Dexamethasone, High-dose Cytarabine, and Carboplatin (DAC) Combination is Effective for Childhood Advanced Large-Cell Non-Hodgkin Lymphoma

Abstract

Background

To purpose of this study was to evaluate the activity and toxicity of dexamethasone, high-dose cytarabine, and carboplatin (DAC) combination therapy in children with newly diagnosed large-cell non-Hodgkin lymphoma (NHL) and to estimate the event-free and overall survival rates achieved when DAC is incorporated into a conventional regimen.

Patients and Methods

From 1991 to 1997, 20 boys and 5 girls aged 4.2 to 17.7 years who had stage III (n=21) or stage IV (n=4) large-cell NHL were treated on this study. DAC therapy was administered at the beginning of the induction phase in 2 sequential cycles and incorporated throughout a continuation phase (modified from the ACOP+ regimen) with doxorubicin, cyclophosphamide, vincristine and dexamethasone. The total duration of treatment was approximately 10 months.

Results

DAC therapy yielded a response in 22 of 25 patients (88%, 95% CI 68%-97%): complete remission in 13 cases (52%) and partial response in 9 (36%). After additional treatment with doxorubicin, cyclophosphamide, vincristine, and dexamethasone, complete remission was attained in 18 patients (72%) and partial remission in 3 (12%). The event-free survival rate (±SE) was 64% ± 9% and the overall survival rate was 80% ± 8% at 5 years.

Conclusion

The DAC regimen is well tolerated and effective for pediatric large-cell NHL.

INTRODUCTION

The non-Hodgkin lymphomas (NHLs) of childhood are primarily high-grade lesions, as defined by the National Cancer Institute (NCI) Working Formulation.(1) When classified according to the more recent WHO classification system, the pediatric NHLs comprise the Burkitt, lymphoblastic, and large-cell subtypes.(2) Significant progress has been made in improving treatment outcomes for children with these tumors. The most effective treatments for the Burkitt lymphomas are intensive, cyclophosphamide-based regimens given over a relatively short period (4-8 months),(3-13) whereas those for lymphoblastic disease are often derived from regimens used to treat high-risk acute lymphoblastic leukemia (ALL), which feature intensive chemotherapy given over a prolonged period (18-30 months).(14-20) Determining the optimal therapy for large-cell NHL has proved problematic, partly because of the biologic heterogeneity of these tumors and because of the wide spectrum of treatment strategies reported.(14, 21-33) For example, in Europe, treatment has been based on immunophenotype, whereas in the United States, patients have historically been treated according to histologic findings and disease stage, regardless of immunophenotype.(10, 12, 33, 34) The histology-directed treatment approach has resulted in a 50%-70% long-term event-free survival rate for children with advanced-stage large-cell NHL.(14, 19, 21, 28-32) (see Table 1) The histology-based treatment approaches rely heavily on anthracyclines and alkylating agents and are frequently CHOP-based regimens; however, some regimens include other agents such as methotrexate, mercaptopurine, bleomycin, and cytarabine.

Table 1.

Reported Treatment Outcomes After Histology-Directed Therapy for Advanced-Stage Large-Cell NHL

Regimen	No. of Patients	Disease Stage	Outcome	Reference
COMP	30	III and IV	3 yr FFS = 46%	(14)
LSA2L2	16	III and IV	3 yr FFS = 44%	(14)
Modified LSA2L2	13	III	3 yr DFS = 70%	(19)
ACOP+	22	III and IV	4 yr DFS = 67%	(30)
APO vs. ACOP+	62	III and IV	5 yr EFS = 72% ± 6%	(31)
	58	III and IV	5 yr EFS = 62% ± 7%	(31)
COMP vs.aD-COMP	106	I – III	10 yr EFS = 48% ± 4.9%	(28)
	14	IV	10 yr EFS = 44%
CHOP	21	III and IV	3 yr EFS = 62% ± 11%	(21)
MACOP-B	11	III and IV	3 yr EFS = 55% ± 16%	(32)
DAC+	25	III and IV	3 yr EFS = 64% ± 10%	Present study

^aRandomized trial: no significant difference in EFS between COMP and D-COMP.(28)

^*COMP: cyclophosphamide, vincristine, methotrexate, and prednisone; LSA2L2: cyclophosphamide, vincristine, prednisone, daunorubicin, cytarabine, L-asparaginase, BCNU, hydroxyurea, methotrexate, thioguanine; ACOP+: doxorubicin, cyclophosphamide, vincristine, prednisone, methotrexate (low dose), 6-mercaptopurine, and L-asparaginase; D-COMP: daunorubicin-COMP; CHOP: vincristine, prednisone, cyclophosphamide, and doxorubicin; MACOP-B: methotrexate, doxorubicin, cyclophosphamide, vincristine, prednisone, and bleomycin; DAC: dexamethasone, cytarabine (high dose), and carboplatin. DFS, disease-free survival; EFS, event-free survival; FFS, failure-free survival; POG, Pediatric Oncology Group.

We and others have demonstrated that the combination of dexamethasone, high-dose cytarabine, and cisplatin (DHAP) is active against relapsed large-cell NHL.(32) In an effort to improve treatment outcome while reducing the total anthracycline and cyclophosphamide exposure, we incorporated a modified DHAP regimen into the treatment for patients with newly diagnosed large-cell NHL; we used carboplatin instead of cisplatin to minimize ototoxicity and nephrotoxicity. The combination of dexamethasone, cytarabine, and carboplatin (DAC) was administered at the beginning of the induction phase as two sequential courses and incorporated throughout a continuation phase modified from the ACOP+ regimen, which features doxorubicin, cyclophosphamide, vincristine, and prednisone.(30) Here we report the activity of the initial courses of DAC given during induction, as well as the treatment efficacy and toxicity of the entire treatment program.

PATIENTS AND METHODS

Twenty-five children (20 boys, 5 girls; age 4.2-17.7 years) with advanced-stage large-cell NHL, as defined by the NCI Working Formulation, (1) were evaluated and treated at our institution between 1991 and 1997. More recently, classification according to the WHO system was performed in cases for which there was available tissue.(2) Staging workup included CT imaging of neck chest abdomen and pelvis, nuclear imaging (egs., gallium scan, bone scan), bilateral posterior iliac crest bone marrow aspiration and biopsy examination and lumbar puncture for examination of cerebrospinal fluid examination. Upon completion of this workup, lymphomas were staged according to the St. Jude system described by Murphy.(35) The treatment protocol was approved by our institutional review board, and informed consent permission for treatment was obtained for all research participants or their legal guardians.

The treatment scheme including dosing and scheduling of therapy, is summarized in Figure 1 and Table 2. Of note, the carboplatin was delivered as a continuous intravenous infusion over 24 hours to achieve a systemic exposure (AUC) of 8 mg/ml min.(36) We calculated the carboplatin dose (in mg/m²) from: AUC × [(0.93 × GFR) + 15]; the GFR (glomerular filtration rate: ml/min/m²) was estimated from ^99mtechnetium DTPA (Tc99)-based serum clearance studies.

Figure 1.

Treatment scheme for the DAC regimen. Treatment was given over a 10-month period. If there was no response to DAC during induction therapy, the patient was taken off the study.*

Table 2.

Outline of the DAC Regimen

Induction: 9 weeks
Block 1 (DAC)
Dexamethasone	Days 1-4	40 mg*, i.v.
Cytarabine	Day 2	2 g/m2 per dose, i.v., ×2, 12 h apart
Carboplatin	Day 1	To achieve a systemic exposure† of 8 mg/mL.min; CI over 24 h
G-CSF	Day 4 onward	10 μg/kg per day, subcutaneously, daily
Block 2 (DAC)
Same as Block 1
Block 3 (CHOD)
Cyclophosphamide	Day 1	800 mg/m2, i.v.
Doxorubicin	Day 1	75 mg/m2, i.v.
Vincristine	Day 1	1.5 mg/m2, i.v. (max. 2 mg)
Dexamethasone	Days 1-14	12 mg/m2, i.v. or orally, in 3 divided doses, daily
Consolidation (VAML): 3 weeks
Vincristine	Day 1	1.5 mg/m2, i.v.
Doxorubicin	Day 1	30 mg/m2, i.v.
Mercaptopurine	Days 1-5	225 mg/m2, orally, in 3 divided doses, daily
L-Asparaginase	3 times per week (e.g., every MWF)	10,000 U/m2, intramuscularly, ×6 (i.e., for 2 weeks)
Continuation (×3): 27 weeks
Block 1 (DAC)
Dexamethasone	Days 1-4	40 mg*, i.v.
Cytarabine	Day 2	2 g/m2 per dose, i.v., ×2, 12 h apart
Carboplatin	Day 1	To achieve a systemic exposure† of 8 mg/mL.min; CI over 24 h
Block 2 (VMMD)
Vincristine	Day 1	1.5 mg/m2, i.v.
Mercaptopurine	Days 1-5	225 mg/m2, orally, in 3 divided doses, daily
Methotrexate	Day 1	60 mg/m2, i.v.
Dexamethasone	Days 1-5	40 mg*, i.v.
Block 3 (CHOD)
Vincristine	Day 1	1.5 mg/m2, i.v.
Cyclophosphamide	Day 1	800 mg/m2, i.v.
Doxorubicin	Day 1	30 mg/m2
Dexamethasone	Days 1-5	40 mg*, i.v.
Prophylaxis and Treatment for CNS Involvement
Bone marrow is positive or there are primary tumors in the head or neck	Give methotrexate, IT, on days 1, 8, 22, and 43 of induction and on day 1 of each continuation cycle	Dose by age: < 1 yr, 6 mg 1-2 yrs, 8 mg 3-8 yrs, 10 mg ≥ 9 yrs, 12 mg
CNS is positive at diagnosis	Give additional doses of methotrexate, IT, on induction days 15, 29, and 36

i.v.: intravenously; CI: continuous intravenous infusion; G-CSF: granulocyte colony-stimulating factor; MWF: Monday, Wednesday, Friday; IT: intrathecally.

Give therapy only when the absolute neutrophil count (ANC) is ≥ 300 × 10⁶/L (300/mm³) and the platelet count is > 100 × 10⁹/L (100,000/mm³) before each block.

^*20 mg for children < 5 years old.

^†Area under the concentration vs time curve (AUC).

Overall survival time was measured from the date of diagnosis to the date of death or date of last contact. Event-free survival was measured from the date of diagnosis to the date of induction failure, relapse, death or last follow-up examination. Overall survival and event-free survival were estimated with Kaplan-Meier methods;(37) the associated standard errors were calculated by the method of Peto and Pike.(38)

Toxicity was graded according to the NCI’s toxicity criteria. Complete blood counts and blood chemistry values were checked before starting each course of chemotherapy, which was delayed if the absolute neutrophil count (ANC) was < 0.3 × 10⁹/L or the platelet count was < 100 × 10⁹/L . We assessed renal function by measuring serum creatinine and blood urea nitrogen concentration and by estimating the GFR from Tc99 clearance measurements. Echocardiography and electrocardiography were performed before each cycle of cyclophosphamide, doxorubicin, vincristine and dexamethasone (CHOD) to evaluate for anthracycline-induced cardiac toxicity. Audiograms were performed before each dose of carboplatin to evaluate ototoxicity. After completion of therapy, we performed a physical examination, blood chemistry analyses, complete blood counts and diagnostic imaging studies (as outlined above for initial staging), to evaluate remission status and treatment-related toxicity. These evaluations were made monthly for the first 6 months, every 2 months for the next 6 months, every 3 months in the second year after therapy, and annually thereafter. Echocardiograms, electrocardiograms, and audiograms were obtained annually to screen for chemotherapy-related organ dysfunction.

RESULTS

Twenty-one patients had stage III disease and 4 had stage IV disease. The clinical features, including sites of disease at the time of diagnosis, are summarized in Table 3.

Table 3.

Clinical Features of 25 Children Given Treatment with the DAC Regimen

No.	Disease Stage	Sites of Disease	Pathology	Response
				DAC	DAC + CHOD	Relapse	Status
1	III	Bones, soft tissue	DLBCL	CR*	--	--	A
2	III	Lymph nodes	DLBCL	CR	--	--	A
3	III	Mediastinum, lung	MLBCL	PR	PR	Mediastinum	A
4	IIIa	Intestine	DLBCL	CR	--	--	A
5	III	Mediastinum, lymph nodes	LC-T,NOS	PRb	PRc	--	D
6	III	Mediastinum, bone, pancreas	LC-T,NOS	CR	--	--	A
7	III	Skin, lymph nodes, bone	ALCL	PRc	--	--	A
8	III	Mediastinum, bones, lymph nodes, soft tissue	ALCL	PR	CR*	Bones, lung	D
9	III	Mediastinum, abdominal mass, lymph nodes	ALCL	PR	CR*	--	A
10	III	Paraspinal mass, bones	ALCL	PR	PRc	--	D
11	III	Mediastinum, soft tissue, bone	ALCL	NRc	--	--	D
12	III	Lymph nodes, pleural mass	ALCL	NR(mixed)c	--	--	A
13	III	Lung, lymph nodes, skin, bones	ALCL	CR	--	--	A
14	IV	Lymph nodes, bone, spleen, bone marrow	ALCL	CR*	--	--	A
15	III	Lymph nodes, spleen	ALCL	CR*	--	--	A
16	III	Lymph nodes, spleen	ALCL	CR	--	--	A
17	III	Lymph nodes	ALCL	CR	--	--	A
18	III	Bones, mediastinum, skin, lymph nodes, kidney	ALCL	CR	--	Skin, lymph node	D
19	III	Mediastinum, lymph nodes	LC-B, NOS	NRc	--	--	A
20	III	Mediastinum	MLBCL	PR	CR*	--	A
21	IV	Bone, mediastinum, lymph nodes, bone marrow	LC-B, NOS	PR	CR*	Mediastinum, bone, bone marrow	D
22	III	Lymph node, bone, soft tissue	DLBCL	CR*	--	--	A
23	III	Mediastinum, lymph nodes, bones	LC, NOS	CR	--	--	A
24	IVa	Brain	LC, NOS	CR	--	--	A
25	IV	Mediastinum, pericardium, bone marrow	LC, NOS	PR	CR*	--	A

^*NR, no response; PR, partial response; CR, complete response (9 patients in whom diagnostic imaging revealed minimal abnormality of questionable significance at the primary site of disease were categorized provisionally as having a CR); A, alive; D, dead; DLBCL, diffuse large B-cell lymphoma; MLBCL, mediastinal large B-cell lymphoma; ALCL, anaplastic large cell lymphoma; LC-T NOS, T-large cell not otherwise specified; LC-B NOS, B-large cell not otherwise specified (with atypical features); LC NOS, large cell not otherwise specified.

^aIncomplete resection (microscopic residual).

^bEmergency mediastinal irradiation at diagnosis.

^cTreatment failure (NR or PR with progressive disease); patient taken off study

Twenty-two of the 25 patients (88%; 95% CI 68%- 97%) had a complete (n=13) or partial (n=9) response to the initial 2 sequential courses of DAC. With additional therapy (the CHOD regimen), some of the patients who had had partial responses to DAC had complete responses, resulting in a complete remission rate of 72% and partial remission of 12%. The 5-year event-free survival rate (± SE) was 64% ± 9% (median follow-up, 9.1 years; Figure 2). With retrieval therapy, including autologous hematopoietic stem cell transplantation (6 patients), a 5-year overall survival rate of 80% ± 8% was achieved (Figure 2).

Figure 2.

Event-free survival and overall survival for 25 children treated with the DAC regimen. The triangles denote patients still at risk for adverse events.

Grade 4 hematologic toxicity was observed in 73% of patients. Overall, episodes of fever with neutropenia occurred after 1 of every 3 blocks of DAC administered. There were 4 episodes of grade III mucositis and grade III transaminase elevation and 1 episode of grade III hyperbilirubinemia - none of these episodes followed treatment with DAC. There was no evidence of substantial renal toxicity as shown by stable serum creatinine concentration and Tc99 clearance rates. No patients had a significant decrease in the shortening fraction on the echocardiogram. Six patients had loss of high-frequency hearing, with decreased acuity above 2000 Hz (n=2), 6000 Hz (n=2), or 8000 Hz (n=2). None of these patients required hearing amplification.

DISCUSSION

Improving the treatment outcome for children with large-cell NHL, while reducing morbidity and the risk of late adverse effects, remains a challenge. Various approaches to address this problem have been investigated. For example, the former Pediatric Oncology Group piloted a regimen that incorporates intermediate-dose methotrexate and cytarabine into the APO regimen, which features doxorubicin, vincristine, and prednisone.(39) The addition of intermediate-dose methotrexate and cytarabine, however, failed to significantly improve outcome.(39) High-dose methotrexate and agents such as ifosfamide have been featured in successful European trials performed by the German (BFM), and French (SFOP) cooperative groups.(40-45) In an attempt to find new, active drug combinations that have fewer acute and late adverse effects, we piloted this study in which DAC was given during both induction, and continuation therapy. The CHOP component of the ACOP+ regimen is widely accepted as being active against large-cell lymphoma in adults, but reports of the activity of low-dose methotrexate, mercaptopurine, and L-asparaginase against large-cell NHL are lacking. Nonetheless, since these drugs were components of the successful ACOP+ regimen, we maintained their use in this regimen.

In this study, the DAC combination was very active, as shown by the 88% rate of complete or partial response after 2 cycles of treatment. However, after completion of the entire treatment regimen (induction, consolidation, and continuation), the 5-year event-free survival rate was 64% ± 10%, which is similar to those obtained with more conventional histology-directed CHOP-based regimens (see Table 1). Likewise, Fisher, et al.(46) made a similar observation that the incorporation of additional agents in the treatment of large-cell NHL in adults did not improve long-term outcome over that obtained with CHOP treatment alone. To obtain a more complete evaluation of the potential benefit of adding agents to the CHOP regimen for children with large-cell NHL, further studies are clearly needed.

Although the inclusion of DAC in our study did not appear to improve the event-free survival rate, it is important to note that with this regimen, we achieved similar results to those obtained previously, but with lower cumulative doses of anthracycline and cyclophosphamide. The most severe acute toxicity associated with this DAC-based regimen was myelosuppression with associated febrile neutropenia, which occurred after one-third of the DAC courses administered. For this reason, G-CSF (filgrastin) was given after the first two courses of DAC, during induction. Rarely, grade III mucositis and grade III transaminase elevation occurred, but not after the DAC cycles. Of note, the use of carboplatin did not produce significant renal toxicity, as determined by serum creatinine and Tc99 clearance studies. A few patients developed mild ototoxicity, which resulted primarily in hearing loss involving high-frequency tones; such loss would not affect conversational speech and hearing.

Sequelae of existing CHOP-based therapy that arouse the most concern include cardiac toxicity,(47-56) infertility,(57) and second malignancies. Cyclophosphamide and other alkylating agents result in a dose-related depletion of germinal cells which is generally more severe in males than in females. Recovery of spermatogenesis after cyclophosphamide-induced azoospermia is dose-related. Sterility is likely at cumulative doses ≥ 7.5 g/m², while fertility is usually maintained at doses < 4 g/m².(57) Our regimen prescribed a cumulative cyclophosphamide dose of 3.2 g/m², which should permit the preservation of fertility in most patients. The Pediatric Oncology Group published the results of a study comparing APO with ACOP+ that demonstrated that cyclophosphamide could be eliminated if the regimen maintained a relatively anthracycline-rich backbone.(31)

The desire to avoid anthracycline-related cardiac toxicity is an important factor influencing protocol development in pediatric oncology patients. Whereas cumulative doses of doxorubicin of 550 mg/m² are generally well tolerated in adults, cumulative doses of doxorubicin of 45 mg/m² to 300 mg/m² may cause abnormalities of ventricular after load and contractility in children.(48) Factors predictive of cardiac dysfunction include a high cumulative anthracycline dose and higher intensity of anthracycline dosage, female gender, young age, use of mediastinal irradiation, and long time interval since completion of therapy.(47, 49, 50, 53-56) All of our patients had normal cardiac function according to routine echocardiography and electrocardiography during and after completion of therapy. Our regimen prescribed a total cumulative dose of 195 mg/m² of doxorubicin, which we anticipate will be associated with a low risk of late-onset cardiac toxicity. However, continued monitoring of cardiac function in these children is essential to evaluate delayed effects that may emerge with time.

Other groups have examined the need for anthracyclines as well as novel ways to preserve cardiac function when anthracyclines are used. The former Children’s Cancer Group’s randomized trial of COMP versus D-COMP (see Table 1) demonstrated that the addition of doxorubicin did not improve outcome compared to that achieved with COMP alone.(28) Nevertheless, as suggested by the Pediatric Oncology Group’s Study,(31) anthracyclines remain an important class of agents in large cell NHL treatment. While cardioprotectants, such as Zinecard (dexrazoxane), may reduce cardiotoxicity in patients who receive a high cumulative dose of anthracyclines, careful monitoring of such use is recommended because of the potentially increased risk of therapy-related myeloid neoplasms.(58)

Previous exposure to alkylating agents, anthracyclines, epipodophyllotoxins, and radiation has been associated with the development of second malignancies. For patients treated with the DAC regimen, we anticipate a low risk of treatment-related secondary carcinogenesis, because this regimen comprises relatively low cumulative doses of cyclophosphamide and doxorubicin, and does not incorporate any epipodophyllotoxins or involved-field radiation therapy. Additional follow-up of this cohort is required to establish this supposition.

Studies suggest that an immunophenotype-directed approach may be more effective in the management of large-cell NHL in children.(10, 12, 41, 42, 59) In a Pediatric Oncology Group study comparing the APO and ACOP+ regimens, children with B-cell tumors had a significantly better treatment outcome than did those with a non-B-cell immunophenotype; however, the sample size in that study was relatively small.(59) Among advanced-stage cases, the 3-year event-free survival rate was 96% ± 5% (SE) for B-cell cases, 67% ± 12% for patients with T-cell lymphomas, and 74% ± 13% for those with non-T, non-B-cell immunophenotypes. The French cooperative group (SFOP) reported an excellent result for the treatment of B-cell large-cell NHLs with their LMB89 regimen, which they use for all B-cell NHL, including Burkitt lymphoma.(10, 42) The LMB89 regimen includes courses of cyclophosphamide, doxorubicin, vincristine, and prednisone as well as courses of high-dose methotrexate (3 g/m² per dose) and low-dose cytarabine in most cases. The German BFM cooperative group, which also uses an immunophenotype-directed approach, has reported excellent results for treatment of B-cell-large cell NHL.(12, 41) All six patients with diffuse large B-cell lymphoma in our study (2 with mediastinal large B-cell), are currently alive and disease-free (one is in second remission after salvage therapy). Therefore, it appears that a novel or more intensive regimen may be necessary to treat the large-cell lymphomas of non-B-cell immunophenotype.

In summary, the DAC combination is active in previously untreated pediatric large cell lymphoma and the entire DAC+ regimen is effective and well tolerated. Although DAC+ did not produce a result superior to other CHOP-based regimens, the low cumulative dosages of cyclophosphamide and anthracycline should be associated with preservation of fertility and a low rate of clinically significant cardiac toxicity. Many of the patients who experienced treatment failure were successfully salvaged with autologous hematopoietic stem cell transplantation, resulting in an excellent overall survival rate (80% ± 8% at 5 years).