Double Relapsed and/or Refractory Multiple Myeloma: Clinical Outcomes and Real World Healthcare Costs

Double relapsed and/or refractory multiple myeloma (DRMM), MM that is relapsed and/or refractory to bortezomib and lenalidomide, carries a poor prognosis. The healthcare costs of DRMM have not previously been reported. We analyzed detailed medical resource utilization (MRU) costs, drug costs and outcomes for 39 UK patients receiving standard DRMM therapy. Median OS in this cohort was 5.6 months. The mean cost of DRMM treatment plus MRU until death was £23,472 [range: £1,411–£90,262], split between drug costs £11,191 and other resource use costs £12,281. The cost per assumed quality-adjusted life year (QALY) during DRMM was £66,983. These data provide a standard of care comparison when evaluating the cost-effectiveness of new drugs in DRMM.


Introduction
Patients with treatment-refractory malignancy have poor outcomes and high healthcare costs. In Multiple Myeloma (MM), the introduction of the proteasome inhibitor bortezomib and immunomodulatory drugs (IMiDs) thalidomide and lenalidomide has improved survival over the last decade [1], but increased the cost of treatment. While these drugs can result in remission, most patients will relapse with increasing symptom burden and worsening prognosis [2]. Double relapsed and/or refractory multiple myeloma (DRMM), MM that is relapsed and/or refractory to bortezomib and lenalidomide [3], carries a poor prognosis and therapeutic options remain limited. An IMWG retrospective analysis of patients who relapsed following bortezomib and at least one of the IMiDs showed a median overall survival (OS) and progression-free survival (PFS) of 9 months and 5 months respectively [4]. Only those potentially eligible for further clinical trials with Eastern Cooperative Oncology Group Performance Status (ECOG PS) 0-2 were included, indicating survival may have been overestimated relative to the entirety of this heavily treated population.
Studies of healthcare costs for relapsed and/or refractory MM patients receiving bortezomib and/or lenalidomide based regimens have been undertaken in a variety of countries, using 'real-world' data and/or economic modelling, often from patients enrolled in clinical trials [5][6][7][8][9]. However, none of them address MRU for patients relapsed after bortezomib and lenalidomide. The third-generation IMiD pomalidomide and second-generation proteasome inhibitor carfilzomib have recently been licensed and have efficacy in DRMM [10]. The cost burden and clinical outcomes outside of trials of this phase of the disease has never been established, so new therapies with efficacy in DRMM have no published benchmark against which to judge cost-effectiveness. To ensure resources are allocated appropriately, the cost-effectiveness evaluation of these therapies in the DRMM setting must involve non-clinical trial, real world MRU data from relevant patients. Our data provides a standard of care comparison when evaluating cost-effectiveness of new drugs in DRMM.

Ethics Statement
All patients whose data were included in this study had provided written consent for the use of their anonymised data for the purposes of audit and service improvement by the Thames Valley Cancer Network, UK. This study was appropriately registered using the clinical audit project proposal system at Oxford University Hospitals NHS Trust, UK. The Clinical Audit Lead reviewed and approved the proposal, and in line with the UK NHS National Research Ethics Service (NRES) guidance, deemed it not to require IRB or ethics committee approval.
Anonymised data on clinical outcomes, anti-myeloma therapies prescribed and MRU were obtained for 39 DRMM patients pre-treated with or intolerant to bortezomib and lenalidomide in the Thames Valley Cancer Network, UK from 2011 to 2014 (Table A in S1 File). Based on the UK National Institute for Health and Care Excellence (NICE) guidelines, bortezomibbased therapy was used for first relapse unless contraindicated. Lenalidomide-dexamethasone combination was approved for second and subsequent relapse. Relapsed myeloma patients were identified using pharmacy-generated lists of all sequential lenalidomide recipients between January 2011 and July 2013 at Oxford University Hospitals and the Royal Berkshire Hospital, Reading, UK. The strategy of using hospital pharmacy dispensing data to identify subjects ensured all lenalidomide recipients during the study period were included in the analysis, as long as they had Multiple Myeloma and had progressed on or were refractory to lenalidomide, according to IMWG criteria [11]. 34 (87%) patients were pre-treated with lenalidomide and bortezomib, and had relapsed following, or failed to tolerate both therapies. 5 patients had not received bortezomib due to pre-existing peripheral neuropathy, sufficiently severe to contraindicate its use.
OS and PFS were calculated from the start date of the first therapy following relapse after lenalidomide ('1 st DRMM therapy') until either IMWG criteria for progression/relapse were reached or death. The data censor date was 10 January 2014 for patients still alive. For each patient the following occurrences of MRU were retrieved from health care records from the start of each successive DRMM therapy until death or censoring: detailed drug regimens, outpatient clinic and chemotherapy unit attendances, inpatient/hospice admissions, supportive therapies, medical procedures, radiological investigations, blood product transfusions and blood tests. Costings were calculated using NHS reference costs 2012-13. These were combined to give one MRU cost from the start of DRMM therapy to death or censoring, using a microcosting approach. Drug costs were separately calculated for each successive DRMM therapy. . BresMed Health Solutions provided support in the form of salaries for authors AB, JE and DL, but did not have any additional role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. The specific roles of these authors are articulated in the 'author contributions' section.
Competing Interests: AB, JE and DL are employees of BresMed, which receives consultancy fees from Celgene. SG, IL, MS, PR, JW, ED, KR are members of a hospital department that has received an unrestricted educational grant from Celgene. KR is in receipt of honoraria from Celgene. No author has any direct financial interests whose value could be affected by this publication. This does not alter the authors' adherence to PLOS ONE policies on sharing data and materials.

Results
When offered DRMM therapy, 67% of patients preferred active treatment to palliative care ( Table 1). The median age of those who chose palliative care was 73.2 years compared to 59.9 for active therapy (p<0.001). The first DRMM treatment regimen was typically bendamustine, thalidomide and dexamethasone (43.6%) as published previously [12]. Retreatment with bortezomib (15.4%) or lenalidomide (33.3%) based regimens was used if poor bone marrow reserve precluded bendamustine use and suitable clinical trial alternatives were lacking. Regimen choice was made by the treating clinician and based on ECOG PS and response/ side effect profile of prior therapies. Third-generation IMiDs and second-generation proteasome inhibitors were not routinely available to this cohort, excepting pomalidomide in one patient. Two patients who had not previously received bortezomib due to neuropathy did receive it at DRMM with no recorded worsening of neuropathy; one had progressive disease despite its use and the other had a partial response of 6 months duration.
Median PFS was 5.2 months and median OS was 5.6 months from start of DRMM therapy (Fig 1). These statistics reflect a steep drop in survival early on with a few patients surviving significantly longer. 24/39 (61.5%) patients had died by the end of follow-up, of whom 79% died in hospital/ hospice. The cohort was deemed too small and heterogeneous for analysis of regimen effect on survival.
MRU was high in this cohort in comparison with that usually observed at earlier lines of therapy (Table 2). 60 inpatient hospital admissions occurred during DRMM therapy in 39 patients. Admissions lasted 9.3 days on average, for indications including pain management, renal failure and most commonly neutropenic fever, with some cases of culture-confirmed septicaemia. Prolonged inpatient admissions, frequent AEs and high transfusion requirements reflect a poor quality of life (QoL) among these patients ( Table 3). The most common recorded Grade 3-4 adverse events (AEs) were anaemia (43.6%), thrombocytopenia (28.2%) and bone pain (33.3%) ( Table 3, Table B in S1 File).

Conclusion
Although the small sample size of this cohort limits the ability to draw definitive survival conclusions, PFS and OS were poor, with a wide range due to the inclusion of all patients whether treated with active or palliative intent, but were similar to published examples [4]. Despite the poor outcomes, up to two thirds of DRMM patients want to be treated with active intent to improve survival, reiterating the need to develop therapies that give patients an improved prognosis, whilst being cost-effective and well tolerated in a heavily pre-treated patient group. The heterogeneity of this cohort is acknowledged but intentional, representing a typical real world hospital cohort of patients, where a range of therapeutic options must be employed, constrained by varying patient-related factors such as ECOG PS, drug tolerance, social situation and patient choice. This is the first report of non-clinical trial based 'real-world' MRU cost analysis in the setting of DRMM. These patients have high MRU costs. A comparable cost analysis report relates to a subset of 54 '4 th line' real world relapsed/refractory patients in a Netherlands study, recruited from a previous trial cohort [7]. However, it is likely that this cohort (data collected  Compared with a decade ago, the price range of new anticancer agents has more than doubled [14], and the cost of care analysis of these agents is imperative. New MM therapies carfilzomib and pomalidomide have been priced significantly higher than currently available antimyeloma drugs. Any subsequent cost benefit analysis comparisons performed in DRMM patients must be set in the context of the high background MRU as observed in our cohort. It is highly relevant that the cost per assumed QALY in this cohort is double that usually accepted by the UK National Institute for Health and Care Excellence [15]. In addition to improving survival, therapies that induce higher response rates or arrest disease progression could potentially increase therapy costs, but lower MRU costs and improve QoL if progression is halted. Biomarkers that focus use of new drugs to cohorts of patients where maximum benefit is obtained would improve cost-effectiveness further. New treatments should be compared with real-world non-trial outcomes such as that provided here, to give a realistic picture of the value of these new drugs.
Supporting Information S1 File.