Introduction

The parasitic disease loiasis, caused by infection with Loa loa, is receiving more and more attention from researchers and public health stakeholders alike [1–3]. Traditionally the infection is known for its two main symptoms: eyeworm and Calabar swellings (CS). The former is the migration of an adult (macro) filaria under the conjunctiva, causing irritation, pain and visual disturbances. The latter are localized, non-pitting edemas typically occurring on the upper limbs, particularly on the back of the hands and the lower half of the forearms. Although less common, CS may also appear on the face, presenting as unilateral orbital swelling [4]. Regardless of its location, the edema is sharply demarcated and can reoccur several times per month [5]. In literature, a median duration of the edema of 3–5 days is described, though longer episodes may occur. CS may be itchy, painful, or both at the same time [3,5,6]. While erythema has occasionally been described, neither wounds nor puncture sites suggesting an infection site, nor lymphangitis or accompanying lymphadenopathy are present. Importantly, CS affects the mobility of the wrist and metacarpophalangeal joints, leading to functional limitations and significant disease-associated morbidity [3,7]. CS is considered to be localized allergic reactions to worm antigens, resulting in episodes of angioedema [4,5,8,9]. Adult L. loa filaria are quite motile and migrate extensively through the host’s tissue. In post-mortem dissections high numbers of adults have been found between the fascia of the lower arms, however the reason for the preference to the upper extremities (tropism) is unclear [10]. Interestingly, the genome of L. loa encodes more chemoreceptors compared to other filarial parasites infecting humans, likely facilitating its migrations [11]. Yet, their precise migration pathways, locations of adults during CS episodes, and tissue tropism remain unclear. Recent advances in portable ultrasound devices have expanded their use to resource-limited settings without continuous electricity [12–14]. Building on these advances, this project aimed to assess CS by ultrasound to better understand their pathophysiology and underlying mechanisms.

Methods

Results

Study population/overview

Overall, 22 individuals had visible swellings identified by the fieldworker and were further examined clinically and by ultrasound. Details on the study population are shown in Table 1, and Fig 1 provides pictures of “typical CS” observed during the study. A more comprehensive overview, including detailed descriptions of the swellings is provided in S1 Table.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Overview on the study population including baseline demographics, reported history of eyeworm and of CS, L. loa microfilarial densities, presence of typical CS, presence of other swellings not typical for CS, and appearance of nodules. Notably, two individuals (8 and 16) presented themselves as having swellings but were assessed as having only nodules by the clinician. * typical CS defined as non-pitting transient swellings on the upper extremities or near the orbit, ** other swellings defined as all not consistent with the morphology of the above-mentioned description, *** CS near the orbita. M = male, F = female, MD = missing data.

https://doi.org/10.1371/journal.pntd.0014240.t001

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. “Typical” Calabar swellings seen during the study of hands (A-E) and face (F) as well as an ultrasound image (G: white arrow indicating the bone, black bracket indicating the edema) of the orbital swelling depicted in (F).

Pictures A to F were taken by LV.

https://doi.org/10.1371/journal.pntd.0014240.g001

Based on clinical examination and provided history of the swelling, 8 of the 22 individuals (36.4%) had swellings that were classified as other swellings “not typical for CS” and 12 (54.5%) were classified as “typical CS”. Of those, two individuals had a facial CS presenting as unilateral orbital swellings (0.9%). Additionally to the “typical CS”, 14 (63.6%) individuals had subcutaneous nodules on the forearms and wrists. Simultaneous presence of CS and subcutaneous nodules was seen in 8 individuals (36.4%). Two individuals had only subcutaneous nodules (0.9%), which they referred to themselves as swellings. These nodules presented as bulging elastic lesions of several cm in length and about one cm in diameter and an elevation of some millimeters above the surrounding skin. Clinically, the “nodules” were reported as painful, beginning with an itchy sensation, appearing repeatedly at similar locations, affecting the same extremity, and disappearing after a few days. In all cases the dominant hand was affected.

Clinical and ultrasound results

A total of 56 ultrasound videos from 22 individuals were assessed for relevant findings and pathologies. During this review, it became evident that the ultrasound videos of the “typical CS” showed non-specific subcutaneous and muscular edema, with no further remarkable findings, such as foreign bodies or effusions. In contrast, the ultrasound videos of the nodules showed more complex findings, namely subcutaneous retentions of varying echogenicity. These videos were assessed by both specialists (PM, HP). Finally, videos from three individuals, namely Patient 5, 16 and 17, were selected for in-depth evaluation as they showed the best quality. Both specialists provided written descriptions and interpretations of these videos as well as an overarching summary, blinded to each other’s results.

Summarizing, the experts agreed that these videos showed subcutaneous retentions with structures of varying echogenicity inside. The retentions are transfascial, in that they penetrate the superficial fascia representing holes in this anatomic structure (see Figs 2–4). The expansion of the retentions through the fascia is disregarding the anatomical boundaries and is characteristic of pathological rather than physiological structures (see still images captured from videos in Figs 2C, 2D, 3C–3F and 4E–4H). In one video, several transfascial gaps are visible (S3 Video, see Fig 3D). Within the retentions, hyperechoic structures are visible, which have a thread-like shape. The hyperechoic structures are clearly visible inside the otherwise anechoic retentions. In two videos, from Patient 5 (S2 Video) and Patient 17 (S5 Video) respectively, autonomous movement of the hyperechoic structures inside the retentions is visible. The measured diameters correspond in size to L. loa macrofilariae, whose length is typically 22–36 mm (male) and 38–65 mm (female), and whose maximum diameter is 0.4-0.5 mm [17]. Of the three individuals, two had simultaneous “typical CS” (5 and 17) and the third (16) reported that the CS had just eased the day before. For details on ultrasound descriptions and clinical complaints, see individual patient descriptions below. Plain and edited videos are provided in the S1–S5 Videos files, edited versions include a time display, circles indicating lesions and asterisks when movement can be seen.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Findings of Patient 5: A) Typical Calabar swelling on the left hand; B) Subcutaneous nodule on the volar aspect of the left distal forearm; C) Corresponding ultrasound image revealing a subcutaneous, anechoic lesion presenting as retention, with trans-fascial gap and hyperechoic structure inside the retention.

The thread like structure is less than 0.5 mm in diameter (compare to the 2 mm scale bar, still image from S2 Videos at second 14); D) Corresponding ultrasound image with annotations. Legend: RT = retention; F = filaria; SC = subcutaneous tissue; FAP = palmar antebrachial fascia (orange); FDS = flexor digitorum superficialis muscle. Pictures A and B were taken by LV.

https://doi.org/10.1371/journal.pntd.0014240.g002

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. Findings of patient 16: A) Subcutaneous nodule on the right volar forearm indicated by white arrow; B) the corresponding ultrasound image showing a transversal cut of the subcutaneous nodule presenting as an anechoic retention in the ultrasound, with trans-fascial gap and hyperechoic structure inside (still image from S3 Video at second 15) and D) showing another transversal cut of the same nodule depicting multiple trans-fascial gaps (still image from S4 Video at second 6); C and E) Corresponding ultrasound images with annotations; F) a schematic depiction of the findings on ultrasound.

Legend: RT = retention; F = filaria; SC = subcutaneous tissue; FAP = palmar antebrachial fascia (orange); FDS = flexor digitorum superficialis muscle. Picture A was taken by LV.

https://doi.org/10.1371/journal.pntd.0014240.g003

Download:

• PNG
  larger image
• TIFF
  original image

Fig 4. Findings of patient 17: A + B) typical Calabar swelling on the left hand: both hands in an expanded position (A) and with fists closed (B); C + D) subcutaneous nodule on the left palmar forearm, indicated by black arrow (D); E + F) the corresponding ultrasound images showing the subcutaneous nodule presenting as a retention in the ultrasound, with trans-fascial gap and hyperechoic structure inside in longitudinal (E) and transversal scan (F) (both show still images from S5 Video at seconds 3, respectively); and G + H) Corresponding ultrasound images with annotations.

Legend: RT = retention; F = filaria; SC = subcutaneous tissue; FAP = palmar antebrachial fascia (orange); FDS = flexor digitorum superficialis muscle. Pictures A to D were taken by LV.

https://doi.org/10.1371/journal.pntd.0014240.g004

Discussion

This project aimed to assess a typical sign of loiasis, the so-called Calabar swelling (CS), using ultrasound. CS has previously been described as a major factor in loiasis-related morbidity as it occurs frequently and significantly limits hand function [3,6,7]. Today, loiasis is increasingly seen as a disease of relevance and may be recognized as a neglected tropical disease by the World Health Organization in the future [1]. Understanding its pathophysiology is thus becoming increasingly important to develop adequate and effective treatment to truly be able to improve the daily life of affected individuals.

CS have been described early on in loiasis case reports and are named after a Nigerian town, where this sign was first reported in Western literature in 1895 [18]. Traditionally, they are described as “transient, non-pitting edema occurring at hands and wrists” and are considered to occur as a localized allergic reaction to the parasite antigen [8,9]. However, the exact mechanism, and why CS tends to occur above the wrists and hands have never been fully understood. In endemic regions CS are often treated by homemade or traditional remedies, including hot wrappings, hot water or creams as well as anti-inflammatory drugs [19]. Currently there are no specific treatment guidelines available for CS. A better understanding of the mechanisms associated with CS pathology may help develop and evaluate new specific treatment strategies.

Advances in ultrasound technology allow higher resolution of soft tissue and have been used to visualize macrofilariae of various species. Live and moving macrofilariae of Wuchereria bancrofti and Brugia species can be visualized in the lymphatic system by ultrasound - a finding also known as the filarial dance sign [20,21]. Several case reports on ultrasound detection of subcutaneously residing Dirofilaria spp. macrofilariae have been published [22,23] and ultrasound of Onchocerca nodules also showed moving filaria [24], specifically echogenic filarial structures in fluid-filled retentions. Otherwise, there is only limited literature of L. loa macrofilariae residing in subcutaneous tissue; one case report described a pipeline-shaped lesion as an ultrasound finding in the subcutaneous tissue at the calf of a patient with loiasis consistent with the morphology of a macrofilaria [25].

Apart from case reports, CS themselves have never been systematically examined by ultrasound. Postmortem autopsies performed in 1905 on a L. loa-infected individual showed 34 L. loa macrofilariae residing in the fascial layers of the palmar forearms [10]. Here, based on previous descriptions, we hypothesized that macrofilariae may be detected by ultrasound around or next to CS [25–28]. Importantly, during the project, physical examination revealed subcutaneous nodules on the palmar side of the wrist in a relevant proportion of patients suffering from typical CS. The nodules were described as painful, acutely occurring at the volar distal forearm. Patients stated that the occurrence of the nodules coincides with swelling of the dorsal and palmar side of the hand, which is consistent with the classic clinical picture of CS [4]. All affected patients reported that the nodules would occur several times per year. To our knowledge these nodules have not been analyzed by ultrasound or systematically described in the literature before. In ultrasound these nodules presented as anechoic retentions with hyperechoic filarial like structures inside. In two ultrasound videos we were able to detect autonomous movement of filaria-like structures through holes in the fascial layers of the forearm. The ultrasound findings of our study suggest that these structures actively penetrate fascial layers and thus cross anatomical barriers. The structures seem to move from the subcutaneous tissue to deeper layers of tissue, e.g., tendon sheaths, away from the transducer while applying pressure. The structures were not visualized in their entire length, but their estimated size would be 0.5 mm in diameter and a minimum of 6 mm in length, probably longer. This would be consistent with the size of a L. loa macrofilaria [17,27]. We therefore hypothesize that L. loa macrofilariae cause these described nodules by penetrating the superficial fascia of the forearm, thereby causing pain and functional disability. Fig 5 schematically displays this finding.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 5. A schematic presentation of the trans-fascial migration of an adult L. loa parasite(macrofilaria), including fascial lesions and a fluid retention.

The figure is original work created by the authors. The authors are the sole copyright holders and grant permission for publication of the figure under the Creative Commons Attribution 4.0 (CC BY 4.0) license.

https://doi.org/10.1371/journal.pntd.0014240.g005

This has been further supported by the small-scale reliability study, which found moderate to good reliability among a group of international experts on small-part ultrasound. While the overall reliability was limited, the agreement on the finding in the video from patient 5 (S2 Video) was 100% in both rounds, and false positivity was mainly driven by two videos, which limited the overall outcome of the reliability study.

Trying to understand the pathology behind this finding, it is important to note that the affected region is an anatomical site with special features. The tight situs in the volar wrist region causes a close relation of fasciae, nerves and muscles with hardly any space in between [29]. Thus, a filarial migration might lead to increased pain and functional restriction. Interestingly, all affected individuals reported occurrence of both CS and nodules on their dominant hand. The affected individuals are all residents of rural areas working in subsistence farming. This implies hard physical labor, which puts significant strain daily on their wrists and hands. It seems reasonable that physical stress may additionally force the L. loa filariae to move. Another aspect could be that individuals may have chronic tendinopathies, such as recurrent tenosynovitis possibly predisposing these areas for worm migration and facilitating further damage by the parasite. The classical presentation of CS may be a secondary reaction to the process, driven by local inflammation. Further supporting this theory, it has been described that CS can be induced by a knock at the site, or even by rubbing the skin, which may damage the worm, leading to the release of antigens and subsequently causing the angioedema [8,9]. The extensive tissue swelling is, however, not directly above the subcutaneous nodules. This may be explained by the fact that the soft tissue lying superficially directly above the carpal tunnel is extremely thin in contrast to the superficial palmar areas where muscles and subcutaneous tissue are thicker and constitute more space for swelling [29].

Anecdotal reports from affected individuals describe the impact of physical pressure on adult worm (macrofilarial) movement in loiasis, including the prevention of eyeworm migration if pressure is applied, thereby providing a physical barrier and preventing migration of the adult worm into the subconjunctival space (personal communication). Further, in highly endemic areas, patients are encouraged to keep their heads down to keep the worm under the conjunctiva while waiting for surgical removal [5,30]. In our study, the filarial structures tended to move away from the ultrasound transducer from the more superficial tissues towards the deeper layers. Since the ultrasound transducer must be applied to the skin with a certain amount of pressure, this also corresponds to the image of a moving filaria reacting to physical impact.

We did not surgically remove the macrofilariae seen in the ultrasound examination due to limited resources and ethical considerations and thus cannot prove that the documented structures are definitely L. loa macrofilariae. However, the affected individuals reside in highly endemic areas for loiasis and, among the three individuals with convincing ultrasound findings, two experienced simultaneous typical CS and nodules, and the third reported a swelling that had just eased. While some experts would consider CS as diagnostic for the infection, we propose to discuss further possible differentials. Other filariae such as Onchocerca volvulus, Wuchereria bancrofti and Mansonella spp. are of low endemicity in the Lékoumou district. The clinical presentation of these macrofilariae is different, as they cause subcutaneous stationary, persistent nodules in onchocerciasis or reside in the lymphatics in the case of lymphatic filariasis, respectively [15,31]. Mansonella spp. may be similar in size but are of low endemicity in the study area. Dirofilaria spp. have been described to cause subcutaneous nodules similar to the ones described here; however, they are considered to be non- endemic in the study area. Non-parasitological differential diagnoses would be thrombosis or inflammatory phenomena such as gouty tophi, other infections, or destructive processes. However, due to the documented autonomous movements in two different cases and the slender shape of the structures within the retentions, these are very unlikely.

Next to the missing parasitological confirmation of the finding, there are other limitations that warrant discussion. First, for microfilaria detection we used calibrated thick blood smears but no concentration method, which may have led to missing some microfilaremic individuals. Next, we assessed swellings at the upper extremities and face but not on other body parts. This was done for simplicity of the field work and may have missed CS on other body parts. However, this study did not aim to assess the true prevalence of CS but ultrasound findings of the typical clinical presentation. Further, the quality of the ultrasound videos is limited as we used a handheld transducer connected to a smartphone. Thus, during the ultrasound, pictures were only seen on a small screen on the phone, making direct interpretation in the field difficult. Further, we experienced some issues of connection, and some ultrasound videos were stalled several times. Unfortunately, the Doppler function of the handheld ultrasound transducer was not working. Future studies should use better ultrasound machines to allow recording of more detailed videos. Finally, while the design of the reliability study was based on previous work from the Outcome Measures in Rheumatology Ultrasound Working Group [32], the video dataset was limited by the small sample size of the study. Next steps include the development of consensual definitions of the filaria-like lesions on ultrasound, a potential scoring system - binary, semiquantitative, or quantitative - and the replication of the reliability study on a larger video dataset. This would be followed by a reliability exercise on patients, including the acquirement of videos and real-time scoring of lesions [33].

Conclusion

Considering the ultrasound findings and clinical complaints reported by the patients, we believe that our results are most consistent with a trans-fascial migration of adult L. loa parasites, associated with fascial lesions and fluid retentions. The classical clinical picture of Calabar swelling seems to appear in close relation but not directly above these lesions. Loiasis is a complex disease, with typical clinical manifestations caused by migration of adult worms and Calabar swellings are not just benign swellings. They may indicate an underlying destructive and disabling process and therefore need to be addressed by further studies and public health programs to provide adequate treatment to affected individuals.

Supporting information

Acknowledgments

We sincerely thank all study participants for their invaluable contribution to this research. We are grateful to the field workers and MorLo Study team for their support in participant recruitment and sample collection. We thank Philip J. Budge and Peter U. Fischer for their valuable input on interpretation on the study findings.