Sister Mary Jospeh's nodule as metastasis of colorectal cancer. Systematic review of the literature and meta-analysis

Abstract

Introduction and importance

Metastatic cancer of the umbilicus is an uncommon and rare presentation.

Case presentation

Our interest for the clinical outcomes of umbilical metastases from colon cancer arose after a 60-years old lady with ulcerated umbilical lesion came to our clinic. She was seen in several other clinics, and the diagnoses of infection of the umbilical region and/or of umbilical hernia were made. She was asymptomatic and in good clinical conditions. A complete evaluation led to the diagnosis of adenocarcinoma in the caecum with umbilical metastasis. During the hospital admission she underwent emergency colectomy for acute obstruction. An uneventful right colectomy was performed, but the lady died 21 months after surgery for diffuse metastases.

Clinical discussion

We performed a literature review of reports describing patients with umbilical metastases. Median survival rate was 7 months from the time of diagnosis of the umbilical metastasis (5 months in clinical reports and 8 months in autopsy studies). Observed survival rates were higher for patients with primary ovarian cancer (18 months), and endometrium (9 months). Median survival rate was 8 months in case of primary colon cancer. Chemotherapy and surgery allowed acceptable survival and quality of life in 8 patients with umbilical metastasis from colon cancer.

Conclusion

Clinical experience suggests that an aggressive approach may offer to selected groups of patients with umbilical metastasis from abdominal cancer acceptable quality of life and improved survival probabilities.

Highlights

• •Metastatic cancer of the umbilicus is an uncommon and rare presentation.
• •We performed a literature review of reports describing patients with umbilical metastases.
• •Median survival rate was 8 months in case of primary colon cancer. The worst prognosis was for primary pancreatic cancer (3 months).
• •Chemotherapy and surgery allowed acceptable survival and quality of life in 8 patients with umbilical metastasis from colon cancer.

1. Introduction

Even if umbilical metastases from cancer of the ovary were already described, the best-known description of umbilical metastasis was published in 1928 by William James Mayo (1861–1939), son of William Worrall Mayo (1815–1911), the founder of the Mayo Clinic in Rochester, Minnesota. In a meeting with his colleagues in Rochester, Mayo described the umbilical metastasis as a dismal prognostic factor of gastric cancer, and an index of not operability similar with the presence of a supra-clavicular metastatic lymph node [1]. This clinical sign and its negative prognostic value were pointed out to Mayo by the head surgical nurse Sister Mary Joseph (1856–1939- Born Susanne Dempsey) who was his surgical assistant, prepping and draping the abdomen of patients before surgery. Sir Hamilton Bailey coined the term “Sister Joseph's nodule” for an umbilical metastasis in his textbook “Clinical Physical Signs in Clinical Surgery” (1949) [2,3]. Sister Joseph, a Catholic nun, daughter of Irish immigrants, worked at Mayo Clinic for 30 years, leading the nursing teaching school. The original surgical building has been named “Joseph Building” in her memory. It is the most important eponym in medicine dedicated to a nurse.

Our interest for the clinical outcomes of umbilical metastases from colon cancer arose after a 60-years old lady with ulcerated umbilical lesion came to our clinic. She was asymptomatic and in good clinical conditions. A complete evaluation led to the diagnosis of adenocarcinoma in the caecum with umbilical metastasis. During the hospital admission she underwent emergency colectomy for acute obstruction. An uneventful right colectomy was performed, but the lady died 21 months after surgery for diffuse metastases [4,5].

We performed a systematic review to analyze the clinical outcomes of patients with umbilical metastases in the last 20, giving special attention at cases of umbilical metastases from colon cancer.

2. Methods

The work has been reported according with the SCARE criteria [6]. We performed a literature review of reports describing patients with umbilical metastases. A literature search was performed in December 2023 by 2 investigators who conducted a review of papers reported in PubMed, Embase, MEDLINE, and Cochrane Database. The strings “Umbilical Metastasis”, “Skin Metastasis”, and “Sister Mary Nodule” were used. There was no language restriction and screened reports were published from December 2003 to December 2023. Primary outcome was to identify the primary cancer. Secondary outcomes were actuarial survival rates of patients. Specific attention was given to patients with umbilical metastases from colon cancer. The research was register in the PROSPERO register (CPD4201808691). The study was approved by the Ethical Board of the Department of Surgery and from the University of Rome Sapienza (N112, January 2022). Written informed consent was obtained from the patient for publication and any accompanying images. A copy of the written consent is available for review by the Editor-in-Chief of this journal on request.

3. Results

Five hundred and fifty papers were screened. We found 168 papers describing the clinical outcomes of 336 patients with umbilical metastases (Supplementary 1, Supplementary 2,3). All analyzed papers were retrospective studies. Complete information were available for 225 patients. Another two papers reported reviews of autopsy studies, including 860 cases [7,8]. Overall, 1196 cases were collected. The organ involved by the primary tumor was unknown for 254 cases (21.2 %).

3.1. Site of primary tumor and time at the diagnosis

Table 1 shows the organ involved by the primary cancer. The ovarian carcinoma was the most frequent site of primaries (25.4 %), followed by colon (17.9 %), pancreas (6.5 %), and stomach (5.9 %). Breast cancer was the most common among the extra-abdominal primary cancers (2.9 %). The fifth and sixth decade were the age groups most represented (mean age 61 + −6 yeas). Diagnosis of umbilical metastases was more common in females with a ratio of (1.7: 1) for the high prevalence of umbilical metastases in patients with ovarian cancer. Diagnosis of colorectal cancer was more common in men. In 839 cases (839/1196 72 %), timing of umbilical metastasis was known: in 73.4 % of the cases the umbilical lesion was the first sign of malignancy, whereas in 26.6 % the umbilical lesion was evaluated after the diagnosis and/or treatment of the primary tumor. As concern timing for metachronous metastases, in most cases the diagnosis was made within four years from the diagnosis of the primary tumor. In cases of primary tumor located in the breast, colon, and endometrium there was a statistically significant longer interval between time between diagnosis of the primary tumor and the diagnosis of the umbilical metastasis (mean 38 months vs 18 months p < 0,0.01).

Table 1.

Primary tumor - umbilical metastasis.

Organ involved by the primary cancer	Clinical reports 336	Autoptic reports 860	Total 1196
Gastrointestinal organs	117	328	445
Colon	32	182	214
Stomach	31	40	71
Gallbladder	13	9	22
Pancreas	26	52	78
Rectum	4	13	17
Appendix	4	11	15
Esophagus	2	9	11
Small intestine	2	6	8
Duodenum	1	1	2
Biliary system	1	4	5
Anus	1	1	2
Genitourinary organs	76	331	407
Prostate	7	7	14
Ovary	49	255	304
Endometrium	10	45	55
Bladder	3	9	12
Kidney	3	2	5
Tuba	2	12	14
Urachus	1	1	2
Penis	1	0	1
Other organs	32	58	90
Pulmonary	3	15	18
Lymphoma	6	0	6
Neuroendocrine	4	0	4
Peritoneum	10	7	17
Breast	3	32	35
Liver	2	1	3
Skin	0	3	3
DSRCT (desmoplastic small round cell tumor)	4	0	4
Unknown primary	111	143	254

3.2. Clinical characteristics of the umbilical metastasis

Umbilical metastasis nodule presented a painful lump of hard-fibrous consistency, irregular edges, attached to the anterior abdominal wall in 90 % of the patients. The surface appeared ulcerated and necrotic, with hematic, serous, purulent, or mucous discharge in 20 % of the patients. The size of the lesion ranged between 0.5 and 10 cm (mean 1.8 cm). In 15 % the umbilical lesion was mis-diagnosed as an umbilical hernia. Histological findings show that most cases were metastasis of adenocarcinoma.

3.3. Survival rates

Median survival rate was 7 months from the time of diagnosis of the umbilical metastasis (5 months in clinical reports and 8 months in autoptic studies). Observed survival rates were higher for patients with primary ovarian cancer (18 months), and endometrium (9 months). Median survival rate was 8 months in case of primary colon cancer. The worst prognosis was for primary pancreatic cancer (3 months). Longer survivals were registered for women, related with better prognosis for umbilical metastases from ovary and endometrial cancer. Median survival was slightly better when the umbilical metastasis was diagnosed before the primary tumor (9,8 months) in comparison to when the umbilical lesion was diagnosed after primary tumor treatment (7,5 months).

3.4. Surgery and chemotherapy in selected patients with colon cancer

The umbilical metastasis was the first diagnosed sign in almost 50 % of the reported patients with colorectal cancer. Considering only clinical reports (32 patients), aggressive surgery and chemotherapy was the chosen therapeutic approach with curative intent in 7 patients (Table 2) [4,[9], [10], [11], [12], [13], [14]]. In five patients without extensive intrabdominal cancer spread, acceptable survival rates were obtained. Even in a patient with diffuse intrabdominal cancer, Chemotherapy and surgery were followed by a 5-year survival rate. In another FOUR patients [[15], [16], [17], [18]], chemotherapy and/or surgery were chosen only as a palliative intent. One patient died after 11 months (only surgical resection), one patient is alive and well 6 months from diagnosis (intestinal bypass of the cancer mass for acute obstruction, post-operative chemotherapy with FOLFOX), one is alive and well 7 months from diagnosis (only chemotherapy: 5-fluorouracil, oxaliplatin, and bevacizumab), and one patient (only chemotherapy: oxaliplatin(SOX)plus bevacizumab) is alive and well 22 months from the diagnosis.

Table 2.

Clinical outcomes of surgery and chemotherapy with curative intent in selected patients with umbilical metastases from colon cancer.

Authors (year)	Age/sex site	Liver met.	Peritoneal met.	Other met.	Pre-operative chemotherapy/post-operative chemotherapy	Synchronous/metachronous (months interval from diagnosis primary)	Surgery	+ nodes	Clinical outcome
Gabriele et al./2004 [4]	61/F Caecum	No	No	No	No pre-operative chemotherapy Yes post-operative chemotherapy-fluorouracil	Synchronous	Emergency right colectomy for acute obstruction while in hospital for diagnosis	Yes	Death 21 months from diagnosis and surgery. Disease progression
Wu et al./2010 [8]	37/M Caecum	No	No	Inguinal metastases	No pre-operative chemotherapy No post-operative chemotherapy (refused by the patient)	Synchronous	Elective right colectomy. Removal inguinal lymph nodes and umbilical area.	Yes	Death 4 months.
Chen and Liu/2015 [9]	61/M Right colon	No	No	No	No preoperative chemotherapy Yes postoperative chemotherapy- FOLFOX	Synchronous	Elective right colectomy and removal umbilical metastasis.	Yes	Alive and well 2 months
Grossi et al./2019 [10]	60/M Sigmoid	No	No	No	No preoperative chemotherapy Yes postoperative chemotherapy- Irinotecan-bevacizumab After 9 months diagnosis of lung metastases. Then oxaliplatin for 7 months Then capecitabine for 12 months	Synchronous	Elective sigmoid resection and portion of adherent small bowel.	Yes	Death 48 months from diagnosis and surgery. Disease progression
Iwata et al./2019 [11]	42/F Right colon	No	Yes	No	Yes preoperative fluorouracil-leucovorin-oxaliplatin-bevacizumab 12 months Yes postoperative same regimen 4 months	Synchronous	Elective right colectomy and removal umbilical metastasis.	Yes	6 months after surgery diagnosis of adenocarcinoma colonic anastomosis-resection. Alive and well 14 months after new surgery (50 months from initial diagnosis)
Majdouby et al./2021 [12]	48/M Sigmoid	No	No	No	No preoperative Yes oxaliplatin and capecitabine 6 months	Synchronous	Elective sigmoid resection-Removal of adherent portion bladder wall (no tumor invasions) and umbilical area.	No	Alive and well 18 months
Tarik et al./2022 [13]	41/F pregnant 7 months. Transverse colon/ Direct Invasion	No	No	No	No Preoperative No postoperative	Synchronous Direct invasion umbellic by tumor	Two months post delivery of health baby. Elective resection left transverse colon and umbilical area.	No	Alive and well 2 months.

4. Discussion

Umbilical nodules are rare, and they can be classified as benign or malignant. Benign nodules are more common [5,6]. Primary umbilical malignancies are less frequent than umbilical metastases by a ratio of 1 to 8. More frequently (85 % of cases) umbilical malignant lesions are metastases from intrabdominal neoplasms. The presence of an umbilical metastasis is a poor prognostic factor, and in most patients a palliative treatment is advisable. In selected patients, a more aggressive approach including surgery and chemo-radiotherapy may be indicated. Survival rates in this clinical setting depend on several factors, including type of primary cancer, clinical characteristics of the patient, and the presence or not of associated diffuse cancer spread in the abdomen. Conceptually, the absence or presence of diffuse cancer spread in the abdominal cavity might indicate different mechanisms at the basis of the diffusion of cancer cells to the umbilical area. Hugen et al. [7] collected autopsy data from National Netherland Cancer Registry. A total of 806 cases of persons dying with umbilical metastasis were included. There were 210 male (26.1 %) and 596 female (73.9 %) patients. Umbilical metastases most frequently originated from the ovaries in female patients (38.8 %) and from the colon in male patients (43.8 %). Patients with umbilical metastases had a dismal prognosis, with a median observed survival of 7.9 months (95 % CI 6.7–9.1) from the date of diagnosis of the umbilical metastasis. Univariable survival analysis demonstrated that the worst survival was seen in patients in whom the primary tumor originated from the pancreas (median survival 3.3 months, 95 % CI 1.88–4.79). Patients with a diagnosis of ovarian and endometrial cancer had the best prognosis with a 3-year actuarial survival rate of around 32 %.

Median survival was longer when the umbilical metastasis was diagnosed before the primary tumor (9,7 months) comparing with the clinical situation when the umbilical lesion developed and diagnosed after primary tumor treatment (7,6 months).

Independently by the possibility of complete curative resection or not, surgery associated with chemo-radiotherapy may have a major role to reduce symptoms and complications in patients with colorectal cancer, namely when the umbilical metastasis is not associated with widespread intrabdominal cancer spread [[19], [20], [21], [22]].

Several possibilities exist explaining the involvement of the umbilical node, without diffuse intrabdominal and peritoneal involvement. Direct extension of the cancer of the transverse colon to the umbilical region is a possibility [14]. Other possibilities, explaining a direct involvement of the umbilical region without diffuse intra-abdominal cancer diffusion, relate with the complex cross-roads of arterial, venous, and lymphatic represented by the umbilical region and eventual embryologic and developmental anomalies [12,[22], [23], [24]].

The results of our analysis highlight the dismal prognosis of patients with umbilical metastasis. In selected patients with colon or ovarian cancer, chemotherapy followed by surgery may results in relief of symptoms and improved survival [25].

The following is the supplementary data related to this article.

Supplementary 1

Clinical case reports.

Ethical approval

Approved Ethical Board University.

Funding

No funds were received for this work.

Author contribution

R Gabriele. Conceptualization; data curation, statistics; validation; reviewing manuscript.

M Campagnol data curation, statistics; validation; software; reviewing manuscript.

V Borrelli data curation, statistics; validation; software; reviewing manuscript.

I Iannone data curation, statistics; validation; software; reviewing manuscript.

P Sapienza data curation, statistics; validation; software; supervision, reviewing manuscript.

A Sterpetti data curation, statistics; validation; software; supervision, writing original draft manuscript.

Guarantor

Raimondo.gabriele@uniroma1.it.

Registration of research studies

1. Name of the registry: PROSPERO.

2. Unique identifying number or registration ID: (CPD4201808691).

Declaration of competing interest

The authors declare no conflicts of interest or commercial ties.

Data availability

Data are available from raimondo.gabriele@uniroma1.it.