Short-Term Thalidomide Incorporated Into Double Autologous Stem-Cell Transplantation Improves Outcomes in Comparison With Double Autotransplantation for Multiple Myeloma

  1. Michele Baccarani
  1. From the Dipartimento di Ematologia e Scienze Oncologiche “Seràgnoli,” Istituto di Ematologia “Seràgnoli,” Università di Bologna; Ematologia, Università di Catania; Ematologia, Università di Udine; Ematologia, Taranto; Ematologia, Avellino; Ematologia, Cagliari; Ematologia, Nocera Inferiore; Ematologia, Ascoli Piceno; Ematologia, Università di Ancona; Ematologia, Treviso; and Ematologia, Università di Genova, Italy.
  1. Corresponding author: Michele Cavo, MD, Istituto di Ematologia “Seràgnoli,” via Massarenti 9, 40138, Bologna, Italy; e-mail: michele.cavo{at}unibo.it.

Abstract

Purpose To assess potential benefits with thalidomide incorporated into double autologous stem-cell transplantation (ASCT) for younger patients with newly diagnosed multiple myeloma (MM).

Patients and Methods One hundred thirty-five patients who received thalidomide from induction until the second ASCT were retrospectively analyzed in comparison with an equal number of pair mates treated with double ASCT not including thalidomide.

Results On an intention-to-treat basis, the addition of thalidomide to double ASCT effected a significant improvement in the rate (68% v 49%; P = .001) and duration (62% v 33% at 4 years; P < .001) of at least very good partial response (VGPR), time to progression (TTP; 61% v 41% at 4 years; P < .001) and progression-free survival (PFS; 51% v 31% at 4 years; P = .001). A trend was also noted for extended overall survival (OS) among thalidomide-treated patients (69% at 5 years v 53% for the control group), although the difference between the two groups was not statistically significant (P = .07). Benefits with thalidomide in increasing the rate of VGPR or better response, TTP, and PFS were confirmed in a multivariate analysis. Median OS after relapse was 24 months for patients receiving thalidomide added to double ASCT and 25 months for the control group. Overall, 17% of patients discontinued thalidomide, including 8% because of drug-related adverse events.

Conclusion In comparison with double ASCT, the addition of first-line thalidomide to double ASCT improved clinical outcomes. Short-term thalidomide was generally well tolerated and had no adverse impact on postrelapse survival.

INTRODUCTION

The introduction into the clinical practice of novel agents targeting myeloma cells in their bone marrow microenvironvement has broadened treatment options for younger patients with multiple myeloma (MM) who are candidates for autologous stem-cell transplantation (ASCT).1,2 Additive efficacy of dexamethasone and thalidomide, the first drug with documented antimyeloma activity since the discovery of melphalan and prednisone,3 formed the basis for the design of phase II Bologna 2002 study incorporating thalidomide and dexamethasone into melphalan-based double ASCT for patients with newly diagnosed MM. To assess potential benefits from this trial design, a case-match comparison with the previous Bologna 96 study of double ASCT not including thalidomide4 was performed. Results reported herein showed that the addition of thalidomide to double ASCT significantly improved clinical outcomes in comparison with the control study. Short-term thalidomide therapy given from induction until the second ASCT was generally well tolerated, as reflected by less than 10% discontinuation rate due to drug-related adverse events, and had no adverse impact on survival after relapse.

PATIENTS AND METHODS

Patients and Criteria of Matching

One hundred thirty-five of 158 patients who were randomly assigned to the double transplant arm of Bologna 96 study4 were retrospectively analyzed with an equal number of pair mates who were enrolled on Bologna 2002 study. Case matching was performed with respect to the following patient characteristics at baseline: age (± 2 years), disease stage according to the Durie-Salmon system, and serum β2-microglobulin levels (± 1 mg/L).

Patient Eligibility

Eligible patients had to have a confirmed diagnosis of symptomatic MM5 and to be previously untreated. The upper age limit with Bologna 96 and Bologna 2002 studies was 60 and 65 years, respectively. Written informed consent was obtained from all patients. Both studies were approved by each local ethics committee of all participating centers and were performed in accordance with the precepts established by Helsinki declaration.

Treatment Regimens

Details of treatment regimens with individual agents and their doses in the two studies are summarized in Table 1. In Bologna 2002 study, thalidomide 200 mg daily was administered from the outset until the second course of ASCT-supported melphalan 200 mg/m2 (MEL-200). Thalidomide was transiently discontinued on the day before administration of both cyclophosphamide (CTX) to collect peripheral blood stem cells (PBSCs) and the first course of ASCT-supported MEL-200. On completion of PBSC harvest and recovery of hematopoiesis after the first ASCT, thalidomide was then resumed. Dexamethasone 40 mg daily for 4 days every 28 days was given in combination with thalidomide. Two added 4-day courses of dexamethasone were administered on days 9 through 12 and 17 through 20 during the first and third cycle of induction therapy.6 Thalidomide and/or dexamethasone dose adjustments were permitted for toxicity at physician's discretion. Instead of four 28-day courses of thalidomide and dexamethasone as induction therapy in Bologna 2002 study, patients enrolled on Bologna 96 trial received four 28-day cycles of vincristine, doxorubicin, and dexamethasone; in addition, they received high-dose therapy with busulfan 12 mg/kg and melphalan 120 mg/m2 (Bu-Mel) before the second ASCT.4 Notably, the planned total dose of dexamethasone during the induction phase, as well as the doses of CTX and melphalan before the first ASCT were identical in the two studies.

Table 1.

Treatment Details of Bologna 2002 and Bologna 96 Studies

Maintenance therapy with recombinant interferon alfa was offered after ASCT(s) to patients enrolled in both clinical trials.

Criteria for Response and Relapse or Disease Progression

Response to therapy was evaluated by means of monoclonal protein concentration and bone marrow plasma cell infiltration. Monoclonal protein evaluations were performed monthly during the entire treatment program and every 3 months during follow-up. The rate of response to the individual treatment phases was assessed after induction therapy, at day +45 after CTX and at day +90 after the first course of MEL-200. For those patients who completed the planned therapy, the maximal response achieved after double ASCT was considered. For patients who were unable to complete the entire treatment program, the maximal response ever achieved was considered. Patients who died before response could be established were considered as experiencing treatment failures. Criteria for response were those proposed by the European Group for Blood and Marrow Transplant,7 with the addition of very good partial response (VGPR)8 category. European Group for Blood and Marrow Transplant criteria were also used to define relapse from complete response (CR) or disease progression.

Toxicity and Statistical Analysis

Adverse events in both study treatments were assessed according to the National Cancer Institute Common Toxicity Criteria, version 2.

Analyses of outcomes of patients enrolled on the two studies were performed on an intention-to-treat basis. Time to progression (TTP) was calculated from the start date of induction therapy to the date of relapse or progression. Progression-free survival (PFS) was defined as the interval between the start of induction therapy and progression or death for any cause, whichever occurred first. Overall survival (OS) and postrelapse survival were measured from the start date of induction therapy or salvage therapy, respectively, until death from any cause. The duration of at least VGPR was computed from the onset of VGPR or better response. Patients lost to follow-up and survivors, including those who experienced no event, were censored at the time of last contact. Differences in the frequency of at least VGPR were evaluated using the χ2 test. Curves of TTP, PFS, OS, and postrelapse survival were plotted according to the Kaplan-Meier method and were compared using the log-rank test. All statistical tests were two tailed. Cox regression analysis was used to examine multivariate models of prognostic factors, including at least VGPR which was evaluated as a time-dependent covariate.

At the cutoff date of this analysis, the median follow-up times for patients enrolled on Bologna 2002 and Bologna 96 studies were 45 months (range, 1 to 74) and 62 months (range, 4 to 133), respectively. The corresponding values for survivors were 52.5 months (range, 29 to 74) months and 102 months (range, 50 to 133), respectively.

RESULTS

Patient Characteristics at Baseline

Baseline characteristics were similar between patients treated on the two studies (Table 2). Comparison between patients enrolled on Bologna 2002 study and a subset of 81 patients treated on Bologna 96 trial for whom baseline cytogenetic data by fluorescence in situ hybridization (FISH) were available, showed similar frequencies of chromosome 13 deletion (46% v 49%, respectively) and t(4;14)(18% v 23%).

Table 2.

Baseline Characteristics of Patients Enrolled on Bologna 2002 and Bologna 96 Studies

Treatment Received

The two groups were comparable with respect to their probabilities of receiving different treatment phases. More details on treatments received and reasons for discontinuing treatment are given in Figure 1. Overall, 60% of patients enrolled on Bologna 2002 study received the two planned autotransplants, as compared with 64% of those included in the previous trial. Time to completion of sequential treatment phases (eg, induction, CTX to collect PBSCs, first and second ASCT) were similar in the two treatment groups (data not shown). Median duration of interferon alfa maintenance therapy was 17 months in Bologna 2002 study and 20 months in the previous Bologna 96 study.

Fig 1.

Patient disposition. CTX, cyclophosphamide; ASCT, autologous stem-cell transplantation; SC, stem cell; SCT, stem-cell transplantation.

Response to Treatment Regimens

The probability to attain VGPR or better response in relationship to the individual treatment phases for all the 270 patients who were enrolled on the two studies is shown in Table 3. In comparison with Bologna 96 study, incorporation of thalidomide and dexamethasone into Bologna 2002 study significantly increased the rate of at least VGPR after induction therapy (15% v 30%; P = .003), CTX (24% v 41%, respectively; P = .002), and the first ASCT (30.5% v 60%, respectively; P < .001). The probability to achieve VGPR or better response after the second ASCT, regardless of whether it was actually received, was 68% for patients enrolled on Bologna 2002 study as compared with 49% for the control group (P = .001).

Table 3.

Response to Therapy in Relationship to the Individual Treatment Phases in Bologna 2002 and Bologna 96 Studies

TTP, Relapse-Free Survival, and OS

On an intention-to treat basis, the addition of thalidomide to double ASCT significantly prolonged both TTP (4-year estimates, 61% v 41% for the control group; P < .001; Fig 2A) and PFS (4-year estimates, 51% v 31% for patients treated on Bologna 96 study; P = .001; Fig 2B). Benefit with incorporation of thalidomide into double ASCT was evident also in terms of extended duration of at least VGPR, the 4-year estimate being 62% as compared with 33% for patients assigned to receive double ASCT alone (P < .001; Fig 2C). Comparison between the two study groups revealed a trend for improved OS with the addition of thalidomide to double ASCT, although the difference did not reach the level of statistical significance (5-year estimates, 69% v 53% for the control group; P = .07; Fig 3A).

Fig 2.

Kaplan-Meier estimates of (A) time to progression (TTP), (B) progression-free survival (PFS), and (C) duration of at least very good partial response (VGPR) in the intention-to-treat populations enrolled onto the two studies. In comparison with the control study, incorporation of thalidomide (Thal) into double autologous stem-cell transplantation significantly prolonged the 4-year estimates of TTP (41% v 61%; P < .001), PFS (31% v 51%; P = .001), and duration of at least VGPR (33% v 62%; P < .001).

Fig 3.

Kaplan-Meier estimates of overall survival (OS) from (A) start of treatment and from (B) relapse or progression in the intention-to-treat populations enrolled onto the two studies. In comparison with the control study, a trend for improved OS was noted with the addition of thalidomide (Thal) to double autologous stem-cell transplantation (ASCT; 5-year estimates, 53% v 69%; P = .07). Postrelapse survival was similar whether patients received Thal added to double ASCT or not (median, 24 v 25 months; P = .2).

Appreciating that the two trials were conducted over different time periods, a multivariate Cox regression analysis of all 270 patients was performed that included the two studies along with baseline features as prognostic variables. The addition of thalidomide to double ASCT was the single most relevant and independent predictor of attainment of VGPR or better response after both induction therapy and double ASCT (P = .001 for each of the two treatment phases). Variables significantly related to extended TTP and PFS included incorporation of thalidomide into double ASCT and baseline platelet count greater than 150,000/μL (Table 4). Attainment of at least VGPR and a value of serum β2-microglobulin at diagnosis lower than 3.5 mg/L were additional good risk factors for PFS (Table 4). Prolonged OS was significantly related to the following independent variables (Table 4): achievement of VGPR or better response, hemoglobin higher than 10 g/dL, monoclonal protein immunoglobulin G isotype, and age younger than 60 years.

Table 4.

Multivariate Analyses of Prognostic Variables Associated With Clinical Outcomes in Bologna 2002 and Bologna 96 Studies

Thalidomide-Related Toxicity

Grade 3 to 4 adverse events seen in more than 5% of patients who were treated with thalidomide from the outset until the second ASCT included venous thromboembolism (VTE) (13%), constipation (10%), infections (8%), fatigue (7%), and peripheral neuropathy (4%). All episodes of VTE were registered during induction therapy. Overall, 23 patients (or 17%) discontinued thalidomide, 11 of them due to adverse events (early death = 4; VTE = 3; neuropathy = 3; skin rash = 1); in the remaining patients additional causes of thalidomide discontinuation included disease progression (n = 4), patient refusal (n = 4), protocol violation (n = 3), or inadequate stem-cell collection (n = 1).

Treatments and Outcomes After Relapse

The proportion of patients rescued with novel agents after post-transplantation relapse or progression was comparable in the two studies. In particular, regimens including either thalidomide or bortezomib were offered to 48% of patients on Bologna 2002 study and 53% of patients accrued on Bologna 96 trial. Median duration of postrelapse survival was 24 months for patients on Bologna 2002 study and 25 months for the control study (Fig 3B).

DISCUSSION

Recognition of marked efficacy of thalidomide in all phases of MM3,6,9,10 provided the rationale for Bologna 2002 study incorporating this agent from induction until the second ASCT in newly diagnosed MM patients younger than 66 years of age. To evaluate whether the addition of thalidomide to double ASCT improved outcomes, we performed a case-match comparison with the previous Bologna 96 study of double ASCT not including thalidomide.4 Since the analysis was retrospective and potentially susceptible to time and treatment bias, an effort was undertaken to account for differences in relevant prognostic features by matching patients closely for age, disease stage according to the Durie-Salmon system, and serum β2-microglobulin. Additional prognostic variables, including the presence of chromosome 13 deletion and t(4;14), were also similar in patients treated on the two studies, although an exhaustive cytogenetic comparison between the two groups was precluded by lack of data in approximately one third of patients accrued to Bologna 96 study.

Recognizing that Bologna 2002 study was designed and conducted after Bologna 96 trial, outcome comparisons between the two study groups should be cautiously interpreted. On an intention-to-treat basis, the rate of at least VGPR was significantly higher for patients treated with thalidomide. This benefit was observed following induction therapy, a finding not recognized in our initial report4 but subsequently confirmed in a larger series of patients who entered the study,11 as well as after postinduction treatment phases, including ASCT(s). Overall, in comparison with double ASCT not including thalidomide, patients enrolled on Bologna 2002 study had an absolute increase of 20% in the final probability to achieve VGPR or better response, regardless of whether the entire treatment program of double ASCT was actually received. Since the positive effects of the incorporation of thalidomide into double ASCT were evident along all treatment phases, it is unlikely that superior outcomes observed with Bologna 2002 study in comparison with the control group were linked to the different regimens used in preparation for the second ASCT (eg, MEL-200 in Bologna 2002 study and Bu-Mel in Bologna 96 study). Results of a French trial designed to prospectively compare thalidomide and dexamethasone with vincristine, doxorubicin, and dexamethasone as induction therapy in preparation for ASCT12 confirm the superior rate of at least VGPR reported in the present analysis with thalidomide and dexamethasone. However, differently from the French study, in our trial prolonged administration of thalidomide offered a gain in terms of increased VGPR or better response also after ASCTs, a finding consistent with results of Total Therapy 2 incorporating into double ASCT thalidomide given from the outset until progression or undue toxicity.13

Importantly, the superior rate of VGPR or better response effected by thalidomide added to double ASCT translated into a significant increase in the 4-year estimate of TTP (61% v 41% for the control group), PFS (51% v 31%, respectively), and duration of at least VGPR (62% v 33%, respectively). Beneficial effects of thalidomide and double ASCT in terms of increased TTP, PFS, and duration of VGPR were further confirmed in a multivariate analysis that also identified attainment of VGPR or CR as the dominant prognostic factor for prolonged OS. These data support the notion that achievement of CR is a relevant early trial end point for improved outcomes after ASCT.1416 Efforts to further increase the rate of CR are currently being focused on the use of the novel agents thalidomide, bortezomib, and lenalidomide as part of induction therapy and/or consolidation/maintenance therapy after autotransplantation.2,13,1723

In Bologna 2002 study, a trend was noted for improved OS in comparison with the control group (5-year rates: 69% v 53%, respectively), although the difference did not reach the level of statistical significance. Similar results were initially reported with Total Therapy 2 and were attributed to the shorter survival after relapse in the thalidomide arm of the study, due to a higher failure rate with salvage therapy.13 In our trial, short-term exposure to thalidomide was not related to increased tumor resistance at the time of relapse. Indeed, the estimated 2-year OS rate from the time of relapse was 49% for patients treated on Bologna 2002 study as compared with 50% for the control group. Possible reasons for the lack of a survival gain for patients treated on Bologna 2002 study may be threefold. The first, but simplest, is that sample size was not powered to detect a statistically significant difference between the two study groups. Secondly, the actual follow-up of patients enrolled on Bologna 2002 trial might not permit as yet to recognize a late divergence between survival curves, as recently reported by Barlogie et al20 with a median follow-up of 6 years and limited to patients with cytogenetic abnormalities. Thirdly, benefit with novel agents for the treatment of post-transplant relapse2428 may have minimized the potential survival gain offered by incorporation of thalidomide into double ASCT.

Because of concern about toxicity, mainly neurological,29 related to long-term thalidomide administration,30 in Bologna 2002 trial the intended duration of thalidomide therapy was approximately 8 months. Differently from previously reported studies in which long-term exposure to thalidomide (eg, in excess of 1 year) led to drug discontinuation in a range between 40% and more than 60%10,13,31,32 of patients, in our trial fewer than 10% of patients discontinued thalidomide due to drug-related adverse events. Similar values were observed in a study of thalidomide given as maintenance therapy for 6 months after a single ASCT.33 Questions remain about the optimal timing of thalidomide therapy, its duration, as well as the selection of patients who mostly benefit from this agent. In Total Therapy 2, both the duration of thalidomide therapy and the cumulative dose that was actually received had no impact on OS, suggesting that a shorter exposure to the drug might result in similar outcomes, while sparing toxicities and avoiding the possibility of increased tumor resistance at the time of relapse.13 In that study, patients with abnormal metaphase cytogenetics who were random assigned to receive thalidomide had a significant longer duration of both OS and CR than comparable patients not treated with thalidomide.20 In contrast, in another phase III study comparing no maintenance therapy with thalidomide given as maintenance therapy, no benefit with this novel agent was seen for patients with chromosome 13 deletion.11 Differences between these studies with respect to the methods used to evaluate genetic abnormalities (eg, conventional cytogenetics or gene expression profile v FISH) may explain these controversies.

In conclusion, data from Bologna 2002 study show that the addition of thalidomide to double ASCT effected superior frequency and duration of at least VGPR, TTP, and PFS in comparison with double ASCT not incorporating thalidomide, a benefit confirmed also in a multivariate Cox regression analysis. Short-term exposure to thalidomide, as planned in this study, was associated with a low toxicity rate and, importantly, had no adverse impact on postrelapse survival.13,34 Incorporation of the novel agents thalidomide, bortezomib, and lenalidomide into ASCT has produced a dramatic shift in the treatment paradigm of patients with MM.35 Newer induction regimens with various combinations of these drugs are now offering increased rates of CR and VGPR in comparison with conventional treatments,2,17,18 a benefit which is further enhanced by autotransplant(s) and, importantly, translates into prolonged TTP, PFS, and OS.2,13,17,18,20,23 Whether these novel agents may delay or challenge ASCT in younger patients with MM is a still unresolved issue which should be properly addressed in the context of prospective, randomized clinical trials.

AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

The author(s) indicated no potential conflicts of interest.

AUTHOR CONTRIBUTIONS

Conception and design: Michele Cavo, Elena Zamagni, Patrizia Tosi, Michele Baccarani

Provision of study materials or patients: Francesco Di Raimondo, Francesca Patriarca, Antonio Francesco Casulli, Silvestro Volpe, Antonio Ledda, Catello Califano, Catia Bigazzi, Massimo Offidani, Piero Stefani, Filippo Ballerini

Collection and assembly of data: Elena Zamagni, Paola Tacchetti, Giulia Perrone, Michela Ceccolini, Annamaria Brioli

Data analysis and interpretation: Michele Cavo, Elena Zamagni, Paola Tacchetti, Mauro Fiacchini, Antonio de Vivo, Patrizia Tosi

Manuscript writing: Michele Cavo

Final approval of manuscript: Michele Cavo, Elena Zamagni, Patrizia Tosi, Michele Baccarani

Acknowledgment

We thank Katia Vitali for valuably supporting data collection and analysis.

Appendix

The following is a list of additional investigators who participated in the Bologna 96 and Bologna 2002 studies.

Bologna 96 study.

Sante Tura, R.M. Lemoli, M.R. Motta, S. Rizzi, (Istituto di Ematologia ed Oncologia Medica “Seràgnoli,“ Bologna); R. Fanin (Ematologia, Udine); R. Giustolisi (Ematologia, Catania); N. Cantore (Ematologia, Avellino); L. Gugliotta, L. Masini (Ematologia, Reggio Emilia); G. Broccia, E. Angelucci (Ematologia, Cagliari); G. Torelli, F. Narni (Ematologia, Modena); B. Rotoli (Ematologia, Napoli); F. Dammacco, V.M. Lauta (Scienze Biomediche, Bari); F. Lauria (Ematologia, Siena); L. Cavanna, Giuseppe Civardi (Medicina, Piacenza); P. Leoni (Ematologia, Ancona); P. Gentilini (Oncologia-Ematologia, Forlì); M. Gobbi (Ematologia, Genova); M. Longinotti, F. Dore (Ematologia, Sassari); L. Guardigni, S. Pasini (Ematologia, Cesena); F. Ricciuti (Ematologia, Potenza); D. Vertone (Ematologia, Potenza); P. Mazza (Ematologia, Taranto); M.G. Michieli (Oncologia Medica, Aviano); R. Battista (Ematologia, Chioggia); D. Mamone (Ematologia, Messina); G. Lalli (Semeiotica Medica, Teramo); A. Zaccaria, A.L. Molinari (Ematologia, Ravenna); A. Polacco (Ematologia, Mestre); C. Ciabatta (Ematologia, Latina); V. Forcellini (Medicina, Repubblica di San Marino); M. Monaco, C. Ferrandina (Ematologia, Foggia); F. Gherlinzoni (Ematologia, Treviso); L. Castagna (Istituto Humanitas, Rozzano); C. Musolino (Ematologia, Messina); E. Pasquini (Oncologia, Cattolica); F. Russo (Ematologia-Oncologia, Napoli); A. Abbadessa (Ematologia, Napoli); A. Bononi (Oncologia, Rovigo); P.P. Fattori, M. Imola (Oncologia-Ematologia, Rimini); A. Corvetta (Medicina Interna, Rimini).

Bologna 2002 study.

R.M. Lemoli, M.R. Motta, S. Rizzi, (Istituto di Ematologia ed Oncologia Medica “Seràgnoli,” Bologna); R. Giustolisi (Ematologia, Catania); R. Fanin (Ematologia, Udine); P. Mazza (Ematologia, Taranto); B. Rotoli, L. Catalano (Ematologia, Napoli); N. Cantore (Ematologia, Avellino); G. La Nasa (Ematologia/CTMO, Cagliari); L. Gugliotta, L. Masini (Ematologia, Reggio Emilia); E. Angelucci, A. Carubelli (Ematologia, Cagliari); A. D'Arco (Ematologia, Nocera Inferiore); P. Galieni (Ematologia, Ascoli Piceno); F. Lauria, A. Gozzetti (Ematologia, Siena); A. Zaccaria, C. Cellini (Ematologia, Ravenna); G. Torelli, F. Narni (Ematologia, Modena); F. Dammacco (Science Biomediche, Bari); P.P. Fattori, E. Pasquini (Ematologia-Oncologia, Rimini-Cattolica); F. Gherlinzoni (Ematologia, Treviso); G. Visani, G. Leopardi (Ematologia, Pesaro); S. Ronconi, D. Cangini (Ematologia, Forlì); L. Cavanna, A. Lazzaro (Ematologia, Piacenza); P. Leoni (Ematologia, Ancona); D. Vertone (Ematologia, Potenza); R. Canistro (Chioggia); L. Guardigni, S. Pasini (Ematologia, Cesena); M. Brugiatelli, D. Mamone (Ematologia, Messina).

Footnotes

  • Supported in part by Università di Bologna, Ricerca Fondamentale Orientata (M.C.), Fondazione Carisbo, and BolognAIL.

  • Written on behalf of Bologna 96 and Bologna 2002 clinical study groups.

  • A list of additional investigators who participated in Bologna 96 and Bologna 2002 studies is reported in the online-only Appendix.

  • Authors' disclosures of potential conflicts of interest and author contributions are found at the end of this article.

  • Clinical trial information can be found for the following: NCT00378222.

  • Received February 24, 2009.
  • Accepted May 21, 2009.

REFERENCES

| Table of Contents
  • Advertisement
  • Advertisement
  • Advertisement