Clinicopathological and Immunohistochemical Analysis of Recurrent Endometrial Carcinomas: A Retrospective Study from an Indian Cohort.

Endometrial carcinoma (EC) ranks as the third most common gynecologic malignancy in India, with an estimated crude incidence rate of 3.9 per 100,000 reported in 2022, by GLOBOCAN.[ 1 ] In contrast to ovarian cancers, ECs are often diagnosed at an early FIGO stage due to their typical clinical presentation with postmenopausal bleeding and are usually low-grade tumors. The cornerstone of treatment is surgical procedures like endometrial staging surgery (ESS) or total abdominal hysterectomy with bilateral salpingo-oophorectomy. Adjuvant therapies such as radiotherapy (RT) or chemotherapy (CT) are tailored based on disease stage and individual risk factors.[ 2 ] The key prognostic factors for EC include patients’ age, histological subtype, FIGO grade and stage, depth of myometrial invasion, and the presence or absence of lymphovascular space invasion (LVSI). Recent advancements in risk stratification categorize EC into low-, intermediate-, and high-risk groups, integrating molecular characteristics with specific histological parameters like tumor type, LVSI, and depth of invasion.[ 3 ] Notably, the majority of EC cases fall into the low-, or intermediate-risk categories, with high-risk tumors constituting <15% of cases.[ 4 ] Survival outcomes are encouraging with an overall 5-year survival rate of 84%, which rises to 94% for patients with low-grade tumors.[ 5 6 ] Despite early diagnosis, appropriate management, and favorable histological features, a small proportion of EC recur.[ 7 ] Numerous studies have investigated the risk factors for tumor recurrence, but their findings remain inconsistent, and no definitive consensus has been established.[ 8 ] The prognosis for patients with recurrent disease is poor, underscoring the need for updated management strategies for EC. Historically, RT was reserved for high-risk endometrial cancer (EC) patients, but recent guidelines now recommend extending its use to intermediate-risk cases. Vaginal brachytherapy is now advised for low-stage (Stage IA) patients with risk factors, replacing the previous approach of observation alone. While clinical trials support the use of external beam radiation therapy in intermediate-risk patients, concerns about complications remain.[ 9 10 11 12 ] Some studies question the validity of risk factors for vaginal recurrence, suggesting these factors may also be present in patients without recurrence, indicating that vaginal recurrence may not always reflect aggressive tumor biology.[ 7 ] Recent evidence suggests that vaginal recurrence may be linked to tumor cell presence in the vaginal vault during surgery, rather than recurrence driven by aggressive behavior, questioning the need for administration of vaginal brachytherapy.[ 4 ] While multi-institutional studies on EC in Western populations have debated various risk factors for recurrence, there is a significant gap in comprehensive analyses focused on the Indian population. Given the differences in tumor characteristics and behavior across demographic groups, population-specific studies are essential for generating relevant insights and more universally applicable data on recurrence. The main aim of this study was to analyze the pathological features associated with recurrence in EC within the Indian population and to study the difference between vaginal and extravaginal recurrences.

There were 17 cases of recurrent ECs reported between January 2012 and December 2024. The slides were available for review in all cases. All except one patient had surgical resection of the primary tumor. One patient was managed by palliative CT and RT due to medical comorbidities. Similarly, all recurrences were biopsied except one patient with multiple metastases, where palliative CT and RT were given. The patients’ ages at the time of first diagnosis ranged from 40 to 84 years, with a median of 60 years. The median age for recurrence was 62 years, and the median time for first recurrence was 24 months. In 4 cases, recurrence was seen after 3 years (delayed recurrence), and three of them were endometrioid FIGO Grade 1. The majority were postmenopausal women ( n = 16, 94.1%). The most common clinical presentation was postmenopausal bleeding ( n = 14, 82.3%), followed by abdominal pain ( n = 2, 11.8%) and intermenstrual bleeding ( n = 1, 5.9%). No significant family history was noted to suggest genetic associations. Medical conditions included hypertension ( n = 4, 23.5%), diabetes mellitus with hypertension ( n = 3, 17.6%), hypothyroidism, diabetes mellitus, and bronchial asthma in 2 patients each (11.8%), and one patient had Non-Hodgkin lymphoma. The primary tumors were managed with ESS ( n = 13) or total hysterectomy with bilateral salpingo-ophorectomy ( n = 2). One case did not undergo surgery due to comorbidities. Post-surgery, six patients received sandwich therapy, four underwent adjuvant CT alone, two received adjuvant RT alone, and five did not receive any adjuvant therapy. The most common site of recurrence was the vagina ( n = 8, 47.1%), followed by the liver and abdominal wall ( n = 2 cases each, 11.8%). Recurrence was also noted in the omentum, bladder, peritoneum, and para-vertebral region, with one case each (5.9%) and one case having multiple extravaginal site metastasis. The disease-free period ranged from 1 to 7 years, with a median of 2 years. Management of recurrent tumors included surgical resection in 6 cases (35.5%), CT alone in 3 cases (17.6%), and a combination of CT and RT in 4 cases (23.5%). FIGO staging was available for 16 cases, and the majority belonged to FIGO stage IA (n = 7,43.8%). Five patients were lost to follow-up, 7 were alive (41.2%), and 5 patients (29.4%) died due to disease. Table 1 provides the clinical details of these patients. Clinical profile of patients with recurrent endometrial carcinomas FIGIO: International Federation of Gynecology and Obstetrics, BSO: Bilateral salpingo-oophorectomy Macroscopic examination ( n = 16) revealed that most tumors were located in the uterine body ( n = 15, 93.8%), with only one case centered in the LUS (6.25%). The predominant growth pattern was proliferative/polypoidal ( n = 13, 81.3%), with only two tumors having an infiltrative pattern (12.5%). Tumor sizes ranged from 2.5 cm to 9 cm, with a median size of 5 cm. Peritoneal fluid cytology, available in eight cases, showed malignant cells in two. Macroscopic adnexal involvement was noted in one case, while gross omental and peritoneal involvement was observed in another. Microscopically, endometrioid carcinoma was the most common histological type ( n = 11, 64.7%), with mucinous differentiation observed in three cases. Seven of the eleven cases were low-grade (FIGO grades 1 and 2). Superficial myometrial invasion was prevalent, seen in 10 cases (62.5%), and only one case had premalignant changes in the adjacent endometrium. Cervical stromal involvement was rare, occurring in only three cases. The diffuse pattern of invasion was the most common (62.5%), with an inflammatory stromal response noted in 43.8%. None of the cases exhibited a MELF pattern of invasion. Parametrial and vaginal cuff margin positivity was identified in one case. Adnexal involvement was present in two cases, one of which also had peritoneal and omental deposits. LVSI was identified in seven cases (43.8%), with substantial involvement observed in three. Retroperitoneal lymph node dissection (RPLND) revealed nodal metastases in five cases. Extranodal extension was noted in two cases. Recurrent tumors retained the same histological type as their primary counterparts, with endometrioid tumors exhibiting mucinous differentiation, displaying consistent features at the recurrence site. Notably, one mixed tumor and one carcinosarcoma showed recurrence of the serous component. Detailed histopathological findings are summarized in Table 2 . According to the ESMO-ESGO-ESTRO classification, nine tumors fell into the high-risk category, whereas two were classified as high-intermediate risk. Microscopic parameters studied in primary endometrial carcinomas FIGIO: International Federation of Gynecology and Obstetrics, LUS: Lower uterine segment, LVSI: Lymphovascular space invasion, MELF: Microcystic, Elongated and Fragmented, EIC: Endometrial intraepithelial carcinoma IHC analysis of low-grade endometrioid carcinomas revealed strong ER-positive staining in both primary and recurrent tumors. PgR expression was robust in the primary tumors but showed reduced intensity in the recurrences, with one case exhibiting complete loss. In three primary cases, p53 IHC was available and showed a wild-type staining pattern, with a similar pattern in recurrent tumors. High-grade endometrioid carcinomas also demonstrated ER positivity in both primary and recurrent tumors, albeit with reduced expression levels in the recurrences. PgR expression was absent in the recurrent tumors, and one case showed a p53 mutation-type overexpression in the recurrence, contrasting with its wild-type pattern in the primary tumor. Serous carcinomas, carcinosarcoma, mixed tumors, and clear cell carcinoma exhibited consistent IHC profiles between primary and recurrent tumors. Her-2 testing, conducted in two cases, was negative. The clear cell carcinoma showed Napsin A positivity while being negative for other markers. Detailed IHC profiles are outlined in Table 3 . Comparison of immunohistochemistry profile of primary and recurrent endometrial carcinomas IHC: Immunohistochemical, LG: Low grade, HG: High grade, ER: Estrogen receptor With near equal numbers of vaginal and extravaginal recurrences, we compared the clinicopathological parameters between the two groups. Tumors with vaginal recurrence were predominantly endometrioid, with two mixed tumors, low-grade, and superficial myometrial invasion. A significant proportion (57%) did not exhibit any stromal response ( P = 0.06). In contrast, patients with extravaginal recurrences more frequently had involvement of the uterine serosa, LUS, cervical stroma, and omentum. LVSI was present in both groups in near equal proportions. RPLND revealed positive lymph node status only in the extravaginal recurrence group ( P = 0.03) and not in the vaginal recurrence group. A significant number of cases (62.5%) did not receive any form of adjuvant therapy in the vaginal recurrence group ( P = 0.009). Table 4 outlines the clinicopathological variations between tumors with vaginal and extravaginal recurrences. Comparison of clinicopathological parameters between vaginal and extravaginal recurrence in endometrial carcinomas * p - value calculated using Chi-square test FIGIO: International Federation of Gynecology and Obstetrics, LUS: Lower uterine segment, LVSI: Lymphovascular space invasion

This study provides valuable insights into the recurrence patterns of EC within our population, despite being limited by the small sample size. Our findings suggest that risk stratification, rather than individual prognostic parameters, serves as a more reliable predictor of recurrence. Loss of PgR expression in recurrent tumors highlights its potential role in disease progression. The predominance of low-grade endometrioid tumors with superficial myoinvasion, negative nodal status, and absence of stromal response in vaginal recurrences suggests that vaginal recurrence may not necessarily reflect more aggressive tumor biology, although we could not establish statistical significance for some of these parameters due to small sample size. The intrinsic biology of tumors dictates recurrence patterns in EC. We stress the importance of molecular techniques to better understand tumor biology and the microenvironment. In addition, with 25% of tumors recurring after 36 months, extended and more frequent follow-up is essential. There are no conflicts of interest.

ECs are often detected early, respond effectively to treatment, and typically carry a favorable prognosis. Yet, a notable percentage of these seemingly well-behaved tumors recur, posing the intriguing question of what drives recurrence in such cases. Although various prognostic factors have been identified, definitive guidance for using additional targeted therapies in high-risk cases remains elusive. Compounding the challenge, existing literature frequently offers conflicting conclusions. We analyzed 17 cases of recurrent EC from our hospital database, focusing on several prognostic factors. The recurrence rate was 11.04%, slightly lower than the 17%–21% reported in other populations.[ 14 15 ] Patient profiles, including median age and common presentations, aligned with prior data from the same population.[ 13 ] However, the median age at recurrence was a decade younger than in Western studies.[ 14 16 ] Median time to first recurrence (24 months) exceeded the 16 months reported in French cohorts,[ 14 ] though high-grade tumors uniformly recurred within 24 months, consistent with Huijgens report.[ 8 ] Conventional histological parameters such as FIGO grade, DOI, and LVSI were not significant predictors of recurrence, with tumors evenly distributed between grades and predominantly showing superficial invasion. Premalignant changes in adjacent endometrium, positive margins, and positive peritoneal cytology were rare. Mucinous differentiation was noted in three cases. Interestingly, the MELF invasion pattern, previously linked to mucinous differentiation in low-grade EC[ 17 ] and poor prognostic factors,[ 18 ] was absent in all recurrent cases in our study. This aligns with findings by Kihara et al ., who reported no association between the MELF pattern and recurrence-free or disease-free survival,[ 19 ] and Pavlakis et al ., who found no disease progression or mortality linked to MELF.[ 20 ] Conversely, a diffuse invasion pattern, commonly observed in ECs,[ 13 17 ] was more frequent among these cases. Stromal reaction patterns showed no clear predominance. Although conventional parameters assessed individually did not significantly predict recurrence, the risk stratification groups outlined by the ESMO-ESGO-ESTRO classification proved effective in identifying patterns of recurrence. In our study, 68.8% of recurrent cases fell into the high-risk or high-intermediate-risk categories, closely aligning with the 51% frequency reported by Bendifallah et al .[ 15 ] Other investigators have included the immune cells in the prediction model for recurrences,[ 21 ] which was beyond the scope of this study. Few studies have investigated recurrence patterns in ECs. Siegenthaler et al . linked recurrence patterns to molecular classifications, noting abdominal recurrences in p53-mutated tumors and loco-regional recurrences in MMRd tumors, with 55% of late recurrences in NSMP types.[ 14 ] Although we lacked molecular data, 80% of our p53-mutated tumors (by IHC) showed extravaginal recurrence, including three abdominal cases. In addition, 75% of recurrences after 36 months were endometrioid Grade 1, aligning with Siegenthaler et al .’s findings.[ 14 ] Our study observed a near-equal distribution of vaginal and extravaginal recurrences. Moschiano et al . found no significant risk factors for vaginal recurrence except cervical involvement, suggesting contamination during surgery rather than tumor aggressiveness as the cause.[ 7 ] Similarly, we noted vaginal recurrences in low-grade endometrioid tumors with superficial myometrial invasion. None of them had a positive nodal status, and a significant proportion did not receive any adjuvant therapy (statistically significant). Notably, unlike other studies reporting vaginal recurrence alongside extravaginal relapse,[ 16 ] we did not have any such case. These findings reinforce that vaginal recurrence may not be truly related to tumor biology. IHC revealed loss of PgR expression in 56% of recurrent tumors, aligning with its association with aggressive disease and poor survival in endometrioid carcinomas.[ 22 ] Tumors with strong, diffuse PgR expression in the primary lesion showed complete loss in recurrence, consistent with progesterone’s tumor-suppressive role in EC.[ 23 ]

This retrospective study examined cases of recurrent EC diagnosed in the Department of Pathology at our institution between January 2012 and December 2024. Cases were identified through departmental records, with clinical and treatment details retrieved from the Gynecologic Oncology Department at our institution. Clinical data included the following: (a) age at diagnosis, (b) menstrual status (premenopausal/postmenopausal), (c) family history, (d) medical history, (e) presenting symptoms, (f) primary treatment details, (g) administration of adjuvant CT and RT, (h) follow-up details, (i) time to recurrence, including the interval from diagnosis to recurrence, (j) disease-free interval, (k) recurrence site (locoregional/systemic), (l) treatment for recurrence, (m) cause of death (disease-related/other), and (n) FIGO 2018 stage. Macroscopic features were obtained from pathology reports for these cases. Recorded characteristics of the primary tumor included: (a) location (body, lower uterine segment [LUS], or cervix), (b) tumor size, (c) growth pattern (infiltrative, proliferative, or polypoidal), (d) gross involvement of the uterine serosa, adnexa, pelvic or abdominal peritoneum, and (e) fluid cytology positivity, or omental involvement. Histopathology and immunohistochemistry (IHC) slides were retrieved from departmental archives and reviewed for the following parameters: (a) tumor type, (b) FIGO grade, (c) depth of myometrial invasion, (d) findings in adjacent endometrium, (e) myometrial features (leiomyoma/adenomyosis), (f) pattern of invasion, (g) stromal response, (h) LVSI (absent, focal, or substantial [≥5 vessels involved]), (i) cervical involvement, (j) margin status (parametrium/vaginal cuff), (k) adnexal involvement, (l) peritoneal involvement, (m) omental involvement, and (n) retroperitoneal lymph node status. The pattern of invasion was categorized as described in our previous study: [ 13 ] (a) absent (confined to the endometrium), (b) diffuse (small glandular clusters with irregular contours or individually dispersed glands infiltrating the myometrium, with or without a desmoplastic stromal response), and (c) broad front/pushing expansile (broad clusters of neoplastic glands with well-defined margins, pushing into the myometrium, with or without a desmoplastic stromal response). The presence of the MELF pattern of invasion was also documented. The stromal response was similarly classified as outlined in our earlier study:[ 13 ] (a) fibromyxoid (desmoplastic), (b) inflammatory (dense lymphocytic infiltrate surrounding invasive tumor nests), or (c) absent (no stromal response). In addition, tumors were stratified using the ESMO-ESGO-ESTRO classification into low-risk, intermediate-risk, high-intermediate-risk, and high-risk categories based on stage, histological type, FIGO grade, depth of invasion, and LVSI status.[ 3 ] For recurrent tumors, only the histological type was recorded. The immunohistochemical (IHC) profile of the primary tumor and recurrent tumor was analyzed, enabling a comparison of staining patterns. Institutional Ethics Committee approval was obtained (Reference number: 233/2023).