From the LymeLight Newsletter of the Lyme Disease Foundation

---

When in 1994 the Centers for Disease Control and Prevention (CDC) and the Association of State and Territorial Public Health Laboratory Directors (ASTPHD) developed interpretive guidelines as to what constitutes a positive Western blot, activists, frontline physicians and many scientists criticized the criteria as too stringent and, in some aspects, unscientific.

According to the guidelines, known as the CDC or Dearborn criteria, patients who have Lyme-like symptoms longer than 4 weeks must present with at least 5 of 10 specific IgG (later-stage) antibodies for a positive test. Such stringent guidelines require the Western blot to be an extremely precise test, capable of definitively identifying the exact antibodies that show up in a patient's serum sample.

At the Lyme Disease Foundation's 14th International Scientific Conference on Lyme Disease and Other Tick-Borne Disorders, researcher Paul T. Fawcett, PhD, Alfred I. duPont Hospital for Children, gave attendees a comprehensive presentation on how "some groups" (CDC and ASTPHD come to mind) have unrealistic expectations of the Western blot's scientific capabilities for routine use, and that applying the CDC criteria to the test is not the best way to determine who has Lyme disease (LD).

In theory, according to Fawcett, the Western blot linearly separates all known proteins of the LD spirochete, Borrelia burgdorferi (Bb). This is accomplished by adding a detergent to a piece of the bacterium that breaks it into little pieces.

This mixture is then placed at one side of a gelatin-like strip and charged with an electrical current. This causes the various proteins, known as antigens or protein markers, to separate linearly by weight.

Lighter proteins move more quickly to one end and each is separated into kilodaltons (kD), a designation of their molecular mass. A very thin sheet of paper is placed on top, and the proteins are transferred to the paper, which is then cut into strips.

Patient serum is then added to the strips to detect antibodies, and the patient's serum sample is added over the paper. Any bacterium-specific antibodies then bind to the corresponding bacterium pieces below. A reagent is added that causes a color change when antibodies are present.

Dr. Fawcett's said that during this process it is important that the proteins separate, transfer, and line up where they are supposed to be. He said subtleties in the separation process, however, often cause the proteins to become improperly aligned with an incorrect molecular weight (e.g. 41kD) when a template identifying each band is placed over the blot.

Such migration subtleties can lead to antibodies being incorrectly identified upon Western blot analysis, Fawcett said.

"A lab cannot be sure that within a region of one millimeter if its looking at a 21, 22, or a 23 kD band," he said. "Apply rigorous interpretive criteria like the CDC criteria, and you have the inability of a lab to be absolutely sure all bands are properly aligned [to assure a truly accurate result]."

Dr. Fawcett said it is possible, but not economically feasible in a commercial laboratory setting, to make sure all bands are properly aligned with their proper molecular weight value on the Western blot template.

While presenting his analysis of a serologic evaluation of 100 pediatric patients with Lyme arthritis, Fawcett said he found that the 10 most common bands (antibodies) to show up were not the 10 bands that make up the CDC criteria. Of the 10 most frequent bands expressed, Fawcett said only 41 kD, was present 100% of the time. Only one other band, 39 kD, was present in more than half of the patients, he said.

This reveals another obstacle LD patients face when they try to obtain a diagnosis.

"There is a high degree of variability in the human immune response" to the LD bacteria that the CDC criteria does not take into account, Fawcett said. "There are no magic bands."

Scientists believe a variety of reasons, including strain variation, the fact Bb is polymorphic (able to change the proteins on its outer surface, thus causing different antibodies to be produced) and the health of the patient's immune system are some explanations for this variability.

Illustrating this variability, Dr. Fawcett reviewed test results of pediatric patients with long-standing Lyme disease, many of whom exhibited severe joint involvement.

"Some patients had 17, 18 bands show up on their blots," he said. "Yet they were negative by Dearborn standards."

(This is also significant because it often result s in infected patients being denied treatment because they are declared "bacteria-free", and it excludes infected patients from studies that require positive Western blots as entry criteria.)

Dr. Fawcett, head of DuPont's immunology laboratories, is an expert in Lyme disease serology. He participated in the Dearborn "consensus" meeting, where he discussed criteria for LD serology.

At the meeting, he said he voiced his (and many other scientists') concerns that there were other antibodies that often show up on Western blots in LD patients, and that the committee was "ignoring" those bands. (As was reported in LymeLight Vol.2, 1995, many scientists in attendance felt the committee's criteria and the conclusions to the meeting were predetermined, and dissenting opinions were not seriously considered.)

His study, "Use of Western blot and Enzyme-Linked Immunosorbent Assays in the Diagnosis of Lyme Disease" (Pediatrics, 1991:88:465-470) is cited in the study that became the basis for the CDC criteria.

Dr. Fawcett went on to review DuPont's own algorithm for interpreting the Western blot. It requires the 41kD and one band in the 60 - 69 range in addition to 2 other bands of any kD be present for a positive result.

"If we don't see any of these bands, we're highly suspicious" that the patient doesn't have Lyme disease, he said. He also said virtually all LD patients, with the exception of an extremely small minority with chronic, late-stage neurologic disease, exhibit the 41kD antibody in his lab.

Using DuPont's algorithm, combined with a thorough analysis of clinical symptoms to make a diagnosis, Fawcett said he has obtained 100% percent specificity (ability to detect only patients with LD) in LD diagnosis and that, compared to the CDC criteria, he observed "no significant" change in sensitivity (ability to detect all patients with LD).

Analysis of each patient groups' response to antibiotic treatment, he said, revealed responders were seropositive, while non-responders among this pediatric group were with few exceptions seronegative.

Running identical serum samples on the two Food and Drug Administration - approved Western blot kits, Dr. Fawcett found each test provided different results, a situation many Lyme-literate physicians say they often encounter.

"There are some real problems out there," he said.

Despite such problems in Western blot testing, Dr. Fawcett told the audience he still believes the test may still be the best available tool for diagnosing LD.

"If you understand the Western blot and know its limitations and potential problems, it can perform as well as it was originally touted," he said. "[The problem is] some groups have unrealistic expectations as to what it can do, some of which go beyond the scientific capabilities of this particular assay format."

To subscribe to the LymeLight