Pharmacoeconomics of Stem Cell Mobilization Using Cyclophosphamide and G-Csf.

Autologous stem cell mobilization (SCM) is conventionally done using chemotherapy and G-CSF (G). Although safe and effective, patients (pts.) are exposed to risks of cytotoxic chemotherapy and myelosuppression. Plerixafor (P) and G mobilizes stem cells (SC) without any myelosuppression. Understanding the pharmacoeconomics (PE) of SCM is important so that the optimal approach can be used.

We studied PE of high dose cyclophosphamide (CY) and G SCM in 241 pts. (1/2004 to 3/2008) undergoing first autologous stem cell transplant and compared outcomes to projected costs using P+G. CY was dosed at 3 gm/m² and G started the next day (10 mcg/kg) for 10 days with a planned first day of collection on day 11. Dose escalations were per institutional protocol. Pheresis was initiated if peripheral blood (PB) had >15 CD34+ cells/uL. PE analyses were done to compute total cost and included cost of CY, G, pheresis, product processing, and clinical events. Cost was based on Medicare part B physician, laboratory, and ancillary fee schedule and was calculated from review of random patient records. This approach removed the bias of inter-institutional cost variations. Ideal Outcome (IO) was defined as >2×10⁶ CD34+ cells/kg collected on the planned day of collection in 1 or 2 apheresis without a preceding negative event that lead to additional evaluation in clinic or inpatient.

141 (61%) were males; 121 (50%) had myeloma (MM), 115 (48%) had lymphoma (L) and 5 had other diagnoses; 61 (25%) received radiation with prior therapy. Median WBC and neutrophil count prior to CY was 5.3 ×10⁹ /L (range, 1.7 to 46), and 3.3 ×10⁹/L (range, 0.95-31.37). 199 (82.6%) started pheresis on the planned first day of pheresis, 18 (7.9%) had a delay in start of pheresis (range, 1- 7 days) due to low PB CD34 count and 24 (9.9%) pts. did not proceed to pheresis due to low PB CD34+ cell count. The mean final SC dose was 10.22 ×10⁶ CD34/kg (range, 0.45 - 60.18). Pts. with MM collected more than lymphoma (11.98 vs. 6.35, P<0.0001). 6 (2.8%) pts. collected <2 ×10⁶ CD34/kg. Median number of pheresis was 1 (range, 0 - 4) with 41 (17%), and 10 (4.1%) requiring 2 and 3 pheresis, respectively. The target SC dose of >6×10⁶ CD34/kg in MM pts. was collected in 1,2, or 3 pheresis in 84 (69.4%), 98 (80.9%), and 102 (84.2%), respectively. In L pts., the target SC dose of >2×10⁶ CD34/kg was collected in 1, 2, or 3 pheresis in 68 (59%), 85 (73.9%) and 90 (74.3%), respectively. Clinical events included: febrile neutropenia (FN) clinic evaluation (7, 2.9%); FN admission (26, 10.8%); line infection (7, 2.9%); line change (8, 3.3%), gastrointestinal side effects (111, 46%), bone pain with evaluation in clinic (127, 52%) and admission for management of bone pain (9, 3.7%). Forty-three (18%) pts. were hospitalized for clinical events. IO was seen in 48 (20%) pts. 23% of MM and 15.7% of L pts. had an IO. Risk factors including prior radiation, WBC count and ANC prior to CY could not predict ability to collect SC or IO. Mean total cost of CY+G SCM was $10,732 (range, 6988-30827). IO was associated with a lower cost in overall group, (mean, $10,371 vs. $12,870, P=0.001), in MM pts. (mean, $10,511 vs. $12,152, P=0.026), and in L pts. (mean, $10,133 vs. $13,627, P=0.006). Assuming a similar distribution of IO in 100 pts. with MM and L, the projected per pt. cost of SCM would be $11,774 and $13,067 (mean, $12,421) with CY+G. Projected costs of SCM using P+G (based on published phase III data that used G dose of 10 mcg/kg without dose escalation and that the non-mobilizers had a maximum of 4 days of P) for 100 pts. with MM and L would be $12,852 and $8986 (mean, $10,919). These do not take into account costs associated with operational planning and predictability of the date of SCM with P+G, impact of the negative event on pts. quality of life with CY+G, effects of mobilization failure leading to other alternative clinical approaches including an allogeneic stem cell transplant (N=6).

Our study shows that SCM with CY+G is associated with a low incidence of IO. P+G can be justified as upfront method of SCM in pts. with MM and L from a PE perspective without any detrimental impact on SCM efficiency. As P+G is not associated with any negative clinical events related to myelosuppression, it should translate into a better quality of life for pts. SCM using P+G may allow for optimal utilization of resources which again will impact PE. These need to be validated and should be addressed in future studies of SCM.

Jagasia:Genzyme: Research Funding.