Serum Free Light Chain Assays in the Early Detection of Cast Nephropathy in Patients Presenting with Unexplained Acute Kidney Injury (AKI): Results from a Pharmacoeconomic Evaluation

International Kidney and Monoclonal research Group (IKMG) recommendations for screening for acute kidney injury (AKI) secondary to multiple myeloma (MM) include the assessment of patients with the serum free light chain test (sFLC). Here we compare the economic impact of following this recommendation compared to standard serum and urine electrophoretic techniques.

An economic model was constructed using published UK data and centre specific clinician advice. The model compared the following scenarios: (i) serum protein electrophoresis (SPEP) + sFLC (IKMG recommendation), (ii) SPEP alone, (iii) SPEP + urine electrophoresis (SPEP + UPE), with a positive test in scenarios (i)—(iii) proceeding to immunofixation electrophoresis (IFE); (iv) all electrophoretic methods run in parallel (SPEP + UPE + sIFE + uIFE). The key economic drivers in the model were length of in-patient stay and dialysis costs. The incremental costs per QALY gained were modelled over a 365d time horizon.

The economic model utilized a decision tree structure to the time of diagnosis and a Markov model including the health states: dialysis dependence, renal function recovery, other cause of AKI and death. The model structure and assumptions were validated through clinician consultation.

Survival in the diagnostic pathway was estimated using hazard ratios applied to survival of MM based on the stage of AKI, obtained from a retrospective observational study (Han, et al., 2013). Survival within the Markov model post-diagnosis was based on UK data stratified for cast nephropathy patients who had or had not recovered renal function. Post-diagnosis, a probability of renal recovery based on the duration of AKI prior to diagnosis was applied.

Utility values in the decision tree structure were estimated using the Modification of Diet in Renal Disease equation. The Markov model employed previously published utility values in MM with a decrement applied for dialysis dependence.

Costs for medical resource use, medical management, dialysis, adverse events associated with dialysis and terminal care were derived from the 2012-2013 NHS Reference Cost and British National Formulary.

Key user defined values for the comparison included a presenting patient distribution of 10% stage 2 AKI, 45% stage 3 AKI (dialysis independent) and 45% stage 3 AKI (dialysis dependent), 1% incidence of cast nephropathy as a cause of AKI, and a 20% incidence of renal biopsy following a positive diagnostic result.

User amendable fields ensure the model is amenable to refinement as new data on assay performance and patient pathways become available.

The incremental costs and QALYs for each comparator are shown in Table 1. The model predicts the pathway with SPEP + sFLC to be dominant; SPEP + sFLC accrues the most QALYs and the least costs. The savings in the diagnostic decision tree (£7, £31, -£5, respectively) are driven by reduced dialysis and in-patient bed stay costs. In the treatment pathway savings (£34, £37, £41, respectively) are driven by reduced dialysis costs. Probabilistic sensitivity analysis showed the model results to be robust - the probability of SPEP + sFLC being cost-effective at a £20,000 per QALY willingness to pay threshold was 98.6%, 93.8%, 92.6%, respectively.

This model supports the recommended international guidelines for screening for AKI secondary to MM. Inclusion of the sFLC assay improves the probability of renal recovery and survival by reducing time to diagnosis and treatment.

Cost savings and QALY gains are found in both the diagnostic stage and treatment stage of the patient pathway. The SPEP + sFLC pathway is therefore cost-effective at a willingness to pay threshold of £20,000 per QALY with a high probability of cost-effectiveness against all comparator pathways from probabilistic sensitivity analysis.

The model template may be used in future health economic models to assess the contribution of new assays into diagnostic algorithms.