Herpes zoster in Germany: Quantifying the burden of disease

Infection with the varicella-zoster-virus (VZV) usually occurs in childhood and causes chickenpox [1]. The virus persists lifelong in the dorsal roots of the cranial and spinal ganglia of humans. Especially in the elderly, the virus can reactivate as herpes zoster (HZ) due to decreasing VZV-specific T-cell-immunity [2, 3]. Besides a 20-30% risk of developing HZ over lifetime, the risk of HZ increases distinctly from the fifth decade in life [1, 4‐6]. The main complication of HZ is postherpetic neuralgia (PHN), a long-lasting and occasionally recurring pain [3, 6, 7]. Both HZ and PHN cause limitation of the quality of life of affected individuals [8‐11]. Thitherto the current treatment options of HZ and the prevention of PHN are not always optimal [12‐14].

One vaccine for the protection against the manifestation of HZ and PHN is currently in late-stage development; another HZ vaccine has recently been licensed in the US and in Europe [15, 16]. The licensed vaccine has demonstrated its efficacy in preventing HZ and PHN in a randomized double-blind, placebo-controlled trail involving almost 40,000 individuals 60 years of age and older. HZ and PHN-incidences were reduced in this study by approximately 51% and 67%, respectively [17]. As of yet the duration of protection was demonstrated to last at least 7 years [18]. Furthermore, the vaccine has demonstrated an acceptable safety profile in individuals 50 years and above and was as immunogenic as in people 60 years and older [19]. Based on these findings, the European Medicine Agency approved the first HZ vaccine in 2006 for immune-competent individuals sixty years of age and older, which was later changed to 50 years of age and older [16].

Germany does not have a national surveillance system in place for HZ, and therefore country-level data on the epidemiology of HZ and PHN are scarce. To support the decision making process related to a potential HZ-vaccination recommendation in the elderly we aimed to estimate annual HZ and PHN disease burden in people aged ≥ 50 years in Germany by utilizing various data sources. Since routine childhood varicella vaccination has been carried out in Germany since 2004, baseline HZ disease burden data are also urgently needed to monitor the impact of varicella vaccination on HZ epidemiology [20, 21].

Our study provides the first nationwide estimates of annual disease burden caused by HZ and PHN in Germany. We derived these estimates from a comprehensive physician database that covers approximately 60% of the total German population 50 years of age and older. With an annual incidence of almost 1% in persons aged ≥ 50 years, HZ poses a considerable burden to the health care system in Germany both in terms of outpatient and inpatient services. The increase in disease incidence and HZ-associated mortality with age reconfirms the remarkable affection of the elderly. And due to the demographic aging in Germany it is likely that HZ disease burden will further increase in the future.

The magnitude of incidence observed in our study is consistent with previously published reports from Germany and other European countries. In 1992/93, HZ-incidence in the district of Ansbach, Germany, was found to be 3.41 per 1,000 PY in people ≥ 50 years, and a study conducted in the German federal state of Hesse revealed an incidence of 9.40 per 1,000 PY in 2004 [28, 32]. Studies from other European countries demonstrated incidences ranging from 5.23 to 8.93 per 1,000 in the same age-group [25, 26, 31, 33, 34]. In Australia, an incidence range of 9,70 to 10.10 HZ-cases per 1,000 PY in people 50 years and over was reported [35]. A randomised clinical trial, conducted 1998-2001 in the US to assess the efficacy of a HZ candidate vaccine, revealed an incidence of 11.12 per 1,000 PY but including only individuals aged ≥ 60 years [17]. A systematic literature review concerning HZ incidence in people 60 years and over identified an incidence range of 3.6 to 14.2 per 1,000 PY [36]. Even though the incidences observed in different study settings were in a rather similar magnitude, the burden posed to the healthcare systems in the individual countries might vary considerably due to difference in the systems.

The remarkable increase in HZ-incidence by age can be explained with waning VZV-immunity [2, 3]. The observed difference in sex-distribution remains, however, unclear. Several other studies reported in accordance to our assessment a higher incidence in females, e.g. in France, Germany, Italy, and the US [17, 25, 26, 28, 34, 37, 38]. It has been suggested, that there might be a difference in the immune response to latent viral infection among women and men, but the exact mechanisms are unknown [38].

An average annual 66 deaths due to HZ was reported within the FHM database we utilized, resulting in a mortality of 0.02 deaths per 100,000 PY (age-group 50-54) to 3.86 deaths per 100,000 PY (in 90 years and over), which is lower than estimated in France, where mortality ranges from 0.072 (age-group 55-64 years) to 19.48 deaths per 100,000 PY in people over the age of 94 [26]. Differences in the reporting system might be a reason for this discrepancy.

Our assessment of QALYs is limited by the fact that HZ-related QoL estimates for Germany were not available. However, the countries from where the QoL estimates were derived seem fairly similar to Germany as far as demographics and health care standard are concerned [24, 29, 31]. Values concerning QALY loss may differ intensely as they depend on epidemiology, population, evaluation technique, and the general setting in which those values were surveyed [39]. We accounted for the resulting uncertainty by calculating a range consisting of lower and upper limit estimates. Annual QALY losses caused by other major diseases in Germany have been published previously. For example, osteoporosis-attributable hip fractures were estimated to cause more than 66,000 QALYs lost in Germany in 2002, and obesity almost 370,000 QALYs [40, 41]. Due to the fact that high disease incidence leading to considerable QALY loss, HZ can be regarded - even without considering QALY loss caused by PHN - as a relevant public health problem in Germany.

The use of the ASHIP and FHM databases may, however, cause some limitations. None of the two databases provided information on PHN, which is the most important complication of HZ and therefore crucial for estimating HZ disease burden [6, 31]. By utilizing published literature we compensated this data gap. The true PHN₃-incidence for the population in Germany is likely to be within the range identified by using the results of the published studies. Since ASHIP-data were also available for 2006 and patients can be tracked by unique identifier, we were able to identify and include only acute HZ cases in 2007 to prevent overestimation of HZ-incidence. However, for 28% of HZ-cases in 2008 the classification as an acute case was not possible as we were not able to track them in 2007 due to the different regional datasets being available. This could cause a slight overestimation of HZ-incidence in 2008. Finally, the inpatient incidence might be slightly overestimated when using the FHM database, since we were not able to exclude non-acute HZ cases. An underestimation might have occurred on the other hand, since inpatients with HZ as a secondary diagnosis were not included in our analysis, but still can contribute to prolonged hospitalizations and additional costs. However, the strength of our study lies within the data sources we used. The ASHIP database provides high-quality outpatient information from a considerable proportion of the German population and can be utilized in a cohort-design fashion; the FHM-System provides data from a number of databases in the German health care system.

Germany is one of the few countries worldwide where routine childhood varicella vaccination is recommended, with its initiation in 2004 and a change from a recommended one-dose to a two-dose vaccination schedule in 2009 [42]. Mathematical modelling suggests that routine childhood varicella vaccination will reduce VZV-circulation in nature, which will decrease the number of natural boosters of VZV-immunity in the population. According to mathematical models, this might lead to a substantial increase in HZ-incidence in the population for 40 to 50 years after the initiation of routine varicella vaccination [43, 44]. Since mathematical models are based on many assumptions, it is important to closely observe whether this phenomenon will occur in nature. Since 2005 a physician sentinel-system is in place in Germany to monitor the effects of the current varicella vaccination recommendation [42]. As of today no clear increase in reported HZ-cases per doctors' practices has been observed. But due to the lack of a proper denominator, HZ incidences can not be determined [42]. In the United States, where a varicella vaccination program is in place since 1995, various studies have as of yet also not shown a clear effect of the program on HZ-incidence, even though the incidence of varicella declined rapidly [21, 45].

To monitor potential changes in HZ-epidemiology, countries with routine varicella vaccination should assess HZ-incidence on a periodical basis. HZ incidence and disease burden data presented in this manuscript will serve as important baseline data, and subsequent follow-up incidence data will not only be of critical importance to the German health system, but also for other countries who have recently introduced or are currently considering the introduction of routine childhood varicella vaccination.

Our study confirms the considerable disease burden that HZ and PHN place on the German health care system. With an aging population it is likely that HZ disease burden will continue to increase in Germany, and the implemented routine childhood varicella vaccination program might contribute to an increase in disease burden in the future as well. Our study provides important baseline estimates of HZ-incidence in different age-groups in Germany, which will be important also for the development of a HZ vaccination recommendation. Follow-up periodical analyses of ASHIP-data will be performed to monitor the impact of varicella and herpes zoster vaccination in Germany.