Melanosis Coli Involving Pericolonic Lymph Nodes Associated With the Herbal Laxative Swiss Kriss: A Rare and Incidental Finding in a Patient With Colonic Adenocarcinoma

Melanosis coli is a frequently reported incidental finding on colonic biopsies and resection specimens. The use of anthracene laxatives and an associated anthraquinone-induced apoptosis of colonic epithelial cells have been implicated in the pathogenesis of the condition.1–3 In spite of its relatively common occurrence, the presence of melanosis pigment in lymphoid tissue is uncommon, and to our knowledge the involvement of pericolonic lymph nodes has been reported in only 4 other patients in the English literature.4 We report the case of a patient with a history of long-term use of the over-the-counter herbal laxative Swiss Kriss (prepared from dried leaves of senna), who underwent a left hemicolectomy for colonic adenocarcinoma and was found incidentally to have melanosis coli, not only in his colonic mucosa and submucosa, but also in his pericolonic lymph nodes. This case implies that the melanosis pigment–laden macrophages formed in the lamina propria of the colon pass to the regional lymph nodes and may explain the observation of similar pigment-laden macrophages in other sites.4–7 In addition, the coexistence of melanosis coli and colonic adenocarcinoma associated with the use of the over-the-counter herbal laxative Swiss Kriss needs to be investigated further, since there may be an association between the two.

An 83-year-old man was admitted to the hospital for fatigue, abdominal pain, anemia, and cardiomegaly. A complete blood count and iron profile revealed severe iron deficiency anemia. A colonoscopy revealed a large, fungating mass in the sigmoid colon.

A surgical consult was obtained, and the patient was deemed to be a candidate for a left hemicolectomy. The patient underwent a left hemicolectomy and tolerated the procedure well. Pathologic examination revealed a colonic adenocarcinoma, stage B2 according to the Astler-Coller modification of the Dukes classification (stage II, T3 N0 M0). Postoperatively, the patient's course was unremarkable. The patient remained well and apparently free of disease 18 months after the operation.

The specimen was fixed in 10% neutral buffered formaldehyde solution, and samples were embedded in paraffin. Five-micrometer sections were processed for routine histology and were stained with hematoxylin-eosin. Selected sections were stained with the Gomori iron, periodic acid–Schiff (with or without digestion), Masson-Fontana, Fite (modified Ziehl-Neelsen) and Congo red stains. Deparaffinized and rehydrated sections were subjected to autofluorescence using an Olympus (Melville, NY) BX41 microscope equipped for reflected fluorescence illumination, as described elsewhere.8 Immunohistochemistry was performed on formalin-fixed, paraffin-embedded tissue sections using an avidin-biotin-horseradish peroxidase system after the tissue was deparaffinized and rehydrated according to standard protocol on an automatic immunostainer (Dako Autostainer, Universal Staining System, Dako Corporation, Carpinteria, Calif). The following antibodies were used: antibodies against S100, HMB-45, and CD68 (Dako), and cytokeratins (a cocktail of AE1/ AE3, CAM 5.2, and 35βH11) (Ventana Medical Systems, Tucson, Ariz).

The resection specimen contained a 3-cm, fungating mass centrally. The remainder of the specimen contained a smaller mass distally. The mucosa was abnormally dark brown to black, grossly consistent with melanosis coli, with sparing of the 2 masses. Sectioning of the colon revealed that the main tumor mass grossly appeared to penetrate through the wall of the colon into pericolonic fat. Several lymph nodes were noted in the pericolonic fat, none of which grossly appeared to contain obvious tumor metastasis. The cut surface of some of the lymph nodes was noted to be dark brown to black in areas, similar to the appearance of the colonic mucosa.

The patient's tumor was a moderately differentiated adenocarcinoma, measuring 2.2 cm in greatest dimension. The tumor invaded through the muscularis propria to reach, but did not penetrate, the pericolonic fat. Pigment- laden macrophages consistent with melanosis coli were present in the lamina propria, submucosa, and submucosal lymphatic vessels of the colon (Figures 1 and 2), with sparing of the areas within the tumor. Thirteen benign pericolonic lymph nodes were present, all containing similar pigment-laden macrophages in sinuses (Figure 3). Gomori iron stains in the sections of colon and lymph nodes were negative, as were immunohistochemical stains for S100, keratin, and HMB-45. CD68 highlighted the cytoplasm of the colonic macrophages containing the pigment, and similarly stained the lymph nodes (Figure 4). Moreover, the staining pattern of macrophages containing the pigment in both colonic mucosa and lymph nodes was identical for periodic acid–Schiff (with or without digestion), Masson-Fontana, Fite, and Congo red stains and showed similar autofluorescence, conforming to the reactions described for lipofuscin or lipofuscin-like material.8,9 

In 1978, Hall and Eusebi4 described 4 cases in which spindle-shaped, yellow-brown bodies were seen in the mesenteric lymph nodes of patients with melanosis coli. They compared the staining reactions and ultrastructural appearances of the so-called spindle bodies in the lymph nodes with the melanosis pigment in the colon, and reasoned that they were related. Beyond this observation in 4 patients, we know of no other reports in the English literature of melanosis pigment in the pericolonic lymph nodes of patients with melanosis coli. Although the first description of spindle-shaped, yellow-brown bodies was described by Hamazaki6 in a lymph node and similar yellow-brown bodies were observed in peritoneal lymph nodes of patients with sarcoidosis,5 it was not known whether these cases were associated with melanosis coli. In spite of the frequency with which one sees melanosis coli within colon biopsies and resections, the pigment has been reported infrequently to involve adjacent lymph nodes. Perhaps this finding is partially due to the fact (at least in our institutional experience) that simply by the sheer number of colonic biopsies received compared to resection specimens, the majority of cases of melanosis coli are diagnosed in colonic biopsies. Obviously, therefore, in most cases the pericolonic lymph nodes cannot be simultaneously examined for involvement by melanosis coli. Nevertheless, it is interesting that given the number of partial or complete colonic resections that occur every year in the United States and other countries, and the likely subset of cases that have melanosis coli as an incidental finding, that involvement of pericolonic lymph nodes has not been reported more frequently.

The presence of extensive melanosis pigment in macrophages of the colonic mucosa, submucosa, submucosal lymphatics, and sinuses of pericolonic lymph nodes in this case implies that the melanosis pigment–laden macrophages formed in the lamina propria of the colon pass to the regional lymph nodes and may explain the observation of similar pigment-laden macrophages in other sites, including in sarcoidosis-like conditions, as predicted by Hall and Eusebi.4 

Another interesting observation in this patient is the long-term (5-year) use of the over-the-counter herbal laxative Swiss Kriss. Although no definitive associations have been found in the use of laxatives, melanosis coli, and colonic adenocarcinoma,10–12 increased incidence of colorectal adenomas has been found in melanosis coli.13 Moreover, in a previous report, the development of tubulovillous adenoma with high-grade dysplasia was described in a patient who took an over-the-counter anthranoid-containing laxative.14 Thus, we feel that further studies are needed to determine whether there is a relation between the prolonged use of this herbal laxative and colonic adenocarcinoma.