Annals of Internal Medicine. 1991;114:

Serologic Response to Treatment of Infectious Syphilis Barbara Romanowski, MD; Ruth Sutherland, DPH, RN; Gordon H. Fick, PhD; Debbie Mooney, BSc; and Edgar J. Love, MD, PhD Objective: To evaluate the serologic response to treatment of patients with infectious syphilis. Design: Historical cohort study of all cases of infectious syphilis in Alberta from 1981 to 1987. Patients: A total of 1090 patients were entered; 857 with primary syphilis, 183 with secondary syphilis, and 50 with early latent disease. Two hundred and eight patients were excluded who either were pregnant, had negative serologic results before treatment, had clinical relapse, were treatment failures, or were lost to follow-up. Interventions: All 882 evaluable patients were treated with a recommended antibiotic regimen for infectious syphilis and returned for re-assessment including repeat serologic testing. Measurements and Main Results: Seventy-two percent (95% CI, 66% to 77%) and 56% (CI, 43% to 70%) of patients with initial episodes of primary or secondary syphilis had seroreverted according to rapid plasma reagin (RPR) test results by 36 months. A 2- and 3-tube decline was seen by 6 and 12 months in primary and secondary syphilis. Early latent syphilis resulted in only a 2-tube decrease at 12 months. Serologic response was not affected by sex, age, race, or sexual orientation. Patients with their first infection were more likely to experience RPR seroreversal than those with repeat infections. The RPR reversal rates also depended on the pretreatment titer and stage of disease. At 36 months, 24% (CI, 20% to 28%) of patients had nonreactive fluorescent treponemal antibody absorption tests (FTA-Abs), and 13% (CI, 11% to 15%) had nonreactive microhemoglutination tests for Treponema pallidum (MHA-TP). Conclusions: Adequate therapeutic response for syphilis must be based on illness episode and the pretreatment RPR titer. Treponemal tests can demonstrate seroreversion after 36 months, and a negative treponemal test does not rule out a past history of syphilis. Annals of Internal Medicine. 1991;114:1005-1009. From STD Control, Alberta Health, Edmonton, Alberta; and the University of Calgary, Calgary, Alberta, Canada. For current author addresses, see end of text. Phy sicians diagnose infectious syphilis on the basis of clinical findings, examination of lesion material for Treponema pallidum, and serologic tests for syphilis. Penicillin and tetracycline regimens are highly effective for treatment of early syphilis (1, 2). No simple test is available that determines that syphilis has been cured within days or weeks after treatment. To assess treatment, the patient is asked to return for repeat serologic testing and clinical evaluation at varying intervals, usually at 3, 6, and 12 months. It has been reported that seroreversion is more rapid after therapy if the duration of infection or clinical lesions is short and if the initial nontreponemal titer is low. It has also been stated, however, that the rate of decline of high-titer reactivity is more rapid after treatment than the rate of decline of lower-titer seroreactivity (3). When the serologic test becomes negative, the patient is considered cured. However, when it remains positive, a clinical dilemma arises. In patients presenting with their first episode of early syphilis, titers can be expected to decline fourfold (2 dilutions) within 3 months of therapy and eightfold (3 dilutions) by 6 months (4). In patients with previous syphilis who become re-infected or in patients with late syphilis, the serologic response is often slower. We were able to examine the serologic response to treatment of patients with infectious syphilis because from 1983 to 1985 the province of Alberta experienced a sudden, sustained epidemic of infectious syphilis that primarily affected the heterosexual community. Patients and Methods In Alberta, standard guidelines for the diagnosis, treatment, and follow-up of sexually transmitted diseases are regulated and enforced by Sexually Transmitted Disease Control, Alberta Health. All patients with syphilis are diagnosed and staged on the basis of clinical and serologic findings. The serum specimens are processed by the Provincial Laboratory of Public Health in either Edmonton or Calgary. The rapid plasma reagin (RPR) test, the microhemoglutination test for Treponema pallidum (MHA-TP), and the fluorescent treponemal antibody absorption (FTA-Abs) test are done in accordance with standard procedures. Over 90% of patients with infectious syphilis receive benzathine penicillin (2.4 million units as a single dose), and the remainder receive either tetracycline or erythromycin (500 mg orally four times daily for at least 15 days). All patients are advised to return for follow-up serologic testing. Sexual partners, when identified and located, are treated prophylactically. Our study sample included patients with infectious syphilis (primary, secondary, or early latent syphilis) diagnosed and treated in the Province of Alberta from 1981 to 1987. Seventeen women who were treated for infectious syphilis either during pregnancy or postpartum were excluded because one would expect them to respond differently due to immunologic changes that occur during pregnancy. Primary syphilis was diagnosed on the basis of the presence of Treponema pallidum on darkfield microscopy from mucocu- 1991 American College of Physicians 1005

Table 1. Cumulative Rapid Plasma Reagin (RPR) Seroreversion after Treatment for Infectious Syphilis by Stage and Illness Episode Stage Illness Episode At 3 At 6 At 12 At 24 At 36 Months Months Months Months Months Primary Secondary Early latent Initial* Cumulative proportion ± SE 12.8 ± 1.4 29.6 ± 2.1 43.9 ± 2.5 60.0 ± 2.7 71.9 ± 2.7 Patients, «t 603 392 224 142 61 Repeat Cumulative proportion ± SE 0 10.5 ± 4.1 20.0 ± 5.4 22.1 ± 5.7 34.3 ± 7.4 Patients, n 66 60 43 34 25 Initial* Cumulative proportion ± SE 0.7 ± 0.7 6.1 ± 2.2 21.6 ± 4.2 41.9 ± 5.7 56.3 ± 6.7 Patients, n 161 123 74 45 18 Initial Cumulative proportion ± SE 3.0 ± 2.9 6.9 ± 4.7 12.7 ± 7.2 12.7 ± 7.2 26.1 ± 13.8 Patients, n 38 28 17 12 7 * Difference between stages for initial illness episode over 36 months using the Lee-Desu statistic; P < 0.001. t Number of patients with a reactive RPR titer entering the time interval. t Difference between stages for initial illness episode over 36 months using the Lee-Desu statistic; P = 0.006. taneous, genital, or rectal lesions or recent seroconversion of syphilis serologic tests together with clinical findings of a primary chancre. Secondary syphilis was diagnosed in patients with a reactive serologic test who had either typical macular or papular rashes on their palms and soles, mucous membrane lesions, or darkfield positive condylomata lata. Asymptomatic patients with reactive serologic findings and a history of sexual contact within the past year with a person who had primary or secondary syphilis were designated as having early latent disease. The patients with early latent disease may also have had a clinical history of untreated primary or secondary syphilis in the past year. To be included in this study, all patients must have had a reactive serologic test at or no more than 21 days before the start of treatment and one or more serologic tests after treatment. The relevant demographic, clinical, and serologic data were collected by a retrospective chart review of records held by Sexually Transmitted Disease Control, Alberta Health. All data were analyzed using the SPSS Statistical Analysis System on the mainframe computing facilities of the University of Calgary. Except for the cross-tabulations and frequency distributions, all analyses were based on life-table methods (5). In these analyses, the terminal event was either seroreversal or a specified decrease in quantitative RPR dilutions. Censored cases included those patients lost to follow-up, those who had clinical relapse with or without re-exposure, and those who failed treatment. Treatment failure was defined as a fourfold rise in the RPR titer after therapy without re-exposure. For descriptive purposes, the survival curves were compared using the Lee-Desu statistic as a test of significance. Confidence intervals (CIs) of 95% are provided where appropriate. Results From 1981 through 1987, the province of Alberta recorded 1090 cases of infectious syphilis. Of these, 857 patients had primary syphilis, 183 had secondary syphilis, and 50 had early latent disease. Most of the cases (834, or 76%) were recorded during the epidemic from 1983 to 1985. Homosexual and bisexual men accounted for 186 (17%) of the total cases from 1981 to 1987 and for 102 cases (12%) during the epidemic years. The human immunodeficiency virus (HIV) status of these men is unknown, as is the proportion of intravenous drug users. Change in RPR Titers This analysis was restricted to 882 cases of infectious syphilis (primary syphilis, 602; secondary syphilis, 161; early latent syphilis, 38). The remaining 208 patients were not analyzed on the basis of previously outlined exclusion criteria. At the end of 36 months, the cumulative proportion showing seroreversal was 63% ± 3%. The seroreversion rate was highest in patients experiencing their first illness episode with a diagnosis of primary syphilis: 72% at the end of 36 months (Table 1). The seroreversion rate was lower in those persons with a repeat episode of primary syphilis (34% at 36 months) and in those experiencing their initial illness episode with either secondary syphilis (56% at 36 months) or early latent syphilis (26% at 36 months). None of the patients experiencing a repeat illness with either secondary syphilis (13 patients) or early latent syphilis (2 patients) seroreverted. Among those patients having their first illness episode, the cumulative proportion of seroreversion varied directly with the initial RPR titer (Table 2). This effect was most marked in those with a diagnosis of primary syphilis. In patients with secondary syphilis, the proportion with seroreversion was comparable to those with an initial RPR titer of less than or equal to 1:8 or (60% and 68%, respectively), and very low (19%) in those with an initial titer of 1:256 or greater. Seroreversal in patients with early latent syphilis only occurred when the initial titer was 1:8 or less. Significant differences were found by sex and race in the cumulative proportion of seroreversion among those with their first illness episode with primary syphilis. Stratification by initial RPR titer, however, eliminated these differences (data not shown). Neither age nor sexual orientation influenced the cumulative proportion of seroreversal. There were too few patients in the other diagnostic groups to allow similar analysis. Because of the large number of patients with infectious syphilis who did not achieve seroreversal, we examined the decrease in RPR dilutions (Table 3) in patients with their initial illness episode and in patients with secondary or early latent syphilis only when the number of cases was adequate. Although these life-table analyses were carried out for the entire 36 months of follow-up, the diminishing sample size limits meaningful discussion beyond 24 months. For patients with primary syphilis (regardless of the 1006 15 June 1991 Annals of Internal Medicine Volume 114 Number 12

initial RPR titer) more than 75% had a 2-tube or greater (dilutions) reduction in RPR titer by 6 months (Table 3). By 24 months, 97% of patients had reached this end point. Three- and four-tube decreases occurred more slowly: Approximately 66% of patients achieved a 3-tube or greater decrease by 6 months but at 24 months the cumulative proportions were comparable. In patients with secondary syphilis, the decrease in RPR titers occurred somewhat earlier, by initial RPR titer, than in those with primary syphilis: Eighty percent of patients achieved a 3-tube or greater decrease by 6 months. The proportion of patients with early latent syphilis had low rates of titer decrease. Change in Confirmatory Treponemal Tests The confirmatory treponemal tests were less likely to serorevert (Table 4). All patients whose MHA-TP or FTA-Abs test results seroreverted were experiencing their first episode of illness. No patients with early latent syphilis seroreverted. All but three of the MHA-TP and all of the FTA-Abs seroreversions occurred in persons with primary syphilis. For those with their first illness episode of primary syphilis, only 8% of those with reactive MHA tests and 11% of those with reactive FTA tests had become negative at 12 months. After 3 years of follow-up, these figures were only 13% and 24% for the MHA-TP and FTA-Abs tests, respectively. The seroreversion rate was not influenced by sex, age, race, or by sexual preference among men. Discussion Clinicians depend on serologic evaluation to determine the efficacy of syphilis therapy. The patient may come for only one or two follow-up visits, however, and little information on temporal trends is available. This study provides data on seroreversion rates and titer decreases by stage of disease, initial RPR titer, and disease episode. We found that sex, age, race, and sexual orientation did not affect the serologic response. In our province, HIV infection has primarily affected homosexual and bisexual men. In 1985, the prevalence of HIV infection in this group was very low. However, we did find that illness episode was significant in predicting serologic response. Individuals with their first infection were more likely to serorevert than those with re-infections regardless of the stage of disease. Patients with first infections also seroreverted earlier. Similar findings have been described by Brown and colleagues (4). The RPR seroreversal rates depend on the pretreatment titer as well as on the stage of infection. Depending on the initial RPR titer, patients with primary syphilis had seroreversion rates of 6% to 70% at 12 months. Secondary syphilis seroreversion rates were much slower, reaching a high of 68% at 36 months. The higher the initial titer, the less likely seroreversion was to occur in primary and secondary syphilis. For persons with early latent disease, only those whose initial titer was 1:8 or less had significant seroreversal rates, and Table 2. Cumulative Rapid Plasma Reagin (RPR) Seroreversion after Treatment for Initial Episode of Infectious Syphilis by Stage and Pretreatment RPR Dilutions Stage Pretreatment RPR Dilution At 3 Months At 6 Months At 12 Months At 24 Months At 36 Months Primary* < 1:8 Cumulative proportion ± SE 26.3 ± 2.9 53.9 ± 3.6 70.4 ± 3.4 80.7 ± 3.1 86.9 ± 2.8 Patients, nt 251 138 56 35 18 Cumulative proportion ± SE 2.7 ± 1.0 13.4 ± 2.3 26.2 ± 3.2 47.4 ±4.1 64.1 ± 4.5 Patients, n 299 213 139 87 35 Cumulative proportion ± SE 0 2.7 ± 2.7 5.9 ± 4.1 19.7 ±8.1 31.2 ± 10.2 Patients, n 48 39 27 21 9 Secondary* < 1:8 Cumulative proportion ± SE 20.0 ± 17.9 20.0 ± 17.7 60.0 ± 21.9 60.0 ± 21.9 60.0 ± 21.9 Patients, n 5 4 2 1 0 Cumulative proportion ± SE 0 7.8 ±3.1 27.1 ± 5.5 52.7 ± 6.8 67.6 ± 7.3 Patients,/? 118 86 49 28 12 Cumulative proportion ± SE 0 0 0 7.7 ± 7.4 18.5 ± 12.1 Patients, n 34 31 22 15 6 Early latent < 1:8 Cumulative proportion ± SE 6.9 ± 6.7 16.2 ± 10.7 31.4 ± 16.3 31.4 ± 16.3 65.7 ± 25.6 Patients, n 16 12 6 3 2 Cumulative proportion ± SE 0 0 0 0 0 Patients, n 19 14 9 7 4 Cumulative proportion ± SE 0 0 0 0 0 Patients, n 3 2 2 2 1 * Difference between pretreatment RPR dilution over 36 months using the Lee-Desu statistic; P < 0.001. t Number of patients with a reactive RPR titer entering the time interval. 15 June 1991 Annals of Internal Medicine Volume 114 Number 12 1007

these rates were not significant until the 36-month follow-up visit. However, the sample size after 6 months is quite small. Our results are strikingly different from those previously reported (6-12). Some of these studies are based on VDRL results rather than on results from the RPR. The agreement ± 1 dilution between these two tests is reported to be 50%, with the RPR having equal or better sensitivity and specificity (13). Widely quoted studies by Fiumara (6, 7) on the response to syphilis therapy indicate that the rate of seroreversal for primary and secondary syphilis at 12 and 24 months is 100%. Ninety-five percent of patients with early latent disease have also been reported to have seroreverted. An important limitation of these studies was that they reported results on patients treated with therapeutic regimens that differed from those generally accepted. In addition, they defined as re-infected those patients who had a fourfold or greater increase in serologic titers and failed to consider that some of these responses may have been a manifestation of treatment failure. Fiumara has also proposed that any patient with a reactive RPR card test 1 or 2 years after treatment for primary or secondary syphilis should be re-treated (6, 7). Schroeter and colleagues (11) reported that 97% of cases of primary syphilis and 77% of secondary cases seroreverted by 2 years. This study included both initial and repeat infections. Petersen and coworkers (12) similarly reported extremely high seroreversal rates of 97%, 84%, and 73% at 12 months for primary, secondary, and early latent cases, respectively. None of these studies analyzed the serologic response by pretreatment RPR titer. The RPR titer data indicate that successful treatment for primary and secondary syphilis results in a 2- and 3-tube decrease by 6 and 12 months, respectively, for all patients regardless of the initial titer. Patients with early latent syphilis respond much slower with only a 2-tube decrease at 12 and 24 months. These results differ from those of Brown and colleagues (4) who described a 2-tube decline in VDRL titer at 3 months and a 3-tube decline at 6 months. Their study sample included patients both with primary and secondary syphilis who had initial VDRL titers between 1:4 and 1:128. They based their findings on patients who were cured after treatment, with cure defined as the patient being symptom-free after therapy. However, in the natural history of this disease individuals will become symptom-free regardless of therapy. Again, they used therapeutic regimens different from those that are generally accepted. The results of their study (4) concur with ours only in the rate of decline in titer with a more rapid serologic response occurring when the initial titer is high. Seroreversion of treponemal tests was only observed in patients with their first episode of illness. All but three of these seroreversions occurred in patients with primary syphilis. The FTA-Abs test was more likely to become negative then was the MHA-TP test, although the difference was not statistically significant at any time interval. Schroeter and associates (11) reported 6% seroreversion in 80 patients with primary or secondary syphilis at 2 years. In another group of 115 patients with primary syphilis, they showed that the rate of seroreversion depended on the reactivity of the VDRL. Patients with "seronegative VDRL" primary syphilis had a 67% seroreversion rate at 12 months compared with only 11% if the VDRL was initially positive. In another group of 93 patients with secondary syphilis, no FTA-Abs seroreversions were noted. The paper by Schroeter and coworkers reviews patients with initial infection or re-infection and does not differentiate between the two groups (11). The extremely high rate of seroreversion in patients with darkfield-positive, VDRLnegative primary syphilis is difficult to explain, but includes only 18 patients of whom 4 did not have a reactive FTA-Abs test before treatment. In our study, we only included patients with reactive serologic tests. Haas and colleagues (14) recently suggested that the Table 3. Rapid Plasma Reagin (RPR) Titer Decrease by Initial RPR for First Illness Episode by Stage and Time* Stage Initial Titer Time, mo Percentage of Titer Decrease by Number of Tubes (Dilutions) > 4 (Number)t > 3 (Number)t > 2 (Number)t Primary 1:8 6 46.3 ± 6.7 (43) 65.9 ± 6.2 (34) 76.8 ± 5.4 (28) 12 62.2 ± 6.7 (22) 78.5 ± 5.5 (13) 86.7 ± 4.4 (8) 24 76.9 ± 6.0 (14) 92.2 ± 3.7 (5) 96.2 ± 2.6 (3) 6 48.4 ± 3.2 (174) 69.7 ± 2.9 (143) 79.4 ± 2.5 (101) 12 75.9 ± 2.9 (66) 86.0 ± 2.3 (48) 90.5 ± 1.9(35) 24 90.6 ± 2.1 (31) 96.6 ± 1.3(17) 98.2 ± 0.9 (8) 6 49.1 ± 7.7(35) 71.0 ± 6.8(26) 88.2 ± 4.9 (15) 12 74.2 ± 7.2(14) 85.5 ± 5.4 (10) 90.3 ± 4.6 (6) 24 89.2 ± 4.6 (7) 92.7 ± 4.0 (4) 97.6 ± 2.3 (2) Secondary 6 61.8 ± 5.1 (74) 79.1 ± 4.1 (56) 87.8 ± 3.3 (35) 12 76.6 ± 4.7 (22) 88.3 ± 3.4 (14) 93.7 ± 2.6 (6) 24 87.1 ± 3.9(12) 95.3 ± 2.3 (8) 98.7 ± 1.2(4) 6 59.9 ± 8.7 (26) 81.6 ±6.8 (19) 85.3 ± 6.1 (14) 12 83.3 ± 6.8 (6) 93.9 ± 4.2 (3) 94.1 ±4.0(2) Early latent 6 18.0 ± 9.5 (14) 30.4 ± 11.4(13) 50.1 ± 11.9(12) 12 27.1 ± 12.0(9) 40.3 ±13.5 (7) 77.5 ± 10.8 (4) 24 37.5 ± 14.1 (7) 40.3 ± 13.5 (6) 85.0 ± 9.4 (3) * RPR titer decreases are expressed as cumulative proportions ± SE. t Number of patients entering time interval. 1008 15 June 1991 Annals of Internal Medicine Volume 114 Number 12

Table 4. Cumulative MHA-TP and FTA-Abs Seroreversion after Treatment for First Illness Episode of Primary Syphilis* Test 3 Months 6 Months 12 Months 24 Months 36 Months MHA-TP Cumulative proportion ± SE 3.5 ± 0.8 5.4 ± 1.0 8.3 ± 1.3 9.3 ± 1.4 13.0 ± 2.0 Patients, wt 616 495 361 261 126 FTA-Abs Cumulative proportion ± SE 3.5 ± 1.2 7.3 ± 1.8 10.6 ± 2.3 15.1 ± 2.9 23.8 ± 4.3 Patients, A?t 259 190 141 103 55 * MHA-TP = microhemoglutination test for Treponema pallidum; FTA-Abs = fluorescent treponemal antibody absorption test. t Number of patients with a reactive test entering the time interval. FTA-Abs and MHA-TP tests do not reliably identify previous syphilis in HIV-infected individuals as their immune dysfunction progresses and their CD4 counts fall. In a group of 104 HIV-infected men, 18 (17%) showed loss of reactivity of treponemal tests. Univariate predictors for loss of reactivity were symptomatic HIV infection, T4/T8 ratio of 0.6 or less, a single episode of syphilis, and an initial VDRL titer of 1:32 or less. The age of the individual or duration of infection was not significant. However, in a group of 19 HIVnegative men, all had persistent reactivity of treponemal tests up to 8.8 years after infection. The authors did not differentiate between the MHA-TP and FTA-Abs tests, but their results for HIV-negative men are strikingly different from our 3-year follow-up, with 13% and 24% seroreversion of the MHA-TP and FTA-Abs tests, respectively. We stress that treponemal tests will become negative in immunocompetent persons and that negative treponemal tests do not necessarily rule out a past history of syphilis. It should also be recognized that either of the treponemal tests can serorevert individually. Current dogma states that after adequate treatment for primary or secondary syphilis, there should be at least a 2-tube decline in titer by 3 or 4 months and a 4-tube decline by 6 or 8 months, with little or no reaction after the first year. We have shown that the decline in RPR dilutions depends on the illness episode and on the initial titer. For primary syphilis, a 2-tube decline can be expected at 6 months and a 3- and 4-tube decline at 12 and 24 months, respectively. After adequate treatment for secondary syphilis, 3- and 4-tube declines should be observed at 6 and 12 months, respectively. Patients with early latent syphilis should not be expected to have declines of 2 tubes until 12 months. Further, RPR seroreversion at 1 year can be expected in 50% of patients with an initial episode of only primary syphilis. Seroreversion of treponemal tests will be seen after 2 to 3 years of follow-up in up to 24% of patients. Acknowledgments: The authors thank Rose N. Cymbaluk for preparation of the manuscript. Grant Support: By a grant from Health and Welfare Canada. Requests for Reprints: Barbara Romanowski, MD, STD Control, 4th Floor Executive Building, 10105-109 Street, Edmonton, Alberta T5J 1M8, Canada. Current Author Addresses: Drs. Romanowski and Sutherland and Ms. Mooney: STD Control, 4th Floor Executive Building, 10105-109 Street, Edmonton, Alberta T5J 1M8, Canada. Drs. Fick and Love: Department of Community Health Services, Faculty of Medicine, University of Calgary, 3330 Hospital Drive N.W., Calgary, Alberta T2N 4N1, Canada. References 1. WHO Scientific Group on Treponemal Infections. Treponen.1 Infections: Report of a WHO Scientific Group, technical repo t series 674. Geneva: World Health Organization; 1982. 2. Willcox RR. Treatment of syphilis. Bull WHO, 1981;59:655-63. 3. Division of Venereal Disease: Evaluation of antisyphilitic therapy with intensive follow-up. J Vener Dis 1951;32:355-79. 4. Brown ST, Zaidi A, Larsen SA, Reynolds GH. Serological response to syphilis treatment. A new analysis of old data. JAMA. 1985;253: 1296-9. 5. Armitage P, Berry G. Statistical Methods in Medical Research, 2d ed. London: Blackwell Scientific Publications; 1987:424-8. 6. Fiumara NJ. Treatment of seropositive primary syphilis: an evaluation of 196 patients. Sex Transm Dis. 1977;4:92-5. 7. Fiumara NJ. Treatment of secondary syphilis: an evaluation of 204 patients. Sex Transm Dis. 1977;4:96-9. 8. Fiumara NJ. Treatment of early latent syphilis under 1 year's duration: serologic response to treatment of 368 patients. J Am Acad Dermatol. 1986;5:1059-61. 9. Fiumara NJ. Reinfection primary, secondary, and latent syphilis: the serologic response after treatment. Sex Transm Dis. 1980;7: 111-5. 10. Fiumara NJ. Treatment of primary and secondary syphilis. Serological response. JAMA. 1980;243:2500-2. 11. Schroeter AL, Lucas JB, Price EV, Falcone VH. Treatment for early syphilis and reactivity of serologic tests. JAMA. 1972;221:471-6. 12. Petersen CS, Jorgensen BB, Pedersen NS. Treatment of early infectious syphilis in Denmark. A retrospective serological study. Dan Med Bull. 1984;31:70-2. 13. Perryman MW, Larsen SA, Hambie EA, Pettit DE, Mullally RL, Whittington W. Evaluation of a new rapid plasma reagin card test as a screening test for syphilis. J Clin Microbiol. 1982;16:286-90. 14. Haas JS, Bolan G, Larsen SA, Clement MJ, Bacchetti P, Moss AR. Sensitivity of treponemal tests for detecting prior treated syphilis during human immunodeficiency virus infection. J Infect Dis. 1990; 162:862-6. 15 June 1991 Annals of Internal Medicine Volume 114 Number 12 1009