Response to Igarashi, et al, cost-effectiveness analysis for PCV13 in adults 60 years and over with underlying medical conditions which put them at an elevated risk of pneumococcal disease in Japan

KEYWORDS:

• Pneumococcal
• vaccine
• immunity

To the Editorial Board,

We read with interest your recently published article, Cost-effectiveness analysis for PCV13 in adults 60 years and over with underlying medical conditions which put them at an elevated risk of pneumococcal disease in Japan, by Igarashi, et al [1]. For the selection of currently available vaccines for use in adult national immunization programs, cost-effectiveness is an important consideration. Under the base-case scenario utilized in this analysis, results demonstrated that PCV13 is a more cost-effective alternative to PPSV23 in this study population. However, we would like to raise the following questions regarding model assumptions that could significantly impact these conclusions.

Firstly, the paper utilizes an assumption for duration of immunity for PPSV23 that is not supported by available evidence and far shorter than assumptions utilized in other models that assume a linear decline in vaccine effectiveness to 0% in 10 years, in some cases 15 years [2]. This is supported by evidence demonstrating that PPSV23 offers clinical protection for approximately 10 to 15 years in most healthy adults and may decline with increasing age and time [3–5]. The authors cite Suzuki et al [6] for their waning immunity assumption. However, duration of protection was not a primary outcome of this study; time since vaccination was considered in stratified analyses, but results were only presented in the appendix and noted to be imprecise due to wide confidence intervals. Additionally, it should be noted that the model did not evaluate different assumptions for waning immunity in sensitivity analyses.

Secondly, the paper uses an estimate of VE for IPD of only 39%. This estimate is based upon an analysis led by the National Institute of Infectious Diseases that is included in a referenced fact sheet [7], but it is lower than VE estimates from currently available evidence. The VE assumption is also inconsistent with estimates used in other models. Recent reviews demonstrate that PPSV23 has 73% efficacy against all serotype Invasive Pneumococcal Disease (IPD) and up to 77% effectiveness against vaccine-type (VT)-IPD in healthy older adults [8–11]. It should also be noted that the CAPITA trial referenced for PCV13 effectiveness includes a study population whose age distribution is younger than the study population from which PPSV23 effectiveness estimates are derived [12]. Additionally, the WHO suggests that in the setting of a mature PCV program, as is the case in Japan, the impact of PPSV23 will depend upon VE against PPSV23 unique serotypes [13]. There are several observational studies that have evaluated that specific outcome including Kim et al. demonstrating a VE of 78% for PPSV23 unique STs in adults 65 years of age and older [14]. Gutierrez-Rodriguez et al and Su et al also reported similar results for PPSV23 unique serotypes (VE 64.4% and 72%, respectively) [15,16]. The inclusion of a VE of only 39% therefore likely underestimates the potential impact of PPSV23.

When considering the impact that cost-effectiveness models can have on policy decisions and immunization recommendations, it is critical that such models include assumptions supported by a complete assessment of current evidence to provide reliable results. We would therefore recommend caution in the interpretation of this analysis. PPSV23 is the foundation of adult pneumococcal immunization programs worldwide due to its demonstrated cost-effectiveness [17] and broad serotype coverage.

Declaration of interest

K Feemster and K Johnson are employees of Merck Sharp & Dohme Corp., a subsidiary of Merck & Co., Inc., Kenilworth, NJ, USA, who may own stock and/or hold stock options in Merck & Co., Inc., Kenilworth, NJ, USA. Y Kim, M Abe, S Sasaki are employees of Merck Sharp & Dohme Corp., KK. Japan. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript.

Acknowledgments

We would like to thank Karyn Davis, BS of Merck Sharp & Dohme Corp., a subsidiary of Merck & Co., Inc., Kenilworth, NJ, USA for editorial assistance.

Additional information

Funding

This paper was funded by Merck Sharp & Dohme Corp., a subsidiary of Merck & Co., Inc., Kenilworth, NJ, USA.