Replacement for 30-Milliliter Flat-Bottomed, Glass-Stoppered, Round Bottles Used in VDRL Antigen Preparation

The VDRL test was developed in the 1940s (3, 4) and was first included in the Manual of Tests for Syphilis in 1949 (7). The VDRL antigen, which is a mixture of cardiolipin, lecithin, and cholesterol, is the basis for all nontreponemal tests. The working VDRL antigen has to be prepared each day and is good only for that day (5). With various modifications, the antigen has been stabilized for use in the unheated serum reagin test (USR), the rapid plasma reagin 18-mm-circle card test (RPR), and the toluidine red unheated serum test (TRUST) by adding stabilizers and pigments. Many laboratories in the United States have switched to these modified tests, but the VDRL test is the only test approved for testing spinal fluids to diagnose neurosyphilis. In addition, because the VDRL antigen is still the least expensive in terms of material cost, many laboratories in resource-poor settings use the VDRL test rather than USR, RPR, or TRUST.

The current procedure for preparing a working dilution of the VDRL antigen requires the use of 30-ml flat-bottomed, glass-stoppered, round bottles (5). The original procedure specified only that a glass-stoppered or screw-capped 30-ml bottle be used for preparing the antigen emulsion (3, 4). The need for either a flat- or concave-bottomed bottle rather than a convex-bottomed bottle was first included as a note in the 1964 Serologic Tests for Syphilis, which also specified glass-stoppered, narrow-mouth bottles (2). The use of flat-bottomed bottles was first specified in the 1969 manual (6) and has been the standard ever since. The use of flat-bottomed glass bottles was specified to ensure that pooling of the liquid at either the periphery or center of the bottle, which could result in rough antigens, did not occur.

We have received calls from users indicating that the specified bottles are no longer being made and asking whether there is another source or a substitute. Since we knew of no other source, a substitute was needed. Our objective with this study was to compare VDRL antigens made in the traditional 30-ml glass-stoppered, flat-bottomed, round bottles with those made in 25-ml glass-stoppered Erlenmeyer flasks with a slightly convex bottom to determine whether the flasks were an acceptable replacement for the discontinued bottles. Twenty-five-milliliter Erlenmeyer flasks were chosen because the volume and diameter of the flask are similar to those of the 30-ml flat-bottomed bottle.

Reference VDRL antigen (Centers for Disease Control and Prevention [CDC], Atlanta, Ga.) made in a traditional bottle and in five 25-ml glass-stoppered Erlenmeyer flasks were tested in parallel with 52 reactive serum samples (38 of which were strongly reactive and 14 of which were either reactive or weakly reactive at a 1:1 dilution) and 54 nonreactive serum samples. These serum samples were from individual bulk sera which are used to prepare control sera. In addition, one reactive reference control serum (CDC), one reactive and one nonreactive serum from a serum control panel (CDC), and one weakly reactive commercial control serum (Cenogenics Corporation, Morganville, N.J.) were tested with two different lots of commercial antigen (Lee Laboratories, Grayson, Ga.) and two reference antigens (CDC), each made up in two traditional bottles and two 25-ml Erlenmeyer flasks. During preparation of the working dilution of the VDRL antigen, 0.4 ml of VDRL-buffered saline is added to the 30-ml flat-bottomed bottle, covering the bottom of the bottle. When 0.5 ml of VDRL antigen is added drop-wise, with continuous rotation on a level surface, the antigen mixes with the saline and forms the emulsion. The VDRL test was performed according to standard techniques (5) by using the antigen made in either the flat-bottomed bottles or the Erlenmeyer flasks. All serum dilutions were made in test tubes as master dilutions to remove any variability due to differences in dilution of the serum samples.

When we tested the 52 reactive serum samples, antigens prepared in the flasks had the same endpoint titers as the antigens prepared in the flat-bottomed bottles approximately 90% of the time (range, 88 to 94%) (Table 1). When the endpoint titers differed, the titers obtained with the flask-prepared antigens were more apt to be lower (range, 4 to 10% of the samples) than higher (0 to 6% of the samples) than those obtained with antigen prepared in the traditional bottles. In either case, there was never more than a 1-dilution difference between endpoint titers obtained with antigen prepared by either method. This difference is within the allowable reproducible error (± 1 doubling dilution) for the test. No reactive serum was nonreactive with any of the antigen preparations.

With the nonreactive sera, any roughness in the 1:1 dilution in the samples was mostly associated with the serum sample rather than method of antigen preparation, except for two sera which had roughness with the antigen prepared in the flasks and not with the antigen prepared in the bottle. Of the 54 nonreactive serum samples, 6 had roughness in all six antigen preparations. None of the nonreactive sera were reactive with any of the six antigen preparations.

In reactions with the first group of sera (52 reactive, 54 nonreactive), all antigens were prepared with one lot of antigen. To determine if working dilutions of antigen prepared from different lots of antigen might be affected by the use of the convex-bottomed Erlenmeyer flasks, we tested four control sera with three additional lots of antigen and prepared each in two bottles and two flasks, respectively. Control sera were used because they had known reactivity and titers. No differences in reactions could be attributed to use of either a bottle or a flask for antigen preparation (Table2). Working antigen from one antigen (CDC 97-0036) was slightly more sensitive with one serum (Cenogenics 09020) than were working antigens prepared from lots of the other two antigens. Because the difference in reaction results was reactive versus weakly reactive at the 1:1 dilution, we do not consider this difference to be significant.

When the 25-ml Erlenmeyer flask is used, the convex bottom causes the 0.4-ml addition of buffered saline to be displaced to the circumference of the flask. When the VDRL antigen is added drop-wise, it hits the center of the convex portion of the flask. However, since the flask is continuously and gently agitated while the antigen is being added, the emulsion is formed just as effectively as when the flat-bottomed bottle is used. If the flask is not agitated adequately, pooling at the periphery of the flask occurs and may result in a rough antigen. This may make detection of prozone reactions difficult or may cause results to be interpreted as weakly reactive. A couple of alternatives were not addressed here. One is to double the volume of reagents, as was noted in the 1959 edition of Serologic Tests for Syphilis(1). This would increase the volume enough to help alleviate the effect of the convex bottom in the flask. The other is to use 50-ml Erlenmeyer flasks, which have a slightly flatter bottom than do the 25-ml Erlenmeyer flasks. Each laboratory should do its own evaluation to determine which alternative system works best for its situation.

This study suggests that 25-ml glass-stoppered Erlenmeyer flasks can be used as a substitute for 30-ml glass-stoppered, flat-bottomed, round bottles. Care needs to be taken to ensure that the glassware is clean and well rinsed and that the flask is rotated enough during preparation to form the emulsion. The diameter of the bottom of the flask is approximately the same as that of the traditional bottle, which allows for adequate mixing when the flask is rotated during addition of VDRL antigen to the 0.4 ml of buffered saline.

• Copyright © 1999 American Society for Microbiology