 10. Byrd JC, Rai K, Peterson BL, et al: Addition of rituximab to fludarabine may prolong progression-free survival and overall survival in patients with previously untreated chronic lymphocytic leukemia: An updated retrospective comparative analysis of CALGB 9712 and CALGB 9011. Blood 105:49-53, 2005 11. Keating MJ, O'Brien S, Albitar M, et al: Early results of a chemoimmunotherapy regimen of fludarabine, cyclophosphamide, and rituximab as initial therapy for chronic lymphocytic leukemia. J Clin Oncol 23:4079-4088, 2005 12. Montillo M, Tedeschi A, Miqueleiz S, et al: Alemtuzumab as consolidation after a response to fludarabine is effective in purging residual disease in patients with chronic lymphocytic leukemia. J Clin Oncol 24:2337-2342, 2006 13. Wierda W, O'Brien S, Wen S, et al: Chemoimmunotherapy with fludarabine, cyclophosphamide, and rituximab for relapsed and refractory chronic lymphocytic leukemia. J Clin Oncol 23:4070-4078, 2005 14. Cheson BD, Horning SJ, Coiffier B, et al: Report of an international workshop to standardize response criteria for non-Hodgkin's lymphomas: NCI Sponsored International Working Group. J Clin Oncol 17:1244, 1999 15. Harris NL, Jaffe ES, Diebold J, et al: World Health Organization classification of neoplastic diseases of the hematopoietic and lymphoid tissues: Report of the Clinical Advisory Committee Meeting Airlie House, Virginia, November 1997. J Clin Oncol 17:3835-3849, 1999 16. Therasse P, Arbuck SG, Eisenhauer EA, et al: New guidelines to evaluate the response to treatment in solid tumors. J Natl Cancer Inst 92:205-216, 2000 17. Muntañola A, Bosch F, Arguis P, et al: Abdominal computed tomography predicts progression in patients with Rai stage 0 chronic lymphocytic leukemia. J Clin Oncol 25:1576-1580, 2007 18. Maslak P, Caravelli J, Chanan-Khan AA: "Complete" remission in chronic lymphocytic leukemia (CLL) is strongly affected by the use of CT scanning: Results from a prospective randomized trial of fludarabine plus cyclophosphamide (Flu/Cy) with or without genasense (Oblimersen) in patients with relapsed or refractory disease. Proc Am Soc Hematol 108:802, 2006 (abstr 2834) 19. Bosch F, Muntanola A, Ferrer A, et al: Computed tomography (CT) predicts response duration in patients with chronic lymphocytic leukemia (CLL) in complete response (CR) by conventional, NCI-working group, criteria. Proc Am Soc Hematol 108:802, 2006 (abstr 2835) 20. O'Brien SM, Kantarjian HM, Cortes J, et al: Results of the fludarabine and cyclophosphamide combination regimen in chronic lymphocytic leukemia. J Clin Oncol 19:1414-1420, 2001 21. Eichhorst BF, Busch R, Hopfinger G, et al: Fludarabine plus cyclophosphamide versus fludarabine alone in first-line therapy of younger patients with chronic lymphocytic leukemia. Blood 107:885-891, 2006 Submitted April 11, 2007; accepted August 1, 2007.
Related Editorial
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2007 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
5 or 
2 test, these differences were not significant (P = .227). 
