Long-term follow-up results of hyperfractionated cyclophosphamide, vincristine, doxorubicin, and dexamethasone (Hyper-CVAD), a dose-intensive regimen, in adult acute lymphocytic leukemia

Authors


Abstract

BACKGROUND

Modern intensive chemotherapy regimens have improved the prognosis for patients with adult acute lymphocytic leukemia (ALL). With these regimens, the complete response rates are now reported to be > 80%, and the long-term survival rates range from 30% to 45%. The current analysis updated the long-term results with the original hyperfractionated cyclophosphamide, vincristine, doxorubicin, and dexamethasone (Hyper-CVAD) program, with a median follow-up time of 63 months.

METHODS

Between 1992 and 2000, 288 patients were treated with Hyper-CVAD. The median age of the patients was 40 years, and 59 patients (20%) were ≥ age 60 years. The incidence of Philadelphia chromosome (Ph)-positive ALL was 17%, and the incidence of of T-cell ALL was 13%.

RESULTS

A complete response (CR) was achieved in 92% of patients. The induction mortality rate was 5% (2% if the patient's age was < 60 years, and 15% if the patient's age was ≥ 60 years). With a median follow-up time of 63 months, the 5-year survival rate was 38% and the 5-year CR duration rate was 38%. Multivariate analysis of prognostic factors for CR duration identified the following adverse factors: age ≥ 45 years, leukocytosis ≥ 50 × 109/L, poor performance status (an Eastern Cooperative Oncology Group score of 3–4), Ph-positive disease, French–American–British L2 morphology, > 1 course to achieve CR, and Day 14 bone marrow blasts > 5%. Patients were divided into low-risk (risk score 0–1; 37%), intermediate risk (risk score 2–3; 36%), and poor-risk groups (risk score ≥ 4; 27%) with 5-year CR duration rates of 52%, 37%, and 10%, respectively.

CONCLUSIONS

Compared with the previous VAD regimens, Hyper-CVAD was associated with significantly better CR rates, CR duration, and survival. The long-term follow-up results of Hyper-CVAD were favorable. Comparison of Hyper-CVAD with other established adult ALL regimens is warranted. Cancer 2004. © 2004 American Cancer Society.

Modern intensive chemotherapy regimens have improved the prognosis for patients with adult acute lymphocytic leukemia (ALL). With these regimens, the complete response rates are now > 80%, and the long-term survival rates range from 30% to 45%.1–6 For the most part, the results have been confirmed in selected study groups, and the outcome is better for younger patients treated with these regimens.7 The median age of the patients in older adult ALL series was 25–30 years. In contrast, the median age of patients in more recent studies is 40–45 years,2, 5 reflecting the true adult age incidence in oncology community practice. Areas of significant concern to patient prognosis include older age (e.g., age 50 to ≥ 60 years), Philadelphia chromosome (Ph)-positive ALL, severe leukocytosis in patients with non–T-cell ALL (T-ALL), and slow response to induction therapy. Initial results concerning the Hyper-CVAD dose-intensive regimen, administered to adult patients with ALL were encouraging.8 The current analysis updates our long-term results with the original Hyper-CVAD program, with a median follow-up time of 63 months.

MATERIALS AND METHODS

Study Group

Adults with previously untreated ALL were treated with Hyper-CVAD between February 1992 and March 2000. The protocol was approved by the institutional review board at The University of Texas M. D. Anderson Cancer Center (Houston, TX). Informed consent was obtained from patients according to institutional guidelines.

Entry criteria were age ≥ 15 years, and the absence of other active malignancy and expected consequent death within 12 months, as well as the absence of human immunodeficiency virus-1–positive status. No exclusions were made because of performance status or because of cardiac, hepatic, or renal function or concomitant active infection.

The diagnosis of ALL required the presence of lymphoid-looking blasts that were < 3% positive for myeloperoxidase and strongly positive (≥ 40%) for terminal deoxnucleotidyl transferase or demonstrated block positivity for periodic acid–Schiff (except for mature B-cell ALL [B-ALL]). Immunophenotyping and cytogenetic studies complemented the diagnosis in difficult cases.8

Pretreatment workup included history and physical examination; complete blood counts, differential, and platelet counts; serum chemistries (sequential multiple analysis 12/60), including liver and renal function studies; bone marrow aspiration for morphologic analysis and staining, and biopsy; cytogenetic analysis; and immunophenotyping. Classification of patient groups by karyotype and immunophenotype was reported in earlier studies.8–10 Follow-up bone marrow studies were performed on Days 14 and 21 of disease remission induction, every 1–3 courses during consolidation, and every 3–6 months during maintenance. A positive immunophenotype referred to a cluster designation (CD) of ≥ 20%. Positive T-ALL referred to ≥ 2 T-cell markers (CD1–CD8). Common acute lymphoblastic leukemia antigen (CALLA) referred to expression of CD10 in addition to expression of CD19 or CD20. Precursor B-ALL referred to expression of CD19 or CD20. Mature B-ALL referred to expression of surface immunoglobulin (SIg) or clonal kappa or lambda. Patients were also tested for myeloid markers CD13, CD14, and CD33. A positive myeloid marker required expression of one or more CDs. Five patients known to have Ph-positive ALL also had a ≥ 20% positive SIg. They also were positive for CD10 and CD19 or CD20. We therefore considered them to be separate from the mature B-ALL category and in the CALLA ALL category.

Therapy

The regimen was comprised of two phases.

Dose-intensive phase

The dose-intensive phase included 8 cycles of dose-intensive therapy courses of Hyper-CVAD (Courses 1, 3, 5, and 7) alternating with high-dose methotrexate (MTX) and cytosine arabinoside (HD-MTX–Ara-C; Courses 2, 4, 6, and 8).

Hyper-CVAD.

Hyper-CVAD was comprised of 300 mg/m2 of cyclophosphamide administered intravenously (i.v.) over 2–3 hours every 12 hours for 6 doses on Days 1–3, with sodium mercaptoethanesulfonate given at twice the total dose as cyclophosphamide but given by continuous infusion starting with cyclophosphamide and ending 12 hours after the last dose; 2 mg of vincristine administered i.v. on Days 4 and 11; 50 mg/m2of doxorubicin administered i.v. over 2 hours on Day 4; and 40 mg of dexamethasone daily on Days 1–4 and on Days 11–14.

HD-MTX-Ara-C.

HD-MTX–Ara-C was comprised of 1 g/m2 of MTX administered i.v. over 24 hours on Day 1; 15 mg of citrovorum factor rescue was initiated 12 hours after the completion of MTX infusion every 6 hours × 8, and increased to 50 mg i.v. every 6 hours if MTX levels were > 20 μmol/L at 0 hours, or were > 1.0 μmol/L at 24 hours, or were > 0.1 μmol/L at 48 hours after the end of MTX infusion, until levels were < 0.1 μmol/L; 3 g/m2 of Ara-C was administered i.v. over 2 hours every 12 hours × 4 on Days 2 and 3; and 50 mg of methylprednisolone was administered i.v. twice daily on Days 1–3.

Central Nervous System Prophylaxis

Patients with a high risk for central nervous system (CNS) disease (lactate dehydrogenase level > 600 U/L [normal range, 25–225 U/L], or with a proliferative index [percent S+G2M] ≥ 14% received 16 intrathecal treatments (IT). Patients with low risk (neither elevated) received four IT. Patients with unknown risk (measurements were not available) received eight IT. Patients with mature B-ALL were included in the CNS high-risk category. CNS prophylaxis was given with 12 mg of MTX IT (6 mg if via the ommaya reservoir) on Day 2 and 100 mg of Ara-C IT on Day 8 of each cycle for 16 IT, 4 IT, or 8 IT as described. Patients at low or unknown risk for CNS disease received four or eight IT treatments on Days 2 and 8 of the first two or four cycles of therapy. Patients with CNS disease at the time of diagnosis received IT therapy twice weekly until cerebrospinal fluid study findings were negative and treatment then was received according to the schedule of the protocol. Patients with cranial nerve root involvement received 24–30 gray (Gy) of radiotherapy in 10–12 fractions, directed to the base of the skull or to the whole brain, during the disease remission induction.

Antibiotic prophylaxis was given during the dose-intensive (induction–consolidation) phase as follows: ciprofloxacin given at 500 mg orally twice daily or levofloxacin given at 500 mg daily; fluconazole given at 200 mg orally daily; and acyclovir given at 200 mg orally twice daily or valacyclovir given 500 mg orally daily.

Supportive care with granulocyte—colony-stimulating factor (G-CSF), 10 μg/kg daily, was given in 2 divided doses starting 24 hours after the end of chemotherapy (i.e., on Day 5 of Hyper-CVAD therapy and on Day 4 of HD-MTX–Ara-C therapy). Subsequent courses of chemotherapy were given as soon as the leukocyte count was > 3 × 109/L and the platelet count was > 60 × 109/L.

Dose Modifications

Vincristine was reduced to 1 mg if the bilirubin level was > 2 mg/dL. The doxorubicin dose was reduced by 25% if the bilirubin level was 2–3 mg/dL, by 50% if it was 3–4 mg/dL, and by 75% if it was > 4 mg/dL. The MTX dose was reduced by 25% when creatinine levels were 1.5–2 mg/dL and by 50% when levels were higher. The Ara-C dose was reduced to 1 g/m2 in patients age ≥ 60 years, if the creatinine level was > 2.0 mg/dL or if the MTX level at the end of the MTX infusion (0 hours after the completion of MTX therapy) was ≥ 20 μmol/L after repeat assays were performed.

Hyper-CVAD treatment courses (i.e., Courses 3, 5, and 7) did not usually require dose reductions for serious toxicities. With the HD-MTX–Ara-C treatment courses, serious toxicities (usually Grade 3–4 myelosuppression-associated complications other than neutropenia or thrombocytopenia) required subsequent dose reductions of 25–33% (i.e., the MTX dose was reduced to 750 mg/m2 and then to 500 mg/m2 and 250 mg/m2 and the Ara-C dose was reduced to 2 g/m2 and then to 1.5 g/m2 and 1 g/m2. Toxicities were graded according to the National Cancer Institute Common Toxicity Criteria (version 2.0).

Mediastinal irradiation

Patients with T-ALL and mediastinal disease at the time of diagnosis underwent mediastinal irradiation at the end of intensive courses of chemotherapy and before maintenance. Patients received 30 Gy (range, 26–39.6 Gy; fraction size 1.5– 3 Gy) over a period of 3–4 weeks. Mediastinal irradiation was recommended regardless of the status of mediastinal disease after achievement of bone marrow complete remission (CR).

Maintenance phase

Patients with mature B-ALL received no maintenance therapy. Patients with Ph-positive ALL who were candidates for allogeneic stem cell transplantation (SCT) and had a matched related (or one antigen mismatch) donor, or who had a matched unrelated donor, underwent allogeneic SCT as soon as possible while in CR (without continuing the intensive phase).

All other patients received maintenance therapy with mercaptopurine (6-MP), MTX, vincristine, and prednisone (POMP) for 2 years. Between 1992 and 1995, oral POMP was given: 6-MP at a dose of 50 mg orally 3 times daily (on an empty stomach), MTX at a dose of 20 mg/m2 orally weekly, vincristine at a dose of 2 mg i.v. monthly, and prednisone at a dose of 200 mg orally daily × 5 every month (with vincristine). Since 1995, i.v. POMP was administered: 6-MP 1 g/m2 given i.v. over 1 hour daily × 5 every month, MTX at a dose of 10 mg/m2 i.v. given over 1 hour daily × 5 every month (with 6-MP), and vincristine and prednisone given monthly as just described. The 6-MP and MTX doses were reduced by 25% (to 750 mg/m2and 7.5 mg/m2, respectively) when moderate toxicity developed and by 50% (to 500 mg/m2 and 5 mg/m2, respectively) when severe toxicity occurred. Mucositis and hepatic dysfunctions were found to be related more to MTX therapy, and the MTX dose generally was reduced selectively before 6-MP dose reductions were considered.

Antibiotic prophylaxis given during the maintenance phase was comprised of trimethoprim-sulfamethoxazole given twice daily on weekends, and acyclovir at a dose of 200 mg or valacyclovir at a dose of 500 mg daily or 3 times weekly, for the first 6 months, to reduce the probability of Pneumocystis infection, herpes zoster, or varicella.

Response and Toxicity Criteria and Statistical Methods

Survival was calculated from the initiation of therapy, and the CR duration was calculated from achievement of CR until there was evidence of leukemia recurrence (e.g., ≥ 10% of lymphoblasts in the bone marrow, or CNS or extramedullary disease recurrence). Survival and CR duration distributions were estimated using the Kaplan–Meier method, and were compared using the log-rank test. Differences in response rates were analyzed using a chi-square test. Significant cutoff points for prognostic factors were according to previously accepted categories, as well as visualization of the Martingale residual plots. A proportional hazards model was used to evaluate independent prognostic factors for survival and CR duration. Logistic regression analysis was used to identify independent prognostic factors for CR. In the design of the simplified multivariate-derived risk models, the 5-year survival or CR duration rates for each independent prognostic factor were assigned the following corresponding scores: > 35%, 0; 25–35%, 1; and < 25%, 2. Risk groups were then divided according to the aggregation/closeness of outcome in difference cumulative score categories.

Hyper-CVAD versus VAD

To evaluate the potential benefit of Hyper-CVAD therapy, the results were compared with results achieved with three previous VAD regimens used between 1982 and 1991 in patients with newly diagnosed adult ALL.11 Entry criteria were similar during the two study periods.

RESULTS

Study Group

The characteristics of the 288 patients treated with Hyper-CVAD are summarized in Table 1. The median age of the patients was 40 years (mean, 42 years; range, 15–92 years), and 104 patients (36%) were age ≥ 50 years and 59 patients (20%) were age ≥ 60 years. The leukocyte count was ≥ 30 × 109/L in 73 patients (25%).

Table 1. Characteristics of the Study Group (n = 288)
CharacteristicsNo. of patients (%)
  1. ECOG: Eastern Cooperative Oncology Group; CNS: central nervous system; CR: complete response; FAB: French–American–British; CALLA: common acute lymphoblastic leukemia antigen; MDACC: M. D. Anderson Cancer Center.

Age (yrs) 
 < 40147 (51)
 40–5982 (28)
 ≥ 6059 (20)
Performance score (ECOG scale) 
 3–419 (7)
Splenomegaly 
 Yes76 (26)
Hepatomegaly 
 Yes44 (15)
Lymphadenopathy 
 Yes91 (32)
CNS disease at diagnosis 
 Yes19 (7)
Mediastinal mass 
 Yes20 (7)
Leukocyte count (× 109/L) 
 < 5105 (36)
 5–49135 (47)
 ≥ 5048 (17)
Platelet count (× 109/L) 
 ≤ 80193 (67)
Hemoglobin level (g/dL) 
 < 10206 (72)
Lactic dehydrogenase level (IU/L) 
 > 600147 (51)
Alkaline phosphatase level (U/L) 
 ≥ 80202 (70)
Creatinine level (mg/100 mL) 
 ≥ 1.340 (14)
Bilirubin level (mg/100 mL) 
 ≥ 1.335 (12)
Karyotype 
 Diploid66 (23)
 Ph positive48 (17)
 t(8;14), t(2;8), t(8;22)15 (5)
 6q−; 14q +19 (7)
 Insufficient metaphases59 (21)
 Hyperdiploid14 (5)
 Hypodiploid13 (5)
 Other48 (17)
 Not done6 (2)
FAB classification 
 L193 (32)
 L2156 (54)
 L331 (11)
 Not classified/missing8 (3)
Immunophenotype 
 Mature B27 (9)
 T38 (13)
 Precursor B21 (7)
 T-CALLA—precursor B8 (3)
 CALLA159 (55)
 Null12 (4)
 Not done23 (8)
Myeloid markers 
 Positive128 (44)
 Negative112 (39)
 Not performed/unknown48 (17)
Systemic risk at CR (MDACC) 
 Low73 (25)
 High212 (74)
Hoelzer risk at CR 
 Low67 (23)
 High221 (77)
Risk for CNS disease 
 High158 (55)
 Low62 (22)
 Unknown68 (24)

Therapy Results

Two hundred sixty-four patients (92%) achieved a CR, 14 (5%) died during induction, and 10 (3%) had resistant disease. Two hundred thirteen patients (81%) achieved a CR after 1 course of therapy. Induction death was caused by infections in 14 patients (i.e., fungal infection in 2 patients, bacterial infection in 3 patients, and multiple infections in 9 patients). The median time to CR was 22 days.

The median follow-up time was 63 months (range, 5–137 months). The median survival period was 32 months, the 5-year survival rate was 38%, and the 5-year CR rate was 38% (Figs. 1 and 2). At the time of last follow-up, 93 patients were alive without disease and 178 patients had died (i.e., induction deaths [n = 14], deaths due to disease recurrence [n = 145], and deaths while patients were in CR [n = 19]). The 19 deaths among patients in CR were caused by postallogeneic SCT complications in 3 patients with Ph-positive ALL, by infection-associated problems in 11 patients (sepsis in 9 patients and other causes in 2 patients), by organ failures in 3 patients (hepatic failure in 1 patient and neurotoxicity in 2 patients), and by old age in 2 patients. At the time of last follow-up, 11 patients (all with Ph-positive ALL) have undergone allogeneic SCT in first CR from related (n = 7) or unrelated donors (n = 4). Two (18%) of these patients were alive without disease for ≥ 64 and ≥ 74 months from SCT, respectively.

Figure 1.

Survival with hyperfractionated cyclophosphamide, vincristine, doxorubicin, and dexamethasone (Hyper-CVAD) versus vincristine, doxorubicin, and dexamethasone (VAD).

Figure 2.

Disease remission duration with hyperfractionated cyclophosphamide, vincristine, doxorubicin, and dexamethasone (Hyper-CVAD) versus vincristine, doxorubicin, and dexamethasone (VAD).

Response by pretreatment characteristics is shown in Table 2. The CR rate for patients age < 30 years was 99%, which is similar to the CR rate for childhood ALL. Patients age ≥ 60 years were found to have a lower CR rate (80%), mostly because of a higher induction mortality. Leukocytosis and karyotypic abnormalities were not associated with differences in CR rates. Patients with Ph-positive disease had a CR rate of 92%. Twenty patients had leukocytosis of ≥ 100 × 109/L, including 15 patients with non–T-ALL—19 (95%) of these patients achieved CR. Twelve patients had null-cell immunophenotypes. Of these, only 8 patients (67%) achieved a CR. Eight patients had a T-CALLA—precursor B phenotype (i.e., two or more positive T-cell markers [CD1–8] in addition to a positive CD10 and CD19 or CD20). Of these 8 patients, 6 (75%) achieved a CR.

Table 2. Response and Survival by Pretreatment Characteristics
CharacteristicsNo. of CRs / TotalCR (%)P value5-Yr survival rate (%)P value
  1. CR: complete response; ECOG: Eastern Cooperative Oncology Group; FAB: French–American–British; CALLA: common acute lymphoblastic leukemia antigen.

Age (yrs)     
 < 40140/14795 51 
 40–5977/8294< 0.00130< 0.001
 ≥ 6047/5980 17 
Performance score (ECOG)     
 0–1191/20195 40 
 260/6888< 0.00137< 0.001
 3–413/1968 12 
Hepatomegaly     
 No228/244930.01400.014
 Yes36/4482 26 
Leukocyte count (× 109/L)     
 ≤ 5100/10893 40 
 5.1–49.9121/132920.82144< 0.001
 ≥ 5043/4890 16 
Platelet count (× 109/L)     
 < 2024/2983 24 
 20–80149/164910.07233< 0.001
 > 8091/9596 51 
Albumin level (mg/100 mL)     
 < 342/52810.002400.39
 ≥ 3222/23694 38 
Bilirubin level (mg/100 mL)     
 < 1.3235/252930.008390.21
 ≥ 1.328/3580 29 
FAB classification     
 L187/9394 48 
 L2143/156920.525290.038
 L327/3187 53 
Immunophenotype     
 Mature B-cell ALL24/2789 36 
 T-cell ALL36/3895 48 
 T-CALLA—precursor B-cell ALL6/8750.005160.18
 Precursor B-cell ALL21/21100 45 
 CALLA149/15994 38 
 Null8/1267 25 
Karyotype     
 Ph positive44/4892 12 
 t(8;14), t(2;8), t(8;22)14/15930.9667< 0.001
 Other200/21991 41 
Myeloid markers     
 Positive115/12890 34 
 Negative104/112930.595410.66
 Unknown45/4894 42 

There were no differences noted in CR rates by gender; presence of splenomegaly, adenopathy, mediastinal disease, or CNS disease; degree of anemia; presence of peripheral blasts; or presence of elevated levels of creatinine, lactate dehydrogenase, alkaline phosphatase, or β-2-microglobulin (data not shown). Patients with a poor performance status (Eastern Cooperative Oncology Group [ECOG] scale), hypoalbuminemia, hepatomegaly, or hyperbilirubinemia were found to have lower CR rates (68–82%), mostly because of high induction mortality rates: poor performance status, 40%; hypoalbuminemia, 24%; hepatomegaly, 16%; and hyperbilirubinemia, 19%. Among the 14 patients who experienced induction death, 9 (64%) were age ≥ 60 years, 6 (43%) had a poor performance status (ECOG score of 3–4), 9 (64%) had albumin levels < 3 mg/dL, 6 (43%) had elevated bilirubin levels, and 6 (43%) had hepatomegaly.

Multivariate analysis identified older age (age ≥ 60 years), thrombocytopenia (< 100 × 109/L), and T-CALLA—precursor B phenotype to be independently associated with worse CR rates (P < 0.05).

Survival and Disease Remission Duration

The overall survival of patients who received Hyper-CVAD is shown in Figure 1. Pretreatment characteristics associated with significant differences in survival rates are shown in Table 2. A multivariate analysis identified the following factors to be independent poor prognostic factors for survival (P < 0.01): older age, Ph-positive disease, leukocytosis, thrombocytopenia, poor performance (ECOG score of 3–4), and hepatomegaly. Patients were divided into good-risk (risk score 0–1; 37%), intermediate risk (risk score 2–3; 38%), or poor-risk groups (risk score ≥ 4; 26%) with estimated 5-year survival rates of 62%, 34%, and 5%, respectively (Tables 3 and 4; Fig. 3).

Table 3. Multivariate Analysis of Prognostic Factors for Survival rate (%)
Prognostic factorNo. of patientsScore assignment5-Yr survival rate (%)Multivariate analysis P value
  1. ECOG: Eastern Cooperative Oncology Group; Ph: Philadelphia chromosome.

Age (yrs)    
 < 40147051 
 40–5982130< 0.0001
 ≥ 6059217 
Performance score (ECOG)    
 0–2269039< 0.0001
 3–419212 
Hepatomegaly    
 No2440400.006
 Yes44126 
Leukocyte count (× 109/L)    
 < 502400420.0006
 ≥ 5048216 
Platelet count (× 109/L)    
 < 2029224 
 20–801641330.0004
 > 8095051 
Karyotype    
 Other2340420.0004
 Ph48212 
Table 4. Survival According to Risk Model
Prognostic model riskScoreNo. of patients (%)Survival
Median (mos)5-Yr survival rate (%)
Low0–1104 (37)10962
Intermediate2–3106 (38)3234
High≥ 472 (26)115
Figure 3.

Survival with hyperfractionated cyclophosphamide, vincristine, doxorubicin, and dexamethasone with the presence of none or one, two to three, or four or more 4 adverse factors.

Disease remission duration with Hyper-CVAD is shown in Figure 2. Among patients achieving a CR, pretreatment factors associated with a worse disease remission duration are shown in Table 5. The need for > 1 course to achieve a CR, and the presence of bone marrow blasts > 5% on Day 14 also were found to be adverse factors. Multivariate analysis identified the following factors to be adverse independent factors for CR duration (P < 0.05): age ≥ 45 years; leukocytosis ≥ 50 × 109/L; poor performance (ECOG score of 3–4); Ph-positive disease; FAB L2 morphology; > 1 course to CR; and bone marrow blasts > 5% on Day 14. Patients were divided into good-risk (risk score 0–1; 37%), intermediate risk (risk score 2–3; 36%), or poor-risk groups (risk score ≥ 4; 27%) with estimated 5-year CR duration rates of 52%, 37%, and 10% respectively (Tables 6 and 7; Fig. 4).

Table 5. Complete Disease Remission Duration by Patient Characteristics
CharacteristicsNo. of patients5-Yr CR rate (%)P value
  1. CR: complete responses; ECOG: Eastern Cooperative Oncology Group; CALLA: common acute lymphoblastic leukemia antigen; Ph: Philadelphia chromosome; FAB: French–American–British; MDACC: M. D. Anderson Cancer Center.

All patients26438
Age (yrs)   
 < 3091400.008
 30–446548 
 45–596131 
 ≥ 604727 
Performance score (ECOG)   
 0–2251390.005
 3–4139 
Leukocyte count (× 109/L)   
 ≥ 510035< 0.001
 5.1–309747 
 > 30–492438 
 ≥ 504319 
Platelet count (× 109/L)   
 < 2024320.223
 20–8014934 
 81–992150 
 ≥ 1007044 
Immunophenotype   
 Mature B24430.221
 T-CALLA Pre-B625 
 T3655 
 Precursor-B2158 
 CALLA14932 
 Null833 
Karyotype   
 Ph positive4414< 0.001
 t (8;14), t (2;8), t (8;22)1476 
 Other20039 
FAB   
 L187440.002
 L214328 
 L32767 
No. of courses to CR   
 121342< 0.001
 > 15118 
Systemic risk group (MDACC)   
 Low69440.016
 High19336 
Hoelzer risk group   
 Low67450.020
 High19735 
Day 14 bone marrow blasts > 5%   
 Yes81250.003
 No15542 
Table 6. Multivariate Analysis of Prognostic Factors Associated with Disease Remission Duration
Prognostic factorNo. of patientsScore assignment5-Yr CR durationMultivariate analysis P value
  1. CR: complete response; ECOG: Eastern Cooperative Oncology Group; FAB: French–American–British.

Age (yrs)    
 < 451560430.0006
 ≥ 45108129 
Performance score (ECOG)    
 0–22510390.002
 3–41329 
Leukocyte count (× 109/L)    
 < 50221041< 0.0001
 ≥ 5043219 
FAB morphology    
 Other1140490.02
 L2143128 
Karyotype    
 Other2140410.002
 Ph44214 
No. of courses to CR    
 12130420.04
 ≥ 251218 
Day 14 bone marrow blasts (%)    
 ≤ 51550420.01
 > 581125 
Table 7. Duration of Remission According to Risk Model
Prognostic model riskScoreNo. of patients (%)CR duration
Median (mos)5-yr survival rate (%)
  1. CR: complete response.

Low0–183 (37)12052
Intermediate2–382 (36)4037
High≥ 462 (27)910
Figure 4.

Duration of disease remission with hyperfractionated cyclophosphamide, vincristine, doxorubicin, and dexamethasone with the presence of none or one, two to three, or four or more adverse prognostic factors.

Comparison of Hyper-CVAD and VAD

The characteristics of patients who received Hyper-CVAD and VAD were similar except for a higher incidence of CNS disease at the time of presentation (P = 0.048), and a lower hemoglobin level (P = 0.08) with Hyper-CVAD.

The CR rates were found to be significantly better with Hyper-CVAD (Table 8), as were the survival and disease remission duration rates (Figs. 1 and 2).

Table 8. Response with Hyper-CVAD Versus VAD
ResponseHyper-CVAD (n = 288) (%)VAD (n = 218) (%)P value
  1. Hyper-CVAD: hyperfractionated cyclophosphamide, vincristine, doxorubicin, and dexamethasone; VAD: vincristine, doxorubicin, and dexamethasone.

Complete response   
 Overall9275< 0.001
 After Course 181730.05
Induction mortality55NS
Resistant disease320< 0.001
Bone marrow blasts > 5% on Day 1434480.005

There were no differences in survival or duration of disease remission duration noted with Hyper-CVAD based on whether patients received oral (n = 103) or i.v. POMP maintenance (n = 83). The 5-year survival rates were 53% and 55%, respectively (P = 0.56).

CNS Disease

Nineteen patients presented with CNS disease; 17 achieved both systemic and CNS disease remission and 11 patients developed disease recurrence (8 patients developed systemic disease and 3 patients developed CNS disease).

Among 269 patients without initial CNS leukemia, 10 (4%) developed later CNS disease: none before systemic disease recurrence, 6 concomitant with bone marrow recurrence, and 4 after bone marrow recurrence. These included 4 (7%) of 60 patients with low-risk CNS disease, 2 (1%) of 145 patients with high-risk CNS disease, and 4 (6%) of 64 patients with unknown risk CNS.

Outcome in Subjects of Interest with ALL

Of 38 patients with T-ALL, 95% achieved a CR. Their 5-year continuous CR rate was 55%, compared with 35% for the other patients (P = 0.04). The 5-year survival rates were 48% and 36%, respectively (P = 0.05).

Twelve patients (4%) had T-ALL and mediastinal disease. One patient without mediastinal disease died in CR after receipt of Course 8 of HD-MTX–Ara-C therapy. Two patients developed disease recurrence before the time of mediastinal irradiation. Mediastinal irradiation was performed for a total of 39.6 Gy, delivered in 22 fractions over 4.5 weeks (5 days on and 2 days off). Of the nine patients eligible for irradiation, seven received mediastinal irradiation and two of these seven patients developed disease recurrence (one in the bone marrow only and one in the mediastinum and bone marrow). Two of the nine patients who did not receive mediastinal irradiation developed disease recurrence (one in the bone marrow only and one in the bone marrow plus mediastinum and lymph nodes).

Among patients with Ph-positive ALL, 44 (92%) achieved a CR. Their 5-year survival rate was 12%. Eleven patients (25%) were eligible for and underwent allogeneic SCT; 6 developed disease recurrence after SCT, 3 died in CR after SCT, and 2 were alive in CR ≥ 68 months and ≥ 81 months, respectively, after SCT. Among the remaining 33 patients who achieved CR, 2 died in CR during consolidation, 12 developed disease recurrence during consolidation, 12 developed disease recurrence during maintenance therapy (6 after receiving interferon-α therapy), and 3 developed disease recurrence after maintenance therapy was completed. Thus, only 4 patients remained in CR after a median of 37 months (range, 21+ to 74+ months). These included two of five patients who also had SIg-positive, Ph-positive ALL.

Two patients (1%) had t(4;11)(q21;q23) translocations. Both patients achieved a CR but developed disease recurrence at 10 months and 16 months, respectively. Five patients had t(1;19). All 5 patients achieved a CR—1 patient died in CR at 3 months while receiving consolidation therapy, 1 patient developed disease recurrence at 9 months, and 3 patients were still alive in CR at the time of last follow-up at 57+ months, 73+ months, and 77+ months, respectively.

Side Effects

With induction chemotherapy (first course of Hyper-CVAD therapy), myelosuppression-associated complications were common. The median time to recovery of a granulocyte count > 109/L was 19 days, and was 22 days to a platelet count > 100 × 109/L. Hospitalization for side effects was required for 54% of patients. Side effects included documented infections (sepsis in 11% of patients, pneumonia in 16% of patients, fungal infection in 4% of patients, and other minor infections in 8% of patients) and fever of unknown origin in 37% of patients. Other significant side effects included neurotoxicity, which was mostly steroid related (4%), moderate to severe mucositis (4%), moderate to severe diarrhea (3%), ileus (1%), and disseminated intravascular coagulopathy requiring therapy (3%).

The side effects during the Hyper-CVAD consolidation courses were modest. The median dose intensity delivery was 100%. Myelosuppression-associated side effects included documented infections (10%: sepsis, 4%; pneumonia, 2%; herpes, 0.5%; cytomegalovirus [CMV] infection, 1%; minor infections, 4%) and fever of unknown origin (8%). Hospitalization for side effects was required after receipt of 16% of courses. Other significant side effects included neurotoxicity (7%), mostly steroid-associated mood changes or depression, mucositis and diarrhea (1%), cardiac complications (1%), and G-CSF therapy-associated bone aches (7%). The median time to recovery of counts and delivery of the next course was 20 days.

Myelosuppression-associated complications were more frequent with HD-MTX–Ara-C therapy. These included sepsis (8%), pneumonia (5%), fungal infections (0.5%), herpes (2%), CMV infection (1%), minor infections (4%), and fever of unknown origin (22%). Other side effects were renal and hepatic toxicities (1.5%), neurotoxicity (5%), skin rashes (4%), rash and desquamation of the palms and feet (2%), mucositis (4%), diarrhea (1%), Ara-C therapy-associated fever (4%), and G-CSF therapy-associated bone aches (1%). Reversible renal failure (creatinine levels ≥ 3 mg/100 mL) from MTX therapy occurred in 9 patients. Hospitalization for side effects was needed after receipt of 36% of the courses.

The median dose intensity delivered was 100%. MTX dose reductions were required—to 75% in 16% of courses, to 50% in 12% of courses, and to 25% in 2% of courses. Dose reductions of ara-C were required in 32% of courses, mostly for patients age > 60 years (1 g/m2; 16% of courses). The median time to recovery and delivery of the next course was 22 days. The median time to delivery of all 8 courses was 5.8 months (range, 4–9 months).

With the POMP maintenance, the median dose of 6-MP delivered was 150 mg daily, the median MTX dose was 15 mg/m2 weekly, and the median vincristine/prednisone delivery was 100% monthly. Three patients developed late acute myeloid leukemia (n = 1) or myelodysplastic syndrome (n = 2) after 34 months, 40 months, and 47 months, respectively.

Unusual side effects included herpes zoster or varicella (n = 17 [6%]), CMV infection (n = 10 [4%]), and Pneumocystis infection (n = 3 [1%]). Fungal infections occurred in 10 patients (4%) during induction, and in 3 patients (1%) during consolidation maintenance. Death during disease remission occurred in 19 patients (the causes were detailed earlier in the text).

DISCUSSION

The long-term follow up results with Hyper-CVAD are encouraging. Two hundred sixty-four of 288 patients (92%) who were treated achieved CR. The induction mortality rate was very low overall (5%), but was higher in patients age > 60 years compared with younger patients (15% vs. 2%; P < 0.001). With a median follow-up time of 63 months, the 5-year survival rate was 38% and the 5-year CR duration rate was 38%. Side effects were as expected, mostly attributed to myelosuppression. No unexpected long-term toxicities occurred.

Prognostic factors for long-term outcome were similar to those reported in previous studies.1–7 Older patients had a worse outcome because of both increased induction–consolidation-associated mortality and disease recurrence. Ph-positive ALL, leukocytosis, and residual disease after disease remission induction were adverse prognostic factors for long-term outcome. Based on our update, we propose simplified prognostic models for both survival and disease remission duration (Tables 3, 4, and 6). The latter divided patients into low-risk (37%), intermediate risk (36%), and high-risk (27%) groups with corresponding 5-year disease remission rates of 52%, 37%, and 10%, respectively (Fig. 4).

Ph-positive ALL remains a poor-risk disease with chemotherapy alone.9–12 In patients eligible to undergo allogeneic SCT, such a procedure should be performed as soon as possible while patients are in first CR.13–15 However, even under optimal conditions, the results are modest. In the study by Goldstone et al.,14 267 of 1389 patients (19%) in the UKALL12/ECOG Intergroup study had Ph-positive ALL. In that study, the CR rate was 76%, the 5-year survival rate was 23%, and the event-free survival (EFS) rate was 18%. The 5-year EFS rate was 37% for patients with Ph-positive ALL who had related donors and 27% for those who did not have related donors.14 Barrett et al.15 reported that the 5-year EFS rate was 22–35% for patients with Ph-positive ALL who had undergone allogeneic SCT in first CR or after disease recurrence. Recent data with Hyper-CVAD plus imatinib mesylate are encouraging.10 Among 16 patients treated with this regimen for newly diagnosed Ph-positive ALL, the CR rate was 100% and the estimated 2-year survival rate was 85%.

The role of allogeneic SCT in first CR is evolving. Earlier studies had demonstrated no advantage of allogeneic SCT versus chemotherapy in the first CR overall,17 but a potential advantage in patients with high-risk ALL (defined as Ph-positive ALL, a longer time to CR, or leukocytosis in non–T-ALL).18 A recent Medical Research Council (MRC)-ECOG Intergroup trial in adult ALL suggested a benefit from allogeneic SCT in first CR in all risk categories for patients ages 15–50 years, but to our knowledge the final results have not been reported.19 Our current institutional recommendation is to offer allogeneic SCT to patients with high-risk ALL in first CR with acceptable risk from SCT. In the future, this may change to include all patients ages 15–50 years in first CR depending on the final results of the MRC-ECOG trial.

The method of CNS prophylaxis is also changing in adult ALL. In the current study, CNS prophylaxis with IT risk-targeted therapy and high-dose chemotherapy was associated with an overall CNS disease recurrence rate of only 4%. Similar experiences were reported in childhood ALL.20, 21 Thus, future studies will likely replace craniospinal irradiation with IT plus high-dose systemic chemotherapy, which alleviates some of the irradiation-associated complications.22, 23

The value of individual components of the total adult ALL programs is being clarified. Recent studies in childhood ALL have confirmed the superiority of dexamethasone to prednisone in relation to lower incidences of CNS and systemic disease recurrences.24–26 Using more dose-intensive regimens and increasing the frequency of delivery of nonmyelosuppressive drugs (vincristine, steroids, and asparaginase) was found to improve the cure rates in childhood ALL, but these agents are difficult to deliver to adults with ALL, particularly asparaginase.27–30 Two recent analyses compared pediatric with adult ALL regimens delivered to adolescents and young adults with ALL (age range, 15–20 years).31, 32 Both suggested a survival advantage for pediatric regimens. In a study comparing the outcome of the pediatric FRALLE-93 with the adult LALA-94 French studies in adolescents with ALL (ages, 15–20 years), the reported CR rates were 94% versus 83% (P = 0.04), and the estimated 5-year EFS rates were 67% versus 41% (P < 0.0001).32 Similarly, a comparison of the pediatric Children's Cancer Group and adult Cancer and Leukemia Group B (CALGB) studies in adolescents with ALL (ages, 16–21 years) demonstrated CR rates of 96% versus 93% (P values were not significant), but the estimated 6-year EFS rates were 64% versus 38% (P < 0.01).31 This is perhaps due to better regimen intensity tolerance and delivery, better compliance of younger patients/pediatric oncologists to regimen delivery, or, quite likely, less intensity maintenance in adult ALL programs with the underappreciated but quite effective four-drug POMP combination. Analysis of the total doses of these four drugs delivered in pediatric versus adult regimens may help to clarify this issue.33–35 Conversely, the MRC studies using identical regimens for pediatric and adult ALL (UKALL trials X and XA) found significant survival differences by age. The 10-year survival rates by age group were 0-9, 62%; 10–14, 49%; 15–19, 35%; 20–39, 29%; and ≥ 40, 15%.36 In contrast to the previous opinions, they concluded that pediatric regimens were not adequate for the treatment of adult ALL.3, 36 With Hyper-CVAD, 92 patients (32%) were age ≤ 30 years. Their CR rate was 99% and their 5-year survival rate was 51%, which is similar to the pediatric ALL regimens, despite a significantly older median age of the patients (22 years vs. 15 years and 16 years, respectively).

Also of interest is the outcome of patients with ALL at the other age extreme, namely patients age ≥ 60 years. Many studies exclude these patients specifically,7, 37 some experts advocate only supportive care for these older patients,38 and trials that include them report 3-year survival rates of < 15% for patients age > 55–60 years.2–5 The 5-year survival rate for patients age ≥ 60 years who received Hyper-CVAD was 17% and the 5-year CR duration rate was 27%. This suggests a definite benefit for Hyper-CVAD in such patients, which may be superior to other regimens.

The duration of maintenance therapy was investigated recently in studies from Japan with more dose-intensive regimens.39 Shorter maintenance durations (i.e., lasting 1 year) were associated with worse outcome than longer maintenance durations (i.e., lasting 2–3 years). Intravenous POMP was not found to be superior to oral POMP therapy in our programs, and may have yielded worse results in pediatric studies.40, 41

Comparison of different adult ALL regimens is difficult because of different entry criteria (Table 9). There were no exclusions by age with Hyper-CVAD. Patients age ≥ 65 years were excluded from the early Berlin-Frankfurt Munster (BFM) studies,37 and those age ≥ 60 years were excluded in the Linker et al. regimens.7 The median age in our study group was 40 years, compared with median age ranges of 25–32 years in other adult ALL regimens.37 The incidence of patients age ≥ 60 years was 22% with Hyper-CVAD versus 9% in the recent CALGB studies.42 The more recent CALGB studies, in which the median age of the patients currently is approximately 40 years, yielded worse long-term results.5, 43 In a recent update of the latest CALGB study using the A-B-C regimen (anthracycline dose-intensive induction, high-dose MTX, and Ara-C consolidations), 163 adults with ALL were treated. Their median age was 41 years. The CR rate was 78% and the induction mortality rate was 11%. The median survival period was 1.6 years and the median disease-free survival period was 1.5 years. This may be due to inclusion of worse prognosis patients in the more recent CALGB studies.43 In the current study, there were no exclusions for poor performance staus, organ dysfunction, or Ph-positive ALL (the latter category was excluded from some programs [e.g., the Linker et al. regimen7] whereas the BFM regimen included only a “few” such patients37). Nevertheless, Hyper-CVAD was associated with favorable results, such as a CR rate of 92%, an induction mortality rate of 5%, a median survival period of 32 months, and a 5-year survival rate of 38%. Future studies that compare, in randomized trials, the Hyper-CVAD regimen with other established programs should be conducted.

Table 9. Summary of Large Adult ALL Studies
Study (reference)No. entered / evaluableMedian age (yrs)Percent age > 60 yrsCR (%)Mortality (%)Median survival/CR duration (mos)Survival/CR duration percent (at X yr)
  • ALL: acute lymphoblastic leukemia; CR: complete response; UK: United Kingdom; ECOG: Eastern Cooperative Oncology Group; A-B-C regimen: anthracycline, methotrexate, and cytosine arabinoside; Hyper-CVAD: hyperfractionated cyclophosphamide, vincristine, doxorubicin, and dexamethasone.

  • a

    Not available.

Hoelzer et al.37384/368250741128/2439 (5)/37 (5)
Durrant et al.3       
 UK ALL 1X266aa87aa—/26 (5)
 UK ALL XA618aa88aaa/28 (5)
 UK ALL—ECOG1330(age 15–55)089aaa/38 (5)
Thiebaut et al.4634/57233076918/—a27 (10)/—a
Larson5/Stock et al.43       
 801119732 85a28/2943 (5)/42 (5)
 911119835 82726/24a
 931182/7835 85aa45 (3)/41 (3)
 A-B-C regimen16341 781119/18a
Hyper-CVAD28840 92532/3238 (5)/38 (5)

The Hyper-CVAD experience has cross-fertilized other areas of research in hematologic cancers. First studied in patients with adult ALL, Hyper-CVAD later demonstrated benefit in several other hematologic malignancies including lymphoblastic lymphoma,44 mantle cell lymphoma-leukemia,45, 46 multiple myeloma,47 Richter transformation of chronic lymphocytic leukemia,48 and other lymphomas. Longer-term follow-up for patients with these other malignancies receiving the Hyper-CVAD regimen will help to place the currently reported data in context.

Current and future investigations should focus on improving the results in patients with adult ALL. This may be accomplished with more dose-intensive programs, the use of allogeneic SCT during the first CR in all eligible patients, combining chemotherapy plus imatinib in the treatment of patients with Ph-positive ALL, incorporating monoclonal antibody therapy in patients whose leukemic cells express the appropriate target (e.g., rituximab if CD20 positive, alemtuzumab if CD52 positive, and gemtuzumab if CD33 positive), including new chemotherapy agents into ALL regimens (e.g., C506U in T-ALL), prolonging the duration of maintenance therapy, and discovering new agents with anti-ALL activity.

Ancillary