In October 1994 at a weekend conference, a group of experts developed criteria for the the definition of Lyme surveillance testing. The outcome of this conference led to a maelstrom of controversy which continues.
The test has never been re-evelated even though two key bands: 34 OspB and 31 OspA were ommitted.
It was decided that a two tier protocol should be followed. The criteria for the IgM Western Blot came from the work of Engstrom et al. Three strains of Lyme were found to perform equally well. The 23 band, identified as the OspC band, the 39 band, and the Fla (41) band were chosen for the assay. A positive result was 2/3 of these bands. It was noted that the 37 band could be added and that 2/4 of these bands would increase the accuracy of the test. (not accurate)
Dressler et al proposed a different criteria for IgM scoring. A different strain of Lyme was used which did not express the 39 band well. A positive result was the presence of 2/8 bands: 18,21,28,37,45,58 and 93. This test performed about as well as the above test.
Two radically different test and test criteria were found to produce almost the same results!
Padula et al found the 23 band to be of diagnostic significance, especially in early Lyme - and that "this protein may be poorly expressed in a number of North American strains of Borrelia burgdoreri." Weinstein and Johnsone report: "Despite these limitations, the proposed criteria for a positive immunoblot in late Lyme disease - at least 5 or the 10 specified IgG bands - seemed to stand up reasonably well in other laboratories...The criteria (will) be used..pending further studies."
Where did the 37 band go?
Stony Brook reports it.
Ah, the missing band that would have made the CDC test more accurate by its own admission.
Two labs prepare their own Lyme Western Blot kits: IgeneX and Stony Brook. IgeneX has developed criteria for a positive result (IgG and IgM) - sounds like the Dressler criteria. They report a positive result if there is a reaction with two highly specific bands, inclding: 23,31,41,34,39,and 93.
Different assays have produced different bands. Dressler reported a 21 band. Stony Brook only reports a 20 band.
Are you as confused as I am?
In addition, in my experience I have found reported Western Blot results differ amongst commonly used "LLMD" laboratories.
In 1994 Gubler reports the two tier test is not the gold standard. (culture is). He states that "in the not too distant future, the development of new tests that use a cocktail of recombinant antigens or chimeric antigens(will increase)the sensitivity and specificity of serological tests for B. burdorferi."
Its been more that 17 years. We are still waiting.