Clinico-Morphological Correlations with Ki-67 and p53 Immunohistochemical Expression in High-Grade Gastrointestinal Neuroendocrine Neoplasms

Abstract

Background/Objectives: The 2019 WHO classification redefined high-grade gastrointestinal neuroendocrine neoplasms (GI NENs), encompassing not only poorly differentiated neuroendocrine carcinomas (NECs), but also well-differentiated grade 3 neuroendocrine tumors (NETs G3). Since both subtypes share a Ki-67 index > 20%, distinguishing them based solely on morphology is challenging. Prior studies have shown TP53 alterations in NECs but not in NETs. This study aimed to evaluate clinico-morphological parameters and the immunohistochemical (IHC) expression of p53 in high-grade GI NENs to identify relevant correlations. Methods: Tumors were stratified by Ki-67 index into two groups: >20–50% and >50%. p53 IHC expression was assessed as “wild-type” (1–20% positive tumor cells) or “non-wild-type” (absence or >20% positivity). Correlations were analyzed between Ki-67, p53 status, and various pathological features. Results: Significant correlations were found between the Ki-67 index and maximum tumor size, pT stage, lymphovascular invasion, perineural infiltration, and diagnostic classification. Similarly, p53 immunohistochemical status was significantly associated with lymphovascular invasion, lymph node metastasis, and tumor classification (NET G3 versus NEC, including NEC components of MiNENs). Conclusions: The findings support the value of Ki-67 and p53 as complementary biomarkers in the pathological evaluation of high-grade GI NENs. Their significant associations with key morphological parameters support their utility in differentiating NETs G3 from NECs, particularly in cases showing overlapping histological features. The immunohistochemical profile of p53 may serve as a useful diagnostic adjunct in routine practice.

1. Introduction

Neuroendocrine neoplasms (NENs) are rare tumors in the general population, with malignant potential and variable risks of local recurrence and metastasis. These risks are influenced by tumor size and specific histopathological features, including the degree of differentiation, depth of invasion, lymphovascular involvement, and proliferative index [1]. Tumors originating from neuroendocrine cells within the gastrointestinal (GI) tract pose distinct diagnostic and therapeutic challenges. Although they often exhibit slow growth, NENs may become life-threatening because of unregulated secretion of bioactive substances such as monoamines and peptide hormones, which are normally produced by the cells from which these tumors arise [1].

The 2019 WHO classification of neuroendocrine neoplasms revised the categorization of high-grade NENs, recognizing that these include not only poorly differentiated neuroendocrine carcinomas (NECs) but also well-differentiated grade 3 neuroendocrine tumors (NETs G3) [2]. Differentiating between these two entities remains challenging, as morphological criteria alone are often insufficient. Both NECs and NETs G3 share a Ki-67 proliferation index exceeding 20%, making histological assessment alone inadequate for conclusive diagnosis [3,4,5,6]. Moreover, standard platinum-based chemotherapy regimens used for NECs have shown limited efficacy in NET G3 cases, underscoring the clinical importance of accurately distinguishing between these high-grade subtypes [7,8].

Recent research, particularly in pancreatic NEN cohorts, has shown that p53 immunohistochemical (IHC) expression can serve as a surrogate marker for genetic alterations commonly found in NECs, but typically absent in NETs. Most NECs demonstrate a non-wild-type p53 staining pattern on IHC analysis [9,10,11,12,13].

This study aimed to evaluate clinico-morphological parameters and the immunohistochemical profile of p53 in high-grade gastrointestinal NENs in order to identify statistically significant correlations and contribute to ongoing efforts towards data standardization and a deeper understanding of these neoplasms.

2. Results

2.1. Description of the Study Group

A total of 42 patients were included: 22 (52.4%) women and 20 (47.6%) men. The age at diagnosis ranged from 43 and 81 years, with a mean of 62.04 years (SD ± 13.25; SE = 2.04) and a median of 67 years. Female patients had a higher mean age (68.54 ± 9.98, median 73), with most diagnosed in the seventh decade (54.5%). Male patients had a lower mean age (54.9 ± 12.61, median 50.5), with 50% diagnosed in their 40s. Among the studied cases, 16 were diagnosed as NET G3 (38.1%), 10 (23.8%) as NEC, comprising 6 small-cell neuroendocrine carcinomas (SCNECs) and 4 large-cell neuroendocrine carcinomas (LCNECs), and 16 as mixed neuroendocrine–non-neuroendocrine neoplasms (MiNENs) with an NEC component (38.1%; all LCNECs). In women, NETs G3 were more frequent (45.5%), followed by MiNENs (36.3%) and NECs (18.2%). In men, MiNENs predominated (40%), followed by NECs and NETs G3 (30% each). The most common tumor sites were the colorectum (61.9%), stomach (33.3%), and small intestine (4.8%). In female patients, tumors were most often located in the stomach (54.5%), followed by the colorectum (36.4%) and small intestine (9.1%). Male patients had predominantly colorectal tumors (90%), with no enteric involvement. Large-cell NECs accounted for 76.9% of all NEC-type cases. SCNECs were mostly colorectal (83.3%), and LCNECs were also frequently colorectal (65%), particularly as components within MiNENs (80%). Tumor size exceeded 5 cm in most cases (64.3%). Colorectal high-grade NENs larger than 5 cm were most common (76.9%), whereas all enteric tumors were ≤5 cm. Gastric tumors were equally distributed (50% ≤ 5 cm, 50% > 5 cm). A summary of the clinical features is provided in

Table 1. Evaluated clinical features in the study group.

	NET G3	NEC	MiNEN
Age at diagnosis (%)
<60 yo	50	10	40
≥60 yo	27.2	36.4	36.4
Tumor site (%)
Stomach	42.9	14.2	42.9
Small intestine	100	0	0
Colorectum	30.8	30.8	38.4
Maximum tumor size (%)
≤5 cm	80	0	20
>5 cm	14.8	37.1	48.1

.

2.2. Clinico-Pathological Features and Correlations with Ki-67 and p53 Status

The Ki-67 index was evaluated in all cases. Most NET G3 tumors (66.7%) had values ranging from 20 to 50%, whereas NECs and NEC components of MiNENs exhibited Ki-67 indices greater than 50% in 83.3% of cases (Figure 1).

The overall mean Ki-67 index was 38.6% for NET G3 and 68.1% for NECs and NEC components of MiNENs. When analyzed by tumor site, the mean Ki-67 index for NETs G3 was 24.33% in gastric tumors, 21% in small intestinal tumors, and 53.75% in colorectal tumors. In comparison, NECs showed mean values of 67.14% in the stomach and 69.38% in the colorectum; no NECs were identified in the small intestine (

Table 2. Diagnostic and topographical distribution of Ki-67 mean values.

	NET G3	NEC and NEC Component of MiNEN
	Mean Ki-67 Index (±SD; SE)
Stomach	24.33% (±1.63; 0.66)	67.14% (±8.09; 2.86)
Small Intestine	21% (±0; 0)	-
Colorectum	53.75% (±25.4; 9)	69.375% (±22.42; 5.28)

).

As expected, a Ki-67 index of 20–50% was more frequently observed in NETs G3 (12/16, 75%). In contrast, higher Ki-67 values (>50%) were predominant in NECs (6/10, 60%) and NEC components of MiNENs (14/16, 87.5%), with a statistically significant correlation (r = 0.510, p < 0.001; χ² = 10.904, p < 0.001; V = 0.510). Histopathological features evaluated in the study group are summarized in

Table 3. Overview of pathological characteristics in the study cohort.

	NET G3	NEC	MiNEN
pT stage (%)
pT1–2	100	0	0
pT3–4	31.6	26.3	42.1
Lymphovascular invasion (%)
Absent	75	0	25
Present	29.4	29.4	41.2
Perineural infiltration (%)
Absent	100	0	0
Present	27.8	27.8	44.4
Lymph node metastasis (%)
Absent	66.7	0	33.3
Present	30.3	30.3	39.4
Ki-67 index (%)
>20–50%	66.7	22.2	11.1
>50%	16.7	25	58.3

.

Further significant correlations were identified between the Ki-67 index and several pathological parameters: tumor size (r = 0.459, p = 0.002; χ² = 8.849, p = 0.003; V = 0.459); tumor stage (pT) (r = 0.375, p = 0.015; χ² = 5.895, p = 0.015; V = 0.375); lymphovascular invasion (r = 0.315, p = 0.042; χ² = 4.169, p = 0.041; V = 0.315); and perineural infiltration (r = 0.471, p = 0.002; χ² = 9.333, p = 0.002; V = 0.471).

Notably, all tumors staged as pT1–2 exhibited Ki-67 index values ranging from 20 to 50%, while 63.2% of those classified as pT3–4 had Ki-67 values > 50%. Furthermore, all tumors without perineural infiltration showed Ki-67 values in the range 20–50%, whereas 66.7% of tumors with perineural infiltration had Ki-67 values > 50%.

2.3. p53 Immunohistochemical Status

Immunohistochemical analysis of p53 revealed wild-type expression in 87.5% of NET G3 cases and only 7.7% of NEC/MiNEN cases. Conversely, non-wild-type p53 expression was detected in 92.3% of NEC/MiNEN cases and only 12.5% of NET G3 cases (

Table 4. p53 immunohistochemical profile in high-grade gastrointestinal neuroendocrine neoplasms.

p53 IHC Status	NET G3	NEC and NEC Component of MiNEN
	No. (%)	No. (%)
Wild-type	14 (87.5)	2 (7.7)
Non-wild-type	2 (12.5)	24 (92.3)

; Figure 2).

A strong correlation was found between p53 IHC status and diagnostic category (NET G3 vs. NEC, including MiNEN) (r = 0.798, p < 0.001; χ² = 26.751, p < 0.001; V = 0.798).

Additional significant associations were identified between the p53 expression profile and several pathological parameters. All tumors staged as pT1–2 (100%) exhibited wild-type p53 expression and were classified as NETs G3. In contrast, 68.4% of pT3–4 tumors showed non-wild-type p53 expression, with 84.6% of these diagnosed as NECs or NEC components of MiNENs (r = 0.414, p = 0.006; χ² = 7.184, p = 0.007; V = 0.414). Regarding lymphovascular invasion, 70.6% of positive cases demonstrated non-wild-type p53, and 91.7% of these were NECs or NEC components of MiNENs. Conversely, 75% of tumors without vascular invasion exhibited wild-type p53, all of which were NETs G3 (r = 0.369, p = 0.016; χ² = 5.707, p = 0.017; V = 0.369). Similarly, wild-type p53 was observed in 66.7% of node-negative cases, all classified as NETs G3, while 69.7% of node-positive tumors showed non-wild-type p53, 91.3% of which were NECs or NEC components of MiNENs (r = 0.307, p = 0.048; χ² = 3.917, p = 0.046; V = 0.307). The statistically significant correlations between clinico-pathological features and Ki-67/p53 IHC status are shown in

Table 5. Statistically significant associations between clinico-pathological parameters and Ki-67 / p53 status.

Parameter	Bio Marker	Correlation (r)	p-Value	Chi-Square (χ2)	Cramér’s V	Key Observations
Tumor size	Ki-67	0.459	0.002	8.849	0.459	Higher Ki-67 values in tumors > 5 cm
Tumor stage (pT)	Ki-67	0.375	0.015	5.895	0.375	Ki-67 > 50% more frequent in pT3–4
Lymphovascular invasion	Ki-67	0.315	0.042	4.169	0.315	Ki-67 > 50% associated with vascular invasion
Perineural infiltration	Ki-67	0.471	0.002	9.333	0.471	High Ki-67 in tumors with perineural spread
Diagnostic category	Ki-67	0.510	<0.001	10.904	0.510	Ki-67 20–50% in NET G3; >50% in NEC and MiNEN
Tumor stage (pT)	p53	0.414	0.006	7.184	0.414	Wild-type p53 in all pT1–2; non-wild-type in 68.4% of pT3–4 (mostly NEC/MiNEN)
Lymphovascular invasion	p53	0.369	0.016	5.707	0.369	Non-wild-type p53 associated with vascular invasion
Lymph node metastasis	p53	0.307	0.048	3.917	0.307	Wild-type p53 absent in node-positive cases
Diagnostic category	p53	0.798	<0.001	26.751	0.798	Wild-type p53 in NET G3; non-wild-type in NEC/MiNEN

.

3. Discussion

The global incidence of gastrointestinal neuroendocrine neoplasms remains difficult to quantify due to inconsistent data reporting, highlighting the need for replication studies and international data sharing [14,15]. Large-scale studies have reported significant site and gender differences. For example, in a multicenter analysis of 760 NEN cases, rectal tumors accounted for 94.5% of cases, and the majority were NETs G1 (88.4%) [16]. Another cohort of 92 cases found NECs predominantly gastric (66.7%) and more frequent in men (68.4%), whereas NETs were more evenly distributed by sex and most commonly colorectal (54.3%) [14].

In comparison, our study revealed a higher frequency of colorectal NENs (61.9%) and a slight female predominance (52.4%). Notably, the mean age of male patients (54.9 years) was considerably lower than that of female patients (68.5 years). MiNENs with NEC components represented 38.1% of cases, a particularly high proportion compared to prior reports. Moreover, 76.9% of NECs in our cohort exhibited large-cell morphology, in line with findings from larger studies [17].

Although the distinction between NETs and NECs relies primarily on morphological evaluation, both the mitotic count and Ki-67 index provide essential support in the classification of high-grade NENs. NETs typically show lower Ki-67 proliferation indices (median around 5%), with most cases remaining below the 40% threshold. In contrast, NECs usually present significantly higher values, often above 40% [1,10,18,19,20]. It is recognized that intratumoral heterogeneity in the Ki-67 proliferation index and biomarker expression is common in gastroenteropancreatic NENs and can impact grading, classification accuracy, and prognostic assessment [21,22]. A Ki-67 index greater than 55% has been proposed as a more reliable threshold for identifying NECs, which tend to be rapidly progressive and responsive to platinum-based chemotherapy [1,3].

In our study, 75% of cases with Ki-67 >50% were classified as NECs or NEC components of MiNENs. The mean Ki-67 index in these groups was 68.1%, exceeding 60%, whereas NETs showed a mean index of 38.6%, remaining below the 40% threshold reported in the literature. Several studies have demonstrated that NENs with Ki-67 indices > 50% are significantly more aggressive than those with values between 20% and 50% [1,3,23]. Consistent with these findings, our study revealed statistically significant correlations between Ki-67 index and key pathological parameters such as tumor size, pT stage, lymphovascular invasion, and perineural infiltration.

The increasing recognition of well-differentiated NENs with high proliferation indices has heightened interest in identifying biomarkers capable of distinguishing these tumors from NECs. Among the most studied is p53, which is frequently altered in NECs but generally retains wild-type expression in NETs G3, particularly in pancreatic tumors [4,9,11]. While most studies have focused on pancreatic NENs (PanNENs), TP53 and RB1 alterations have also been described beyond the pancreas [11,12,19,24]. However, immunohistochemical and biomolecular data regarding p53 expression in gastrointestinal NENs remain limited, with most available studies including mixed cohorts [25]. Importantly, p53 IHC is most useful when interpreted qualitatively—wild-type versus aberrant patterns—rather than semi-quantitatively by intensity, as studies have shown that staining intensity does not reliably enhance diagnostic or prognostic value [22].

Our findings are consistent with previous reports showing that aberrant p53 IHC patterns—characterized by either complete absence of nuclear staining or diffuse overexpression in >20% of tumor cells—strongly correlate with TP53 mutations in poorly differentiated NECs [11,14,24]. These patterns reflect underlying molecular events: loss-of-function TP53 mutations typically result in undetectable protein, while gain-of-function mutations lead to the accumulation of a stabilized mutant protein in the nucleus [26,27,28]. In contrast, NETs G3 often retain a wild-type p53 staining pattern, with only scattered nuclear positivity (1–20%), consistent with the rapid degradation of wild-type p53 protein under physiological conditions and the low prevalence of TP53 mutations in this group [14,24]. This distinction reinforces the utility of p53 IHC as a surrogate marker for TP53 status and as a valuable ancillary tool for differentiating NETs G3 from NECs, particularly in morphologically ambiguous cases or when molecular testing is not feasible.

In our cohort, non-wild-type p53 immunostaining was observed in 92.3% of NECs and NEC components of MiNENs, while 87.5% of NETs demonstrated wild-type p53 expression. Moreover, we found statistically significant correlations between p53 IHC status and diagnostic category, lymphovascular invasion, and lymph node metastasis.

To our knowledge, this is one of the few studies focusing exclusively on high-grade gastrointestinal NENs, with no inclusion of lower-grade cases, and evaluating both p53 immunohistochemical patterns and Ki-67 proliferation indices in correlation with key clinico-morphological parameters using inferential statistical analysis. This comprehensive approach reinforces the practical diagnostic value of these biomarkers in differentiating NETs G3 from NECs, particularly in gastrointestinal sites.

Although SSTR immunohistochemistry was not included in our analysis, its diagnostic and theranostic relevance in high-grade neuroendocrine neoplasms is well established. NETs G3 often retain SSTR2 expression, supporting the use of somatostatin analogs and peptide receptor radionuclide therapy, whereas NECs typically lack SSTR expression and are managed with cytotoxic chemotherapy [22,29].

The rationale behind our study lies in the critical need to accurately differentiate high-grade neuroendocrine neoplasms, as treatment strategies differ significantly between NETs G3 and NECs. It has already been demonstrated that NETs G3 exhibit limited responsiveness to standard platinum-based chemotherapy, which remains the cornerstone of NEC treatment [7,8].

Given the rarity of clinical studies focused specifically on well-differentiated high-grade neuroendocrine tumors and the absence of standardized therapeutic protocols for this group, further research is urgently needed to better characterize their biomolecular profile and guide treatment decisions. Current therapeutic options for NETs G3 currently include somatostatin analogs and chemotherapy, but clear guidelines are still lacking [7,30].

While our study provides valuable insights, we acknowledge certain limitations. These include the relatively small sample size and the fact that classification and pathological assessment were performed within a single institution without centralized or external review. Nevertheless, all cases were evaluated by experienced board-certified pathologists in accordance with the 2019 WHO classification. Future multicenter studies with larger cohorts and external validation would be instrumental in strengthening diagnostic robustness and confirming our findings. In addition, assessing p53 staining intensity and exploring intratumoral heterogeneity of IHC reactivity could provide further diagnostic and prognostic insights.

4. Materials and Methods

The study focused on high-grade gastrointestinal neuroendocrine neoplasms (NENs) diagnosed both morphologically and immunohistochemically over a four-year period at the Clinical Pathology Service of “Saint Apostle Andrew” Emergency County Hospital in Constanţa, Romania. Tumor sites included the stomach, small intestine, and colorectum. Diagnosis was established through the correlation of histopathological findings with immunohistochemical results, in accordance with WHO recommendations. At least two neuroendocrine markers were used to confirm tumor lineage, and the Ki-67 proliferation index was assessed for grading.

The inclusion criteria required informed patient consent, gastrointestinal primary tumor localization, no prior oncologic treatment, surgical specimen availability, a Ki-67 index > 20%, confirmed immunohistochemical diagnosis using at least synaptophysin and chromogranin A, and sufficient paraffin-embedded tissue for further analysis. Cases were excluded if any of these conditions were not met, including extra-gastrointestinal localization, biopsy without subsequent surgery, incomplete or inappropriate IHC testing, or inadequate tissue material.

Following the review of histopathological examination request forms, the following parameters were collected and entered into a database: patient age at diagnosis, sex, tumor site, maximum tumor size, pT category, lymph node metastasis, pTNM stage, lymphovascular invasion, perineural infiltration, NEC subtype (SCNEC or LCNEC), and final diagnosis confirmed through IHC. The Ki-67 proliferation index was retrieved from hospital pathology databases, where it had been determined during routine diagnostic procedures. No additional Ki-67 testing or re-evaluation was performed as part of this study. Given that the Ki-67 index in NEC is often significantly higher than 20% and seldom falls below 50% [9,31], two ranges were selected for statistical analysis: >20–50% and >50%.

In MiNEN cases, tumor size was recorded as a global dimension, as it was not possible to quantify the NEC component separately. For pathological features such as lymphovascular invasion, perineural infiltration, and lymph node metastases, only cases showing direct involvement of the NEC component (confirmed morphologically) were included in the analysis.

A total of 42 eligible cases were subjected to complementary p53 immunohistochemical testing at the Center for Research and Development of Morphological and Genetic Studies in Malignant Pathology, Constanţa, Romania. For each case, a representative tissue block was selected—free of tissue necrosis that could compromise immunohistochemical quality and containing a substantial tumor component. Immunostaining for p53 was performed using a human anti-p53 monoclonal antibody (clone SP5/rabbit—RTU, Master Diagnostica, Madrid, Spain). Formalin-fixed, paraffin-embedded tumor blocks were sectioned at 4 µm thickness and mounted on slides. The staining protocol followed the manufacturer’s guidelines and included heat-induced epitope retrieval (HIER), peroxidase blocking, antibody incubation, detection, and hematoxylin counterstaining.

Nuclear staining was considered positive regardless of staining intensity. p53 immunostaining was evaluated manually by two independent pathologists using light microscopy, with a minimum of 500 tumor cell nuclei assessed per case. Any discrepancies in interpretation were resolved through joint review to reach consensus. The p53 IHC profile was classified as non-wild-type if nuclear staining was either completely absent or observed in more than 20% of tumor cells; staining in the 1–20% range was interpreted as wild-type [14,24].

In this study, NECs and the NEC components of MiNENs were considered as a single diagnostic category for p53 immunohistochemical evaluation. This approach was supported by recent studies showing that NECs and MiNENs with NEC components associated with adenocarcinoma often share molecular alterations, suggesting a common pathogenetic pathway with conventional adenocarcinomas [11,32,33,34]. Moreover, molecular features of the NEC component are generally similar to those of the associated carcinoma, reinforcing the concept that NEC represents a neuroendocrine variant along the adenocarcinoma progression pathway, distinct from NETs [1].

All clinico-morphological and IHC data were compiled in a Microsoft Excel database (version 2019, Microsoft Corporation, Redmond, WA, USA), and statistical analysis was conducted using SPSS software (version 26.0, IBM Corporation, Armonk, NY, USA). Continuous variables were described using the mean, standard deviation (SD), and standard error (SE), while categorical variables were expressed as percentages.

To evaluate associations between clinico-morphological features and p53 immunohistochemical status, Pearson’s correlation coefficient and the Chi-square test (or Fisher’s exact test, where applicable) were applied. A p-value < 0.05 was considered statistically significant. Cramér’s V was calculated to assess the strength of associations, with values ≥ 0.30 indicating a moderate association and values ≥ 0.50 reflecting a strong association.

5. Conclusions

Our study supports the role of Ki-67 and p53 as valuable biomarkers in the pathological assessment of high-grade gastrointestinal neuroendocrine neoplasms. In our cohort, most tumors with Ki-67 > 50% were classified as NECs or NEC components of MiNENs, and higher proliferation indices correlated with more aggressive pathological features, while p53 immunohistochemical status (wild-type vs. non-wild-type) showed statistically significant correlations with critical pathological features, including pT stage, lymphovascular invasion, lymph node metastasis, and, most notably, the microscopic diagnostic category.

These findings reinforce the diagnostic utility of Ki-67 and p53 in differentiating between NET G3 and NEC, particularly in morphologically equivocal cases where both entities share a proliferation index > 20%. The immunohistochemical profile of p53 has demonstrated strong potential for distinguishing between well-differentiated and poorly differentiated neoplasms, offering a practical tool for improving diagnostic accuracy.

In addition, the high proportion of MiNENs with NEC components observed in our cohort highlights the importance of recognizing mixed histologies, which may influence prognosis and therapeutic decision-making.

Our results align with and support the broader scientific effort to standardize diagnostic criteria and promote international data sharing in the field of neuroendocrine pathology.

Given the limited response of NETs G3 to platinum-based chemotherapy regimens—which are effective in treating NECs—our findings underscore the need for accurate classification to ensure appropriate treatment strategies. Although limited by sample size, this study provides a compelling rationale for including p53 in routine immunohistochemical panels and calls for further multicenter studies to validate these results and support the development of evidence-based therapeutic guidelines for high-grade GI NENs.