Journal of Neuro-Ophthalmology, December 1990, Volume 10, Issue 4

The Bascom Palmer Eye Institute Lyme/syphilis survey.

Serologic screening of patients for Lyme borreliosis began at the Bascom Palmer Eye Institute (BPEI) in September 1987. This report reviews the data on 641 sera from that date up to January 1, 1990. Initially only immunofluorescent (IFA) IgG and IgM titers were obtained. Because of increasing numbers of borderline and positive IFA tests, a Lyme enzyme linked immunosorbent assay (ELISA) was added in April 1988. Also, because of significant serologic cross reactivity in patients exposed to Treponema pallidum, rapid plasma reagin (RPR) and fluorescent treponemal antibody absorption (FTA-ABS) tests were added to the serologic screening panel. Of all sera tested, 10% showed reactive RPR tests and 22% showed reactive FTA-ABS tests. Lyme IFA IgG titers were greater than or equal to 1:64 in 17% of the sera, and Lyme ELISA tests were greater than 1.25 in 15% of the sera. Our experience agrees with reports that serum RPR or VDRL tests are nonreactive in Lyme borreliosis, and that false positive FTA-ABS tests can occur in Lyme borreliosis. The importance of getting all four tests--RPR, FTA-ABS, Lyme IFA IgG and IgM, and Lyme ELISA--in all patients suspected of spirochetal disease is emphasized.

Journal of Clinical Neuro- ophthalmology 10( 4): 255- 260, 1990. The Bascom Palmer Eye Institute Lyme/ Syphilis Survey J. Lawton Smith, M. D., Bret C. Crumpton, and Jan Hummer, a. D. © 1990 Raven Press, Ltd., New York Serologic screening of patients for Lyme borreliosis be­gan at the Bascom Palmer Eye Institute ( BPEI) in Sep­tember 1987. This report reviews the data on 641 sera from that date up to January 1, 1990. Initially only im­munofluorescent ( IFA) IgG and IgM titers were ob­tained. Because of increasing numbers of borderline and positive IFA tests, a Lyme enzyme linked immunosor­bent assay ( ELISA) was added in April 1988. Also, be­cause of significant serologic cross reactivity in patients exposed to Treponema pallidum, rapid plasma reagin ( RPR) and fluorescent treponemal antibody absorption ( FTA- ABS) tests were added to the serologic screening panel. Of all sera tested, 10% showed reactive RPR tests and 22% showed reactive FTA- ABS tests. Lyme IFA IgG titers were ~ 1: 64 in 17% of the sera, and Lyme ELISA tests were > 1.25 in 15% of the sera. Our experience agrees with reports that serum RPR or VDRL tests are nonreactive in Lyme borreliosis, and that false positive FTA- ABS tests can occur in Lyme borreliosis. The im­portance of getting all four test&-- RPR, FTA- ABS, Lyme IFA IgG and IgM, and Lyme ELISA- in all patients sus­pected of spirochetal disease is emphasized. Key Words: Lyme disease- Lyme ELISA- False positive FTA- ABS tests. From the Bascom Palmer Eye Institute, Department of Oph­thalmology, University of Miami School of Medicine, Miami, Florida. Address correspondence and reprint requests to Dr. J. Law­ton Smith, Bascom Palmer Eye Institute, P. O. Box 016880, Mi­ami, FL 33101, U. S. A. 255 Serologic screening of patients at the Bascom Palmer Eye Institute ( BPEI) for Lyme borreliosis began in September 1987, Initially the test per­formed was the immunofluorescent assay ( IFA) for IgG and IgM antibodies. With experience it be­came evident that a significant serologic cross re­activity was occurring in patients with present or past exposure to Treponema pallidum, Accord­ingly, it became routine to also order both RPR ( rapid plasma reagin) and FTA- ABS ( fluorescent treponemal antibody absorption) tests in all cases, Because of increasing numbers of borderline and positive Lyme IFA tests, a Lyme ELISA ( enzyme linked immunosorbent assay) was later added to the survey. A pilot study was done during a I- week interval ( late March- early April 1988) on a group of 75 pa­tients seen at the BPEI. This group consisted of 35 preoperative cataract patients, 15 outpatient neuro- ophthalmic return patients, and 25 consec­utive consenting patients seen in the triage/ eye emergency clinic. This group was considered rep­resentative of the total population being encoun­tered in the University of Miami ophthalmology department, The data from that study was re­ported elsewhere in detail ( 1). As a result of the pilot study, through coopera­tion with the Microbiology Reference Laboratory ( MRL) ( Cypress, CA, U. s, A.) a combined serologic protocol was established, Clinicians ordered a " Lyme/ treponemal panel" and the following tests were then performed on all cases: RPR with titer, FTA- ABS, Lyme IFA IgG and IgM, and Lyme ELISA, The techniques and methodology for these tests are similar to those reported elsewhere ( 2), MRL repeated all borderline or conflicting data be­fore issuing a final report, and kept all submitted sera frozen for 3 months to enable repeat testing if deemed clinically necessary. In a few select cases, 256 J. L. SMITH ET AL. TABLE 2. Positive Lyme IgM test titers TABLE 1. Positive Lym,~ :;,~ test :,~ ers No. patients Lyme ELISA Lyme ELISA tests were obtained on 417 sera, or 65% of the series. It should be noted that the ELISA was added routinely to the serologic panel only after the April 1988 pilot study, therefore yielding the smaller number of ELISA tests avail­able in this review. MRL initially reported an IgM Lyme IFA IgM titers were reported as positive ( i. e., ~ 1: 16) in only 24 cases, or 3.7% of the series. The titers are shown in Table 2. The IgM titer was negative ( i. e., < 1: 16) in 616 sera ( 96.1 %) of the series. Interestingly, only four specimens were re­ported as showing both IgG and IgM titers simul­taneously positive in any IFA specimen. The data on these patients are seen in Table 3. Two points are evident from Table 3. Three of the four sera had IgM titers of 1: 16, which is a low titer at the exact cut- off point below which the IgM titer would be called negative. This may indicate the need for a borderline or equivocal zone in the def­inition of IFA IgM seropositivity as exists with IgG. The other point is that the data strongly suggest that three of these four patients have had syphilis, thus probably representing serologic cross reactiv­ity in these instances. There were 10 equivocal IFA IgG titers with pos­itive IgM titers ( 1.6%), and 107 equivocal IFA IgG titers with negative IgM titers. The IFA tests ( IgG and IgM) were both negative in 406 sera, or 63.3% of the series. 10 1 422 2 1 2 24 45 18 24 5 16 108 No. patients 1: 16 1: 20 1: 32 1: 40 1: 64 1: 80 1: 128 1: 160 Total 1: 64 1: 128 1: 256 1: 512 1: 1024 Total DATA The following reviews all the Lyme/ treponemal serologic panel data done at the BPEI from Sep­tember 16, 1987 to January 1, 1990. During this 27- month interval, 641 sera were evaluated on 570 patients. Sex was determined from names on the requisition as being 269 males and 258 females. This showed an equal sex distribution in the study. One serum specimen was drawn on 524 patients and multiple or serial data were obtained as fol­lows: two tests = 30 patients; three tests = 10 patients; four tests = 4 patients; and five and six tests done on one patient each. IgG sera were also sent to other laboratories to confirm borderline or variable results. This study to date has made it increasingly evi­dent that the number of cases with ocular and neu­rologic involvement due to Borrelia burgdorferi is significant even in south Florida, which might oth­erwise be considered a nonendemic area ( 3,4). Fur­thermore, clinical evidence of seronegative Lyme borreliosis ( 5,6) has become a problem of such magnitude that other tests have necessarily been added to the routine serum panel when IFA and ELISA tests were both nonreactive. These have in­cluded serum Western blot and polymerase chain reaction ( 7) tests, as well as a lymphocyte antigen stimulation test ( 5). More recently, a Lyme urinary antigen test is also being introduced into the in­vestigative protocol in selected cases. These newer studies are undergoing careful evaluation to deter­mine usefulness in diagnosing Lyme borreliosis. Because only a small number of patients in this study had these newer tests, their data are not reviewed in this report. Lyme Immunofluorescent Assay Lyme IFA tests were obtained on 641 sera in this study ( 100%). MRL considers the Lyme IFA IgG negative if < 1: 16, equivocal if 1: 16- 1: 32, and pos­itive if ~ 1: 64. The Lyme IgG was reported as pos­itive ( i. e., ~ 1: 64) in 108 instances or 16.8% of the sera. The titers of the positive IgG tests are shown in Table 1. An equivocal Lyme IgG ( 1: 16- 1: 32) was noted in 117 sera, or 18.3% of the series. This was 1: 16 in 69 instances and 1: 32 in 48 others. The Lyme IgG was negative ( i. e., < 1: 16) in 416 sera, or 64.8% of the series. '~" I BPEI LYME/ SYPHILIS SURVEY 257 TABLE 3. Data on patients with both IFA IgG and IgM positive IgG titer IgM titer Patient ELISA FTA RPR 1: 1024 1: 64 TB 3.01 R3+ R 1: 64 1: 1024 1: 16 PJP 2.20 R 3+ R 1: 32 1: 256 1: 16 CB 1.24 R3+ R 1: 1 1: 128 1: 16 AS 1.51 NR NR ELISA, enzyme linked immunosorbent assay: FTA, fluores-cent treponemal antibody; RPR, rapid plasma reagin. series. The titers of these reactive serum reagin tests are seen in Table 4. The serum RPR was nonreactive in 416 instances or 90% of the series. Another reagin test, the Ve­nereal Disease Research Laboratory test ( VORL), was obtained in only nine instances. Four were nonreactive, and five were reactive ( two at 1: 1, two at 1: 4, one at 1: 128). TABLE 4. Reactive rapid plasma reagin tests Syphilis Serology- Reagin Tests RPR tests were obtained on 462 sera or 72% of the series. These were reactive in 46 or 9.9% of the ELISA < 1.0 as negative, and ~ 1.0 as positive. Us­ing this criterion, of the 417 ELISA tests, 104 were reported as positive (~ 1.0), or 25% of the series. Reported as negative were 313 serum ELISA tests « 1.0), or 75% of the series. However, MRL now reports ELISA < 0.85 as negative, 0.85- 1.25 as equivocal, and > 1.25 as pos­itive. A review of our data in this regard revealed that the ELISA was 0.85-< 0.9 in 9 sera, 0.9-< 1.00 in 29 sera, or 2.2% of the series. The ELISA was 1.0 in 10 cases ( 2.4%); > 1.0-< 1.2 in 20 cases ( 4.8%), and 1.2- 1.25 in 13 cases ( 3.1 %). Thus, a total of 81 sera ( 19.4%) met this definition ( serum Lyme ELISA 0.85- 1.25) of being an equivocal or border­line reaction. There were 275 sera showing ELISA titers < 0.85, and thus 65.9% were negative. There were also 61 sera showing ELISA titers> 1.25, and thus 14.6% were positive. The higher titered Lyme ELISA tests were of great interest. Lyme ELISA titers were > 1.25-< 1.5 in 20, 1.5-< 2.0 in 26, and 2.0- 3.0 in 11 sera. Only four sera had an ELISA titer > 3.0: 3.01, 3.05, 4.17, and 5.07 in those in­stances. It is of interest that three of the latter four specimens were from patients with definite sero­lOgic evidence of syphilis. Titer 1: 1 1: 2 1: 4 1: 8 1: 16 1: 32 1: 64 1: 128 1: 256 1: 512 1: 1024 Total No. patients 11 11 4 2 1 5 3 5 2 1 1 46 Fluorescent Treponemal Antibody Absorbed Tests FTA- ABS tests were done on 467 sera, or 72.8% of the series. This test was reactive in 103 sera or 22% ( 103/ 467) of those tested. No equivocal FTA­ABS reports were given, and the test was reported as nonreactive or negative in 364 sera ( 364/ 467) or 77.9% of those tested. At this juncture, it should be noted that the yield of seroreactivity for syphilis was over twice as high using the treponemal ( FTA­ABS) test: 22% compared to the nonspecific reagin ( RPR) test's 9.9%. It is preferable in our experience to use the FTA- ABS test to make the serodiagnosis of syphilis ( particularly in neuro- ophthalmologic patients), and to follow the course of therapy in early infectious syphilis with the reagin test titer. In many hospitals, if one orders an RPR or VORL test, the result is reported as either nonreactive or reactive, and one must specifically request a titer. The latter is quite important in following the course of the disease and in the effectiveness of therapy. ANALYSIS OF DATA Immunofluorescent Assay Versus ELISA There have been differences of opinion as to whether the IFA or the ELISA test is preferred in the serologic diagnosis of Lyme borreliosis. It should be recalled that both a positive IgG and a positive IgM occurred in only four instances i~ ~ he total series. Our data showed, however, a posItIve serum IFA IgG with a negative ELISA test in 22 sera. Conversely, a positive serum ELISA test with a negative serum IgG was found in only eight sera. Thus, if only a single test for Lyme borreliosis had been obtained ( i. e., either the IFA or the ELISA), the serologic diagnosis would have been missed in 22 cases if only the ELISA were obtained. But the diagnosis would have been missed in 8 cases if only the IFA had been obtained. The difference may appear of a small order of magnitude in total numbers, but has proven to be of definite clinical significance. The following brief illustrative cases document this point. I eli" Neuro~ ophthalmol, Vol. 10. No. 4, 1990 258 J. L. SMITH ET AL. Case one ( LR # 031- 83- 610) An ll- year- old girl was well until May 1988 when she developed pain and redness in her left eye. Shortly thereafter the right eye became simi­larly involved. The patient saw Dr. W. W. Culbert­son in this department on September 21, 1988. He found an interesting uveitis syndrome consisting of bilateral vitritis, bilateral optic disc edema, and associated anterior segment inflammation. He im­mediately considered the possibility of ocular Lyme borreliosis. He elicited the history that al­though she did not recall an antecedent erythema migrans lesion, she had lived in Wisconsin in an endemic area until 18 months earlier. A Lyme IFA on September 29, 1988, was reported as negative ( IgG < 1: 16 and IgM < 1: 16). The patient continued to have active iritis and chronic pars planitis, not controlled with topical medications. A serum panel on February 22, 1989, revealed: RPR = NR, FTA- ABS = NR, Lyme IFA IgG < 1: 16 and IgM < 1: 16, but Lyme ELISA 2.1. The latter was re­peated twice and was consistently positive. She was treated with oral erythromycin 400 mgm qid for 3 weeks in May 1989 and showed no significant improvement. The girl was then given a course of intravenous Rocephin, 1 g daily for 28 days in July 1989. When Dr. Culbertson saw her again on August 30, 1989, her eyes showed dramatic improvement in her chronic iridocyclitis. She did well for four months after the intravenous Rocephin, but then had re­currence of anterior uveitis treated with oral dox­ycycline. Comment An 11- year- old girl from Wisconsin presented with a chronic ocular inflammatory triad consisting of iridocyclitis, plus a chronic pars planitis ( 8,9)- like picture with snowball deposits in the vitreous, and mild optic disc edema. This pic­ture suggested Lyme borreliosis to Dr. W. W. Cul­bertson. Two Lyme IFA titers were absolutely neg­ative. However, the Lyme ELISA test was strongly positive at 2.1. This case illustrates the importance of getting both IFA and ELISA tests because the diagnosis could not be serologically confirmed with IFA, but was definitely positive with ELISA testing. Case two ( SG # 349703) A 24- year- old woman had several tick bites while camping in North CarL, IIllCl in June 1987. A prolonged illness ensued characterized by palpita­tions, chroni'= fJtJc~~ q ', c" nrent polyarthralgias, ill paresthesias in the feet, apr] . -, tes. Short courses of oral antibiotics gal';:: 11( 1 fcaet. A complete neuro- ophthalmologic examination on December 18, 1989, was normal. A serum Lyme/ treponemal panel drawn at that time re­vealed: RPR = NR, FTA- ABS = NR, Lyme ELISA 0.53, but IFA IgG 1: 256 and IgM < 1: 16. A serum Western blot showed a strong 41 kd band, a weak 66 kd band, and a few other minor bands. This Western blot pattern has been seen with early or late Lyme disease, antibiotic treated disease, or se­rologic cross reactivity. A course of intravenous Rocephin therapy was advised. Comment The 24- year- old woman gave a history consistent with chronic Lyme borreliosis. Al­though her Lyme ELISA test was absolutely nor­mal, a Western blot test was interpreted as " suspicious" and the IFA IgG titer of 1: 256 con­firmed the clinical diagnosis. In this instance, in contrast to the previous case, the serologic diagno­sis would have been missed if only an ELISA test had been obtained. The definitely positive IFA IgG was necessary for serodiagnosis. We therefore advise that both the IFA IgG and IgM and ELISA tests be obtained in all cases being evaluated for Lyme borreliosis. Furthermore, be­cause of serologic cross reactivity with syphilis, the importance of obtaining all four tests ( RPR, FTA­ABS, Lyme IFA IgG and IgM, and Lyme ELISA) on all patients being investigated for Lyme disease must be stressed. Fluorescent Treponemal Antibody Absorption Findings Correlated with Lyme Data False positive serum FTA- ABS tests have been reported in patients with known Lyme disease ( 10). Although this was apparently a low titered ( 1: 5) response, since the FTA- ABS is not usually reported in numeric titers in the United States ( but only nonreactive, and reactive 1-- 4 +), this could cause a problem in clinical practice. Because of a known previously reported high incidence of reac­tive FTA- ABS tests in the BPEI clinic population ( 11), all cases in which IFA, ELISA, FTA- ABS, and RPR were obtained were reviewed. The pertinent data are found in Table 5. DISCUSSION A review of 641 serological tests for Lyme bor­reliosis performed at the BPEI in the 27- month pe­riod from September 1987 to January 1, 1990, has been performed. Since April 1988, because of dif­ferences between IFA and ELISA reports, and be- BPEI LYME/ SYPHILIS SURVEY TABLE 5. Review of 408 patients who had Lyme IgG, ELISA, FTA- ABS, and RPR tests 259 IgG ELISA FTA- ABS RPR No. Clinically suspect cases Pas Pas Pas Pas Active syphilis 18 Pas Pas Pas Neg Late/ treated syphilis with false + Lyme tests; or Lyme disease with 10 false + FTA- ABS Pas Equiv Pas Pas Syphilis with false + Lyme IgG 3 Pas Equiv Pas Neg Late/ treated syphilis with false + IgG; or Lyme with false- ELISA 2 Pas Neg Pas Pas Syphilis 2 Pas Neg Pas Neg Late syphilis or Lyme with false+ FTA- ABS 4 Pas Pas Neg Pas Abnormal serology 0 Pas Pas Neg Neg " True" Lyme 14 Pas Equiv Neg Pas Abnormal serology 0 Pas Equiv Neg Neg Lyme ( IFA better than ELISA) 13 Pas Neg Neg Pas Abnormal serology 0 Pas Neg Neg Neg Normal, nonspecific IgG, or possible Lyme 16 Equiv Pas Pas Pas Syphilis 1 Equiv Pas Pas Neg Late/ treated syphilis with false + Lyme serologies; or Lyme disease with 0 false + FTA- ABS Equiv Equiv Pas Pas Syphilis 4 Equiv Equiv Pas Neg Late syphilis or Lyme disease with false + FTA- ABS 7 Equiv Neg Pas Pos Syphilis 2 Equiv Neg Pas Neg Late/ treated syphilis; or Lyme with equiv IgG and false + FTA- ABS 6 Equiv Pas Neg Pas Abnormal serologic pattern; BFP 0 Equiv Pas Neg Neg ? " True" Lyme disease ( ELISA better than IFA) 9 Equiv Equiv Neg Pas Abnormal serologic pattern; BFP 0 Equiv Equiv Neg Neg Probably normal; borderline Lyme 21 Equiv Neg Neg Pos Abnormal serologic pattern; BFP 0 Equiv Neg Neg Neg Probably normal 43 Neg Pas Pas Pas Syphilis with false + ELISA 2 Neg Pos Pas Neg Late syphilis or Lyme with false+ FTA- ABS 0 Neg Equiv Pas Pas Syphilis 1 Neg Equiv Pas Neg Late syphilis 6 Neg Neg Pas Pas " Pure" syphilis 2 Neg Neg Pas Neg Late syphilis 15 Neg Pas Neg Pos Abnormal serology 0 Neg Pas Neg Neg Normal, nonspecific ELISA, or possible Lyme 6 Neg Equiv Neg Pas Abnormal serology 1 Neg Equiv Neg Neg Probably normal 31 Neg Neg Neg Pas Biologic false + 2 Neg Neg Neg Neg Normals 167 FTA- ABS, fluorescent treponemal antibody absorption; RPR, rapid plasma reagin; IFA, immunofluorescent assay; ELISA, emzyme linked immunosorbent assay; BFP, biologic false positive. Lyme IFA IgG positive is ;;. 1 : 64; IgG equivocal is 1: 16 or 1: 32; IgG negative is < 1 : 16. ELISA positive is > 1.25; equivocal is 0.85- 1.25; negative is < 0.85. cause of frequent serologic cross reactivity with syphilis, it became routine to obtain a battery of four tests ( RPR, FTA- ABS, Lyme IFA IgG and IgM, and Lyme ELISA) on each specimen. Table 5 pre­sents the possible serologic diagnostic interpreta­tions on the 408 cases that had the complete Lyme/ treponemal panel. Certain conclusions can be made to date from this study: Ten percent of all sera tested showed a reactive serum RPR test. Twenty- two percent of all sera tested showed a reactive serum FTA- ABS test. These findings again emphasize the significantly greater yield in serodiagnosis of the treponema­toses by using the treponemal ( FTA- ABS) test rather than the nonspecific ( RPR or VDRL) tests. There are two important considerations, how­ever. It has been reported that reagin ( RPR or VDRL) tests are nonreactive in Lyme borreliosis ( 12) and our experience to date agrees with this tenet. It has also been reported that false positive FTA- ABS tests can occur in Lyme borreliosis ( 10), and we have also confirmed this. The following case is an example. Case 3 ( FB # 320634) A 71- year- old woman visited Martha's Vine­yard, Massachusetts, in September 1988 and spent some time in the woods. Within 3 weeks, she de­veloped an erythematous skin lesion on her left forearm and an internist suspected she had a spi­der bite. She then developed fever of 102° and a flu- like illness followed by the onset of oblique ver­tical diplopia, numbness in the right side of her face, and a right peripheral VII nerve palsy. Neuro- ophthalmologic examination revealed a re- J Clin Neuro- ophthalmol, Vol. 10, No. 4, 1990 260 J. I. SMITH ET AI. solving erythema migrans on the left forearm and involvements of cranial nerves IV, V, and VII on the right. The clinical diagnosis was Lyme borreli­osis. Serologic testing on October 21, 1988, re­vealed: RPR = NR, FTA- ABS = R, Lyme ELISA 2.3, Lyme IFA IgG 1: 256, and IgM < 1: 16. The pa­tient was treated with oral doxycycline 100 mg three times a day for 2 weeks. Her diplopia cleared in 2 to 3 days, and the facial palsy cleared totally within a week. A repeat serum panel on Novem­ber 1, 1988, revealed: RPR = nonreactive, FTA­ABS = reactive 3 +, Lyme ELISA 2.1, and Lyme IFA IgG 1: 512 and IgM < 1: 16. Because of the re­peatedly reactive FTA- ABS tests in this patient, se­rum was sent to the Institut Alfred Fournier in Paris, the French National Center of Reference for the Treponematoses. Through the courtesy of Mme. A. Paris- Hamelin, paired serum testing on this patient revealed: VDRL, syphilis IFA IgM, TPHA, FTA- ABS, and TPI tests = negative. In France, the Lyme IFA IgG was positive at 1: 128 and IgM negative, and a Lyme Western blot was reported as positive. It was concluded on the basis of the history, clinical findings, serologic testing, and response to therapy that this patient had Lyme borreliosis, had a false negative FTA- ABS in one laboratory on two occasions as a serologic cross reaction, and had not had syphilis. If we consider the data in Table 5 and adhere to the belief that reactive reagin tests do not occur with Lyme borreliosis, it is seen that 50 cases of the 408 ( 12.2%) who had the complete panel are sero­logically suspicious of having had either active or late/ treated syphilis. Likewise, if we consider pa­tients who are positive for Lyme IFA and/ or ELISA antibodies, but who are nonreactive to both FTA­ABS and RPR tests, we have 42 cases ( 10.3%) who are serologically suspicious for having either active or late Lyme borreliosis. It is evident, however, because of the definite occurrence of seronegative cases of both syphilis and Lyme disease, that a final statistical differentiation of these spirocheto­ses cannot be made solely on serologic parameters. In order to further define these cases, a clinical chart review, more specific historical question­naire, repeat serologic testing, refined studies ( e. g., Western blot, PCR, lymphocyte stimulation test, urine antigen studies), and, when indicated, further examination of these patients may be nec-essary. It is hoped that ir ,(~~ r2r:'~'.( we ca. n correlate these clinical finding~ with the serolOgiC data reported here. It is evident, however, from both clinical and serologic experience to date that both Treponema pallidum and Borrelia burgdorferi organisms are significant causes for neuro- oph­thalmologic disease in a south Florida eye hospital population. The importance of obtaining a sero­logic battery of at least four tests ( RPR, FTA- ABS, Lyme IFA IgG and IgM, and Lyme ELISA) in all patients suspected of having either ocular syphilis or ocular Lyme borreliosis is emphasized. Acknowledgment: We thank Drs. Alan MacDonald, Kirk Winward, W. W. Culbertson, Richard Porschen, Mme. A. Paris- Hamelin, and the laboratory staff of Bas­com Palmer Eye Institute for all of their help with this investigation. REFERENCES 1. Smith JL, Parsons TM, Paris- Hamelin AJ, et al. The preva­lence of Lyme disease in a nonendemic area. A comparative serologic study in a south Florida eye clinic population. ! Clin Neuro- ophthalmol 1989; 9: 148- 55. 2. Barbour AG. Laboratory aspects of Lyme borreliosis. Clin Microbial Rev 1988; 1: 399- 414. 3. Winward KE, Smith JL. Ocular disease in Caribbean pa­tients with serologic evidence of Lyme borreliosis. ! Clin Neuro- ophthalmol 1989; 9: 65- 70. 4. Winward KE, Smith JL, Culbertson WW, et al. Ocular Lyme borreliosis. Am! Ophthalmol 1989; 108: 651- 7. 5. Dattwyler RJ, Volkman OJ, Luft BJ, et al. Seronegative Lyme disease. Dissociation of specific T- and B- lymphocyte responses to Borrelia burgdorferi. N Engl ! Med 1988; 319: 1441-<>. 6. Schutzer SE, Coyle PK, Belman AL, et al. Sequestration of antibody to Borrelia burgdorferi in immune complexes in seronegative Lyme disease. Lancet 1990; 335: 312- 5. 7. Eisenstein BI. The polymerase chain reaction. A new method of using molecular genetics for medical diagnosis. N Engl! Med 1990; 322: 178- 83. 8. Gass JDMG. Retinitis and vitritis with vitreous base exuda­tion and organization ( pars planitis, peripheral uveitis, or chronic cyclitis). In: Stereoscopic atlas of macular diseases. Di­agnosis and treatment. 3rd ed. St. Louis: CV Mosby, 1987: 524-<>. 9. Pederson JE, Kenyon KR, Green WR, et al. Pathology of pars planitis. Am! Ophthalmol 1978; 86: 762- 4. 10. Hunter EF, Russell H, Farshy CE, et al. Evaluation of sera from patients with Lyme disease in the fluorescent trepo­nemal anhbody- absorption test for syphilis. Sex Transm Dis 1986; 13: 232. 11. Harner RE, Smith JL, Israel CWo The FTA- ABS test in late syphilis. lAMA 1967; 78: 284- 8. 12. Magnarelli LA. Laboratory diagnosis of Lyme disease. In: Rheumatic disease clinics of North America: Lyme disease; vol 15. Philadelphia: WB Saunders, 1989: 735- 46. I Olfl ," t. ,. jQ( J