Capsule endoscopy: The voyage is fantastic—will it change what we do?
2002; Elsevier BV; Volume: 56; Issue: 3 Linguagem: Inglês
10.1016/s0016-5107(02)70063-1
ISSN1097-6779
Autores Tópico(s)Gastric Cancer Management and Outcomes
ResumoThe video capsule endoscope (VCE) was approved by the Food and Drug Administration (FDA) in August 2001. Some would say that its eventual development was foretold by the 1966 movie Fantastic Voyage, based on the science fiction story by Isaac Asimov and screenplay by Harry Kleiner. This action-packed story begins with the defection of a Russian scientist whom Russian spies attempt to assassinate. He is injured in the attack and suffers a potentially fatal cerebral thrombosis. In an attempt to salvage whatever secrets the scientist might be harboring, the Pentagon assembles a team led by agent Grant (Stephen Boyd) and his beautiful assistant Ms. Peterson (Racquel Welch) to save his life. The 5-member team and their ship are zapped to microscopic size by vague and unspecified scientific means (the Methods section is weak) and injected into the scientist's blood stream so they can destroy the blood clot with a laser. Although Fantastic Voyage may have heightened interest in the concept of a miniaturized medical device that could travel through the body, the scientific discovery and development precedes that movie and dates to technology that began in 1954. The present design is attributed to the brilliant work of Iddan et al.1Iddan G Meron G Gluckhovsky A Swain P. Wireless capsule endoscopy.Nature. 2000; 405: 417Crossref PubMed Scopus (2489) Google Scholar In this issue of Gastrointestinal Endoscopy, Lewis and Swain2Lewis BS Swain P. Capsule endoscopy in the evaluation of patients with suspected small intestinal bleeding, a blinded analysis: the results of a pilot study.Gastrointest Endosc. 2002; 56: 349-353Abstract Full Text Full Text PDF PubMed Scopus (533) Google Scholar report on the use of capsule endoscopy in the evaluation of patients with suspected small intestinal (SI) bleeding. The authors chose the term “suspected SI bleeding” because the patients had all had gastrointestinal bleeding (GIB) and had undergone colonoscopy (COL), upper endoscopy (EGD), and small bowel series without discovery of a source. Such GI bleeding is often categorized as obscure, and it can also be subdivided into those cases in which the bleeding is overt, but obscure (i.e., when the patient has unexplained melena or hematochezia) or cases in which the bleeding is occult, but obscure (i.e., when the patient has a positive fecal occult blood test [FOBT]). Another group of patients who have iron deficiency anemia (IDA) but a negative FOBT often have obscure GIB. Data are sparse on the frequency and natural history of obscure GIB. Some investigators report that it occurs in 3% to 5% of patients with GIB.3Szold A Katz L Lewis B. Surgical approach to occult gastrointestinal bleeding.Am J Surg. 1992; 163: 90-93Abstract Full Text PDF PubMed Scopus (142) Google Scholar, 4Thompson JN Salem RR Hemingway AP Rees HC Hodgson HJ Wood CB et al.Specialist investigation of obscure gastrointestinal bleeding.Gut. 1987; 28: 47-51Crossref PubMed Scopus (73) Google Scholar There is some variation in etiology by age with angioectasias (preferred term for telangectasia, vascular malformations), medication-induced ulcers, and watermelon stomach (more common in patients older than 40 years), and small bowel tumors, Crohn's disease, and polyposis more common in younger patients, although all lesions can occur in either group. Less common causes of obscure GIB include Meckel's diverticulum, Dieulafoy's lesion, portal gastropathy, aortoenteric fistulae, amyloidosis, and pancreatic and biliary lesions. A number of diagnostic studies can be used to evaluate patients in whom COL and EGD do not reveal the cause of obscure GIB. These include small bowel follow-through (SBFT), enteroclysis, red blood cell (nuclear) scan, Meckel's scan, angiography, enteroscopy, and intraoperative endoscopy. A detailed discussion of this topic is beyond the purview of this editorial, but a few salient points are these. SBFT leads to the diagnosis of obscure GIB in approximately 5% of patients, and with enteroclysis the yield is still under 10%.5Bashir RM Al-Kawas FH. Rare causes of small intestinal bleeding.Gastrointest Endosc Clin North Am. 1996; 6: 709-738PubMed Google Scholar, 6Rex D Lappas L Maglinte D. Enteroclysis in the evaluation of suspected small intestinal bleeding.Gastroenterology. 1989; 97: 58-60PubMed Google Scholar Nuclear medicine studies are theoretically appealing but are notoriously disappointing in “real life.” Angiography may be applied in the acute setting or when a patient is not thought to be actively bleeding and it may be used with or without provocative agents. Enteroscopy has now become a standard part of the evaluation of obscure GIB although there are numerous variables, as for example which instrument is used; whether fluoroscopy is necessary for guidance and documentation; whether an overtube is required; and whether antiperistaltic or prokinetic agents are helpful. Push enteroscopy yields a diagnosis in 30% to 50% of patients7Foutch P Sawyer R Sanowski R. Push enteroscopy for diagnosis of gastrointestinal bleeding of obscure origin.Gastrointest Endosc. 1990; 36: 337-341Abstract Full Text PDF PubMed Scopus (184) Google Scholar, 8Rutgeerts P Broeckaert L Willemse P Gevers AM Van Cutsem E Vantrappen G et al.Push enteroscopy has a higher yield than Sonde enteroscopy.Gastrointest Endosc. 1993; 39 ([abstract]): 257Google Scholar, 9Barkin JS Chong J Reiner D. First generation video-enteroscope.Gastrointest Endosc. 1994; 40: 743-747PubMed Google Scholar, 10Pennazio M Arrigoni A Risio M Spandre M Rossini FP. Clinical evaluation of push enteroscopy.Endoscopy. 1995; 27: 164-170Crossref PubMed Scopus (138) Google Scholar, 11Zaman A Katon R. Push enteroscopy yields a higher incidence of proximal lesions within the reach of the standard endoscope.Gastrointest Endosc. 1998; 47: 372-376Abstract Full Text Full Text PDF PubMed Scopus (179) Google Scholar when using either colonoscopes (length 160 cm) or specifically designed SI endoscopes (length 200-270 cm). Sonde enteroscopy uses an instrument 270 to 400 cm in length, which is designed to be carried by peristalsis to reach most of the small intestines. The diagnostic yield ranges between 26% and 77%.12Lewis B Waize JD. Chronic gastrointestinal bleeding of obscure origin.Gastroenterology. 1988; 94: 1117-1120Abstract PubMed Google Scholar, 13Gostout CJ Schroeder KW Burton D. Small bowel enteroscopy.Gastrointest Endosc. 1991; 37: 5-8Abstract Full Text PDF PubMed Scopus (69) Google Scholar The technique has not gained wide acceptance. Intraoperative endoscopy, which is underused, is the most invasive of the diagnostic options. However, it has the advantage of allowing therapy if an abnormality is found and it has a high yield.14Lopez M Cooley J Petros J Sullivan JG Cave DR. Complete intraoperative small-bowel endoscopy in the evaluation of occult gastrointestinal bleeding using the sonde enteroscope.Arch Surg. 1996; 131: 272-277Crossref PubMed Scopus (55) Google Scholar, 15Lewis B Wenger J Waye J. Intraoperative enteroscopy.Am J Gastro. 1991; 86: 171-174PubMed Google Scholar The study by Drs. Lewis and Swain compared capsule endoscopy to push enteroscopy for patients with obscure GIB.2Lewis BS Swain P. Capsule endoscopy in the evaluation of patients with suspected small intestinal bleeding, a blinded analysis: the results of a pilot study.Gastrointest Endosc. 2002; 56: 349-353Abstract Full Text Full Text PDF PubMed Scopus (533) Google Scholar The yield of push enteroscopy was 30% (6/20). VCE detected those 6 lesions and found an additional 5 (55%) beyond the reach of the enteroscope. The difference between the two did not quite reach statistical significance (p = 0.0625). The authors chose a study design that compared VCE with push enteroscopy. I believe that this was the correct comparison. One could argue that VCE had an unfair advantage because it would provide an opportunity to see much more of the SI and therefore a higher yield was predictable. They could have compared VCE with Sonde enteroscopy or intraoperative endoscopy, which visualizes more of the SI, but the former has little relevance for the practicing physician and the latter is more invasive. To assess the importance of the study and the relevance to one's own practice, the reader would need to note the patient population in the study, examine the methodology, and analyze the interpretation of the results. What about the patients included in the study? Are they typical of those seen in the practice of a gastroenterologist or surgeon who might evaluate obscure GI bleeding? For the most part they are, in that the source of bleeding was not uncovered with COL, EGD, and SBFT. However, they may be a group with more severe bleeding. Those 20 patients had been bleeding for an average of 3 years; had been transfused an average of 28 units of packed red cells; and had been hospitalized an average 2.9 times. Because this bleeding was substantial, it could be argued that the likelihood of a diagnosis being made by either VCE or push enteroscopy was high. Because these patients had already undergone 23 SBFTs, 22 push enteroscopies, 12 nuclear scans, and 3 angiograms with no cause being determined, it could also be asserted that the yield would be low. It also must be appreciated, however, that just as a rose is not a rose is not a rose, obscure bleeding is not obscure bleeding is not obscure bleeding. The evaluation of patients who had one episode of severe unexplained GIB may have a yield different from that for patients who had no overt bleeding but were only a positive FOBT. And does it matter if the latter group was iron deficient and does it matter if they were taking aspirin or nonsteroidal anti-inflammatory drugs (NSAIDs) at the time of the bleeding? And how important is the age of the patient or other known risk factors (e.g., previous aortic aneurysm repair)? Lewis and Swain provide some of this information, but they do not tell us whether the yield was different in the different subgroups mentioned above. It is also important in endoscopic studies to see whether the experience and technique of the endoscopist would be relevant to one's own practice. Dr. Lewis is an acknowledged expert, so his experience with push enteroscopy would be greater than most; however, the technique is widely practiced by most endoscopists and reasonably straightforward. In this study, an overtube was not used for push enteroscopy and some who believe it is helpful argue that this does not give push enteroscopy its best chance to compete with VCE. Because there is no firm scientific information to support this position and because in my own experience an overtube does not offer a distinct advantage, I would downgrade this argument. In trying to reach a conclusion about the value of VCE in assessing patients with obscure GIB, based on the Lewis-Swain study, an analyst might take either of the two “half-cup positions.” The optimist would say that the diagnosis was made in more than 50% of the cases and that no lesions seen on enteroscopy were missed. The pessimist, whose position is far more intriguing to examine, would say that VCE was not diagnostic in almost half of the cases (9 of 20) and that since VCE does not allow the option of therapy, as does push enteroscopy, one is still obligated to perform surgery if a lesion is found beyond the range of push enteroscopy. And if surgery must occur anyway, why not just do intraoperative enteroscopy instead of VCE? The pessimist could also argue that although VCE was given credit for discovering a bleeding site in 11 patients, in 4 of those 11, active bleeding was seen. So although the site was discovered, in those 4 the actual cause was not defined, and this is not as useful as discovering the site and the cause. Let us look at the 45% of patients in whom no diagnosis was made. Why was no diagnosis made? There could be several explanations. First, it is possible that no SI source of bleeding existed—ever, or at the time of the VCE. It is possible that apparent colonic angioectasias were not seen, or that a Dieulafoy's lesion was not apparent. It is possible that the FOBT was falsely positive. It is possible that NSAID-induced ulcers had healed by the time the last VCE was done. Obscure bleeding is not obscure bleeding is not obscure bleeding, and it is necessary to assess pertinent diagnostic modalities for each group. Second, it is possible that a lesion was present but that it was not seen by VCE. The device captures two pictures per second, and it does not matter whether the camera is looking forward or backward as it passes through the small bowel, but it is possible that pathology might not be captured. There are blind spots; there may be debris in the lumen even if the patient has fasted; and because there is no air insufflation, an abnormality may not be fully exposed. Third, a lesion may be missed if it is distal to the area that the capsule traverses before its battery expires. The battery is supposed to last 8 hours, but in our personal experience at Mayo Clinic Scottsdale it lasts an average of 7 hours and 3 minutes. In the current study, the capsule did not reach the colon in 3 patients. Fourth, the diagnostic abnormality must be observed by the physician who reads the study. There are some pitfalls. The video is usually reviewed on fast speed (~20 frames/second) so if a lesion is only captured for a few seconds, it could be overlooked. As mentioned, food debris and/or luminal secretions could obscure the pathology. And there is tedium associated with reading the studies. Without concentration on the part of the physician, a lesion could be missed. In the future, computers could be taught to spot and mark variations from normal, based on color or topography to shorten reading times or even to perform analysis without a human being (i.e., computerized electrocardiogram readings). It is to the investigators' credit that they required a second blinded reading for VCE and they found complete concurrence in their observations. Although the results of this study did not reach statistical significance, there was a strong trend suggesting a higher diagnostic yield with VCE than push enteroscopy. A larger study may have shown statistical significance, but it is comforting that several preliminary studies presented at an international conference on capsule endoscopy in Italy in March 2002 showed similar results to the study of Lewis and Swain published in full in this issue of Gastrointestinal Endoscopy.16Pennazio M Santucci R Rondonotti E Abbiati C Beccari G Luchetti R et al.Wireless capsule endoscopy in patients with obscure gastrointestinal bleeding: results of an Italian multicenter experience.in: Proceedings of the First Given Conference on Capsule Endoscopy, Rome, Italy, May 17-19, 2002 Rahash Printing, Haifa (Israel)2002: 61-62Google Scholar, 17Ell C Remke S May A Helou L Henrich R. A prospective controlled trial comparing wireless endoscopy to push enteroscopy in chronic GI bleeding.in: Proceedings of the First Given Conference on Capsule Endoscopy, Rome, Italy, May 17-19, 2002 Rahash Printing, Haifa (Isreal)2002: 27-28Google Scholar, 18Van Gossum A Francois E Schmitt A Deviere J. A prospective comparative study between push enteroscopy and wireless video capsule in patients with obscure digestive bleeding.in: Proceedings of the First Given Conference on Capsule Endoscopy, Rome, Italy, May 17-19, 2002 Rahash Printing:, Haifa, Israel2002: 31-32Google Scholar, 19Demedts K Van Assche G Hiele M Tack J Rutgeerts P. A prospective comparative study of capsule versus push enteroscopy in obscure GI bleeding.in: Proceedings of the First Given Conference on Capsule Endoscopy, Rome, Italy, May 17-19, 2002 Rahash Printing:, Haifa, Israel2002: 37-38Google Scholar, 20Gay G Savrin J Gaudin J et al.Comparison of wireless endoscopy and push enteroscopy in patients with obscure occult/overt digestive bleeding.in: Proceedings of the First Given Conference on Capsule Endoscopy, Rome, Italy, May 17-19, 2002 Rahash Printing:, Haifa, Israel2002: 51-52Google Scholar, 21Selby W Sheckele N. A comparison between the M2A capsule and push enteroscopy for obscure GI bleeding.in: Proceedings of the First Given Conference on Capsule Endoscopy, Rome, Italy, May 17-19, 2002 Rahash Printing:, Haifa, Israel2002: 53-54Google Scholar Although it is encouraging to learn that VCE can increase the diagnostic yield, there are 2 important caveats. The first is that there has been no reference standard measurement in any of the studies. A difficult but more scientifically rigorous study would be to compare VCE with intraoperative endoscopy. The second point is that making a presumed diagnosis is only the first step. None of the studies to date have demonstrated that the higher diagnostic yield alters outcome for patients in a positive way. It will be important to learn whether it leads to management that reduces further bleeding, eliminates further diagnostic studies, or provides a patient peace of mind by establishing a diagnosis even if no effective treatment can be rendered. When any new technology emerges, practicing physicians generally consider whether they wish to use that technology in their practice. Capsule endoscopy is somewhat unique among recently emerging endoscopic procedures in that one does not need to learn how to perform the procedure. However, there are other considerations. The reasons why physicians might want to begin offering capsule endoscopy are that (1) it seems clear that the procedure will benefit patients; (2) it is not clear that it is beneficial, but they wish to investigate the procedure; (3) it will enhance the image that theirs are cutting edge practices; (4) they are concerned they will lose patients if they do not perform VCE; and (5) either VCE will be profitable or their institutions are willing to take a loss. Reasons why physicians may not want to begin to offer capsule endoscopy are that (1) it is not clear that VCE is an important medical advance; (2) their practices do not see enough patients with obscure GI bleeding; (3) the time required to read the studies (~60-90 minutes) does not make economic or practice sense; (4) they would prefer to send such patients to a referral center; and (5) there may be malpractice concerns if a complication occurs or a diagnosis is missed. There are also issues that exist whenever a first-generation product reaches the market that will likely be addressed with subsequent renditions. Such issues exist with the VCE. Currently physician reading time is an hour or more. Advances in software development will hopefully scan the video and designate areas of concern. Battery time is now 8 hours or less. Longer lasting batteries are being developed. If an abnormality is seen, there is not a good way to define its location in the SI. The pylorus is usually seen, and in many patients the ileocecal valve can be demonstrated, but apart from a rough estimate linked to “time beyond the pylorus” or “time in front of the ileocecal valve,” specific localization is not possible. Prototype products use sensors similar to EKG leads to help define the location and a cartoon pathway can be generated. Reimbursement for the procedure is not ideal for such a time-consuming activity. Some private insurance companies pay a professional and/or faculty fee, but Medicare currently offers no reimbursement. The main application for VCE to date has been for the evaluation of obscure GI bleeding of suspected SI origin. Are there other applications for diseases of the SI? Numerous studies are underway assessing other conditions that affect the small bowel. Several centers are evaluating the capsule for the diagnosis of Crohn's disease, hereditary hemorrhagic telangiectasias, polyposis syndromes, and SI tumors. There is also preliminary work evaluating the SI in patients with acquired immunodeficiency syndrome, celiac disease, immunodeficiency syndromes, small bowel transplants, and those taking medications such as NSAIDs. Patients with irritable bowel syndrome, unexplained abdominal pain, and diarrhea are undergoing evaluation. Although the FDA approved VCE for “visualization of the small bowel mucosa as an adjunctive tool in the detection of abnormalities of the small bowel,” what about its utility for conditions that affect the esophagus, stomach, or colon? The esophageal transit time is brief (<5 seconds in patients studied at Mayo Clinic Scottsdale) when patients ingest the capsule with water in an upright position. In some, however, the squamocolumnar (SC) junction is vividly seen. By having the patient ingest the capsule lying horizontally, the SC junction can be seen. If the cost of the product/procedure were reduced, VCE could possibly be used to screen for Barrettapos;s esophagus. Currently the colon is not well seen with VCE. There are several limitations. Stool obscures observation of colonic mucosa. The battery wears out before much of the colon is seen. And even if those two problems were overcome, it is probable that the entire colon, with a large diameter and less peristaltic activity, would not be well seen. Having said that, studies are beginning to assess the possible role of the VCE for colon cancer screening. In conclusion, what can be said about the role of the VCE for obscure GI bleeding? Although the study by Lewis and Swain does not prove with statistical significance that VCE has a higher diagnostic yield than push enteroscopy for obscure bleeding, the trend from their study, my view of other reports,16Pennazio M Santucci R Rondonotti E Abbiati C Beccari G Luchetti R et al.Wireless capsule endoscopy in patients with obscure gastrointestinal bleeding: results of an Italian multicenter experience.in: Proceedings of the First Given Conference on Capsule Endoscopy, Rome, Italy, May 17-19, 2002 Rahash Printing, Haifa (Israel)2002: 61-62Google Scholar, 17Ell C Remke S May A Helou L Henrich R. A prospective controlled trial comparing wireless endoscopy to push enteroscopy in chronic GI bleeding.in: Proceedings of the First Given Conference on Capsule Endoscopy, Rome, Italy, May 17-19, 2002 Rahash Printing, Haifa (Isreal)2002: 27-28Google Scholar, 18Van Gossum A Francois E Schmitt A Deviere J. A prospective comparative study between push enteroscopy and wireless video capsule in patients with obscure digestive bleeding.in: Proceedings of the First Given Conference on Capsule Endoscopy, Rome, Italy, May 17-19, 2002 Rahash Printing:, Haifa, Israel2002: 31-32Google Scholar, 19Demedts K Van Assche G Hiele M Tack J Rutgeerts P. A prospective comparative study of capsule versus push enteroscopy in obscure GI bleeding.in: Proceedings of the First Given Conference on Capsule Endoscopy, Rome, Italy, May 17-19, 2002 Rahash Printing:, Haifa, Israel2002: 37-38Google Scholar, 20Gay G Savrin J Gaudin J et al.Comparison of wireless endoscopy and push enteroscopy in patients with obscure occult/overt digestive bleeding.in: Proceedings of the First Given Conference on Capsule Endoscopy, Rome, Italy, May 17-19, 2002 Rahash Printing:, Haifa, Israel2002: 51-52Google Scholar, 21Selby W Sheckele N. A comparison between the M2A capsule and push enteroscopy for obscure GI bleeding.in: Proceedings of the First Given Conference on Capsule Endoscopy, Rome, Italy, May 17-19, 2002 Rahash Printing:, Haifa, Israel2002: 53-54Google Scholar and my own experience leads me to that conclusion. However, I do not feel that it obviates the need for push enteroscopy, but quite the contrary, sets up the following algorithm. For the patient whose GI bleeding is not explained by standard COL and EGD, and in whom there is no acute overt bleeding, VCE should be the next study. It should replace SBFT, enteroclysis, nuclear scans, and angiography for this purpose. If VCE is positive for an abnormality in the proximal SI, enteroscopy should be performed with intent to obtain biopsy specimens and/or treat the lesion. If VCE is negative and the patient is stable, an enteroscopy may not be necessary. Having laid out this algorithm, which I think to be correct, I do not feel that there are sufficient data to establish this as a standard of care. Therefore the publication of complete data from the studies reported in the preliminary abstract form to which I referred16Pennazio M Santucci R Rondonotti E Abbiati C Beccari G Luchetti R et al.Wireless capsule endoscopy in patients with obscure gastrointestinal bleeding: results of an Italian multicenter experience.in: Proceedings of the First Given Conference on Capsule Endoscopy, Rome, Italy, May 17-19, 2002 Rahash Printing, Haifa (Israel)2002: 61-62Google Scholar, 17Ell C Remke S May A Helou L Henrich R. A prospective controlled trial comparing wireless endoscopy to push enteroscopy in chronic GI bleeding.in: Proceedings of the First Given Conference on Capsule Endoscopy, Rome, Italy, May 17-19, 2002 Rahash Printing, Haifa (Isreal)2002: 27-28Google Scholar, 18Van Gossum A Francois E Schmitt A Deviere J. A prospective comparative study between push enteroscopy and wireless video capsule in patients with obscure digestive bleeding.in: Proceedings of the First Given Conference on Capsule Endoscopy, Rome, Italy, May 17-19, 2002 Rahash Printing:, Haifa, Israel2002: 31-32Google Scholar, 19Demedts K Van Assche G Hiele M Tack J Rutgeerts P. A prospective comparative study of capsule versus push enteroscopy in obscure GI bleeding.in: Proceedings of the First Given Conference on Capsule Endoscopy, Rome, Italy, May 17-19, 2002 Rahash Printing:, Haifa, Israel2002: 37-38Google Scholar, 20Gay G Savrin J Gaudin J et al.Comparison of wireless endoscopy and push enteroscopy in patients with obscure occult/overt digestive bleeding.in: Proceedings of the First Given Conference on Capsule Endoscopy, Rome, Italy, May 17-19, 2002 Rahash Printing:, Haifa, Israel2002: 51-52Google Scholar, 21Selby W Sheckele N. A comparison between the M2A capsule and push enteroscopy for obscure GI bleeding.in: Proceedings of the First Given Conference on Capsule Endoscopy, Rome, Italy, May 17-19, 2002 Rahash Printing:, Haifa, Israel2002: 53-54Google Scholar and others will be needed to confirm the results of initial studies such as that of Lewis and Swain in this issue. In addition, outcome data are necessary. What happens to the patient in whom capsule endoscopy led to a diagnosis, and what happens to those in whom it was not diagnostic? Two final points: First, the importance of this new technology should not be underappreciated. The ability to visualize the mucosal surface of the entire small bowel and to observe its motility is a revolutionary advance. The opportunity to study the anatomy and function of the normal intestine will be as important as viewing pathology. Second, capsule endoscopy provides a new paradigm for endoscopy and for medicine. It differs from other endoscopic procedures in that it does not require the patient to be tethered to a physician or a medical facility while they are being tested. That has important implications. It means that GI endoscopy does not necessarily depend on long black flexible instruments. It means that the individual performing endoscopy and evaluating the GI tract may not be a GI endoscopist. It could be an internist or a family physician. Or a technician. So far the voyage with capsule endoscopy has been fantastic. At this time, however, there is little evidence that it affects outcome or changes what we do. In the future, I think that it will.
Referência(s)