Subcutaneous infection with Dirofilaria spp. nematode in human, France.

To the Editor: The article by Foissac et al. titled Subcutaneous infection with Dirofilaria immitis nematode in human, France () presents an interesting and challenging diagnostic dilemma. The paper described, but did not illustrate, the worm as having a strongly ridged external surface of the cuticle—a feature known not to exist on Dirofilaria immitis, the dog heartworm. However, molecular sequencing of the specimen demonstrated much closer similarity to D. immitis than to D. repens, the most common cause of zoonotic subcutaneous dirofilariasis infection in Europe.

Well-described morphologic features of parasites, including in tissue sections, have long been the standard for diagnosis. More recently, molecular diagnostics have helped in many of these difficult cases. However, in some cases, the morphology and molecular diagnosis are discordant. On the basis of the data in the article, the worm does not seem to represent D. repens. A more likely possibility is some other species for which no sequences are yet available for comparison. In such a worm, the regions sequenced must be similar to D. immitis, and distinct from D. repens, to achieve the observed results.

When one encounters a case such as this, where well-validated morphologic features (Figure) are contradictory to the molecular analysis, one must exercise caution in arriving at a final diagnosis. One disadvantage of morphologic and molecular diagnostics is an absence of information on poorly described and characterized pathogens or new pathogens that have yet to be identified. No good algorithm exists to resolve these conflicts other than to explore all possibilities. The diagnosis in the described case is probably best left as a Dirofilaria species of the Dirofilaria (Nochtiella) type, members of which exhibit marked cuticular ridging, and not D. (Dirofilaria) immitis type, members of which have as a feature an absence of cuticular ridging.

Figure

Cross-section of the filarial nematode seen in the subcutaneous nodule on the thigh of a woman in France. The features, as described in the original report (), include prominent, longitudinal ridging of the cuticle (arrows), 2 reproductive tubes, and the intestine (asterisk). Scale bar indicates 50 µm. Image courtesy of Jean-Philippe Dales.

In Response: We agree with Eberhard () that it is difficult to make a species identification when data derived from morphologic examinations do not correlate with those of molecular diagnostics. Errors may be the result of poor indexing of sequences deposited in sequence databases or inaccurate estimation of the degree of genomic polymorphisms within a species and between closely related species. On the other hand, a morphologic difference between 2 organisms, if it is associated with only 1 characteristic, should not be considered sufficient to classify them as 2 distinct species. Such a phenotypic variation may be the result of a single mutation or deletion. Consequently, the absence of a certain character does not exclude the categorization of an organism as a given species.

Molecular identification of the Dirofilaria spp. worm in our clinical case was made on the basis of 2 distinct sequences, each of which exhibited marked differences between D. immitis and D. repens (). The first sequence targeted internal transcribed spacer regions of ribosomal genes and revealed up to 100% homology with D. immitis sequences from GenBank, whereas a maximum homology of 80% was observed with D. repens sequences from GenBank. The second sequence targeted the cytochrome oxidase 1 gene and showed 100% homology with D. immitis, whereas <90% homology was observed for D. repens. For both analyzed targets, GenBank contained several sequences for D. immitis and D. repens that were deposited by various investigators, and all sequences yielded consistent results. Therefore, there is no basis to suggest that the sequences deposited in GenBank were incorrect.

Nevertheless, we agree that an alternate hypothesis is possible. The worm reported in our article could conceivably belong to a species that differs slightly from both D. immitis and D. repens, displaying morphologic similarities with D. repens but being more closely associated with D. immitis at the genomic level.