Patients with high-risk aggressive lymphoma treated with frontline intensive chemotherapy and autografting

Evidence of marked differences in outcome between patients with age-adjusted International Prognostic Index scores 2 and 3

Authors

  • Alessandra Cuttica MD,

    Corresponding author
    1. Dipartimento di Medicina Oncologia Sperimentale, Divisione Universitaria di Ematologia, Azienda Ospedaliera San Giovanni Battista, Turin, Italy
    • Divisione Universitaria di Ematologia, Azienda Ospedaliera S. Giovanni Battista, Via Genova 3, 10126 Torino, Italy
    Search for more papers by this author
    • Fax: (011) 39 011 696 3737

  • Francesco Zallio M.D.,

    1. Dipartimento di Medicina Oncologia Sperimentale, Divisione Universitaria di Ematologia, Azienda Ospedaliera San Giovanni Battista, Turin, Italy
    Search for more papers by this author
  • Marco Ladetto M.D.,

    1. Dipartimento di Medicina Oncologia Sperimentale, Divisione Universitaria di Ematologia, Azienda Ospedaliera San Giovanni Battista, Turin, Italy
    Search for more papers by this author
  • Massimo Di Nicola M.D.,

    1. Divisione di Oncologia Medica, Istituto Nazionale Tumori, Milan, Italy
    Search for more papers by this author
  • Daniele Caracciolo M.D.,

    1. Dipartimento di Medicina Oncologia Sperimentale, Divisione Universitaria di Ematologia, Azienda Ospedaliera San Giovanni Battista, Turin, Italy
    Search for more papers by this author
  • Michele Magni M.D.,

    1. Divisione di Oncologia Medica, Istituto Nazionale Tumori, Milan, Italy
    Search for more papers by this author
  • Carlo Marinone M.D.,

    1. Divisione Medicina Generale, Azienda Ospedaliera San Giovanni Battista, Turin, Italy
    Search for more papers by this author
  • Maria Dell'Aquila Ph.D.,

    1. Dipartimento di Medicina Oncologia Sperimentale, Divisione Universitaria di Ematologia, Azienda Ospedaliera San Giovanni Battista, Turin, Italy
    Search for more papers by this author
  • Michela Rosace Ph.D.,

    1. Dipartimento di Medicina Oncologia Sperimentale, Divisione Universitaria di Ematologia, Azienda Ospedaliera San Giovanni Battista, Turin, Italy
    Search for more papers by this author
  • Alessandro Pileri M.D.,

    1. Dipartimento di Medicina Oncologia Sperimentale, Divisione Universitaria di Ematologia, Azienda Ospedaliera San Giovanni Battista, Turin, Italy
    Search for more papers by this author
  • Mario Boccadoro M.D.,

    1. Dipartimento di Medicina Oncologia Sperimentale, Divisione Universitaria di Ematologia, Azienda Ospedaliera San Giovanni Battista, Turin, Italy
    Search for more papers by this author
  • Alessandro M. Gianni M.D.,

    1. Divisione di Oncologia Medica, Istituto Nazionale Tumori, Milan, Italy
    Search for more papers by this author
  • Corrado Tarella M.D.

    1. Dipartimento di Medicina Oncologia Sperimentale, Divisione Universitaria di Ematologia, Azienda Ospedaliera San Giovanni Battista, Turin, Italy
    Search for more papers by this author

Abstract

BACKGROUND

The goal of the current study was to evaluate the impact of presentation with an age-adjusted International Prognostic Index (aaIPI) score of 2 or 3 on patients with high-risk aggressive lymphoma who are treated with frontline intensive chemotherapy and autografting.

METHODS

Sixty-nine consecutive patients (median age, 40 years) with either B-cell (n = 60) or non–B-cell (n = 9) aggressive lymphoma were treated with high-dose sequential (HDS) chemotherapy and peripheral blood progenitor cell (PBPC) autografting. The patients who were examined had poor prognoses, with aaIPI scores of 2 (n = 37) or 3 (n = 32). The original treatment regimen, sequential delivery of cyclophosphamide, methotrexate, and etoposide, followed by PBPC autografting (o-HDS), was used in the first 32 patients; the program was intensified by the addition of a course of high-dose cytosine arabinoside (C-HDS) in the next 37 patients.

RESULTS

There were 4 toxicity-related deaths—2 in each aaIPI subgroup (treatment-related mortality, 5.8%). The complete remission rate was significantly higher among patients with an aaIPI score of 2 (n = 32 [86%]) compared with those with an aaIPI score of 3 (n = 13 [41%]; P < 0.001). Patients with an aaIPI score of 2 had significantly better outcomes than did patients with an aaIPI score of 3 in terms of both overall survival (78% vs. 34% at 8 years; P < 0.001) and event-free survival (72% vs. 28% at 8 years; P < 0.001). Similar results were observed when the analysis was limited to the 60 patients with B-cell-derived lymphoma. No significant differences in outcome between patients receiving o-HDS and patients receiving C-HDS were observed. Multivariate analysis demonstrated that an aaIPI score of 3 was the only parameter that was significantly associated with poor overall and event-free survival.

CONCLUSIONS

Age-adjusted International Prognostic Index score is applicable to patients with aggressive lymphoma who are treated with frontline intensive chemotherapy and autografting. In addition, upfront use of HDS chemotherapy appears to be beneficial to patients with an aaIPI score of 2 but not to those with an aaIPI score of 3. Cancer 2003;98:983–92. © 2003 American Cancer Society.

DOI 10.1002/cncr.11610

Patients with aggressive non-Hodgkin lymphoma (NHL) who receive conventional chemotherapy have an overall cure rate of approximately 40%.1 The outcome of these patients is influenced by several clinical features at presentation. The International Prognostic Index (IPI) is the most widely used prognostic factor model for identification of patients who are unlikely to be cured with standard therapy.2, 3 For patients age less than 60 years, the simplified age-adjusted IPI (aaIPI) scoring system is used. Patients with an aaIPI score of 0 or 1 have a favorable prognosis, with an approximately 70–80% probability of long-term survival, whereas patients with an aaIPI score of 2 or 3 have a poorer prognosis, with less than a 50% chance of long-term survival.

In an effort to improve cure rates for patients with aggressive NHL, intensive treatments involving high-dose chemotherapy and autologous stem cell transplantation (ASCT) have been investigated extensively over the last decade, mostly in patients with poor prognoses who were younger than age 60 years.4, 5 Several Phase II studies and a small number of randomized Phase III studies, in fact, suggest that ASCT may be beneficial to this subset of patients, improving response rate and long-term survival relative to conventional chemotherapy.6–12 Significant differences between conventional therapy and ASCT, however, were not evident in other prospective studies.13, 14

The heterogeneity of patient populations probably is the main reason for the discrepant results observed in ASCT study programs to date.15 A wide variety of prognostic parameters have been used to select patients for ASCT programs. In addition, although aaIPI score has been used relatively frequently as a prognostic variable, intermediate-high-risk patients generally have been considered as a single group, without separate evaluation of patients with a score of 2 and patients with a score of 3.

Several reports suggest that ASCT is most effective when performed after maximal cytoreduction.15, 16 All randomized studies that favor the upfront use of ASCT to treat poor-prognosis aggressive NHL used autografting as consolidation therapy after prolonged or intensified induction chemotherapy.7, 8, 11, 12 In contrast, abbreviated standard induction management followed by early ASCT was shown to be ineffective.13, 14 Therefore, adequate chemotherapy-based debulking probably is the best prerequisite for beneficial use of high-dose therapy plus autografting. The high-dose sequential (HDS) chemotherapy regimen is based on this concept of early high-dose chemotherapy followed by autografting.17, 18 We initially reported in a randomized trial that HDS was superior to MACOP-B for use as upfront therapy in a subset of patients with diffuse large cell lymphoma (DLCL).8 However, the patients in that trial were selected using narrow criteria: those with bone marrow (BM) involvement, transformed histology, or non–B-cell subtype were excluded from the trial. Furthermore, aaIPI scores were not available at the time of that study.

We have since used the original HDS regimen (o-HDS) and its modified version (C-HDS), which included a 6-day course of high-dose cytosine arabinoside (ara-C), to treat a series of 69 consecutive patients with an aaIPI score of 2 or 3. The long-term outcome observed in these 69 patients is superior to the outcome observed in conventionally treated historical control patients; however, a comparison of patients with an aaIPI score of 2 and patients with an aaIPI score of 3 revealed a significantly better outcome for the former group in terms of both overall and event-free survival. These results suggest that patients with an aaIPI score of 2 and those with an aaIPI score of 3 should be evaluated separately when analyzing the results of ASCT programs for treating aggressive NHL. Upfront treatment using a highly intensive program such as HDS may be beneficial to patients with an aaIPI score of 2, but it probably is not beneficial to those with an aaIPI score of 3.

MATERIALS AND METHODS

Patients

Between January 1992 and August 2000, two consecutive studies were performed at the University of Turin (Turin, Italy) and the Istituto Nazionale Tumori (Milan, Italy). The goal of both studies was to evaluate the efficacy of HDS chemotherapy in conjunction with peripheral blood progenitor cell (PBPC) autografting as upfront therapy for treating all subtypes of aggressive NHL in intermediate-risk and high-risk patients age < 60 years. The two studies differed only in terms of the treatment schedule followed: the o-HDS regimen was used in the first study (1992–1996), whereas the modified C-HDS regimen was used in the second one (1997–2000).19

The eligibility criteria for both studies were as follows: 1) age 15–60 years; 2) ability to provide informed written consent; 3) negative serologic markers for human immunodeficiency virus, hepatitis B virus (HBV), and hepatitis C virus (HCV) infection, except for patients with documented negativity of HBV-DNA and HCV-RNA; 4) no severe heart, lung, kidney, or liver failures, except for those that were lymphoma-related; 5) no other concomitant or previous neoplasia; 6) biopsy-proven diagnosis of aggressive NHL (including B-cell DLCL [B-DLCL]; peripheral T-cell lymphoma, unspecified type [PTCL-U]; nucleophosmin–anaplastic lymphoma kinase–negative, CD30-positive anaplastic large cell lymphoma [ALCL]; B-DLCL with histologic transformation [transformed lymphoma]; and CD56-positive large cell lymphoma [natural killer cell lymphoma])20, 21; and 7) advanced-stage disease and aaIPI score ≥ 2. Thirty-two patients were enrolled in the o-HDS study, and 37 were enrolled in the C-HDS study. Patient characteristics are summarized in Table 1. The histologic subtypes found in each patient group are listed in Table 2; B-DLCL was the most common subtype by far.

Table 1. Patient Characteristics
CharacteristicAll patients (n = 69)aaIPI 2 (n = 37)aaIPI 3 (n = 32)
  • aaIPI: age-adjusted International Prognostic Index Score; M: male; F: female; o-HDS: original high-dose sequential chemotherapy regimen; C-HDS: cytosine arabinoside–supplemented high-dose sequential chemotherapy regimen; LDH: lactate dehydrogenase.

  • a

    Involvement of one or more extranodal sites.

  • b

    Mediastinal mass with a width greater than one-third the transthoracic diameter or a nonmediastinal lesion greater than 10 cm in cross section.

Age (yrs)   
 Median403941
 Range18–6318–5827–63
Gender (M/F)41/2822/1519/13
Treatment schedule (%)   
 o-HDS32 (46)13 (36)19 (58)
 C-HDS37 (54)24 (64)13 (42)
Extranodal sitesa (%)39 (56)12 (30)29 (85)
 Bone marrow involvement14 (20) 3 (6)11 (36)
Bulky lesionb (%)40 (58)22 (61)18 (54)
High LDH level (%)55 (79)24 (67)32 (100)
Table 2. Main Histologic Subtypes of Lymphoma
HistotypeaNo. of patients (%)
All patients (n = 69)aaIPI 2 (n = 37)aaIPI 3 (n = 32)
  • aaIPI: age-adjusted International Prognostic Index score.

  • a

    According to the Revised European–American Lymphoma Classification criteria.

Diffuse large B-cell57 (83)31 (83.5)26 (82)
Transformed histology 3 (4) 2 (5.5) 1 (3)
Anaplastic CD30-positive large cell 6 (9) 2 (5.5) 4 (15)
Natural killer cell 1 (1)— — 1 (3)
Peripheral T-cell, unspecified 2 (3) 2 (5.5)

Treatment Plans

o-HDS regimen

The o-HDS regimen consisted of debulking with 1 reduced-dose course of APO (doxorubicin: 50 mg/m2, 1 or 2 administrations with 21 days between administrations; vincristine: 2 mg on Days 1 and 21; and prednisone: 50 mg/m2 daily on Days 1–21),22 followed by the high-dose phase, which involved sequential administration at 10–15-day intervals of 1) cyclophosphamide (CY; 7 g/m2 administered intravenously [i.v.]); 2) methotrexate (8 g/m2) plus vincristine (2 mg i.v.); and 3) etoposide (2 g/m2). PBPC harvesting was scheduled to occur at the time of hemopoietic recovery after CY administration; a second harvest was performed after etoposide administration if post-CY collections were poor. The regimen concluded with myeloablative treatment and PBPC autografting. The conditioning regimen included a combination of high-dose mitoxantrone (MITO; 60 mg/m2 i.v. on Day–5) and melphalan (L-PAM; 180 mg/m2 on Day–2).23

C-HDS

The C-HDS regimen began with a debulking phase involving 3 courses of APO on Days 1, 15, and 21 (doxorubicin dose 50 mg/m2 [Day 1] or 75 mg/m2 [Days 15 and 21]); vincristine (2 mg on Days 1, 15, and 21) and prednisone (50 mg/m2 daily on Days 1–21) also were administered. The high-dose phase consisted of sequential administration at 10–15-day intervals of 1) CY (7 g/m2 i.v.); 2) ara-C (2 g/m2 twice daily for 6 consecutive days); and 3) cisplatin (100 mg/m2) plus etoposide (2.4 g/m2). To reduce hematologic toxicity, 1–3 × 106 CD34-positive cells (collected after CY administration) per kilogram were infused following ara-C administration, and a second progenitor cell harvest was performed after ara-C administration to collect as many progenitors as possible.19 The protocol concluded with PBPC autografting after administration of MITO and L-PAM, as in the o-HDS regimen.

In both regimens, consolidation radiotherapy (RT) was administered to bulky sites or sites of disease persistence approximately 2 months after autografting.

Collection and Evaluation of Hemopoietic Progenitors

PBPC were mobilized and collected at hematologic recovery after high-dose CY, etoposide, or ara-C administration. To predict the number and timing of leukaphereses, circulating CD34-positive cell counts and blood cell counts were measured daily starting on Day 9 after chemotherapy and ending at the completion of harvesting. CD34-positive cell counts were evaluated with flow cytometry–direct immunofluorescence analysis of whole blood samples performed according to published procedures.18, 24 Peripheral blood buffy coat cells were collected when the white blood cell count reached 1000/μL and the peripheral blood CD34-positive cell count was > 10/μL. Before cryopreservation, PBPC counts in the leukapheresis product were estimated by evaluating both CD34-positive cell and granulocyte-macrophage colony-forming unit (CFU-GM) counts.19 For a given patient, counts of CD34-positive cells (× 106/kg) and CFU-GM (× 104/kg) were determined by multiplying the concentrations (per mL) of these cells by the total volume of cryopreserved cell suspension and dividing by body weight.

Supportive Care

All patients had indwelling central venous catheters. The entire pretransplant program was conducted in ordinary, unprotected rooms. Autografting was performed in a dedicated inpatient BM transplant unit. Sulfamethoxazole was given twice a week as Pneumocystis carinii prophylaxis during debulking with APO. During pancytopenia and after autografting, patients were managed with a common prophylaxis protocol consisting of oral ciprofloxacin (500 mg twice daily) plus fluconazole (150 mg) plus acyclovir (250 mg i.v. 3 times daily). In the event of fever > 38 °C, blood culture and chest X-ray were performed, and patients were treated empirically with i.v. ceftriaxone plus amikacin; vancomycin was added only if surveillance culture indicated the presence of methicillin-resistant gram-positive cocci. If fever was of undetermined origin and persisted for 36–48 hours, ceftriaxone was replaced by imipenem; i.v. amphotericin was added if fever persisted for longer than 36–48 hours. Antibiotic administration was continued until body temperature reverted to normal for at least 48 hours and absolute neutrophil count was > 500/μL. Irradiated, leukocyte-filtered, single-donor platelet concentrates or, less frequently, irradiated, multiple-donor platelet concentrates were given if platelet count was < 20,000/μL; irradiated, leukocyte-filtered packed red blood cells were given if hemoglobin concentration was less than 8 g/dL.

Response to Treatment and Assessment of Long-Term Outcome

Clinical response was assessed by full restaging 2 months after autografting and then once every 3 months for the first year and once every 6 months thereafter. Complete remission (CR) was defined by Cheson et al.25 as the absence of any clinical sign of disease, and partial remission (PR) was defined as a reduction of ≥ 50% in tumor size. Progressive disease (PD) was defined as the appearance of any new abnormal nodes or an increase of ≥ 50% in the size of a previously identified abnormal node relative to the minimum size of that node. Responses that fell between PR and PD were considered to be stable disease (SD). All patients who started the treatment were considered evaluable for response and outcome on an intention-to-treat basis.

Statistical Analysis

The proportion of patients with a given characteristic was compared across subgroups; the statistical significance of the difference was evaluated with the Fisher exact test. Overall survival (OS) was measured from the start of treatment to the date of death or the last follow-up at which the patient was known to be alive. Event-free survival (EFS) was calculated from the start of treatment to the first adverse event (i.e., recurrence or progression, diagnosis of a secondary malignancy, or treatment-related death) or to the last follow-up at which the patient was known to be alive. The closing date for the analysis was October 31, 2002. OS and EFS were calculated according to the method of Kaplan and Meier.26 The log-rank test was used to compare survival curves.27 All P values were two-sided, and the cutoff level for significance was P = 0.05. Several clinical variables were evaluated for possible prognostic significance in the univariate analysis; these variables included age (≤ 45 years vs. > 45 years), gender, B symptoms, BM involvement, extranodal involvement, phenotype (B-cell vs. non–B-cell), treatment schedule (o-HDS vs. C-HDS), and aaIPI score (2 vs. 3). The significance of these variables also was evaluated in the multivariate analysis, using the stepwise Cox regression model.28

RESULTS

Treatment Feasibility, Toxicity, and Clinical Response

Data regarding treatment feasibility, early toxic death, and clinical response are reported in Table 3. Sixty patients (87%) completed the entire program, including RT after autografting. Interruptions resulted from toxic death during PR (n = 4 [5.7%]), progression of disease to the central nervous system (n = 1; patient had PTCL-U), and mycotic lung infiltrates (during CR) requiring prolonged antifungal therapy (n = 1). There were no significant differences in feasibility between the two aaIPI score groups.

Table 3. Feasibility of and Response to Treatment
ParameterNo. of patients (%)
All patients (n = 69)aaIPI 2 (n = 37)aaIPI 3 (n = 32)
  • aaIPI: age-adjusted International Prognostic Index score; CR: complete remission; PR: partial remission.

  • a

    Causes of death were systemic aspergillosis, septic shock, small bowel strangulating obstruction, and persistent pancytopenia following postgraft abdominal radiotherapy. (see text for details).

Treatment completion60 (87)34 (93)26 (81)
CR45 (65)32 (86)13 (41)
PR disease progression20 (29) 3 (8)17 (53)
Toxicity-related deatha 4 (5.8) 2 (5.4) 2 (6.2)

Of the patients with an aaIPI score of 2, two patients receiving o-HDS had fatal toxicities: one patient died of small bowel strangulating obstruction after high-dose etoposide administration, and one patient died of prolonged pancytopenia during consolidation abdominal RT. Among patients with an aaIPI score of 3, there were two fatal toxicities, both of which occurred in patients receiving C-HDS: one patient died of systemic aspergillosis with central nervous system involvement before autografting, and one patient died of septic shock during postautograft pancytopenia.

At 2 months after autografting, 45 cases of lymphoma (65%) were in CR. There was a significantly higher rate of CR among patients with an aaIPI score of 2 (32 of 37 [86%]) compared with patients with an aaIPI score of 3 (13 of 32 [41%]; P < 0.001).

Hematologic and extrahematologic toxicities following high-dose CY and ara-C administration and autografting are described in Table 4. Duration of neutropenia and thrombocytopenia and incidence of infectious complications were greater after high-dose ara-C administration than after high-dose CY administration. All evaluable patients achieved complete hematologic recovery after PBPC autografting. The median time needed to reach an absolute granulocyte count > 0.5 × 103/L was 10 days (range, 8–21 days), and the median time required to attain a self-sustaining platelet count of 50 × 103/L was 15 days (range, 10–150 days). Consolidation RT was feasible for all 33 candidate patients. Four patients had prolonged pancytopenia after RT and required additional infusion of backup PBPC; 1 of these patients died with persistent pancytopenia, whereas the other 3 achieved full hematologic recovery within 1 month. There were no significant differences in the incidence of hematologic toxicities between patients with an aaIPI score of 2 and those with a score of 3. Overall, hematologic and extrahematologic toxicities were similar to those recorded in previous HDS trials.8, 19, 23

Table 4. Hematologic and Extrahematologic Toxicities
ParameterMedian (range)
Post-CY (n = 69)Post-ara-C (n = 37)Postautograft (n = 66)
  • CY: high-dose cyclophosphamide; ara-C: high-dose cytosine arabinoside; WBC: white blood cell count; plt: platelet; RBC: red blood cell.

  • a

    WHO grading system

Days to reach WBC > 1000/μL11 (8–16)12 (9–15)10 (8–21)
Days to reach plt count > 50,000/μL11 (0–17)14 (11–20)15 (10–150)
No. of RBC transfusions (units) 1 (0–5) 3 (1–7) 2 (0–6)
No. of plt transfusions (units) 0 (0–3) 2 (0–8) 1 (1–4)
% of patients with fever406369
Severe infections (n)242
% of patients with Grade III–IV mucositisa0652

Two cases of secondary acute myeloid leukemia (AML) occurred in the C-HDS group. One patient, who had an aaIPI score of 2 and lymphoma that was in CR, developed AML after 1 year; this patient received an allogeneic BM transplant from a human leukocyte antigen–identical sibling but died of early transplant-related toxicity. The other patient, who had an aaIPI score of 3 at presentation, developed acute promyelocytic leukemia (APL) approximately 2 years after autografting; this patient was treated with induction chemotherapy supplemented with all-trans retinoic acid and currently is experiencing clinical and molecular remission of APL, 22 months after its onset.

Long-Term Outcome

After a median follow-up period of 5 years, the projected 8-year OS and EFS rates were 58% and 51%, respectively (Fig. 1A, B). Long-term outcome by aaIPI score also was evaluated. Figure 2 shows projected 8-year OS rates of 78% and 34% for patients with aaIPI scores of 2 and 3, respectively (P < 0.001), and Figure 3 shows projected 8-year EFS rates of 72% and 28%, respectively (P < 0.001). Long-term outcome data are summarized in Table 5; regardless of treatment regimen, better outcomes consistently were observed in patients with an aaIPI score of 2 compared with those with an aaIPI score of 3.

Figure 1.

(A) Overall and (B) event-free survival curves following high-dose sequential chemotherapy for 69 patients with high-risk aggressive lymphoma. The median follow-up period for censored patients was 5 years.

Figure 2.

Overall survival following high-dose sequential chemotherapy, according to age-adjusted International Prognostic Index (aaIPI) score, for 69 patients with high-risk aggressive lymphoma. P < 0.001 for the difference between score groups.

Figure 3.

Event-free survival following high-dose sequential chemotherapy, according to age-adjusted International Prognostic Index (aaIPI) score, for 69 patients with high-risk aggressive lymphoma. P < 0.001 for the difference between score groups.

Table 5. Long-Term Follow-Up
StatusNo. of patients (%)
All patients (n = 69)aaIPI 2 (n = 37)aaIPI 3 (n = 32)
  1. aaIPI: age-adjusted International Prognostic Index score; o-HDS: original high-dose sequential chemotherapy regimen; C-HDS: cytosine arabinoside–supplemented high-dose sequential chemotherapy regimen.

All patients   
 Alive40/69 (58)28/36 (78)12/33 (36)
 Alive without disease36/69 (52)26/36 (72)10/33 (30)
o-HDS   
 Alive16/32 (50) 9/13 (69) 7/19 (37)
 Alive without disease14/32 (44) 8/13 (62) 6/19 (32)
C-HDS   
 Alive24/37 (65)20/24 (83) 4/13 (36)
 Alive without disease22/37 (59)19/24 (78) 3/13 (28)

Long-term outcome subsequently was evaluated according to major histologic subgroup (i.e., B-cell-derived lymphoma [n = 60] vs. non–B-cell-derived lymphoma [n = 9]). Patients with B-cell lymphoma had a better outcome in terms of both OS and EFS, but these results were not statistically significant (data not shown). The outcome of patients with B-cell-derived lymphoma then was evaluated for correlations with aaIPI score. Patients with an aaIPI score of 2 had a significantly better outcome in terms of both OS (76% vs. 41% at 8 years; P < 0.01) and EFS (69% vs. 35% at 8 years; P < 0.005) compared with patients with an aaIPI score of 3 (Figs. 4, 5).

Figure 4.

Overall survival following high-dose sequential chemotherapy, according to age-adjusted International Prognostic Index (aaIPI) score, for 60 patients with high-risk, B-cell-derived aggressive lymphoma. P < 0.01 for the difference between score groups.

Figure 5.

Event-free survival following high-dose sequential chemotherapy, according to age-adjusted International Prognostic Index (aaIPI) score, for 60 patients with high-risk, B-cell-derived aggressive lymphoma. P < 0.005 for the difference between score groups.

Prognostic Parameters

Several clinical characteristics were evaluated with univariate analysis for their ability to predict OS and EFS. As detailed in Table 6, two variables, bone marrow involvement and aaIPI score = 3, demonstrated significant prognostic value; however, only aaIPI score = 3 maintained its independent adverse prognostic value (for both OS and EFS) after multivariate analysis (Table 7).

Table 6. Univariate Analysis
Clinical variableNo. of patientsProjected 5 yr OS (%)PProjected 5 yr EFS (%)P
  1. F: female; M: male; BM: bone marrow; o-HDS: original high-dose sequential chemotherapy regimen; C-HDS: cytosine arabinoside–supplemented high-dose sequential chemotherapy regimen; aaIPI: age-adjusted International Prognostic Index.

Age (yrs)     
 < 454663 56 
 > 4522450.08400.14
Gender     
 F2857 50 
 M41580.68530.31
B symptoms     
 No1540 33 
 Yes54620.15560.08
BM involvement     
 No4866 57 
 Yes21340.03380.13
Extranodal involvement     
 No2369 59 
 Yes46520.09470.19
Phenotype     
 B-cell6060 53 
 non–B-cell9440.15440.29
Treatment regimen     
 o-HDS3250 43 
 C-HDS37640.74580.60
aaIPI score     
 23778 72 
 33234< 0.000128< 0.0001
Table 7. Relative Risk in Multivariate Analysis
CharacteristicOverall survival as endpointEvent-free survival as endpoint
Risk ratio95% CIP valueRisk ratio95% CIP value
  1. CI: confidence interval; aaIPI: age-adjusted International Prognostic Index.

aaIPI score = 33.8761.5–9.40.00034.031.7–9.070.0007

DISCUSSION

For several years, we have been using high-dose sequential chemotherapy with PBPC autografting as an upfront treatment regimen for patients younger than age 60 years with aggressive lymphoma and an aaIPI score of 2 or 3. Results obtained over the last 10 years using this treatment strategy are reported in the current article. The 69 consecutive intermediate-risk and high-risk patients who received either the original or ara-C-supplemented HDS regimen had an outcome that was better than the previously reported outcome resulting from conventional chemotherapy. However, when patients were evaluated separately according to aaIPI score (2 vs. 3), marked differences between the two score groups were observed; only patients with a score of 2 obtained substantial benefit from the use of intensive upfront treatment. This finding may have a critical impact on the interpretation of reports regarding the treatment of aggressive lymphoma using regimens that involve ASCT.

The HDS regimens, both the original schedule and, to an even greater degree, the ara-C-containing version, represent an intensive approach that is particularly conducive to evaluation of the potential advantages of using ASCT in the management of aggressive lymphoma. The two schedules were feasible, and there were no significant differences between them in terms of major toxicity. The 4 toxicity-related deaths (2 in each treatment group) yielded a treatment-related mortality rate of 5.6%, which is slightly higher than the rate expected when intensive chemotherapy with autografting is used5, 29, 30; this difference may be due in part to the strict selection of patients with advanced disease and poor prognosis in the current study. Response rates in the current study were satisfactory, with 65% of patients achieving CR. This percentage is higher than the values reported by Shipp3 (56% and 47% for aaIPI scores 2 and 3, respectively) and the values reported in other series of patients with poor prognosis who received conventional chemotherapy (approximately 50%).31–33

In an earlier randomized trial, the o-HDS regimen was shown to be superior to MACOP-B; in terms of CR rate, progression-free survival, and EFS; for use as an upfront treatment regimen for B-cell DLCL.8 However, in that study, the response and survival rates were higher than those reported in the current study, due to differences in patient selection methods. Data on aaIPI scores had not been published at the time of the earlier randomized trial, and patient selection was made using less reliable parameters. In addition, only patients with de novo B-cell DLCL and without BM involvement were considered eligible for the earlier study. In the current study, aaIPI score was used as a strict eligibility criterion, and patients with transformed or non–B-cell histology, as well as patients with BM involvement, were included; patients in these subgroups have very poor outcomes after conventional treatment.33–37

Although patient outcome was not as good as in the previous randomized study, the findings of the current study indicate that some benefits result from the upfront use of the HDS approach to treat aggressive lymphoma. In the long term, patients in the current study had a projected 8-year OS rate of 59%, which compares favorably with the projections of less than 50% that were reported for patients with poor prognosis after conventional management.2, 3, 31–33 The results of the current study therefore corroborate studies suggesting that in young patients presenting with intermediate-risk or high-risk aggressive lymphoma, intensive chemotherapy followed by autografting offers a greater chance of disease control and long-term survival compared with more conventional approaches.15 In particular, our results are consistent with those from the large, randomized study performed by the Group d'Etude des Lymphomes de l'Adulte.7, 12

The aaIPI score was developed to allow diagnostic characterization of patients with aggressive lymphoma who were receiving conventional chemotherapy. More recently, its value has been recognized specifically in the case of patients with B-cell-derived DLCL.38 In addition, the aaIPI model also may apply to patients with recurrent disease who are undergoing salvage treatment with ASCT.39 In the current study, we had the opportunity to evaluate the use of aaIPI score in patients receiving ASCT as their primary treatment; in this setting, the model proved to be effective. In fact, only patients with an aaIPI score of 2 were found to benefit from intensive chemotherapy and ASCT; the projected 8-year OS and EFS rates for these patients were 78% and 72%, respectively, which represent a definite improvement compared with the 5-year survival rate of 46% that was reported in conventionally treated patients with an aaIPI score of 2.2, 3 In contrast, patients with an aaIPI score of 3 had a poor outcome, with an overall long-term survival rate of 34%, which is consistent with the rate that results from the use of conventional treatment. Overall, the incidence of toxicity was similar in both aaIPI score groups; this finding rules out the possibility that the differences in outcome might simply result from differences in treatment tolerance. Furthermore, patients with an aaIPI score of 2 consistently had better outcomes than did patients with an aaIPI score of 3, regardless of treatment schedule or disease subtype. Indeed, aaIPI score = 3 was identified by multivariate analysis as the only parameter to have independent adverse prognostic value for both OS and EFS. Thus, the aaIPI model allows the identification of subgroups of patients with aggressive lymphoma and poor prognosis who have markedly different expected responses to upfront intensive chemotherapy and ASCT.

The role of ASCT in the treatment of newly diagnosed patients with aggressive lymphoma has long been an issue of debate. Several studies have been performed, most often in patients with poor prognosis, who are less likely to be cured with standard combination chemotherapy. It has been suggested that differences in the prognostic characteristics of patients enrolled in trials evaluating the efficacy of ASCT against aggressive lymphoma may explain the conflicting results that have been reported.3 Our observation of markedly different outcomes in patients with an aaIPI score of 2 and patients with an aaIPI score of 3 confirms the need for clear-cut identification of prognostic subgroups in the evaluation of novel treatment strategies. New methods based on microarray techniques recently have been introduced; these methods can be used to identify prognostic subgroups based on peculiar gene expression patterns.40, 41 This innovative diagnostic approach will allow better identification of patients who are candidates for experimental treatments such as intensive chemotherapy and ASCT.

The introduction of the anti–B cell monoclonal antibody rituximab represents a major advance in the management of B-cell lymphoma. Rituximab can enhance the efficacy of conventional chemotherapy, as has been shown by the randomized study of Coiffier et al.,42 who evaluated the CHOP regimen (cyclophosphamide, doxorubicin, vincristine, and prednisone) with and without rituximab in treating aggressive lymphoma in elderly patients. CHOP plus rituximab was found to be more effective than CHOP alone in all prognostic subgroups, although the difference was more pronounced in patients with low IPI scores. The efficacy of CHOP plus rituximab in both low-risk and high-risk patients with aggressive lymphoma also has been reported in a recent multicenter Phase II study.43 The addition of rituximab also may improve ASCT programs by intensifying the in vivo purging effect before PBPC harvesting, as has been reported by our group and others.44–46 Our preliminary experience with rituximab-supplemented C-HDS has yielded promising results with respect to both salvage and first-line treatment of aggressive lymphoma.47 Future studies will aim to clarify whether rituximab-containing conventional chemotherapy substantially improves patient outcome in high-risk subgroups (such as patients with an aaIPI score of 3) or whether investigational programs, including rituximab-supplemented ASCT, should continue to be considered.

Acknowledgements

The authors thank the laboratory staff and nurses from the Divisione Universitaria di Ematologia and the Centro Dipartimentale Trapianto Midollo, San Giovanni Battista Hospital (Turin, Italy) for their help and for providing patient care.

Ancillary