Clinicopathological characteristics and surgical value of primary gastrointestinal lymphoma

Abstract

BACKGROUND

Primary gastrointestinal lymphoma (PGIL) is a relatively uncommon clinical entity, exhibiting distinctive features including occult primary sites, nonspecific clinical presentations, and considerable diagnostic and therapeutic difficulties. Consequently, comprehensive clinical investigations into its clinicopathological characteristics and surgical intervention value are warranted to enhance diagnostic and therapeutic proficiency.

AIM

To investigate the clinicopathological characteristics and surgical significance of PGIL from a surgical perspective, providing a theoretical basis for optimizing diagnostic and therapeutic strategies.

METHODS

This study included 50 cases of PGIL treated by the General Surgery Department of the Chinese PLA Air Force Medical Center from June 2001 to March 2025. Data were extracted from the Electronic Medical Record system for retrospective analysis. A retrospective analysis was conducted on their epidemiological, clinical manifestations, imaging, pathological features, and treatment outcomes. Descriptive statistics were applied for data summarization, with continuous variables presented as frequencies and percentages. Correlations between variables were assessed using the Spearman rank correlation coefficient.

RESULTS

All cases had the gastrointestinal tract as the primary site. Abdominal pain was the most common initial symptom (52.0%), with 80.0% of patients experiencing pain during the course of the disease, and 38.0% experiencing hematochezia/melena or anemia. Computed tomography diagnosis exhibited a high overall sensitivity (94.3%); the endoscopic detection rate was 91.5%. Diffuse large B-cell lymphoma was the most common subtype (52.0%). The improvement rate was higher in the surgery combined with chemotherapy group than in the chemotherapy only group. The incidence of postoperative complications was 26.5%, all occurring in patients with tumors > 5 cm.

CONCLUSION

Diffuse large B-cell lymphoma is the primary PGIL subtype. Imaging and endoscopic biopsy are diagnostic essentials. Surgery aids in resection, complication management, and pathologic diagnosis. Multidisciplinary, individualized strategies are recommended, necessitating further prospective molecular studies.

Key Words: Gastrointestinal lymphoma; Multimodal diagnostics; Clinical characteristics; Surgical operation; Multimodal diagnostics

Core Tip: Primary gastrointestinal lymphoma predominantly manifests as diffuse large B-cell lymphoma, with abdominal pain as the leading symptom. Comprehensive diagnosis relies on imaging (computed tomography sensitivity: 94.3%) and endoscopic biopsy (detection rate: 91.5%). Surgery is pivotal for definitive diagnosis (via complete specimen acquisition), emergency management (e.g., obstruction/perforation), and primary lesion resection, particularly for tumors > 5 cm (26.5% complication rate). While combined surgery and chemotherapy showed higher improvement rates than chemotherapy alone, statistical significance was not reached. Integration of surgical strategies into multimodal therapy may optimize outcomes, emphasizing individualized approaches based on tumor size, stage, and complications.

INTRODUCTION

Lymphoma is a malignant tumor originating from the lymphohematopoietic system, posing a substantial global health burden. Among its subtypes, primary gastrointestinal lymphoma (PGIL) refers to lymphoma arising from the gastrointestinal tract and underlying mucosa-associated lymphoid tissue (MALT). In recent years, the incidence of PGIL has been rising. As one of the most common types of extranodal lymphomas, PGIL accounts for 30%-40% of all extranodal lymphomas and 1%-4% of all gastrointestinal malignancies. The stomach is the most frequently involved site, followed by the small intestine and colorectum[1]. The etiology and pathogenesis of PGIL remain incompletely understood. Current evidence suggests that multiple factors may contribute to its development, including radiation exposure, autoimmune diseases, microenvironmental dysregulation, and adverse lifestyle factors[2]. Furthermore, PGIL is relatively uncommon in clinical practice. Definitive diagnosis and subtyping often require surgical biopsy to obtain sufficient tissue samples, highlighting the critical role of surgical intervention. Moreover, the disease is prone to severe complications such as gastrointestinal bleeding, obstruction, and perforation. Therefore, a systematic analysis of its clinical characteristics and surgical value from a surgical perspective is of considerable importance.

Based on data from 50 patients with PGIL, a retrospective study was conducted to provide clinical insights for further optimizing the comprehensive management strategy of this disease.

MATERIALS AND METHODS

Objects

Patients diagnosed with PGIL treated by the General Surgery Department of the Chinese People’s Liberation Army Air Force Medical Center (Beijing, China) from June 2001 to March 2025 were selected. This was a single-center retrospective study. The medical record data were cases with relatively complete traceable medical records and examination data in our hospital, and the earliest medical records that met the requirements were June 2001 to the time of the study, that is March 2025. Inclusion criteria were: (1) Confirmation of diagnosis by histomorphology and/or immunohistochemistry; and (2) Completion of at least one full treatment course at our institution. Exclusion criteria comprised: (1) Presence of concurrent other malignant tumors; and (2) Cases with in complete clinical information. These criteria were applied to minimize potential confounding effects from major comorbidities and inadequately documented cases. Ultimately, 50 patients were included in the study. Among them, 34 patients who underwent surgical resection with chemotherapy were assigned to the experimental group, whereas 16 patients receiving chemotherapy alone served as the control group.

Research methods

General information such as sex, age, and lifestyle habits of the enrolled patients were extracted from the electronic medical record system. The chief complaints, concurrent symptoms, duration and severity of symptoms at presentation, and detailed medical and family history, were systematically documented. Past medical history, including family history, was also documented. Additionally, images and reports of ultrasound, X-ray, computed tomography (CT), positron emission tomography (PET)/CT, magnetic resonance imaging, and other examinations were collected, recording the location, size, and invasion of the tumor. Endoscopic findings, including tumor size, biopsy site, and biopsy results, were also recorded. Pathological data included histological types and immunohistochemistry test results. Postoperative complications were statistically analyzed. This study only included patients who continued follow-up treatment at our hospital after surgery. The final treatment outcomes were categorized as improved, stable, or deteriorated based on clinical symptoms, physical signs, and imaging or PET/CT findings.

Statistical analyses

Clinical data were statistically organized and descriptively analyzed using Excel, SPSS 24.0, and Origin Pro 2024 software. Categorical variables are presented as the frequencies and percentages, either overall or stratified by group. Intergroup comparisons for count data were performed using the χ² test or Fisher’s exact test, as appropriate. Associations between variables were assessed using Spearman’s rank correlation coefficient. P < 0.05 was considered statistically significant.

RESULTS

General information

The age of onset of patients in this group ranged from 15 years to 89 years, with a median age of 59 years. Patients over 60 years of age accounted for a higher proportion (40.0%). Males comprised the majority of this group (64.0%). The most common primary site for the lymphoma was the intestine (70.0%). Abdominal pain was the most frequent initial symptom (52.0%) and concurrent symptom (80.0%). Tumors larger than 5 cm were found in 46.0% of the patients. Additional data are summarized in Table 1.

Table 1 General patient information.

Item	Category	Number of cases	Percentage
Basic characteristics
Sex	Male	32	64%
	Female	18	36%
Family history	Yes	16	32%
Behavioral risk factors	Smoking	5	10%
	Alcohol consumption	5	10%
Tumor characteristics
Primary site	Stomach	15	30%
	Intestine	35	70%
Maximum tumor diameter	> 5 cm	23	46%
	≤ 5 cm	16	32%
	Infiltrative growth difficult to evaluate	11	22%
Clinical manifestations
Initial symptom	Abdominal pain	26	52%
	Abdominal mass	8	16%
	Mass with pain	2	4%
	Hematochezia or melena	7	14%
	Others	7	14%
Concurrent symptoms	Pain	40	80%
	Hematochezia/melena	19	38%
	Nausea and vomiting	16	32%
	Weight loss	18	36%
	Anemia	19	38%
Comorbidities	Cardiac disease	8	16%
	Chest disease	20	40%
	Gastritis	13	26%
	Liver disease	17	34%
	Gallbladder disease	8	16%
	Other abdominal diseases	28	56%
	Central nervous system disease	5	10%
	Hypertension	10	20%
	Diabetes	7	14%
Metastasis to other sites		30	60%

Auxiliary examinations

Ultrasonography is often less sensitive due to the varying primary locations and deep-seated nature of the lesions. Different tumor sites may exhibit distinct features. For example, lymphoma in the small intestine may present as thickening of the intestinal wall, which can extend to the mesentery, abdominal wall, or retroperitoneum (Figure 1). Most cases are also associated with abnormally enlarged lymph nodes. CT typically reveals masses closely related to the gastrointestinal tract or abnormal thickening of the intestinal wall, with the highest overall diagnostic rate of 94.3%. In cases of metastasis, PET/CT was employed for post-diagnostic assessment of metastasis, demonstrating lymph nodes with abnormal metabolic activity (Figure 2). The number of PET/CT-positive cases in Table 2 corresponds to the number of patients with confirmed metastasis. Endoscopic examinations commonly reveal gastrointestinal ulcers, intestinal stenosis, or space-occupying masses (Figure 3).

Figure 1 Ultrasonographic findings. A: Multiple hypoechoic lesions are visualized in the peritoneal cavity. The largest, located in the perigastric region, measures 1.5 cm × 1.1 cm, demonstrates well-defined borders and a regular morphology, and exhibits homogeneous, markedly hypoechoic internal echogenicity; B: Localized thickening of the intestinal wall was observed in the terminal ileum, with a maximum thickness of approximately 0.5 cm; the layered structure of the wall remained relatively clear; C: Multiple lymph node echoes were detected in the retroperitoneum, with the largest measuring approximately 2.3 cm × 1.0 cm. These lymph nodes had clear boundaries and a regular shape; D: Lymph node echo was observed in the retroperitoneum, with the larger one located at the umbilical level in the right mid-abdomen, measuring 15 cm × 2.2 cm. The lymph node had clear boundaries and a regular shape, but the hilum structure was unclear; E: Multiple hypoechoic areas were found in the intestinal interspace, with the largest measuring 0.7 cm × 1.0 cm; F: Multiple enlarged lymph node echoes were detected in the mesentery, with the largest measuring approximately 0.8 cm × 1.8 cm. These lymph nodes exhibited relatively clear boundaries, an irregular shape, and an unclear hilum structure.

Figure 2 Positron emission tomography/computed tomography findings. Segmental circumferential thickening and mass formation were noted in the second segment of the small intestine, extending from the left upper to mid-abdomen, with significantly increased fluorodeoxyglucose uptake. Multiple enlarged lymph nodes were observed in the abdominal cavity surrounding the small intestinal lesion, also demonstrating markedly elevated fluorodeoxyglucose uptake. Arrow: Segmental small bowel wall thickening and a mass, respectively. The circled areas highlight regions with increased fluorodeoxyglucose uptake. The solid lines mark the measured lengths.

Figure 3 Endoscopic features of lymphoma. A: An ulcer was observed at the junction of the gastric fundus and body, with a white coating in the center. The surrounding mucosa exhibited congestion, edema, and convergence, with a rigid sensation upon insufflation, poor peristalsis, and slight oozing; B: A circumferential ulcerative and proliferative lesion was found at approximately 300 cm of the small intestine, with an uneven surface covered by dirty coating, leading to intestinal stenosis; C: A circumferential ulcerative mass causing luminal stenosis was detected 160 cm from the pylorus in the small intestine. The surrounding mucosa was congested and reddened, with local mucosal rigidity and friability, making it prone to bleeding; D: A circumferential ulcerative and proliferative lesion was observed at approximately 300 cm of the small intestine via anal approach, with an uneven surface covered by dirty coating, leading to intestinal stenosis; E: A mucosal mass was found 50 cm from the ileocecal junction in the small intestine, with surface erosion and fresh blood stains. Mild luminal stenosis was noted at the lesion site; F: A circumferential space-occupying lesion was detected in the ascending colon during endoscopy, showing surface ulceration and a hard texture.

Table 2 Auxiliary examination results.

Characteristics	Number of examinations	Positive diagnoses	Percentage
Ultrasound	49	18	36.7%
CT	35	33	94.3%
PET-CT	36	24	66.7%
Endoscopy	47	43	91.5%

CT: Computed tomography; PET-CT: Positron emission tomography/computed tomography.

Histological examination

Diffuse large B-cell lymphoma (DLBCL) represents the most prevalent histologic subtype of PGIL. In this study, DLBCL constituted the highest proportion (52.0%) of cases, which is consistent with established literature. Immunophenotypic positivity rate serves as a critical indicator for determining cellular origin and differentiation, and reliably reflects both proliferative activity and malignant potential. The immunophenotypic positive rates of various subtypes were as follows: (1) DLBCL (26 cases), with high positive rates for cluster of differentiation 20 (CD20) (84.6%, 22/26), paired box protein 5 (80.7%, 21/26), and multiple myeloma protein 1 (65.4%, 17/26); (2) MALT lymphoma (12 cases), with high positive rates for CD20 (100%, 12/12), B-cell lymphoma 2 (Bcl-2) (75%, 9/12), and CD79a (66.7%, 8/12); (3) T-cell lymphoma (TCL) (5 cases), with high positive rates for Bcl-2 (100%, 5/5), CD3, and CD5 (80%, 4/5), as well as CD8, CD20, and CD56 (60.0%, 3/5); (4) Follicular lymphoma (FL) (5 cases), with high positive rates for CD20, CD21, and Bcl-2 (80%, 4/5); and (5) Transformed lymphoma (MALT to DLBCL) (1 case), positive for CD3, CD5, CD10, CD20, CD23, CD79a, Bcl-2, and Bcl-6 (Figure 4).

Figure 4 Histopathological features of lymphoma (hematoxylin and eosin staining). A: Gastric diffuse large B-cell lymphoma; B: Jejunal diffuse large B-cell lymphoma; C: Duodenal follicular lymphoma (low-grade, grade 1-2); D: Axillary follicular lymphoma (grade II); E: Intestinal monomorphic T-cell lymphoma of the small intestine; F: Small intestinal extranodal marginal zone lymphoma of mucosa-associated lymphoid tissue.

The results in this group showed that B-cell-derived lymphomas (DLBCL, FL, MALT) shared the common feature of CD20 positivity, whereas TCL was mainly characterized by Bcl-2 positivity. Transformed lymphoma exhibited overlapping immunophenotypic features of both the primary lesion and high-grade transformation.

The Lugano staging system was adopted in this study to guide treatment selection and prognostic evaluation, based primarily on the extent of disease involvement (localized gastrointestinal tract, regional lymph nodes, or disseminated spread) and the presence of systemic “B symptoms”. In our cohort, stage II disease accounted for the largest proportion (44.0%), with the median stage being II. The International Prognostic Index was applied as one of the most widely utilized and critical prognostic tools in lymphoma. It integrates clinical features reflective of the patient’s overall condition and tumor biological behavior to accurately predict survival outcomes and treatment response. According to the International Prognostic Index scoring, the low-risk category represented the largest subgroup (58.0%), as detailed in Table 3.

Table 3 Staging and risk.

Characteristics	Classification	Number of cases	Percentage
Lugano stage	I	10	20%
	II	22	44%
	III	6	12%
	IV	12	24%
IPI risk	Low	29	58%
	Low-intermediate	10	20%
	High-intermediate	9	18%
	High	2	4%

IPI: International Prognostic Index.

Treatment and prognosis

Among the surgery group, 9 cases (22.5%) developed postoperative complications, including poor incision healing, infection, and fever. The occurrence of complications was not significantly correlated with staging or risk. However, all tumors in these cases were larger than 5 cm. Prognostic data were available for 22 cases in the surgical group. Prognostic analysis showed no statistically significant difference in outcomes between the surgery plus chemotherapy group and the chemotherapy only group, although the improvement rate was higher in the surgical group (Table 4). Correlation analysis of outcomes revealed no significant correlations, suggesting the need for further analysis with a larger sample size (Table 4).

Table 4 Treatment outcomes.

Characteristics	Classification	Number of cases	Percentage
Surgery group	Improvement	9	40.9%
	Stable	5	22.7%
	Deterioration	8	36.3%
Non-surgery group	Improvement	4	25.0%
	Stable	6	37.5%
	Deterioration	6	37.5%

DISCUSSION

This study was designed to analyze the clinicopathological characteristics of PGIL and offer clinical insights for its management. PGIL is a common site for extranodal lymphoma[3-6]. Lymphoma involving the gastrointestinal tract not only compromises nutritional status but may also lead to life-threatening complications such as hemorrhage or obstruction as the disease progresses. Therefore, elucidating the distribution of involved sites, histologic subtypes, and clinical manifestations of PGIL is critical for guiding clinical decision-making, facilitating early diagnosis, and improving treatment outcomes. In the present case series, intestinal involvement was more frequently observed. The distribution of gastrointestinal pathological subtypes is site-specific, with MALT and DLBCL predominantly occurring in the stomach, while FL and intestinal disease-associated TCL are more common in the small intestine[4,7]. In this group, abdominal pain was the predominant initial and concurrent symptom. This was often accompanied by nonspecific manifestations such as abdominal distension, nausea, vomiting, gastrointestinal bleeding, weight loss, and anemia. From a clinical perspective, patients often present with diverse and nonspecific symptoms due to the deep anatomical location and the lack of characteristic manifestations. Abdominal pain was the most common initial and concomitant symptom in this cohort, accompanied by nonspecific presentations such as bloating, nausea, vomiting, gastrointestinal bleeding, weight loss, and anemia. These nonspecific manifestations frequently contribute to diagnostic challenges, including misdiagnosis or delayed detection. Therefore, recognizing these symptom patterns may enhance clinical vigilance toward this disease and help reduce diagnostic delays.

Given the diagnostic challenges associated with PGIL, its accurate identification should be guided by the principle of multi-site integrated and complementary evaluation. Ultrasound has a low sensitivity for PGIL and should be combined with CT and other examinations such as endoscopic biopsy. In this group, the mass detection rate by CT was 94.3%, while the diagnostic confirmation rate by endoscopy was 91.5%. In particular, lesions causing unexplained gastrointestinal bleeding are frequently found in the small intestine, and comprehensive oral and anal enteroscopy is essential to assess the entire digestive tract. Direct endoscopic biopsy can enhance the diagnostic accuracy. However, there remains a risk of missed diagnosis in clinical practice. Some patients with small intestinal lymphoma in this group underwent only gastroscopy or colonoscopy, while others with colonic lymphoma only received gastroscopy. Therefore, in patients with unexplained gastrointestinal bleeding but negative initial examinations, the possibility of tumors in occult locations should always be considered. Comprehensive multi-site examination should be supplemented. Additionally, PET/CT provides critical guidance in evaluating metastasis, determining surgical feasibility, and restaging following treatment[8]. It frequently reveals occult metastatic lesions, thereby helping to compensate for the limitations of preoperative mono-imaging assessment[8].

Histological biopsy remains the gold standard for the diagnosis of PGIL[9]. The primary methods for obtaining tissue specimens include puncture, endoscopic sampling, and surgical resection biopsy[10]. Surgery offers the advantage of fully preserving the lesion’s original appearance, providing ample tissue for histological examination, which is a significant advantage over fine-needle aspiration or core needle biopsy[11]. Biopsy specimens obtained via fine-needle aspiration are sometimes insufficient for pathological evaluation, and core needle biopsy is only performed when excisional biopsy is not feasible[12,13]. This highlights the critical role of excisional biopsy in accurate pathological assessment and molecular testing.

Regarding histopathological subtypes, all cases in this group were non-Hodgkin's lymphoma, with DLBCL accounting for up to 52%. DLBCL is typically the most common pathological type of gastrointestinal lymphoma[12], with a minority originating from MALT transformation[14]. FL and MALT transformations were also observed in this study. Similarly, primary thyroid lymphoma predominantly consists of DLBCL (50%-70%) and MALT lymphoma (10%-50%)[15]. These findings further confirm the dominant role of these histopathological subtypes in gastrointestinal lymphoma, which enables more tailored therapeutic strategies depending on the specific subtype. In the pathological findings, this study showed that B-cell-derived lymphomas (DLBCL, FL, MALT) share the common feature of CD20 positivity, which is consistent with findings from previous literature[12]. Genetic testing and molecular pathology are driving the advancement of diagnostic and treatment models. Looking ahead, the integration of genetic testing and AI analysis into precision diagnostic and treatment models will become the direction for development[16,17].

Regarding therapeutic strategies, preoperative surgery for PGIL is demonstrated to be safe and feasible[14]. Furthermore, a combined modality of surgery and chemotherapy may lead to improved survival outcomes in affected patients. The results of this group showed that the surgical intervention plays a crucial role throughout the diagnosis, treatment, and management of life-threatening complications. Surgery facilitates complete resection of the primary lesion, providing rapid symptomatic relief (e.g., pain, bleeding, dysfunction) and reducing life-threatening complications. In cases of life-threatening complications such as obstruction, bleeding, or perforation, active surgical intervention may be the only life-saving emergency option available to patients. However, the patients’ ability to tolerate anesthesia and surgery depends on their overall health status, including cardiopulmonary function. For certain sites, such as gastric MALT lymphoma, the eradication of Helicobacter pylori is the first-line treatment[18]. In contrast, primary thyroid lymphoma is primarily treated with a combination of radiotherapy and chemotherapy[19]. For patients with advanced-stage disease, a multimodal approach involving radiotherapy, chemotherapy or targeted therapy is generally required. Moreover, surgery combined with chemotherapy significantly improves the survival outcomes of patients with primary extranodal lymphoma[20,21]. Current prognostic models aim to promote individualized treatment by integrating clinical, pathological, and molecular variables[22]. Postoperative complications are generally associated with factors such as advanced age and late-stage disease. Further research is needed to explore strategies that balance treatment efficacy and safety. The combination of surgery and chemotherapy has proven beneficial in improving survival outcomes, particularly for certain lymphoma subtypes and advanced-stage patients[23]. Even in cases of histological transformation, which typically carry a poor prognosis such as FL transforming into DLBCL)[24], complete remission can be achieved through surgical resection of the transformed lesion combined with chemotherapy. However, treatment selection should remain individualized[6,25].

In summary, through a systematic analysis of clinical data from patients with PGIL, this study has delineated the clinical characteristics, diagnostic approaches, treatment strategies, and prognostic factors associated with this disease, thereby providing valuable insights for clinical practice. Looking forward, advances in precision medicine - incorporating molecular profiling, genetic testing, and artificial intelligence - are expected to facilitate more individualized management of PGIL, ultimately leading to improved diagnostic accuracy and patient outcomes. However, as a retrospective study with a limited sample size from a single center, the present findings may be subject to potential regional and selection biases. Therefore, future large-scale, randomized, prospective studies are warranted to validate and extend these observations.

CONCLUSION

PGIL is commonly associated with abdominal pain as the main symptom, with DLBCL being the most prevalent subtype. Diagnosis requires a combination of imaging techniques, including ultrasound, CT, and endoscopic biopsy, with PET/CT playing a crucial role in evaluating metastasis. Surgery is essential for diagnostic sampling, lesion resection, symptom relief, and the management of life-threatening complications, and can improve the prognosis in combination with radiotherapy and chemotherapy.