Early economic evaluation of chelation therapy in kidney transplant recipients with high-normal lead

Abstract

Background: Kidney transplant recipients (KTR) with high-normal lead have a higher risk of graft failure (GF). Clinically, chelation therapy using meso-2,3-dimercaptosuccinic acid (DMSA) removes lead. Despite the proposal that chelation therapy can prevent GF through lead removal, evidence is lacking. To guide research efforts, we conducted an early economic evaluation, aiming to explore the economic feasibility of screening for and implementing chelation therapy with oral DMSA for high-normal plasma lead concentrations in KTR (i.e., the intervention) compared to standard of care.

Methods: A Markov model simulated the life course of 10,000 KTR in the Netherlands from a societal perspective. Transition probabilities were estimated using the data from TransplantLines Food and Nutrition Biobank and Cohort study. Costs and utilities were sourced from publications and public data. Model robustness was investigated through deterministic and probabilistic sensitivity analyses. Various administration strategies were tested. Five-year costs were calculated from a healthcare payer's perspective. Value of information was assessed.

Results: The intervention was cost-saving and improved health, leading to a dominant incremental cost-effectiveness ratio. The result was most sensitive to transition probabilities (led by GF, followed by death with functioning graft and after graft failure). The probability of the intervention being cost-effective was 60%. Chelation strategies did not affect the result. The intervention applied to the Dutch KTR population could save €27 million in the initial five years. Further research is desirable if the cost of obtaining perfect information on GF survival is approximately below €4,000/KTR (all uncertainties under €5,000/KTR).

Conclusion: The cost-effectiveness of the intervention is robust in KTR, except when considering the uncertainties around (graft) survival probabilities. Applying chelation therapy in the new setting we studied holds significant potential. However, trials that systematically assess the efficacy, administration strategies, and impacts on survival are crucial in updating the current evaluation and informing policies.

Fig 1. Markov model for progression of disease.

The dashed arrow indicates primary non-function. From each state, it is possible to transition to the death state. FG1, functioning graft 1; FG2, functioning graft 2; GF1, graft failure 1; GF2, graft failure 2.

Fig 2. The timeline and procedure of screening and implementing chelation therapy in base case.

C, Course; C1, Course 1 (i.e., one course); DMSA, meso-2,3-dimercaptosuccinic acid; BT, blood test; KTR, kidney transplant recipient; Y, year.

Fig 3. Deterministic sensitivity analysis.

The tornado diagram represents the effect of the lower and upper limit of every parameter on (A) incremental costs and (B) incremental effectiveness. The 15 most influential parameters are shown. The red dashed line represents either $\Delta C=0$ or $\Delta E=0$ Meanlog and sdlog are the parameters from lognormal distribution and were estimated through the parametric survival analyses. These parameters are used to calculate the transition probabilities for graft failure in different subgroups of KTR with high, medium and low plasma lead concentrations. $\Delta C$ incremental costs; $\Delta E$ incremental effectiveness; QALY, quality adjusted life year; TP, transition probability; GF, graft failure; DwGF, death with graft failure; DWFG, death with functioning graft; FG, functioning graft; HD, haemodialysis; PD, peritoneal dialysis; CHD, center haemodialysis; CAPD, continuous ambulatory peritoneal dialysis; KTx, kidney transplantation; DMSA, meso-2,3-dimercaptosuccinic acid.

Fig 4. Probabilistic sensitivity analysis.

The cost-effectiveness planes show the PSA results of (A) varying all parameters and (B) varying all parameters except for the estimated parameters which calculated transition probabilities. (C) The resultant probabilities of the intervention being cost-effective at differing willingness to pay thresholds are plotted. KTR, kidney transplant recipients; $\Delta C$ incremental costs; $\Delta E$ incremental effectiveness; PSA, probabilistic sensitivity analysis; TP, transition probability, specifically with the endpoints of graft failure, death with functioning graft and death with graft failure; BC, base case; WTP, willingness to pay; QALY, quality adjusted life year.

Fig 5. Visualization of the base case and scenario results.

$\Delta E$ incremental effectiveness; QALY, quality adjusted life year; $\Delta C$ incremental costs; BC, base case.

Fig 6. EVPI and EVPPI by willingness to pay.

EVPI, expected value of perfect information; EVPPI, expected value of partial perfect information; PSA, probabilistic sensitivity analysis; TP, transition probability; GF, graft failure; DwGF, death with graft failure; DWFG, death with functioning graft; KTR, kidney transplant recipient; QALY, quality adjusted life year.