Small cell neuroendocrine carcinoma (SCNEC) is a highly malignant and aggressive neoplasm. It sometimes occurs in the uterus, however, it can be difficult to distinguish whether it originated in the cervix or the corpus because it often occupies a large part of the uterus when discovered. However, adenocarcinoma is often found in the uterine cervix (1-5).

In recent years we experienced two cases of uterine SCNEC coexisted with minor components of endocervical adenocarcinoma (ECAC). Both were large tumors occupying most of the uterus, and the origins of them were unclear. We performed immunohistochemistry to compare expression patterns in two different histological types. In both cases, SCNEC components were diffusely positive for synaptophysin, somatostatin receptor type 2 (SSTR2) and p16 (1, 4, 6-8), and ECAC components were negative for synaptophysin and SSTR2, and diffusely positive for p16 (9). In one of those, since both SCNEC and ECAC components were diffusely and strongly positive for p53, we performed mutation analyses for TP53 (4, 6, 8). A common point mutation was detected in both components, but an additional point mutation was also seen only in the SCNEC component. Based on these findings, we present our opinions on the histogenesis, progression and treatment strategies of uterine SCNEC. When formulating therapeutic strategies for uterine SCNEC, including not only somatostatin analogue (SSA) administration (10, 11) but also future therapy targeting SSTR2 down-regulation (12), the possibility of coexisting histological subtypes, such as ECAC, should be considered.

Case 1. A 51-year-old woman had a previous illness of breast cancer approximately 10 years ago. She had no significant change in her abdomen on computed tomography (CT) examination five months ago. A mass of uterine corpus, 8 cm in the greatest diameter, was observed during scrutiny for irregular bleeding, and cytopathology indicated malignancy. After improving her general condition, total hysterectomy and bilateral adnexal resection were performed (Figure 1A and B). Histologically, the tumor was composed of monotonous small cells with enlarged ovoid hyperchromatic nuclei, finely stippled chromatin, inconspicuous nucleoli, and scanty cytoplasm, along with a small number of ECAC tubules (Figure 1C and D). We diagnosed SCNEC with minor ECAC component. FIGO classification was Stage IVB. Despite postoperative cisplatin and etoposide therapy, residual tumor grew. The patient died seven and a half months after surgery.

Immunohistochemistry was performed with monoclonal antibodies, CD56 (clone MRQ-42, Histofine, Tokyo, Japan), chromogranin A (polyclonal, Histofine), synaptophysin (clone MRQ-40, Roche, Tokyo, Japan), SSTR2 (clone EP149, Histofine), p16 (clone E6H4, Roche), p53 (clone DO-7, Roche) and Ki-67 (clone MIB-1, Agilent, Hachioji, Japan), using standard methods of automated immunostaining system (VENTANA BenchMark GX, Roche). SCNEC cells were diffusely positive for synaptophysin (Figure 1E), SSTR2 (Figure 1F), p16 (block-type pattern) (Figure 1G) and p53 (mutant pattern) (Figure 1H), focally positive for CD56, and negative for chromogranin A. MIB1-index was 90% in SCNEC cells (Figure 1I). ECAC tubules were diffusely positive for p16 (block-type pattern) (Figure 1G and K) and p53 (mutant pattern) (Figure 1H and L), and negative for CD56, chromogranin A, synaptophysin (Figure 1E) and SSTR2 (Figure 1F). MIB1-index was 50% in ECAC tubules (Figure 1I).

TP53 mutation analyses were performed by the PCR direct sequencing method using materials macrodissected separately from paraffin sections of broad SCNEC areas, and ECAC tubules only seen at the edge of the bulky tumor (Figure 1J-L) (Genome Analysis Department, LSI Medience Co., Ltd., Tokyo, Japan). Mutations, exon 4 p. P72R and exon 7 p. R248Q, were detected in SCNEC, while one mutation, exon 4 p. P72R, was detected in ECAC.

Case 2. A 53-year-old woman presented a pelvic cavity tumor and bilateral hydronephrosis revealed by CT examination during fracture scrutiny. Most of the uterus had been replaced with the tumor and a cervical biopsy was performed. Histologically, most of the tumor was composed of small, atypical cells described as in Case 1. Immunohistochemistry was performed with the same methods as in Case 1. The tumor cells were diffusely positive for synaptophysin, SSTR2 and p16 (block-type pattern), indicated cervical SCNEC (Figure 2A-D). FIGO classification was Stage IIIC1. The tumor shrank by cisplatin and etoposide therapy, and total hysterectomy and bilateral adnexal resection were performed four months after the lesion occurred. The postoperative course was good.

The therapeutic effect against the carcinoma was very good on the resected specimen, however, we found not only the residual SCNEC but also newly appeared foci of ECAC (Figure 2E-G, K). SCNEC cells were diffusely positive for p16 (block-type pattern), synaptophysin, SSTR2 (Figure 2H-J) and Ki-67 (MIB-1 index 90%), focally positive for CD56, sporadically positive for p53 (wild-type pattern), and negative for chromogranin A. ECAC cells were diffusely positive for p16 (block-type pattern) (Figure 2L) and Ki-67 (MIB-1 index 90%), sporadically positive for p53 (wild-type pattern), and negative for CD56, chromogranin A, synaptophysin and SSTR2 (Figure 2M and N).

Neuroendocrine cells are present in almost every organ and tissue, not just endocrine organs, and neuroendocrine neoplasm (NEN) is a concept that broadly covers tumors with neuroendocrine differentiation. NEN consists of neuroendocrine tumor (NET) G1, NET G2, NET G3, SCNEC and large cell neuroendocrine carcinoma (LCNEC) (6). SCNEC can occur anywhere in the gynecological tract, most commonly in the uterine cervix (1).

The presence of neuroendocrine differentiation can be confirmed by examining expression of neuroendocrine markers, which include components of endocrine granules that store peptide hormones such as chromogranin A, components of the synaptic endoplasmic reticulum and endocrine granules such as synaptophysin, and receptors involved in the regulation of neuroendocrine cells such as SSTR (6). SSTR has the subtypes SSTR1, 2a, 3, 4, and 5. SSTR2a is one of the seven transmembrane G protein-coupled receptors that binds to somatostatin 14 and somatostatin 28, and is involved in neurotransmission, endocrine regulation, and inhibition of cell proliferation. Volante et al. reported that score 2/score 3, in which SSTR2 was positive on the cell membranes of tumor cells, was highly correlated with tumor detection by somatostatin receptor scintigraphy, and 75% of these tumors showed a therapeutic response to SSA (7). In addition, SSTR2 has been reported to be a biomarker for poor prognosis in small cell lung carcinoma (SCLC) (12).

Synaptophysin is usually expressed in almost all tumor cells showing neuroendocrine differentiation. However, the expression of chromogranin A and SSTR2 is usually strong in many NETs, and weak or absent in many NECs (6). In this study, both SCNEC components were positive for synaptophysin and SSTR2, and negative for chromogranin A. These findings suggest that SCNEC in the uterus may be more likely to express SSTR2 than those in other organs. Actually, some institutions have reported SSTR2-positivity (score 2/score 3) rates of 43% to 50% in cervical NEC (13, 14). Eskander et al. reported that a large cohort of patients with cervical SCNEC demonstrated a genomic landscape distinct from SCLC, calling into question the biologic and therapeutic relevance of the histologic similarities between the entities (15). Therefore, the addition of SSA to the treatment strategy for the uterine SCNEC should be considered (14).

Our two SCNEC cases coexisted with minor components of ECAC. The incidence of cervical mixed tumors, SCNEC and other pathologic types, is 26.8% to 40%, suggesting that a high incidence of mixed tumor phenotype is a common feature of SCNEC of the uterine cervix (16). Masuda et al. reported that mixed SCNEC and cervical tumors had a clonal origin and were characterized by an ambiguous and flexible differentiation state (16). In Cases 1 and 2, p16 positivity was common to both SCNEC and ECAC components. In Case 1, a TP53 point mutation in exon 4 was common to both SCNEC and ECAC components, and another one was found in exon 7 in only the SCNEC component. These findings suggest that two components have a common origin in both cases and that SCNEC may have been developed as a result of neuroendocrine differentiation of ECAC in Case 1. The difference in MIB1-index between both components in Case 1 may indicate that SCNEC cells exceeded ECAC tubules in a volume when the cancer was detected.

In SCNEC components of the two cases, both MIB1-indexes were 95%. Sorbye et al. reported that gastroenteropancreatic NENs with Ki-67 >55% had a significantly shorter survival, but were more responsive to platinum-based chemotherapy (17). Miyoshi et al. also reported that etoposide and platinum tend to have a better prognosis compared to other regimens used for other subtypes of cervical cancer (18). In Case 2, preoperative cisplatin and etoposide therapy was also highly effective against the SCNEC component, but it did not lead to complete disappearance, suggesting the need for additional treatment such as SSA based on the SSTR2 positivity of the cancer cells and/or immune checkpoint inhibitors (19). In recent years, there have also been reports re-evaluating surgery following neoadjuvant chemotherapy for locally advanced neuroendocrine tumors (20). Furthermore, the ECAC component was identified after surgery, suggesting that additional resection was one of the correct choices because cisplatin and etoposide therapy is ineffective against this component. When developing drug treatment strategies for uterine SCNEC, including not only SSA administration (10, 11) but also future therapy targeting SSTR2 down-regulation (12), the possibility that other coexisting histological subtypes, such as ECAC, remain should be considered because other chemotherapies and/or surgical treatments may be required in some cases.

Comprehensive genetic testing, such as next-generation sequencing, is expected to continue providing important information in the future (13, 21, 22). However, when comparing two histological types, analyses can be difficult if the sample volume obtained from important microlesions is very small. In this study, detailed histological examination, immunohistochemistry, and TP53 mutation analysis (in Case 1) were performed on our two cases to obtain insights into the histogenesis and progression of cervical SCNEC that can lead to the development of treatment strategies.

These cases suggest that uterine SCNEC may coexist with ECAC, and in some cases, may even arise from this tumor. SSTR2 expression in uterine SCNEC may have prognostic and therapeutic significance, supporting consideration of SSA administration and future therapy targeting SSTR2 down-regulation. Because coexisting histological subtypes such as ECAC may persist after treatment directed at SCNEC, therapeutic planning should include careful pathological assessment and consideration of multimodal treatment.

The Authors have no conflicts of interest to declare in relation to this study.

Masamichi Bamba: Writing – original draft, Resources, Conceptualization. Tomo Namura: Writing - review & editing, Resources, Conceptualization. Toshikazu Kato: Writing - review & editing, Resources. Masanori Shiohara: Writing - review & editing. Kenji Isogawa: Writing - review & editing. Mao Uemura: Writing - review & editing, Resources. Shiro Wakinoue: Writing - review & editing. Hiroko Yomo: Writing - review & editing, Resources. Tetsuya Nakagawa: Writing - review & editing, Resources. Tetsuya Katsumori: Writing - review & editing, Resources. Ryoji Kushima: Writing - review & editing, Supervision. Suzuko Moritani: Writing - review & editing, Supervision, Resources, Conceptualization. All Authors read and approved the final version of the manuscript.

The Authors thank the patient’s husband of Case 1 and the patient of Case 2 for consenting in writing to the publication of this study and all the accompanied images.

None.

No artificial intelligence (AI) tools, including large language models or machine learning software, were used in the preparation, analysis, or presentation of this manuscript.