Treatment of 72 newly diagnosed Waldenström macroglobulinemia cases with oral melphalan, cyclophosphamide, and prednisone

Results and cost analysis

Abstract

BACKGROUND

Current treatment regimens for Waldenström macroglobulinemia (WM) are based on the use of oral alkylating agents. Recently, however, other more costly agents have been proposed for the treatment of WM. In the current study, the authors report on results obtained using oral melphalan, cyclophosphamide, and prednisone (MCP) to treat 72 patients with WM, and they compare these results (and the associated costs) with those observed using more aggressive protocols.

METHODS

Between July 1973 and April 2002, the authors documented overexpression of the immunoglobulin M paraprotein in 317 consecutive patients. Of these, 100 had newly diagnosed WM, and the 72 who were symptomatic were treated using the MCP protocol. Response rate, overall survival (OS), response duration, freedom from progression (FFP), event-free survival (EFS) duration, toxicity, and cost per course in Euro and U.S. dollars were evaluated for patients receiving this regimen.

RESULTS

Seventy-one of 72 patients (99%) were evaluable. Of these patients, 55 (77%) achieved a response; 7 others (10%) experienced disease stabilization, and the remaining 9 (13%) experienced disease progression. After a median follow-up of 72 months (range, 3–195 months), the median durations of EFS, FFP, response, and OS were 47, 55, 64, and 66 months, respectively. No World Health Organization Grade III or IV toxicities were observed, and side effects were limited to transient nausea, emesis, and mild neutropenia. The cost per course of the MCP regimen was $16, similar to that of standard protocols involving chlorambucil and much less than that of more aggressive protocols (price range, $91–11091) proposed for the treatment of WM.

CONCLUSIONS

Like chlorambucil-based protocols, the MCP regimen is a cost-effective and safe option for the treatment of patients with WM. Furthermore, the results obtained do not appear to be inferior to those yielded by more expensive, aggressive, and toxic intravenous protocols. Cancer 2005. © 2004 American Cancer Society.

Waldenström macroglobulinemia (WM) is a rare lymphoproliferative disorder characterized by high levels of immunoglobulin M (IgM [macroglobulin]) and lymphoplasmocytic bone marrow infiltration with or without associated lymphadenopathy and/or splenomegaly.1–5 High levels of IgM may be responsible for hyperviscosity syndrome but are generally associated with typical fundoscopic findings. The median age for patients with WM is 61 years, with a slight predominance of males over females.3, 6 Due to the low prevalence of WM, information on it treatment is scarce, and most published series deal with limited numbers of patients. Nonetheless, the standard treatment option for patients with symptomatic macroglobulinemia is the administration of oral alkylating agents such as chlorambucil, melphalan, and cyclophosphamide.4, 5, 7 These agents are able to induce responses in approximately 50% of patients (median survival duration, ∼5 years). Recently, in an effort to improve upon these results, new agents such as fludarabine and cladribine have been proposed for the treatment of WM.8–15 Although these agents have yielded higher response rates, at present, the associated survival results appear to be similar to those obtained using less aggressive protocols involving oral alkylating agents.3, 16 According to the Consensus Panel Recommendations of the Second International Workshop on Waldenström's Macroglobulinemia, alkylating agents, nucleoside analogs (e.g., cladribine, fludarabine), and the monoclonal antibody rituximab are all reasonable choices for the first-line treatment of WM, and currently available data do not point to the use of one of these types of agents over another.17

In 1989, we published results obtained using an oral combination regimen involving melphalan, cyclophosphamide, and prednisone (MCP) to treat 34 patients with newly diagnosed WM.18 In the current study, we report results obtained using this protocol in a larger series of patients with a longer follow-up duration and compare these results (and the associated costs) with those obtained using similar or more aggressive treatment schedules.

MATERIALS AND METHODS

Between July 1973 and April 2002, we documented 317 consecutive patients with IgM paraprotein overexpression. Of these, 187 (59%) had a monoclonal gammopathy of undetermined significance (MGUS), IgM type; 114 (36%) had WM; 11 (3.4%) had non-Hodgkin lymphoma (NHL); and 5 (1.5%) had IgM myeloma. In our previous study,18 the diagnosis of WM required the presence of monoclonal IgM levels ≥ 1.5 g/dL along with > 30% lymphoplasmocytic cells in bone marrow aspirates; the presence of lymphadenomegaly and/or splenomegaly was not required. Of the 114 cases of WM that were encountered, only 79 (72%) were newly diagnosed, whereas the remaining 35 (28%) had been treated previously. During follow-up, of the 187 cases of MGUS, 21 (11%) and 1 (0.53%) evolved into WM and NHL, respectively. Thus, as of April 2002, of the original 317 patients who exhibited IgM paraprotein overexpression, a total of 100 had newly diagnosed WM. Of these patients, 28 were completely asymptomatic, whereas the remaining 72 were symptomatic and, as a result, eligible for treatment.

A patient was considered to be symptomatic if he or she had at least one of the following: anemia; hyperviscosity syndrome; hemorrhagic manifestations; granulocytopenia and thrombocytopenia; or advanced tumor mass with extensive lymphadenopathy, splenomegaly, or hepatomegaly. The clinical and laboratory characteristics of the study cohort at the time of treatment are reported in Table 1.

Table 1. Patient Characteristics at the Time of Treatment

Characteristic
IgM: immunoglobulin M; Hb: hemoglobin.
No. of male/female patients	46/26
Median age (range) in yrs	61 (42–87)
Median monoclonal IgM level (range) in g/dL	3.8 (1.5–6.7)
Median Hb level (range) in g/dL	12 (6.5–16)
Hb level < 10 g/dL (%)	37
Lymphadenopathy (%)	10
Splenomegaly (%)	37
Hepatomegaly (%)	39
Hyperviscosity syndrome (%)	20

Treatment

Antineoplastic treatment consisted of melphalan (6 mg/m² on Days 1–7), cyclophosphamide (125 mg/m² on Days 1–7), and prednisone (40 mg/m² on Days 1–7), all administered orally. Courses were repeated every 4–6 weeks for a maximum of 12 courses. Thereafter, patients with responsive or stable disease received daily maintenance treatment, which consisted of chlorambucil (3 mg/m²) and prednisone (6 mg/m²), until progression. Only patients who completed the first 7 days of treatment were evaluable for response.

Response Criteria

To avoid biases, the response criteria available at the beginning of the study period (i.e., 1973) were used. Therefore, objective response and disease stabilization were defined as a decrease of ≥ 50% and a decrease of 25–50%, respectively, in baseline levels of monoclonal IgM in association with the disappearance of all symptoms present at diagnosis and in association with reductions in lymphadenopathy and splenomegaly, provided that these changes persisted for at least 3 months during induction treatment. Objective and stable disease required confirmation on two consecutive assessments made at least 30 days apart from one another. Refractory disease was defined as the absence of a response (< 25% decrease in serum monoclonal IgM levels) at the end of induction treatment. An increase of > 50% in monoclonal IgM levels in association with the reappearance or worsening of laboratory and clinical symptoms during induction treatment was defined as disease progression; the occurrence of such changes during maintenance therapy was considered indicative of recurrent disease.

Clinical and Laboratory Monitoring

All patients underwent physical examination and acquisition of a complete history. Baseline assays included complete blood count with full differential; a panel of chemistry assays (including assays of renal and hepatic function); serum electrophoresis with immunoelectrophoresis (1973–1990) or immunofixation studies (1990 onward); quantitative immunoglobulin assessment; measurement of serum viscosity and serum lactate dehydrogenase levels; bone marrow aspiration; and, in some patients, bone marrow biopsy. Clinical and laboratory assays were carried out monthly, with the exception of bone marrow aspiration and bone marrow biopsy, which were performed only at response evaluation and when clinically required.

Comparative Cost Evaluation

A comparative evaluation of cost per unit surface area per course for each of the various protocols published in the medical literature for the treatment of WM was performed. This evaluation was not based on total medical expenses, which reflect full hospital costs (including overhead costs), but rather on the cost in Euro of the agents used in each protocol (as reported in the 2003 Italian National Formulary). These costs were then converted to U.S. dollars using the following exchange rate: €1 = $1.23 (the official Italian exchange rate on September 30, 2004).

Statistical Considerations

Freedom from progression (FFP), overall survival (OS), response duration (RD), and event-free survival (EFS) duration were analyzed using the method of Kaplan and Meier. Follow-up data were current as of September 30, 2003. To account for cases of early treatment failure, FFP was calculated from the time of registration to the time of progression or recurrence. OS was calculated from the time of treatment initiation to the time of most recent follow-up or death due to any cause, and EFS was calculated from the time of treatment initiation to the time of progression, advancement to a more aggressive neoplasm, or death due to any cause. Finally, RD was calculated from the date of response evaluation to the date of recurrence. Patients who died due to causes unrelated to WM were censored in the analysis of response duration.

RESULTS

Of the 72 patients who entered the current study, only 1 was lost to follow-up, with this loss occurring a few days after the start of therapy. Therefore, 71 of 72 patients (98%) were evaluable for response. A total of 55 evaluable patients (77.4%) achieved an objective response, and 7 others (9.8%) had stable disease. As a consequence, a total of 62 patients (87.4%) received maintenance treatment. The remaining nine patients, including one who experienced a histologic transformation to high-grade NHL, had disease that was resistant to treatment. Of the 55 objective responders, 9 (14.6%) experienced a reduction of ≥ 75% in IgM levels relative to baseline, and 46 (85.4%) experienced a reduction of 50–75%.

Toxicity

Treatment was well tolerated, with acute side effects being limited to transient nausea, emesis, and mild neutropenia. Hematologic toxicity was generally reversible within a few weeks of the end of treatment. Also noteworthy was that no World Health Organization (WHO) Grade III or IV toxicities were documented in the current cohort.

EFS, FFP, RD, and OS

As of September 30, 2003, after a median follow-up period of 72 months (range, 3–195 months), the median durations of EFS, FFP, response, and OS were 47, 55, 64, and 66 months, respectively (Fig. 1A–D). The observed 24- and 48-month OS rates were 84% and 60%, respectively. For comparison, the OS rate for the 28 asymptomatic patients who had never been treated was 63% at 125 months of follow-up (Fig. 1D). Of the 62 patients who had responses, 42 subsequently experienced recurrences, 10 remain alive with sustained responses, and 10 died while still experiencing their responses.

Figure 1. Analysis of (A) event-free survival (median, 47 months), (B) freedom from progression (median, 55 months), (C) response duration (median, 64 months), and (D) overall survival (median, 66 months).

Causes of Death

To date, 58 of 71 evaluable patients have died: 10 during response, 39 during recurrence, and 9 (all with resistant disease) during induction. Causes of death for the 10 responders were as follows: adenocarcinoma of the lung (n = 2), acute myeloid leukemia (n = 1), myelodysplasia (n = 1), severe hemolytic anemia (n = 1), pulmonary embolism (n = 1), respiratory failure (n = 1), cardiac failure (n = 1), and unknown causes (n = 2). Causes of death for the 39 patients who died during recurrence included progressive disease (n = 37), septic shock (n = 1), and respiratory failure (n = 1). Finally, for the 9 patients who died during induction, the causes of death were progressive disease (n = 4), severe infection (n = 3), progression to high-grade NHL (n = 1), and cardiac failure (n = 1).

Cost Evaluation

Costs per unit body surface area per course, response rates, and OS durations for the various therapeutic protocols used to treat newly diagnosed WM are reported in Table 2. According to the 2003 Italian National Formulary, the cost of MCP chemotherapy is €13 ($16)/m² per course. Thus, the cost of MCP, along with the associated response and OS rates, is similar to that of regimens containing chlorambucil alone or chlorambucil plus prednisone (Table 2). Despite the greatly elevated cost of fludarabine and cladribine, the overall response rates yielded by these agents (range, 34–85%) are similar to those yielded by MCP and chlorambucil-based protocols. Although conclusive data regarding survival are still lacking, when available, survival data do not appear to favor these costlier regimens over other less expensive and less toxic protocols (Table 2).9 Finally, with regard to the traditional intravenous polychemotherapy protocols used to treat multiple myeloma and NHL (namely, M2 [vincristine, carmustine, cyclophosphamide, melphalan, and prednisone] and CHOP [cyclophosphamide, doxorubicin, vincristine, and prednisone]), the cost per course is €74 ($91)/m² for M2 and €238 ($293)/m² for CHOP (assuming a body weight of 70 kg); the response rates yielded by these regimens are 82% and 65%, respectively, and the corresponding median OS durations are 4.2 years and 7.3 years, respectively (Table 2).

Table 2. Analysis of the Costs, Response Rates, and Durations of Response, Event-Free Survival, and Overall Survival Associated with the Therapeutic Protocols Used To Treat Patients with Newly Diagnosed Waldenström Macroglobulinemia

Treatment	Reference(s)	No. of patients	Cost/m^2/course in Euro	Response rate (%)	Median RD (yrs)	Median EFS (yrs)	Median OS (yrs)	Cost/m^2/course in (U.S.$)
RD: response duration; EFS: event-free survival; OS: overall survival; NA: not available; MCP: melphalan, cyclophosphamide, and prednisone; M2: vincristine, carmustine, cyclophosphamide, melphalan, and prednisone; CHOP: cyclophosphamide, doxorubicin, vincristine, and prednisone; PDN: prednisone; 2-CdA: cladribine; CTX: cyclophosphamide. a Monthly cost for a patient weighing 70 kg. b Rituximab dose: 375 mg/m^2 weekly for 4 weeks (cost calculations are based on 4 weeks of treatment).
Chlorambucil	16	24	24	64	NA	NA	5.4	30
Chlorambucil	16	22	34a	75				42
Chlorambucil + PDN	3	77	18a	57	NA	NA	5.0	22
MCP	Current study	71	13	87	5.3	3.9	5.5	16
M2	19	33	74a	82	NA	NA	4.2	(91
CHOP	3	20	238	65	NA	NA	7.3	293
Fludarabine	8	20	756	79	3.3	NA	Not yet reached	930
Fludarabine	9	118	907	38	49% at 5 yrs	NA	62% at 5 yrs	1117
2-CdA alone	10–14	63	2795–3261	44–90	NA	NA	NA	3439–4011
2-CdA alone	15	16	3260	94	NA	NA	6	4010
2-CdA + PDN	15	20	3262	60	NA	NA	3.4	4012
2-CdA + CTX	15	37	2798	84	NA	NA	NA	3440
2-CdA + CTX + rituximabb	15	17	9668	94	NA	NA	NA	11,091
Rituximabb	21	15	6871	40	NA	1.3	NA	8451
Thalidomide	22	10	NA	30	NA	NA	NA

DISCUSSION

The standard primary treatment option for patients with symptomatic WM is the administration of oral alkylating agents such as chlorambucil, melphalan, and cyclophosphamide.4, 5, 7 In a trial recently conducted at the Mayo Clinic (Rochester, MN), 46 patients with previously untreated WM were randomized to receive 0.3 mg/kg oral chlorambucil daily for 7 days with repetition every 6 weeks or 0.1 mg/kg continuously administered chlorambucil; the observed response rates were 64% and 75%, respectively, the median survival duration was 5.4 years, and no differences were found between the two regimens.16 In another study, a combination of chlorambucil (8 mg/m²) and prednisone (40 mg/m²) administered orally for 10 days and repeated every 6 weeks produced a response rate of 57% and a median survival duration of 5 years in 77 patients with newly diagnosed WM.3 In the current study of 72 patients with newly diagnosed WM, the observed response rate was 87.4%, with 9 of 71 evaluable patients (12.6%) achieving a reduction of > 75% in IgM paraprotein levels; the corresponding median EFS, FFP, and OS durations were 47, 55, and 66 months, respectively, and the median RD was 64 months. These results confirm that the oral administration of alkylating agents, either alone or (as in the current study) in combination, is a good standard treatment option for patients with newly diagnosed, symptomatic WM. In an attempt to improve upon results achieved using standard treatment, some investigators have proposed the use of more aggressive intravenous protocols or new agents for the treatment of newly diagnosed WM. In particular, when the CHOP regimen was administered to 20 patients with previously untreated WM, the resulting response rate was 65%, with a median OS duration of 7.3 years, 2 years longer than the OS achieved using standard oral protocols.3 Such promising results were not found, however, in a study of 33 patients with newly diagnosed WM who were treated with the M2 protocol, which originally was used in patients with multiple myeloma. Although patients receiving M2 for WM had a high overall response rate (82%), their median OS was only 4.2 years.19

The availability of novel nucleotide analogs such as fludarabine and cladribine, coupled with the remarkable activity of these agents against other low-grade lymphoproliferative disorders, has prompted the evaluation of these agents in the treatment of WM. In a recent European multicenter trial, which followed the first known report on the use of fludarabine to treat MW (n = 11),20 20 patients with newly diagnosed WM received 25 mg/m² intravenous fludarabine daily for 5 consecutive days every 4 weeks. In that trial, objective responses occurred in 79% of patients, and after a median follow-up of 18 months, the median time to progression was 40 months, the median OS had not been reached, and the median RD was > 3 years.8 In another study, the Southwest Oncology Group examined 118 patients with newly diagnosed WM who were treated with fludarabine (30 mg/m² daily for 5 days) and found an overall response rate of 38%, with 5-year OS and PFS rates of 62% and 49%, respectively.9 With regard to other nucleotide analogs, to date, a combined total of 153 patients with newly diagnosed WM have been treated with cladribine in 6 separate trials, with response rates ranging from 44% to 94%10–15; however, data on survival and RD are available for only one of these studies.15 In that study, 90 previously untreated patients with symptomatic WM were treated with one of the following cladribine-based regimens: 1) cladribine alone (0.1 mg/kg for 7 days) as a continuous infusion over 24 hours (Regimen A; n = 16); 2) Regimen A plus 60 mg/m² oral prednisone for 7 days (Regimen B; n = 20); 3) cladribine (1.5 mg/m²) administered via subcutaneous injection every 8 hours plus cyclophosphamide (40 mg/m²) administered orally twice daily, with both agents being administered for a total of 7 days (Regimen C; n = 37); and 4) Regimen C plus weekly rituximab (375 mg/m²) for 4 weeks (Regimen D; n = 17). The overall response rate was 94% in Arm A, 60% in Arm B, 84% in Arm C, and 94% in Arm D; the median OS durations in Arm A and Arm B were 73 months and 41 months, respectively, whereas median OS had not yet been reached in Arms C and D.15 Overall, cladribine doses and administration modalities differed from study to study.10–15

Recently, Dimopoulos et al.21 administered the anti-CD20 monoclonal antibody rituximab intravenously at a dose of 375 mg/m² for 4 weeks to treat 15 patients with newly diagnosed WM. Of these patients, 40% experienced responses, and the median time to progression was only 16 months.21 The use of thalidomide at a starting dose of 200 mg daily in 10 patients with previously untreated WM yielded similar results, with responses observed in only 3 patients and recurrence documented in 1 patient after 1 year.22

The administration of oral alkylating agents rarely produces severe (WHO Grade III–IV) side effects, and when present, these side effects are limited to nausea and emesis, as was the case in the current study. In contrast, in studies involving nucleotide analogs, patients have experienced adverse effects such as myelosuppression and immunosuppression, sometimes leading to severe cytopenia and infection (including herpes zoster infection), as well as neurologic toxicities.8–10, 12, 14 Furthermore, treatment with these novel nucleoside analogs may cause a sustained reduction in monocyte and CD4-positive lymphocyte levels, with these reductions potentially contributing to an increase in the risk of fatal opportunistic infection.11

In conclusion, for patients with WM, a very rare low-grade lymphoproliferative disorder with a median age at presentation in the seventh decade of life, more aggressive regimens have not yet proven to be superior to standard regimens involving oral alkylating agents. Furthermore, cost-benefit ratios and toxicity profiles must also be carefully weighed before more expensive agents are considered for use in the first-line treatment of WM. In the future, promising results obtained in preliminary trials combining nucleoside analogs with cyclophosphamide and/or rituximab must be confirmed and subjected to further testing in randomized trials.