Patterns of Treatment Failure in Pediatric and Young Adult Patients With Hodgkin's Disease: Local Disease Control With Combined-Modality The
http://www.100md.com
《临床肿瘤学》
the Departments of Radiological Sciences, Biostatistics, and Hematology-Oncology, St Jude Children's Research Hospital, Memphis, TN
ABSTRACT
PURPOSE: Refinement in managing pediatric Hodgkin's disease (HD) requires understanding of factors associated with local treatment failure. This study defines the cumulative incidence (CI) of local failure (LF) and prognostic factors for pediatric patients treated for HD with combined-modality therapy (CMT).
PATIENTS AND METHODS: We enrolled 195 patients onto two prospective studies at St Jude Children's Research Hospital between 1990 and 2000. Patients received CMT with chemotherapy (vinblastine, doxorubicin, methotrexate, and prednisone [VAMP]; vinblastine, etoposide, prednisone, and doxorubicin; or VAMP/cyclophosphamide, vincristine, and procarbazine) and involved-field radiation therapy delivered to initial site(s) of disease on the basis of early response. Sites of disease involvement, treatment, and sites of failure were confirmed from the patients' medical record, imaging, and radiotherapy treatment records. We estimated the overall survival, event-free survival, and CI of LF.
RESULTS: With a median follow-up of 7.6 years, the CI of LF was 10.9% and 11.6% at 5 and 10 years, respectively. Twenty-seven (14%) of 195 patients experienced recurrence of HD, and 22 (81%) of those experienced LF. Bulky mediastinal disease greater than one third transthoracic diameter predicted a higher incidence of LF, but did not predict failure in the mediastinum. Male sex, low initial hemoglobin, and bulky mediastinal disease were prognostic indicators of LF. Attenuation of radiation dose to 15 Gy based on response provides excellent infield control.
CONCLUSION: CMT provides excellent local disease control in children and young adults with HD. LF remains a primary site of disease recurrence, with male sex, low initial hemoglobin, and bulky mediastinal disease predicting for LF.
INTRODUCTION
The management of pediatric Hodgkin's disease has evolved over the past 20 years from high-dose extensive nodal irradiation or aggressive multiagent systemic chemotherapy to combined-modality therapy.1-6 This shift has taken place because standard-dose irradiation and alkylator-based chemotherapy have resulted in significant treatment effects, which can often be attenuated by combining more limited chemotherapy with reduced-dose limited-volume radiation therapy.7-12
Since the early 1990s, we have carried out two sequential risk-adapted studies in collaboration with other institutions. In these studies, combined-modality therapy was adapted on the basis of the extent of disease, prognostic factors, and response to therapy.9,11-13 To continue to refine therapy for both early- and advanced-stage Hodgkin's disease, we have investigated the sites and patterns of treatment failure to further evaluate the effectiveness of current therapies and plan future approaches.
In this report, we describe the local disease outcome and patterns of treatment failure observed in pediatric and young adult patients with Hodgkin's disease treated on two large prospective trials at St Jude Children's Research Hospital.
PATIENTS AND METHODS
HOD90 and HOD94 Study Design
The St Jude Children's Research Hospital HOD90 study opened in January 1990. This study enrolled 76 institutional patients into two treatment strata based on the risk classification outlined below. The study closed in 1993 because it met its stopping criteria in the high-risk strata.11 The second study, HOD94, opened in October 1993 and closed in February 2000.9,12 This study enrolled 119 patients based on the same two treatment strata used in HOD90. These protocols were approved by the institutional review board at St Jude Children's Research Hospital and concurrently opened at collaborating institutions. In total, this risk-based approach was used to treat Hodgkin's disease in 195 patients at St Jude Children's Research Hospital. The event-free outcome for these patients has previously been reported with the collaborating institutions as part of protocol-specific reports.9,11,12 Only patients accessioned to the studies at St Jude Children's Research Hospital were included in this analysis of patterns of failure based on the detailed classification of nodal site involvement, dose-specific irradiation of defined nodal regions, and extensive imaging correlations of sites of failure with initial sites of disease involvement and radiation treatment.
All patients enrolled on HOD90 or HOD94 underwent initial staging evaluations that included a biopsy-proven diagnosis of Hodgkin's disease, a complete initial clinical history, and a physical examination in which sites and extent of nodal disease and splenic involvement were documented. Laboratory evaluations included a complete blood cell count and chemistry panel. Imaging studies, which have evolved over time, included chest x-ray; gallium scan; computed tomography (CT) scan of the chest, abdomen, and pelvis; and lymphangiogram.
After two cycles of chemotherapy, disease response was evaluated by clinical examination, nuclear imaging, and CT scan. Complete response was defined as resolution of all tumor-related constitutional symptoms and disappearance or normalization of all measurable nodal or nonmeasurable but assessable disease. Partial response was defined as at least 50% reduction in the sum of the products of two perpendicular diameters of all measurable lesions and disappearance of constitutional symptoms if initially present. For purposes of this analysis, mixed response was defined as a complete response in some nodal regions, but only a partial response in other nodal or non-nodal sites of disease. Biopsy confirmation of disease progression or relapse was required. Follow-up was conducted every 3 to 6 months after completion of therapy, and cases were coded for any site of treatment failure, secondary malignancy, or other causes of death.
Risk Classification
Patients with high-risk or unfavorable disease included those with Ann Arbor stage III or IV disease and patients with stage I or II disease who had bulky lymphadenopathy. (Bulky lymphadenopathy was defined as either a mediastinal mass of > one third of the transthoracic diameter on an upright chest radiograph or a peripheral lymph node of > 6 cm in maximal axial diameter.) The protocol was amended in July 1998 to include all patients with "B" symptoms in the high-risk group. All other patients were considered as having favorable disease and assigned to the low-risk group (ie, stages IA to IIA without bulky disease).
Systemic Therapy
Treatment included four to six cycles of chemotherapy and involved-field radiation therapy; the prescribed radiation dose was based on tumor response after the second course of chemotherapy. Patients in the low-risk group received vinblastine, doxorubicin, methotrexate, and prednisone (VAMP) chemotherapy for four cycles on both the HOD90 and HOD94 studies. VAMP chemotherapy was administered as follows: vinblastine 6 mg/m2 intravenously (IV) on days 1 and 15, doxorubicin 25 mg/m2 IV on days 1 and 15, methotrexate 20 mg/m2 IV on days 1 and 15, and prednisone 40 mg/m2 per day orally on days 1 through 14.
Systemic therapy for patients in the high-risk group on the HOD90 study included vinblastine, etoposide, prednisone, and doxorubicin (VEPA) for six cycles administered as follows: vinblastine 6 mg/m2 IV on days 1 and 15, etoposide 200 mg/m2 IV on days 1 and 15, prednisone 40 mg/m2 per day PO on days 1 through 14 (omitted after mediastinal radiation therapy), and doxorubicin 25 mg/m2 IV on days 1 and 15. Patients with high-risk disease treated on HOD94 received three cycles of VAMP chemotherapy, as described above, alternated with three cycles of cyclophosphamide, vincristine, and procarbazine (COP) chemotherapy. COP chemotherapy was administered as follows: cyclophosphamide 600 mg/m2 IV on days 1 and 8, vincristine 1.4 mg/m2 IV on days 1 and 8 (maximum of 2 mg), and procarbazine 100 mg/m2 per day PO on days 1 through 14. Cycles were repeated every 4 weeks, as permitted by blood count recovery. Six patients received VEPA alternated with VAMP and/or COP (VEPA hybrid) on the HOD90 study before the opening of HOD94.
Radiation Therapy
Radiation therapy was prescribed as consolidative treatment for all patients. Patients received involved-field irradiation to all sites of initial disease, nodal and extra-nodal, with the exception of bone marrow. Radiation therapy was initiated after the second course of chemotherapy except for patients with extensive disease involvement who received radiation after completion of all chemotherapy; patients requiring more than one treatment volume received further radiation after the fourth or sixth course of chemotherapy. Treatment was concurrently delivered to all upper-torso sites of disease, then to sites involving the abdomen, and subsequently to sites involving the pelvis. The prescribed dose was 25.5 Gy delivered at 1.5 Gy per fraction, one fraction per day over 17 treatment days.
Patients who experienced a complete response at all sites of disease after the initial two courses of chemotherapy subsequently received 15 Gy at 1.5 Gy per fraction, one fraction per day, over 10 treatment days to all sites of disease. All sites of residual disease (measured after two courses of chemotherapy) and all initially bulky nodal sites received 25.5 Gy.
Radiation was delivered through anterior and posterior fields with 4-MV photons. Some patients with superficial extranodal extension of disease received electron energies in fields adjacent to the nodal disease sites. Radiation was delivered to the entire lymph node–bearing region if involvement of any lymph node within that region was documented at diagnosis. Extensions of disease to an extra-nodal site designated as a so-called E lesion were encompassed within the adjacent nodal treatment fields. Extra-nodal sites involving the lung classified as stage IV disease were planned to receive 8 Gy whole-lung irradiation delivered with partial transmission blocking over the entire course of treatment (approximately 0.47 Gy/d). Involvement of the bone was treated with individual treatment fields of 15 to 25.5 Gy at 1.5-Gy per fraction. Bone marrow involvement was not treated with radiation therapy.
Analysis of Patterns of Failure
The initial sites of disease involvement, initial lymph node and extra-nodal regions irradiated, and subsequent sites of treatment failure were all recorded and confirmed from the patients' medical record, pretreatment diagnostic imaging, radiation therapy treatment records and portals, and diagnostic imaging at the time of treatment failure. The patterns of treatment failure were related to the initial sites of disease and regions of irradiated nodal and extra-nodal tissue. Local failure was defined as recurrence of Hodgkin's disease in an initial site of disease involvement. Local failures were further reviewed by two radiation oncologists and a radiologist (M.K., L.K., and S.K.) to determine their relationship to the radiation therapy treatment fields using CT imaging from diagnosis and recurrence in conjunction with radiotherapy simulation and treatment portal films. Local recurrences were further categorized as in-field or marginal. Disease recurrence was considered in-field when the epicenter of the recurrence was located within the treatment fields away from the block edge 1 to 2 cm, ensuring that the recurrence was not potentially related to an underdose of radiation. Recurrences that occurred at or under a treatment block edge were classified as marginal. Recurrences in other locations were considered distant. Nodal as well as extra-nodal sites of disease were evaluated for treatment failure. Detection of disease in the bone marrow after completion of treatment was always classified as a distant site of failure.
Statistical Analyses
Descriptive statistics derived from the study population included patient characteristics, treatment, and prognostic factors. Values used to categorize WBC count, hemoglobin concentration, and erythrocyte sedimentation rate were based on values found to be significantly predictive of outcome in previously published literature.13 The Kaplan-Meier method14 was used to estimate the overall survival (OS), event-free survival (EFS), and cumulative incidence (CI) of local failure for the entire cohort and for subgroups of patients, as determined by the presence of individual prognostic factors by using the Kaplan-Meier method. Survival differences were compared using the unweighted log-rank test.15 The OS time was determined as the time from enrollment on protocol until death or last follow-up evaluation. The EFS time was determined as the time from enrollment on protocol until the first adverse event (ie, disease progression, second malignancy, or death from any cause). The CI of local failure was determined as the time from enrollment on protocol until local recurrence, either with or without simultaneous distance recurrence and with other events classified as competing risks. To determine the independent prognostic significance of pretreatment factors on local failure, we conducted a multivariable analysis using the Cox proportional hazards regression method.16 Fisher's exact test was used to test for independence. All calculations were performed with SAS statistical software (version 9.2; SAS Institute Inc, Cary, NC).
RESULTS
Patient Characteristics
The pretreatment characteristics of the 195 patients are listed in Table 1. The median age at diagnosis was 15.3 years (range, 3.1 to 20.9 years). The median follow-up time was 7.6 years for the entire cohort and 7.8 years for patients who were alive at the time of the analysis (range, 2.4 to 11.8 years).
OS and EFS
The 5-year OS for the entire cohort was 93.1%, and the 10-year OS was 90.3%. The 5-year EFS was 83.8%, and the 10-year EFS was 80.8% (Fig 1). Thirty-four patients died, experienced disease recurrence, or an adverse treatment-related event (eg, secondary malignancy). Four of the 34 patients developed a second malignancy. Two secondary cancers arose within the radiation treatment fields: a low-grade mucoepidermoid carcinoma of the parotid was diagnosed at 2.4 years in one patient, and follicular carcinoma of the thyroid was diagnosed at 8.4 years in the other. The other two second malignancies arose outside the radiation volume: Ewing's sarcoma, which developed 4 years post-therapy, and acute myeloid leukemia, which developed at 1.8 years.
Incidence and Patterns of Local Failure
Twenty-seven (13.8%) of 195 patients experienced recurrence of Hodgkin's disease. Local failure was a component of relapse in 22 (81.5%) of these patients, including a nodal site of disease in 21 cases and lung in one. In 14 of the 22 cases local failure alone was the only site of recurrence. One patient with stage IIB disease progressed locally while on systemic therapy in a nonbulky site of disease before radiation therapy and is classified as a local failure. All 23 patients with stage IV pulmonary involvement at diagnosis received low-dose lung irradiation to the whole involved lung(s). Eleven patients received unilateral lung irradiation and 12 received bilateral lung irradiation. Two of the 23 patients receiving lung irradiation experienced local failure within the lungs; both cases presented with initial bilateral lung involvement. Three additional patients without initial pulmonary involvement experienced failure in the lung.
Sixty-one patients presented with bulky mediastinal disease, all of which received 25.5 Gy to the mediastinum as part of protocol-directed therapy. Local failure was a primary component of failure in this cohort, with 13 of 14 patients experiencing treatment failure in an initially involved nodal site. Only four of 13 patients had a component of failure in the mediastinum, and only one of these patients failed in the mediastinum alone.
Four of the 22 cases of local failure were classified as marginal as follows: two patients with recurrent disease in the lower mediastinum at the lateral edge of the treatment field, one patient with recurrence in the axilla adjacent the tip of the lung block, and one patient with para-aortic nodal recurrence at the lateral edge of the treatment field. The CI of local failure for the cohort was 10.9% (SE, 2.3%) at 5 years and 11.6% (SE, 2.4%) at 10 years (Fig 2).
Prognostic Factors for Local Failure
Univariable analysis of factors predicting local treatment failure is shown in Table 2. Patients with extranodal sites of disease experienced an increased rate of local failure; eight (20.0%) of 40 patients experienced disease recurrence locally in an initially involved site of disease; local recurrence in three of these eight patients was extranodal in nature. Nodular sclerosis histology also predicted for local failure, with 20 of 22 local failures occurring in patients with this histologic subtype of disease; the other two failures were in patients with mixed cellularity histology. Treatment-related factors, including the type of chemotherapy and the response to the initial two courses of chemotherapy, did not predict for local treatment failure.
At 5 years, the CI of local failure did not differ between the subgroup of patients who received 15 Gy of dose-attenuated radiation therapy (10.3%; n = 69) and those who received 25.5 Gy (12.1%; n = 126; P = .61; Fig 3).
The risk of local failure after 5 years was five-fold higher for patients with bulky mediastinal disease (21.6%) than for those with nonbulky mediastinal disease (4.0%; P = .005; Fig 4). To determine whether the site within the radiated volume related to the bulky site of mediastinal disease, we calculated the incidence of treatment failure within a bulky mediastinum. Four (6.6%) of 61 patients with bulky mediastinal disease experienced relapse with a component of disease in the mediastinum. Nine (4.8%) of the 61 patients experienced a local recurrence that did not involve the mediastinum, instead involving the neck (four patients), axilla (two patients), neck and axilla (one patient), and para-aortic nodes or spleen (two patients). No significant association was found between the location of failure sites (local mediastinal failure v local nonmediastinal failure) in patients with bulky mediastinal disease (P = .267).
Multivariable Analysis
Multivariable analysis of prognostic factors that independently predicted local failure revealed that male sex, bulky mediastinal disease, and hemoglobin concentration less than 11 g/dL at the time of diagnosis were significant predictors of local failure, and extranodal and bulky mediastinal disease demonstrated a trend toward significance (Table 3).
DISCUSSION
The cumulative incidence of local treatment failure in a cohort of 195 pediatric patients with Hodgkin's disease treated at St Jude Children's Research Hospital was 10.9% at 5 and 11.6% at 10 years. Seventy percent of all local recurrences were seen within the first 2 years. Although the incidence of local treatment failure was quite low, the patterns of failure indicated local failure as a component of failure in 82% of patients with recurrent disease. One half of the treatment failures (14 of 27) were local without evidence of distant recurrence.
Despite the high proportion of local failures among patients who experienced disease recurrence, the overall incidence of treatment failure was low. Therefore, we analyzed prognostic factors to define patient groups at high risk for local failure. Several prognostic factors are well defined in the literature for patients receiving combined-modality therapy for Hodgkin's disease. "B" symptoms, clinical stage, histology, and bulky mediastinal disease have been reported as significant factors for decreased EFS, disease-free, or OS.2,13,17 In our patient population, several of these clinical factors, as well as several laboratory measures, were significantly predictive of a higher CI of local failure.
A multivariable model for local failure defined male sex, hemoglobin concentration less than 11 g/dL, and bulky mediastinal disease as significantly predictive factors of local failure. However, the ability of these factors to predict the specific site of in-field failure is limited. Bulky mediastinal disease seems to increase the risk of local failure. Despite our concerns that bulky mediastinal disease may not be adequately controlled with 25.5 Gy of radiation, we found that only four of the 61 patients with bulky disease experienced recurrence in the mediastinum, and no association was seen between bulky mediastinal disease and risk of local failure within the mediastinum. The relationship between bulky disease and local disease control with radiation therapy has been investigated by several groups in the context of local control for both pediatric and adult patients.7,17-19 The German DAL-HD-90 study, in which bulky disease was defined as a tumor site greater than 4 cm x 5 cm whose volume was larger than 50 cm3, determined that bulky disease predicted disease recurrence, but the site of recurrence was not correlated with the initial volume of disease.17 In the DAL-HD-90 study, as well as the current GPOH-HD 95 trial, patients with residual disease after chemotherapy received a boost to a cumulative dose of 30 to 35 Gy. Patients with bulky disease more frequently had larger residual masses and received higher radiation doses (20 to 25.5 Gy); thus the rate of local control could not be uniformly defined for those patients.7,17 Lung involvement at diagnosis was treated as part of protocol therapy with low-dose irradiation to the involved lungs during radiation therapy to other sites of disease in the chest. The local failure rate in the involved lungs (two of 23) seems similar to the overall local failure rate for all sites of disease in this study (22 of 195). Though the pulmonary control rate seems favorable, the design of the studies included in this report prevent definitive conclusions regarding the benefit of lung irradiation. Four of 33 adult patients with initial lung involvement experienced pulmonary relapse after Stanford V systemic therapy, a compacted, dose intensive multiagent chemotherapy alternating combinations of myelosuppressive and nonmyelosuppressive agents weekly for 12 weeks.20 Histology has also been noted to be a significant predictor of event-free survival in univariate analysis and multivariate analysis in pediatric Hodgkin's disease studies.13,18 Nodular sclerosing histology was a predictor of local failure on univariable analysis, but did not remain significant in the multivariable model likely because of the small number of events and the overwhelming effect of other, more significant factors.
Adult Hodgkin's disease studies, which often accrue patients as young as 16 years, typically include as a component of combined-modality therapy the delivery of substantially higher doses of radiation therapy to sites of bulky disease; these doses are often as high as 39.6 Gy. Local failure rates in the mediastinum were comparable when this approach was used to treat a patient population that included adolescents and older adults.18 Although clearly effective, this approach in a pediatric and young adult population could increase the risk of late adverse events and contradicts the rationale of the risk-adapted approach.21 It seems that for a selected group of patients, 15-Gy radiation therapy may provide adequate local tumor control, while allowing further tailoring of therapy to the risk of disease. It is gratifying that more than one third of our patients received 15-Gy radiation therapy to one or more nodal regions based on a complete or mixed response to therapy. Although local control rates with 15-Gy irradiation are excellent, this study was not powered to detect small differences in the incidence of local failure between the patients receiving 15 Gy or 25.5 Gy.
Whether patients whose disease shows an early response to chemotherapy can avoid radiation therapy altogether is not known, but this question is being addressed in our current consortium study where patients with early-stage disease receive VAMP chemotherapy and no radiation therapy if they achieve a complete response after two courses of chemotherapy. Investigators in the recently completed German Hodgkin's study GPOH-HD 95 tested the hypothesis that patients who experienced complete response after completion of chemotherapy may avoid adjuvant radiation therapy.7,21 Although the overall survival was unchanged, patients with intermediate or advanced disease treated on GPOH-HD 95 experienced a higher rate of treatment failure requiring salvage therapy.
The outcome of Hodgkin's disease is among the most favorable of the pediatric malignancies. In the current era, delivery of risk-adapted combined-modality therapy based on initial prognostic factors and disease response to initial treatment is the standard of care; this approach allows the delivery of limited therapy to many patients, thereby potentially decreasing the cumulative incidence of late effects. Current prognostic factors define groups of patients at increased risk of disease recurrence, but those factors do not indicate the site of recurrence; thus they do not facilitate the modification of dose or volume of radiation therapy. Intensification of systemic therapy for patients with advanced disease22 and incorporation of functional imaging studies to monitor disease response23,24 or potential biologic differences in disease may increase our ability pinpoint regions at higher risk of recurrence.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
NOTES
Supported in part by the American-Lebanese-Syrian Associated Charities.
Presented at the 46th Annual Meeting of the American Society for Therapeutic Radiology and Oncology, October 2-7 2004, Atlanta, GA.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Shah AB, Hudson MM, Poquette CA, et al: Long-term follow-up of patients treated with primary radiotherapy for supradiaphragmatic Hodgkin's disease at St. Jude Children's Research Hospital. Int J Radiat Oncol Biol Phys 44:867-877, 1999
Ekert H, Toogood I, Downie P, et al: High incidence of treatment failure with vincristine, etoposide, epirubicin, and prednisolone chemotherapy with successful salvage in childhood Hodgkin's disease. Med Pediatr Oncol 32:255-258, 1999
Hunger SP, Link MP, Donaldson SS: ABVD/MOPP and low-dose involved-field radiotherapy in pediatric Hodgkin's disease: The Stanford experience. J Clin Oncol 12:2160-2166, 1994
Weiner MA, Leventhal B, Brecher ML, et al: Randomized study of intensive MOPP-ABVD with or without low-dose total-nodal radiation therapy in the treatment of stages IIB, IIIA2, IIIB, and IV Hodgkin's disease in pediatric patients: A Pediatric Oncology Group study. J Clin Oncol 15:2769-2779, 1997
Nachman JB, Sposto R, Herzog P, et al: Randomized comparison of low-dose involved-field radiotherapy and no radiotherapy for children with Hodgkin's disease who achieve a complete response to chemotherapy. J Clin Oncol 20:3765-3771, 2002
Hudson MM, Greenwald C, Thompson E, et al: Efficacy and toxicity of multiagent chemotherapy and low-dose involved-field radiotherapy in children and adolescents with Hodgkin's disease. J Clin Oncol 11:100-108, 1993
Ruhl U, Albrecht M, Dieckmann K, et al: Response-adapted radiotherapy in the treatment of pediatric Hodgkin's disease: An interim report at 5 years of the German GPOH-HD 95 trial. Int J Radiat Oncol Biol Phys 51:1209-1218, 2001
Schwartz CL, Constine LS, London W, et al: POG 9425: Response-based, intensively timed therapy for intermediate/high stage (IS/HS) pediatric Hodgkin's disease. Proc Am Soc Clin Oncol 21:389a, 2002 (abstr 1555)
Donaldson SS, Hudson MM, Lamborn KR, et al: VAMP and low-dose, involved-field radiation for children and adolescents with favorable, early-stage Hodgkin's disease: Results of a prospective clinical trial. J Clin Oncol 20:3081-3087, 2002
Donaldson SS, Link MP: Combined modality treatment with low-dose radiation and MOPP chemotherapy for children with Hodgkin's disease. J Clin Oncol 5:742-749, 1987
Friedmann AM, Hudson MM, Weinstein HJ, et al: Treatment of unfavorable childhood Hodgkin's disease with VEPA and low-dose, involved-field radiation. J Clin Oncol 20:3088-3094, 2002
Hudson MM, Krasin M, Link MP, et al: Risk-adapted combined modality therapy with VAMP/COP and response-based, involved-field radiation for unfavorable pediatric Hodgkin's disease. J Clin Oncol 22:4541-4550, 2004
Smith RS, Chen Q, Hudson MM, et al: Prognostic factors for children with Hodgkin's disease treated with combined-modality therapy. J Clin Oncol 21:2026-2033, 2003
Kaplan EL, Meier P: Nonparametric estimation from incomplete observations. J Am Stat Assoc 53:457-481, 1958
Kalbfleisch JD, Prentice RL: The Statistical Analysis of Failure Time Data. New York, NY, John Wiley, 1980, pp 163-178
Cox DR: Regression and life-tables. J Roy Stat Soc B 34:187-220, 1972
Dieckmann K, Potter R, Hofmann J, et al: Does bulky disease at diagnosis influence outcome in childhood Hodgkin's disease and require higher radiation doses? Results from the German-Austrian Pediatric Multicenter Trial DAL-HD-90. Int J Radiat Oncol Biol Phys 56:644-652, 2003
Chronowski GM, Wilder RB, Tucker SL, et al: Analysis of in-field control and late toxicity for adults with early-stage Hodgkin's disease treated with chemotherapy followed by radiotherapy. Int J Radiat Oncol Biol Phys 55:36-43, 2003
Mendenhall NP, Rodrigue LL, Moore-Higgs GJ, et al: The optimal dose of radiation in Hodgkin's disease: An analysis of clinical and treatment factors affecting in-field disease control. Int J Radiat Oncol Biol Phys 44:551-561, 1999
Horning SJ, Hoppe RT, Breslin S, et al: Stanford V and radiotherapy for locally extensive and advanced Hodgkin's disease: Mature results of a prospective clinical trial. J Clin Oncol 20:630-637, 2002
Bleyer WA: Adolescents and young adults with cancer: A neglected population. American Society of Clinical Oncology Educational Book, pp 125-132, 2001
Dorffel W, Luders H, Ruhl U, et al: Preliminary results of the multicenter trial GPOH-HD 95 for the treatment of Hodgkin's disease in children and adolescents: Analysis and outlook. Klin Padiatr 215:139-145, 2003
Krasin MJ, Hudson MM, Kaste SC: Positron emission tomography in pediatric radiation oncology: Integration in the treatment-planning process. Pediatr Radiol 34:214-221, 2004
Hudson MM, Krasin MJ, Kaste SC: PET imaging in pediatric Hodgkin's lymphoma. Pediatr Radiol 34:190-198, 2004
Submitted December 14, 2004; accepted August 9, 2005.(Matthew J. Krasin, Shesh )
ABSTRACT
PURPOSE: Refinement in managing pediatric Hodgkin's disease (HD) requires understanding of factors associated with local treatment failure. This study defines the cumulative incidence (CI) of local failure (LF) and prognostic factors for pediatric patients treated for HD with combined-modality therapy (CMT).
PATIENTS AND METHODS: We enrolled 195 patients onto two prospective studies at St Jude Children's Research Hospital between 1990 and 2000. Patients received CMT with chemotherapy (vinblastine, doxorubicin, methotrexate, and prednisone [VAMP]; vinblastine, etoposide, prednisone, and doxorubicin; or VAMP/cyclophosphamide, vincristine, and procarbazine) and involved-field radiation therapy delivered to initial site(s) of disease on the basis of early response. Sites of disease involvement, treatment, and sites of failure were confirmed from the patients' medical record, imaging, and radiotherapy treatment records. We estimated the overall survival, event-free survival, and CI of LF.
RESULTS: With a median follow-up of 7.6 years, the CI of LF was 10.9% and 11.6% at 5 and 10 years, respectively. Twenty-seven (14%) of 195 patients experienced recurrence of HD, and 22 (81%) of those experienced LF. Bulky mediastinal disease greater than one third transthoracic diameter predicted a higher incidence of LF, but did not predict failure in the mediastinum. Male sex, low initial hemoglobin, and bulky mediastinal disease were prognostic indicators of LF. Attenuation of radiation dose to 15 Gy based on response provides excellent infield control.
CONCLUSION: CMT provides excellent local disease control in children and young adults with HD. LF remains a primary site of disease recurrence, with male sex, low initial hemoglobin, and bulky mediastinal disease predicting for LF.
INTRODUCTION
The management of pediatric Hodgkin's disease has evolved over the past 20 years from high-dose extensive nodal irradiation or aggressive multiagent systemic chemotherapy to combined-modality therapy.1-6 This shift has taken place because standard-dose irradiation and alkylator-based chemotherapy have resulted in significant treatment effects, which can often be attenuated by combining more limited chemotherapy with reduced-dose limited-volume radiation therapy.7-12
Since the early 1990s, we have carried out two sequential risk-adapted studies in collaboration with other institutions. In these studies, combined-modality therapy was adapted on the basis of the extent of disease, prognostic factors, and response to therapy.9,11-13 To continue to refine therapy for both early- and advanced-stage Hodgkin's disease, we have investigated the sites and patterns of treatment failure to further evaluate the effectiveness of current therapies and plan future approaches.
In this report, we describe the local disease outcome and patterns of treatment failure observed in pediatric and young adult patients with Hodgkin's disease treated on two large prospective trials at St Jude Children's Research Hospital.
PATIENTS AND METHODS
HOD90 and HOD94 Study Design
The St Jude Children's Research Hospital HOD90 study opened in January 1990. This study enrolled 76 institutional patients into two treatment strata based on the risk classification outlined below. The study closed in 1993 because it met its stopping criteria in the high-risk strata.11 The second study, HOD94, opened in October 1993 and closed in February 2000.9,12 This study enrolled 119 patients based on the same two treatment strata used in HOD90. These protocols were approved by the institutional review board at St Jude Children's Research Hospital and concurrently opened at collaborating institutions. In total, this risk-based approach was used to treat Hodgkin's disease in 195 patients at St Jude Children's Research Hospital. The event-free outcome for these patients has previously been reported with the collaborating institutions as part of protocol-specific reports.9,11,12 Only patients accessioned to the studies at St Jude Children's Research Hospital were included in this analysis of patterns of failure based on the detailed classification of nodal site involvement, dose-specific irradiation of defined nodal regions, and extensive imaging correlations of sites of failure with initial sites of disease involvement and radiation treatment.
All patients enrolled on HOD90 or HOD94 underwent initial staging evaluations that included a biopsy-proven diagnosis of Hodgkin's disease, a complete initial clinical history, and a physical examination in which sites and extent of nodal disease and splenic involvement were documented. Laboratory evaluations included a complete blood cell count and chemistry panel. Imaging studies, which have evolved over time, included chest x-ray; gallium scan; computed tomography (CT) scan of the chest, abdomen, and pelvis; and lymphangiogram.
After two cycles of chemotherapy, disease response was evaluated by clinical examination, nuclear imaging, and CT scan. Complete response was defined as resolution of all tumor-related constitutional symptoms and disappearance or normalization of all measurable nodal or nonmeasurable but assessable disease. Partial response was defined as at least 50% reduction in the sum of the products of two perpendicular diameters of all measurable lesions and disappearance of constitutional symptoms if initially present. For purposes of this analysis, mixed response was defined as a complete response in some nodal regions, but only a partial response in other nodal or non-nodal sites of disease. Biopsy confirmation of disease progression or relapse was required. Follow-up was conducted every 3 to 6 months after completion of therapy, and cases were coded for any site of treatment failure, secondary malignancy, or other causes of death.
Risk Classification
Patients with high-risk or unfavorable disease included those with Ann Arbor stage III or IV disease and patients with stage I or II disease who had bulky lymphadenopathy. (Bulky lymphadenopathy was defined as either a mediastinal mass of > one third of the transthoracic diameter on an upright chest radiograph or a peripheral lymph node of > 6 cm in maximal axial diameter.) The protocol was amended in July 1998 to include all patients with "B" symptoms in the high-risk group. All other patients were considered as having favorable disease and assigned to the low-risk group (ie, stages IA to IIA without bulky disease).
Systemic Therapy
Treatment included four to six cycles of chemotherapy and involved-field radiation therapy; the prescribed radiation dose was based on tumor response after the second course of chemotherapy. Patients in the low-risk group received vinblastine, doxorubicin, methotrexate, and prednisone (VAMP) chemotherapy for four cycles on both the HOD90 and HOD94 studies. VAMP chemotherapy was administered as follows: vinblastine 6 mg/m2 intravenously (IV) on days 1 and 15, doxorubicin 25 mg/m2 IV on days 1 and 15, methotrexate 20 mg/m2 IV on days 1 and 15, and prednisone 40 mg/m2 per day orally on days 1 through 14.
Systemic therapy for patients in the high-risk group on the HOD90 study included vinblastine, etoposide, prednisone, and doxorubicin (VEPA) for six cycles administered as follows: vinblastine 6 mg/m2 IV on days 1 and 15, etoposide 200 mg/m2 IV on days 1 and 15, prednisone 40 mg/m2 per day PO on days 1 through 14 (omitted after mediastinal radiation therapy), and doxorubicin 25 mg/m2 IV on days 1 and 15. Patients with high-risk disease treated on HOD94 received three cycles of VAMP chemotherapy, as described above, alternated with three cycles of cyclophosphamide, vincristine, and procarbazine (COP) chemotherapy. COP chemotherapy was administered as follows: cyclophosphamide 600 mg/m2 IV on days 1 and 8, vincristine 1.4 mg/m2 IV on days 1 and 8 (maximum of 2 mg), and procarbazine 100 mg/m2 per day PO on days 1 through 14. Cycles were repeated every 4 weeks, as permitted by blood count recovery. Six patients received VEPA alternated with VAMP and/or COP (VEPA hybrid) on the HOD90 study before the opening of HOD94.
Radiation Therapy
Radiation therapy was prescribed as consolidative treatment for all patients. Patients received involved-field irradiation to all sites of initial disease, nodal and extra-nodal, with the exception of bone marrow. Radiation therapy was initiated after the second course of chemotherapy except for patients with extensive disease involvement who received radiation after completion of all chemotherapy; patients requiring more than one treatment volume received further radiation after the fourth or sixth course of chemotherapy. Treatment was concurrently delivered to all upper-torso sites of disease, then to sites involving the abdomen, and subsequently to sites involving the pelvis. The prescribed dose was 25.5 Gy delivered at 1.5 Gy per fraction, one fraction per day over 17 treatment days.
Patients who experienced a complete response at all sites of disease after the initial two courses of chemotherapy subsequently received 15 Gy at 1.5 Gy per fraction, one fraction per day, over 10 treatment days to all sites of disease. All sites of residual disease (measured after two courses of chemotherapy) and all initially bulky nodal sites received 25.5 Gy.
Radiation was delivered through anterior and posterior fields with 4-MV photons. Some patients with superficial extranodal extension of disease received electron energies in fields adjacent to the nodal disease sites. Radiation was delivered to the entire lymph node–bearing region if involvement of any lymph node within that region was documented at diagnosis. Extensions of disease to an extra-nodal site designated as a so-called E lesion were encompassed within the adjacent nodal treatment fields. Extra-nodal sites involving the lung classified as stage IV disease were planned to receive 8 Gy whole-lung irradiation delivered with partial transmission blocking over the entire course of treatment (approximately 0.47 Gy/d). Involvement of the bone was treated with individual treatment fields of 15 to 25.5 Gy at 1.5-Gy per fraction. Bone marrow involvement was not treated with radiation therapy.
Analysis of Patterns of Failure
The initial sites of disease involvement, initial lymph node and extra-nodal regions irradiated, and subsequent sites of treatment failure were all recorded and confirmed from the patients' medical record, pretreatment diagnostic imaging, radiation therapy treatment records and portals, and diagnostic imaging at the time of treatment failure. The patterns of treatment failure were related to the initial sites of disease and regions of irradiated nodal and extra-nodal tissue. Local failure was defined as recurrence of Hodgkin's disease in an initial site of disease involvement. Local failures were further reviewed by two radiation oncologists and a radiologist (M.K., L.K., and S.K.) to determine their relationship to the radiation therapy treatment fields using CT imaging from diagnosis and recurrence in conjunction with radiotherapy simulation and treatment portal films. Local recurrences were further categorized as in-field or marginal. Disease recurrence was considered in-field when the epicenter of the recurrence was located within the treatment fields away from the block edge 1 to 2 cm, ensuring that the recurrence was not potentially related to an underdose of radiation. Recurrences that occurred at or under a treatment block edge were classified as marginal. Recurrences in other locations were considered distant. Nodal as well as extra-nodal sites of disease were evaluated for treatment failure. Detection of disease in the bone marrow after completion of treatment was always classified as a distant site of failure.
Statistical Analyses
Descriptive statistics derived from the study population included patient characteristics, treatment, and prognostic factors. Values used to categorize WBC count, hemoglobin concentration, and erythrocyte sedimentation rate were based on values found to be significantly predictive of outcome in previously published literature.13 The Kaplan-Meier method14 was used to estimate the overall survival (OS), event-free survival (EFS), and cumulative incidence (CI) of local failure for the entire cohort and for subgroups of patients, as determined by the presence of individual prognostic factors by using the Kaplan-Meier method. Survival differences were compared using the unweighted log-rank test.15 The OS time was determined as the time from enrollment on protocol until death or last follow-up evaluation. The EFS time was determined as the time from enrollment on protocol until the first adverse event (ie, disease progression, second malignancy, or death from any cause). The CI of local failure was determined as the time from enrollment on protocol until local recurrence, either with or without simultaneous distance recurrence and with other events classified as competing risks. To determine the independent prognostic significance of pretreatment factors on local failure, we conducted a multivariable analysis using the Cox proportional hazards regression method.16 Fisher's exact test was used to test for independence. All calculations were performed with SAS statistical software (version 9.2; SAS Institute Inc, Cary, NC).
RESULTS
Patient Characteristics
The pretreatment characteristics of the 195 patients are listed in Table 1. The median age at diagnosis was 15.3 years (range, 3.1 to 20.9 years). The median follow-up time was 7.6 years for the entire cohort and 7.8 years for patients who were alive at the time of the analysis (range, 2.4 to 11.8 years).
OS and EFS
The 5-year OS for the entire cohort was 93.1%, and the 10-year OS was 90.3%. The 5-year EFS was 83.8%, and the 10-year EFS was 80.8% (Fig 1). Thirty-four patients died, experienced disease recurrence, or an adverse treatment-related event (eg, secondary malignancy). Four of the 34 patients developed a second malignancy. Two secondary cancers arose within the radiation treatment fields: a low-grade mucoepidermoid carcinoma of the parotid was diagnosed at 2.4 years in one patient, and follicular carcinoma of the thyroid was diagnosed at 8.4 years in the other. The other two second malignancies arose outside the radiation volume: Ewing's sarcoma, which developed 4 years post-therapy, and acute myeloid leukemia, which developed at 1.8 years.
Incidence and Patterns of Local Failure
Twenty-seven (13.8%) of 195 patients experienced recurrence of Hodgkin's disease. Local failure was a component of relapse in 22 (81.5%) of these patients, including a nodal site of disease in 21 cases and lung in one. In 14 of the 22 cases local failure alone was the only site of recurrence. One patient with stage IIB disease progressed locally while on systemic therapy in a nonbulky site of disease before radiation therapy and is classified as a local failure. All 23 patients with stage IV pulmonary involvement at diagnosis received low-dose lung irradiation to the whole involved lung(s). Eleven patients received unilateral lung irradiation and 12 received bilateral lung irradiation. Two of the 23 patients receiving lung irradiation experienced local failure within the lungs; both cases presented with initial bilateral lung involvement. Three additional patients without initial pulmonary involvement experienced failure in the lung.
Sixty-one patients presented with bulky mediastinal disease, all of which received 25.5 Gy to the mediastinum as part of protocol-directed therapy. Local failure was a primary component of failure in this cohort, with 13 of 14 patients experiencing treatment failure in an initially involved nodal site. Only four of 13 patients had a component of failure in the mediastinum, and only one of these patients failed in the mediastinum alone.
Four of the 22 cases of local failure were classified as marginal as follows: two patients with recurrent disease in the lower mediastinum at the lateral edge of the treatment field, one patient with recurrence in the axilla adjacent the tip of the lung block, and one patient with para-aortic nodal recurrence at the lateral edge of the treatment field. The CI of local failure for the cohort was 10.9% (SE, 2.3%) at 5 years and 11.6% (SE, 2.4%) at 10 years (Fig 2).
Prognostic Factors for Local Failure
Univariable analysis of factors predicting local treatment failure is shown in Table 2. Patients with extranodal sites of disease experienced an increased rate of local failure; eight (20.0%) of 40 patients experienced disease recurrence locally in an initially involved site of disease; local recurrence in three of these eight patients was extranodal in nature. Nodular sclerosis histology also predicted for local failure, with 20 of 22 local failures occurring in patients with this histologic subtype of disease; the other two failures were in patients with mixed cellularity histology. Treatment-related factors, including the type of chemotherapy and the response to the initial two courses of chemotherapy, did not predict for local treatment failure.
At 5 years, the CI of local failure did not differ between the subgroup of patients who received 15 Gy of dose-attenuated radiation therapy (10.3%; n = 69) and those who received 25.5 Gy (12.1%; n = 126; P = .61; Fig 3).
The risk of local failure after 5 years was five-fold higher for patients with bulky mediastinal disease (21.6%) than for those with nonbulky mediastinal disease (4.0%; P = .005; Fig 4). To determine whether the site within the radiated volume related to the bulky site of mediastinal disease, we calculated the incidence of treatment failure within a bulky mediastinum. Four (6.6%) of 61 patients with bulky mediastinal disease experienced relapse with a component of disease in the mediastinum. Nine (4.8%) of the 61 patients experienced a local recurrence that did not involve the mediastinum, instead involving the neck (four patients), axilla (two patients), neck and axilla (one patient), and para-aortic nodes or spleen (two patients). No significant association was found between the location of failure sites (local mediastinal failure v local nonmediastinal failure) in patients with bulky mediastinal disease (P = .267).
Multivariable Analysis
Multivariable analysis of prognostic factors that independently predicted local failure revealed that male sex, bulky mediastinal disease, and hemoglobin concentration less than 11 g/dL at the time of diagnosis were significant predictors of local failure, and extranodal and bulky mediastinal disease demonstrated a trend toward significance (Table 3).
DISCUSSION
The cumulative incidence of local treatment failure in a cohort of 195 pediatric patients with Hodgkin's disease treated at St Jude Children's Research Hospital was 10.9% at 5 and 11.6% at 10 years. Seventy percent of all local recurrences were seen within the first 2 years. Although the incidence of local treatment failure was quite low, the patterns of failure indicated local failure as a component of failure in 82% of patients with recurrent disease. One half of the treatment failures (14 of 27) were local without evidence of distant recurrence.
Despite the high proportion of local failures among patients who experienced disease recurrence, the overall incidence of treatment failure was low. Therefore, we analyzed prognostic factors to define patient groups at high risk for local failure. Several prognostic factors are well defined in the literature for patients receiving combined-modality therapy for Hodgkin's disease. "B" symptoms, clinical stage, histology, and bulky mediastinal disease have been reported as significant factors for decreased EFS, disease-free, or OS.2,13,17 In our patient population, several of these clinical factors, as well as several laboratory measures, were significantly predictive of a higher CI of local failure.
A multivariable model for local failure defined male sex, hemoglobin concentration less than 11 g/dL, and bulky mediastinal disease as significantly predictive factors of local failure. However, the ability of these factors to predict the specific site of in-field failure is limited. Bulky mediastinal disease seems to increase the risk of local failure. Despite our concerns that bulky mediastinal disease may not be adequately controlled with 25.5 Gy of radiation, we found that only four of the 61 patients with bulky disease experienced recurrence in the mediastinum, and no association was seen between bulky mediastinal disease and risk of local failure within the mediastinum. The relationship between bulky disease and local disease control with radiation therapy has been investigated by several groups in the context of local control for both pediatric and adult patients.7,17-19 The German DAL-HD-90 study, in which bulky disease was defined as a tumor site greater than 4 cm x 5 cm whose volume was larger than 50 cm3, determined that bulky disease predicted disease recurrence, but the site of recurrence was not correlated with the initial volume of disease.17 In the DAL-HD-90 study, as well as the current GPOH-HD 95 trial, patients with residual disease after chemotherapy received a boost to a cumulative dose of 30 to 35 Gy. Patients with bulky disease more frequently had larger residual masses and received higher radiation doses (20 to 25.5 Gy); thus the rate of local control could not be uniformly defined for those patients.7,17 Lung involvement at diagnosis was treated as part of protocol therapy with low-dose irradiation to the involved lungs during radiation therapy to other sites of disease in the chest. The local failure rate in the involved lungs (two of 23) seems similar to the overall local failure rate for all sites of disease in this study (22 of 195). Though the pulmonary control rate seems favorable, the design of the studies included in this report prevent definitive conclusions regarding the benefit of lung irradiation. Four of 33 adult patients with initial lung involvement experienced pulmonary relapse after Stanford V systemic therapy, a compacted, dose intensive multiagent chemotherapy alternating combinations of myelosuppressive and nonmyelosuppressive agents weekly for 12 weeks.20 Histology has also been noted to be a significant predictor of event-free survival in univariate analysis and multivariate analysis in pediatric Hodgkin's disease studies.13,18 Nodular sclerosing histology was a predictor of local failure on univariable analysis, but did not remain significant in the multivariable model likely because of the small number of events and the overwhelming effect of other, more significant factors.
Adult Hodgkin's disease studies, which often accrue patients as young as 16 years, typically include as a component of combined-modality therapy the delivery of substantially higher doses of radiation therapy to sites of bulky disease; these doses are often as high as 39.6 Gy. Local failure rates in the mediastinum were comparable when this approach was used to treat a patient population that included adolescents and older adults.18 Although clearly effective, this approach in a pediatric and young adult population could increase the risk of late adverse events and contradicts the rationale of the risk-adapted approach.21 It seems that for a selected group of patients, 15-Gy radiation therapy may provide adequate local tumor control, while allowing further tailoring of therapy to the risk of disease. It is gratifying that more than one third of our patients received 15-Gy radiation therapy to one or more nodal regions based on a complete or mixed response to therapy. Although local control rates with 15-Gy irradiation are excellent, this study was not powered to detect small differences in the incidence of local failure between the patients receiving 15 Gy or 25.5 Gy.
Whether patients whose disease shows an early response to chemotherapy can avoid radiation therapy altogether is not known, but this question is being addressed in our current consortium study where patients with early-stage disease receive VAMP chemotherapy and no radiation therapy if they achieve a complete response after two courses of chemotherapy. Investigators in the recently completed German Hodgkin's study GPOH-HD 95 tested the hypothesis that patients who experienced complete response after completion of chemotherapy may avoid adjuvant radiation therapy.7,21 Although the overall survival was unchanged, patients with intermediate or advanced disease treated on GPOH-HD 95 experienced a higher rate of treatment failure requiring salvage therapy.
The outcome of Hodgkin's disease is among the most favorable of the pediatric malignancies. In the current era, delivery of risk-adapted combined-modality therapy based on initial prognostic factors and disease response to initial treatment is the standard of care; this approach allows the delivery of limited therapy to many patients, thereby potentially decreasing the cumulative incidence of late effects. Current prognostic factors define groups of patients at increased risk of disease recurrence, but those factors do not indicate the site of recurrence; thus they do not facilitate the modification of dose or volume of radiation therapy. Intensification of systemic therapy for patients with advanced disease22 and incorporation of functional imaging studies to monitor disease response23,24 or potential biologic differences in disease may increase our ability pinpoint regions at higher risk of recurrence.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
NOTES
Supported in part by the American-Lebanese-Syrian Associated Charities.
Presented at the 46th Annual Meeting of the American Society for Therapeutic Radiology and Oncology, October 2-7 2004, Atlanta, GA.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Shah AB, Hudson MM, Poquette CA, et al: Long-term follow-up of patients treated with primary radiotherapy for supradiaphragmatic Hodgkin's disease at St. Jude Children's Research Hospital. Int J Radiat Oncol Biol Phys 44:867-877, 1999
Ekert H, Toogood I, Downie P, et al: High incidence of treatment failure with vincristine, etoposide, epirubicin, and prednisolone chemotherapy with successful salvage in childhood Hodgkin's disease. Med Pediatr Oncol 32:255-258, 1999
Hunger SP, Link MP, Donaldson SS: ABVD/MOPP and low-dose involved-field radiotherapy in pediatric Hodgkin's disease: The Stanford experience. J Clin Oncol 12:2160-2166, 1994
Weiner MA, Leventhal B, Brecher ML, et al: Randomized study of intensive MOPP-ABVD with or without low-dose total-nodal radiation therapy in the treatment of stages IIB, IIIA2, IIIB, and IV Hodgkin's disease in pediatric patients: A Pediatric Oncology Group study. J Clin Oncol 15:2769-2779, 1997
Nachman JB, Sposto R, Herzog P, et al: Randomized comparison of low-dose involved-field radiotherapy and no radiotherapy for children with Hodgkin's disease who achieve a complete response to chemotherapy. J Clin Oncol 20:3765-3771, 2002
Hudson MM, Greenwald C, Thompson E, et al: Efficacy and toxicity of multiagent chemotherapy and low-dose involved-field radiotherapy in children and adolescents with Hodgkin's disease. J Clin Oncol 11:100-108, 1993
Ruhl U, Albrecht M, Dieckmann K, et al: Response-adapted radiotherapy in the treatment of pediatric Hodgkin's disease: An interim report at 5 years of the German GPOH-HD 95 trial. Int J Radiat Oncol Biol Phys 51:1209-1218, 2001
Schwartz CL, Constine LS, London W, et al: POG 9425: Response-based, intensively timed therapy for intermediate/high stage (IS/HS) pediatric Hodgkin's disease. Proc Am Soc Clin Oncol 21:389a, 2002 (abstr 1555)
Donaldson SS, Hudson MM, Lamborn KR, et al: VAMP and low-dose, involved-field radiation for children and adolescents with favorable, early-stage Hodgkin's disease: Results of a prospective clinical trial. J Clin Oncol 20:3081-3087, 2002
Donaldson SS, Link MP: Combined modality treatment with low-dose radiation and MOPP chemotherapy for children with Hodgkin's disease. J Clin Oncol 5:742-749, 1987
Friedmann AM, Hudson MM, Weinstein HJ, et al: Treatment of unfavorable childhood Hodgkin's disease with VEPA and low-dose, involved-field radiation. J Clin Oncol 20:3088-3094, 2002
Hudson MM, Krasin M, Link MP, et al: Risk-adapted combined modality therapy with VAMP/COP and response-based, involved-field radiation for unfavorable pediatric Hodgkin's disease. J Clin Oncol 22:4541-4550, 2004
Smith RS, Chen Q, Hudson MM, et al: Prognostic factors for children with Hodgkin's disease treated with combined-modality therapy. J Clin Oncol 21:2026-2033, 2003
Kaplan EL, Meier P: Nonparametric estimation from incomplete observations. J Am Stat Assoc 53:457-481, 1958
Kalbfleisch JD, Prentice RL: The Statistical Analysis of Failure Time Data. New York, NY, John Wiley, 1980, pp 163-178
Cox DR: Regression and life-tables. J Roy Stat Soc B 34:187-220, 1972
Dieckmann K, Potter R, Hofmann J, et al: Does bulky disease at diagnosis influence outcome in childhood Hodgkin's disease and require higher radiation doses? Results from the German-Austrian Pediatric Multicenter Trial DAL-HD-90. Int J Radiat Oncol Biol Phys 56:644-652, 2003
Chronowski GM, Wilder RB, Tucker SL, et al: Analysis of in-field control and late toxicity for adults with early-stage Hodgkin's disease treated with chemotherapy followed by radiotherapy. Int J Radiat Oncol Biol Phys 55:36-43, 2003
Mendenhall NP, Rodrigue LL, Moore-Higgs GJ, et al: The optimal dose of radiation in Hodgkin's disease: An analysis of clinical and treatment factors affecting in-field disease control. Int J Radiat Oncol Biol Phys 44:551-561, 1999
Horning SJ, Hoppe RT, Breslin S, et al: Stanford V and radiotherapy for locally extensive and advanced Hodgkin's disease: Mature results of a prospective clinical trial. J Clin Oncol 20:630-637, 2002
Bleyer WA: Adolescents and young adults with cancer: A neglected population. American Society of Clinical Oncology Educational Book, pp 125-132, 2001
Dorffel W, Luders H, Ruhl U, et al: Preliminary results of the multicenter trial GPOH-HD 95 for the treatment of Hodgkin's disease in children and adolescents: Analysis and outlook. Klin Padiatr 215:139-145, 2003
Krasin MJ, Hudson MM, Kaste SC: Positron emission tomography in pediatric radiation oncology: Integration in the treatment-planning process. Pediatr Radiol 34:214-221, 2004
Hudson MM, Krasin MJ, Kaste SC: PET imaging in pediatric Hodgkin's lymphoma. Pediatr Radiol 34:190-198, 2004
Submitted December 14, 2004; accepted August 9, 2005.(Matthew J. Krasin, Shesh )