ISMC is a recently recognized distinct type of invasive cervical adenocarcinoma [3, 8], which was described by Lastra et al [3] as a morphologic variant of endocervical adenocarcinoma. SMILE is believed to be a putative precursor of ISMC because of their similar morphology. ISMC is an uncommon malignant cervical tumor. The mean age of ISMC among our cases was 48 years (range 40–51 years). As reported by Horn et al [7], the average age of ISMC is similar to that of invasive squamous cell and other histologic subtypes of invasive cervical adenocarcinoma. The symptoms of ISMC have no specific characteristics that distinguish them from usual invasive cervical carcinoma [9]. The ISMCs of our cases manifested as vaginal bleeding. However, owing to the ignorance of SMILE and ISMC, the thin-prep cytology test resulted in a misdiagnosis, namely, H-SIL, among our four cases, and the preoperative colposcopy biopsy was misdiagnosed as mucinous adenocarcinoma and squamous cell carcinoma in our five cases. Therefore, routine histological sections and immunohistochemical staining of ISMCs are particularly important for diagnosis.
ISMC is recognized as a mucinous subtype of HPV-associated adenocarcinoma by the International Endocervical Adenocarcinoma Criteria and Classification (IECC) system [9]. Similar to other histologic subtypes of cervical adenocarcinomas [7], HPV high-risk DNA can be detected in ISMCs [7–10]. The analysis of our cases demonstrated high-risk HPV in all informative cases and p16 overexpression in all tumors, supporting the hypothesis that high-risk HPV is associated with tumor development, mainly HPV 18. Immunohistochemical staining showed that most tumor cells were positive for p16, ck8, ck18, and CEA expression; had a high Ki-67 index; and were negative for CK5/6 and p63. IMP3 is a member of the insulin-like growth factor protein family and is a useful marker for several carcinomas [11]. Onishi [8] reported that IMP3 was more strongly and diffusely positive in invasive cervical adenocarcinoma than in intraepithelial lesions. IMP3 expression may be helpful in the differential diagnosis of ISMC and SMILE.
In general, most ISMCs show a pure stratified mucin-producing histology, whereas mixed cases are also reported to be accompanied by adenocarcinoma of the usual, endometroid or colloidal type or squamous cell carcinoma [3,7–8,12−13]. Among our cases, four were pure ISMCs, and four were ISMCs mixed with usual-type endocervical adenocarcinomas. The five cases were preoperatively misdiagnosed as mucinous adenocarcinoma and squamous cell carcinoma. As described by Stolnicu et al [12], ISMC can display a wide morphologic spectrum that can impart an ambiguous appearance and make it difficult to classify the tumor, especially to distinguish it from traditional poorly differentiated HPV-associated usual-type and mucinous-type adenocarcinoma.
There was a wide range of tumor sizes, with a mean size of 33 mm (ranging from 25–50 mm) among our cases. In the presence of the endophytic type and neck type, most of the cases were diagnosed with tumors > 20 mm. Typically, ISMC shows a cauliflower-like, papillary and ulcerous appearance. In our cases, most ISMCs had a depth of stromal invasion of more than half (5/8), vascular invasion (5/8), neutral invasion (3/8), uterine segment involvement (3/8), parametrial involvement (1/8), and lymph node metastasis (2/8). Based on the pathology findings of the ISMCs, ISMC may be more invasive in the stroma and vasculature.
The clinical stage of cervical carcinoma is of great significance to the prognosis and treatment. The higher the clinical stage is, the greater the risk of death will be among patients with cervical cancer. Survival in cervical carcinoma depends on stage, with 99% five-year survival for stage IA1, 65% for IIB and 43% for IIIB disease [14]. The presence of pelvic/para-aortal lymph node metastases, local extension of the disease and tumor size are also well-established prognostic factors of cervical carcinomas [16–17]. The mean tumor size of six recurrent patients was 5.7 cm (range 4.1 to 6.4 cm), as provided in published reports [3, 7]. Five patients showed pelvic lymph node metastases reported by Horn et al [7]. During the follow-up period of 6 weeks to 28 months, all of them developed pelvic recurrence. In our study, one patient developed vaginal recurrence, and the depth of stromal invasion of the tumor was full-thickness, accompanied by vascular and neutral invasion, cervical junction, vaginal fornix and vaginal cuff infiltration.
Several studies have suggested that ISMC is potentially more aggressive, with worse outcomes than usual-type endocervical adenocarcinomas [3, 7, 15] and with a substantial risk of distant metastatic disease, especially to the lung [7]. All five patients with ISMCs reported by Horn et al [7] showed pelvic recurrence within 6 weeks to 12 months during follow-up, and four patients had died. In the report of Lastra et al [3], two of the five patients with ISMCs presented with parametrial involvement and histologically proven pulmonary metastases after 9 months and 36 months, respectively. One patient developed widespread peritoneal disease at the time of diagnosis and died 1.5 months later. Nonetheless, all three patients with ISMCs in the report of Onishi et al [8] were alive without recurrence during a mean follow-up of 59.3 months. Three patients reported by Lei et al [9] were alive with no recurrence from 19 to 27 months of follow-up. Our eight patients were alive from 8 to 21 months of follow-up, and only one developed vaginal wall recurrence at 15 months. Until this study, there were 17 cases reported [3, 7–9, 13]; the addition of the 8 cases in our study brings the total to 25 cases. Eight of the 25 patients with information developed recurrent disease from 6 weeks to 36 months, and four patients developed distant metastases.