<<Disclaimer: Verify this information before applying it to your situation.>> Testing for CD -------------- Many different tests have been used or proposed for CD screening: * Routine blood count, serum chemistry, etc. These are general health-screening tests. They are only useful if they come back abnormal, and for many celiacs these tests come back normal. * Absorption tests. There is a xylose test in which the patient drinks a "foreign" sugar solution and then the level of this sugar in the blood or urine is monitored. If it is properly absorbed then the level should rise at a certain known rate. You need to have a normal intestine in order for it to be absorbed properly, so it is not a bad test. It is not painful and picks up many celiacs but not all. * Antibody blood tests. These are fairly specific to CD. One disadvantage is that these tests don't pick up other causes of malabsorption, which may be why they have not been as popular in the USA as they have been in other countries. * The small intestine biopsy remains the "gold standard" for diagnosing CD. At the Finland symposium there was a whole session devoted to debating how many biopsies are needed to diagnose CD. There are still experts who believe that three biopsies are needed, while many others think that one is enough. However, none of the 392 celiac experts at the symposium thought CD could be diagnosed without a biopsy. (Dr. Murray thinks that in the next 4-5 years blood tests may become reliable enough to diagnose CD in some circumstances without a biopsy.) * Small bowel x-rays. This is where you swallow a white chalky liquid. This test is pretty much useless for diagnosing CD. (30-40 years ago the barium solution used behaved differently so that it wasn't a bad test for CD, but with modern barium it is not useful.) * An empirical trial of the GF diet. Why not just go on the GF diet and see if you get better? Well, suppose you try a GF diet for six months and you feel better. But you decide the diet is a real hassle, so you want to be sure you really have CD. You then visit Dr. Murray to get a definite diagnosis. What you are faced with is a gluten challenge, which is not fun for anybody. Otherwise, there is no way to know if you are a celiac or not. If you have a biopsy after six months on a GF diet, the pathologist is probably not going to be able to tell you anything definitive. So the moral of the story is: Get a biopsy and a confirmed diagnosis before going on a GF diet. There are very few exceptions to this rule. On rare occasions where the patient is not physically well enough to undergo a biopsy, or the health care company completely refused to pay for it and the patient can't afford to pay for it themselves, Dr. Murray has diagnosed CD on the basis of a positive antibody test followed by a good cli nical response to the GF diet. But this is absolutely the last alternative to pursue. In the past, three biopsies were required to diagnose CD. The European Society for Pediatric Gastroenterology and Nutrition (ESPGAN) developed criteria in 1969-1970 for defining CD. They said you needed three biopsies: The initial biopsy to show the damaged villi, a second biopsy after being on a GF diet showing that the villi have healed, and a third biopsy showing damaged villi after a gluten challenge. In 1990 ESPGAN revised their criteria. They now say that in most cases only one biopsy is needed. If there is a clinical response to a GF diet (meaning the symptoms go away), so that there is some definable improvement, then no additional biopsies and no gluten challenge are necessary to make the diagnosis. Now there occasionally may still be a second biopsy to show healing, but it is not specifically to confirm diagnosis. The second biopsy is usually for a different reason, such as looking for a complication or making sure that healing has occurred (and sometimes it doesn't occur despite the best efforts of doctor and patient). The gluten challenge is now reserved for less than 10% of the cases that Dr. Murray sees; usually those that have minimal damage and in whom it is not easy to determine if the GF diet has really helped. Older patients may respond to the GF diet but not completely heal. Why not? Well, they may be healing fully further down the small intestine but not in the area near the stomach, which has been exposed to the most gluten for the longest time. However, it is only in the area near the stomach that you can effectively take small intestine biopsies. So a biopsy might not show the true extent to which healing has occurred. What about all the atypical cases; people who present, but don't yet have symptoms? Where do they fit in the ESPGAN criteria? And do milder degrees of damage cause problems? The other thing is the response to a gluten challenge. Some people don't get sick at all; they don't get any symptoms. They may have damage in their gut but no symptoms, sometimes even after six months. There was one patient that went 14 years on a gluten challenge before they developed intestinal damage. This tells us that if a celiac eats gluten there won't necessarily be any reaction, so you can't use how you react to judge whether or not an item is GF. On the other hand, Dr. Murray said that some people will get intestinal changes within two hours of ingesting gluten; this has been shown by taking a biopsy of a fast-reacting celiac before and two hours after some gluten was ingested. If a celiac is having a gluten challenge Dr. Murray looks for symptoms to begin occurring again, or if not then he follows the antibody tests until they become positive. After the return of symptoms, or a positive antibody test, or after about six months, then Dr. Murray generally performs a biopsy. There are two basic serological (blood) antibody tests: antigliadin (antibodies directed against a component of gluten) and connective tissue (antibodies directed against our own tissues). The antiendomysial test is an example of the second type of antibody test. It is very operator-dependent (unfortunately). A study reported in Gut in 1992 suggests that the celiac antibody blood tests are 100% sensitive and 99% specific to CD. Some HMOs have pointed to these studies as proof that a biopsy is unnecessary. But there are some problems with that study. It was based on a serum bank saved from previously-diagnosed celiacs. These were celiacs who were originally found because of screening with these blood tests. So negative results would automatically not be found in the serum bank. In September the University of Iowa tested seven reference labs in the USA. They took 20 serum samples from untreated celiacs, split them seven ways, and sent them to seven different labs across the country for analysis. All of the untreated celiacs were diagnosed by biopsy and had not previously had the antibody tests. They also sent 20 serum samples from people who were biopsy-proven to NOT have CD. Dr. Murray just got the data back in the last week. Each lab tested the samples blindly, using their own methods. The results: * The endomysial antibodies at all the labs were 100% specific. In other words, none of the 20 non-celiacs tested positive. * The sensitivity of the endomysial antibodies, meaning how often is the test positive for celiacs, varied a lot: from 55% to 90%. So with this test alone, you would not find all the celiacs. In other words, every lab had some false negatives on this test. * The sensitivity of the IgA and IgG antigliadin tests was quite variable. If you combined this test with the endomysial antibody test, the sensitivity is about 95-100%. In other words, if both of these tests came back negative from the same lab, the chances of having CD are less than one in twenty. So these tests are not bad. They aren't quite as good as some European studies have suggested, but they are not bad. (This is reassuring.) However, there is one caveat with these test results: the 20 celiac samples were "classic" celiac cases. These were not patients with mild damage. So patients with mild damage might not have abnormal antibody levels. [Yet these are the very patients that are hard to diagnose, and for which better tests are needed!--editor] Some of the potentials for these antibody tests include: * Screening whole populations, or higher-risk populations such as diabetics. * Monitoring the response to a GF diet. These tests should become negative after a celiac has been on a GF diet for awhile. There are problems with the blood tests. For example, there is no standardization between the various labs. None of the labs do these tests the same way, and none of them report the results the same way. A "high" positive from one lab is not necessarily going to be a "high" positive from another. That means you may not be able to compare results from one year to the next if the samples are sent to different labs. (This often happens as HMO's change contracts.) Who should be screened? Dr. Murray usually screens people with Type I diabetes, unexplained iron-deficiency anemia (Dr. Murray does not accept heavy menstrual flow in women as an explanation), unexplained bone disease, functional diarrhea, irritable bowel syndrome, history of "transient" gluten-sensitivity ("Well, I had it as a child and outgrew it..."--you DON'T outgrow CD), or lactose intolerance (if Caucasian). How should close relatives of a known celiac be screened? Obviously, a biopsy would be ideal as it would determine for sure whether or not villi damage is present. But a more reasonable first step is to use the celiac antibody blood tests to screen close relatives. Tissue typing could also be done to determine whether or not a relative is at risk for developing CD, but there is not enough information on the American population. The type of tissue you see in celiacs also occurs in about 20% of Caucasians. You also have to follow up intermittently, since a relative might not develop CD until after the first screening. For example, suppose you are a celiac and you have a child. Dr. Murray recommends keeping the child GF the first year, then putting them on a gluten-containing diet. At the age of two, or at the appearance of any symptoms, they would then be screened for CD using the antibody tests. Then check them again at age 9 or 10, before they go into that pubertal growth spurt.