Surgically treated primary malignant tumor of small bowel: A clinical analysis

Abstract

AIM: To evaluate the clinical presentation, treatment and survival of patients with primary malignant tumor of small bowel (PMTSB).

METHODS: Clinicopathologic data about 141 surgically treated PMTSB patients (91 males and 50 females) at the median age of 53.5 years (range 23-79 years) were retrospectively analyzed.

RESULTS: The most common initial clinical features of the patients were intermittent abdominal discomfort or vague abdominal pain (67.4%), abdominal mass (31.2%), bowel obstruction (24.1%), hemotochezia (21.3%), jaundice (16.3%), fever (14.2%), coexistence of bowel perforation and peritonitis (5.7%), coexistence of gastrointestinal bleeding and shock (5.0%), and intraabdominal bleeding (1.4%). Ileum was the most common site of tumor (44.7%), followed by jejunum (30.5%) and duodenum (24.8%). PMTSB had a nonspecific clinical presentation. Segmental bowel resection (n = 81) was the most common surgical procedure, followed by right hemi-colectomy (n = 15), pancreaticoduodenectomy (n = 10), and others (n = 19). Twenty-seven adenocarcinoma patients and 13 malignant lymphoma patients received adjuvant chemotherapy with 5-fluorouracil and cyclophosphamide, adriamycin, vincristine and prednisone, respectively. Information about 120 patients was obtained during the follow-up. The median survival time of PMTSB patients was 20.3 mo. The 1-, 3- and 5-year survival rate was 75.0% (90/120), 40.0% (48/120) and 20.8% (25/120), respectively. Adenocarcinoma was found in 73.7% (42/57), 21.1% (12/57) and 15.8% (9/57) of the patients, respectively. Gastrointestinal stromal tumor was observed in 80.0% (20/25), 72.0% (18/25) and 36.0% (9/25) of the patients, respectively. Carcinoid was detected in 100.0% (15/15), 80.0% (12/15) and 46.7% (7/15) of the patients, respectively. Malignant lymphoma was demonstrated in 69.2% (9/13), 30.8% (4/13) and 0% (0/13) of the patients, respectively.

CONCLUSION: En bloc resection is the principal therapy for most PMTSB and chemotherapy is the important treatment modality for malignant lymphoma and other malignant tumors of small bowel which cannot be radically removed.

Key Words: Small bowel; Malignant tumor; Diagnosis; Surgical treatment; Chemotherapy

INTRODUCTION

Primary malignant tumor of the small bowel (PMTSB) accounts for 2% of all gastrointestinal (GI) tumors and 1% of GI tumor-related deaths[1-5]. The small bowel is relatively resistant to carcinogenesis although it is considerably long (accounting for 70% of the whole digestive tract) and exposed to a wide variety of potentially noxious substances. In a review of over 11 000 primary GI malignant tumors, Martin[6] found that only 2.4%, 10.8%, 16.4%, and 70.3% of primary malignancies are originated from the small bowel, esophagus, stomach, and colorectum, respectively. However, the diagnosis of PMTSB is difficult, because its symptoms and signs are nonspecific at presentation. Therefore, it is usually discovered at its advanced stage and often needs a surgical intervention due to acute complications. Sometimes, it is occasionally found during other surgical procedures[6-11]. This retrospective study was to evaluate the clinical presentation, treatment and survival of PMTSB patients.

MATERIALS AND METHODS

Medical records of PMTSB patients at the tumor registry of our hospital between January 1988 and December 2007 were analyzed. Patients with tumor at the ampulla of Vater, pancreatic head, and ileocecal valve and metastatic cancer were excluded from the study. Only those with primary tumor arising from the duodenum, jejunum, or ileum were included.

One hundred and forty-one patients entered the study. Data on demography, clinical presentation, diagnosis, surgical treatment, histopathological findings, postoperative course, and survival time were collected from each patient. The TNM staging classification (AJCC system, 6th edition)[12] was used to classify the extent of adenocarcinoma and carcinoid based on the histopathological and surgical reports. Stage I (T1-2, N0) was defined as tumor extending to muscularis propria, stage II (T3-4, N0) as tumor extending to subserosa, mesentery, and adjacent viscera, stage III (any T, N1) as tumor with regional lymph node metastasis, and stage IV (any T, any N, M1) as tumor with distant metastasis. Adenocarcinoma at stages I-IV was detected in 3, 14, 37 and 7 patients, respectively. Carcinoid at stages I-IV was found in 4, 8, 5 and 2 patients, respectively. Operation reports on the type, extent, and necessity of en bloc resection were reviewed. Operation was defined as radical if the tumor was completely removed both grossly and microscopically, and as palliative if the patients had distant metastasis at presentation, gross residual tumor at surgery, or positive margins microscopically. The World Health Organization standard grading system (well differentiated, moderately differentiated, poorly differentiated, and undifferentiated) was used to classify the histological types[12]. In addition, Ann Arbor staging classification[13] was used to classify the extent of malignant lymphoma. The diagnosis of gastrointestinal stromal tumor (GIST) was made as previously described[14-16].

Statistical analysis

Data were analyzed using the SPSS software (version 13.0; SPSS, Inc., Chicago, IL). Patient records were reviewed during the follow-up or by direct contact with the patients, their relatives, or office visit. Survival rate was calculated on the day of histological diagnosis until death or last follow-up with the Kaplan-Meier method for analysis of censored data. Cox regression analysis was performed to assess the independent prognostic significance of parameters. P < 0.05 was considered statistically significant.

RESULTS

Clinical and diagnostic features

Of the 141 PMTSB patients, 91 were male and 50 were female. Their median age was 53.5 years (range 23-79 years). Ileum was the most common site of PMTSB (44.7%), followed by jejunum (30.5%) and duodenum (24.8%). The most common clinical features of PMTSB patients at initial presentation were intermittent abdominal discomfort or vague abdominal pain (67.4%), abdominal mass (31.2%), bowel obstruction (24.1%), hemotochezia (21.3%), jaundice (16.3%), fever (14.2%), coexistence of bowel perforation and peritonitis (5.7%), coexistence of gastrointestinal bleeding and shock (5.0%), and intraabdominal bleeding (1.4%). Other symptoms were loss of appetite, diarrhea, anemia and loss of bodyweight (Table 1). The median time of symptoms was 2 mo (range 0-41 mo).

Table 1 Clinical symptoms and signs of PMTSB patients.

Symptoms	n (%)
Abdominal pain	95 (67.4)
Abdominal mass	44 (31.2)
Bowel obstruction	34 (24.1)
Hemotochezia	30 (21.3)
Jaundice	23 (16.3)
Fever	20 (14.2)
Bowel perforation coexistent peritonitis	8 (5.7)
Gastrointestinal bleeding and coexistent Shock	7 (5.0)
Intraabdominal bleeding	2 (1.4)

Others symptoms: Loss of appetite, diarrhea, anemia and loss of bodyweight. PMTSB: Primary malignant tumor of small bowel.

The preoperative diagnostic rate was decreased to 91.7% (11/12) at the duodenum, 70.6% (36/51) at the jejunum, and 60.3% (47/78) at the ileum, respectively. The most commonly used diagnostic techniques were ultrasonography (US) of the abdomen (90.1%), followed by computed tomography (CT) of the abdomen (80.1%), upper gastrointestinal radiography (31.9%) and upper endoscopy (25.5%). Additional techniques included ultrasonography, endoscopic retrograde cholangiopancreatography (ERCP), superior mesenteric arteriography, colonoscopy and bone scanning.

All the patients were diagnosed histopathologically after operation. Of the 141 patients, 61 (43.3%) were diagnosed as adenocarcinoma, 28 (19.8%) as GIST, 17 (12.1%) as carcinoid, 14 (9.9%) as malignant lymphoma, 10 (7.1%) as leiomyosarcoma, 6 (4.3%) as malignant melanoma, 3 (2.1%) as malignant neurilemmoma, and 2 (1.4%) as fibrosarcoma, respectively (Table 2). Twenty-three of the 61 patients (37.7%) who underwent curative resection were found to have lymph node metastases after surgery, which were not suspected before operation.

Table 2 Histopathological type of PMTSB (n = 141).

Histopathological type	n (%)
Adenocarcinoma	61 (43.3)
GIST	28 (19.8)
Carcinoid	17 (12.1)
Malignant lymphoma	14 (9.9)
Leiomyosarcoma	10 (7.1)
Malignant melanoma	6 (4.3)
Malignant neurilemmoma	3 (2.1)
Fibrosarcoma	2 (1.4)

GIST: Gastrointestinal stromal tumor.

Surgical procedures

Of the 141 patients who underwent surgical intervention, 31 (22.0%) had emergency operation and 110 (78.0%) had selective operation. The emergency indications included bowel obstruction (n = 24), gastrointestinal bleeding (n = 4) and bowel perforation (n = 3). The most commonly used surgical procedure was segmental bowel resection (65.3%), followed by right hemicolectomy (10.6%), and pancreaticoduodenectomy (7.1%). Other procedures included gastric bypass (n = 13), biopsy only (n = 5), feeding jejunostomy (n = 2), biliary bypass (n = 2), and enteric bypass (n = 2).

Of the 141 PMTSB patients, 104 (73.8%) received a radical resection, 37 (26.2%) underwent diagnostic or palliative operation. Of the 32 patients who underwent a palliative resection, 15 had synchronous distant metastasis (liver metastasis in 10 and peritoneal dissemination in 5) in small bowel and its mesentery (n = 11), retroperitoneum (n = 5) and ovary (n = 1) (Table 3).

Table 3 Surgical procedure for PMTSB patients.

Procedure	n (%)
Operative intervention
Emergent procedure	31 (22.0)
Elective procedure	110 (78.0)
Disease for emergent procedure
Bowel obstruction	24 (77.4)
Gastrointestinal bleeding	4 (12.9)
Bowel perforation	3 (9.6)
Surgical procedure
Segmental bowel resection	92 (65.3)
Right hemi-colectomy	15 (10.6)
Pancreaticoduodenectomy	10 (7.1)
Gastric bypass	13 (9.2)
Biopsy at laparotomy	5 (3.6)
Feeding jejunostomy	2 (1.4)
Biliary bypass	2 (1.4)
Enteric bypass	2 (1.4)
Radicality of procedure
Radical1	104 (73.8)
Palliative2	37 (26.2)

¹Radical resection: Negative margin, resection of all gross diseases, en bloc local resection when indicated;

²Palliative resection: Positive margin, gross residual disease.

Postoperative complications

Postoperative complications occurred in 21 (14.9%) patients, including pancreatic anastomotic leak in 7 (5.0%), wound infection in 6 (4.3%), prolonged gastric emptying in 3 (2.1%), subphrenic abscess in 3 (2.1%), and gastrointestinal bleeding from gastrojejunostomy in 2 (1.4%) as shown in Table 4. The median hospital stay time of patients was 13.2 d (range 8-60 d).

Table 4 Complications after operation.

Complications	n (%)
Pancreatic anastomotic leak	7 (5.0)
Wound infection	6 (4.3)
Prolonged gastric emptying	3 (2.1)
Subphrenic abscess	3 (2.1)
Gastrointestinal bleeding	2 (1.4)
Total	21 (14.9)

Postoperative adjuvant therapy

Of the 141 PMTSB patients, 40 (28.4%) received adjuvant chemotherapy after operation. However, adenocarcinoma was treated with 5-fluorouracil (5-FU) and malignant lymphoma was treated with cyclophosphamide, adriamycin, vincristine and prednisone (CHOP).

Recurrence patterns

Recurrence of the tumor was found in 32 (22.7%) of the 104 patients after radical resection (at a single site in 13 and at multiple sites in 19). The most common sites of recurrence were liver and lung (65.6%), peritoneal carcinomatosis (21.9%) and intestinal mesentery (12.5%). Nine patients (28.1%) underwent further operative intervention, 8 (25.0%) received chemotherapy and/or radiotherapy, and 4 (12.5%) received no further treatment. Of the 9 patients who underwent a second operation, 7 received a palliative procedure and died of the disease progression at a median time of 10 mo (range 2-18 mo) after operation.

Survival rate of PMTSB patients according to histology

Information was obtained during the follow-up of 120 patients with PMTSB including adenocarcinoma (n = 57), GIST (n = 25), carcinoid (n = 17), malignant lymphoma (n = 17), leiomyosarcoma (n = 3) and malignant melanoma (n = 1). The median survival time of PMTSB patients was 20.3 mo. The 1-, 3- and 5-year survival rate was 75.0% (90/120), 40.0% (48/120) and 20.8% (25/120), respectively. Adenocarcinoma was detected in 73.7% (42/57), 21.1% (12/57), and 15.8% (9/57) of the patients, respectively. GIST was observed in 80.0% (20/25), 72.0% (18/25) and 36.0% (9/25) of the patients, respectively. Carcinoid was found in 100.0% (15/15), 80.0% (12/15) and 46.7% (7/15) of the patients, respectively. Malignant lymphoma was shown in 69.2% (9/13), 30.8% (4/13) and 0% (0/13) of the patients, respectively. In addition, 3 leiomyosarcoma patients had a survived time of 15, 39 and 71 mo, respectively. One malignant melanoma patient survived for 18 mo (Table 5).

Table 5 Survival rate of PMTSB patients n (%).

Histopathological type	Loss of follow-up cases (n)	1-yr survival rate	3-yr survival rate	5-yr survival rate
Adenocarcinoma (n = 61)	4	42 (73.7)	12 (21.0)	9 (15.8)
GIST (n = 28)	3	20 (80.0)	18 (72.0)	9 (36.0)
Carcinoid (n = 17)	2	15 (100.0)	12 (80.0)	7 (46.7)
Lymphoma (n = 14)	1	9 (9.25)	4 (30.8)	0
Other tumors(n = 21)	11	3 cases of leiomyosarcoma survived (15, 39 and 71 mo, respectively) and 1 case of malignant melanoma survived 18 mo

DISCUSSION

Presentation and diagnosis of PMTSB

PMTSB is a rare malignancy. Most PMTSB patients have nonspecific clinical symptoms and signs[1-6]. In this series, the most frequent symptoms were abdominal pain (67.4%), abdominal mass (31.2%) and bowel obstruction (24.1%), followed by hemotochezia (21.3%), jaundice (16.3%), fever (14.2%), coexistence of bowel perforation and peritonitis (5.7%), coexistence of gastrointestinal bleeding and shock (5.0%), and intraabdominal bleeding (1.4%). The symptoms are similar to the reported findings[1-3,10,11]. The median time of delayed diagnosis in our series was 2 mo.

Preoperative diagnosis of PMTSB is often difficult. In our series, the preoperative diagnostic rate was decreased to 91.7% at the duodenum, 70.6% at the jejunum, and 60.3% at the ileum, respectively. The most commonly used diagnostic techniques were US, CT, upper gastrointestinal radiography and upper endoscopy. Endoscopic biopsy proved that most duodenal tumors (91.7%) in our study were malignant before surgery, suggesting that the more distal the tumor is, the more difficult the preoperative diagnosis is[16,17]. Since the accuracy of CT staging for small bowel adenocarcinoma is 47%-61%, it is only used in the detection of mesenteric infiltration and regional lymphadenopathy[16-20].

Hatzaras et al[1] showed that carcinoid tumor is the most common intestinal cancer, followed by adenocarcinoma. Our data demonstrate that ileum is the most common site (44.7%), followed by jejunum (30.5%), duodenum (24.8%), and that the most prevalent histological type is adenocarcinoma (43.3%), followed by GIST (19.8%), carcinoid (12.1%), malignant lymphoma (9.9%), leiomyosarcoma (7.1%), malignant melanoma (4.3%), malignant neurilemmoma (2.1%) and fibrosarcoma (1.4%), indicating that carcinoid tumor is more frequently found in ileum than adenocarcinoma in duodenum[1,8,21-23].

Laparoscopy can help to stage small bowel malignant tumor; serosal infiltration, retroperitoneal fixation, lymph node metastasis and ascites[23]. Laparoscopy can accurately assess and stage gastric adenocarcinoma[24,25], thus avoiding unnecessary laparotomy for those with no indication for palliative surgery. However, laparoscopy is not appropriate for patients with obstruction or bleeding.

Treatment strategy and surgical procedure

Treatment of PMTSB is mainly based on its histopathological type, location and extent. If the tumor is located at jejunum or ileum, an aggressive segmental resection and primary anastomosis are indicated[2]. If the tumor is located at ileum around the ileocolon junction, ileocolonic resection or right hemi-colectomy is indicated[3]. The optimal resection extent of duodenal tumor has not been defined. Some authors advocate pancreaticoduodenectomy (PD) for all patients with malignant tumor of the duodenum, including those located at the third and fourth portions to ensure adequate en bloc resection[6]. If the tumor is an adenocarcinoma or a carcinoid, en bloc resection and systemic lymph node dissection are indicated[7]. If the tumor is a histopathologically proven malignant lymphoma before operation, systemic chemotherapy is the first choice of treatment and surgery selection is only indicated for those with bowel obstruction, perforation and bleeding[26-29]. The value for routine extensive resection of adenocarcinoma or carcinoid has been recently challenged[2,3,7,8,26,27]. Others support PD for proximal duodenal carcinoma, but segmental resection for tumor of the third and fourth portions[27-29]. In our patients, the most commonly used surgical procedure was segmental bowel resection (65.3%), followed by right hemicolectomy (10.6%), PD (7.1%) and others.

Of the 61 patients (37.7%) who underwent curative resection, 23 were found to have lymph node metastasis after surgery, which was not suspected before operation. It has been shown that laparoscopy can help to stage GI cancer, thus avoiding unnecessary laparotomies. However, laparoscopy is not appropriate for patients with obstruction or bleeding or for those with no indication for palliative surgery[24,25]. In this series, 31 (22.0%) patients required urgent laparotomy for intestinal obstruction, gastrointestinal bleeding, or perforation.

It has been reported that chemotherapy may be beneficial for PMTSB, but optimal chemotherapy and the degree of benefit remain to be defined[27,30,31]. Bakaeen et al[30] showed that chemotherapy with lomustine (CCNU), 5-FU, either alone or in combination with other therapies, can considerably improve symptoms and hormone level, and moderately inhibit tumor regression, particularly in patients with metastatic gastroenteropancreatic neuroendocrine tumors with minimal adverse effects[30,31]. In our study, 40 patients (28.4%) received adjuvant chemotherapy after operation. However, adenocarcinoma and malignant lymphoma were treated with 5-FU and CHOP, respectively.

Surgical outcome

It has been reported that the 5-year survival rate of PMTSB patients after curative resection is 32%-47%[1,20-22,32,33]. Howe et al[31] have reported that the 5-year survival rate of patients after curative resection of localized, regional and distant metastatic PMTSB is 47.6%, 31% and 3.9%, respectively, which is similar to that of our patients. The median survival time of the 141 PMTSB patients was 20.3 mo. The 1-, 3- and 5-year survival rate was 75.0%, 40.0% and 20.8%, respectively. Adenocarcinoma was detected in 73.7%, 21.1% and 15.8% of the patients, respectively. GIST was found in 80.0%, 72.0% and 36.0% of the patients, respectively. Carcinoid was observed in 100.0%, 80.0% and 46.7% of the patients, respectively. Malignant lymphoma was demonstrated in 69.2%, 30.8% and 0% of the patients, respectively.

The site, clinical stage, and histological type do not influence the survival time of PMTSB patients. Howe et al[31] reported that the median survival time of patients with tumors of duodenum, jejunum and ileum is 16.9 and 28-31 mo, respectively. In our study, 32 of 104 patients (22.7%) had distant metastasis or intra-abdominal carcinomatosis at presentation, which is consistent with the reported findings[1,5,7,11,26]. Ito et al[7] reported that the 5-year survival rate of T1/T2 and T3/T4 tumor patients is 82% and 58%, respectively (P < 0.05). In contrast, Bakaeen et al[30] found that T stage can not predict the survival time of PMTSB patients. Curative resection, however, may not be possible owing to the late diagnosis of PMTSB. Dabaja et al[29] also reported that the 5-year survival rate of patients with PMTSB at stage IV is 5%, which is much lower than that of those with PMTSB at stages I-III (36%). Although carcinoid is usually silent and diagnosed at its advanced stage, carcinoid patients have a good prognosis and a long survival time after effective treatment[2,5,7,34,35].

In summary, en bloc resection is the principal procedure for most PMTSB patients and chemotherapy is the important treatment modality for malignant lymphoma and other small bowel malignant tumors with no indication for radical resection.

COMMENTS

Background

Although small bowel is considerably long and exposed to a wide variety of potentially noxious substances, it is relatively resistant to carcinogenesis. Primary malignant tumor of the small bowel (PMTSB) accounts for 2% of gastrointestinal (GI) tumors and 1% of gastrointestinal tract cancer- related deaths. In a review of over 11 000 primary GI malignant tumors, Martin found that only 2.4%, 10.8%, 16.4% and 16.4% are originated from the small bowel, esophagus, stomach, and colorectum, respectively. However, the diagnosis of PMTSB is difficult, because its symptoms and signs are nonspecific at presentation. Therefore, PMTSB is usually discovered at its advanced stage and often needs surgical intervention due to acute complications. Sometimes, it is occasionally found during other surgical procedures. The objective of this study was to evaluate the clinical presentation, treatment and survival of PMTSB patients.

Research frontiers

How to improve the early diagnosis and treatment of PMTSB is a hotspot in recent studies.

Innovations and breakthroughs

This study evaluated the clinical presentation, treatment and survival time of PMTSB patients.

Applications

The retrospective analysis of the clinical presentation, treatment and survival time of PMTSB patients showed that PMTSB can be made early diagnosed and can thus be rationally treated.

Peer review

This is a rather large series of PMTSB patients from a single center. The data provided in this study contribute to the early diagnosis and treatment of PMTSB.