Stem Cell Transplantation Economic/Cost-Effectiveness Studies
The guideline developer reviewed a number of published cost analysis studies.
One group of researchers performed a prospective, non-randomized, population-based, multicenter study of 274 patients with multiple myeloma (MM) comparing autologous peripheral blood stem cell transplantation (PBSCT) with melphalan (MEL) conditioning and interferon alpha (IFNa) maintenance with 274 historical controls pooled from 5 randomized trials of conventional chemotherapy, as previously described. Additional researchers reported a companion study that collected data on costs, resource consumption, and health-related quality-of-life (HRQoL) at baseline and during periodic follow-up of the prospective trial; the same method had been used in one of the historical trials including patients treated with melphalan and prednisone (MP) for induction (n=70). In the PBSCT group, 221 patients (78%) participated in the HRQoL study, of whom 201 patients (73%) completed all questionnaires. In the MP group, 66 patients (94%) participated and 61 patients (87%) completed all questionnaires. Quality-adjusted life-years (QALYs) were calculated with the assumption of a mean 1.5-year gain in survival at the cost of a 6-month reduction in the HRQoL.
The PBSCT group had significantly prolonged median overall survival (OS) compared with the MP group (62 versus 44 months). In the PBSCT group, resource consumption included medical costs, hospital stay (including intensive care unit days), personnel costs (physicians and nurses), leukapheresis, and transfusions, and involved a cost of $24,400 (all costs are in year 2000 United States [US] dollars). Indirect costs measuring lost production (estimate of 104 days of lost unpaid employment per person) were estimated at $7,900, for a total societal cost per PBSCT patient of $32,300. The cost-utility ratio for PBSCT over MP was $27,000 per QALY, and by sensitivity analysis ranged from $20,200 to $40,000 per QALY.
Another group of researchers retrospectively calculated the treatment costs of 26 patients with MM who received MEL (n=11) or MEL plus granulocyte-colony stimulating factor (G-CSF) (n=7) compared with autologous transplantation with G-CSF mobilized peripheral blood stem cell (PBSC) re-infused after MEL (n=8). Costs included hospital days (personnel, supplies, medical services, and overhead), diagnostics, pharmacy, laboratory, insertion of central venous catheters, and transfusions. The PBSCT group had significantly lower costs for hospital days (US $7,335 versus $16,747; P<0.005), antibiotics ($2,454 versus $6,476; P<0.01), parenteral nutrition ($229 versus $2,148; P<0.001), transfusions ($1,065 versus $2,762; P<0.05), and total treatment costs ($17,908 versus $32,223; P<0.005) compared with the MEL +/-G-CSF group. The PBSCT group had significantly higher costs for G-CSF ($5,293 versus $1,393; P<0.01) compared with the MEL +/-G-CSF group. The article does not state what calendar year the costs in US dollars represent, although the article was submitted to the journal in 1993.
Analysts retrospectively compared the survival, quality of life, and therapy costs of 12 patients with MM stage III treated with MEL as induction and mobilization, PBSC collections, and autologous transplantation (group 1) with 10 patients with similar characteristics but treated with conventional chemotherapy (group 2) with 15 patients with MM stage II treated with conventional chemotherapy (group 3). The conventional chemotherapy regimen consisted of at least 6 cycles of either VAD (Vincristine [continuous infusion for 4 days], Adriamycin [continuous infusion for 4 days], and dexamethasone [orally, varying schedules]) or M2 (BCNU, eldisine, Cy, and MEL); the conditioning regimen was MEL (140 mg/m2) plus total body irradiation (TBI). Group 1 patients did not receive any maintenance therapy post-PBSCT; they were treated at time of relapse with one of the conventional chemotherapy regimens listed above or with subcutaneous IFNa plus pulse dexamethasone. Group 2 patients surviving at 6 months post-induction therapy received maintenance therapy with the regimen they did not initially receive (VAD or M2). Group 3 patients were treated with MP as maintenance therapy in case of disease response and conventional chemotherapy in case of disease progression.
The average total costs (all in 1993 US dollars) for each group including all therapy as defined above was significantly higher in group 1 ($56,700) versus group 2 ($46,555; P<0.05) versus group 3 ($37,430; P<0.02). The average total costs of therapy based on mean survival duration in group 1 was significantly lower ($350/wk) compared with group 2 ($1,862/wk: P<0.0001) but significantly higher than group 3 ($225/wk; P<0.05). When these values were adjusted for quality of life, group 1 cost $74/wk more than group 2 and $966/wk more than group 3.
A cost-minimization analysis was performed of 51 patients with MM comparing autologous bone marrow transplant (BMT) (n=14) versus PBSCT (n=37). All patients received induction therapy with VAMP (vincristine, Adriamycin, and methylprednisolone), C-VAMP (cyclophosphamide, vincristine, Adriamycin, and methylprednisolone), or verapamil, Cy (cyclophosphamide), vincristine, adriamycin, and methylprednisolone (VC-VAMP) followed by MEL (200 mg/m2) and infusion of either bone marrow (BM) or PBSC. The PBSCT group had a significantly faster time to neutrophil engraftment (16 versus 22 days; P=0.0019) and time to platelet recovery (19 versus 27 days; P=0.0019), which resulted in a shorter duration of intravenous antibiotics (12 versus 19 days; P<0.0001), reduced number of platelet transfusions (12 versus 31.5 units; P=0.0005), and shorter hospital length of stay (19 versus 27.5 days; P<0.0001). The total cost of PBSCT was 27.5% less than autologous BMT (actual costs are stated in British pounds with no conversion to US dollars and no calendar year indicated).
In another study, 91 patients with MM who received a total of 118 transplants as outpatients were compared with 160 patients with MM who received 218 transplants as inpatients. Patients treated as outpatients were younger, had a higher percentage of CD34+ cells in the apheresis product, and were more likely to have a normal serum albumin level, low B2M (Beta2-microglobulin) level, and chemotherapy-sensitive disease than inpatients. There was no significant difference in the hematologic recovery between inpatients and outpatients. Twenty-one percent of patients who underwent outpatient transplantations required admission after transplantation for nausea, vomiting, diarrhea requiring parenteral alimentation and/or severe mucositis requiring narcotic analgesics (28%), bacteremia or pneumonia (28%), febrile neutropenia and gastrointestinal toxicity (24%), persistent fever for more than 3 days (12%), or were admitted at the discretion of the physician (8%). B2M >2.5 mg/L was the only significant risk factor for hospital admission in the outpatient transplant group (58% versus 24%; P<0.001). Median hospital length of stay was 9 days for outpatients versus 15 days for inpatients (P=0.0001).
Total charge for the transplantation procedures included physician, hospital, and clinic charges. A multivariate analysis assessing age, gender, prior response to therapy, time from diagnosis to first transplantation, immunoglobulin (Ig) isotype, disease stage, number of CD34+ cells infused, serum creatinine, albumin, B2M, and lactate dehydrogenase (LDH) was performed to identify factors associated with savings. Outpatient transplantation was the only factor associated with savings. Total average adjusted charges were $13,172 (1994 US dollars) lower for outpatients compared with inpatients. Specifically, outpatients had lower hospitalization charges (50% of overall savings), pharmacy charges (42%), and pathology/laboratory charges (37%). Outpatients had higher miscellaneous charges (-30% of overall savings) including housing and caregiver costs.
The treatment and follow-up costs were calculated in a retrospective study of 29 patients with newly diagnosed MM. Costs included those for hospitalization, outpatient visits, laboratory, pharmacy, pathology, imaging (X-rays, computed tomography, etc.), apheresis, transfusions, insertion of central venous catheters, personnel, supplies, medical equipment, and overhead. All prices are stated in 1995 US dollars. Each patient was scheduled to be treated and followed in 8 phases (mean cost for each phase and number of patients completing that phase): VAD or VAMP induction ($8,400; n=29), follow-up I ($425; n=29), MEL plus whole blood rescue ($11,000; n=29), follow-up II ($1,825; n=26; 3 patients died during this phase), PBSC collections ($9,350; n=21), follow-up III ($1,250; n=17; 1 patient died during this phase), autologous PBSCT with busulfan plus Cy conditioning ($15,125; n=15; 2 patients died during this phase), and follow-up intravenous (IV) until 3 months post-hospital discharge after PBSCT ($2,400; n=13). The total mean costs of treatment and follow-up for the 13 patients who completed the program as scheduled was $44,800 and for the 16 who did not complete the entire program or who required additional therapy was $57,025.
A meta-analysis of 5 clinical trials published between 1993 and 1996 with at least 100 patients with newly diagnosed untreated MM per treatment arm was conducted, and the cost-effectiveness ratio was determined. The trials included 4 comparing MP +/- IFNa and one comparing autologous BMT versus conventional chemotherapy. Survival data were abstracted and pooled (where more than 1 trial evaluated the treatment) from the published trial data and used to calculate the mean lifetime survival (MLS) for each therapy. Costs also were abstracted from the published literature of autologous transplantation and estimated at $60,000 (1995 US$) per patient, however a sensitivity analysis of the transplantation cost data used the range of $20,000 to $120,000 as the most extreme published values. Four of the clinical trials published cost data on MP as induction therapy, which averaged $2,700 per patient; no sensitivity analysis was performed for MP because of the high precision of these data.
The pooled MLS values for the MP versus MP plus IFNa were not significantly different (3.47 versus 3.74 years; P>0.05). Autologous BMT had a significantly longer survival than the MP group (MLS 7.28 years; P<0.05). The cost-effectiveness ratio was calculated by dividing the difference in costs between MP and BMT by the difference in life years gained (LYG) per patient. Using the $60,000 estimate for BMT, the cost per LYG (cost-effectiveness ratio [CER]) was $25,710 and ranged from $7,773 to $52,616 by the sensitivity analysis.
Another group of researchers identified from the literature 1 randomized controlled trial and 2 case series of high-dose therapy with autologous SCT versus conventional-dose chemotherapy as first-line treatment of MM. Examined outcomes were LYG and event-free LYG. Cost estimates for SCT were based on out-of-area treatment costs for Central Sheffield University Hospitals and included costs for mobilization, stem cell harvest, 3-week inpatient hospital stay, outpatient follow-up, and pharmacy costs. The overall average treatment cost for SCT was £12,460 per patient. Cost estimates for conventional chemotherapy were based on the pharmacy costs of 6 to 9 courses of ABCM(Adriamycin, BCNU, cyclophosphamide, and MEL) and additional outpatient visit costs, yielding an average treatment cost of £1,980 per patient. The randomized trial data resulted in a mean 5-year survival benefit of 0.7 LYG for SCT patients, an additional 0.7 event-free LYG for SCT patients, and a CER of £14,970 per LYG. A sensitivity analysis using the case series data with information on 10-year survival rates determined the survival benefit to be 1.7 LYG and a CER of £6,160 per LYG. Fitting a mathematical Weibull curve to the survival data points yielded a 10-year survival benefit of 2.3 LYG, a CER of £4,553 per LYG, and a 20-year survival benefit of 3.8 LYG.