Our patient presented with the association of PG and adenocarcinoma of the rectosigmoid junction, which is rarely described. However, this case report is distinguished by the appearance of PG lesions before the clinical expression of cancer and their disappearance after tumor excision.
PG is a dermatosis that can occur in isolation. However, in more than 75% of cases, it is associated with an underlying pathology, including inflammatory bowel disease; endocrine, hematological, and rheumatological disorders; and nonhematological neoplasia [
4]. Few cases of PG related to colorectal cancer have been described. Those that have been reported have been either a single skin lesion located on the trunk or the upper limbs for some [
7‐
9] or ulcerative colitis that was associated with rectal cancer, which made a direct causal relationship between rectal cancer and PG questionable [
10]. Cailhol
et al. [
11] described a cutaneous (lower limb location) and extracutaneous (lung location) PG with secondary appearance during neoplastic progression after the first line of chemotherapy. Our patient’s case is distinguished from previous cases by its premature appearance before the diagnosis of adenocarcinoma without any other associated disease, the multiple lesions of PG, and their location in the lower limbs. The appearance of the first PG lesions 6 months before the cancer diagnosis, their recovery without reappearance of new lesions 4 months after resection of the tumor, and the absence of any other underlying condition made us consider its paraneoplastic character. Our patient’s case is similar to the one described by You
et al. [
5], which makes our patient’s case, to our knowledge, the second observation of paraneoplastic PG on colorectal adenocarcinoma appearing before any clinical manifestation of the cancer. From a clinical point of view, PG is of variable expression, but there are four major forms: ulcerous, pustular, bullous, and vegetative [
12]. However, of all of these forms, the paraneoplastic form, which is the most frequent, is ulcerous, as was the case in our patient [
11]. The pathophysiological mechanism of paraneoplastic PG remains debatable and is not yet well understood. Some suggest a hapten common to the tumor and the dermis responsible for a cross-immune reaction against the dermis [
13,
14]. For Adachi
et al. [
15], this is an abnormal immune surveillance that can result in neutrophilic dysfunction, chemotaxis defects, or hyperreactivity.
The prognosis of PG is marked by a long-term evolution that remains unpredictable [
1]. The relapse rate is high for commonly used drugs: 70% for prednisolone and 66% for cyclosporine [
16]. Some studies have even reported a mortality rate of more than 30% [
17]. In the case of paraneoplastic PG, the evolution seems to depend on the tumor, as suggested by our observation marked by a recurrence of PG lesions and their healing after the excision of the tumor, reinforcing the hypothesis of a cross-immune reaction against the dermis. This evolutionary dependence of PG on the tumor provides an opportunity for surveillance as an element of recurrence or evolution of the tumor in the event of reappearance. Furthermore, although few cases of paraneoplastic PG in colon cancer have been reported, the possibility of early appearance even before the clinical manifestation of PG may allow early diagnosis of colorectal cancer. Colonoscopy for neoplastic lesions should therefore be part of the etiological assessment in cases of PG in a patient at risk.