EDITORIAL

 

Capsule endoscopy - are we making the most of all its benefits?

La endoscopia con cápsula, ¿aprovecharemos todas sus ventajas?

 

 

E. Pérez-Cuadrado Martínez

Service of Digestive Diseases. Hospital Morales Meseguer. Murcia, Spain

 

 

Since small-bowel (SB) endoscopic exploration beyond Treitz's angle (usually the farthest point reached by upper digestive endoscopy (UDE) with a gastroscope) was introduced, upper digestive tract (UDT) lesions overlooked by UDE were reported, diagnosed "de novo" initially by oral push enteroscopy (OPE) (2), and subsequently by small-bowel capsule endoscopy (SBCE) (3) and then double-balloon enteroscopy (DBE). All these endoscopic explorations went through segments already examined by prior UDEs that had yielded negative results even when performed by the same clinician. Why is this?

In this issue Benito Velayos and colleagues (4) study this, and report their retrospective UDT findings following diagnosis with SBCE after a single negative UDE procedure. They specifically refer to obscure gastrointestinal bleeding (OGIB). In a substantial 17% of patients the authors diagnosed UDT conditions significant enough to indicate a second UDE, with a high impact on patient management changes.

Consistency in so-called "stable" mucosal lesions such as polyps is easier with endoscopy as compared to changing conditions such as digestive bleeding, where the time factor has proven critical for diagnostic effectiveness. In the natural history of hemorrhage CE may document initial bleeding findings (often not revealing the underlying cause, made invisible by blood itself), the responsible lesion, or finally a scarred or healed lesion with no pathological findings whatsoever. The dominant lesion, the true origin of bleeding, must be differentiated from simple findings. Angiodysplasia, the "princeps" lesion in OGIB, may also be a clinically non-significant finding, particularly if small in size and no bleeding is seen to originate in it. This is probably the case with the finding by the authors of gastro-duodenal angiodysplasia using SBCE in a patient where an initial UDE provided no diagnosis and a second UDE procedure finally found gastric cancer.

Our concern is intrinsically complex when the lesion also is vascular in nature - despite attempts at nomenclature standardization for these lesions, a single endoscopic exam may provide various diagnostic interpretations regarding characterization, whereas a second exploration, even by the same clinician, may set the record straight.

OGIB has been recently defined by the 2007 world consensus meeting in Berlin (5) as any gastrointestinal bleeding event whose cause remains unknown after UDE and colonoscopy with ileoscopy. Bleeding should be located in the SB if its origin in the UDT and colon was effectively excluded, which has given rise to a new emerging concept: mid gastrointestinal bleeding (MGIB) (6), with particular characteristics versus the traditional lower gastrointestinal bleeding (LGIB) beyond Treitz's angle. In MGIB the origin of bleeding lies within the SB, between the papilla of Vater and Bahuin's valve, and this condition has a special behavior regarding severity, transfusional requirements, stays, and overall cost, even if it only represents 5-10% of all bleeding events in the GI tract. MGIB has occupied a portion of the proximal territory of the still-accepted LGIB for the SB.

According to the world consensus on SBCE (7), this is the first-line diagnostic tool for the study of SB, most particularly for OGIB (the primary indication for SBCE). Only in selected OGIB cases is LDE initially indicated in the first line, including patients with Roux-en-Y anastomosis (8), SB stricture or obstruction (a contraindication for CE), massive bleeding, etc., as recommended by recent ESGE guidelines on flexible enteroscopy (9).

From a cost-effectiveness point of view a recent study by Gerson et al. (10), based on the analysis of a computerized model, considers LDE the preferred first-line strategy for OGIB, albeit the authors recognize that a different arm in their research --SBCE first, followed by SBCE-guided LDE (11)-- showed fewer complications and saves "blank" LDE procedures at a higher cost; this decreases morbidity, which is low but not negligible with this invasive technique (12). The indication for SBCE must be adequately selected.

Following the above overall review on the current endoscopic management of OGIB, we shall now turn to analyze the weaknesses and strengths of the two techniques compared by this study: UDE and SBCE within the UDT area.

SBCE only captures 2 endoscopic images per second (3) versus flexible endoscopy (both UDE and LDE), which takes 25; there is no insufflation (which accounts for a poorer view of the stomach) or lavage (on occasion only blood may be seen, but no underlying lesion), palpation or instrumentation capabilities, hence it is inferior to flexible endoscopy in terms of optical diagnosis. Why then at least 10% of lesions are overlooked by UDE that are diagnosed with SBCE? (3).

Several answers may apply here. In so-called "dark" or difficult-to-access areas, in the stomach (fundus-incisure) and also the duodenum (pyloric canal, knees, ampullary area) lateral was thought to provide diagnoses overlooked by UDE a few years ago; some teams performed a duodenoscopy for uncertain upper digestive bleeding (UGIB) following ordinary UDE, but the yield was low. This is now limited to highly selected patients given the improved flexibility and angled view of new-generation flexible videoendoscopes, which additionally have "intelligent" chromoendoscopy to better characterize vascular lesions. Few areas remain inaccessible to current frontal UDE, with maybe the exception of the ampullary area. SBCE may fortuitously reveal some of these dark spots in the stomach, despite lack of insufflation, because of the time factor.

The time factor (time of endoscopy and its duration in minutes) is highly relevant for the endoscopic diagnosis of bleeding in the context of its natural history. Thus, extensive CE series obtain a higher percentage of diagnoses when the exploration is performed within a short time after the initial event, with no delay, particularly for OGIB. Recurrent bleeding and performing the exam at the time this occurs are often fortuitous factors, but may positively influence the diagnosis of a digestive bleeding source, and SBCE usually remains longer in the UDT than UDE. An UDE procedure negative for UGIB lasts only a few minutes as compared to CE.

On the other hand, bleeding still not documented as OGIB often prompts an emergency UDE exam under conditions that may certainly be suboptimal --no fasting, emergency staff, no adequate tools accessible in the emergency room, etc. There is some heterogeneity when performing UDE under such conditions in terms of staff and materials-- an endoscopy-trained nursing staff is also a key factor. UDE for UGIB may be highly complex, with many influencing factors; not so much for SBCE, where there is homogeneity regarding materials and methods (a capsule progresses with no handling or manipulation). As the "human" factor plays a greater role in UDE, some teams recommend that the quality of the index exploration be assessed, prior to a second UDE, in patients with OGIB, even for the same explorer, most particularly in the presence of adverse environmental or evolutional conditions. Regarding cost-effectiveness (10), repeat UDE, in contrast with the US, has an acceptable cost in Europe, particularly when more costly techniques such as SBCE may be avoided.

Should we perform one or rather two UDE procedures to pinpoint a source of bleeding in a patient with UGIB? Well, that depends.

Unnoticed lesions are in the esophagus, as seen by the good consistency exhibited by esophageal specific CE studies versus UDE (13), but routine clinical practice teaches, however, that CE provides findings where UDE yielded false negative results in the stomach; thus the authors diagnosed two "watermelon" stomachs with SBCE that had been overlooked by the first but not a second UDE, which confirmed the diagnosis under CE guidance. This has been corroborated by other authors highly experienced in SBCE (14).

It is the stomach where SBCE may reach a higher diagnostic yield versus UDE for selected patients with UDT conditions. In the study by Velayo et al. (4), of 14 UDT diagnoses using SBCE with a negative initial UDE, 10 were related to the stomach. This may be influenced by the time capsules spend in the stomach (with no optical steering anyway), far longer than UDE, around 30 minutes in this paper, consistent with other reports by this same group (15). In fact, during SBCE delayed gastric emptying is defined when the capsule stays in the stomach for longer than 60 minutes. This may result in an opportunity to occasionally see blood emerging from a previously overlooked lesion within a few pictures (3). Undoubtedly, careful routine UDE procedures in Japan (for gastric cancer screening) provide more diagnoses overall (early lesions lost to Western endoscopy) by using longer observation periods, insufflation, and longer UDE duration. This would be applicable to initial UDE procedures for UGIB, where no findings are detected during the few minutes UDE may last in our setting.

CE passes the duodenum relatively fast, but UDE also fails to adequately reach Treitz's angle on occasion, as may be seen by correlating the radiographic position of the gastroscope's distal end to said angle. This may explain two author-reported cases of duodenal stenosis (inflammatory and neoplastic) unnoticed by the first UDE procedure, which were later diagnosed by SBCE. Both were also confirmed by a second, guided UDE procedure. CE has the advantages of positive patient acceptance (16), low morbidity -retention is its main complication (17)- and diagnostic technique sparing in OGIB (1,3).

However, while falling outside the scope of this study by Velayos et al., the cecum also provides challenging cases (because of accessibility and the common cecal condition in LGIB --angiodysplasia), and LGIB is also commonly diagnosed following several colonoscopies, particularly in case of angiodysplasia (frequently multiple lesions). In the colon CE has also provided positive diagnoses after unproductive colonoscopies (3). We should bear in mind that the primary cause of OGIB in Western series is angiodysplasia (in contrast to Eastern series, where ulcerative lesions predominate), and this lesion's changing nature makes it management challenging both in the SB and colon. Thus, synchronic lesions are usual in various gastrointestinal segments including the UDT, and so is also the case with metachronic lesions regarding the underlying causal condition, which persists after local therapy.

Therefore, in OGIB's diagnostic algorithm, should we also repeat colonoscopies? The clinical setting and common sense should rule here. It might be necessary for some patients.

To adequately manage a satisfactorily defined OGIB case following negative UDE (one or more studies) and colonoscopy (one or more studies), CE should be performed as soon as possible (particularly for OGIB with macroscopic blood). This CE will record more than 50,000 pictures for longer than 7 hours, and the reading may be tedious and fall under the influence of activity as usual at an endoscopy unit; hence the reading strategy should be counted in when assessing CE quality -- mean duration should be one hour or slightly longer, with rests not to reduce the image processing capacity by the brain after so many repeat sequences (3), and of course a thorough analysis of the whole recording. The latter includes the UDT and also the colon. To adequately analyze CE images visualization speed should be moderate to minimize false negative results. Above 15-25 images/sec we may miss lesions anywhere along the gut. Lesions such as OGIB- originating angiodysplasia have been allegedly diagnosed in only one frame (3). Appropriate visualization conditions require well-trained personnel and enough time. Visualization should also occur at an early time -- should CE diagnose findings unnoticed (or out of reach) during flexible endoscopy, a therapeutic procedure and/or sample collection should be indicated soon (UDE, colonoscopy, LDE, surgery, etc.). It is therefore clear that to obtain the maximum from SBCE recordings should be visualized carefully all throughout.

Common CE has only one lens whereas colonic capsules have dual bipolar lenses, and may thus detect more lesions with a higher number of frames per second and a wider vision field. In addition to conventional visualization following battery depletion and image download, SBCE may be performed online with current equipments. While this is difficult to implement with a full-time (more than 7 hours) clinician, the nursing staff may play a key role in this respect (sensor placement, capsule administration, marking of anatomical sites, blood detection, etc.), and further studies are needed to assess the cost-effectiveness of real-time CE. The overall diagnostic-therapeutic process in OGIB my increase in efficiency if performed "live", provided its coordination with other techniques such as LDE or UDE for active MGIB or UGIB, respectively, for which no significant wait list should exist. The SBCE wait list itself, particularly in reference sites such as the authors', makes this process difficult to implement. What could then be done?

Probably, to optimize the diagnostic-therapeutic process for gastrointestinal bleeding. First of all, in the presence of digestive bleeding (even for suspected LGIB) UDE should be performed as an early or urgent procedure with the highest quality; if negative, colonoscopy should follow as soon as possible. We should be in the "optimal diagnostic time" to most adequately address the dilemma posed by the authors: a second UDE procedure before SBCE? The answer to this question should be guided by patient status and the estimated quality of the initial UDE exam (18), in terms of whether any limitations occurred.

For the time being, our decision will be made on an individual basis until there is evidence enough to support that systematically performing a second UDE exploration is superior to just one.

In summary, we should make the most of the important resource capsule endoscopy represents, and improve our circuits for the diagnostic challenge posed by gastrointestinal bleeding. Shall we know how to do it?

 

References

1. Pérez-Cuadrado E. Acceso oral al intestino delgado. Rev Esp Enferm Dig 2002; 94: 423-6.        [ Links ]

2. Pérez-Cuadrado E. Videoenteroscopia oral: estudio prospectivo de 30 casos. Rev Esp Enferm Dig 1996; 88: 9-15.        [ Links ]

3. Pérez-Cuadrado E. Cápsula endoscópica. Madrid: Ed. Entheos; 2004.        [ Links ]

4. Velayos B, Herreros de Tejada A, Fernández L, Aller R, Almaraz A, Del Olmo L, et al. Upper gastrointestinal findings detected by capsule endoscopy in obscure gastrointestinal bleeding. Rev Esp Enferm Dig 2009; 101: 11-9.        [ Links ]

5. Pohl J, Blancas JM, Cave D, Choi KY, Delvaux M, Ell C, et al. Consensus report of the 2nd International conference on double balloon enteroscopy. Endoscopy 2008; 40: 150-60.        [ Links ]

6. Ell C, May A. Mid-gastrointestinal bleeding: capsule endoscopy and push-and-pull enteroscopy give rise to a new medical term. Endoscopy 2006; 38: 73-5.        [ Links ]

7. Pennazio M, Eisen G, Goldfarb N. ICCE Consensus for Obscure Gastrointestinal Bleeding. Endoscopy 2005; 37: 1046-50.        [ Links ]

8. Chacón S, Pérez-Cuadrado E. Abordaje endoscópico en cirugía bariátrica. Papel de la EDB. Rev Esp Enferm Dig 2008; 100: 365-6.        [ Links ]

9. Pohl J, Delvaux M, Ell C, Gay G, May A, Mudler C, Pennazio M, Pérez-Cuadrado E, Vilmann P and the ESGE Clinical Guidelines Committee. European Society of Gastrointestinal Endoscopy (ESGE) Guidelines: flexible enteroscopy for diagnosis and treatment of small-bowel diseases. Endoscopy 2008; 40: 609-18.        [ Links ]

10. Gerson L, Kamal A. Cost-effectiveness analysis of management strategies for obscure GI bleeding. Gastrointest Endosc 2008; 68(5): 920-36.        [ Links ]

11. Gay G, Delvaux M, Fassler I. Outcome of capsule endoscopy in determining indication and route for push and pull enteroscopy. Endoscopy 2006; 38: 49-58.        [ Links ]

12. Mensink P, Haringsma J, Kucharzik TF, Cellier C, Pérez-Cuadrado E. Complications of double balloon enteroscopy: A multicenter survey. Endoscopy 2007; 39: 613-5.        [ Links ]

13. Eisen GM, Eliakim R, Zaman A, Schawartz J, Faigel D, Rondonotti E, et al. The accuracy of PillCam ESO capsule endoscopy versus conventional upper endoscopy for the diagnosis of esophageal varices: a prospective three-center pilot study. Endoscopy 2006; 38: 31-5.        [ Links ]

14. Mascareñas-Saraiva M, Lopes L, Mascareñas-Saraiva A. Watermelon stomach seen by wireless -capsule endoscopy. Endoscopy 2003; 35: 100.        [ Links ]

15. Velayos B, Fernández L, Aller R, De la Calle F, Del Olmo L, Arranz T, et al. Estudio de los tiempos intestinales con cápsula endoscópica. Gastroenterol Hepatol 2005; 28: 315-20.        [ Links ]

16. Velayos B, Fernández L, Aller R, De la Calle F, Del Olmo L, Arranz T, et al. Public opinion survey after capsule endoscopy: patient's point of view on its utility. Rev Esp Enferm Dig 2006; 98: 436-48.        [ Links ]

17. Fernández Urien I, Carretero C, González B, Pons V, Caunedo A, Valle J, et al. Capsule endoscopy complications: retrospective analysis de 4656 cases. Oral communication. 16th United European Gastroenterology Week. Viena; 2008.        [ Links ]

18. Peter S, Heuss LT, Beglinger C, Degen L. Capsule endoscopy of the upper gastrointestinal tract. The need for a second endoscopy. Digestion 2005; 72: 242-7.        [ Links ]