Immunostaining for Treponema pallidum: Caution in its Evaluation

To the Editor: We would like to report some examples that have made us evaluate the immunostaining for Treponema pallidum with extreme caution, especially out of an appropriate clinical context. Most laboratories use the antibody T. pallidum (Biocare), which has given good results in the diagnosis of mucocutaneous syphilitic lesions.1-6 With this antibody, we saw immunostaining of what seemed to be bacilli in cases in which acid alcohol resistant bacilli (AARB) had previously been demonstrated. We have observed this phenomenon in several cases. Figure 1 (top left), for instance, shows a biopsy from the larynx of an 82-year-old man who presented nodules in the larynx. Luetic serology was negative. The biopsy from the larynx showed necrotizing granulomas (Fig. 1, top left) with AARB which were easily found (Fig. 2, top left). After immunostaining for T. pallidum, some positive structures which were very similar to the AARB were found in the biopsy (Fig. 2, top right).FIGURE 1: Several examples in which we have found bacilli-like structures when immunostaining for T. pallidum. Top left: bronchial biopsy in which many granulomas can easily be seen. Top right: bronchial biopsy with a necrotizing granuloma with caseum. Bottom: cutaneous lesions with a sporothricoid pattern (left) in a case of M. marinum infection; the biopsy demonstrated a granulomatous dermatitis, with central abscesification (right).FIGURE 2: First row: bronchial biopsy of the granulomatous bronchitis: AARB were easily found in the stain with Ziehl-Neelsen (left). The immunostaining for T. pallidum showed some positive structures which seemed to be bacilli (right). Second row: biopsies from a patient with “swimming-pool” granulomas. The Ziehl-Neelsen staining showed some bacilli (left). The immunostaining for T. pallidum showed similar bacillar structures which expressed the marker (right). Third row: high power of the caseum of a bronchial biopsy (left). AARB were mainly found in the necrotized area which is remarked in the square and which is shown in the middle after staining with Ziehl-Neelsen. The picture on the right shows an immunostaining for T. pallidum under oil immersion which is very similar to the structures highlighted in the Ziehl-Neelsen staining. Fourth row: this is a typical example of chronic gastritis due to H. pylori. The microorganism is highlighted with Giemsa stain (left) but also shows positivity with the immunohistochemical stain for T. pallidum (right).We have also observed the same phenomenon in cases of atypical mycobacteria infection. Figure 1 (bottom), for instance, shows a cutaneous infection by Mycobacterium marinum in a 50-year-old woman (swimming-pool granuloma). Clinically, the lesion shows a sporothricoid appearance (Fig. 1 bottom, left). The biopsy demonstrated a granulomatous dermatitis (Fig. 1, bottom right), with a central abscess in the granuloma. Under oil immersion examination, some AARB were seen among the neutrophils with Ziehl-Neelsen staining (Fig. 2, second row, left). The woman had an aquarium at home, and the lesion easily healed with a combination of trimethoprim-sulfamethoxazole and minocycline. Although no serologic investigation was made in this case for syphilis, we retrospectively performed an immunohistochemical study for T. pallidum. As Figure 2 (second row, right) shows, the antibody highlighted some positive long structures (nonspirochetae) which were very similar to the AARB. Figure 1 shows (top right) another case which is more typical of what a general pathologist sees in cases of tuberculosis. A 51-year-old woman presented with hemoptoic sputum and some bronchial exophytic lesions which were biopsied. The biopsy showed a necrotic granuloma with caseum (Fig. 1, top right). In the central areas of the granuloma, several AARB were found with Ziehl-Neelsen staining (Fig. 2, third row, left). Figure 2 (third row, right) shows the immunostaining for T. pallidum in this case. Given the difficulties in distinguishing these positive structures from bacilli (if they are not real bacilli), we suggest caution when evaluating a positive immunostaining for T. pallidum, especially out of a clinical context of syphilis. Some cases of nodal inflammatory pseudotumor caused by luetic infection have been reported,7 but in them, the microorganisms presented showed a typical corkscrew morphology, which could be a crucial requisite before interpreting any immunostaining for T. pallidum as positive. This is even more important, because it is well known how inflammatory pseudotumor can be associated to infections by mycobacterium. Examples have been reported in the spleen,6,8 the lung,9 the lymph nodes,10 and the soft tissues.11 A spindle-cell tumor due to mycobacterium has also been described in the skin.12,13 Furthermore, the antibody for T. pallidum immunostains Helicobacter pylori, which typically has a corkscrew morphology (Fig. 2, bottom right). However, these two microorganisms can be easily differentiated utilizing a Giemsa stain; Helicobacter are positive with this stain (Fig. 2, bottom left). Being aware of this fact is crucial, in order not to over diagnose gastritis due to Helicobacter as syphilitic gastritis, which is uncommon but it can occur.14-17 Angel Fernandez-Flores, MD, PhD Service of Cellular Pathology, Clinica Ponferrada, Ponferrada, Spain