Single-agent high-dose melphalan salvage therapy for Hodgkin's disease: Cost, safety, and long-term efficacy

D. A. Stewart 2 5 D. Guo 1 2 5 J. A. Sutherland 2 5 B. A. Ruether 0 2 5 A. R. Jones 0 2 5 C. deMetz 2 5 J. Klassen 2 5 A. Chaudhry 2 5 C. B. Brown 2 3 5 J. A. Russell 2 3 5 M.-C. Poon 0 2 5 0 Department of Apheresis 1 Department of Epidemiology 2 'Department of Medical Oncology 3 Alberta Bone Marrow Transplant Program, Tom Baker Cancer Centre, Foothills Hospital, and University of Calgary , Calgary, Alberta , Canada 4 Department of Radiation Oncology 5 Douglas Stewart, MD Tom Baker Cancer Centre 1331-29 Street, N.W. Calgary, Alberta Canada T2N4N2 Summary Background: Few data are available on the cost, safety, and long-term efficacy of single-agent high-dose melphalan (HDM) followed by autologous bone marrow (ABMT) or blood stem cell (ABSCT) transplantation in the salvage therapy of Hodgkin's disease (HD). Patients and methods: From February 1981 to September 1996, 23 patients with relapsed (n - 15) or refractory (n = 8) HD received salvage therapy with HDM 140-200 mg/m2 followed by non-cryopreserved ABMT (n = 18) or cryopreserved ABSCT (n = 5). The cost of HDM/ABSCT in 1996, from initial consultation until transfer back to referring physician, was determined and compared to the estimated costs of two multi-agent regimens commonly used for HD. Introduction High-dose chemotherapy (HDCT) and autologous bone marrow (ABMT) or blood stem cell (ABSCT) transplantation cures approximately 25%-60% of selected patients with relapsed or refractory Hodgkin's disease (HD) [1]. Although the optimal HDCT regimen is unknown, most involve combinations of non-cross-resistant agents. We participated with another centre in a phase III study investigating single-agent high-dose melphalan (HDM) salvage therapy for HD [2], and elected to continue this study locally because the treatment was relatively easy to deliver and the results were encouraging. Few long-term data are published regarding such single-agent HDCT. The purposes of this single-centre study were: 1) to determine the outcome of 23 consecutive HD patients treated with HDM, and 2) to determine the 1996 cost of HDM/ABSCT in order to estimate the current cost savings of this regimen relative to the multi-agent HDCT regimens used more commonly world-wide. Patients and methods Between February 1981 and September 1996, 23 patients with relapsed (n = 15) or refractory (n = 8) HD, aged 17-45 years (median 29 years) were treated with HDM and ABMT or ABSCT. The pre-transplant characteristics of these patients are listed in Table 1. This report includes eight patients from our centre originally reported by Russell et al. [2]. Of patients who had an initial complete remission duration of more than one year, three received HDM after failing just one chemotherapy regimen (two of the three had also failed prior radiotherapy). This HDCT protocol received ethical approval by our Institutional Review Board, and all patients gave written informed consent prior to receiving HDM. Melphalan 140-200 mg/m2 was infused on day - 1 over five minutes, and was followed immediately with i.v. hydration and diuresis. One patient received 140 mg/m2 and a second patient received 160 mg/m2 as part of the initial phase I study [2]. The remaining 21 patients received 200 mg/m2 of melphalan. Non-cryopreserved autologous marrow (n = 18) or cryopreserved autologous blood stem cells (ABSQ (n = 5) were reinfused on day 0, 24 hours or less after HDM. The median number of nucleated marrow cells harvested was 1.9 x 108/kg (range 1.2-2.7 x 108/kg). ABSC were used after 01/93 and were collected following mobilization with cyclophosphamide 2 g/m2 on day 1 plus granulocyte colony-stimulating factor (G-CSF) 300 ug or 12(7-55) 59(14-144) 480 ug (5-8 ug/kg) subcutaneously daily from day 8-12 (n = 4), or following G-CSF alone from day 1-4 {n = 1). The median number of CD34 positive ABSC infused was 6.2 x 106/kg (range 2.0-16.0 x 106/ kg). Isolation procedures have never been used for patients receiving ABMT/ABSCT at our centre. Eight patients received post-transplant involved field radiotherapy to sites of residual abnormality or sites of previous bulky disease. Overall survival (OS) and progression-free survival (PFS) distributions from the time of stem cell infusion were estimated using the method of Kaplan-Meier [3]. Non-hematologic toxicity was graded by the method described by Bearman et al. [4]. A costing initiative was undertaken in 1996 by a group of health care professionals from the departments of administration, finance, nursing, and medicine because of restructuring in our health system and Government need to determine cost-effectiveness of expensive medical therapies. Costs were determined for 15 Hodgkin's and nonHodgkin's lymphoma patients receiving HDM/ABSCT in 1996 in Calgary. Inpatient and outpatient activity were recorded in detail and costed from the time of initial consultation, through work-up, blood stem cell mobilization, apheresis, hospitalization for ABSCT, and outpatient follow-up visits until the time of transfer back to the referring physician. Contributing to the costs were inpatient and outpatient nursing, unit clerk, psychosocial, dentistry, blood bank, laboratory, procedures, pharmacy, supplies and equipment. Because transplant physicians are on salary, their costs per case could not be determined and were not included in this analysis. Post-transplant radiotherapy costs were also not included. Overhead costs were added at 25% of the direct costs to cover such things as maintenance of building and equipment, nutrition, laundry, health records, finance, and administration for support areas (laboratory, pharmacy, etc.). CBV (cyclophosphamide, BCNU, VP-16) and BEAM (BCNU, etoposide, ara-C, melphalan) are the two most commonly used high dose multi-agent regimens for HD worldwide [5, 6]. Since these regimens have never been used in Calgary, their costs could not be determined directly. We attempted to estimate their costs by adding the chemotherapy costs and the expected number of extra pancytopenic days to our melphalan data. HDM was found to be feasible, well tolerated, and produced a five year PFS rate of 50% for patients with relapsed/refractory HD. Although this series of 23 patients is too small to draw firm conclusions regarding relative efficacy, this PFS rate is similar to the reported 23%-52% PFS rates of the multi-agent H D C T regimens in similar patient populations [1, 5-11]. No grade III or IV regimen-related toxicity occurred following HDM. A recent review also concluded that HDM very rarely causes severe toxicity to major organs such as the heart, lungs, liver or kidneys [12]. In contrast, comparable toxic death rates during the 1980s and early 1990s of multi-agent HDCT regimens for HD were approximately 10% (0%-23%) [1, 5-11]. Because HDM is given on day - 1 , neutropenia generally does not begin before day +6. Multi-agent HDCT regimens start on day - 6 and produce neutropenia by day 0. Thus, compared to other regimens, HDM is a (...truncated)