Sister Mary Joseph’s nodule: a clue to finding pancreatic cancer in a patient previously affected by gastric cancer

ARTICLE

Auteur(s) : Toru Inadomi

Department of Dermatology, 180 Nishi-Araijuku, Kawaguchi, Saitama 333-0833 Japan

accepté le 4 Janvier 2005

Case report

Case: 82-year-old male

A small nodule at the umbilicus of the patient appeared in autumn 2002, and the patient visited our Department in December.

A 2 cm skin colored nodule with a black crust was observed on the navel wall. The tumor had a wide base without any constriction ( (figure 1) ). A part of the tumor was taken for histological examination.

Histopathological findings showed hyperkeratosis and acanthosis of the epidermis without any atypia. In the dermis, scattered duct structures were seen ( (figure 2) ) which consisted of one to two layers of endothelial cells, showing a high degree of atypia with large amounts of chromatin.

Laboratory findings were within the normal range, including tumor markers such as CA19-9 and CEA.

Based on these clinical and histological findings, we diagnosed the case as SMJN, although its origin was undecided.

At that time, we performed several immunohistochemical stainings to examine the possibility that the earlier gastric cancer could be the origin of the SMJN. The results are summarized in table 1( Table 1 ). CA19-9 staining was negative in the gastric cancer but strongly positive in SMJN ( (figure 3) ). p53 staining was negative in the gastric cancer but strongly positive in SMJN. PCNA staining was weakly positive in the basal layer of the gastric cancer but strongly positive in SMJN. Based on apparent differences in staining observed between these two specimens, we concluded that the first gastric cancer could not be the origin of the SMJN and we continued the search for responsible lesions, finding a small space occupying lesion (SOL) in the pancreas by magnetic resonance imaging (MRI) in January, 2003. The SOL was very small in January and making a diagnosis was difficult. However, his CA19-9 level began to increase from the beginning of February and the tumor had become apparent in the corpus of the pancreas ( (figure 4) ), and a diagnosis of pancreatic cancer was made. He died in March. An autopsy was not allowed.
Table 1 Results of immunohistochemical staining

Discussion

Considering its characteristic clinical features, it would seem to be easy to diagnose the disease. However, it is sometimes difficult to identify the lesion of origin [3], since there are several pathways for dissemination. Almost 20-30% of the cases of SMJN have not had their origins identified [2, 8].

Generally speaking, the stomach is the most common lesion of origin in many countries [1]. It seems that the pancreas is not an uncommon site of origin: Majmudar et al. reported 23 out of 259 cases of SMJN were derived from pancreatic cancers [4] and Ishizawa et al. reported 15 out of 80 SMJNs had the pancreas as the site of origin [1]. In our case, we were inclined to speculate that the earlier gastric cancer, which had been removed prior to the patient’s first visit to our clinic, was the lesion of origin. However, it was difficult to believe that such an earlier gastric cancer could cause a metastatic skin lesion like a SMJN.

Several types of immunohistochemical staining were performed to examine different characteristics between cells of the previously existing gastric cancer and that of SMJN. Only infiltrating cells in the gastric cancer were positive for bcl-2, which regulates apoptosis [9], but both types of tumor cells were negative. CA19-9 staining, which is generally positive in pancreatic cancers, was negative in this gastric cancer and was strongly positive in the SMJN. The p53 tumor suppressor gene [9] was negative in the gastric cancer but was strongly positive in the SMJN. PCNA, which is known as a cell proliferation maker, was weakly positive in the basal layer of the gastric cancer but was strongly positive in the SMJN.

These clearly different reactivities between the specimens from the gastric cancer and the skin lesion suggested that the gastric cancer could not be the primary site, although circumstances, such as the incidence of sun exposure, could have had some influence on cell reactivity [10-12]. Chen et al. pointed out that, being different from other cutaneous metastatic lesions, histological sampling from SMJN often helps to determine the origin of the intra-abdominal primary malignancy [13].

The nodule together with pleural internal malignancies has rarely been seen. Besides our case, Benson et al. reported a case of SMJN which was affected with both an adenocarcinoma of the ascending colon and a carcinoma of the prostate [3]. The nodule’s origin was believed to be the former due to its negative staining for prostate specific antigen.

Our case demonstrates that a detailed examination is sometimes required to determine the lesion responsible for the skin manifestation. Since the nodule appears prior to the diagnosis of the original lesion in more than two thirds of cases of SMJN [1], dermatologists have a strong responsibility to make a correct diagnosis, which includes identification of the primary lesion.

Acknowledgement

References

2 Ive FA. The umbilical, perianal and genital regions. In: Champion RH, Burton JL, Burns DA, Breathnach SM, eds. Textbook of Dermatology 6^th. Edn. Oxford: Blackwell Science Ltd., 1998: 3163-5.

3 Benson JR, Singh S, Thomas JM. Sister Mary Joseph’s nodule: a case report and review. Eur J Surg Oncol 1997; 23: 451-4.

4 Majmudar B, Wiskind AK, Croft BN, Dudley AG, et al. The sister (Mary) Joseph Nodule: Its significance in gynecology. Gynecol Oncol 1991; 40: 152-9.

5 Schneider V, Smyczek B. Sister Mary Joseph’ syndrome. Acta Cytologica 1990; 34: 555-8.

6 Skellchock LE, Goltz RW RW. Umbilical nodule. Acta Dermatol 1992; 128: 547-52.

7 Michiwa Y, Kamata T, Hayashi H, Hayashi Y, Minatoya G, Onishi I, Takeda T, Kannno M, Ueda Y. Complete response of Sister Mary Joseph Nodule from gastric adenocarcinoma treated with combination chemotherapy of Low-dose S-1 and cisplatin. J Exp Clin Cancer Res 2002; 21: 609-11.

8 Evans JW, Dutton J, Ng C, Arends M. Calcified “Sister Mary Joseph” umbilical metastasis from ovarian cystadenocarcinoma seen on an MDP bone scintigram. Clin Nucl Med 2002; 27: 134-5.

9 Vidal D, Matias-guiu X, Alomar A. Efficacy of imiquimod for the expression of Bcl-2, Ki67, p53 and basal cell carcinoma apoptosis. Br J Dermatol 2004; 151: 656-62.

10 Clingen PH, Berneburg M, Petit-Frere C, Woolions A, Lowe JE, Arlett, et al. Contrasting effects of an ultraviolet A tanning lamp on interleukin-6, tumor necrosis factor-α and intercellular adhesion molecul-1 expression. Br J Dematol 2001; 145: 54-62.

11 Murphy M, Mabruck MJEMF, Lenane P, Liew A, MCCann P, Buckley A, et al. The expression of p53, p21, Bax and induction of apostisis in normal volunteers in response to different doses of ultraviolet radiation. Br J Dematol 2002; 147: 110-7.

12 El-Domyati MM, Attia SK, Saleh FY, Ahmad HM, Gasparro FP, Uitto JJ. Effects of topical tretinoin, chemical peeling and dermabrasion on p53 expression in facial skin. Eur J Dematol 2003; 13: 443-8.

13 Chen P, Middlebrook MR, Goldman SM, Sandleret CM. Sister Mary Joseph nodule from metastatic renal cell carcinoma. J Comput Assist Tomogr 1998; 22: 756-7.