Prevalence and intensity of Loa loa infection over twenty-three years in three communities of the Mbalmayo health district (Central Cameroon)

Loiasis, better known as the African eye worm, is a parasitic disease caused by the filarial nematode Loa loa and transmitted to humans by tabanids belonging to the genus Chrysops. Chrysops dimidiata and C. silacea, commonly known as deer flies and mango or mangrove flies, are the two main anthropophilic vector species, only found in Africa [1]. The disease is restricted to western and central Africa where a geostatistical analysis of history of eye worm has mapped high risk levels of loiasis in 10 countries where an estimated 14.4 million people live [2]. Although most cases of loiasis are asymptomatic [3], the most striking clinical manifestations of this filarial infection are the migration of the adult worm under the bulbar conjunctiva, and the migratory transient edema called Calabar swellings. In addition, loiasis has been historically ranked as the second or third most common reason for medical consultation after malaria and pulmonary diseases in some communities of Congo and Gabon [4, 5]. Data reporting implication of loiasis on human health are very scanty, and its clinical impact is yet to be clarified [6]. Loiasis is still considered a benign disease eliciting very little attention, although a recent survey has reported an excess mortality associated with the infection [7].

Due to the benign nature attributed to this tropical disease, no large-scale control program is dedicated to loiasis though four chemotherapeutic agents (diethylcarbamazine, ivermectin, albendazole and mebendazole) have been proven effective against the parasite [6]. However, it is now accepted that the most effective drugs on L. loa infection are not really safe when individuals are heavily infected, and the safer ones are not completely efficient [6, 8]. Indeed, although benzimidazoles are safer for the treatment of loiasis, their effects on microfilarial loads are quite modest or transitory [8, 9]. In contrast, diethylcarbamazine and ivermectin are very efficient against L. loa, but therapeutic accidents with sometimes fatal outcomes have been reported in some individuals harboring high microfilarial loads (>8000 microfilariae per milliliter, mf/mL) after treatment with each of these two drugs [10].

In areas endemic for loiasis and where onchocerciasis is meso- (> 40% mf or > 20% nodule prevalence) or hyper-endemic (> 60% mf or > 40% nodule prevalence), ivermectin treatments have been recommended since the clinical impact and consequences associated to onchocerciasis outweighed the risk of post-treatment severe adverse events (SAEs), under specific surveillance system for early detection and management of SAE cases [11]. As a consequence of the collateral impact of ivermectin on Loa loa, the transmission of infection has been reduced in some areas under community directed treatment with ivermectin (CDTI) [12]. However, ivermectin is not recommended in hypo-endemic areas for onchocerciasis (< 40% mf or < 20% nodules prevalence) that are co-endemic for loiasis [11]. One can anticipate an increase in transmission and morbidity as well as life expectancy reduction of infected individuals [7].

The objective of the present study was to follow up the prevalence and intensity of Loa loa infection over 23 years in three neighboring communities of the Mbalmayo health district (Central Cameroon) in order to investigate the long-term trends in infection parameters in an area where no mass anti-filarial treatment was ever distributed.

The survey was carried out in three small neighboring communities (Abang, Akometam and Ngat-Bane) of the Mbalmayo health district. A total of 271 individuals were examined, among whom 140 (51.6%) were females. The median age of the population was 50 years old (interquartile range: 28–63).

Intensity of infection

Regarding the intensity of infection, the mf density arithmetic mean of was 1922.7 (sd: 6623.2; maximum: 59,540) mf/mL. Like prevalence, intensity of infection was similar between communities (p = 0.528), whereas a high heterogeneity was reported between genders and age groups (Table 1). The intensity of Loa infection was higher among males (p < 0.0001) and a convex in form trends was observed among age groups (p = 0.032). The maximum microfilarial load was reached in individuals aged 35 to 49 years old, both in males and females (Table 1; Fig. 1). Besides these observations, 7.4% (7.7% amongst females) and 1.5% (1.9% amongst females) of enrollees harbored microfilarial loads exceeding 8000 mf/mL and 30,000 mf/mL, respectively, with similar trends between males and females (p > 0.816).

23-years change in Loa loa endemicity

Among the 271 individuals tested in 2016, 39 (of whom 27 males) were already present in 1993. A positive association was observed between the prevalence of Loa microfilariaemia and the duration of residence in the community (OR: 1.016; 95% CI: 1.005–1.027; p = 0,003). The infection rate, when considering only those 39 individuals, was relatively higher in 2016 (41.0% (95% CI: 27.1–56.6) than in 1993 (35.9%; 95% CI: 22.7–51.6), but the difference was not significant (Chi-square: 0.22; p = 0.639). Microfilarial densities were also similar between 1993 (mean: 2280.5; sd: 6074.0) and 2016 (mean: 3080.0; sd: 5872.3), though a slight increase was observed in 2016 (Wilcoxon signed ranks tests, p = 0.351) (Fig. 2). Importantly, 50% of heavily infected (microfilarial loads exceeding 8000 mf/mL) enrollees already exhibited high microfilarial densities 23 years ago.

This repeated cross-sectional study was carried out to investigate the long-term trends in prevalence and intensity of Loa loa infection in an area that has never benefitted from community preventive chemotherapy. The overall prevalence of loiasis was 27.3% in the three targeted communities; a figure similar to what was previously observed in the same communities [13–15] and indicating that this area is highly endemic for African eye worm [20]. The prevalence of the African eye worm, estimated by history of either migration of adult worm under the bulbar conjunctiva or the Calabar swelling, was not accurately predicted as previously demonstrated [18, 21]. Several reasons can explain this discrepancy: (i) in the present study, the interview was quicker and lighter (no image was shown) than in previous study; (ii) it was demonstrated that people presenting with loiasis associated clinical signs (migration of adult worm under the bulbar conjunctiva and the Calabar swelling) rarely harbor L. loa mf [22, 23]; (iii) finally the sample size of our study was small (only 177 individuals were interviewed for logistical constraints), thus reducing the power of statistical tests, and the likelihood to accurately capture this relationship. However, a positive association was found between Loa microfilarial prevalence and histories of both migration of eye worm and Calabar swelling.

The mean intensity of infection was 1922.7 (sd: 6623.2; maximum: 59,540) mf/mL, with 7.4 and 1.5% of enrollees harboring microfilarial loads exceeding 8000 mf/mL and 30,000 mf/mL, respectively. The relationship between prevalence and intensity of infection observed in the present study fits with previous observations [20]. Indeed, it was demonstrated that prevalence of Loa microfilariaemia between 20 and 30%, can predict approximately 5–9% of adults harboring microfilarial loads exceeding 8000 mf/mL, and 1–3% of adults with > 30,000 mf/mL [20]. Since severe adverse events have been correlated with high Loa microfilarial loads [10], these results (i) indicate that the implementation of ivermectin treatments in the Mbalmayo health district needs to be done with caution, and (ii) confirm why CDTI is yet to be implemented in this health district [11].

Loa loa prevalence and intensity of infection were found to increase while age was increasing (OR: 1.018; 95% CI: 1.005–1.032; p = 0.007). A convex in form trend was observed among age groups, the maximum microfilarial load being reached in individuals aged 35 to 49 years old (Table 1; Fig. 1). These results might indicate that the exposure to loiasis is increasing with age, the disease appearing to accumulate in infected individuals. In case of accumulation of the parasite consecutive to new infective bites by the vector, one would have expected an increase in prevalence and intensity of infection over time, particularly in the absence of any intervention targeting either the parasite or the vector. However, the convex in form profile displayed in the relationship between age and intensity of infection might suggest that this accumulation is somehow relative and might reverse in advanced age group. Similar results have been reported in previous studies, suggesting a noncumulative nature of this filarial disease [14]. Such observations have also been described in Onchocerca volvulus [24], another filarial nematode closely related to L. loa, and was explained by the fact that in younger individuals the parasite population has not reached its equilibrium, the number of maturing adult worms in these individual hosts exceeding that of dying worms. On the contrary, in older individuals the number of dying adult worms exceeds that of maturing worms as a consequence to their reduced exposure to chrysops’ bites because of their indoor habit related to their age. This latter hypothesis supports the decreasing trend in intensity of infection observed in advanced age.

A follow up of individuals already examined 23 years ago have shown that both the rate and intensity of infection remained almost unchanged (Fig. 2), indicating that the transmission of loiasis is stable over time. This stability in L. loa transmission was already observed in Gabon [25], Congo [26], and Cameroon (same site like in our study) [14] over a shorter interval (1–3 years). The hypotheses underpinning this stability were suggested to be inherent both to the parasite (mf production by adult worms) and to the host (immune defense mechanism). This long-term stability observed in L. loa infection is consistent with a noncumulative status of the disease observed in the present study. A 13-year follow up of loiasis transmission in the Lékié division (Central Cameroon), has also revealed a stability in entomological indices, including (but not limited to) chrysops infection and potential infective rates [27]. Despite these interesting findings, the sample size of this follow-up study was small (only 39 individuals examined in 1993 were found in 2016), and one can question representativeness and the generalizability of this small sample with regards to the general population. It is worth mentioning that similar observations have been previously reported elsewhere, even though within a shorter time frame, suggesting that these findings and associated interpretations are not unique to our study.

This 23-year repeated cross-sectional study confirmed that the prevalence and intensity of Loa loa infection is stable over time, indicating that this filarial disease might be noncumulative [28]. It has been argued that infection with Loa loa can hinder the success of onchocerciasis and lymphatic filariasis control and/or elimination programs as a consequence of post-ivermectin SAEs. Given that this study has demonstrated that loiasis might be noncumulative and microfilarial densities are somehow stable over time, an increased risk of post-ivermectin SAEs in these untreated communities is unlikely or not expected, and the caution with which mass drug administration has to be implemented in loiasis endemic areas will not fundamentally change.