Introduction

Loiasis, caused by the filaria Loa loa, is a parasitic disease endemic in rural West and Central Africa [1]. It is transmitted by tabanid flies of the genus Chrysops and 14 million people are estimated to be infected. In hyperendemic regions prevalence often exceeds 60% in adults [2]. Loiasis is commonly known for the pathognomonic sign of the “eye worm”, which is caused by adult L. loa filariae transiently migrating under the patient’s conjunctiva. This distinct clinical feature has been used as an epidemiological marker to assess community prevalence of loiasis in endemic settings (“RAPLOA” survey) [2].

Adult L. loa filariae can live more than 15 years in their definitive hosts, continuously migrating through the soft tissue and reproducing sexually. Female worms release microfilariae (mf) which can be found in the peripheral blood of patients, in sometimes very high numbers (up to 300,000 mf/mL) [1,3]. Loiasis without detectable microfilaremia is commonly referred to as “occult loiasis”. Determinants for the presence or level of microfilaremia are incompletely understood and may include host genetics, varying immune responses, parasite factors such as number of adult worms, worm fecundity or sex ratio [4–6]. The majority of clinical data in scientific literature stem from reports on travelers and temporary residents of endemic regions [1,7–11]. Only few studies systematically evaluated the clinical picture of loiasis in long-term residents of endemic regions, not least due to the supposedly benign disease course [4,12–17]. Chronic infection has been associated with unspecific manifestations, such as itching, body pains or fatigue and importantly, individuals with high microfilarial densities are known to be at risk for spontaneous or treatment-related severe manifestations and excess mortality [1,7,12,13,18,19]. Eosinophils are known to be important immune-effector cells in parasitic infections and an association between loiasis and eosinophilia has been described in short term residents, but their role in chronic loiasis has barely been studied [8,9,12].

While loiasis is in general increasingly understood as an infection of clinical relevance, many questions on the clinical features and manifestations in the most affected populations are still unanswered. To improve our understanding of the disease, we evaluated clinical and hematological findings associated with loiasis in a large cohort of individuals living in rural Gabon [18].

Methods

Statistical analysis

Statistical analysis was performed using the software package R (Version 4.0.5, R Foundation for Statistical Computing, Vienna, Austria). Participants were categorized into “infection states”, defined by presence/absence of eye worm history and presence/absence of detectable microfilaremia. Thus, the main L. loa infection states were: 1) No sign of L. loa infection, i.e. absence of microfilaremia and no history of eye worm (LN), 2) detectable microfilaremia but no history of eye worm (MF), 3) positive history of eye worm but no detectable microfilaria (EW) and 4) history of eye worm and detectable microfilaremia (EWMF). To assess differences according to the level of L. loa microfilaremia, microfilaremic participants were categorized into three groups: low (1–7,999 mf/mL; LMF), high (8,000–19,999 mf/mL; HMF) and hyper-microfilaremia (≥20,000 mf/mL; HYMF) (see Table 1). Occurrence of reported symptoms during the previous three months and hematological parameters were analyzed and compared between infection states. Calabar swelling was not considered as diagnostic criterion as it has been shown to be of limited specificity but was included as symptom in data analysis if it occurred during the previous year [2].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Definitions of the classified infection states described in the paper.

https://doi.org/10.1371/journal.pntd.0010793.t001

Categorical variables were compared using the χ² test or the Fisher exact test as appropriate. Multi-group comparisons were performed and corrected for multiple comparisons using the false discovery rate method. Continuous variables were expressed as median and interquartile range (IQR). To correct for demographic parameters and filarial co-infections, multiple regression models were run and adjusted for the following possible confounders: age group, sex and presence of M. perstans microfilariae. In case of logistic regression analysis, adjusted prevalence ratios (aPR, 95% confidence interval, CI) were computed.

Results

Study population

Data on history of eye worm migration and microfilaremia were available for 1,232 participants. First, the cohort was categorized into the above described four mutually exclusive infection states LN, MF, EW and EWMF. MF, EW and EWMF are thus the subgroups of individuals with definite signs for loiasis (LP).

Within the loiasis positive infection states (LP, n = 626), EW was the most common infection state (52.4%, n = 328), followed by EWMF (30.7%, n = 192) and MF (16.9%, n = 106) (see Table 2). Inter-group comparisons between the infection states revealed significant differences between age and sex distribution. There was no difference in the distribution of extent of microfilaremia between women and men, nor between the adult age groups when comparing the microfilaremic categories (see Table 3).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Distribution of subjects according to the infection state within each age and sex category and comparisons of these distributions between infection state categories.

Distributions by age categories were compared using the Fisher’s exact test, and distributions by sex were compared using the χ2 test.

https://doi.org/10.1371/journal.pntd.0010793.t002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Distribution of subjects according to their microfilaremia category within each age and sex category and comparisons of these distributions between microfilaremia categories.

Distributions by age and sex categories were compared using the Fisher’s exact test.

https://doi.org/10.1371/journal.pntd.0010793.t003

Loiasis infection states and clinical presentation

Occurrence of the disease manifestations were compared between the infection states and results were adjusted to possible confounding factors including sex, age and M. perstans microfilaremia (see Tables A, B, C in S1 Appendix). Fig 1 displays the frequency of reported manifestations by infection state, and adjusted p-values of intergroup comparisons if significant. In comparison to LN, each of the two groups with history of eye worm (EW and EWMF) reported several signs and symptoms significantly more often. This was the case for Calabar swelling-like manifestations (adjusted prevalence ratio [95% confidence interval]: (1.40 [1.10–1.79] for EW and 2.05 [1.56–2.69] for EWMF), paresthesia (1.51 [1.19–1.92] and 1.73 [1.32–2.26]), transient paralysis of extremities (2.63 [1.86–3.72] and 1.78 [1.15–2.76]), and fatigue (1.81 [1.43–2.30] and 1.49 [1.12–2.01]). In addition, individuals in the EW category reported myalgia and severe headache more often compared to LN (1.36 [1.01–1.82] and 1.33 [1.15–1.54], respectively). The frequency of the reported manifestations was similar between MF and LN subjects.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Frequencies of signs and symptoms within the three loiasis infection states and in uninfected individuals as bar plots and adjusted p-values of inter group comparisons, if significant.

All intergroup comparisons are adjusted to sex, age and Mansonella PCR. * p-adj. <0.05; **p-adj. <0.01; *** p-adj. <0.001; * LN = No sign of loiasis infection; MF = detectable microfilaremia but no history of eye worm; EW = positive history of eye worm but no detectable microfilaria; EWMF = positive history of eye worm as well as detectable microfilaremia.

https://doi.org/10.1371/journal.pntd.0010793.g001

Comparing the 3 LP categories among each other, EW and EWMF subjects reported some manifestations more frequently than MF individuals. This was the case for myalgia (1.79 [1.08–2.97] and 1.76 [1.04–2.99], respectively) and Calabar swelling-like manifestations (1.55 [1.00–2.40] and 2.27 [1.45–3.55]). Further, EW subjects reported several disease manifestations more often than the MF subjects; this was the case for transient swellings of any body part (other swellings 1.81 [1.05–3.10]), transient paralysis of extremities (2.84 [1.48–5.47]), urticaria (2.38 [1.23–4.59]) and fatigue (1.53 [1.02–2.30]). Comparing the EW and EWMF categories, patients in the EWMF group reported significantly fewer Calabar swelling-like manifestations (0.68 [0.51–0.91]), whereas patients in the EW category indicated more often urticaria (p = 0.062, 1.48 [0.97–2.24]). HMF and HYMF individuals reported more often an episode of eye worm migration during the previous year compared to LMF (58.2% versus 35.5%; p-adj. = 0.003).

In summary, the infection states comprising participants with history of eye worm were significantly associated with the occurrence of clinical complaints. Contrarily, microfilaremic individuals reported complaints in a similar proportion as persons without signs of loiasis infection. The proportion of reported manifestations was independent of extent of microfilaremia, with exception for occurrence of eye worm.

Loiasis infection states and hematological phenotypes

Hematological findings were analyzed following the same categorization and adjustments as described above (see Table D and E in S1 Appendix). Data for differential blood count were incomplete or missing for 133 (10.8%) participants. There was no statistical difference in hemoglobin values between LP and LN nor between the different loiasis infection states. However, LP had higher total WBC counts (median [inter quartile range]: 7.4 [5.9–8.9] vs. 6.7 [5.4–8.4] x10³ leucocytes/μL, p-adj.<0.001), as well as higher absolute eosinophils/μL (1.1 [0.5–1.8] vs. 0.6 [0.25–1.3] x10³, p-adj.<0.001) and relative eosinophil counts (15.0% [8.0–22.0] vs. 9.2% [4.1–17.0] of total WBCC, p-adj.<0.001) compared to LN. Differences in WBC counts and eosinophilia were also found when comparing the loiasis infection states individually to LN and among each other. Fig 2 displays the results graphically including adjusted p-values.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2.

Distribution of hematological findings including median and interquartile range of hemoglobin (A), total white blood cell count (B), absolute (C) and relative eosinophil counts (D) within the four infection states. Adjusted p-values of intergroup comparisons are provided if <0.05. All intergroup comparisons are adjusted to sex, age and positivity of Mansonella PCR. * p-adj. <0.05; **p-adj. <0.01; *** p-adj. <0.001; LN = No sign of loiasis infection; MF = detectable microfilaremia but no history of eye worm; EW = positive history of eye worm but no detectable microfilaria; EWMF = positive history of eye worm as well as detectable microfilaremia.

https://doi.org/10.1371/journal.pntd.0010793.g002

Comparing all individuals with microfilaremia (MF and EWMF) to patients with history of eye worm only (EW), the two former categories were characterized by higher absolute eosinophil (1.3 [0.8–1.85] and 1.4 [0.8–2.0] vs. 0.8 [0.34–1.6] 10³/μL; p-adj. = 0.076 and 0.003) and relative eosinophil counts (18.0% [11.0–23.0] and 17.0% [12.0–24.0] vs. 13.0% [6.0–21.0]; p-adj. = 0.016 and 0.002). To further investigate this association between microfilaremia and eosinophilia, comparisons were made between the microfilaremic groups. The HYMF group had higher absolute eosinophil counts than HMF and LMF (1.70 [0.60–2.70] vs.1.40 [0.90–1.90] and 1.30 [0.80–1.90] 10³/μL; p-adj. = 0.034 and p-adj. = 0.021), respectively (see Fig A in S1 Appendix). Of note, of the 56 individuals in the HMF and HYMF groups, none had an absolute eosinophil count within normal range (≤0.5x10³/μL). Next, a simple linear regression analysis of the microfilaremic sub-population of the study, with at least 2 microfilariae per mL revealed an increase in absolute eosinophil count with every 10-fold increase in parasitemia (p-adj. = 0.012, ß-estimate: 0.17 [0.04–0.31], see Fig B in S1 Appendix).

In summary, loiasis was associated with absolute and relative eosinophilia. Extent of eosinophilia was higher in microfilaremic individuals in comparison to only eye worm positives. The level of eosinophilia was associated with the quantity of L. loa microfilaremia.

Discussion

Here we describe the clinical and hematological findings associated with L. loa infection in a large population of individuals living in a highly endemic region. We found that distinct infection states of loiasis were associated with specific clinical and hematological findings (see Fig 3A). The loiasis positive infection states can be classified into three mutually exclusive categories based on diagnostic criteria. These categories consist of individuals having a positive history of eye worm (EW), microfilaremic individuals (MF) and individuals having both (EWMF) (Fig 3B). The associated clinical and hematological outcomes do overlap in individuals who had both diagnostic criteria but seem to be associated with the specific infection states (see Fig 3C and 3D).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Graphic presentation of the hematological and clinical findings within the study population.

3A is a diagram depicting the distribution and overlap of the diagnostic findings of history of eye worm, microfilaremia and eosinophilia in the study population (Eosinophilia data missing for 133 (10.8%) participants, see supporting information Table D). 3B depicts the distribution of the three mutually exclusive infection states in the same diagram based on diagnostic criteria. These include microfilaremic (MF), positive history of eye worm (EW), and individuals positive for both (EWMF). 3C displays the clinical outcome of eye worm positive loiasis and 3D the hematological outcome of microfilaremic loiasis in this schematic display. Note that these outcomes are not mutually exclusive, but overlap in individuals who are positive for both diagnostic criteria.

https://doi.org/10.1371/journal.pntd.0010793.g003

Up to now clinical and hematological findings associated with loiasis have been described predominantly in travelers and temporary residents of endemic regions. However, differences in the clinical and immunological outcomes between short term residents and individuals living in endemic regions, as in other parasitic infections, are evident [1,3,4,7–13,15]. While it is known that loiasis causes various manifestations, including signs and symptoms of a rather “allergic type” and varying degrees of immune-activation, these effects have been described to be more predominant in travelers and short term residents than in residents of endemic regions with chronic infections [10,11]. However, knowledge on disease outcomes in individuals living in endemic regions is scarce and improved understanding of the disease in this patient population is now increasingly warranted [8–10,19,22–26].

In the here presented comprehensive dataset, participants of both sexes were found to be similarly often infected with L. loa, however microfilaremic loiasis was found disproportionally more often in men and only eye worm positive loiasis in women. Within the adult population the occurrence and distribution of microfilaremia remained rather stable in the different age groups, a pattern that has repeatedly been described and seems to be consistent across various endemic areas [1,5,27].

Comparing the loiasis infection states between each other showed that history of eye worm migration was associated with the presence of a wide range of signs and symptoms, independently of presence of microfilaremia. At the same time, individuals reporting only history of eye worm had only moderate elevation of absolute eosinophil counts. Consequently, individuals having a high symptom load, often lack diagnostic features of infection in the blood which would facilitate definitive diagnosis, such as microfilaremia or extensive eosinophilia. This constellation may hamper diagnosis, and could cause inappropriate clinical management and treatment while these patients are indeed in need of appropriate clinical care to improve their symptoms. Additionally, these findings underline that the omission of patients reporting only history of eye worm does not only significantly underestimate the true disease prevalence, but also the extent of associated symptoms, and thus the perceived and calculated associated disease burden [1,4,14,18,24]. Importantly, we found that neurological symptoms, including paresthesia and transient paralysis of extremities, were strongly associated with loiasis, specifically with eye worm positive loiasis and not with microfilaremia itself. While these symptoms have been described in reports on returning travelers, data on their occurrence in individuals living in endemic regions are scarce [7,28,29]. However, in the here described population these neurologic findings were quite frequent in the loiasis positive individuals and should thus be further investigated in future clinical studies. In addition to the neurological symptoms, a range of unspecific symptoms, such as fatigue or headache, were quite common and should therefore be taken into account by clinicians as diagnostic hints for loiasis.

Contrarily, microfilaremic individuals reported signs and symptoms in a similar frequency as individuals with no sign of loiasis infection, indicating that microfilaremia was not in itself associated with the queried symptoms. A potential explanation for this unapparent disease presentation could be a more Th-2 driven immune response leading to a lack of “reactive” symptoms in these individuals. A small study comparing microfilaremic to amicrofilaremic individuals provided data supporting this hypothesis, but clearly, this theory still needs to be investigated in future studies [22].

Interestingly, in more historic descriptions of loiasis, when only microfilaremic individuals where diagnosed and eye worm positive loiasis was not taken into account, the infection was described to be rather asymptomatic—an observation that still shapes the perception of this disease today [7,18,19,24]. The here described finding, that microfilaremia per se is not associated with symptoms, could thus serve as an explanation for this discrepancy between current and historical disease descriptions.

Eosinophils are paramount for host defense against helminth infections and significant eosinophilia can be caused by various helminthic pathogens [30,31]. In loiasis, differences in the extent of eosinophilia and eosinophil activation have been noted between temporary residents and residents of endemic regions [8–11]. While L. loa-associated eosinophilia has been described in smaller cohorts, to our knowledge this is the first analysis of eosinophilia by distinct loiasis infection states in a large group of individuals from an endemic region [13]. Here, all loiasis positive infection states were associated with higher absolute eosinophil counts compared to individuals without signs of loiasis. Importantly, eosinophilia was higher in microfilaremic individuals than in amicrofilaremic loiasis. Furthermore, the extent of eosinophilia seemed to be associated with L. loa microfilarial density and, conclusively all 56 individuals with ≥8,000mf/ml, had eosinophil counts above the normal range (≤0.5x10³/μL). The difference of absolute eosinophil counts between individuals with and without signs of loiasis and between microfilaremic and amicrofilaremic loiasis patients was statistically significant and may also be of clinical relevance. Of note, microfilaremia with related blood eosinophilia was not associated with subjective symptoms.

It is established that chronic eosinophilia can cause organ damage due to chronic inflammation [32]. The here described finding of an association between extent of eosinophilia and microfilaria counts, might therefore pathophysiologically explain the previously described excess mortality in highly microfilaremic individuals [13,19]. However, this theory needs further investigation to be verified or refuted.

Further, one could hypothesize that combining the subjective well-being of hyper-microfilaremic individuals and their previously described risk for excess mortality, could lead to inappropriately low health care seeking behavior in those most at risk for severe disease outcomes.

An important limitation of our analysis is the potential confounding of eosinophilia by other parasitic infections prevalent in the study region [3,15,20]. But, despite this co-endemicity of various parasites, hematological changes were consistently found to be associated with different infection states of loiasis, supporting the hypothesis that eosinophilia is indeed primarily caused or at least significantly augmented by L. loa infection. In line with this data, is the fact that the occurrence and extent of eosinophilia has long been known to depend on infection states in other filarial diseases, such as lymphatic filariasis and onchocerciasis, further supporting the here presented findings [33–35].

To understand the relevance of eosinophilia in loiasis in more detail, more clinical and immunological data are needed, especially as it is known that the provoked pathology depends on the activation of these effector cells [36]. Different immune activation mechanisms may underlie the different disease manifestations. Further longitudinal studies investigating the association of specific immune responses, including eosinophilic activation markers with infection states, would therefore be of interest.

Importantly, the cross-sectional study design does not allow a direct inference of a cause-effect relationship between associated factors, nor the analysis of natural evolution of single versus repeated infections or assessments of more clinically serious outcomes such as organ lesions, organ damage or death [37]. Of note, due to the cross-sectional study design, laboratory findings collected at a single time point, including microfilaremia and eosinophilia, and retrospectively reported findings, including history of eye worm and occurrence of symptoms, were compared. This is a clear limitation of this study. Therefore, it is evident that the reported findings and raised hypotheses need to be addressed in prospective, clinical studies for further verification. However, due to the long duration of infections and the rare appearance of worms in the eye—a major diagnostic criterion—longitudinal studies of loiasis are extremely challenging.

In general, there are several pitfalls in loiasis diagnosis and associated difficulties have repeatedly been discussed [38]. Even in individuals where no microfilariae can be found, it is important to state that microfilaremia might still be present but just below the limit of detection of the used method. Here we used a combination of microscopy of thick blood smears and an additional concentration step assessing 1 mL of blood in order to raise the sensitivity of microfilariae detection. A qPCR based on previously published methods by Fink et al., 2011 was performed on all samples, but did not improve sensitivity and therefore results were not used for classification in this analysis (manuscript in preparation) [39]. However, very low microfilaremic individuals might have been missed and thus might have been wrongly classified as individuals without microfilaremia.

Serology has been shown to be cross-reactive and seroprevalence is very high in the study population as shown by a recent analysis (manuscript under preparation). Thus, currently available serological methods seem inappropriate to definitively distinguish active from past infection and counting all seropositive individuals as loiasis positive would likely overestimate the true disease prevalence [38,40,41]. In the absence of an adequate diagnostic tool, proving or refuting an active infection in amicrofilaremic individuals is difficult, as adult filaria might not always become evident as eye worm. In fact, it is not known which percentage of actively infected individuals never experience an eye worm migration. On the other hand, it is not known how long after a reported eye worm migration an individual remains actively infected. While it would be possible that the parasite died shortly after appearing as an eye worm, based on clinical reports it is estimated that adult L. loa filaria can live more than 15 years in the host [23]. Combining these factors, it becomes evident that history of eye worm is an imprecise diagnostic feature and does not allow to differentiate active from past infection. This may have led to misclassifications of individuals into incorrect infection states in the presented analysis. Nevertheless, the standardized questionnaire for history of eye worm (RAPLOA) remains one of the most important diagnostic tools for loiasis until better methods for detection of occult loiasis are developed. Therefore, while not allowing to exclude an infection, RAPLOA was used together with microscopy for microfilaremia, for diagnosis of loiasis in this study [2]. In conclusion, while there are obvious limitations of the data due to the cross-sectional study design, our findings are based on the best available diagnostic methods.

Further, the here provided findings were described in a large group of individuals who are likely to be re-exposed and chronically infected over several years. This underlines the significant clinical and hematological impact of loiasis, and refute once more the notion of loiasis being an inconspicuous, asymptomatic infection [1,7,12–14,19,23,37]. Contrarily, loiasis apparently shares important similarities to other filarial diseases, for which the respective infection states leading to various disease outcomes have long been acknowledged [34,35].

These data show that loiasis is a complex disease, manifesting with distinct clinical and hematological outcomes in individuals living in a highly endemic region. The consequences of these findings for the affected individuals or on populations as well as the pathophysiological consequences of correlating eosinophilia are largely unknown. Currently, loiasis remains highly neglected and individuals in endemic regions are often left untreated [18,19,24,25]. High-quality research on loiasis needs to be further fostered to provide adequate treatment options for the most affected populations and to drive the development of adequate elimination strategies.

Supporting information

Acknowledgments

We want to thank all communities that welcomed us for this survey and sincerely thank all volunteers for participation. Further we would like to thank all colleagues who helped during the survey, especially Madame Augustine Epee and Roger Brothier Mouckoga. We thank Nina Hackbarth for establishing and performing the Mansonella perstans PCR and Matthew B.B. McCall for substantial support and help to establish diagnostic procedures.