When is pyoderma gangrenosum not pyoderma gangrenosum?

By Sylvia Hsu, MD, November 1, 2019

In this month’s Clinical Applications column, Editorial Advisory Workgroup member Sylvia Hsu, MD, talks with Emanual Michael Maverakis, MD, about his recent JAMA Dermatology and JAAD articles, ‘Diagnostic criteria of ulcerative pyoderma gangrenosum’ and ‘Peristomal pyoderma gangrenosum: An exceedingly rare and overdiagnosed entity?’

Dr. Hsu: Please explain why pyoderma gangrenosum (PG) is not a diagnosis of exclusion.

Dr. Maverakis: No disease in medicine should be a diagnosis of exclusion. Diagnoses should always be based on a finite set of criteria. In the case of PG, a diagnosis of exclusion would require the physician to rule out all possible causes of an ulcer and subsequently test the patient for these. It would be very difficult to rule out every single possible cause of ulceration.

One of my favorite “PG cases” was a patient with horrible lower extremity ulcerations that came to us on infliximab. She was seen at the local dermatologic society and had been worked up extensively. Since her workup up to this point was entirely negative, she was given the diagnosis of PG. In the end, it turned out that she had end stage scurvy, NOT PG. This case was one of the reasons why I put together the team to establish the PG diagnostic criteria. These criteria are by no means perfect, but having PG as a diagnosis of exclusion was simply unacceptable.

Dr. Hsu: Should the clinician do an additional biopsy for tissue culture?

Dr. Maverakis: Special stains for infection and tissue culture should be considered in every case. Results from the H&E [hematoxylin and eosin] analysis can guide the clinician with regard to how extensively infectious etiologies should be investigated. If numerous PG criteria are met, the importance of a wound culture is less. There is some flexibility in this regard.

Dr. Hsu: If the biopsy of a leg ulcer does not show neutrophilic infiltrates, can it still be PG or treated as PG?

Dr. Maverakis: Unfortunately, PG ulcers go through stages and depending on the stage, a neutrophilic infiltrate can be absent. This was a concern when developing the diagnostic criteria. For example, my own group believes that T cells dominate the infiltrate during the earliest stages of the disease, when the PG papule first forms, literally in the initial hours after its appearance. Such early lesions are extremely hard to capture by biopsy because PG lesions rapidly evolve and by the time a PG patient presents to dermatology, ulcers are usually present. In fact, it would be very difficult to say that someone has PG prior to the appearance of ulcers. There are also reports of T cells being commonly present at the ulcer edge, although this region should also have many neutrophils, regardless of whether T cells are present.

However, there are more common scenarios in which a neutrophilic infiltrate will be absent. The cause of PG is still unknown, but one theory is that PG is an autoimmune disease. As is typical for many types of autoimmunity, lesions can spontaneously resolve, even as new ones appear. If a PG lesion is in the process of spontaneously resolving, then neutrophils may be variably present, even absent.

Possibly the most common reason for absence of neutrophils is that the patient was started on therapy. Treated patients may have ulcers that lack neutrophils. This is because it takes time for ulcers to heal after the aberrant immune response has been successfully suppressed with medications. This is a major problem in my clinic. Most PG patients who are referred to me for management have already been started on immunosuppression. Of course, the referring doctor should start immunosuppression, but it makes it difficult for my clinic to make a definitive diagnosis. We are relying on the referring doctor’s ability to make the diagnosis. Partially treated PG lesions not only lose their neutrophilic infiltrate, but they also lose some of the characteristic clinical features of PG. For example, the nearly pathognomonic undermined borders are usually quickly lost in patients on immunosuppression. The histology of PG lesions of patients on immunosuppression can be very difficult to interpret. At best, they might be helpful to rule out cancer, infection, and a few other things.

Thus, it might be helpful to re-biopsy the patient during an acute flare if the first biopsy was devoid of neutrophils.

Dr. Hsu: What therapies do you consider first-line treatment for PG?

Dr. Maverakis: It is important to note that randomized clinical trials are lacking. The results from the two best controlled trials to date (the infliximab trial and the STOP GAP trial) highlight the fact that PG is very difficult to treat.

I do like treating PG with biologics, but biologics take a long time to take effect and are often insufficient to achieve complete remission. PG can be a rapidly evolving disease, so a patient may continue to develop new debilitating lesions while waiting weeks for the biologic to become fully effective. Additionally, each new lesion dramatically increases the patient’s morbidity. Another obstacle is that it is very difficult to get insurance approval for biologics. If I am confident that the patient’s disease is stable without the appearance of many new ulcers, I might be less aggressive in my approach.

For these reasons, I will often treat PG patients with a multi-drug regimen, at least until the patient can be bridged to a biologic. I like this approach because some of the classic immunosuppressive medications are faster acting than the biologics. Also, we have extensive experience from treating transplant patients and I use very similar regimens for my PG patients. For healthy patients, it is not uncommon for me to treat with cyclosporine (and/or prednisone) together with mycophenolate mofetil, at least until I can get a biologic approved. After, I will overlap treatment with the biologic for a few weeks. I generally shy away from the anti-IL-17A biologics because PG patients might have underlying inflammatory bowel disease. I usually will choose a TNF inhibitor.

Dr. Hsu: Do you recommend that peristomal ulcers should be biopsied to make sure that they are really PG, and not just an ulcer with granulation tissue due to constant trauma in that area?

Dr. Maverakis: I do not biopsy peristomal ulcers unless I have an extremely high suspicion for PG or infection. PG is not a diagnosis that can be easily made with a biopsy. In my experience, peristomal lesions are often not PG. The fact, regardless of the location, is that PG is a rare entity. In contrast, the development of erosive lesions on occluded skin is a common phenomenon. Nearly all patients with a stoma will have an irritant dermatitis.

There are several clues that a peristomal ulcer is not PG. First, be wary of lesions that begin as erosions. PG ulcers typically develop as an inflammatory papule or pustule that then rapidly ulcerates. PG ulcers typically have undermined borders. If an ulcer began as an erosion, then that means the pathology started in the epidermis and then worsened. In contrast, PG lesions usually develop in the deep dermis, which is why the ulcerations have undermined borders; there is a rim of intact epidermis that has not yet died. The pathology is undermining the epidermis. Lastly, if the ulcer(s) recurs at sites of prior ulceration. PG ulcers try to avoid sites of prior ulceration. This is not a perfect rule and once an ulcer heals it has a tendency to break down again. However, actual new PG ulcers recurring at the exact same site is not common. They usually require some adnexal structures to be present and scars are devoid of these, so PG tends to avoid scars.

Dr. Hsu: Do you have a good way of biopsying peristomal ulcers, since the procedure can be quite messy, once the colostomy bag is removed?

Dr. Maverakis: Unfortunately, no. I tend to not biopsy these ulcers unless they have not responded to a comprehensive wound care regimen. Peristomal ulcers are usually very dirty and require an enormous amount of wound care to heal. I have to see these patients in clinic frequently until their ulcers at least begin to close. Afterwards, patients can take over much of the care. Often the ulcers are under-cared for when patients initially present to us because PG was a concern.

Of course, chronic ulcers that are present for a long time should be biopsied. In the cases of a chronic ulcer, an ulcer that failed to respond to a few weeks of wound care, or one that I felt PG was likely, I would do a shave biopsy from the ulcer edge.

Dr. Maverakis serves as associate professor of dermatology at the University of California Davis. Dr. Maverakis has no relevant financial or commercial interests to disclose. His articles appeared in JAMA Dermatology and JAAD.

Disclaimer: The views and opinions expressed in this article do not necessarily reflect those of DW.

Additional DermWorld Resources