- © 2011 by American Society of Clinical Oncology
Long-Term Follow-Up Analysis of HD9601 Trial Comparing ABVD Versus Stanford V Versus MOPP/EBV/CAD in Patients With Newly Diagnosed Advanced-Stage Hodgkin's Lymphoma: A Study From the Intergruppo Italiano Linfomi
- Teodoro Chisesi⇓,
- Monica Bellei,
- Stefano Luminari,
- Antonella Montanini,
- Luigi Marcheselli,
- Alessandro Levis,
- Paolo Gobbi,
- Umberto Vitolo,
- Caterina Stelitano,
- Vincenzo Pavone,
- Francesco Merli,
- Marina Liberati,
- Luca Baldini,
- Roberto Bordonaro,
- Emanuela Anna Pesce and
- Massimo Federico
- Teodoro Chisesi, Ospedale Civile S.S. Giovanni e Paolo, Venezia; Monica Bellei, Stefano Luminari, Antonella Montanini, Luigi Marcheselli, Emanuela Anna Pesce, and Massimo Federico, Università di Modena e Reggio Emilia, Modena; Alessandro Levis, A.O. S.S. Antonio e Biagio, Alessandria; Paolo Gobbi, Università di Pavia, Istituto Di Ricovero e Cura a Carattere Scientifico (IRCCS) Policlinico San Matteo, Pavia; Umberto Vitolo, Azienda Ospedaliera e Universitaria San Giovanni Battista, Torino; Caterina Stelitano, Azienda Ospedaliera “Bianchi-Melacrino-Morelli,” Reggio Calabria; Vincenzo Pavone, Ospedale G. Panico, Tricase, Lecce; Francesco Merli, Azienda Ospedaliera Arcispedale S. Maria Nuova, Reggio Emilia; Marina Liberati, Policlinico Monteluce, Perugia; Luca Baldini, Università di Milano, Fondazione Ospedale Maggiore PoMaRe, IRCCS, Milano; and Roberto Bordonaro, Ospedale Vittorio Emanuele, Catania, Italy.
- Corresponding author: Teodoro Chisesi, MD, Unità Operativa di Ematologia, Ospedale dell'Angelo AUSL 12 Venezia-Mestre-via Paccagnella, 11, 30174 Mestre-Venezia, Italy; e-mail: OCVE.ematologia{at}ulss12.ve.it.
-
Presented in part at the 50th Annual Meeting and Exposition of the American Society of Hematology, December 6-9, 2008, San Francisco, CA.
Abstract
Purpose The Intergruppo Italiano Linfomi HD9601 trial compared doxorubicin, bleomycin, vinblastine, and dacarbazine (ABVD) versus doxorubicin, vinblastine, mechloretamine, vincristine, bleomycin, etoposide, and prednisone (Stanford V [StV]) versus the combination of mechlorethamine, vincristine, procarbazine, prednisone (MOPP) with epidoxorubicin, bleomycin, vinblastine (EBV), lomustine, doxorubicin, and vindesine (CAD) (MOPP/EBV/CAD [MEC]) for the initial treatment of advanced-stage Hodgkin's lymphoma to select which regimen would best support a reduced radiotherapy program (limited to two or fewer sites of either previous bulky or partially remitting disease). Superiority of ABVD and MEC to StV was demonstrated. We report analysis of long-term outcome and toxicity.
Patients and Methods Patients with stage IIB, III, or IV were randomly assigned among six cycles of ABVD, three cycles of StV, and six cycles of MEC; radiotherapy was administered in 76, 71, and 50 patients in the three arms, respectively.
Results Currently, the median follow-up is 86 months; in the prolonged observation period, eight additional failures, including two relapses, both in the StV arm, and six additional deaths in complete response were recorded. The 10-year overall survival rates were 87%, 80%, and 78% for ABVD, MEC, and StV, respectively (P = .4). The 10-year failure-free survival was 75%, 74%, and 49% in the ABVD, MEC, and StV arms, respectively (P < .001). The 10-year disease-free survival of patients treated or not with radiotherapy (RT) showed no difference for ABVD or MEC (85% v 80% and 93% v 68%), and a statistically significant difference for StV (76% v 33%; P = .004). No significant long-term toxicity was recorded.
Conclusion The long-term analysis confirmed ABVD and MEC superiority to StV. The use of RT after StV was established as mandatory. ABVD is still to be considered as the standard treatment with a good balance between efficacy and toxicity.
INTRODUCTION
Hodgkin's lymphoma (HL) is one of the most treatable adult cancers, with long-term cure rates of more than 80% achieved even in patients with advanced disease. The combination of doxorubicin, bleomycin, vinblastine, and dacarbazine (ABVD) is currently considered the standard of care for HL worldwide.
Several trials have been conducted to identify the therapy with the best risk-to-benefit ratio, thus increasing the efficacy of treatment while minimizing the risk of late toxicity. Many of the tested chemotherapy combinations, which were derived from the regimen that was standard in the 1970s and 1980s (mechloretamine, vincristine, procarbazine, and prednisone [MOPP]) and/or from the standard regimen used in the 1980s (ABVD), were compared with various modified schedules of these well-tested models or regimens originally devised for advanced-stage disease.1
Other schedules that include multiple drugs have also demonstrated marked effectiveness in advanced disease, with 5-year failure-free survival (FFS) rates ranging from 72% to 89%.2,3 The Stanford V (StV) regimen (doxorubicin, vinblastine, mechloretamine, vincristine, bleomycin, etoposide, and prednisone), by virtue of continuous exposure to active drugs on a weekly basis, was a way to shorten the duration of chemotherapy to only 3 months.4,5 Moreover, a series of studies was designed to test the combination of mechlorethamine, vincristine, procarbazine, prednisone (MOPP) with epidoxorubicin, bleomycin, vinblastine, lomustine, doxorubicin, and vindesine (MOPP/EBV/CAD [MEC]), an early attempt to intensify and hybridize some well-known and active available chemotherapy regimens to increase their efficacy.6–8 Both MEC and StV produced particularly interesting results.
In 2005, we published the final results of the Intergruppo Italiano Linfomi HD9601 trial on the treatment of advanced-stage HL, comparing six cycles of ABVD versus six cycles of MEC versus 12 weeks (three cycles) of StV.9
PATIENTS AND METHODS
Patients
Previously untreated patients age 16 to 65 years with a biopsy-proven diagnosis of HL and clinical stage IIB, III, or IV disease were eligible for inclusion (Fig 1). Normal cardiac, pulmonary, hepatic, and renal functions were required unless any abnormality was considered to be directly related to the HL. Pregnant or lactating women and patients with a history of malignancy, a positive HIV test, or an Eastern Cooperative Oncology Group (ECOG) performance status greater than 3 were ineligible. Patients with anaplastic, CD30+, large-cell lymphoma were also excluded from the study.
CONSORT diagram. ABVD, doxorubicin, bleomycin, vinblastine, and dacarbazine; MEC, MOPP/EBV/CAD (mechlorethamine, vincristine, procarbazine, and prednisone [MOPP], with epidoxorubicin, bleomycin, and vinblastine [EBV], and lomustine, doxorubicin, and vindesine [CAD]); Stanford V, doxorubicin, vinblastine, mechloretamine, vincristine, bleomycin, etoposide, and prednisone; Tx, treatment.
All patients were clinically staged according to the Cotswold modifications of the Ann Arbor criteria.10 Bulky disease was defined as a mass with a diameter of ≥ 6 cm according to a definition rather more restrictive for therapeutic purposes than that given at the Cotswold meeting.
This study fully complied with all provisions of the Declaration of Helsinki and was conducted in accordance with Good Clinical Practice rules. All patients signed a written informed consent to participate.
Patients were randomly assigned to receive ABVD, MEC, or StV. Drug doses, time schedules, and study procedures are provided in our previous report.9 At the end of chemotherapy, only responding patients were administered radiotherapy (RT) at doses and with modalities summarized in Table 1, which also reports differences with respect to the original Stanford V protocol.
Comparison Between the Stanford V Original Protocol and the Radiotherapy Guidelines Adopted in the HD9601 Study
Statistical Methods
MEC and StV were considered the experimental regimens and were tested against ABVD, primarily in terms of FFS. An intention-to-treat analysis was performed that included all randomly assigned patients except for one who had inconsistent histology. FFS was defined from the start of treatment to last follow-up or to one of the following events: death from any cause, disease progression during treatment, or any response other than complete response (CR) at the end of therapy or relapse. Additional outcome measures were overall survival (OS), computed from the date of diagnosis to the date of last observation or death from any cause; relapse-free survival, measured for complete responders from the date of therapy completion to the date of last observation or relapse; and freedom from progression (FFP), calculated from the start of treatment to disease progression or relapse or to the last follow-up. By using the subset of patients who had attained CR or CR unconfirmed (CRu) after chemotherapy, we explored the role of RT by means of disease-free survival (DFS) measured from the date of end of chemotherapy to the date of relapse or death from any cause or date of last follow-up for censored patients. Late toxicity was assessed according to the intention-to-treat analysis.
RESULTS
Between January 1996 and April 2000, 355 patients were registered. Of the 355, one was rejected for inconsistent histology; 19 patients were excluded from analysis (eight patients withdrew their consent or emigrated to another country, and 11 had no data but the information given at random assignment), allowing assessment of clinical response and toxicity on 335 patients. The main characteristics of the 335 assessable patients are listed in Table 2. The MEC arm contained a lower percentage of patients with stage IIB disease, mediastinal bulk, and nodular sclerosis histology type than the other two arms, but it included a higher proportion of patients with stage III disease.
Patient Characteristics (n = 335)
At the end of the treatment program, CR rate was 89% with ABVD, 94% with MEC, and 76% with StV. Currently, median follow-up for the entire group is 86 months (range, 7 to 146 months).
The 10-year OS was 87% (95% CI, 77% to 92%) for patients treated with ABVD, 80% (95% CI, 67% to 88%) for those treated with MEC, and 78% (95% CI, 68% to 85%) for those treated with StV. After an extended follow-up, no significant difference emerged among the three arms in terms of OS (P = .4; Fig 2A), with a plateau obtained for all three groups.
Kaplan-Meier analysis. Estimates of 10-year (A) overall survival, (B) failure-free survival, (C) freedom from progression, and (D) relapse-free survival by treatment arm. ABVD, doxorubicin, bleomycin, vinblastine, and dacarbazine; MEC, MOPP/EBV/CAD (mechlorethamine, vincristine, procarbazine, and prednisone [MOPP], with epidoxorubicin, bleomycin, and vinblastine [EBV], and lomustine, doxorubicin, and vindesine [CAD]); StV (Stanford V), doxorubicin, vinblastine, mechloretamine, vincristine, bleomycin, etoposide, and prednisone.
The 10-year FFS was 75% (95% CI, 65% to 82%) in the ABVD arm, 74% (95% CI, 61% to 83%) in the MEC arm, and 49% (95% CI, 39% to 58%) in the StV arm (P < .001; Fig 2B).
The 10-year FFP rates were 84% (95% CI, 75% to 90%) in the ABVD arm, 90% (95% CI, 81% to 95%) in the MEC arm, and 68% (95% CI, 58% to 76%) in the StV arm (P < .001; Fig 2C). Difference in FFP between StV and both other arms remains smaller with respect to that shown for FFS after an extended observation.
This analysis confirms that MEC and ABVD better maintain CRs compared with StV (10-year relapse-free survival, 92%, 88%, and 67%, respectively) because of two additional relapses reported during the prolonged follow-up period in the StV arm (Fig 2D).
Eight additional failures were observed, including two relapses in the StV arm and six additional deaths in CR (three in the ABVD arm, two in the MEC arm, and one in the StV arm; Table 3).
Events Recorded in Patients Achieving at Least a PR After Chemotherapy (n = 312)
The outcome of patients who received RT and of those who did not was evaluated in the cohort of patients achieving CR/CRu after chemotherapy and was compared in the three arms (Fig 3). No difference was shown for ABVD or MEC in terms of 10-year DFS (85% v 80% [P = .79] and 93% v 68% [P = .057] with and without RT, respectively). On the contrary, a statistically significant difference in 10-year DFS between patients who underwent RT and those who did not was demonstrated for patients treated with StV (76% v 33%; P = .004; Fig 3).
Disease-free survival of the patients treated with (A) ABVD (doxorubicin, bleomycin, vinblastine, and dacarbazine); (B) MEC (MOPP/EBV/CAD, mechlorethamine, vincristine, procarbazine, and prednisone [MOPP], with epidoxorubicin, bleomycin, and vinblastine [EBV], and lomustine, doxorubicin, and vindesine [CAD]); or (C) Stanford V (doxorubicin, vinblastine, mechloretamine, vincristine, bleomycin, etoposide, and prednisone) in complete response or complete response unconfirmed, after chemotherapy, with or without consolidation radiotherapy (RT).
No difference in late toxicity emerged among the three groups: six late toxicity events were reported, two for each treatment arm, consisting of four metachronous tumors (follicular non-Hodgkin's lymphoma, grade 3b, ABVD arm; B-cell chronic lymphocytic leukemia, StV arm; and prostatic adenocarcinoma and epidermoid carcinoma infiltrating the tonsil, MEC arm), one acute autoimmune thrombocytopenic purpura (ABVD arm), and one late event not otherwise specified (StV arm). No cases of late cardiac toxicity have been reported, but on the whole, five events affecting pulmonary function and leading to fatal outcome occurred.
Death as a result of any reason occurred in 15 (12%), 16 (14%), and 21 (18%) patients treated with ABVD, MEC, and StV, respectively. Among the 10 deaths reported during the extended follow-up period, five were recorded in the ABVD arm (two due to lymphoma 4 and 5 years after the patients completed treatment, two due to pulmonary embolism, and one due to acute myeloid leukemia), two in the MEC arm (one due to pulmonary embolism and one to generalized sepsis), and three in the StV arm (two due to lymphoma 5 years after treatment for both patients, and one to pulmonary embolism).
Late failures are equally distributed among the three treatment arms. Overall, 29 failures were recorded in the ABVD arm (13 early treatment failures, 12 relapses, and four deaths in CR), 20 in the MEC arm (three early treatment failures, seven relapses, and 10 deaths in CR), and 55 in the StV arm (26 early treatment failures, 25 relapses, and four deaths in CR).
DISCUSSION
The main aim of the trial started in 1996 by the Intergruppo Italiano Linfomi was to evaluate which chemotherapy regimen could most favorably be combined with a reduced RT program to decrease toxicity without limiting the efficacy of combined therapy in terms of OS and FFS and to decrease the incidence of second tumors. At that time, MEC and StV were considered the new promising alternatives to be tested against ABVD.
The three regimens were different in dose intensity, duration of treatment, and number of drugs. The choice of ABVD as the reference chemotherapy regimen for advanced HL was mandatory given the large amount of data confirming its relative best efficacy/toxicity ratio. The StV regimen was attractive because of its weekly administration of drugs, according to the theory of Goldie and Coldman,11 and the short duration of the treatment. The MEC chemotherapy represented another way of testing the theory of Goldie and Coldman; it had already yielded interesting results and allowed a strong reduction in RT administration.
The criteria for irradiation that we used in the three arms of this trial were more permissive than those in the original Stanford protocol4 but more restrictive than those in the original MEC study,6 in which RT was administered to only 28% of patients; the hypothesis was that omitting RT should be possible in case of more intense and effective chemotherapeutic regimens without compromising the efficacy in terms of overall response rate or FFS.
In the analysis previously published,9 the FFS was significantly lower for patients treated with StV (54%) compared with those treated with ABVD (78%) and MEC (81%); RT, which was not mandatory, was given in 66% and 60% of patients who were treated with StV and ABVD, respectively, and in only 44% of patients treated with MEC. OS was 90%, 89%, and 82% for ABVD, MEC, and StV, respectively. At that time, the relapses were well distributed in the three groups. These results appeared to be confusing for the lower rate of response in the StV arm. However, our study, to the best of our knowledge, was the first attempt to test the StV regimen in a randomized trial outside Stanford, and the results still need to be confirmed by longer observation and other experiences.
One of the secondary end points of the study was to evaluate the different toxicity among the three regimens, considering the different schedules and dose intensities of the drugs. No differences in terms of toxicity were reported between the three arms after a prolonged follow-up period, but the final results of the trial showed a significant difference in acute hematologic toxicity among the three arms (P < .01), with ABVD being the safest regimen compared with moderate toxicity in terms of anemia for StV and substantial toxicity regarding the white blood cells, platelet counts, and infections for MEC.9 We can confirm that, on the whole, ABVD was better tolerated than StV, which in turn was better tolerated than MEC, thus confirming that ABVD seemed to offer the best and most reliable balance between efficacy and toxicity.
We concluded that, under the study conditions, ABVD and MEC were superior in terms of response rate and FFS, although the OS rates of all three regimens were equivalent because of effective salvage therapies.
At present, only one other recently published large, randomized study12 has compared ABVD with StV. In this multicenter experience, after 5 years of follow-up, the results seemed to be different from those in our study, with no differences between the two regimens in terms of progression-free survival and OS. The estimated 5-year progression-free survival was 76% for ABVD and 74% for StV; OS was superimposable, although the trial was designed to confirm an improvement in outcome after StV therapy, as supported by results from a previous single-institution, single-arm study.5,13 However, ABVD was superior to StV in terms of CR/CRu rate (55% v 36%; P < .001), with less partial responses at completion of chemotherapy (35% v 58%), and with significant difference persisting at the completion of the whole therapeutic program (including RT in 53% and 73% of patients receiving ABVD and StV, respectively), with a CR/CRu rate of 67% and 57% (P = .036) at the end of treatment. Finally, the authors' conclusions were the same as those in our initial study, and they included a recommendation for RT for patients treated with StV who could have some detriment from omission of RT.
The results of this analysis compared with those in our previous reports9,14 indicate that FFS is affected, for the most part, by early treatment failures and by relapses within the first 3 years after completion of therapy, although late events have no influence on this end point. As shown in Figure 2, the curves for ABVD and MEC tend to overlap over time, although StV shows lower rates compared with rates in both other arms, depending on the excess of early failures but maintaining the same shape of the curves.
As already hypothesized,9,14 the different use of RT in our study could explain the lower response rate in the StV arm. However, it is possible that differences in the application of RT in our trial accounted for some of the differences in the outcomes of patients treated with StV. In fact, the original StV protocol was developed as a combined modality treatment with RT considered as part of the regimen, although in our study, we used StV mainly as a chemotherapy regimen with RT delivered only to residual masses or to lesions initially greater than 6 cm. Our adaptation of StV resulted in only 66% of the patients undergoing RT. Among the 107 patients randomly assigned to the Stanford V arm, 12 were given salvage treatment for less than partial response at intermediate restaging or because of treatment toxicity. Of the remaining 95 patients responding to chemotherapy, 71 received RT and 24 did not on the basis of the established guidelines. According to the original Stanford protocol, 16 of 24 patients not treated with RT should have undergone RT—eight because they had a bulky tumor of at least 5 cm, seven because they presented with more than two disease sites at diagnosis, and one because of spleen involvement (stage IIIs). The role of RT in affecting the outcome of patients in terms of DFS was analyzed in the three arms: although our analyses were underpowered, they showed that consolidative RT provided a strong benefit in the StV arm, a moderate benefit in the MEC arm, and no benefit in the ABVD arm. Similar results are described in a study15 that reported the experience on a small number of patients treated with StV and compared with a historical series of patients treated with ABVD. Although the number of patients was small, there was a striking difference in the results obtained with StV, with an estimated 3-year freedom from treatment failure of 66% with respect to 76% of the patients in the ABVD arm, noting that only 44% of patients received RT after StV. The authors also reported that in the StV group, treatment failures were mostly due to disease progression during and soon after completion of chemotherapy.15
Considering the differences among the different regimens, the intensified MEC regimen was expected to have the major impact on the overall response rate, but in our study, no differences were reported when the MEC arm was compared with the ABVD arm. Having two groups of patients with similarities in terms of CR rate at completion of the protocol treatment (94% v 89%) and FFP at 10 years (90% v 84%) suggests that it is not the intensification of therapy but the risk category of treated patients that makes the difference. From our data, it is impossible to draw a conclusion about the superiority of one regimen compared with the remaining two tested in our study. The same conclusions are reported by the authors of the English LY09 study16: from their data, ABVD still seems to represent the standard initial treatment regimen for advanced HL.
The second point is that the randomized study suggests that patients with a CR after multiagent chemotherapy do not benefit from consolidation RT. This does not seem to be applicable to the StV regimen, and patients with initial site of bulky disease need mandatory RT, as suggested in the original Stanford series5 and also confirmed by a recent retrospective analysis conducted at the Memorial Sloan-Kettering Cancer Center17 on a series of patients with advanced or locally extensive HL to whom the Stanford protocol was strictly applied. However, this study hypothesizes that for patients with stage IV disease and/or International Prognostic Score ≥ 4, a more aggressive treatment should be adopted.
As a result of a series of investigations performed in the last 5 years,18–22 the use of positron emission tomography emerged as a promising predictive tool for therapy outcome at an early evaluation; positron emission tomography could help us tailor therapy and limit the use of RT or more intensive regimens to a subset of patients who do not respond to two courses of standard treatment, resulting in a best risk-adapted treatment. Many trials are now under way to validate this hypothesis.
In conclusion, the 10-year follow-up in the HD9601 study confirms the results that were initially published, and the small number of additional registered events equally distributed in the three study groups does not modify the conclusion of substantial similarity of the three chemotherapeutic regimens as far as long-term OS is concerned. We would like to stress that in our trial, we adopted a reduced version of the RT used in the original StV protocol. In our opinion, the long-term analysis of the results of the HD9601 trial suggests that ABVD still remains the treatment with the best balance between efficacy and toxicity.
AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTSOF INTEREST
The author(s) indicated no potential conflicts of interest.
AUTHOR CONTRIBUTIONS
Conception and design: Teodoro Chisesi, Alessandro Levis, Paolo Gobbi, Massimo Federico
Provision of study materials or patients: Teodoro Chisesi, Stefano Luminari, Alessandro Levis, Paolo Gobbi, Umberto Vitolo, Caterina Stelitano, Vincenzo Pavone, Francesco Merli, Marina Liberati, Luca Baldini, Roberto Bordonaro, Massimo Federico
Collection and assembly of data: Monica Bellei, Antonella Montanini, Emanuela Anna Pesce
Data analysis and interpretation: Teodoro Chisesi, Monica Bellei, Stefano Luminari, Luigi Marcheselli, Alessandro Levis, Paolo Gobbi, Massimo Federico
Manuscript writing: Teodoro Chisesi, Monica Bellei, Stefano Luminari, Antonella Montanini, Luigi Marcheselli, Alessandro Levis, Paolo Gobbi, Umberto Vitolo, Caterina Stelitano, Vincenzo Pavone, Francesco Merli, Marina Liberati, Luca Baldini, Roberto Bordonaro, Emanuela Anna Pesce, Massimo Federico
Final approval of manuscript: Teodoro Chisesi, Monica Bellei, Stefano Luminari, Antonella Montanini, Luigi Marcheselli, Alessandro Levis, Paolo Gobbi, Umberto Vitolo, Caterina Stelitano, Vincenzo Pavone, Francesco Merli, Marina Liberati, Luca Baldini, Roberto Bordonaro, Emanuela Anna Pesce, Massimo Federico
Acknowledgment
Supported by grants from the Ministero dell'Universitá e della Ricerca Scientifica e Tecnologica, Roma; the Associazione “A. Serra,” Modena; the Istituto di Ricovero e Cura a Carattere Scientifico Policlinico, S. Matteo; and the Ferrara-Storti Foundation, Pavia, Italy.
Footnotes
- Received June 29, 2010.
- Accepted March 14, 2011.
