August 2003

A 4½-year-old girl was brought to the emergency department (ED) for evaluation of an injured ankle. The ankle turned out to be simply sprained. However, as is often the case, an “oh by the way” complaint becomes the most interesting finding. While in the ED, her parent wanted a rash evaluated. The rash was on the right thigh and had been present for about a week. It began as itchy red bumps, then transformed into an irregular linear pattern. The chief complaint she has about the rash is a sort of painful itching.

Her past medical history (PMHx) was positive for occasional allergic symptoms, including seasonal rhinitis, conjunctivitis and one episode of urticaria. Otherwise, her PMHx and family were unremarkable. No one else at home was sick or had a rash.

Examination revealed normal vital signs and a normal-appearing female child with a swollen ankle and a rash on the right thigh. The rash consisted of small erythematous papules mixed with slightly raised, serpiginous streaks (figures 1 and 2).

No lab tests were done.

What’s Your Diagnosis?

1. Common variable immunodeficiency
2. Transient hypogammaglobulinemia of infancy
3. Chronic granulomatous disease
4. X-linked agammaglobulinemia

Answer

As we are well into the summer, and people are out more in short pants and swimsuits on beaches and other outdoor locations, and in direct contact with the earth, this case seemed timely. The diagnosis made and treated was cutaneous larva migrans (CLM), commonly known as creeping eruption (B). In this country, this is the result of invasion of the skin by the infective larvae of the dog or cat hookworm, Ancylostoma braziliensis, or occasionally, Ancylostoma caninum. The life cycle of this worm begins by the passage of embryonated eggs through the stool of the dog or cat. If the ground environment is favorable (warm and moist), the rhabditiform larvae emerge after one to two days. About a week later the larvae become infective (filariform) and can penetrate the skin of the host whose skin comes in direct contact with these larvae. This usually occurs in areas under houses, sandboxes, beaches — virtually anywhere dogs and cats defecate.

At the point of penetration, the skin develops an erythematous papule. As the larva moves through the superficial layers of the skin, presumably seeking access to adequate lymphatic vessels and/or the bloodstream, an irregular, linear, raised erythematous streak is formed. This gives it a serpiginous or snake-like appearance. This process usually provokes a fairly intense local inflammatory response with variable degrees of itching. The most common complication of this condition is local skin irritation or secondary infection from scratching. The life cycle of the worm ends at this stage since the larvae are unable to penetrate to the deeper layers of the skin to the bloodstream. Therefore, it is a self-limiting infection, requiring only time to resolve. However, this may require weeks to months, and most patients want it gone yesterday. Various methods of therapy have been used with variable results. Cryotherapy has been used for about 80 years, but the success rate is highly variable. Either the type of freezing therapy used was not applied long enough or it missed the larvae altogether. A common mistake is to freeze the visible lesion when the parasite is actually ahead of the edge of the serpiginous lesion. Most experts have gone to recommending topical or oral thiabendazole. An interesting prospective study from Belgium concluded that a single dose of ivermectin (Stromectol, Merck) was safe and effective. Albendazole (Albenza, GlaxoSmithKline) has also been successfully used. As usual, I recommend whatever is in John Nelson’s Pocketbook of Pediatric Antimicrobial Therapy.

When you read about CLM, a certain amount of confusion may exist. It is occasionally said that CLM larvae may rarely migrate to the lung through the bloodstream, resulting in transient pulmonary infiltrates with associated eosinophilia (Loeffler’s syndrome). Indeed, Loeffler’s syndrome has been well documented to occur with CLM, but the exact pathophysiology remains questionable. It has been postulated that this pulmonary complication results from a generalized, allergic reaction to the migrating worms. It is also stated that under heavy infestations, some larvae actually migrate to the lung, where they ultimately die. However, when this is mentioned, even in modern texts and papers, a single reference is usually quoted where larvae were actually recovered from multiple sputum samples of a single patient with CLM in New Orleans in 1952. The problem with this is that there are weaknesses of that paper that make me question its conclusion. Additionally, one subsequent study where 381 sputum samples from 76 patients with CLM (26 with Loeffler’s syndrome) failed to find a single larva.

If the larvae of the CLM hookworm can have a pulmonary phase, it must be very rare, and likely to be associated with a heavy infestation. I would further speculate that the subcutaneous tissue of those patients would probably be abnormal, either naturally or by damage done by a heavy worm burden, since these aberrant hookworms do not possess the enzymes needed to penetrate completely through the epidermis under normal circumstances. In fact, reference #4 by Peter Hotez, MD, PhD, one of the preeminent experts on parasitic diseases in this country, is an excellent review of hookworm disease in general. If you know of a good reference that will scientifically answer the question of systemic migration of the causative worm of CLM, please let me know, and we will add it to the next issue. Till then, I remain skeptical about the true existence of a “pulmonary phase” of this worm.

Other choices

Shingles is, of course, the rash resulting from reactivation of varicella zoster virus infection. It may begin the same as CLM with a stinging sensation, followed by the appearance of a painful, itchy rash. However, the lesions are usually a cluster of discrete vesicles arranged in a dermatomal pattern (figure 3). While CLM can have vesicles associated with it, the linear aspect of CLM is not likely to be seen with shingles.

Contact dermatitis due to poison ivy may have a linear, vesicular pruritic rash. It may be confused with shingles, but does not really resemble the more discrete lesions of CLM. Poison ivy dermatitis will likely have patches of diffuse erythema interspersed among the areas of streaking (figure 4).

Papular urticaria is a localized, delayed hypersensitivity reaction of an insect bite, that usually occurs in young children. It most commonly results from the bite of fleas, mosquitoes and mites. The bite produces discrete, papular lesions on an erythematous base of variable sizes (figure 5). This is what I usually think of when I hear from a parent that their child is “allergic” to insect bites. There are some similarities between the initial lesions of CLM at the entry sites and papular urticaria, but papular urticaria does not progress to producing linear streaks.

At the risk of offending sensitive dog and cat owners everywhere (or as some would have it, dog or cat caregivers or foster parents), I feel I must make a comment on this topic. And don’t get me wrong. My family and I have a cat, and have had dogs and other cats in the past, and loved them all. Our current relationship with our cat is really more like a partnership, as he is independent and, of course, highly intelligent. If he only had a thumb, he probably would not need us at all. But as it is, he is strictly an indoor cat, and relies on us to open his food and clean his litter box. Again, no thumbs. If he were allowed to go outside (which he seems happy not to), before long, he and the yard would be full of fleas. This is a lesson many cat foster parents have to learn the hard way. Cats are flea magnets. This happened to us about 21 years ago, when we lived in Hawaii. Before we realized it, walking across our living room carpet stirred up a cloud of fleas, much like walking across a cow pasture here in Texas stirs up a cloud of grasshoppers. Normally, these fleas are nothing more than a nuisance, but under certain circumstances, they can spread diseases, like plague.

Fleas are one thing, but dog and/or cat feces in public places, or anywhere children play, is something else. In addition to being a stinking, disgusting mess that your toddler may get into, it can also be the source for clinically significant infestations with various hookworms causing CLM, as well as worms associated with the more problematic visceral larva migrans (VLM), such as Toxocara canis and T. cati (toxocariasis). VLM may result in liver and spleen enlargement, asthma exacerbation, neuropsychiatric problems and potential damage to the eye resulting in vision loss. And we are all painfully aware of the risk in having children (and adults) in the same areas as unleashed dogs. The first time I got bitten by a dog, I was 12 years old, simply walking in a public place, and the first thing the dog owner said of his unleashed dog, with a surprised voice, was “he’s never done that before.” Knowing that did not make me feel any better, or numb the sting of the tetanus shot or stitches I received. Nowadays, an attorney would be waiting outside the clinic door for me and my parents to offer his or her “assistance.” So, my appeal to dog and cat (mostly dog) managers, foster parents, partners, mentors, owners or whatever you are considered: please clean up after them in public places and keep them under control with a leash. To do less can be a very expensive lesson to learn, as those responsible for the damage to the child in figure 6 found out the hard way.

For more information:

• Van den Enden E, et al. Treatment of cutaneous larva migrans. N Engl J Med. 1998;339:1246-1247.
• Muhleisen JP. Demonstration of pulmonary migration of the causative organism of creeping eruption. Ann Intern Med. 1953;38:595-600.
• Wright DO, Gold EM. Loeffler’s syndrome associated with creeping eruption. Arch Intern Med. 1946;78:303-312.
• Hotez PJ. Hookworm disease in children. Pediatr Infect Dis J. 1989;8:516-520.