Does the inclusion of the cost and burden of adverse drug reactions associated with drug-resistant TB treatment affect the incremental cost-effectiveness of new treatment regimens?

ABSTRACT

Does the inclusion of the cost and burden of adverse drug reactions associated with drug-resistant

TB treatment affect the incremental cost-effectiveness of new treatment regimens? A case study

from the introduction of bedaquiline in South Africa National TB Programme

Kathryn Lou Bistline

South Africa has one of the world’s highest burdens of TB, HIV/TB co-infection, and drug-resistant TB. Second-

line TB treatment is less effective, more expensive, and more toxic than treatment for drug-sensitive TB.

Nearly 1 in every 5 persons who starts treatment for drug-resistant TB in South Africa will die; 1 in every 3

persons who survives treatments experiences permanent, profound hearing loss. For decades there was little

progress in TB research, however, and so treatment with old regimens continued despite safety concerns.

In 2012 the US and European regulatory authorities approved a new drug, bedaquiline, but only for treatment

in cases with no other options. In 2015, the South African Medicines Control Council approved bedaquiline

for drug-resistant TB, but only a limited number of doses were approved in the 2016/2017 annual

budget and the focus, again, was only for the patients who had no other options. In order to inform policy

makers in planning and budgeting for drug-resistant TB treatment, the aim of this thesis was to determine

whether the simple calculation that bedaquiline was too expensive relative to standard regimens using kanamycin

was too simple. Particularly, given the high burden of adverse drug reactions (ADR) associated with

kanamycin, would the inclusion of the cost and burden of ADR affect the incremental cost effectiveness ratio

of a new treatment regimen where bedaquiline replaces kanamycin?

Analysis of the national drug-resistant TB case register showed that mortality during second-line treatment

was early, primarily in the first 6 months of treatment, even when patients do not have extensive drug resistance.

HIV-positive patients not on anti-retroviral therapy (ART) at initiation of drug-resistant TB treatment

have the highest risk of mortality. The high early mortality is a real risk that clinicians have to balance

when deciding to initiate ART and effective second-line TB treatment both as quickly as possible. Daily injections

coupled with taking more than 10 pills each day are a heavy burden for patient compliance, but also

pose concerns in terms of overlapping and compounding toxicities; this burden was confirmed through a

meta-analysis of the pooled frequency of adverse events among cohorts with at least 25% of the patients

HIV-positive. A competing risk analysis of a cohort of drug-resistant TB patients from Johannesburg – addressing

the reality that patients may not have experienced an ADR because they died rather than because

they were at lower risk – indicated that HIV-infected patients who are not yet stable on ART and second-line

TB treatment are at the highest risk of ADR.

A Markov model built and parameterized using the data from the South African national TB programme indicates

that bedaquiline for all drug-resistant TB led to a small gain in effectiveness at a cost that was under

the costs of the drug itself, due to savings from daily injection visits. While cost-effective, it was not clear

that South African policy makers needed to move beyond the offer of bedaquiline for patients with extensive

drug resistance. However, the calculation, and the decision point, were different once the costs and disability

associated with ADRs was included in the analysis. Bedaquiline-based regimens offer a cost-saving and

more effective alternative to an injection-based regimen for drug-resistant TB patients treated in the public

sector in South Africa.