Small bowel adenocarcinoma and Crohn's disease: Any further ahead than 50 years ago?

Abstract

This review of the literature on small bowel carcinoma associated with Crohn’s disease specifically addresses the incidence, risk factors, and protective factors which have been identified. It also reviews the clinical presentation, the current modalities of diagnosis, the pathology, treatment, and surveillance. Finally, the prognosis and future direction are addressed. Our experience with small bowel adenocarcinoma in Crohn’s disease is reported. Readers will be provided with a better understanding of this rare and often poorly recognized complication of Crohn’s disease.

Core tip: This review of the literature on small bowel carcinoma associated with Crohn’s disease specifically addresses the incidence, risk factors, and protective factors which have been identified. It also reviews the clinical presentation, the current modalities of diagnosis, the pathology, treatment, and surveillance. Finally, the prognosis and future direction are addressed. Our experience with small bowel adenocarcinoma in Crohn’s disease is reported. Readers will be provided with a better understanding of this rare and often poorly recognized complication of Crohn’s disease.

INTRODUCTION

Over the past several decades it has become increasingly recognized that small bowel adenocarcinoma is an undeniable complication of Crohn’s disease of the small intestine. The exact magnitude of this risk is virtually impossible to determine. The first case of small bowel carcinoma in Crohn’s disease was reported by Ginzburg in 1956. Since then, there have been countless published case reports as well as numerous retrospective reviews and cohort studies which have attempted to define the occurrence of small bowel carcinoma in Crohn’s disease.

This review of the literature on small bowel carcinoma associated with Crohn’s disease specifically addresses the incidence, risk factors, and protective factors which have been identified. It will also review the clinical presentation, the current modalities of diagnosis, the pathology, treatment, and surveillance. Finally, the prognosis and future direction will be addressed. Our experience with small bowel adenocarcinoma in Crohn’s disease is reported. Readers will be provided with a better understanding of this rare and often poorly recognized complication of Crohn’s disease.

INCIDENCE

The relative risk of developing carcinoma of the small bowel in patients with Crohn’s disease has been estimated to range from 6 to 320[1]. Jess et al[2] studied a population-based cohort of 374 patients with Crohn’s disease in order to determine the long term risk of intestinal and extra-intestinal malignancies. The risk of small bowel adenocarcinoma was increased by more than 60 fold as compared to the general population. On the basis of 5000 cases of Crohn’s disease reported in the world literature, Amman[3] calculated an association with carcinoma of 0.08%, which, he emphasized, was lower than the incidence of carcinoma of the small bowel in the absence of Crohn’s disease, which he estimated at 0.098% in 137124 autopsies. However, the incidence in Crohn’s disease may be higher than reported as the associated carcinoma may have been missed due to inadequate histopathologic study. These observational series serve to underline the possible fallacies of mass statistics collected from the literature.

Our review of the literature revealed 220 reported cases of adenocarcinoma associated with small bowel Crohn’s disease. Since we do not have the number of patients who suffered from Crohn’s disease of the small bowel from which those patients were extracted, the actual incidence of carcinoma associated with Crohn’s cannot be calculated. These numbers were obtained through a Medline search from the years 1975 to 2013. Some of the reports were reviews and hence there may be some duplication of cases in the series reported.

There is nevertheless a general consensus that the risk of developing small bowel adenocarcinoma is greater in patients with Crohn’s disease than in the general population. The exact magnitude of the increased risk is difficult to determine because information derived from population-based studies, case-controlled studies, physician surveys, and case reports have been combined.

Many reports have documented that adenocarcinoma of the small bowel is a complication of Crohn’s disease and this has been well reviewed by Kerber[4] and Frank[5]. Sometimes the patients who develop an adenocarcinoma are those with small bowel-limited Crohn’s disease but most often there is a combination of those with both small and large bowel Crohn’s disease. Patients with longstanding small bowel Crohn’s disease are thought to have an increased risk of small bowel carcinoma[6-9].

In a meta-analysis by von Roon et al[10], the relative risk of developing small bowel carcinoma in the 9642 patients was 28.37. The incidence rate of small bowel carcinoma in 12740 patients was 1.55 per 100000 patient years. In the review by Von Roon, the mean duration of Crohn’s disease before the onset of carcinoma was 9 years (range 0.8-41). The relative risk of developing small bowel carcinoma compared to the background population was higher in North America (RR = 41.23), the United Kingdom (RR = 40) and Scandinavia (RR = 21.3).

Shaukat et al[11] studied all cases of small bowel carcinoma in persons 67 years and older in the Surveillance Epidemiology and End Results catchment area. They identified 923 cases of small bowel carcinoma and 142273 controls and found a strong association between Crohn’s disease and small bowel carcinoma [odds ratio (OR) = 12.02]. The prevalence of Crohn’s disease in patients with small bowel carcinoma was low (1.6%) so the absolute risk remains low.

RISK AND PROTECTIVE FACTORS

Numerous risk factors for developing small bowel carcinoma in Crohn’s disease have been postulated in the literature. Many purported risk factors surfaced as a result of observed trends across case reports such as previous strictureplasty[12-15] and excluded/bypassed bowel segments[16-18]. Approximately 30% of reported small bowel carcinomas in Crohn’s disease occurred in patients who had bypassed loops[19]. This complication stresses the need to encourage resection rather than bypass. Lashner[20] reported a case control study of carcinoma of the small bowel in Crohn’s disease in which each case was matched to 4 randomly selected controls from an inflammatory bowel disease registry. The following factors were significantly associated with small bowel carcinoma in Crohn’s disease: (1) occupation, with three cases having had exposure to halogenated aromatic compounds and aliphatic amines, asbestos, and cutting oil solvents and abrasives; and (2) 6-mercaptopurine use (OR = 10.8).

Several studies reported that the risk of developing small bowel carcinoma was according to the anatomic location of the Crohn’s disease (Table 1). For example, the odds ratio of developing small bowel carcinoma was found to be much higher in 363 patients whose disease was confined to the small bowel (RR = 158.5) than in the 507 patients with ileocolic Crohn’s disease (RR = 83.8)[10].

Table 1.

Risk factors for small bowel cancer in Crohn’s disease

		Risk factors for small bowel cancer?
		Yes	No
Crohn's related risk factors
1	''Long" duration of CD	Greenstein et al[68] (mean 33.5 yrs)	Jess et al[73] Laukoetter et al[63]
		Greenstein et al[24] (mean 22 yrs)
		Jess et al[2] (median 16 yrs)
		Kamiya et al[69] (one case 18 yrs)
		Kersting et al[65] (one case 16 yrs)
		Kvist et al[70] (mean 29 yrs)
		Mellemkjaer et al[71] (mean 22 yrs)
		Michelassi et al[55] (mean 19.6 yrs)
		Mizushima et al[72] (mean 14 yrs)
		Munkholm et al[8] (mean 13.5 yrs)
		Palascak-Juif et al[27] (median 15 yrs)
		Petras et al[46] (mean 20 yrs)
		Ribeiro et al[1] (mean 26.5 yrs)
		Savoca et al[62] (one case 20 yrs)
		Sigel et al[48] (median 12 yrs)
		Solem et al[26] (median 21 yrs)
		Widmar et al[23] (mean 25.3 yrs)
2	Area of CD inflammation	Jess et al[2]	Laukoetter et al[63]
		Michelassi et al[55]
		Mizushima et al[72]
		Munkholm et al[8]
		Palascak-Juif et al[27]
		Ribeiro et al[1]
		Savoca et al[62]
		Solem et al[26]
		von Roon et al[10]
3	Jejunal CD	Lashner[20]	Palascak-Juif et al[27] Solem et al[26]
4	Strictures	Jaskowiak et al[15]	Solem et al[26]
		Kersting et al[65]
		Lakatos et al[74]
		Petras et al[46]
		Ribeiro et al[1]
5	Fistula	Kersting et al[65]	Lashner et al[20] Solem et al[26]
		Laukoetter et al[63]
		Ribeiro et al[1]
6	Bypassed segment	Greenstein et al[75]	Lashner[20]
			Palascak-Juif et al[27]
			Ribeiro et al[1]
			Solem et al[26]
7	CD Medications	1Lashner[20]	Canavan et al[76]
			Solem et al[26]
8	"Young" age	Freeman et al[60] (mean 45.7 yrs)	Jess et al[73] (median 66 yrs) Munkholm et al[8] (mean 70.5 yrs)
		Hoffman et al[7] (mean 46 yrs)
		Kersting et al[65] (one case 34 yrs)
		Laukoetter et al[63] (20 yrs earlier) Michelassi et al[55] (mean 47.7 yrs)
		Palascak-Juif et al[27] (median 47 yrs)
		Petras et al[46] (mean 46 yrs)
		Savoca et al[62] (mean 38 yrs)
		Sigel et al[48] (median 42 yrs)
		Widmar et al[23] (mean 55.4 yrs)
General risk factors
9	Gender		Jess et al[2] Petras et al[46]
10	Male	Lakatos et al[74]	Lashner[20] Palascak-Juif et al[27]
		Michelassi et al[55]
		Shaukat et al[11]
		Sigel et al[48]
		Ribeiro et al[1]
		Widmar et al[23]
11	Female	Freeman et al[60]
12	Black race	Shaukat et al[11]
13	Past corticosteroid use		Kaerlev et al[67]
14	Past use of radioactive medication		Kaerlev et al[67]
15	Liver disease (cirrhosis/hepatitis)		Kaerlev et al[67]
16	Gallstones		Kaerlev et al[67]
17	Previous cholecystectomy	Chen et al[64]	Kaerlev et al[67]
18	Prior history of peptic ulcer disease	Chen et al[64]	Kaerlev et al[67]
19	Celiac disease	Kaerlev et al[67]
20	Prior malignancy		Solem et al[26]
21	Blood type B		Chen et al[64]
22	Rh type		Chen et al[64]
23	Tobacco	Chen et al[64]	Chow et al[77]
		Lakatos et al[74]	Negri et al[25]
24	Alcohol	Chen et al[64]	Chow et al[77] Negri et al[25]
25	Diet	2Negri et al[25]
		3Chow et al[77]
26	Lower education level	Kaerlev et al[67]
27	Geographic location	4von Roon et al[10]
28	Hazardous occupation	5Lashner[20]
29	Marital status		Chen et al[64]
30	Religion		Chen et al[64]
31	Room type		Chen et al[64]

¹6-MP;

²Bread, pasta, rice, sugar, red meat;

³Red meat, salt-cured foods, smoked foods;

⁴North America higher risk;

⁵Exposure to halogenated aromatic compounds and aliphatic amines, asbestos, solvents, oils, abrasives. CD: Crohn’s disease.

Table 1 displays a review of the studied risk factors for small bowel carcinoma in Crohn’s disease in the literature, highlighting which authors agree and disagree with a purported risk factor. Clinicians should perhaps be more vigilant for small bowel carcinoma in patients with inflammation restricted to the small bowel versus ileocolic inflammation.

Protective factors against development of small bowel carcinoma in Crohn’s disease have been less frequently studied, but available information is cited in Table 2. In a study of 29 patients with Crohn’s disease and small bowel carcinoma Piton et al[21] found that small bowel resection and prolonged use of salicylates may protect against small bowel carcinoma in Crohn’s disease patients.

Table 2.

Protective factors for small bowel cancer in Crohn’s disease

		Protective factor against small bowel cancer?
		Yes	No
1	Diet	1Negri et al[25]	2Chow et al[77]
2	5-ASA	Piton et al[21]
		Solem et al[26]
3	CD medications		Canavan et al[76]
			Solem et al[26] (other than 5-ASA)

¹Coffee, fish, vegetables, fruit;

²Fruits and vegetables. 5-ASA: 5-aminosalicylic acid.

CLINICAL PRESENTATION

Obstruction is the most common presenting manifestation in small bowel carcinoma in Crohn’s disease with symptoms of nausea, vomiting and abdominal pain. Other possible presentations are hemorrhage, fistula, or perforation[22-24]. Unfortunately, all of these symptoms are hard to differentiate from those of a Crohn’s exacerbation, which partly explains the challenge of detecting small bowel carcinoma in this patient population and results in the majority of diagnoses being made at the time of operation or postoperatively. In fact, only a small minority (< 5%) is diagnosed preoperatively[22]. Furthermore, Collier et al[19] described that over 50% of small bowel carcinomas in resected Crohn’s disease segments were unsuspected or incidentally found by the pathologist.

Two important clinical indicators of malignancy are recrudescent symptoms after long periods of relative quiescence and small bowel obstruction that is refractory to medical therapy[23]. Therefore, it is prudent to consider a surgical assessment of patients with longstanding symptomatic Crohn’s disease who fail to respond to conservative management.

The usual age of diagnosis of small bowel carcinoma in Crohn’s disease patients is 45 to 55 years[4,6,7,22,23]. This is in contrast to small bowel carcinoma de novo which is usually diagnosed between 60 and 69 years of age[25]. Crohn’s disease will often predate the carcinoma diagnosis by 20 to 25 years[4,22-24,26].

Palascak-Juif et al[27] studied 20 patients with Crohn’s disease-associated small bowel carcinoma recruited from French university hospitals and compared them to 40 patients with small bowel carcinoma de novo recruited from a population-based registry. Small bowel carcinoma occurred after a median time of 15 years of Crohn’s disease and was located within the inflamed areas of the ileum (19) or jejunum (1), whereas in patients with small bowel carcinoma de novo it was distributed all along the small intestine. The median age of diagnosis of small bowel carcinoma was 47 years (range 33-72 years) in patients with Crohn’s disease and 68 years (range 41-95 years) in those with small bowel carcinoma de novo. The cumulative risk of small bowel carcinoma was 0.2% and 2.2% after 10 and 25 years of small bowel Crohn’s disease, respectively. The diagnosis was made preoperatively in 1 of 20 patients with Crohn’s disease and 22 of 40 patients with small bowel carcinoma de novo. Signet ring cells were found in 35% of Crohn’s disease cancers but not in patients with small bowel carcinoma de novo. Relative survival at 2 and 5 years was not significantly different between these two categories of patients (54% vs 37% and 35% vs 30%; with and without Crohn’s disease, respectively).

In a retrospective review from 1993 to 2009, Widmar et al[23] identified 29 patients with small bowel carcinoma (22 ileal and 5 jejunal) in Crohn’s disease. There were no carcinomas in excluded intestinal loops. The median age of onset of Crohn’s disease symptoms was 25 years and the median age at cancer diagnosis was 55.4 years, for a mean interval of 25.3 years. Widmar found that 75% of carcinomas arose in the terminal ileum, a location that only accounts for 13% of sporadic small bowel adenocarcinomas[28]. Patients with Crohn’s disease developed adenocarcinoma at an average age of 48 years versus 65 in the general population, with a male to female ratio of 3 to 1.

Solem et al[26] described the clinical features, outcomes, and risk factors of small bowel carcinoma in Crohn’s disease. Nine cases (4 males) were identified. The patients presented with abdominal pain (89%), obstruction (89%), and weight loss (78%). The carcinoma was located in the ileum in 8 patients (89%) and in the jejunum in 1 patient (11%). All cases but one had advanced disease with either lymph node involvement or metastases. The mortality rates at 1 and 2 years were 42% and 61%, respectively.

Floch et al[29] reviewed 47 previously reported small bowel carcinomas in Crohn’s disease. The average age of the Crohn’s carcinomas was 46.5 years while that for the de novo group was 55 years. The sexual ratios were 2.46:1 and 2:1 males to females for the respective groups. The de novo carcinomas had a slight predilection for the duodenum (4.7%) while the latter group had a heavy predilection for the ileum (70.8%) and contained no duodenal carcinomas. The prognosis of the Crohn’s group appeared to be much worse than that of the de novo group with five year survivals of 3.7 and 20%-22%, respectively. Late diagnosis in the enteritis group was felt to be the major reason for this.

Hoffman et al[7] also reviewed the literature and found 49 cases and added two of their own. The Crohn’s associated carcinomas differed from carcinoma not associated with Crohn’s in that (1) mean age of carcinoma discovery was less (46 years vs 64 years); (2) more carcinomas arose in the ileum (76% vs 27%); (3) diagnosis and cure were less successful; and (4) they occurred more frequently.

The review by Fresko et al[30] of 59 reported cases of carcinoma of the small bowel in Crohn’s disease, to which they added three of their own cases, revealed that (1) carcinoma develops at a younger age than in carcinoma de novo; (2) there is no difference in incidence of carcinoma in the first, second and third decades after onset of symptoms of Crohn’s disease; (3) 73% of neoplasms arose in the ileum; (4) in all but one case it developed in inflamed segments of bowel; and (5) in 31% of cases carcinoma developed in a bypassed segment of bowel. They concluded that Crohn’s carcinoma is a complication of Crohn’s disease and not a chance co-existence of the two diseases in the patient.

DIAGNOSIS

The clinical diagnosis of small bowel carcinoma in Crohn’s disease patients based on symptoms and physical examination is quite difficult, if not impossible. Indeed, many patients with carcinoma of the small bowel are not suspected of having a malignancy even at time of operation[5,6,31-33]. Most cases of small bowel carcinoma have been in segments involved with Crohn’s disease. These malignancies were indistinguishable radiologically from longstanding Crohn’s disease. In two of the patients described by Kerber[4] and Frank[5], there was “shouldering”, destruction, and a mass.

In general, imaging techniques may miss small lesions and may not be able to differentiate areas of small bowel carcinomas from those of severe Crohn’s disease. Routine computed tomography (CT) exposes patients to radiation, and although magnetic resonance (MR) imaging does not, it is time consuming and costly[34]. Buckley et al[35] found CT staging of small bowel carcinoma to be 47% accurate but errors occurred in patients with Crohn’s disease. Enteroclysis is invasive and requires special training[36]. The usefulness of FDG-PET is limited by the background chronic inflammation of Crohn’s[37].

Video capsule endoscopy is challenged by issues of visualization (i.e., limited field of vision, non-continuous image capture, lesions hidden in folds), inadequate preparation, and the stenosing nature of Crohn’s disease[36] that may prohibit a capsule from passing. Furthermore, lesion localization can be difficult and it does not allow for tissue sampling[38].

Enteroscopic techniques do allow direct visualization and tissue sampling but are invasive, labor intensive, and may be limited by the length and tortuosity of the small bowel. Intraoperative endoscopy is now reserved for lesions which are not accessible by balloon enteroscopy[39].

Despite their disadvantages, there are published cases demonstrating the utility of some of the techniques listed above. Placé and colleagues[40] observed two different patterns using MR-enterography: the first was a long, circumferential, asymmetric, and heterogeneous thickening of the ileum with a visible nodule on free induction echo stimulated acquisition images, and the other was a mass of the terminal ileum showing restricted diffusion on diffusion weighted MR imaging. Soyer et al[41] evaluated 7 patients with small bowel carcinoma, and on CT enterography the carcinoma was visible in five patients. Four different patterns were individualized including small bowel mass (2 patients), long stenosis with heterogeneous submucosal layer (2 patients), short and severe stenosis with proximal small bowel dilatation (2 patients), and sacculated small bowel loop with irregular and asymmetric circumferential thickening (1 patient). Stratification, fat stranding, and comb signs were present in 2, 2, and 1 patient(s), respectively. Nevertheless, adenocarcinoma may be completely indistinguishable from benign fibrotic or acute inflammatory strictures.

A case reported by Kodaira et al[42] highlighted the unique successful combination of PET/CT and double balloon enteroscopy for the diagnosis of small bowel carcinoma in a Crohn’s disease patient. Van Weyenberg et al[43] found both MR enterclysis and video capsule endoscopy useful but they believe that MR enterclysis is the better option. Ultimately, a combination of methods is likely the present day solution.

On the basis of their study of patients in whom adenocarcinoma of the small bowel developed as a complication of Crohn’s disease, Kerber[4] and Frank[5] concluded: (1) the development of adenocarcinoma is more likely to be seen in patients with longstanding disease; (2) classical radiographic appearance of carcinoma may not be seen; (3) a progressive change in radiographic appearance over time with the development of masses, fistulas, strictures and obstruction should raise the suspicion of co-existing carcinoma; (4) malignancy should be considered when there is a longstanding quiescent disease activity followed by a recrudesces of symptoms with concurrent radiographic changes; and (5) fistulas may be associated with carcinoma in two ways: a mass produced by carcinoma or carcinoma arising in chronic fistulas from Crohn’s disease. Contrary to prior recommendations, they suggest that obtaining radiographic examinations to document changing patterns of disease displays an important role in the management of patients with Crohn’s disease and may lead to earlier detection of complicating carcinoma thus improving prognosis in such patients.

PATHOLOGY

In contrast to de novo small bowel carcinomas which are most often in the duodenum (55%)[28], 75% of Crohn’s related small bowel carcinomas are ileal[22,28]. Miller et al[44] noted that all small bowel carcinomas associated with Crohn’s disease were ileal in location. Watanabe et al[45] published a summary of small bowel carcinoma within Crohn’s disease up until 1991 documenting only adenocarcinomas and signet ring cell carcinomas (Table 3). Petras et al[46] reported four patients with small intestinal carcinoma: three with poorly differentiated or signet ring cell type carcinomas and one with mucinous type. All four patients had high grade dysplasia in the mucosa immediately adjacent to the carcinoma, supporting the dysplasia-carcinoma sequence believed to occur in Crohn’s disease as with ulcerative colitis[47,48]. A wider variety of malignancies have been reported and are noted in Table 3, including sarcomas, lymphomas, and carcinoids.

Table 3.

Histopathology of small bowel cancers in Crohn’s disease

Histology	Watanabe et al[45]	Update since 1991
Adenocarcinoma	61 cases reported up until 1991 as quoted by Watanabe	Barwood et al[14]
		Chan et al[78]
		Chen et al[64]
		Christodoulou et al[79]
		Dossett et al[22]
		Feldstein et al[49]
		Fell et al[80]
		Fielding et al[9]
		Gillen et al[81]
		Gusakova et al[82]
		Jaskowiak et al[15]
		Jess et al[73]
		Kamiya et al[69]
		Katsanos et al[83]
		Kersting et al[65]
		Koga et al[84]
		Kronberger et al[85]
		Lindgren et al[86]
		Mellemjaker et al[71]
		Menon et al[12]
		Michelassi et al[55]
		Palascak-Juif et al[27]
		Partridge et al[13]
		Ribeiro et al[1]
		Richards et al[54]
		Rubio et al[87]
		Sammartino et al[88]
		Sigel et al[48]
		Solem et al[26]
Sarcoma		Gollop et al[89] (leiomyosarcoma)
		Jess et al[73] (leiomyosarcoma)
		Fielding et al[9] (reticulum-cell sarcoma)
Local lymphoma		Jess et al[73]
Carcinoid		Chen et al[64]
		Kvist et al[70]
		Mellemkjaer et al[71]
		Savoca et al[62]
Poorly differentiated		Petras et al[46]
		Savoca et al[62]
		Simpson et al[47]
Signet ring	8 cases reported up until 1991 as quoted by Watanabe

TREATMENT

The treatment of choice is wide resection of the small bowel segment harboring the carcinoma as well as resection of the corresponding mesentery and lymph nodes[22]. Pancreaticoduodenectomy for lesions of the second or third portion of the duodenum and right colectomy for carcinoma of the distal ileum would be required[49].

Evidence regarding the value of adjuvant chemotherapy for small bowel carcinoma is sparse and consists mostly of small retrospective reviews. Most available data is from experience in managing ampullary adenocarcinoma. Fishman et al[50] reported response rates upwards of 30% in the palliative setting: 33% with Gemcitabine, 50% with 5-FU or Capcitabine, and 42% with Platinum- or Irinotecan-based therapy.

PROGNOSIS

The prognosis of Crohn’s associated small bowel carcinoma varies among reported studies but has been noted to be poorer than the de novo small bowel carcinomas[24]. Most small bowel carcinomas in Crohn’s disease present at a younger age and are more diffusely and distally located than de novo carcinomas, usually making them undiagnosable at a curable stage. Indeed, two-thirds of cases present with intestinal obstruction. Greenstein[51] reported two year disease survival for small bowel carcinoma in Crohn’s disease as 9% compared with 15%-25% for de novo carcinomas. Mortality for carcinoma in excluded bowel has been reported to be as high as 100%[51]. One report of carcinomas developing in small bowel Crohn’s strictures found only 9 such cases[52]. All patients had Crohn’s disease for more than ten years. The average age of patients was 48 years compared to 65 years for de novo carcinomas. In patients with Crohn’s disease, carcinoma affects the ileum twice as commonly as the jejunum and four times as commonly as the duodenum. Fifty-nine percent of all carcinomas complicating Crohn’s disease were discovered incidentally during pathologic examination of resected specimens. If small bowel alone is considered this figure rises to 70%.

Small bowel carcinomas associated with Crohn’s disease tends to be poorly differentiated and are associated with a poor prognosis[4,5,19,33,53]. Two year survival rates have been found to be as low as 27%[22]. In a report by Richards[54] with three ileal carcinomas, survival ranged from 8 to 44 mo.

Michelassi et al[55] reported 14 cases of intestinal carcinoma complicating Crohn’s disease, 7 occurring in the small intestine and 7 in the large bowel. Two thirds of patients were male. The average age at time of diagnosis of Crohn’s disease and carcinoma was 28 and 48 years respectively. In five patients with small bowel carcinoma the diagnosis was made at laparotomy. In the remaining cases only careful histologic examination revealed the carcinoma. Six small bowel carcinomas were located in the ileum. Two small bowel carcinomas were multi-focal and had surrounding mucosal dysplasia. No patient with regional or distal metastases survived five years in comparison with an 83% five year actuarial survival rate in patients with carcinoma confined to the intestinal wall. Mean survival was 6 mo for patients with small bowel carcinoma.

Hawker et al[53] reported the clinical and pathological details of three cases diagnosed between 1968-1980 with a review of 58 patients from the literature. Of the 61 cases, 41 carcinomas occurred in the ileum, 18 in the jejunum, 1 in the duodenum and ileum, and 1 in the ileum and colon. Eighteen occurred in bypassed intestinal loops. The prognosis was poor: 44 patients (72%) died with a mean interval of only 7.9 mo from the diagnosis of their malignancy.

Widmar et al[23] reported significant differences in the two year survival for node negative versus node positive carcinomas (79.3% vs 49%) and for localized versus metastatic disease (92.3% vs 33.3%). Overall, 36 mo survival was 69.3% compared to 40% among those without excluded loops. Sixteen patients had long periods of quiescent disease before the diagnosis (7-45 years) and 16 required operation for bowel obstruction that was refractory to medical management.

OUR EXPERIENCE

From 1990 to 2013, 10 patients with underlying Crohn’s disease and small bowel adenocarcinoma were treated at our institution. In our series, there were twice as many males as females. The median age of Crohn’s diagnosis was 28 years and the median age of small bowel adenocarcinoma diagnosis was 57 years; this interval is consistent with the literature. In none of the 10 patients was the diagnosis known pre-operatively. Nine patients presented with a clinical picture of obstruction that did not respond to steroid treatment and required operation. Of the nine, two were found to have metastatic disease secondary to small bowel adenocarcinoma at the time of operation, while the diagnosis for the remaining seven was made on final pathology examination. The tenth patient in our series had refractory Crohn’s disease and was incidentally found to have terminal ileal adenocarcinoma on final pathology.

An equal number of patients had Crohn’s disease isolated to their small bowel as concomitant small and large bowel disease. None of our patients had bypassed loops of small bowel. In fact, only three patients had had previous operations for Crohn’s disease. All but one of our patients had terminal ileal adenocarcinomas; the exception was one jejunal carcinoma. Moreover, all patients had either stricturing and/or fistulizing Crohn’s disease. Four of our 10 patients had been treated with 5-ASA, a reported protective factor. All patients had a history of remote immunomodulator use but none were on maintenance immunomodulators at the time of presentation; if they were receiving medical therapy, it was solely high dose steroids. This is consistent with the conclusions made by Kerber[4] and Frank[5] concerning the development of a small bowel adenocarcinoma after a period of quiescent Crohn’s disease. As reported by others, the prognosis for small bowel adenocarcinoma in our series was also quite poor: the carcinoma-related mortality in our series is 70%.

SURVEILLANCE

It has become increasingly recognized that the risk of developing carcinoma of the colon in patients with colonic Crohn’s disease is comparable to those with chronic ulcerative colitis. Hence, regular colonoscopic surveillance in search of dysplastic changes is in order. However, no similar surveillance for patients with small bowel Crohn’s disease is possible.

Greenstein[24] suggests that surveillance should consist of regular abdominal examinations and that the recurrence of obstructive symptoms as well as the development of new symptoms should not be ignored especially after long quiescent periods.

CEA levels, found to be elevated in up to 38% of patients with active Crohn’s disease[56,57], have not been found to be useful in monitoring for small bowel carcinoma[19,58].

LIMITATIONS

Each of the studies included in this descriptive view has its own limitations. A descriptive review such as ours could not possibly be exhaustive if it had strict inclusion/exclusion criteria. Thus we chose to include data and observations from all available studies despite their limitations.

It is difficult to draw conclusions regarding the cumulative incidence of small bowel carcinoma in Crohn’s disease from studies with such a low frequency of event as small bowel carcinoma. Compared to current data, incidence values from older studies may actually be overestimated, as 5-ASA formulations (possibly protective against small bowel carcinoma) were released in the late 70’s and 80’s, late in the observation period of most cohorts[59]. Furthermore, the majority of reports did not have small bowel carcinoma as a primary outcome. Analysis did not routinely address incidence and discussions were sometimes not exclusive to the small bowel (e.g., “intestinal”[60], “upper digestive tract”[61], or discussing risk factors for small bowel and colorectal carcinomas together[62]). Studies rarely controlled for immunosuppressive agents, tobacco or alcohol[63], and those that did examine such exposures did not provide quantitative data[64]. Finally, multiple biases inherent of retrospective and single-centre studies exist in the available literature. For example, a high incidence of small bowel carcinoma in Crohn’s disease may reflect a bias in tertiary hospitals[65] or a surveillance bias due to close monitoring of Crohn’s disease patients[66], and reported risk factors may be a result of recall bias[67]. These limitations are inherent to the challenging problem of a rare disease that is difficult to diagnose.

CONCLUSION

In this review, we highlighted the available current evidence and the gaps of knowledge, technology, and clinical guidelines required for improving care of Crohn’s disease patients at risk of this devastating problem. Although the association of carcinoma in Crohn’s disease and the need to screen Crohn’s disease of the colon is well established, carcinoma associated with Crohn’s disease of the small bowel is difficult to diagnose and indeed is often not identified until operation for what is believed to be an exacerbation or non-response to medical therapy. Sadly, the diagnosis is often made after careful examination of the resection specimen by the pathologist. Over the decades there has been a lack of significant improvement in prognosis. There is a need to elucidate screening modalities to facilitate earlier diagnosis and treatment.