Non-Hodgkin lymphoma: retrospective study on the cost-effectiveness of early treatment response assessment by FDG-PET

Although lymphomas are very chemosensitive, 50% of patients with aggressive non-Hodgkin lymphoma (NHL) are not cured with standard first-line treatment. This consists of six cycles of doxorubicin, vincristine, prednisolone and cyclophosphamide (CHOP), recently complemented with rituximab. Preliminary studies show that PET mid-treatment is a good predictor of the remission status at the end of therapy. As patients with persistent FDG uptake after three cycles are unlikely to gain a complete remission, the remaining three cycles of chemotherapy are useless. We investigated the costs and benefits for the use of PET in this early treatment setting.

We conceived a model using a conventional arm where patients receive the full regimen of six cycles of CHOP [-rituximab] and an experimental algorithm where patients receive either six cycles (PET response) or only three cycles (PET non-response). Based on a patient sample (2004-2006), we calculated the costs for hospitalisation and treatment. We took into account all costs accrued (including overhead costs). We used a sensitivity analysis by varying the most important parameters.

With a PET price of 700 euro and CHOP price (per cycle) of 1,829 euro , we can conclude to cost saving of 1,879 euro per patient. The PET price can increase up to 2,580 euro and the cost for one cycle of CHOP can decrease to 500 euro per cycle before cost savings are nil. The percentage of non-responders may be as low as 10%. The implementation of rituximab in first-line therapy only increases benefit (4,900 euro/pt).

We conclude to substantial cost savings if management of NHL patients is based on mid-treatment PET scan. The economical data we used seem to be comparable to those published in other European studies. Implementation of Mabthera in first line only increases cost savings.