Primary gastrointestinal (GI) tract malignancies in children are extremely uncommon compared with benign tumors. The overall incidence of these 2 categories of tumors is reportedly <5% of all of the pediatric neoplasms (1–3); however, the true incidence of benign tumors in the GI tract remains unknown in both developing and developed countries. In previous reports, intestinal polyps have accounted for the majority of benign alimentary tract tumors, whereas teratomas, lipomas, hemangiomas, and others are relatively rare; however, primary GI malignancies constitute approximately 1% of pediatric cancers (4). According to the Cancer Registry Annual Report of the Department of Health in Taiwan, only an average of 13 new cases of primary GI malignancies per year have been reported in patients younger than 19 years from 1999 to 2008 (5).
Although primary pediatric GI malignancies are exceedingly unusual, early diagnosis is still crucial because the prognosis is poorer compared with benign tumors. The diagnosis of primary GI malignancies is usually delayed or made postoperatively (6). The rare GI malignancies have even a worse prognosis in children because of a high incidence of unfavorable histotypes and advanced clinical stage at onset (7).
The purpose of the present study was to retrospectively review and discuss the different clinical manifestations and characteristics between primary GI malignancies and benign tumors in children, to increase awareness regarding their early detection and differential diagnoses.
The medical records of all of the patients ages 18 years and younger with primary GI tract tumors and treated between January 1984 and December 2009 at Mackay Memorial Hospital, Taipei, Taiwan, were retrospectively reviewed. Mackay Memorial Hospital is a tertiary hospital center in the north of Taiwan with an average of 272,400 outpatients and 77,500 inpatients (18 years and younger) per year. Data extracted from the charts included demographic information, family history, duration from onset to diagnosis, clinical features, laboratory data, diagnostic workup, location and pathology of the tumors, duration of hospitalization, and outcomes. Diagnostic modalities included barium enema, abdominal ultrasonography, upper or lower endoscopy, and computed tomography. The institutional review board registry approved the study.
The patient population was divided into 2 groups: the B group, or all of the children with GI tract benign tumors, and the M group, or those with malignancies. Metastatic tumors and other large infiltrating tumors arising from the retroperitoneum or peritoneum rather than from the alimentary tract were excluded.
All of the statistical analyses were performed using SPSS version 18.0 (SPSS Inc, Chicago, IL). Results were reported as median and range. Correlations between categorical variables in both population groups were compared using the χ2 test or the Fisher exact test, whereas continuous variables were analyzed using the Mann-Whitney U test. A P < 00.05 was considered statistically significant.
Eighty-six patients (50 boys and 36 girls) were diagnosed as having primary GI tract tumors during the 26-year study period (Table 1). Of the 86 patients, 62 (72%) had benign tumors (B group) and 24 (28%) had malignancies (M group). There were 10 (12%) primary tumors in the stomach, 15 (17%) in the small intestine (including the ileocecum), and 61 (71%) in the colorectum (Fig. 1). The median duration of hospitalization was significantly longer in the M group (19 days vs 5 days, P < 00.05). Overall, 61 (98%) patients in the B group were alive by the end of the study. Their median duration of follow-up was 5 months (range 7 days–17 years). One (2%) patient had malignant change to fibrosarcoma after 1 year and died due to subsequent complications.
In contrast, 9 (38%) patients in the M group were alive by the end of the study. Their median duration of follow-up was 9 years (range 1–21 years). Eleven (46%) patients died and 4 (16%) were lost to follow-up after surgery.
Thirty-four (55%) patients were boys and 28 (45%) were girls (male-to-female ratio, 1.2:1). The median age of the patients on diagnosis was 4 years (range 4 months–18 years), with most presenting at younger than age 10 years (45/62, 73%). The median duration of symptoms before diagnosis was 2 months (range 1 day–5 years). Only 1 (2%) patient was diagnosed as having gastric teratoma. The rest of the patients (61/62, 98%) had GI polyps arising from the stomach (5/62, 8%), small intestine (5/62, 8%), and colorectum (51/62, 82%). Of the 62 patients, 48 (77%) presented with hematochezia, compared with only 1 (4%) of 24 in the M group (P < 00.05). None of the patients had body weight loss, anorexia, fever, or constipation. Laboratory data showed that anemia was less common in the B group (47%) than in the M group (63%); however, the difference was not statistically significant (P = 00.191). The median length of hospitalization time was 5 days (range 1–52 days).
Most of the GI polyps in this series were solitary (82%), although 2 patients had 2 polyps, 2 had 3 polyps, and 7 had multiple polyps. Of the 7 patients who met the criteria of GI polyposis syndrome, 4 had juvenile polyposis syndrome, 2 had familial adenomatous polyposis (FAP), and 1 had gastric polyposis. Three of them had a family history of GI polyposis syndrome. Seven patients with polyposis were all considered to have premalignant conditions. Two eventually developed malignancies, including 1 who had gastric polyposis developed gastric adenocarcinoma and the other had FAP developed fibrosarcoma.
Sixteen (67%) patients were boys and 8 (33%) were girls (male-to-female ratio, 2:1). Their median age at diagnosis was 16 years (range 5–18 years). Most of the patients were older than 10 years (19/24, 79%) at diagnosis. The median duration from onset to diagnosis was 1 month (range 1 day–3 years). Categorization of the malignancies by site was shown in Figure 1. Pathology showed 3 gastric adenocarcinomas, 3 small intestinal cancers (including 1 jejunal adenocarcinoma, 1 jejunal fibrosarcoma, and 1 ileal leiomyosarcoma), 8 colorectal adenocarcinomas, 8 GI lymphomas (including 4 Burkitt lymphomas, 3 diffuse large B-cell lymphomas, and 1 diffuse lymphoblastic T-cell lymphoma), and 2 appendiceal carcinoid tumors.
Symptoms such as abdominal pain (79%), palpable abdominal mass (25%), body weight loss (25%), anorexia (25%), and fever (21%) were significantly more likely to occur in the M group than in the B group (P < 00.05). Laboratory data showed that approximately two-thirds of the patients had anemia at diagnosis. Of the 8 patients with colorectal carcinoma, only 2 (25%) were positive for stool occult blood. Moreover, 2 (25%) cases had elevated carcinoembryonic antigen (CEA) levels. Among the patients with GI lymphoma, 4 (50%) had high levels of serum lactate dehydrogenase. The median duration of hospitalization for this group was 19 days (range 4–94 days). Family history was noncontributory in the M group.
Primary GI tract tumors, including benign tumors and malignancies, are unusual in pediatric patients. To date, there is still a lack of well-documented studies. Benign tumors, especially GI polyps, account for the majority of cases (8–11). In contrast to adults, primary GI malignancies are exceedingly rare in children. In Pickett and Briggs's 1967 review, GI malignancies constituted only about 1% of pediatric cancers (4). Bethel et al (2) reported 55 cases of primary intestinal malignancies of 4547 tumor registrants (1.2%) diagnosed during a 44-year period. Zhuge et al and Yang et al (12,13) reported even lower incidences. Despite the rarity of primary GI malignancies, early detection is still important because of their relatively more advanced clinical stage and poorer outcomes than benign tumors.
The present study showed that most patients in the benign group were younger than 10 years, whereas the patients in the malignant group were mostly older than 10 years, and the difference was statistically significant (P < 0.05). Recently published studies report the age distribution in the patients with GI tract malignancies to be more heavily weighted toward adolescents, with almost all of the patients older than 10 years, with 15- to 19-year-olds comprising the largest group at diagnosis (7,12,13). Of the 24 patients in the malignant group, 15 (62.5%) were older than 15 years. In contrast, most of the patients in the benign group, particularly those with intestinal polyps, were diagnosed in the first decade of life, with peak incidence between 2 and 6 years of life. This finding is also consistent with previous studies (9–11).
Bethel et al (2) reported 55 children with primary alimentary tract tumors, with 35 boys and 19 girls (male-to-female ratio, 1.8:1). Khurshed et al (7) reported a male-to-female ratio of approximately 3:1 (44 boys, 16 girls) in their 60 cases. There was a male predominance in both study groups, although the difference was not statistically significant (P > 00.05).
The distribution of tumors in the present study is obviously different between the benign and malignant groups. Macroscopically, most of the cases in the benign group arose from colorectal sites, especially the rectosigmoid colon. In contrast, the location of primary GI tract malignancies in was almost equally distributed between the small intestine and the colorectum.
Duration From Onset to Diagnosis
Some authors have postulated that primary GI malignant tumors in children are more poorly differentiated and advanced in stage than in adults on diagnosis. This is most likely due to a failure to consider potential GI malignancy in pediatric patients, thus leading to delays in diagnosis that range from months to years (14–20). The median duration from onset to diagnosis of the present study is 1 month in the malignant group; however, interestingly, the median duration in the benign group (2 months) is longer than that of the malignant group, although without a statistically significant difference (P > 0.05). One possible explanation is that the symptoms of patients in the benign group are too mild and unapparent to be detected early. The lack of awareness of these patients has therefore led to delays in seeking medical advice. Moreover, most tumors of the malignant group are already advanced and progressive when discovered. If these tumors mimic presentations of intussusception, appendicitis, and peritonitis, for which urgent intervention is required, there is the possibility of earlier diagnosis in patients in the malignant group than in patients in the benign group.
The clinical presentations of GI tumors in children are variable and nonspecific. The presentations that have frequently been described in previous studies are abdominal pain, rectal bleeding, vomiting, and abdominal mass (2,6).
In the present study, the most common symptom of all of the patients was hematochezia (49/86, 57%), 48 of which were in the benign group and only 1 in the malignant group, with a statistically significant difference (P < 0.05). This may be because tumors from the benign group are mostly located in the left colon, especially the rectosigmoid colon. This increases the time and surface area of friction between the tumor and stool, eventually causing the tumor to bleed. In contrast to adult patients with GI malignancies, especially colorectal cancer, patients in the malignant group in the present study do not frequently have hematochezia or rectal bleeding. The next most common presentations of patients in the present study are abdominal pain and vomiting. Palpable abdominal masses account for only 9% of patients. Of patients in the malignant group, 25% have palpable abdominal mass, body weight loss, and anorexia, followed by 21% with fever before diagnosis. These symptoms are significantly different compared with those of the benign group (P < 0.05). Thus, clinicians should pay more attention to patients presenting with these symptoms because their likelihood of having malignancy is higher.
About half to two-thirds of patients in the benign and malignant groups have anemia at diagnosis; however, few patients have to undergo blood transfusions in both groups. In the malignant group, only 25% of patients with colorectal carcinoma have elevated CEA levels. Although CEA has been widely used as a tumor marker in adults, little is known about its clinical value in children (21). Some authors believe that CEA determination possesses neither sensitivity nor specificity for use as a screening test for suspected colorectal carcinoma in children (22,23); however, others suggest a strong correlation between preoperative CEA level and classification of Duke stage. More advanced Duke stage has been associated with significantly higher preoperative CEA levels (24).
In the present series, 87.5% of patients with colorectal carcinoma are Duke stage C or D at diagnosis. It seems, therefore, that a more advanced stage is far less likely to be associated with elevated serum CEA levels in the present study; however, serum CEA level may still be a helpful tool for early suspicion of colorectal carcinoma from benign tumors if the level is elevated at diagnosis (25). On the contrary, half of the study patients with GI lymphoma have high levels of lactate dehydrogenase, which may be directly related to total tumor burden and may be a prognostic factor (6,26). Thus, this serum marker may also be a helpful parameter in differentiating GI tumors.
Other Relevant Polyposis
GI polyposis syndrome is a relatively rare condition. Careful medical and family history evaluations, thorough physical examination, and regular follow-up should be given because of the potential for malignant change (10,27,28). None of the patients with solitary juvenile polyps in the present study had a long-term risk of malignant change. In contrast, 2 (29%) of the patients with polyposis eventually developed malignancies with poor prognosis. One of them was a patient who previously had gastric polyposis and developed gastric adenocarcinoma years later. The other had FAP before developing fibrosarcoma, and died after surgery due to subsequent complications.
Goldthorn et al (1) emphasized that gastric adenocarcinomas develop in 3 ways in pediatric patients: de novo, following treatment of gastric lymphoma, and as part of polyposis syndrome. In polyposis syndromes, there is a predilection for adenomatous polyps to become malignant in certain anatomic sites. To date, there have been only a few cases of FAP-associated sarcoma, and the etiology is still unknown. Genetic factors may be one of the contributory factors (29). Lastly, intestinal fibrosarcoma is also extremely rare, with only a few case reports published to date (30–32).
The study had several limitations. First, it had a relatively small sample size, all coming from a single tertiary center. The study period should also have been longer and included more participating medical institutions. Furthermore, it lacked integrated patients’ clinical information. All of the patient information had been obtained from the patients themselves and based on the experience of physicians; hence, there may be a lack of consistency. Lastly, the patients’ clinical features may be underestimated if these are not documented by physicians on the medical records.
Primary GI tract malignancies are extremely uncommon compared with benign tumors in children. Because of this rarity, a high index of suspicion should be kept in mind when treating children who present with recurrent vomiting and persistent abdominal pain of unknown cause, especially those older than 10 years and associated with alarming symptoms like palpable abdominal mass, body weight loss, anorexia, and nonspecific fever. The different manifestations between malignant and benign GI tumors may help pediatricians detect high-risk patients early to improve their outcomes and survival. Patients with evidence of GI polyposis syndrome should also be carefully investigated and monitored because of their potential risk for malignancy.
1. Goldthorn JF, Canizaro PC. Gastrointestinal malignancies in infancy, childhood, and adolescence. Surg Clin North Am 1986; 66:845–861.
2. Bethel CA, Bhattacharyya N, Hutchinson C, et al. Alimentary tract malignancies in children. J Pediatr Surg 1997; 32:1004–1008.
3. Ladd AP, Grosfeld JL. Gastrointestinal tumors in children and adolescents. Semin Pediatr Surg 2006; 15:37–47.
4. Pickett LK, Briggs HC. Cancer of the gastrointestinal tract in childhood. Pediatr Clin North Am 1967; 14:223–234.
6. Skinner MA, Plumley DA, Grosfeld JL, et al. Gastrointestinal tumors in children: an analysis of 39 cases. Ann Surg Oncol 1994; 1:283–289.
7. Khurshed A, Ahmed R, Bhurgri Y. Primary gastrointestinal malignancies in childhood and adolescence—an Asian perspective. Asian Pac J Cancer Prev 2007; 8:613–617.
8. Jass JR, Williams CB, Bussey HJ, et al. Juvenile polyposis—a pre-cancerous condition. Histopathology 1988; 13:619–630.
9. Corredor J, Wambach J, Barnard J. Gastrointestinal polyps in children: advances in molecular genetics, diagnosis, and management. J Pediatr 2001; 138:621–628.
10. Attard TM, Young RJ. Diagnosis and management of gastrointestinal polyps: pediatric considerations. Gastroenterol Nurs 2006; 29:16–22.
11. Durno CA. Colonic polyps in children and adolescents. Can J Gastroenterol 2007; 21:233–239.
12. Zhuge Y, Cheung MC, Yang R, et al. Pediatric intestinal foregut and small bowel solid tumors: a review of 105 cases. J Surg Res 2009; 156:95–102.
13. Yang R, Cheung MC, Zhuge Y, et al. Primary solid tumors of the colon and rectum in the pediatric patient: a review of 270 cases. J Surg Res 2010; 161:209–216.
14. Contreary K, Nance FC, Becker WF. Primary lymphoma of the gastrointestinal tract. Ann Surg 1980; 191:593–598.
15. Sandlund JT, Downing JR, Crist WM. Non-Hodgkin's lymphoma in childhood. N Engl J Med 1996; 334:1238–1248.
16. Chantada GL, Perelli VB, Lombardi MG, et al. Colorectal carcinoma in children, adolescents, and young adults. J Pediatr Hematol Oncol 2005; 27:39–41.
17. Kravarusic D, Feigin E, Dlugy E, et al. Colorectal carcinoma in childhood: a retrospective multi-center study. J Pediatr Gastroenterol Nutr 2007; 44:209–211.
18. Curtis JL, Burns RC, Wang L, et al. Primary gastric tumors of infancy and childhood: 54-year experience at a single institution. J Pediatr Surg 2008; 43:1487–1493.
19. Ferrari A, Rognone A, Casanova M, et al. Colorectal carcinoma in children and adolescents: the experience of the Istituto Nazionale Tumori of Milan, Italy. Pediatr Blood Cancer 2008; 50:588–593.
20. Saab R, Furman WL. Epidemiology and management options for colorectal cancer in children. Paediatr Drugs 2008; 10:177–192.
21. Wanebo HJ, Rao B, Pinsky CM, et al. Pre-operative carcino-embryonic antigen level as a prognostic indicator in colorectal cancer. N Engl J Med 1978; 299:448–451.
22. Angel CA, Pratt CB, Rao BN, et al. Carcino-embryonic antigen and carbohydrate 19-9 antigen as markers for colorectal carcinoma in children and adolescents. Cancer 1992; 69:1487–1491.
23. Shih HH, Lu CC, Tiao MM, et al. Adenocarcinoma of the colon in children presenting as abdominal pain: report of two cases. Chang Gung Med J 2002; 25:349–354.
24. Wolmark N, Fisher B, Wieand HS, et al. The prognostic significance of pre-operative carcino-embryonic antigen levels in colorectal cancer. Results from NSABP (National Surgical Adjuvant Breast and Bowel Project) clinical trials. Ann Surg 1984; 199:375–382.
25. Taguchi T, Suita S, Hirata Y, et al. Carcinoma of the colon in children: a case report and review of 41 Japanese cases. J Pediatr Gastroenterol Nutr 1991; 12:394–399.
26. Murphy SB, Fairclough DL, Hutchison RE, et al. Non-Hodgkin's lymphomas of childhood: an analysis of the histology, staging, and response to treatment of 338 cases at a single institution. J Clin Oncol 1989; 7:186–193.
27. Barnard J. Screening and surveillance recommendations for pediatric gastrointestinal polyposis syndromes. J Pediatr Gastroenterol Nutr 2009; 48 (Suppl 2):S75–S78.
28. Brosens LA, van Hattem A, Hylind LM, et al. Risk of colorectal cancer in juvenile polyposis. Gut 2007; 56:965–967.
29. Jannasch O, Dombrowski F, Lippert H, et al. Rare coincidence of familial adenomatous polyposis and a retroperitoneal fibromyxoid sarcoma: report of a case. Dis Colon Rectum 2008; 51:477–481.
30. Van Niekerk ML, Nel WA, Slavik T. Infantile fibrosarcoma of the ileum presenting with congenital bowel obstruction. J Pediatr Surg 2010; 45:461–462.
31. Islam S, Soldes OS, Ruiz R, et al. Primary colonic congenital infantile fibrosarcoma presenting as meconium peritonitis. Pediatr Surg Int 2008; 24:621–623.
32. Buccoliero AM, Castiglione F, Degl’Innocenti DR, et al. Congenital/infantile fibrosarcoma of the colon: morphologic, immuno-histochemical, molecular, and ultrastructural features of a relatively rare tumor in an extraordinary localization. J Pediatr Hematol Oncol 2008; 30:723–727.