Clinical, serological and epidemiological features of hepatitis A in León, Nicaragua

Introduction

The hepatitis A virus (HAV) is mainly transmitted by the fecal-oral route (World Health Organization, 2010a). Therefore, the burden of disease correlates with the quality of sanitary conditions, the access to clean water and the socioeconomic status (Jacobsen & Koopman, 2005). After an incubation period of 14 to 50 days the disease might start with unspecific symptoms such as malaise, fatigue, anorexia, vomiting, abdominal discomfort and diarrhea. In the course of symptomatic illness, which ensues more often in older children and adults, dark urine, pale stool and jaundice can also be observed; in children below 6 years of age less than 10% show jaundice (Armstrong & Bell, 2002). Rarely, fulminant hepatic failure develops, associated with a severe disease process and a high case fatality, usually necessitating liver transplantation (Lemon et al., 2018). Interestingly, HAV has been reported to be a major cause for fulminant hepatic failure in children in Pakistan (Talat et al., 2020) and Latin America (Ciocca et al., 2007).

There are many unreported cases of hepatitis A (up to 80% of infections). WHO estimated from 1990 to 2005 an increase in the number of infections and deaths due to hepatitis A from 177 to 212 million and 30,283 to 35,245, respectively (World Health Organization, 2010a). From the 212 million infections only 33 million cases are estimated to present as symptomatic illness (Rein et al., 2014). Different studies showed over the last two decades, however, decreasing hepatitis A seroprevalences all over the world, except for Africa, most likely due to better access to clean water, better sanitary conditions, higher educational level and less crowded housing conditions (Jacobsen & Wiersma, 2010). Levels of hepatitis A endemicity are defined on the basis of seroprevalence: “high (≥ 90% by age 10 years); intermediate (≥ 50% by age 15 years, with <90% by age 10 years); low (≥ 50% by age 30 years, with <50% by age 15); and very low (<50% by age 30 years)” (Jacobsen & Wiersma, 2010). Because HAV infection leads to lifelong immunity there are in high endemic settings hardly any susceptible adults, and clinically manifest HAV infections are rare. However, if sanitary conditions improve, there will be a shift to intermediate endemicity, and with increasingly older age groups remaining susceptible, more severe cases (World Health Organization, 2010a) and outbreaks (Bell et al., 1998) will be observed. For Nicaragua there exist very few published data on the hepatitis A epidemiology (Andani et al., 2020). Hepatitis A infections are not routinely reported by the health authorities in Nicaragua, a still highly endemic country with 36.9 cases per 100,000 inhabitants reported for 2012 (Supplemental File 4).

Universal mass vaccination (UMV) of toddlers with a two-dose regimen is now recommended by WHO for countries with sizable acute hepatitis A incidences and declining endemicity (World Health Organisation, 2012). Two-dose UMV successfully eliminates hepatitis A within a few years, as shown, e.g., in Israel (Levine et al., 2015). Single-dose UMV likewise lowers rapidly the disease burden, as documented for Argentina (Vizzotti et al., 2014) and Brazil (Souto, de Brito & Fontes, 2019). Although single-dose vaccination elicits in >95% of children seroprotection for at least 7–10 years (Mayorga et al., 2016; Urueña et al., 2016; Espul et al., 2020), the long-term protective effectiveness over 20-30 years remains still to be established.

The objectives of our project were: (1) to document with a prospective survey (2006–2010)—in parallel to an ongoing single-dose vaccine effectiveness study (Mayorga et al., 2016)—the sustained circulation of HAV in León, Nicaragua; (2) to describe the clinical features of acute hepatitis A; and (3) to evaluate the HAV infection risk factors in the same study area, for which already for the 1990s and the 2000s a high endemicity had been reported (Mayorga Perez et al., 2014).

Materials & Methods

Results

Study population

A total of 575 children, adolescents and adults with suspected viral hepatitis were screened from May 2006 until June 2010. Eighteen patients had to be excluded from analysis: samples lost (n = 2), anti-HAV IgM serology for April-June 2010 samples lacking (n = 15) or unclear age (n = 1), thus leaving 557 eligible patients (see Study Flow Chart, Fig. 1). Six patients had “gray-zone” results of the anti-HAV IgM test and were counted as non-hepatitis A. The majority (80.6%) of the 315 patients with acute hepatitis A were children ≤10 years of age, two were ≤1 year and five >18 years old (Table 1). With this cohort of acute hepatitis A cases we documented the continuous exposure to HAV of the subjects enrolled into the single-dose hepatitis A vaccination trial carried out by our team in the same area during 2005–2012 (Mayorga et al., 2016).

Table 1:

Characteristics of the study population.

Characteristic	Category	N (%)		p-value
		Acute Hepatitis A (n = 315)	Non Hepatitis A (n = 242)
Demographics
Age				<0.001
	≤1 year	2 (0.6)	16 (6.6)
	1–≤5 years	70 (22.2)	47 (19.4)
	5–≤10 years	182 (57.8)	64 (26.4)
	10–≤14 years	46 (14.6)	32 (13.2)
	14–≤18 years	10 (3.2)	11 (4.5)
	>18 years	5 (1.6)	72 (29.8)
Sex				1.000
	female	157 (49.8)	120 (49.6)
	male	158 (50.2)	122 (50.4)
Socioeconomic
Housing				0.013
	no crowding	141 (44.8)	135 (55.8)
	crowding	174 (55.2)	107 (44.2)
Toilets				1.000
	outside	157 (49.8)	99 (40.9)
	inside	158 (50.2)	143 (59.1)
Water				0.044
	outside	57 (18.1)	43 (17.8)
	inside	258 (81.9)	199 (82.2)
Clinical features
Jaundice				<0.001
	no	14 (4.4)	48 (19.8)
	yes	301 (95.6)	194 (80.2)
Pale stool				<0.001
	no	207 (65.7)	203 (83.9)
	yes	108 (34.3)	39 (16.1)
Dark urine				<0.001
	no	120 (38.1)	162 (66.9)
	yes	195 (61.9)	80 (33.1)
Nausea				<0.001
	no	83 (26.3)	98 (40.5)
	yes	232 (73.7)	144 (59.5)
Vomiting				0.001
	no	81 (25.7)	127 (52.5)
	yes	234 (74.3)	115 (47.5)
Fever				0.014
	no	64 (20.3)	72 (29.8)
	yes	251 (79.7)	170 (70.2)
Anorexic				0.004
	no	55 (17.5)	68 (28.1)
	yes	260 (82.5)	174 (71.9)
Malaise				0.101
	no	46 (14.6)	49 (20.2)
	yes	269 (85.4)	193 (79.8)
Laboratory finding*
Liver enzymes GOT				<0.001
	elevated	281 (89.2)	108 (44.6)
	normal	34 (10.8)	134 (55.4)
Liver enzymes GTP				<0.001
	elevated	289 (91.7)	93 (38.4)
	normal	26 (8.3)	149 (61.6)
Total serum bilirubin				<0.001
	elevated	274 (87.0)	118 (48.8)
	normal	41 (13.0)	124 (51.2)

DOI: 10.7717/peerj.11516/table-1

Notes:

*cut-off definitions are described in the Methods section.

We did not evaluate occupation and educational levels because the vast majority of study subjects were very young or school age children and thus an analysis of these parameters would not have made sense.

Seasonality

All along the observation period May 2006 to June 2010 cases of suspected viral hepatitis were enrolled (∼11.1/month), with slightly more patients presenting themselves at the health centers in the 1st and 3rd quarters of the year (data not shown). 56.6% of suspected cases were diagnosed with acute hepatitis A (∼6.7/month). On average, more cases of acute hepatitis A were diagnosed in the 4th and 1st quarters (autumn & winter) than in the 2nd and 3rd quarters (Fig. 2). The percentage of hepatitis A cases among all persons examined decreased from 63.5% (n = 54/85) in 2006 to 50.0% (n = 32/64) cases in 2010 (p = 0.013, p for trend =0.0016). The age distribution of patients with acute hepatitis A remained, however, similar over the years (data not shown).

Discussion

Study population: In our study 80.6 of the 315 symptomatic hepatitis A cases enrolled were children aged ≤10 years, an age where HAV infection is in the majority asymptomatic (Armstrong & Bell, 2002). This implies that many more subclinical HAV infections must have occurred in parallel in the population surveyed. Overall, our findings indicate that Nicaragua was in 2006–2010 still a country of high endemicity and that the single-dose hepatitis A vaccination study (Mayorga et al., 2016)—for which the present project acted as a kind of community control group—had been conducted during a sustained circulation of HAV.

Seroepidemiology: Quantitative anti-HAV IgG antibody levels could only be measured in 263 of the 557 patients with suspected hepatitis. In the Non-A subpopulation tested (n = 94), the age-specific seroprotection rates rise until the age of 11- ≤16 years, where 100% seroprotection is reached (Table 2), consistent with the 100% found from age 18 years onwards in our 2003 study (Mayorga Perez et al., 2014). The present anti-HAV antibody data cannot be used as cross-sectional population seroprevalence data, as only symptomatic subjects were screened. However, the data gathered allow the conclusion that hepatitis A was still highly endemic in the area investigated: nine of the 11 Non-A patients lacking seroprotection were ≤11 years of age and only two were 18 and 19 years of age. Nevertheless, the incomplete seroprotection of 71.4% observed in the age group 16 to ≤20 years, and the fact that two of the five Non-A infants were seronegative, i.e., their mothers most likely as well, indicates the beginning of a transition away from high endemicity. The most recent hepatitis A data cover the years 2004–2012 and report for the whole country a rise from 19.2 to 36.9 infections per 100,000 inhabitants, and for León a rise from 12.6 to 22.2 (2011), followed by a sharp increase to 86.9/100.000 in 2012 (Supplemental File 4). Our monthly enrolment of an average 6.7 patients with acute hepatitis A matches well with the average of 6.3 cases reported monthly for León during 2006–2010 (Supplemental File 4). Using our 2003 cross-sectional serosurvey data (Mayorga Perez et al., 2014), a recent literature review on hepatitis A epidemiology in Latin American countries classified the endemicity level of Nicaragua as high-intermediate (Andani et al., 2020), based on the age at midpoint of population immunity (Mohd Hanafiah, Jacobsen & Wiersma, 2011) of 5–9 years. Previous studies in the same area in León showed that there was already between 1995–96 and 2003 a halving of the HAV infection risk for 1 to 5 years old children (Mayorga Perez et al., 2014). Although we found with the limited serosurvey data only few serological indications of shifting endemicity among the non-A cases, i.e., very few HAV susceptible school children or young adults (Table 2), 17.8% of the acute hepatitis A cases occurred between 10 and ≤18 years (n = 56) and 1.6% (n = 5) at >18 years of age. In addition, there were two cases in infants who are in hyperendemic settings usually seroprotected by maternal antibodies (Brinkhof et al., 2013). Taken all together, these signs of a beginning epidemiological shift are a call for the health authorities to start monitoring hepatitis A more closely, e.g., with repeated cross-sectional serosurveys (Cutts & Hanson, 2016; Mayorga Perez et al., 2014), in order to not miss the right time to start with public health interventions, such as e.g., universal childhood vaccination.

Clinical features: The clinical outcome of HAV infection is strongly associated with age: below 6 years of age only 10–20% of infections are symptomatic, at the age of 10 years up to 50% might present jaundice or other clinical features and from age 18-20 years upwards 80–90% of infections are symptomatic (Armstrong & Bell, 2002). The severity of disease and fatal outcomes are higher in older age groups, with case fatality rising from 0.1% in children <15 years, to 0.3% among persons 15–39 years of age and reaching 1.8–5.4% above 50 years of age (Lemon et al., 2018). In our study, no severe cases were seen, the laboratory findings were ‘unspectacular’ and corresponded to the overall mild clinical features of the patients enrolled. None of the children or adolescents had to be hospitalized, which is typical for a country where hepatitis A is highly endemic and most infections occur during childhood (World Health Organization, 2010a). However, there are reports from Latin American countries of severe cases with fulminant hepatitis A and acute liver failure (Ciocca et al., 2007; Ferreira et al., 2008), this being an important reason for implementation of UMV of children (Lemon et al., 2018), as e.g., in Argentina (Vizzotti et al., 2014). It is still debated whether differences between various geographic world regions regarding pathogenicity and virulence of HAV are due to genetic host factors in certain ethnic groups or due to nucleotide sequence variations in certain HAV subtypes (Long et al., 2014; Vaughan et al., 2014).

Socioeconomic factors: All over the world the risk of HAV infection is highly correlated with the socioeconomic status (Jacobsen & Koopman, 2005). Although the global number of HAV infections is still on the rise (World Health Organisation, 2012)), declining risks of infections are seen in many countries over the last two decades, mainly due to rising sanitary and economic standards and better access to clean water (Jacobsen & Koopman, 2005), but in some countries also due to UMV, such as in Israel (Levine et al., 2015), Argentina (Vizzotti et al., 2014) and Brazil (Souto, de Brito & Fontes, 2019). This shift from high to intermediate HAV endemicity has been observed in Latin America since the 1990s, a trend which continued in the 2000s (Jacobsen & Wiersma, 2010; World Health Organization 2010b; Andani et al., 2020). The risk for HAV infection is influenced by many different factors. Logistic regression analysis showed in Brazil that anti-HAV seroprevalence correlates with increasing age and level of crowding (Vitral et al., 2014), whereas access to clean water and reinforced concrete buildings were protective factors (Vitral et al., 2012). For age and crowding we can see similar effects in our study, with age being the most strongly correlated factor. For the water source we saw no significant effect, even though this is mentioned in many studies from Latin America (Jacobsen & Wiersma, 2010), probably due to fact that the water was clean in our study area, even if it came from an outside well. In a Brazilian study, an overall low socioeconomic status was identified as risk factor (Vitral et al., 2012). We could not assess this in our study, because we did not evaluate income and education of the families. Thus, the socioeconomic risk factors for HAV identified in our study were only the sanitary conditions, i.e., the place of the toilet, and crowding. Household contacts are an important source of infection (Lima et al., 2014). In our setting, crowding made a significant difference in crude Odds ratios but not in the adjusted ones. Also, we could not see a difference between households hosting different age groups. Another Brazilian study suggests that crowding is confounded by other socioeconomic factors, possibly due to the difficulty in maintaining hygiene in big households (Almeida et al., 2001). In our study this is shown with the adjusted odds ratio, where the p-value for crowding is no longer significant, indicating that there is a confounding by socioeconomic factors, such as the localization of toilets (Table 3). In the logistic regression analysis we saw a correlation between the different risk factors. Independently seen, none of the factors made, however, a significant difference. Therefore, the main risk factor appears to be the low socioeconomic status with all his features.

Seasonality: In our study there were slightly more hepatitis A cases in the 1st and 4th quarter of the year, i.e., during autumn/winter, when the weather is in Nicaragua slightly cooler and drier. There are seemingly conflicting data published on hepatitis A seasonality, some reports indicating peak incidences in the wet season or summer, as for Brazil (Villar, de Paula & Gaspar, 2002), others in the dry season or winter, as in Israel (Green et al., 2001). Many different environmental factors, such as rainfall, ambient temperature, and air humidity, as well as host behavior, source of infection and mode of transmission are playing a role in the seasonality of infectious diseases (Martinez, 2018). A recent review on the seasonality of viral hepatitis explains these seemingly conflicting seasonality patterns for hepatitis A infections with the different modes of transmission being decisive: water-borne during wet summer seasons in case of low hygiene standards and rather food-born in winter, caused e.g., through shellfish consumption (Fares, 2015). Interestingly, even in present-day Spain water-related climate events correlate with the incidence of hepatitis A (Gullón et al., 2017).

Our study has limitations: The data collection was done via passive surveillance, relying on self-reporting of cases (parents/patients). The real number of acute hepatitis A cases was most likely higher, as we collected the more symptomatic cases, and because of the low manifestation index in young children, many more asymptomatic HAV infections must have occurred during the observation period. We collected educational levels and jobs only for the enrolled subjects themselves, and not for their mothers/parents, which would have made more sense because most of the patients were young children. Anti-HAV IgG antibody data were only available for 47% of all 557 cases of assumed viral hepatitis and in only 94 of the 242 Non-A cases, limiting the significance of the age-specific seroprotection data, although they pretty well match previous data from the same area (Mayorga Perez et al., 2014). The number of 315 hepatitis A cases was too small to assess in the given setting with certainty the influence of all socioeconomic factors on contracting acute hepatitis A.

The strengths of the study were: In a hepatitis A endemic setting, a large series of cases with assumed viral hepatitis was carefully ascertained and laboratory confirmed, allowing to accurately describe the clinical features of hepatitis A. Due to the prospective collection of viral hepatitis cases over all seasons of 4 years, the seasonality features of hepatitis A could be assessed and documented for the first time for Nicaragua. The screening test diagnosis of acute hepatitis A was confirmed by a recognized reference laboratory.

Conclusions

Nicaragua is a country of still high hepatitis A endemicity, with all its typical features. The majority of cases were oligosymptomatic with no severe pathology seen in over three hundred documented cases. The main risk factor for acquiring hepatitis A is living in low socioeconomic conditions. There are some first indications that the endemicity level started in the 2000s shifting from high to high-intermediate. A closer monitoring of hepatitis A is warranted in Nicaragua in order to be able to start in time with appropriate public health interventions.

Supplemental Information