What is the Helicobacter pylori global reinfection rate ?

This article was originally presented at a conference entitled "Helicobacter pylori: Basic Mechanisms to Clinical Cure 2002", sponsored by Axcan Pharma, November 10-14, 2002, Maui, Hawaii Stanford University School of Medicine, Stanford, California, USA Correspondence: Dr Julie Parsonnet, Stanford University School of Medicine, Grant Building, Room S156, Stanford University, Stanford, California 94305-5107, USA. Telephone 650-725-4561, fax 650-498-7011, e-mail parsonnt@stanford.edu J Parsonnet. What is the Helicobacter pylori global reinfection rate? Can J Gastroenterol 2003;17(Suppl B):46B-48B.

S ince the initiation of Helicobacter pylori treatment in the 1980s, concerns have been raised that high reinfection rates might mitigate against H pylori treatment.Reinfection has been of particular concern where primary infection rates are also high (ie, in developing countries).In Bangladesh and Brazil, where studies have indicated high annual reinfection rates, investigators have expressed pessimism about long-term treatment benefits (1,2).Even in developed countries such as the United States, reinfection rate has been a concern.Because infection with multiple strains is relatively common, the likelihood of reinfection following treatment is particularly plausible (3).If those who are actively infected can be reinfected, what is to prevent such a phenomenon in recently treated patients?Knowing the reinfection rate provides important data on the natural acquisition of host immunity, providing insights into vaccine development and other immunological therapies.Moreover, from a pathophysiological standpoint, distinguishing reinfection from recrudescence of unsuccessfully treated infection permits us to better delineate the most effective antimicrobial strategies.
The primary incidence rate for H pylori is dependent on four factors: the intrinsic transmissibility of the agent, barriers imposed on transmission (these first two factors combine to define the 'force of transmission'), the innate immunity of the host, and the prevalence of infection within the population (4).Of these factors, only the last -H pylori prevalence -is known.As predicted, in general, the population prevalence of H pylori infection correlates with incidence.Although few incidence studies of adults have been done in high prevalence areas, with rare exception (Table 1), in low prevalence populations, H pylori incidence in adults is uniformly low (typically below 0.5% per person year).The best links of prevalence and incidence can be observed in children (Table 1).Where population prevalence of H pylori is relatively low, incidence in children is also low.Where population prevalence is high, childhood incidence rates tend to be high.However, there are exceptions to this trend.For example, in Japan, where the prevalence of infection remains high in adults, incident infection in childhood and adulthood is rare (5,6).Thus, prevalence does not necessarily drive transmission within populations.Rather, because the organism itself is not changing, barriers to transmission -either innate to the host or environmental barriers -must arise that disassociate prevalence and incidence.
Given that barriers to transmission may arise in populations over time (eg, improved environmental and household hygiene) and with increasing age (improved personal hygiene, nutritional status or innate immunity), one would expect that the risk for secondary infection in any individual following H pylori eradication would be less than their primary risk.This does not mean, however, that a treated individual would be less at risk than a person who has never been infected at all.Having had primary infection may indicate continuing host or environmental risk factors that predispose to H pylori acquisition.Hence, in the absence of acquired immunity, reinfection rates might be expected to be higher in previously infected patients than in naïve hosts of similar age and socioeconomic status.On the other hand, if H pylori infection does induce a protective immune response, previously infected patients may be at lower risk than naïve individuals.Thus, understanding the secondary rates of infection, and being able to compare them with contemporaneous primary rates, provides insights into acquired immunity, innate immunity, and the contribution of environmental barriers to H pylori transmission.If the secondary infection rate for a treated population is lower than the primary infection rate expected if the population had been H pylori naïve, then there is evidence of acquired immunity suggesting both predisposition to infection and lack of acquired immunity.
Unfortunately, given the importance of the topic, few studies have actually been able to document true reinfection rates.Investigations that have been done fall into two categories: those in which breath test or gastric biopsy indicated reinfection following a period of apparent cure; and those in which genotyping of reinfecting strains was performed.In the last five years, a number of studies of the former type have been performed (Table 2).These studies indicate a considerably higher rate of reinfection than one might expect given the data for primary infection rates in similar areas.Although this finding may lead one to believe that no protective immunity exists and that previously infected patients are at increased risk, there is significant likelihood in this type of study that recrudescence of uncured infection was misclassified as reinfection.The evidence for this type of misclassification comes from several sources.First, studies have documented higher rates of 'reinfection' in patients who receive second-line H pylori eradication regimens (7,8) compared with those who receive first-line therapy.Similarly, individuals with high-negative breath tests (9) or continued inflammation at the end of treatment have been found to be 'reinfected' at a higher rate than those with low-negative breath tests or reduced inflammation, again likely indicating failed primary therapy (8).In a 1997 review, Xia et al (10) confirmed the problems with many reinfection studies.Among the factors that caused the greatest discrepancy among studies were inadequacy of the initial test of cure (insensitive tests of cure lead to high 'reinfection' rates) and inadequacy of primary treatment.Most recently, Jeen et al (11) found that eight of 10 patients who had recurrent H pylori infection six to 24 months after apparently successful treatment had identical strains by restriction-fragment length polymorphism analysis to the pretreatment strain.Thus, there is considerable evidence that the majority of 'reinfection' represents primary treatment failure.If Jeen et al's figure of 80% recrudescence is applied to the figures in Table 2, it might be deduced that the reinfection rate closely approximates the primary rate in Table 1.
A mere handful of studies have evaluated the reinfection rate using strain fingerprinting techniques (Table 3).As a whole, these studies indicate that the reinfection rate is similar to the primary infection rate in similar populations.Unfortunately, however, even reinfection rates using fingerprinting can yield false results.No diagnostic test for H pylori is perfect.Lack of sensitivity of diagnostic tests -which typically have a sensitivity ranging from 85% to 95% -can be misinterpreted as continued treatment success, falsely diminishing the true reinfection rate.Of greater concern, however, are the factors contributing to overestimation of the reinfection rate.For example, coexistence of more than one strain in the stomach can result in regrowth of a strain unidentified at first endoscopy.Some strains, including cag-negative strains and strains carrying resistance to antibiotics, are less likely than others to respond to treatment (12,13).Thus, treatment can successfully eliminate a dominant strain, leaving a less-    dominant strain to regrow.Given the high frequency of multiple infection, this is a plausible hypothesis, although difficult to demonstrate clinically.Another reason for falsely identifying natural reinfection is that a new strain could be introduced by endoscopic means during the study.Such endoscopic transmission has been noted in previous investigations (14).In summary, reinfection with H pylori appears to occur at a similar rate as primary infection in the same population.Reinfection is no higher if family contacts are infected with H pylori than if they are not (15,16), and reinfection is not consistently higher in areas of high H pylori prevalence than in areas of low prevalence.Thus, reinfection of adults is unusual, even when there is considerable contact with infected hosts.Although several rigorously conducted studies with strain fingerprinting indicate that there may be acquired immunity to infection (reinfection may be somewhat less common than primary infection), the data are not consistent enough to be termed conclusive.Also, despite the relative consistency of fingerprinting studies showing low reinfection rates, one outlier indicates that reinfection can be considerable.In particular, adults in Bangladesh had a high rate of H pylori reinfection confirmed by molecular typing (2).Because these results do not conform with other studies -even studies from developing countries -it will be important to dissect the underlying meaning.Is there something in the environment in Bangladesh that makes reinfection particularly likely?Are Bangladeshis innately more susceptible to infection?Was endoscopic transmission of infection occurring?Identifying reasons for this particularly interesting result may be the key to unlocking the true nature of H pylori transmission and reinfection.

Global reinfection rate of Helicobacter pylori Can
J Gastroenterol Vol 17 Suppl B June 2003 47B

TABLE 3 Reinfection rate in recent studies with confirmatory strain fingerprinting
*Reinfection confirmed by fingerprinting was 0.7%