In 1959, Colman and Rosenthal modified Sampson’s criteria to apply to carcinomas developing from adenomyosis: (i) carcinoma should be absent from the normally situated endometrium and anywhere else in the pelvis; (ii) the carcinoma should be actually observed to be arising from the epithelium of the areas of adenomyosis and not invading from another source; and (iii) endometrial stromal cells should surround the aberrant glands to support a diagnosis of adenomyosis.17 In this case, co-existence of normal endometrial glands, atypical hyperplastic glands, and cancerous glands were observed both in eutopic endometrium and ectopic endometrium in uterine adenomyoma lesions. In addition, obvious endometrial stromal cells surrounding cancerous glands were observed. These observations were sufficient to prove that eutopic endometrium and ectopic endometrium in myometrium both had malignant transformation and they were both primary. In reported cases of carcinomatous change of uterine adenomyoma or adenomyosis, the probability of simultaneous malignant transformation of eutopic endometrium was approximately 20%.9, 11, 18-20 Bingjian Lu and colleagues reported 3 cases of serous carcinoma in uterine cervical adenomyosis. Among them, one case did not have carcinomatous changes, one case had serous carcinoma, and one case had endometrioid adenocarcinoma in eutopic endometrium.9
The most important feature of this case was that whether the poorly differentiated endometrioid adenocarcinoma in the left parametrial area combined with partial clear cell carcinoma was the carcinomatous changes of the original deep endometriosis or carcinomatous changes and metastasis of uterine adenomyoma or eutopic endometrium. HE staining showed that the structure and morphology of parametrial tumor cells were different from those of adenomyosis and tumor cells in eutopic endometrium. Because normal endometriotic glands surrounding parametrial lesions or in lesions were not found, diagnostic criteria of carcinomatous changes of endometriosis proposed by Sampson in 1925 could not be completely met.21 Therefore, the possibility that this lesion was from metastasis could not be ruled out. Normal endometrial glands could not be found in left parametrial adenocarcinoma that had involvement of rectum, ureter, and right ovary. Thus, the possibility of metastasis of carcinomatous change of adenomyosis to left parametrial was more likely.
However, combined with surgery and pathological evidence, we considered that the left parametrial adenocarcinoma was more likely to be carcinomatous changes of the original deep endometriosis. First, the left parametrial lesion was generally not adjacent to the adenomyoma lesion and the left parametrial adenocarcinoma lesion contained chocolate-like fluid, indicating that this site was likely to have the original deep endometriosis. Next, tumor cells in parametrial adenocarcinoma lesions and endometrial tumor cells in endometrium and adenomyosis had significant differences in differentiation level, structure and morphology. In addition, there were CD10+ endometrial stromal cells surrounding tumor cell masses in parametrial lesions, indicating that tumor cells were from endometrial glands. Finally, normal endometriotic glands were not found in left parametrial adenocarcinoma lesions; which might be associated with insufficient or biased material collection in pathological examination or carcinomatous change of all endometriotic glands in that place.
Immunohistochemical staining of microsatellite instability in left parametrial adenocarcinoma lesions that had clear cell components and eutopic endometrioid adenocarcinoma was performed. MLH1, MSH2, MSH6, and PMS2 were all positive in parametrial lesions, whereas PMS2 was negative in eutopic endometrioid adenocarcinoma lesions (Fig 4). The microsatellite instability levels between these two were not consistent, indicating that the pathogenic mechanisms of these two carcinomatous changes might be different and the parametrial lesion was likely to be the carcinomatous change of the original deep endometriosis. It is considered that obesity and unopposed estrogen use are high-risk factors leading to carcinomatous change of endometriosis.16 However, the patient did not have the above high risk factors and did not have a family history. The eutopic and ectopic endometrium both had carcinomatous changes, suggesting that the genetic factor of the patient might play a critical role in disease pathogenesis.
There is still controversy about the adjuvant treatment of carcinomatous change of endometriosis after surgery. Some scholars consider that radiotherapy can benefit patients more than chemotherapy.19 However, considering the ovarian and colorectal involvement in this patient, chemotherapy was chosen. This patient had a history of endometriotic cysts and dysmenorrhea but did not have menstrual disorders. Therefore, it was diagnosed as deeply infiltrating endometriosis before surgery. Tumors were discovered in pathological examination after hysterectomy. However, retrospective analysis of preoperative examination showed that the patient had CA125 of 183.4U/ml and color ultrasound suggested abundant blood flow in the posterior cervical wall lesion. These results all suggested the possibility of carcinomatous changes.