Essentially, the spectrum of lesions responsible for small bowel bleeding is similar to that seen in other GI tract segments. It includes vascular and inflammatory lesions, as well as polyps/neoplasms. The prevalence of different small bowel findings in patients with OGIB and positive CE is reported in Fig. 3.3.

Large studies identify a definite bleeding source in 50–60 % of patients [37, 38, 69–71]. The DY is usually calculated taking into account only lesion with high bleeding potential; in fact, SB lesions identified at capsule endoscopy are usually divided into three subgroups: (1) highly relevant lesions (P2) such as angioectasia, large ulcerations, tumours or varices; (2) uncertain relevance lesions (P1) such as red spots, small isolated erosions; (3) low relevance lesions (P0) such as visible submucosal veins, non-bleeding diverticula , nodules without mucosal break [72].

In studies concerning CE, the DY is the parameter conventionally used to estimate the diagnostic ability of this examination. That is mostly because, in this setting, it is very difficult to calculate the most common clinical efficacy parameters such as sensitivity , specificity and diagnostic accuracy , as it lacks a true reference standard to which it could be compared with. Actually, in the study of the small intestine, a reference standard should allow to explore the entire small bowel, recognising the presence of even small lesions and obtaining histological confirmation, when necessary. These goals, at the present time, can be achieved only through intra-operative enteroscopy (IOE) or bidirectional DAE. The first requires of course surgical intervention (associated with mortality and morbidity), whereas the latter is logistically feasible only in a small proportion of the patients [73].

A recent study that compared CE with IOE found that the sensitivity , specificity, positive and negative predictive values of CE were 95, 75, 95 and 86 %, respectively. In this study, the gold standard, i.e. IOE, had 100 % DY for ongoing bleeding, but only 70.8 % for previous overt OGIB and less so (50 %) for occult OGIB [74]. These data confirmed results from an earlier retrospective study [48] in which CE was compared with a composite reference standard, taking into account both further diagnostic examinations and the follow-up. Moreover, in a more recent paper, Tenembaum et al. [67] compared CE with DAE combining oral and anal approach, according to the type of the lesion identified. These authors found that although CE and DAE yielded similar results, CE had an excellent accuracy profile for masses/tumours (sensitivity and specificity : 100 %), an intermediate performance for vascular and inflammatory lesions (sensitivity: 58 and 50 % respectively; specificity: 93 and 98 % respectively) and a disappointing accuracy for small bowel diverticulae (sensitivity : 9 %; specificity : 100 %).

On the basis of solid scientific evidence [69, 70, 75], most national and international gastrointestinal scientific societies have issued updated practice guidelines for the diagnosis and management of patients with OGIB, recommending CE as the first-line tool to evaluate the small bowel, after negative bidirectional endoscopy, in the clinical context of OGIB [32–36].

A key factor, which has been demonstrated in several studies, to significantly affect the yield of CE is the timing of the procedure. In fact, several studies [48, 76, 77], mostly focused on patients with overt OGIB, reported a significantly higher DY (75–90 %) in patients undergoing CE in the early stage of their diagnostic work up. Yamada et al. [77] recently showed an inverse linear correlation between the time elapsed from the last episode of overt bleeding and CE.

Unfortunately, less is known about the exact timing for CE in patients with obscure–occult GI bleeding; in these patients, it is often impossible to establish, in a reliable fashion, the length of the clinical history of anaemia; therefore, a more conservative approach can be generally suggested [78]. Recent guidelines [32, 79] suggest first to exclude other causes of persistent/recurrent iron-deficiency anaemia (i.e. malabsorption, menstrual loss) and then an appropriate empiric therapy with iron supplementation (i.e. ferrous sulphate 200 mg twice daily, for 1–3 months) could reasonably precede CE. Another factor, which has been found to have a relevant impact on the yield of CE, is the severity of bleeding. Patients with OGIB and low haemoglobin levels and/or high transfusion requirement are more likely to have small bowel lesions and deserve a small bowel evaluation by means of CE [80, 81].

It should not be underestimated that patients referred for CE can harbour lesions within the reach of standard endoscopes; nevertheless, in a cost-effective analysis as that performed by Vlachogiannakos et al. [30], their incidence is too low [30, 82] to justify a standard policy of second-look endoscopy before CE. Nevertheless, further studies are needed to identify subgroups of patients at increased risk to harbour lesions within the reach of a conventional endoscope in which repeating gastroscopy and/or colonoscopy would be cost-effective. Therefore, at present time, repeating conventional endoscopy before CE has to be decided on an individual basis, taking into account the reason for referral (occult vs. overt OGIB), but also the quality/timing of previous procedures, ongoing/previous therapies and comorbidities.

It is also important to emphasise that, after negative CE, when patients have ongoing or recurrent evidence of bleeding or when clinical suspicion of small bowel pathology is sufficiently high, further investigation with small bowel (total) enteroscopy or cross-sectional imaging is clearly warranted [83].

Although several [48, 71, 75] studies have assessed the yield of CE in OGIB, the exact significance of the lesions identified and their impact on clinical outcome has not been adequately examined. Unfortunately, the majority of studies on CE in OGIB are focused on potential changes in management, rather than on evaluating long-term outcomes. The studies assessing a change in clinical decision-making after CE showed that about two-thirds of patients with OGIB received specific therapeutic interventions or changes in management based on a finding from CE [84, 85].

Some studies [86, 87] reported that a negative CE in patients with OGIB is associated with a low rate of recurrent bleeding; therefore, it is reasonable to take an expectant approach with these patients, thus avoiding the need for unnecessary additional investigations, but keeping these patients under scheduled surveillance.

As far as the long-term outcome is concerned, the majority of studies evaluating the long-term (2–3 years) outcomes in patients undergoing CE report a favourable outcome in most of patients (about 60 %) managed on the ground of capsule endoscopy results [88, 89]. However, recently published studies [90, 91] are casting some doubts on the real impact on long-term outcome , despite a high diagnostic yield of CE. Several factors can probably explain the differences observed in different studies: there are differences in the population characteristics (i.e. medications, type of bleeding, timing of CE, etc.), there are no standardised approaches for patient management after CE and the different policies adopted at different medical centres can introduce relevant bias.

Therefore, although it is well known that CE has the capability to diagnose small bowel lesions and to drive further management [92–96], in the next few years, large multicentre studies with standardised surveillance and/or treatment protocol, with long follow-up, are warranted in order to truly estimate the long-term impact of CE in patients with OGIB.

Crohn’s disease (CD) is a transmural chronic granulomatous inflammatory disorder that can affect any segment of the gastrointestinal tract. Isolated small bowel involvement occurs in up to a third of the patients [97, 98]. Management in these individuals can be challenging due to varying symptoms, absence of raised inflammatory markers and a low sensitivity of small bowel barium study for early mucosal disease [99]. A mean delay in the diagnosis of CD of 1–7 years in published series has been reported [100, 101]. Over the past years, since the introduction of biological therapy for CD, there has been a paradigm shift in the way we manage CD with greater emphasis on mucosal healing rather than symptom control only. Mucosal healing has shown to be associated with improved long-term outcome s for CD and lower rates of hospitalisation and surgery [102]. With the advent of better diagnostic tests and imaging modalities such as capsule endoscopy , early diagnosis of small bowel Crohn’s disease (SBCD) has become ever increasingly important.

Small bowel Crohn’s disease (SBCD) is often a difficult diagnosis to establish clinically. Patients can present with variable symptoms including abdominal pain , fatigue, diarrhoea, weight loss, with or without evidence of bleeding [103]. Patients can also present with complication s of transmural inflammation that include stricturing or fistulating disease.

Apart from GI symptoms, CD can present with extra-intestinal manifestations. These include the following:

Laboratory markers can be helpful in raising the suspicion of small bowel CD. Patients may present with iron-deficiency anaemia , B12 deficiency and/or raised inflammatory markers. It is unclear that combination of symptoms and/or laboratory markers predicts CD [104]. Faecal calprotectin (fC) has been used as a cost-effective screening tool prior to SBCE in patients with suspected CD and prior normal bidirectional endoscopy [105]. Koulaouzidis et al. [105] showed that an fC > 100 μg/g is a good predictor of positive SBCE findings, whilst fC > 200 μg/g confirmed CD in 50 % of cases. There are published guidelines on the use of SBCE in CD for USA, UK and Europe [106–108].

Typical histological features of CD include the presence of granulomas, areas of chronic inflammatory cell infiltrate (lymphocytes, neutrophils, plasma cell infiltrates), skip lesions, transmural inflammation with lymphoid aggregates [109]. It is, however, uncommon to see all the typical histological features of CD in a biopsy specimen [110].

Histology has always been regarded as a cornerstone in the diagnosis of Crohn’s disease . However, obtaining histological confirmation is not always possible as the disease is patchy (such that biopsies near inflamed mucosa may be normal) and may arise in less accessible areas of the intestine, such as the small bowel, leaving the clinician to decide on diagnosis and treatment without histological confirmation.

Coeliac disease is a common autoimmune condition characterised by a heightened immunological response to ingested gluten, with prevalence rates in the USA and European populations estimated to range between 0.2 and 1 % [160, 161]. Furthermore, there is some evidence to suggest that the prevalence of coeliac disease is increasing [162–164]. Finally, clinicians from both China and the Indian subcontinent are now recognising patients with coeliac disease. This had not previously been the case, and one hypothesis is that coeliac disease is emerging due to the introduction of wheat into these ethnic groups (as their diet becomes more westernised). Thus, coeliac disease is a global problem [165, 166]. The current gold standard diagnostic test for coeliac disease is small bowel histology, demonstrating the presence of villous atrophy (VA) (Marsh 3a to 3c) [167]. Corroborative evidence used to support the diagnosis of coeliac disease comes from positive serological tests (tissue transglutaminase (tTG) and endomysial (EMA) antibodies) and a clinical response to a gluten-free diet. Occasionally, when diagnostic uncertainty exists, human leucocyte antigen (HLA) typing is undertaken, which may help to exclude coeliac disease, given the high negative predictive value of this test [168].

Historically, a small bowel biopsy was obtained using a Crosby suction biopsy capsule. With the advent of fiberoptic oesophagogastroduodenoscopy (EGD), investigators were able to demonstrate that endoscopic duodenal biopsy was comparable to the suction biopsy in terms of its ability to detect VA with three or four endoscopic duodenal biopsies, taken at different levels along the duodenum and jejunum [169–171]. As well as providing a more reliable method of obtaining a small bowel sample, EGD has the advantage over Crosby capsule in that it allows direct visualisation of the duodenal mucosa. Investigators are now able to detect the endoscopic markers of VA—reduction or absence of Kerckring’s folds, mosaic mucosal pattern, micronodular pattern, scalloping and possibly duodenal erosions [172–175]. Although these markers can be seen in other conditions that cause VA, they have excellent specificity when combined with positive coeliac serology.

However, there are several potential limitations of EGD as part of this diagnostic pathway. These include its invasive nature and its inability to evaluate small bowel mucosa beyond the duodenum. Changes of coeliac disease are well recognised to be patchy [176], and occasionally, in some patients, the small bowel distal to the reach of a standard gastroscope may be more affected than the proximal bowel where biopsies are taken [177–179]. In addition, duodenal biopsy sampling may be affected by specimen orientation [180]. The distribution of complication s of coeliac disease such as ulcerative jejunitis and enteropathy-associated T-cell lymphomas is also particularly important as these appear to be more commonly seen in the distal small bowel [181–183]. For this reason, other endoscopic modalities such as push or double-balloon enteroscopy (DBE) may be employed in order to allow more extensive evaluation of the small bowel and obtain histology [177, 184]. However, these investigations are offered in relatively few centres that are labour intensive and are more invasive than standard EGD.

Capsule endoscopy (CE) could therefore provide a useful alternative for direct visualisation of the small bowel. CE compares favourably with other small bowel imaging techniques such as magnetic resonance imaging (MRI) enteroclysis (Van Weyenberg et al. 2013). CE is a well-tolerated, minimally invasive test, predominantly utilised for the assessment of obscure gastrointestinal bleeding , inflammatory bowel disease and polyposis syndromes . However, there has been increasing interest in the role CE may have in coeliac disease. With an eightfold magnification power comparable to a dissecting microscope, capsule endoscopy has the potential to detect VA and other small bowel complication s seen in coeliac disease. In this chapter, we will discuss the potential role of capsule endoscopy in the diagnosis of coeliac disease, its use in complicated cases and the potential for future development .

A number of endoscopic markers for coeliac disease have been identified. The presence of these markers at EGD is sometimes used to determine whether duodenal biopsies are indicated. However, at present, capsule endoscopes are unable to take biopsy samples, so what are the endoscopic markers of coeliac disease? How accurate are they in diagnosis? and Are they applicable to CE? Endoscopic markers of coeliac disease include the following: reduction or absence of Kerckring’s duodenal folds; scalloping, which is a notched and nodular appearance of the duodenal folds; increased visualisation of submucosal vasculature; a mosaic pattern, resulting from the cobblestone or micronodular appearance of the mucosal surface; and mucosal fissures and grooves [173]. Endoscopic images of these features are shown in Fig. 3.21. Other endoscopic features such as duodenal erosions have been suggested but are not widely reported and can be seen in multiple other conditions [174]. A wide range of studies have attempted to demonstrate the usefulness of endoscopic markers during EGD with contradictory results. Compared to the gold standard of histology the sensitivity and specificity of all endoscopic markers combined varies from 37 to 94 % and from 92 to 100 %, respectively [185–188]. There are several possible explanations for the absence of endoscopic markers in patients with coeliac disease. Although histology is the gold standard test, it is known that this does not have 100 % sensitivity particularly if small numbers of biopsies are taken or enteropathy is patchy and if specimens are poorly orientated [176, 180]. Also endoscopic markers might actually be absent for degrees of enteropathy milder than subtotal or total VA [185]. Although coeliac disease is the most common cause of VA, particularly in patients with positive coeliac serology, the specificity of endoscopic markers of coeliac disease is also not 100 % and there are a number of differential diagnoses VA [189]. These include infections such as giardiasis and Whipple’s disease or other autoimmune conditions such as Crohn’s disease or autoimmune enteropathy. A full list of differential diagnoses is shown in Table 3.3 with capsule images of some of these differentials shown in Fig. 3.22. Particular consideration should be made to these differentials in CE as there is no confirmatory intestinal biopsy. The lack of a biopsy is also important when we are considering the wide range of reported sensitivities of endoscopic markers for coeliac disease. The potentially low sensitivity of these markers may lead to a number of missed diagnoses. What is the evidence, therefore, that CE is a suitable tool for diagnosing or assessing coeliac disease? The potential indication s for CE in coeliac disease are shown in Table 3.4 and will be discussed in detail below.

Diagnostic EGD with duodenal biopsies has an excellent safety record and is the gold standard test for diagnosing coeliac disease. However, even with adequate sedation, EGD can be poorly tolerated with a small risk of serious complication s such as perforation. As a result, some patients are unwilling or unable to undergo the procedure. CE is much better tolerated by patients and as a result may be a potential imaging modality in these patients. CE may also have some other potential advantages over EGD in assessing macroscopic features of coeliac disease. CE is performed without the air insufflation, required for standard endoscopy, which may obscure subtle changes of VA. The optical dome of the capsule is in close contact with the mucosa, which may allow better visualisation of the mucosa. This has a similar effect to water immersion endoscopy, which has been shown to have excellent sensitivity and specificity during standard endoscopy of unselected patients [190]. Characteristic changes of coeliac disease in some patients may only be seen distally, and CE also allows visualisation of the entire length of the small bowel. Capsule images of villous atrophy are shown in Fig. 3.23.

A small study of 10 coeliac patients with VA and 10 controls was the first to assess the utility of CE in diagnosing coeliac disease [191]. All of the images were reviewed by 4 blinded investigators 2 of who had extensive pre-study experience of reporting CE and demonstrated 100 % accuracy in identifying patients with VA. The 100 % accuracy is particularly impressive given that 4 out of the 10 patients with VA had apparently normal looking duodenal mucosa at EGD. However, there was relatively poor inter-observer agreement between some of the investigators with kappa coefficients (κ) as low as 0.26. When results were combined with those of the less experienced clinicians, they achieved a sensitivity , specificity , positive predictive value (PPV) and negative predictive value (NPV) of 70, 100, 100 and 77 %, respectively. There are several limitations to this study with a small sample size, high degree of ascertainment bias and only one patient included who had partial VA (Marsh 3a). Some of these limitations were addressed in a study of 21 EMA-positive patients and 23 antibody-negative controls [192]. The EMA-positive patients had a range of histology findings from Marsh 1 to 3c including 1 patient with lymphocytic duodenosis (Marsh 1) only and 5 patients with partial VA (Marsh 3a). All of the control patients had normal histology. Seventeen of the 20 patients with VA were correctly identified via CE, and there were no false positives in the control group. A further study of 22 patients with positive serology 8 of whom had normal duodenal histology showed that results were similar when tTG was used for patient selection with sensitivity, specificity, PPV and NPV for CE of 93, 100, 100 and 89 %, respectively [193]. Rondonotti et al. [194] also identified patients with positive serology and showed a sensitivity of 89 % for in 28 patients with confirmed villous atrophy. Finally, in the largest study to date of 37 patients with coeliac disease and 38 gender-matched controls, CE showed a sensitivity of 92 % and specificity of 100 % [178]. Importantly, the disease group included 19 patients with partial VA.

A consistent finding in all of these studies is that the PPV and specificity in the presence of EMA or significantly elevated tTG (usually greater than ten times the upper limit of normal) for the recognition of endoscopic markers of coeliac disease is 100 %. However, the high pretest probability of coeliac disease in all of these studies may again be a potential limitation, leading to an overestimation of CE performance. However, they accurately reflect real-life clinical practice where patients are likely to be selected for CE of the basis of positive serology and suggest that CE may be an appropriate tool for patients who are unable to undergo EGD. A summary of the studies in suspected coeliac disease is shown in Table 3.5.

One area where CE may confer an advantage over standard endoscopy is that CE has the potential to image the entire small bowel. The gluten load is highest in the proximal small bowel but the entire length of the small bowel can be affected as demonstrated in a study of terminal ileal biopsies [195]. Patients with newly diagnosed coeliac disease had significantly higher intraepithelial lymphocyte counts when compared to controls. However, not all coeliac disease patients were affected. It would seem intuitive that the more of the bowel that is affected, the more severe symptoms and the higher the chance of potential complication s. However, this has not been proven mainly because it is difficult to assess the extent of disease. Capsule endoscopy may provide a way of doing this.

Also if CE for newly diagnosed coeliac disease patients was common practice, then there is potential for both assessing disease severity response to treatment. This may be pertinent in patients with persistent symptoms as it is known that the villous architecture recovers more quickly in the distal small bowel and proximal duodenal biopsies may show continuing VA. These patients may be labelled as non-adherent. However, the lack of improvement in histology proximally may not represent a significant improvement in the extent of disease. In a recent study of 38 untreated coeliac disease patients and 38 controls, the authors attempted to assess extent of disease by 2 methods [178]. Firstly, the investigator reviewing the images was invited to make a qualitative assessment of severity based on whether disease was patchy or continuous and extent of disease. A second quantitative assessment of severity was also used. The total length of time with changes of VA and time relative to the small bowel transit time were recorded. The authors were unable to show a relationship between either qualitative or quantitative measurements of extent of disease and severity of clinical presentation; however, a positive EMA was associated with more extensive disease. In the 30 coeliac disease patients who agreed to repeat CE after GFD, the mean time with abnormality reduced from 60 to 12 min. A second more recent study of 12 patients with coeliac disease who had repeat CE after 12 months on a GFD has also demonstrated this improvement. The investigators assessed the extent of disease as a percentage of the total small bowel transit time [193]. Although there was no initial correlation between extent of disease and clinical severity, they did demonstrate a significant reduction in the mean time with VA.

These 2 studies have so far failed to demonstrate any relationship between extent of small bowel involvement and clinical severity of disease. However, as experience with CE in coeliac disease increases, this may become possible. The use of CE to assess small bowel healing does appear to be a promising area; however, this will only become relevant as more patients undergo CE at the time of diagnosis. New technologies as discussed later in this chapter may also help to improve the objective measurement of disease severity.

Another area where CE may play a role is in the investigation of equivocal cases of coeliac disease. As previously discussed, the changes of coeliac disease can be patchy and a duodenal biopsy in patients with positive serology may not demonstrate VA. Lesser degrees of histology that can be associated with coeliac disease are non-specific and are seen in a variety of other conditions. This can leave some patients without a definitive diagnosis. How therefore should we investigate patients with positive coeliac serology but normal or non-specific duodenal histology? An international consensus conference on the use of capsule and double-balloon enteroscopy advocated the potential use of CE in this situation [196]. There is, however, limited data assessing the role of CE in these equivocal cases.

In a large multicentre study, the investigators identified patients with gastrointestinal symptoms and positive EMA, tTG or antigliadin antibodies (AGA) [194]. Duodenal biopsies and CE were performed at the time of presentation, and 11 patients were found to have normal duodenal histology. Of these, 10 patients also had a normal capsule endoscopy ; however, these patients were all positive for AGA only which is non-specific and seen in a variety of other conditions. The one patient that had evidence of VA on capsule endoscopy was also positive for EMA and had biopsy-proven dermatitis herpetiformis making coeliac disease very likely. One patient with Marsh 1 changes and positive EMA had a normal CE; however, the 3 patients with Marsh 2 changes and positive EMA and tTG all had evidence of VA on CE. The authors do not state whether any further confirmatory tests such as HLA genotyping or further small bowel biopsies were taken to confirm coeliac disease but it is likely that these were additional diagnoses that had been missed by conventional EGD and duodenal biopsy.

There is, however, conflicting evidence. In a study of 8 patients with positive serology (EMA or tTG) and a normal duodenal biopsy, CE did not reveal any endoscopic features of coeliac disease [193]. Thus, the investigators concluded that there was no benefit in performing CE for this subgroup of patients. In a further study, 22 irritable bowel syndrome patients with positive AGA, EMA or tTG and normal duodenal histology underwent CE and HLA genotyping [197]. Subtle mucosal abnormalities within in the small bowel, such as mucosal breaks, ulceration or denuded and blunted villi, were seen in 55 % (12/22) of cases. However, the authors felt that none of these features were conclusively characteristic of coeliac disease and these changes were seen in both patients with and without an HLA genotype required for coeliac disease. The majority of the patients only had a positive AGA, which as discussed previously has a low specificity for coeliac disease. The single patient with a positive EMA (the most specific marker for coeliac disease) had a normal CE. Finally, in a study of 30 patients with Marsh 1 or 2 changes, only 6 of whom had positive EMA or tTG, 1 patient was diagnosed with coeliac disease and another with small bowel Crohn’s on the basis of CE appearances [198]. It is clear that further work is required to assess the cost-effectiveness of the use of CE in these equivocal cases if the yield is as low as in this final study. CE use may be justified however, in EMA or tTG positive patients with Marsh 1 or 2 changes or gastrointestinal symptoms particularly, if they are unwilling to undergo further EGD and repeat biopsies.

Another diagnostic challenge is antibody-negative VA. As previously discussed, there is a wide range of differential diagnoses for VA. In the study of equivocal cases by Kurien et al. [198], they also included a group of patients with antibody-negative VA to see whether this increased the diagnostic yield. Patients were extensively investigated for coeliac disease including HLA phenotyping, by monitoring response to GFD and, in some cases, repeat duodenal biopsies. On the basis of CE appearances and other ancillary tests, 7 patients could be diagnosed with coeliac disease and 2 further patients were diagnosed with small bowel Crohn’s as a cause for VA. Again, this is a single small study and further work needs to be done to clarify the role of CE in antibody-negative VA cases. This is particularly important as CE alone is probably insufficient to confirm a diagnosis of coeliac disease as endoscopic markers are not specific to coeliac disease rather they are predictors of mucosal disease [199].

Although the majority of patients with diagnosed coeliac disease improve on a gluten-free diet (GFD), up to 30 % patients do not respond as expected. In many of these patients, other causes for their symptoms such as microscopic colitis, small bowel bacterial overgrowth or pancreatic insufficiency are identified [200]. However, many of these patients undergo repeated endoscopy and duodenal biopsy to assess small bowel healing and look for serious complication s such as refractory coeliac disease (RCD) or enteropathy-associated T-cell lymphoma (EATL). RCD is defined as persistent malabsorptive symptoms and villous atrophy despite strict adherence to a GFD. It is subdivided into types I and II depending on clonality of intraepithelial lymphocytes. RCD type II carries a worse prognosis and is associated with greater progression to EATL [200]. However, changes may not be confined to the proximal small bowel and may be out of the reach of a standard gastroscope. CE may therefore play a role in the investigation of these patients.

Several studies have attempted to delineate a role for CE in non-responsive patients. In blinded comparison of duodenal biopsy and CE in 18 patients who had failed to respond to a GFD, 67 % of those with histological evidence of coeliac disease had abnormal CE [201]. Six patients with normal histology also had a normal CE. Four patients, however, had evidence of persistent histological changes of coeliac disease but had normal CE appearances. Agreement between histology and CE appearances was therefore fairly modest with a κ coefficient of 0.65. Importantly, however, 2 cases of ulcerative jejunitis were identified. Ulcerative jejunitis is usually associated with RCD type II that can result in small bowel stricturing and is associated with a high risk of developing EATL. Early identification of RCD type II may allow effective treatment with immunosuppression and prevent progression to EATL. These findings were replicated in a recent study of 69 patients with coeliac disease and persisting symptoms [198]. Signs of VA were identified in 45 %, and serious complication s were identified in 8 patients including 2 cases of EATL, 1 case of ulcerative jejunitis and 4 RCD type 1. However, although the investigators were blinded to the EGD findings, no control groups were included in either of these studies, leading to a high degree of ascertainment bias, which may have overestimated the usefulness of CE. This was addressed in a recent study of 42 non-responsive patients, 84 controls and 30 patients with uncomplicated coeliac disease who had responded well to a GFD for at least 6 months [182]. Only 9 of 16 patients with villous atrophy on histology had abnormal findings on CE. Four patients with normal histology had apparently abnormal CE appearances. As a result, overall the agreement between histology and CE findings was weak with a κ coefficient of 0.44. However, it must be noted that duodenal histology may be a less than perfect gold standard as changes can be patchy. It may be the findings on CE represented VA that was distal to the reach of EGD and standard duodenal biopsy. Also it must be noted that of the 16 patients with VA, 13 had partial VA (Marsh 3a), which may not be seen as clearly as total VA. All 3 of the most severe lesions were correctly identified as abnormal by CE. Again, a case of RCD type II was identified by CE and confirmed by DBE and biopsy. The use of CE to assess the extent and severity of disease in patients with known RCD may also be helpful as shown in a recent study of 29 patients with RCD and 9 patients with symptomatic coeliac disease [202]. Three cases of EATL and 5 cases of ulcerative jejunitis requiring specific treatment in the RCD cohort were identified. The majority of the RCD patients also underwent DBE, and the authors concluded that 17 patients could have avoided this invasive investigation based on CE findings.

Apart from this final study, where there was an unusually high proportion of patients with RCD, the apparent diagnostic yield for complication s such as EATL and ulcerative jejunitis appears low. However, these diagnoses carry significant rates of morbidity and mortality, which may be reduced by prompt diagnosis. The use of capsule in non-responsive patients may therefore be justified. Patients with ulcerative jejunitis and EATL can have a significant risk of small bowel stricturing. CE should therefore be used with caution, and a patency capsule should always be employed to reduce the incidence of capsule retention . Capsule images of ulcerative jejunitis and EATL are shown in Figs. 3.24 and 3.25.

As previously discussed, the main drawback of CE is the inability to perform biopsy. It may, however, be possible to incorporate this function into future incarnations of capsule technology [203]. In the near future , these advances, however, are only likely to allow non-targeted biopsies of the small bowel. As previously discussed, changes of coeliac disease can be patchy and biopsies may miss affected areas. Another method therefore of increasing the diagnostic yield would be to incorporate other image-enhancing techniques. Multiple techniques such as narrow-band imaging or optical-band imaging, amongst others, have been used in conventional endoscopy with limited improvement in diagnostic accuracy [204]. Capsule has already been shown to be more accurate than standard endoscopy for identifying macroscopic features of VA, and as capsule technology and picture resolution improve, these techniques may be incorporated and may prove beneficial in increasing the yield further.

These technologies are, however, not presently available and may never come to fruition. An interesting development that is presently undergoing clinical trials is the use of digital quantitative analysis of CE images. CE interpretation is currently subjective as has been demonstrated by the significant inter-observer variability in some studies. CE can miss lesions as demonstrated by a pooled analysis of 530 examinations where 10.8 % of lesions picked up on conventional endoscopy were missed by CE [64]. There is also a lack of standardisation of severity of CE findings. An objective computerised analysis of CE images would therefore seem advantageous. This has been developed using current CE technology and involves the production of a 3-dimensional representation of 2-dimensional CE images, mucosal protrusions can then be assessed for height, width and number per image. The images are also converted to greyscale and the intensity of each pixel is measured. These techniques were used in 3 recent studies [205–207]. On analysis of these images, patients with villous atrophy showed blunted protrusions with lower height and greater diameter. The mucosal surface also showed greater variability in pixel greyscale intensity in patients with VA, which the authors postulated was due to fissures, scalloping and mosaic pattern seen in patients with VA. Digital enhancement techniques have also been used to estimate small bowel motility, which is known to be impaired in patients with active coeliac disease [208]. With further development, these techniques may prove invaluable in the assessment of patients with suspected or known coeliac disease. Crucially, this objective measurement may also allow for more accurate quantification of severity and improvement in pathology on a GFD.

At present, duodenal biopsy remains the gold standard for diagnosing coeliac disease; however, CE may play a role in those patients who are unable or unwilling to undergo standard EGD and biopsy. CE is also proving an important diagnostic tool for investigating patients for possible complication s of coeliac disease such as small bowel malignancy, RCD and ulcerative jejunitis. There are potential limitations to CE with high degrees of inter-observer variability, potentially high miss rates for significant lesions and the risk of capsule retention in patients with strictures resulting from complicated coeliac disease. However, technology is evolving and future development s may allow small bowel biopsy or incorporate ‘virtual biopsy’ technologies that will improve the versatility of CE in coeliac patients.

Hereditary gastrointestinal (GI) polyposis syndromes are rare autosomal dominant disorders that are characterised by the presence of GI polyps, extra-GI phenotypic manifestations and an increased risk of GI and extra-GI neoplasia. The histopathology of polyps, nature of other phenotypic manifestations and degree of risk of malignancy are governed by the genetic mutations of the underlying syndrome. Small bowel (SB) capsule endoscopy (SBCE ) has a role to play in the surveillance and management of patients with SB polyposis. Although this is now established for patients with Peutz–Jeghers syndrome (PJS), evidence to support the routine use of SBCE for the surveillance of patients with other polyposis syndromes is lacking.

Peutz–Jeghers syndrome (PJS ) is a high-penetrance autosomal dominant polyposis syndrome, which is associated with a germline mutation in the STK11/LKB1 gene (19p13.3) in up to 94 % of cases [209]. PJS has an incidence of about 1 in 8,500 to 1 in 200,000 live births [210–213] and is characterised by a phenotype that includes the presence of distinct mucocutaneous melanin pigmentation, gastrointestinal (GI) polyposis [214] and a predisposition to GI and extra-GI malignant disease [210, 212–218].

PJS polyps are thought to arise as a result of GI mucosal prolapse [211, 216] and have a characteristic hamartomatous histopathology including ‘arborisation’ of a smooth muscle core with ‘frond-like’ epithelial lengthening (Fig. 3.26). Although PJS polyps may occur anywhere within the GI tract, they occur predominantly within the proximal small bowel (SB) [214, 218, 219]. PJS SB polyps that are ≥1.5 cm in size are considered to be clinically significant and contribute to a major part of the disease burden in patients with this condition, mainly by resulting in intussusception and SB obstruction , often requiring emergency surgery [211–213, 218, 220, 221]. SB complication s arising from large PJS polyps frequently result in multiple laparotomies throughout a patient’s lifetime [222]. These complication s commence in childhood and early youth [211, 223–225], and the cumulative risk of SB intussusception may be as high as 50 % by the age of 20 years [218].

In view of the high risks associated with clinically significant SB polyps (≥1.5 cm), current guidelines recommend that patients with PJS aged ≥8–10 years, should undergo SB surveillance using a minimally invasive modality such as SB capsule endoscopy (SBCE ) and/or radiological diagnostic imaging [magnetic resonance enterography/enteroclysis (MRE)] on a biennial to triennial basis [211, 220, 221] in order to facilitate early detection of large PJS polyps and also to investigate for possible occult SB malignancies. SBCE is very well tolerated by patients and is a less invasive option for the surveillance of the SB in patients with PJS [220]. The limitations relating to SBCE include its potential to miss clinically significant polyps, especially if these are located within the proximal SB [220, 221, 226–228] and challenges relating to the estimation of polyp size and location and (Fig. 3.27), possible ‘double counting’ of polyps [220, 221]. Some of these limitations (such as potential miss rates) may be mitigated by the more advanced technology of newer generation capsules, such as the PillCam SB3® (Given Imaging , Yoqneam, Israel), which also incorporates an ‘adaptive frame rate’, enabling a higher frequency of frame capture (up to 6 frames per second) during rapid capsule transit.

Before the introduction of SBCE and MRE into clinical practice, SB surveillance in patients with PJS depended on SB follow through (SBFT). However, SBFT is limited by low sensitivity , poor spatial resolution (due to its two-dimensional quality), with poor differentiation of overlapping SB loops and also exposes patients to a significant dose of ionising radiation [220, 229–231]. Radiological imaging of the SB in patients with PJS has therefore been superseded by MRE [220, 221]. A recent prospective comparative study of MRE versus SBCE for this indication showed that although patients tolerated SBCE better and preferred it to MRE [220], the 2 modalities were comparable for the detection of clinically significant (≥1.5 cm) PJS SB polyps. Currently, MRE may also allow for more accurate estimation of PJS polyp size and respective polyp location (Fig. 3.28) [220, 221, 232] and therefore appears to be a useful, alternative modality to SBCE for surveillance or for further evaluation of polyps detected by SBCE [220, 221].

Minimally invasive SB surveillance for clinically significant (≥1.5 cm) polyps allows pre-emptive removal, before an episode of intussusception occurs, potentially avoiding the need for emergency surgery and possible SB resection [211, 222]. This minimally invasive approach to the management of PJS SB polyposis has been enhanced by the introduction of device-assisted enteroscopy (DAE). DAE facilitates endoscopic excision of PJS polyps located deep within the SB (Fig. 3.29) [233–243] and therefore has the potential to obviate the need for laparotomy with intra-operative enteroscopy (IOE), which until recently was the sole option available for removal of PJS polyps deep within the SB [244, 245]. Laparotomy with IOE, however, exposes patients to major abdominal surgery and its potential complication s such as post-operative intra-abdominal adhesions and short-bowel syndrome [246–248]. Data on DAE for this indication mainly relate to the double-balloon enteroscopy (DBE)and suggest that this endotherapeutic modality may offer a minimally invasive alternative to laparotomy and IOE for selected patients with PJS [216, 219, 236–238, 240]. The introduction of a DAE-based therapeutic strategy in childhood (before the patient’s first laparotomy has been required) may enhance its success, since DAE may be hindered by the presence of post-surgical intra-abdominal adhesions [249, 250]. In patients who have already developed post-surgical adhesive disease, division of adhesions by mini-port laparoscopic assistance during DAE may still provide a less invasive alternative to laparotomy with IOE for SB polypectomy [251]. However, in certain patients, laparotomy with IOE shall continue to have a major role to play in the management of SB polyps and the overall strategy should be tailored to the requirements of the individual patient. The current recommendations for the surveillance and management of PJS SB polyposis using minimally invasive strategies are presented as an algorithm (Fig. 3.30).