External ophthalmomyiasis caused by Dermatobia hominis. A retrospective study of nine cases and a review of the literature

Abstract.

Purpose: To report nine cases of external ophthalmomyiasis caused by Dermatobia hominis.

Methods: Retrospective, non-comparative, interventional case series. Participants consisted of patients (n = 9) presenting at Cayenne Hospital between 1968 and 2003. The location and number of larvae, the larval stage, and the medical and surgical procedures applied were studied in each case.

Results: Seven patients had palpebral myiasis (including one with three larvae) and two had conjunctival myiasis. Every patient had palpebral oedema. The larval respiratory pore was located on the palpebral skin or free margin or on the conjunctiva. Movements were present within the lesion in at least three patients. Petroleum ointment or ivermectine solution was used in at least four patients to smother or kill the larvae. Extraction under local anaesthesia was possible in six patients, while three required general anaesthesia.

Conclusion: Several larvae may be present in a patient. Topical ivermectine may help to kill the larvae before extraction is attempted.

Introduction

Myiasis is the infestation of living vertebrate animals by fly maggots, that is dipterous larvae (White 1996a, 1996b). Three parts of the body may be attacked: cutaneous tissue, body cavities or gut lumen (White 1996a). Dermatobia hominis causes cutaneous myiasis in humans (Wilhelmus 1986). It may very occasionally cause external ophthalmomyiasis (McGavin 1996), with eyelid (Pradinaud & Rivierez 1968; Cordero-Moreno 1973; Savino et al. 1986; Wilhelmus 1986; Orihel & Ash 1995; Bangsgaard et al. 2000; Emborsky & Faden 2002) and conjunctival (Cordero-Moreno 1973; Goodman et al. 2000) involvement. Myiasis caused by Dermatobia hominis is a tropical disease found in Central and South America (Cordero-Moreno 1973). Myiasis infestation may, however, be encountered worldwide as a result of international travel (Cordero-Moreno 1973; Bangsgaard et al. 2000). We report a series of nine patients with external ophthalmomyiasis caused by Dermatobia hominis and compare these cases with those previously reported in the literature with the aim of better outlining the clinical features and therapeutic management.

Material and Methods

This paper describes a retrospective, non-comparative, interventional case series.

Between 1968 and 2002, at least nine patients at Cayenne Hospital were diagnosed with external ophthalmomyiasis due to Dermatobia hominis. Medical records and photographs were available for six of the patients, allowing a correct analysis of each case. Three other cases were collected from the Dermatology Department's photographic library. It was, unfortunately, impossible to trace the files back from these photographs. Data concerning these three cases (patients 7, 8 and 9) are consequently incomplete.

For each patient, the location and number of larvae, the larval stage, and the medical and surgical procedures applied were studied.

The case of patient 1 was previously reported by Pradinaud & Rivierez (1968).

The data pertaining to patients 4, 5 and 6 are fully reported here. Data concerning all patients are summed up in Table 1.

Table 1.  Data concerning the presented series of nine cases of external ophthalmomyiasis caused by Dermatobia hominis.

Patient	Year	Age	Sex	Myasis origin	Number and location of larvae	Symptoms	Lesion	Discharge	Ointment	Antibiotic treatment	Larva removal	Larval stage
y =years; m = months. * Pradinaud & Rivierez 1968.
1*	1968	25 y	Male	Brazil	1 (right upper eyelid)	?	Furuncular blepharitis	No	Petroleum	No	Traction with forceps and pressure on eyelid	?
2	2000	1.5 m	Female	French Guiana	1 (right lower eyelid)	?	Oedema Pore near inner canthus	Serosanguinous	Ivermectine	No	Surgical (local anaesthesia)	?
3	2001	2 m	Female	French Guiana	3 (1 left upper, 2 left lower eyelids)	?	Movements seen in dome- shaped lesion	Serosanguinous	Ivermectine	No	Surgical (local, then general anaesthesia)	? (piecemeal removal)
4 (Fig. 1)	2002	39 y	Male	French Guiana	1 (right upper eyelid)	Moderate pain, wriggling sensation	Pseudo preseptal eyelid cellulitis	Serosanguinous and purulent after larva death	Petroleum and then ivermectine	Yes (oxacillin)	Surgical (general anaesthesia)	2
5 (Fig. 2)	2002	3 y	Male	French Guiana	1 (right upper conjunctival fornix)	Pruritis	Eyelid oedema Pore in conjunctival fornix	No	No	Yes (amoxicillin/ clavulinate)	Surgical (local and general anaesthesia)	2
6 (Fig. 3)	2003	13 y	Male	French Guiana	1 (right caruncula)	Smarting sensation; movements within lesion	Caruncula larvae intermittently protruding	No	No	No	Surgical (local anaesthesia)	2
7	?	15 y approx	?	French Guiana	1 (left upper eyelid)	?	Dome-shaped lesion Pore on free margin	Serosanguinous	?	No	Surgical (local anaesthesia)	2
8	?	1 y approx	?	French Guiana	1 (right lower eyelid)	?	Erythema and moderate oedema around pore	No	?	No	Surgical (local anaesthesia)	2
9	?	45 y approx	Male	French Guiana	1 (right upper eyelid)	?	Oedema Pore on thickened free margin	Serosanguinous	?	No	Surgical (local anaesthesia)	3

Results

Patient 4

A 39-year-old Brazilian gold-washer presented with an inflamed and swollen right eyelid. Preseptal cellulitis was suspected. The patient's tetanus vaccination was updated. Two different systemic antibiotic treatments (amoxicillin/clavulinate and then ampicillin and gentamycin) failed to improve the patient's condition. The patient believed that he had what he called a palpebral ‘ver macaque’, the French vernacular denomination for ‘Dermatobia hominis larva’, as he felt movements within the eyelid lesion. He was referred to our department 10 days after onset with the suspected diagnosis of ophthalmomyiasis.

An inflamed right palpebral swelling and a right pretragial lymphadenopathy were noted. On closer examination, an aperture oozing a clear discharge was noted just above the free margin (Fig. 1A). Petroleum was applied to the aperture, after which the posterior end of the larva appeared at the aperture and then protruded above the thin petroleum layer (Fig. 1B). An attempt to pull the larva out with forceps under local anaesthesia after grasping the posterior end failed. Ivermectine (10 mg/ml solution) application was carried out three times a day for 3 days. Extraction was completed under general anaesthesia. The larva was dead and the posterior spiracular region, damaged during the procedure, was absent. The larva measured about 15 mm and was identified as a second stage Dermatobia hominis larva based on its pear shape and numerous stout spines (Fig. 1C). A purulent discharge leaked from the extraction site. A sample was collected for bacterial analysis, which demonstrated an oxacillin-sensitive Staphylococcus aureus. Systemic treatment with oxacillin and topical treatment with a steroid-antibiotic ointment allowed the patient to heal within a few days.

Figure 1. Patient 4. (A) Significant inflammatory upper eyelid swelling mimicking preseptal eyelid cellulitis. The respiratory pore is just above the ciliary line. (B) Lacking oxygen, the larva posterior end protrudes shortly after a petroleum ointment has been applied on the respiratory pore. The two posterior spiracles (which allow the larva to breathe) are just contiguous to the two lateral flash reflections. A stamen-like crown surrounds the spiracles. (C) Pear-shaped second stage larva, the anterior end of which is covered with rosethorn-like stout spines. The posterior end tip, damaged during surgery, is absent.

Patient 5

A 3-year-old boy presented with a right palpebral oedema that had appeared 2 days earlier. The patient felt comfortable and only complained of pruritus. Ocular examination disclosed a follicular conjunctivitis and a superior fornix aperture surrounded by a chemosis, through which the end of the larva protruded (Fig. 2). Under slit-lamp magnification, the two posterior spiracles were identified. Extraction under local anaesthesia was incomplete and the anterior end of the larva had to be removed under general anaesthesia. The anterior end had numerous rows of hooklets, which, together with slit-lamp examination findings, allowed identification of Dermatobia hominis. However, the larva was too damaged to allow larval stage identification.

Figure 2. Patient 5. Conjunctival myiasis. The posterior end of the larva protrudes through the respiratory pore into the right superior fornix.

Rifamycine eyedrops and systemic amoxicillin/clavulinate were prescribed for 8 days and the patient recovered promptly.

Patient 6

A patient aged 13 years presented with right inner canthus smarting that had lasted for 3 days and the intermittent sensation of movements that had begun the day before presentation.

Macroscopic examination on the right side showed a mild conjunctival hyperhaemia, an inflamed and swollen semilunar fold and a tiny, moving, caruncular lesion.

On slit-lamp examination, the moving lesion was identified as the posterior end of a Dermatobia hominis larva, intermittently protruding through a caruncular aperture, based on recognition of a stamen-like crown centred by the two posterior spiracles (Fig. 3A) and on previous clinical experience. The rest of the examination was normal.

Figure 3. Patient 6. (A) Caruncular myiasis. The larva's posterior end spontaneously protrudes through the respiratory pore. The two posterior spiracles are located at the very tip. A stamen-like crown surrounds the posterior spiracles. (B). Post-operative photograph. The larva is stage 2, measures 10 mm and is pear-shaped with an enlarged anterior end covered with spines and an elongated posterior end. A polyglactin suture is passed through the middle part of the larva.

The patient was immediately operated on using an operating microscope in the minor interventions room. After topical anaesthesia, an operative field was laid under sterile conditions and a lid speculum was placed. Vitamin A ointment was spread temporal to the lesion and fashioned as a dyke in order to create a pool filled with xylocaine over the caruncula. The larva came out through the pore over the xylocaine layer to breathe and was firmly grasped with a Bonn toothed forceps. After pulling the posterior end upward, an 8.0 polyglactin thread was passed through the larva as far towards the anterior end as possible. The larva partially collapsed – with a brownish fluid oozing through the piercing points – and immediately stopped moving. The thread was used to pull the larva upwards while xylocaine was injected deep into the caruncula, causing a pressure increase. This manoeuvre dislodged the larva. It measured 10 mm and was formally identified as a stage 2 Dermatobia hominis larva (Fig. 3B), based on its size, its pear shape and the presence of numerous backward-directed spines covering the anterior and middle segments. The patient was treated with steroid-antibiotic eyedrops for 1 week and recovered fully.

Discussion

Dermatobia hominis is the most common cause of cutaneous myiasis in humans (Wilhelmus 1986). It may cause external ophthalmomyiasis (McGavin 1996), with eyelid (Pradinaud & Rivierez 1968; Cordero- Moreno 1973; Savino et al. 1986; Wilhelmus 1986; Orihel & Ash 1995; Bangsgaard et al. 2000; Emborsky & Faden 2002) or conjunctival (Cordero-Moreno 1973; Goodman et al. 2000) involvement. This affection is rather infrequent and represents less than 5% of all sites (Cordero-Moreno 1973).

Myiasis caused by Dermatobia hominis is found in Central and South America (Cordero-Moreno 1973), from Mexico to Argentina and Chile (White 1996b), mainly in tropical areas (Wilhelmus 1986) and at the edge of tropical forests (White 1996b). Dermatobia hominis attacks many different hosts, mainly cattle and fowl (White 1996b). Humans are not the preferred hosts, as human skin is apparently an inadequate source of nutrition for larvae and often prevents maturation beyond the first stage (Bangsgaard et al. 2000). A history of travel or residency in a tropical American country is required to make the diagnosis. Nine cases of external ocular myiasis caused by Dermatobia hominis were retrieved from the peer-reviewed English literature (using Pubmed) and textbooks (Table 2). Each case originated in tropical American countries except for the case reported by Emborsky & Faden (2002), which involved a boy aged 4 years, presumably from New York, USA or from the New York vicinity. It is puzzling that such a case could have occurred without the patient travelling in Central or South America. Furthermore, the 7 mm extracted larva, a photograph of which appears in the article, lacked the usual anatomical features of a presumably second stage Dermatobia hominis, in that the narrow end was not elongated and the spicules were very small instead of being stout. In this case, the diagnosis is, in our opinion, dubious.

Table 2.  Data concerning cases of external ophthalmomyiasis caused by Dermatobia hominis published in the peer-reviewed English language literature and in textbooks.

Authors
Year	Patient age	Patient sex	Myasis origin	Number and location of larvae	Symptoms	Lesion	Discharge	Ointment	Antibiotic treatment	Larva removal	Larval stage
y =years; m = months.
Cordero-Moreno 1973
1973	2 y	Female	Venezuela	1 (left upper eyelid)	?	Respiratory pore on inner palpebral border	?	?	?	?	Probably 2
Cordero-Moreno 1973
1973	?	?	Venezuela	1 (right eye)	?	Larva in upper fornix, near inner canthus	?	?	?	?	?
Cordero-Moreno 1973
1973	10 y	Male	Venezuela	1 (upper eyelid) (side not mentioned)	Movement sensation at the beginning	Chalazion-like lesion	?	?	?	Surgical (type of anaesthesia not mentioned)	? (partially disintegrated larva)
Wilhemus 1986
1986	?	Female	Colombia	1 (right upper eyelid)	Pruritis; wriggling sensations under skin	Cellulitis pore near internal canthus	Serosanguinous	Petroleum	No	Surgical local anaesthesia)	3
Savino et al. 1986
1986	6 y	Female	Venezuela	1 (left upper eyelid)	Moderate pain at palpation	Dome-shaped lesion Pore in free margin	Serosanguinous	No	No	Surgical local anaesthesia)	?
Orihel & Ash 1995
1995	?	?	?	1 (left upper eyelid)	?	Oedema Pore near internal canthus	?	?	?	?	?
Goodman et al. 2000
2000	5 y	Male	Honduras	1 (right inferior conjunctival fornix)	?	Chemosis Pore in inferior fornix	?	?	?	Surgical (general anaesthesia)	3
Bangsgaard et al. 2000
2000	46 y	Male	Brazil	1 (left upper eyelid)	Pruritis	Dome-shaped lesion Pore in free margin	Serosanguinous	Pilocarpine	Yes (to treat pyogenic arm infection)	Surgical	3
Emborsky & Faden 2002
2002	4 y	Male	Vicinity of New York, USA	1 (left upper eyelid)	None	Cellulitis Aperture in internal canthus	Clear discharge	No	Yes (oxacillin)	Surgical (local anaesthesia)	? (dubious larva aspect)

The female Dermatobia hominis only mates once (White 1996a). Subsequently, she deposits her eggs (Service 2000) onto a haematophagous insect, usually a mosquito (White 1996b) or a fly (Service 2000), which serves as a vector to convey the eggs to a new host (White 1996a). This transmission system is called ‘phoresis’ (White 1996a). It takes 4–9 days for the attached eggs to mature into first stage larvae (Service 2000). The eggs, apparently stimulated by a temperature increase (Bangsgaard et al. 2000), do not hatch until the carrier insect comes into contact with a warm-blooded animal (Wilhelmus 1986; Service 2000). The larvae (i.e. the maggots) drop from the eggs on to the host skin (Service 2000). Usually one (Cordero-Moreno 1973) but sometimes several larvae penetrate the skin (Service 2000) at the site of a bite on an haematophagous vector or through hair follicles (Bangsgaard et al. 2000) or they painlessly chew through intact skin in 5–10 min (White 1996b). Penetration may take longer and may be delayed, as the larvae are able to survive for weeks on the skin surface (Bangsgaard et al. 2000). Burrowed into the subcutaneous tissue, the 1 mm first stage larva moults twice in 4−18 weeks (Service 2000), developing into a second and then a mature third stage larva (Wilhelmus 1986), about to drop from the host and bury itself in the ground for a 5–12-week pupation that eventually produces an adult fly (Wilhelmus 1986). Third stage larvae measure 18–25 mm, are more globular and bear smaller spines on their anterior end than second stage larvae.

During their development, each larva produces a furuncular lesion called a ‘warble’ (Wilhelmus 1986). At the centre of this lesion lies a skin hole through which the larva intermittently sticks out its posterior end (Cordero-Moreno 1973) in order to breathe through the two posterior spiracles (Wilhelmus 1986). During these breathing manoeuvres, movements may be felt within the lesion. The warble usually oozes a serous fluid containing the dark faeces of the larva (White 1996a). Purulent fluid is rare as the larva faeces seem to involve a bacteriostatic action that prevents overgrowth of pyogenic bacteria (White 1996b) and possible secondary infection.

If eyelids are involved, a swelling is always present. The lesion may be misdiagnosed as a chalazion if the swelling is mild or as preseptal cellulitis if the swelling is significant (Wilhelmus 1986). A wriggling sensation within the lesion and identification of the respiratory pore are both helpful in making the diagnosis (Cordero-Moreno 1973). In cases of conjunctival penetration in the cul-de-sac and caruncula, the respiratory pore may be more difficult to visualize (Cordero-Moreno 1973).

If the larva dies within the cavity, the lesion may be very similar to a chalazion (Cordero-Moreno 1973; Wilhelmus 1986). Nevertheless, when it is opened, an entire or disintegrated larva is found, which allows the diagnosis to be rectified (Cordero-Moreno 1973).

A pretragial lymphadenopathy is sometimes found on the involved side, as in patient 4.

Lesions can be multiple, as in the case of patient 3, who had three eyelid larvae on the same side.

Patients usually complain of discomfort related to the perception of movement within the lesion, of pruritus and sometimes of pain (Orihel & Ash 1995), especially when the lesion is palpated (Bangsgaard et al. 2000). Ocular pruritus may be related to the accumulation of larval excretion in the inferior fornix (Bangsgaard et al. 2000), especially if the respiratory pore is on the free palpebral margin or in a conjunctival fornix. In such cases, an allergic reaction may be seen (Bangsgaard et al. 2000).

During slit-lamp examination, the conjunctival fornices must be carefully inspected (Cordero-Moreno 1973), as the larva is able to develop under the conjunctiva (Cordero-Moreno 1973). In patient 5, the respiratory pore emerged in the superior conjunctival fornix, which presumably prevented the larva from absorbing the necessary oxygen for its subsistence and led to its death. This, incidentally, explains the absence of any feeling of movement within the lesion for patient 5.

Contrary to the suggestions put forward by Savino et al. (1986), the respiratory pore is frequently found on the palpebral free margin (patients 1, 7 and 9 and cases reported by Savino et al. (1986) and Bangsgaard et al. (2000)), among the row of cilia, possibly because the larva may easily penetrate through the skin via an eyelash follicle.

Extraction of the larva is greatly facilitated by smothering it through obstruction of the respiratory pore. A thick layer of ointment, such as petroleum or paraffin oil, may be used. The larva usually comes out through the pore to breathe, which allows it to be grasped with forceps (Pradinaud & Rivierez 1968; Cordero-Moreno 1973). However, the deeply embedded, enlarged anterior end may make removal difficult. Furthermore, the numerous backward-directed (Wilhelmus 1986), rosethorn-like spines covering the anterior end act as hooks and oppose the dislodging of the larva. Some authors suggest instilling lidocaine or pilocarpine in the funnel in order to paralyse the larva if stopping its oxygen supply proves insufficient (Bangsgaard et al. 2000). In conjunctival myiasis, some authors report inactivation of the larva by instilling cocaine through the respiratory pore to facilitate extraction (Cordero-Moreno 1973). Surgical extraction under local anaesthesia is another valid option (Bangsgaard et al. 2000). In recent years, the first step procedure used in our institution for eyelid myiasis has been instillation of ivermectine eyedrops (10 mg/ml) into the respiratory pore followed by surgical extraction under local anaesthesia, consisting of a simple enlargement of the skin opening followed by removal with toothed forceps. In most cases the entire larva can be removed, as its thick, resistant, chitinous wall is able to resist strong traction (Cordero-Moreno 1973). In rare cases, removal has had to be performed under general anaesthesia, either in young children or after a failed extraction under local anaesthesia. In some cases, using a suture passed through the larva to help extraction might be advisable. This procedure kills the larva and reduces its volume (which might prevent enlarging the respiratory pore) and may facilitate traction and dissection if this proves necessary. Failed primary extraction with possible larval death and retention and secondary infection (Bangsgaard et al. 2000) might thus be prevented.

The Dermatobia hominis larva is primarily identified by its size, shape and surface features (especially spines and spiracles). Scanning electron microscopy may be useful in cases where the extracted larva is incomplete or damaged (Bangsgaard et al. 2000).

Traditional methods of extraction used by the French Guiana Creole community rest on the same principle of depriving the larva from its oxygen supply. These methods include, for example, using a blend of pitches and tobacco (Pradinaud & Rivierez 1968) or candle wax. Patients sometimes present only when the lesion location is tricky (eyelid; external auditory canal) or in the case of a failed traditional remedy, when the larva is retained or there is secondary infection (Pradinaud & Rivierez 1968).

White (1996b) asserts that antibiotic prophylaxis is advisable before attempting larva removal. For our part, we believe, like Desruelles et al. (1999), that antibiotic treatment is not warranted, except in cases where a lesion oozes a purulent discharge. In such cases, a pus sample should be collected for bacteriological identification and antibiogram.

Prophylaxis involves taking the usual measures against mosquitoes bites, by using insect repellent (Wilhelmus 1986; Desruelles et al. 1999) and a mosquito net when sleeping in a hammock at night.

Control with insecticides has been implemented in Brazil (White 1996a), while in Curaçao, Dermatobia hominis males sterilized by radiation were used to make the females sterile, which allowed for the eradication of Dermatobia hominis within 2 years (White 1996a).

Cutaneous furuncular myiasis is represented by Dermatobia hominis in Central and South America and by Cordylobia anthropophaga in Africa (Desruelles et al. 1999), south of the Sahara (White 1996b). Cordylobia anthropophaga oviposits on the ground (White 1996a). Clothing laid on the ground to dry may be affected. Larvae can penetrate unbroken skin (White 1996a). Contamination can result from lying on the ground or from clothing (White 1996a). Any part of the body may be affected, whereas Dermatobia hominis tends to affect only exposed body parts (White 1996a; Desruelles et al. 1999), although the larvae may penetrate clothing (White 1996a, 1996b). To the best of our knowledge, no case of external ocular myiasis caused by Cordylobia anthropophaga has ever been reported.

Conjunctival myiasis may involve the sheep botfly, Oestrus ovis (Mazzeo et al. 1987). Patients usually complain of a foreign body sensation (Mazzeo et al. 1987) and slit-lamp examination demonstrates conjunctivitis and may detect larvae (Mazzeo et al. 1987).

Hypoderma, found in Asia and Africa (Bangsgaard et al. 2000), may cause palpebral myiasis (Bal & Amitava 2000), as well as internal ocular myiasis (Fitzgerald & Rubin 1974).

The Cuterebra larva, native to North America, has also been reported in cases involving the eyelid (Rodrigues et al. 1976) or internal ocular myiasis (Custis et al. 1983). As far as we know, Dermatobia hominis has never been reported as a cause of internal ocular myiasis.

Myiasis caused by Dermatobia hominis mostly concerns doctors who work in the American tropical zone, between 18 degrees south and 25 degrees north (Wilhelmus 1986). However, the long larval life (up to 18 weeks), combined with easy air travel to and from the American tropical zone (Cordero-Moreno 1973; Service 2000), means that infected persons may present in any part of the world (Cordero-Moreno 1973; Desruelles et al. 1999; Bangsgaard et al. 2000).

Acknowledgement

The cost of reproducing colour illustrations was supported by l'ADOG (Association pour le Développement de l'Ophtalmologie en Guyane).