The emergence of the fiber optic flexible endoscope as the "veterinarytoy of the 90s" has been fostered by the belief that the ability toobtain biopsies from the gastrointestinal tract is an important tool inthe management of acute and chronic gastrointestinal disease in dogs andcats. Certainly, these endoscopes can do other tasks like visualize tumorsor foreign bodies, but their ability to obtain tissue samples is the bigselling feature.

Much has been said and written by various "gurus" of clinicalgastroenterology about the uses of endoscopy, particularly its role inthe diagnosis of inflammatory bowel disease in dogs and cats. Rarely doesone have the chance to hear or read the opinions of the pathologists who,in the end, must interpret the tiny biopsies obtained. I know of many pathologistswho treat such biopsies with great scepticism, and there certainly is agap between what the clinical gastroenterologist would like us to say,and what we in fact are prepared to say, about these samples.

Over the last 10 years or so, I have looked at over 6,000 gastrointestinalbiopsies. This experience, added to my original training in gastrointestinalpathology, allows me to give you a perspective that is different from whatyou may have heard at the various meetings or read in clinical review articles.

Endoscopy versus full-thickness biopsy:
The greatest advantage of endoscopic sampling, in addition to the obviousadvantages related to the ability to avoid invasive surgery, is the abilityto obtain a large number of samples in a relatively short period of time.Typically, one would take 6-8 gastric samples, 2-3 duodenal samples, andanother half dozen from the colonic mucosa. Endoscopic biopsies are limitedto the mucosa, but are perfectly adequate when one is attempting to diagnosemany types of inflammatory bowel disease, in which the ability to widelysample random areas of mucosa is more important than obtaining full thicknesssamples. In contrast, full thickness biopsies are more arduous and usuallyI receive only a single sample from each region of the gut. Full thicknessbiopsies do have several distinct advantages, however. The most obviousis that you are able to sample jejunum and ileum, which normally are notaccessible to even the most skilled endoscopist. Secondly, there are anumber of conditions for which sampling the full thickness of the gut wallis important. This is particularly true of lymphangiectasia (in which thedilated lymphatics are much more obvious in submucosa and tunica muscularisthan they are in the mucosa), and many examples of intestinal neoplasiain which the transmural spread by the atypical cells is the only certainway to diagnose the malignancy. Inflammatory bowel disease - is it real?
The gigantic supposition that underlies the clinical syndrome known asinflammatory bowel disease is that an increase in leucocytes within theintestinal lamina propria is somehow directly related to the productionof vomiting or diarrhea. The supposition may be correct, but in fact itis nothing more than a supposition that has been repeated so often thatit is now assumed to be true.

The normal cellularity of the dog and cat intestine has never been defined,so our ongoing struggle to pinpoint those patients in which the cellularityis abnormal is based on no foundations whatsoever. I can guarantee youthat the same intestinal biopsy, shown to 10 different pathologists, islikely to generate a 50:50 split in terms of who thinks it is inflammatorybowel disease and who would classify it as falling within the broad rangeof normal. Why, one must ask, has such an obvious question never been addressed?The problem is one of the huge cost for such a study, which would requirevery large numbers of dogs and cats maintained for several years, so thatwe could study the influence of such things as diet, age, hormone fluctuation,concurrent disease, and previous or current subclinical parasitism on intestinalcellularity.

The second flaw in the assumption that increased leucocyte numbers somehowcause disease is more theoretical. A parallel argument would be that anincrease in lymphocyte numbers within lymph node (lymphoid hyperplasia)signifies malfunction of that lymph node. This is clearly preposterous,since we assume that a hyperplastic lymph node is only one that has beenstimulated to do its job. Why then do we assume that the intestine, whichis the largest lymphoid organ in the body, is somehow malfunctioning whenit undergoes lymphoid hyperplasia? A valid approach to solving the problem,as an alternative to very expensive characterization of normal intestinalvariability, would be to examine the cellularity of intestines that arethe site of active disease, and then do a repeat biopsy when the diseasehas spontaneously disappeared or when it has responded successfully totherapy. Alas, it is very difficult to convince owners to submit theirpets to this second series of biopsies just to satisfy our curiosity, andthey are usually grateful that the diarrhea or vomiting has subsided -and wish to proceed no further! Unlike skin disease, where we have learneda great deal by charting the response of the skin to a single therapeuticprotocol, we tend to use a wide variety of therapies when trying to controlgut disease, and thus we cannot even use the clinical benchmark of responseto therapy as we attempt to answer the question about how many cells istoo many, or what kinds of cells are "evil" cells. In my judgement,we have made virtually no progress over the last 10 years in addressingthis most fundamental of all questions about intestinal disease.

What is the correlation between endoscopic observation and histopathology?
In several studies, the correlation between what you see with the endoscopeand what I would eventually see with the microscope is not particularlygood. It varies with the nature of the endoscopic observation. Things likemucosal redness, edema, and granularity are seldom supported by histologicobservation. This is not to say that the endoscopic observations are invalid,but merely that these changes reflect subtle hyperemia and edema that arenot easily perceived microscopically. I think it is true to say that theuniversal rule, used by experienced endoscopists, is that one should biopsya random selection of mucosa regardless of whether it looks normal or abnormal.It is also true, incidentally, that the most dramatic microscopic lesionsare not necessarily found in that portion of gut which seems to be responsiblefor the clinical signs at that particular time. In short, there may bevery good colonic lesions in animals that are vomiting without diarrhea,and conversely the gastric lesions may be very dramatic in animals thatare afflicted primarily with diarrhea.

Is endoscopy difficult to learn?
I certainly am no expert to answer that question, but I can tell you thatthe great majority of private practitioners who have purchased endoscopesare sending me good quality samples on their first or second attempt. Evenexperts will hit the occasional animal that seems most reluctant to giveup any decent mucosal samples, but in general the learning time appearsto be very short. It is not an instrument which one should fear.

If pathologists cannot diagnose inflammatory bowel disease, is it worthgetting an endoscope?
I do not wish to paint too pessimistic a picture. There are many casesin which there are lesions like mucosal fibrosis, villus atrophy, epithelialulceration, or massive eosinophilic or neutrophilic infiltration whichare quite clearly the result of some kind of mucosal inflammatory disease.Endoscopic samples can also pick up unsuspected neoplasia, particularlyintestinal lymphoma, and can diagnose some specific intestinal infectiousdiseases. My comments above were directed specifically at the vague entityknown as inflammatory bowel disease, which is defined as a relatively normalintestinal morphology but with an excess proprial population of lymphocytes,plasma cells, or eosinophils. I stand by my claim that the normal rangefor these particular cellular elements is not known, and that it is thereforevery subjective to claim that this dog or cat has "too many"of any one of these cell types. It is also undoubtedly true that the intestineof each animal is unique in its flora and unique in its cellularity, sothat a biopsy that lies well within the normal range for the whole populationmay well be very abnormal for a particular individual - or vice versa.If I have portrayed the diagnosis of inflammatory bowel disease as a verysubjective one, rather than as being based on well documented and widelyaccepted criteria, then I have portrayed exactly the correct message. Ithink it is very much that the technology to obtain the biopsies is severalyears ahead of our ability to interpret them. What is particularly worrisomeis the lack of evidence of any serious moves, on anyone's part, to overcomeour ignorance in this particular area.

What about bacterial gastritis?
The discovery 10 years ago that the bacterium Helicobacter pylori is amajor contributor to the syndrome of chronic gastric ulceration in peoplespurred a tremendous flurry of activity in the veterinary field. Part ofthe enthusiasm, undoubtedly, is driven by the desire to prove that gastriculceration or chronic gastritis in dogs and cats also has an easily treated,bacterial pathogenesis. It is equally true that another major stimulusfor the flurry of activity is the prospect of the huge dollars that wouldflow towards the researchers that are able to establish dogs or cats asvalid models for the human disease. Based on our current understanding,we know that Helicobacter and similar organisms are part of the normalflora of both dogs and cats (as well as many other mammals). Colonizationby these organisms results in an increase in overall cellularity withinthe lamina propria and in mucosal thickness, but in most experimental modelsthere are no associated clinical signs. A similar phenomenon occurs throughoutthe intestinal tract at weaning, and seems to be a nonspecific responseto increases in intralumenal antigen load. We know that some dogs and catswith gastritis have an increased number of these organisms, but it is byno means clear whether the increase is the cause of, or the result of,the gastritis. There are many examples, throughout the gut, of shifts inflora that are the result of altered local environments, but we persistin assuming that any shift in flora must be the cause of the enteric lesionsrather than simply the result. The jury is still out on this one. A diagnosisof helicobacter gastritis is, in some laboratories, the "flavor ofthe month". At the moment, it is my policy to carefully examine gastricbiopsies for any evidence of overgrowth by spiral bacteria, and to reportthat observation when it occurs. It may be that specific antibiotic therapyin such cases will indeed be beneficial, regardless of whether the organismsare primary pathogens or are simply part of an opportunistic overgrowth.

When in doubt . . . ask!
If you have any doubts about whether endoscopic biopsy would be appropriatefor a specific clinical case, I would welcome the chance to speak withyou. I may not be able to answer your question, but it is sometimes prudentto at least consult the person who in the end will be looking at your biopsies!Perhaps here I can provide answers to a few of the more commonly-askedquestions.

Do I need to put samples from each region in separate bottles?
The answer is no, because it is easy to distinguish the various parts ofthe stomach from one another, and from small intestine or colon.

Do I need to place the biopsies on pieces of cardboard?
The answer again is no, mostly because they all fall off anyway!

Why do I need to take so many samples?
There are two answers for this one. The first is that orientation of thebiopsies is hit-and-miss, so that some of the samples will inevitably bemaloriented and will not give me the information that I need. The second,and more significant, reason is that intestinal disease (particularly gastricdisease) tends to be patchy, and it is quite common to have one very abnormalsample in the midst of 2-3 that are either completely normal or equivocal.

In addition, many gastrointestinal diseases have widespread histologiclesions even if the clinical signs seem to point to very local disease.Histologic changes in colon, for example, may greatly strengthen my faithin the validity of subtle gastric lesions, or vice versa.

Are there any diseases which cannot be diagnosed by endoscopy?
There is no easy answer to this one. Obviously, any disease which is purelyfunctional will not be diagnosed by biopsy of any type. However, amongthose that have histologic lesions, the most difficult to diagnose by endoscopyare those in which the character of submucosal or transmural lesions isan important part of the diagnosis. Under that category, I would put thingslike gastric carcinoma (most of the tumor actually lies in the tunica muscularis)and some cases of malignant lymphoma. Even though lymphoma always has lesionswithin the lamina propria, the neoplastic character of the disease is mosteasily seen as the lymphocytes move through the submucosa and into thetunica muscularis. It is difficult to pick out early neoplastic infiltrateinto a tissue like the lamina propria which already is rich in reactivelymphocytes and other mononuclear cells. Lymphangiectasia can be a difficultdiagnosis based on mucosa biopsies, because the presence of dilated lactealswithin each villus in neither sensitive nor specific. I think it can beinduced by a biopsy artifact, and I know of several cases of absolutelyconvincing lymphangiectasia in which the initial endoscopic biopsies failedto demonstrate the dilated lymphatics. It is worth a try, but you mightfind the answer to be less definitive than you had hoped for.