Abstract
Of cystic fibrosis (CF) patients referred for lung transplant consideration, those who are cytomegalovirus (CMV) IgG+ at referral are 8 years younger than seronegatives, suggesting CMV may have a pathogenic role in CF lung disease http://ow.ly/iSQf30nAxxl
To the Editor:
Cytomegalovirus (CMV) is a betaherpesvirus, the impacts of which are well known to clinicians providing post-transplant cystic fibrosis care. Lung transplant recipients have the highest risk of any solid-organ transplant for CMV reactivation and ganciclovir resistance [1, 2]. Furthermore, CMV reactivation increases the risk of chronic lung allograft dysfunction. However, even in general populations, CMV seropositivity is associated with adverse outcomes including cognitive impairment, frailty, heart disease and all-cause mortality [3–5]. How CMV may contribute to disease is not evident but many streams of evidence suggest CMV replication in inflamed sites contributes to exaggerated inflammation and tissue injury [6]. Individuals with cystic fibrosis experience chronic inflammation within the airways leading to remodelling and eventually respiratory failure. Indeed, inflammatory biomarkers in the sputum and serum of cystic fibrosis patients correlate with short- and long-term outcomes [7]. We hypothesised that CMV may represent an unrecognised contributor to cystic fibrosis lung disease.
We performed a detailed chart review of all Calgary Adult Cystic Fibrosis Clinic patients who were referred for lung transplantation, where CMV IgG testing would be performed. CMV serostatus, demographics, infecting pathogens, markers of nutrition and lung function from last clinical encounter were recorded for those who were transplanted or succumbed to disease. Our primary outcome was a composite end-point of age at lung transplantation/death without transplant. We also analysed the outcomes of death and lung transplant separately by CMV serostatus, and conducted a stratified analysis per time period to account for improvements in care (1990–1999, 2000–2009 and 2010–2017). Ethical approval for this study was supplied from the Conjoint Health Region Ethics Board of the University of Calgary (REB15-2744).
Sociodemographic and clinical characteristics, including referral and listing for transplantation, were summarised. Univariate and multivariable linear regression models were constructed for the primary outcome to compare CMV+ and CMV− patients. We incorporated clinical (disease severity and microbiology) factors that varied significantly between patients by CMV status as well as sex, body mass index (BMI) and educational status (less than high school, high school, technical college or university) (which trended toward significance and were biologically plausible). Nonparametric tests of trend were conducted to compare outcomes between periods. Significance was based on α<0.05, all hypothesis tests were two-sided and statistical analysis was performed using Stata version 14.2 (StataCorp, College Station, TX, USA).
Since 1991, 71 patients (50.7% female) were referred and listed for transplant consideration who either successfully received a lifesaving transplantation or died waiting. Of these, 59 (83%) received bilateral lung transplants and 12 (16.9%) died prior to transplantation. Of the cohort, 15 patients were excluded (six died and nine transplanted) from the analysis as they did not have CMV serology documented. Patients who were excluded were demographically similar but were more likely to represent earlier cohorts (44% in 1990–1999, 14% in 2000–2009 and 0% in 2010–2017; p=0.004).
Of the 56 included, 30 (54.6%) were CMV+ and the prevalence of CMV seropositivity did not differ by time period. When sociodemographic and clinical characteristics were examined, patients who were CMV− were more likely to be F508del homozygous and have Burkholderia cepacia complex infection (all Burkholderia multivorans) but otherwise did not differ from CMV+ individuals (table 1). Lung transplantation occurred a median of 1.25 years (interquartile range (IQR) 0.87–1.78 years) after referral and 0.44 years (IQR 0.14–0.90 years) following listing, and did not differ by CMV status.
For the primary outcome, patients who were CMV+ died or underwent bilateral lung transplantation at a significantly younger age compared to those who were CMV− (27.17 versus 35.11 years; difference 7.95 years (95% CI 3.61–12.29 years), p<0.001). The difference in age at death or lung transplant remained significant in a multivariable model adjusted for sex, BMI, B. cepacia complex infection, genotype and education (6.96 years (95% CI 2.51–11.4 years), p=0.003). When the outcomes of death and transplant were assessed separately, CMV+ was associated with a significantly lower mean age at either death (difference 9.35 years (95% CI 0.89–17.82 years), p=0.03) or lung transplantation (difference 7.36 years (95% CI 2.58–12.13 years), p=0.003). No significant differences in age at death or transplant were noted by CMV serostatus (p>0.05) when outcomes were stratified by time periods.
Our retrospective study demonstrated a significant age disparity at death or lung transplantation in persons with cystic fibrosis, suggesting a deleterious association for CMV. However, we must consider several limitations to these observations. Most importantly, association does not confirm causation [8]. This was at a single-centre study with a limited sample size and limited power to detect potentially meaningful differences in the outcome in the stratified analyses. Although there was no evidence of changing rates of CMV seroprevalence, the study spanned ≥25 years and considerable improvements in cystic fibrosis outcomes have since been realised. However, when we analysed our outcomes stratified by time periods, no significant changes in the association were observed. A portion of patients were excluded from analysis due to missing CMV status leading to potential selection bias but they were demographically similar to the analysed cohort. While we were able to identify CMV serostatus of patients at the time of transplant referral, we were unable to determine when patients were infected and infection duration may have differential effects on clinical outcomes. Within general populations, CMV seropositivity increases over time with rates increasing from 36% in those 6–11 years, 49% in those 20–29 years, 65% at 40–49 years and >91% in those >80 years of age [9]. Accordingly, duration of CMV infection may be a more important predictor of progression to the primary outcome as opposed to serostatus at a single time-point. Furthermore, quantitative levels of CMV IgG may be more sensitive at identifying risk as opposed to current reporting of qualitative results [4]. CMV seropositivity is known to exist disproportionally in socially marginalised groups [10]; socioeconomic status is an important modifier of cystic fibrosis [11], and influences access to lung transplant referral and listing [12]. However, we did not identify significant differences between groups in socioeconomic status using education indicators.
Our knowledge of CMV pathogenesis offers a potential model through which CMV may accelerate disease progression in cystic fibrosis: 1) the lungs are a major site of CMV reactivation [13], and progressive lung disease is responsible for the majority of CF morbidity and mortality; 2) CMV reactivation can be triggered by bacterial infection [14] and bacterial infection is omnipresent in cystic fibrosis; 3) CMV reactivation is associated with reduced immune surveillance [15] and thus, increased susceptibility to infections; and 4) CMV reactivation is associated with an exaggerated immune response [16], potentially enhancing tissue damage. CMV infection could partially explain the varied clinical courses experienced by individuals with cystic fibrosis, and why conventional microbiological markers correlate poorly with pulmonary exacerbation recovery and long-term outcomes [17].
To confirm whether CMV associates with disease progression in cystic fibrosis, we can leverage CMV serostatus data collected as part of transplant databases (e.g. the International Society for Heart and Lung Transplantation Registry) in efforts to understand CMV pathogenesis pre- and post-transplantation, i.e. to confirm the results of this single-centre study. Furthermore, cross-sectional seroprevalence studies of general cystic fibrosis populations captured with national registries will enable us to correlate critical markers of cystic fibrosis disease progression with CMV status such as baseline lung function, changes in lung function and exacerbation frequency. We would advocate for prospective longitudinal studies of cystic fibrosis populations to assess factors associated with incident infections and seroconversion, in order to determine whether acquisition of CMV infection is associated with change in disease status and how infection duration correlates with clinical outcomes. Indeed, a case series of three individuals experiencing exacerbations in the context of CMV seroconversion was recently published [18]. Finally, exploring the role of CMV reactivation/viraemia and its association with exacerbation occurrence and recovery and rates of lung function decline is important, as demonstrated by efforts currently underway to explore CMV pathogenesis in intensive care units [19].
The identification of the association of CMV with cystic fibrosis lung disease progression could have tremendous implications for disease management. If this association is confirmed in larger multicentre studies, strategies exploring the use of pre-emptive and prophylactic antiviral treatments, including acyclovir (which while not having significant CMV activity, can reduce reactivation potential) or valganciclvoir, in specific populations may be in order [20]. Considerable progress in developing an effective vaccine for CMV has been made recently and a large number of products are currently being investigated [21]. The rationale for these vaccines has predominately focused on their role in preventing congenital cases of CMV and avoiding transplant complications in pre-transplant patients. However, it may be that individuals with cystic fibrosis who are seronegative for CMV may derive benefit from effective vaccine strategies both acutely and in the long-term setting of their risk for transplant.
Footnotes
Author's contributions: M.D. Parkins was primarily responsible for the data collection and record maintenance. Statistical analyses were performed by R. Somayaji and M.D. Parkins. M.D. Parkins was responsible for the creation of the initial draft of the manuscript. All authors contributed to development of the final manuscript. M.D. Parkins is the guarantor of this work.
Conflict of interest: M.D. Parkins reports grants from CF Canada, CIHR Canada, Gilead Sciences and the Alberta and NWT Lung Association, outside the submitted work.
Conflict of interest: K.J. Ramos reports grants from the National Institutes of Health, Cystic Fibrosis Foundation and CHEST Foundation, outside the submitted work.
Conflict of interest: C.H. Goss reports grants from the Cystic Fibrosis Foundation, European Commission and NIH (NHLBI, NIDDK and NCRR), during the conduct of the study; and personal fees from Gilead Sciences and Novartis, grants from the NIH and FDA, and serving as the US lead in a phase 2 trial of a novel therapy for cystic fibrosis for Boehringer Ingelheim, outside the submitted work.
Conflict of interest: R. Somayaji reports grants from CF Canada, CIHR Canada and Alberta Innovates, outside the submitted work.
- Received September 14, 2018.
- Accepted January 10, 2019.
- Copyright ©ERS 2019