Response of latent syphilis or neurosyphilis to ceftriaxone therapy in persons infected with human immunodeficiency virus

CLINICALSTUDIES

Response of Latent Syphilis or Neurosyphilis to Ceftriaxone Therapy in Persons Infected With Human Immunodeficiency Virus MARK F'. DOWELL, M.D., PAUL G. ROSS, M.D., DANIEL ROBERT F'. BAUGHN, Ph.D., Houston, Texas

M. MUSHER,

OBJECTIVF~To evaluate the effect of ceftriaxone in treating latent syphilis or asymptomatic neurosyphilis in patients infected with the human immunodeficiency virus (HIV). DESIGN: Follow-up study'of patients treated at two HiV-based clinics during 16 months from 1989 to 1991. PATIEN2~ Patients were those in whom a cllnical diagnosis of latent syphiliA or asymptomatic neurcsyphills was made, who received all recommended doses of antimicrobial therapy, and who returned for follow-up visits for 6 or more months. RESULTS:Forty-three patients were treated with ceftrla~one, 1 to 2 g d~ily for 10 to 14 days. Thirteen underwent lumbar puncture before treatment; 7 (58%) had documented neurosyphills (pleecytosls in 5, elevated protein levels in 6, VDRL reactive in cerebrespinR! fluid [CSF] in 7), and 6 had documented latent syphilis (entirely normal CSF). The remaining 30 were said to have presumed latent syphilis+ There was no relation between the diagnosis and the selected dosage of ceftriaxone. Response rates were similar in these who had documented neuresyphills and documented or presumed latent syphilis. Overall, 28 patients (65%) responded to therapy, 5 (12%) were serofast, 9 (21%) had a serologic relapse, and 1 (2%) who experienced progression to symptomatic neurosyphills was a therapoutie failure. Thirteen patients received benzathlne penicillin for presumed latent syphilis; results were similar to those observed after ceftriRTone therapy, with 8 (62%) responders, 1 (8%) serofast, 2 (15%) relapses, and 2 (15%) failFromthe InfectiousDiseaseSectionand the SyphilisResearchLaboratory, VeteransAffairs MedicalCenter (MED, DMM, REB),The Thomas StreetClinicof the HarrisCountyMedicalDistrict(PGR,TRC),and Baylor Collegeof Medicine(MED, DMM, REB,TRC), Houston,Texas. Requestsfor reprintsshouldbe addressedto DanielM. Musher, M.D., VeteransAffairs MedicalCenter, InfectiousDiseaseSection, Room4B370, Houston,Texas77030. Manuscript submitted February 25, 1992, and accepted in revised form March 30, 1992.

M.D., THOMAS R. CATE, M.D.,

ure~ CD4 cell counts in responders were not different from those who failed to respond. CONCLUSIONS:Even in the absence of neurologic symptoms, hRlf of the HIV-infected persons who have serologic evidence of syphiliA may have neuresyphiliA. Although ceftriaxone achieves high serum and CSF levels, 10 to 14 days of treatment with this drug were associated with a 23% failure rate in HIV-infeeted patients who had latent syphilis or asymptomatic neuresyphiliA. Three doses of benT~thine ponicillln did not have a significantly higher relapse rate and may provide appropriate therapy, at least for documented latent syphiliR in persons co-infected with HIV. Studies comparing ceftriaxone with 10 to 14 doses of procJine penicillin are needed to determine the most cost-effective treatment for asymptoma+tic neuresyphilis or presumed latent syphilis in this group of patients.

exually transmitted diseases (STDs) have long been known to occur together. With the rising Sincidence of both syphilis [1] and acquired immunodeficiency sydrome (AIDS), concurrent infection with Treponema paUidum and the human immunodeficiency virus (HIV) presents a growing public health problem [2-5].Infection with HIV altersthe natural history of syphilisand the response to therapy; specifically,the incidence of neurosyphilis is vastly increased in HIV-infected persons, many of w h o m have previously been treated with recommended doses of benzathine penicillin for early syphilis [6]. Invasion of the central nervous system (CNS) by T. paUidum occurs early in syphilitic infection [7-10]. Although treatment with benzathine penicillin does not yield measurable CSF levels of penicillin [11,12], clinical failure, specifically the appearance of neurosyphilis after benzathine penicillin therapy, was exceedingly rare in the pre-AIDS era [13,14]. In contrast, many cases of early neurosyphilis have been documented after treatment of HIV-infected patients with recommended doses of

November 1992 The American Journal of Medicine Volume 93

481

SYPHILISTREATMENTIN HIV INFECTION/ DOWELLET AL

penicillin [15-23]; this sequence of events is thought to follow the treatment of an immunologically incompetent host with what previously had been regarded as an effective antibiotic regimen [6,24].A number of authorities [5,25,26]have suggested that ceftriaxoneshould be considered for the treatment of syphilisin HIV-infected patients,although others [27,28]remain skeptical.Ceftriaxone achieves high serum levels,effectivelypenetrates the blood-brain barrier [29],and inactivatesT. pallidum at three to five times the concentration of penicillin [30,31]. Preliminary reports have indicated favorable results using ceftriaxoneto treat incubating syphilis [32], primary [32-35] and secondary [32,34,35] syphilis,and, in a singlecase,neurosyphilis [36].In practices that provide care for large numbers of HIV-infected patients, however, syphilis presents as latent disease or asymptomatic neurosyphilis in the great majority of cases (Dowell ME, unpublished observations). We are unaware of published studies of the efficacy of ceftriaxone in these conditions. Despite the paucity of data, physicians who provide care for HIV-infected individuals increasingly have made the decision to treat latent syphilis or neurosyphilis with ceftriaxone rather than with a penicillin regimen. The purpose of this article is to describe the results of ceftriaxone therapy in a sizeable group of patients. PATIENTS A N D METHODS Patient Selection

agents with efficacy against T. paUidum were given during that time. Serologic Tests

RPR, VDRL, and M H A - T P tests were done by clinical laboratories in accord with generally accepted guidelines [37],using R P R antigen suspension (Becton-Dickinson, Lincoln Park, NJ), V D R L test (Difco,Detroit,MI), and M H A - T P kits (Miles Corp., Elkhart, IN). R P R and M H A - T P testswere carried out on serum specimens, and the V D R L reactionon cerebrospinalfluid(CSF). Endpoint dilutions were utilizedfor reactive R P R and V D R L tests but not for the M H A - T P . CSF Analysis

CSF was regarded as abnormal if it contained greater than or equal to 6 white blood cells (WBCs)/mm 3, greater than or equal to 46 mg/dL protein,or lessthan or equal to 45 mg/dL glucose,or ifthe V D R L testwas reactive.These definitionsare essentiallyidenticalto those used by earlierinvestigators [10].Cryptococcal antigen was routinelyassayed, and allspecimens were cultured for bacteria, mycobacteria, and fungi. Treatment

Therapy was selected by individual physicians who staffedthe participatingclinics.As a resultof considerations summarized in the Introduction, some physicians who regularly attended at these clinicstreated syphiliswith ceftriaxoneI g (or rarely 2 g) intravenouslyfor 10 to 14 consecutivedays or I g intramuscularlyon weekdays until10 to 14 doses had been administered. Although these physicians regarded lumbar puncture as part of the routine evaluation of patients with a reactive serologicresult,most patients declined to undergo this procedure, thus explainingthe availabilityof CSF analysisin only a minority of cases.Other physicianswho attended lessregularlyat the clinicsdid not recommend routinelumbar puncture, insteadtreatingpatients with three doses of 2.4 million units benzathine penicillinat weekly intervals.Since assignment of a patient to one or another physician in these clinicsis made by a nonmedical receptionist in random fashion,unrelated to the nature or severity of the illness,the groups treated with ceftriaxone or penicillinare regarded as comparable.

Patients sought treatment for HIV-infection and/or its related complications at the Thomas Street Clinic of the Harris County Hospital District or the Special Medicine Clinic of the Veterans Affairs Medical Center, Houston, clinics that provide total care for HIV-infected persons. A screening rapid plasma reagin (RPR) test is done on all patients at the time of enrollment; if the RPR is reactive, a microhemagglutination assay for antibody to T. paUidum (MHA-TP) is also done. Records of all subjects who were found to have RPR reactive at greater than or equal to 1:4 at the time of admission to the clinics between November 1989 and February 1991 were reviewed. Many of these patients received medical care from one of the authors. Patients were considered for inclusion in this study if the diagnosis of latent syphilis or asymptomatic syphilis was made, and outpatient treatment was prescribed using ceftriaxone or benzathine penicil- Response lin. Data were included if patients received every Based on a review of all pretreatment and postdose of the prescribed antibiotic, if results of sero- treatment clinical observations and serologic relogic tests and clinical observation for 6 months or sults, patients were stratified to one of the following more were available, and if no other antimicrobial four groups: Group I: Response. A fourfold or great-

482

November1992 The American Journal of Medicine Volume93

SYPHILIS TREATMENT IN HIV INFECTION / DOWELL ET AL

TABLE I

TABLE II

Schedulesof Treatment With Ceftriaxone*

Classification

No. of Patients Initial therapy (n = 43) 7 6 30 Relapseor failure (n = 10) 1 3 4 2

lg

2g

Daily 10 d 14 d

Daily 14 d

Asymptomatic NS Documented latent Presumed latent

-1 3

5 5 26

2 -1

Symptomatic NS Asymptomatic NS Documented latent Presumed latent

_ ----

_ ----

_1 3 4 1

Clinical Diagnosis

by Responseto Ceftriaxoneor Penicillin* lg Ceftriaxone

Group Classification I II I11 IV

Response Serofast Relapse Failure

(n=43) 28 (65%) 5 (12%) 9 (21%) 1 (2%)

Daily 10d 14d 2 -2

24 5 7

2g Daily 14d

Penicillin ( n = 13)

2 --1

8 (62%) 1 (8%) 2 (15%) 2 (15%)

*Ceftriaxonedosesare as indicated.All penicillin-treatedpatientsreceivedthree doseseachof 2.4 million unitsbenzathinepenicillinat weeklyintentals.

NS neurosyphilis. *Data aregivenasthe numberof patientswho receivedeachtreatmentschedule. tThe patient who developedsymptomatic neurosyphilisafter his initial ceftdaxonetherapywas treated with 24 million units penicillin daily, and one personleft the clinic without treatment; therefore,the numbersin the ceftriaxonetreatmentcolumnfor patientswith relapseor failuretotal onlyeight. =

er decline in the serum RPR titer that was sustained during the period of follow-up. Group II: Serofast. No change in titer during the follow-up period and no signs of progressive infection. Group III: Serologic relapse. An initial fourfold or greater decline in the serum RPR titer, followed by a subsequent fourfold or greater rise in serum titer. Although relapse cannot definitively be distinguished from reinfection, these patients were believed to have had relapses, based on the early return of elevated serologic titers in the absence of findings of primary or secondary syphilis and the clinicians' knowledge of their social history. Group IV: Treatment failure. An increase in the serum RPR titer (fourfold or greater) without an initial response, persistence of a high RPR titer (1:64 or greater), and/or development of clinical disease attributable to syphilis after treatment.

Statistics Serologic titers were compared by means of a two-tailed t-test using Minitab Statistical Software (Minitab, Inc., Austin, TX). RESULTS Cases in Context During the 16 months under study, 116 patients were diagnosed as having latent syphilis or asymptomatic neurosyphilis at the 2 participating clinics; these cases represent 92.6% of all cases of diagnosed syphilis and 5.3% of all patients seen at those clinics during that time period. Forty-three patients who received ceftriaxono and 13 who received benzathine penicillin met all the criteria for inclusion. Results from the remaining 60 patients were not included because of insufficient follow-up data (29

patients) or failure to comply with the prescribed course of antibiotic therapy and/or concurrent administration of an antibiotic effective against T. paUidum for another indication during the treatment period (31 patients).

Ceftriaxone Treatment Forty-three patients received ceftriaxone. These individuals (38 males, 5 females; mean age: 34.9 years) were free of chancre, rash, or other manifestations of secondary syphilis and had no symptoms suggestive of neurosyphilis at the time they were found to have a reactive RPR. These 43 individuals were asked, as part of the evaluation of their reactive VDRL, to undergo lumbar puncture; 13 assented. The CSF was abnormal in 7 of 13 patients (58%), with pleocytosis in 5, elevated protein levels in 6, and reactive VDRL in 7; these 7 individuals were diagnosed as having documented neurosyphilis. Those six patients whose CSF was normal were diagnosed as having documented latent syphilis. The remaining 30 patients in whom a lumbar puncture was not done were regarded as having presumed latent syphilis, although based on the results in those who underwent lumbar puncture, it is reasonable to believe that as many as half may, in fact, have had neurosyphilis. The schedules of treatment for these 43 patients are summarized in Table I. Twenty of the 43 patients (47%) had been treated with one to three doses of benzathine penicillin in the preceding 3 years without clinical evidence of reinfection thereafter. As shown in Table II, 28 of 43 ceftriaxone-treated patients (65%) responded to therapy and were stratified to Group I. Five patients (12%) were serofast (Group II). Nine (21%) had serologic evidence of relapse (Group III); at the time of relapse, none had lesions of active early syphilis or clinical evidence of neurosyphilis. Therapy failed in one patient (2%) (Group IV); after treatment with 14 daily doses of 2 g ceftriaxone for documented asymptomatic neurosyphilis, he developed symptomatic

November 1992 The American Journal of Medicine Volume 93

483

SYPHILIS TREATMENT IN HIV INFECTION / DOWELL ET AL TABLE III Serum RPR, CSF Abnormalities, Treatment, and Outcome in Patients With Asymptomatic Neurosyphilis*

CD4

RPR

CerebrospinalFluid Protein (cells/mm3) (mg/dL) WBC

Patient

(cells/mm3)

(serum titer)

VDRL

Treatmentt

1

450

2,048 64 (18)*

60 10

40 35

2 2

A

2

590

2,048 64 (10)

37 2

72 80

4 4

A

3

217

1,024 ND

202 0

51 47

4 2

A

4

160

4,096 256 (6)

23

50 Refused lumbar puncture

2

A

5

17

32 2 (6)

1

54 Refused lumbar puncture

1

B

6

767

64 32 (12)

0

58 Refused lumbar puncture

2

B

7

235

128 ~ 128 (6)

10 1

1 8

A

60 70

Outcome Group

IV~

RPR= rapidplasmareagin;CSF= cerebrospinalfluid;WBC= whitebloodcellcount;VDRL= VenerealDiseaseResearchLaboratorytest;ND = notdone. *Thesecondlineof datafor eachpatientindicatesthelastserumRPRwiththetime(in months)aftertreatmentshownin parentheses,andthefollow-up.CSFanalysisdone4 to 6 weeksaftercompletionof ceftriaxonetreatment. tA = ceftriaxone,2 g dailyfor 14days;R = ceftriaxone,1 gdailyfor 14days. tPatient,describedin text,hadacerebrovascularaccidentwith increasingVDRLtiterin CSF.

neurosyphilis, manifested by a cerebrovascular accident and a CSF VDRL titer that increased from 1:1 to 1:8. Of the seven patients with documented neurosyphilis (Table III), five responded to therapy (Group I), documented in two by decreased CSF WBC counts within 4 to 6 weeks and subsequent declines in serum RPR, in one by marked improvement in CSF abnormalities (no repeat RPR done; patient left clinic), and in two who refused to allow a follow-up lumbar puncture by a fourfold or greater decline in the serum RPR. One subject refused a follow-up lumbar puncture and was serofast (Group II), and one, as noted above, developed symptomatic neurosyphilis (Group IV). The rate of response in the six patients who had documented latent syphilis was similar, with five serologic responses (Group I) and one serologic relapse (Group III). As shown in Tables I and II, initial ceftriaxone therapy was considered to have failed in a total of 10 patients (9 in Group III and 1 in Group IV). The nine patients with serologic relapse were all asymptomatic. Seven allowed a lumbar puncture to be done, of whom three (43%) had CSF abnormalities consistent with asymptomatic neurosyphilis (pleocytosis and elevated protein levels in three, VDRL reactive in two). Eight of these Group III patients were re-treated with ceftriaxone 2 g daily for 14 doses. Follow-up of these patients has been for 6 months or less; in this limited time, seven appear to have responded, and one has failed to respond. The

484

patient who developed symptomatic neurosyphilis was treated with intravenous penicillinG, 24 million units dailyfor 14 days, without clinicalor serologic response. One patient left the clinicwithout treatment and was lost to follow-up.

Penicillin Treatment Thirteen patients (11 males, 2 females; mean age: 30.2 years) who were treated with benzathine penicillinmet the criteriafor inclusion.None had symptoms or signs of active syphilis,and, for reasons noted above, none underwent lumbar puncture; these individualswere, therefore,regarded as having presumed latentsyphilis.As shown in Table II, 8 of the 13 (62%) were classifiedas responders (Group I); 1 (8%) was serofast (Group II);2 (15%) had a relapse (Group III);and treatment failedin 2 (15%) (Group IV). The two Group IIIpatientswere asymptomatic at the time of serologicrelapse and consented to lumbar puncture; both had CSF abnormalities including a reactive V D R L and were diagnosed as having neurosyphilis.Both were subsequently treated with ceftriaxone,2 g dally for 14 days, and appear to have responded serologicallyin the limited time of follow-up observation.

Serologic Tests for Syphilis The median baseline RPR titer in subjects who responded to treatment (Group I) or those who relapsed (Group III) was higher than in those who

November 1992 The American Journal of Medicine Volume 93

SYPHILIS TREATMENT IN HIV INFECTION / DOWELL ET AL

were serofast (Table IV). When these data were converted to geometric means, comparison showed that the RPR titers were significantly higher for Groups I and III than for Group II (p <0.05). Every patient in this series had a positive MHA-TP result as a criterion for inclusion; in one, who was thought to have responded to treatment, this test became nonreactive during the period of follow-up. Relation Between HIM Status and Outcome

Nine patients had CD4 cell counts exceeding 500/ mm 3 on initial evaluation, 25 patients had counts between 200/mm3 and 500/ram ~, and 22 patients had counts below 200/mm 3. There was no correlation among the CD4 cell count (Table IV), the clinical state of infection (HIV antibody only, AIDSrelated complex, or AIDS), and the outcome of treatment. Eleven percent of those who responded to therapy (Group I) had CD4 cell counts greater than or equal to 500/ram 3, 44% had counts between 200/mm 3 and 500/ram 3, and 42% had counts below 200/ram 3, compared with 20%, 45%, and 35%, respectively, of the patients who stratified to other groups based on therapeutic response. COMMENTS

Concern over the apparent failure of conventional schedules of penicillin to eradicate syphilis in the presence of HIV infection has led to the suggestion that ceftriaxone be considered in its place [5,25,26], a recommendation that seems eminently reasonable, although based on limited experience in nonHIV-infected syphilitic patients [32-36]. Our study is the first to present data from a series of cases in which ceftriaxone was used to treat latent or asymptomatic neurologic syphilis in HIV-infected individuals, the modes of presentation in the great majority of cases of syphilis in our HIV-based clinics. Treatment with 10 to 14 doses of this antibiotic appeared to arrest the progression of syphilis in 33 of 43 (77%) patients, including 6 of 7 who had documented neurosyphilis. The remaining 10 patients (23%) either failed to respond (1 patient) or had serologic evidence for reactivation of disease after an initial response (9 patients). Differences in dosage or duration of therapy did not appear to affect the outcome. A similar outcome was observed in a small series of patients, treated in the same clinics by physicians who did not recommend lumbar puncture and who elected to prescribe three doses of benzathine penicillin. Only limited data are available on the effectiveness of any antibiotic treatment for syphilis in HIVinfected subjects. With conventional doses of benzathine penicillin for secondary syphilis, a high rate

TABLE IV RelationBetweenOutcomeAfterTherapyand Rapid PlasmaReagin (RPR)Titersand CD4 CellCountsin 56 TreatedPatients Mean CD4 Cells

Group

(/ram3) *

Median

RPR Titer Range

GM

I

281

1:64

1:4-1:16,384

1:70 ~

II

347

1:16

1:8-1:64

1:18

III

263

1..64

1:8-1:2,048

1:106 t

IV

235

--

1:128

1:51

GM = geometricmean. *Differencesin CD4 cell countsare not differentamonggroups. tResultsfor GroupI and GroupIll are eachsignificantlygreaterthan resultsin GroupII, p <0.05. Outlierswereexcludedfor purposesof determiningthe GM.

of failure has been reported as shown by the lack of resolution of skin lesions and/or the failure of the RPR titer to decline [38,39]. Of 40 reported HIVinfected patients who developed neurosyphilis, 16 were known to have been treated with benzathine penicillin, generally a single dose of 2.4 million units [6]. In our current series, 20 of 43 (47%) ceftriaxonetreated patients had previously received a course of benzathine penicillin and had no history of recognized reinfection to account for their reactive serologic results. From such data, one cannot, of course, comment on the incidence of treatment failure with benzathine penicillin, since the total number of HIV-infected patients treated for early syphilis (the denominator) is unknown and is presumably large. Because treatment with benzathine penicillin does not achieve detectable CSF levels of penicillin [40,41], it might seem reasonable to assume that this drug is not effective in treating neurosyphilis. In fact, in the pre-HIV era, the U.S. Public Health Service recommended three doses of benzathine penicillin to treat neurosyphilis, with nearly universal success in presumably immunologically competent hosts [14]. Lukehart et al [10] showed that a single dose of benzathine penicillin failed to clear treponemas from the CSF of HIV-infected, but not HIV-uninfected, individuals who had early syphilis. Three doses of benzathine penicillin did, however, clear the CSF in one patient. In the current series, we documented the presence of neurosyphilis in two patients shortly after three doses of 2.4 million units of benzathine penicillin were given; these individuals had not undergone lumbar puncture and may have already had neurosyphilis prior to therapy. Although three doses of benzathine penicillin should not be used to treat neurosyphilis and probably should not be used for presumed latent syphilis in HIV-infected persons, the efficacy of this regimen in proven latent infection probably deserves

November 1992 The American Journal of Medicine Volume 93

485

SYPHILIS TREATMENT IN HIV INFECTION / DOWELLET AL

further study, based on the overall response rate, which was similar to that for ceftriaxone. Daily administration of procaine penicillin is more likely to achieve effective antitreponemal penicillin levels in the CSF [39,40] and would appear to be preferable to benzathine penicillin; even fewer data are available on this regimen in HIVinfected patients. Lukehart et al [10] successfully re-treated their three patients in whom treatment with benzathine penicillin failed by administering 2.4 million units of procaine penicillin daily, together with 2 g of probenecid. Stoumbos et al [42] treated one patient who had syphilitic retinitis with a similar regimen, and McMillan et al [43] arrested early syphilis in 19 of 20 consecutive HIV-infected patients utilizing 10 daily doses of 600,000 U of procaine penicillin without probenecid. It is worth noting that concomitant administration of probenecid, which significantly increases CSF levels of penicillin [41], also increases the half-life of zidovudine and may cause adverse drug reactions in HIVinfected persons. Our data are consistent with findings from other clinics that see only HIV-infected patients in showing that syphilis is generally diagnosed in its latent stage or as asymptomatic neurosyphilis. Such findings differ from what is usually observed in clinics established to treat other STDs [44]. Unlike the pre-1980 experience in STD clinics, when the CSF was usually found to be normal in persons who underwent examination for reactive serologic tests [28,45], a substantial proportion of asymptomatic HIV-infected patients may have CSF findings consistent with neurosyphilis. The CSF was abnormal in 7 of 13 asymptomatic patients with an initially reactive RPR and 3 of 7 who had serologic failure or relapse after treatment, totaling 10 of 20 patients (50%). Although some CSF abnormalities may be attributable to HIV infection [46], the presence of a reactive CSF VDRL in nearly all of our cases and the documented improvement 4 to 6 weeks after treatment for syphilis support the concept that T. paUidum infection was responsible. The prevalence of neurosyphilis in asymptomatic HIV-infected men who have a reactive serum RPR titer greater than or equal to 1:4 might be used to bolster the recommendation [6,24,47] that evaluation in such cases must include lumbar puncture. Our results, however, show no difference in outcome after ceftriaxone was used to treat proven latent,presumed latent, or proven neurosyphilis,and one could use the data equally to argue against routine lumbar puncture if this approach to therapy is to be followed. Our findings confirm the earlierreport [48] that H I V infection is associated with higher-than-ex486

pected titersof antibody to cardiolipin,a phenomenon that has been attributed to polyclonal activation. Recent observations in our laboratory (unpublished observations)document the presence of cross-reactingantiphospholipidantibodiesin the serum of HIV-infected persons, which may also contribute to the height of the anticardiolipinantibody titer. Eighty-eight percent of our patients with latent syphilishad a serum R P R titergreater than or equal to 1:16,contrastingsharply with the experience of Fiumara [49] in HIV-negative patients, only 4% of w h o m had titersof this magnitude. The lowest serum titerobserved in a patient with neurosyphilis in our serieswas 1:32,whereas 75% of patients with neurosyphilis in the seriesof Burke and Schaberg [50] had R P R titerslessthan or equal to 1:4.Romanowski et al [44] showed that the rate of decline of serologictestresultswas unrelated to the originaltiterand was slower in patients with latent syphilisthan in those who had active disease (a fourfoldversus a twofold decline,respectively,within 6 months of treatment).The high proportion of our patients who exhibited a fourfoldor greater decline in 6 months suggests a greater degree of activityof disease in HIV-infected patients, despite the label of latency used to describe them. Our patients were selected for study based on a reactivenontreponemal serologictest,so we cannot address the question of false negativity of these tests in AIDS. W e [6,24] have previously stated, however, that the rare case of an RPR-negative, HIV-infected patient with syphilis [51] should be regarded as the exception that proves the rule, namely, that R P R is reactivein all but the earliest stages of syphilis.Loss of antibody to treponemal antigens, as detected by the M H A - T P or fluorescent treponemal antibody-absorption (FTA-ABS) tests,has been saidto occur in up to 38% of patients with symptomatic H I V infection [52,53],but these observations were made in persons who had no evidence to suggest a need for further treatment for syphilis.None of our patientshad a negative M H A T P result at the time that they had serologicevidence for ongoing syphiliticinfection,and only one lost antibody as detected by M H A - T P in the relativelyshort time of follow-up.Even in immunocompetent hosts, the M H A - T P result may revert to nonreactive aftertreatment, but that occurs in subjectswith a firstbout of syphilisand slmost always in primary infection [44]. Taken together, these observations show that evaluation and treatment of syphilisin HIV-infected persons present major challenges,with many important questions remaining unanswered. Favoring the use of lumbar puncture in an HIV-infected patientwhen a serologictestfor syphilisis firstfound

November 1992 The American Journal of Medicine Volume 93

_

SYPHILIS TREATMENT IN HIV INFECTION / DOWELL ET AL

to be reactive is our observation that half of such persons may, in fact, have asymptomatic neufbsyphilis. Even if therapy is ultimately determined to be the same whether the CSF is normal or abnormal, repeated lumbar puncture may be indicated to assess the response of neurosyphilis to therapy, in which case an initial, baseline CSF analysis is required. It is possible, however, that a decline in the serum RPR titer may provide sufficient evidence of a response. The optimal antibiotic regimen remains to be determined. For proven latent syphilis, defined by the absence of clinical signs of disease with a reactive serologic result arid a normal CSF, no evidence shows any regimen to be more efficacious than three doses of 2.4 million units of benzathine penicillin at weekly intervals. In our study, the number of failures did not appear greater than in those who were treated far more aggressively. If the CSF is abnormal, the inability of benzathine penicillin to provide measurable CSF levels of penicillin should probably disqiialify this drug, even though three doses at weekly intervals were regarded as effective in the pre-HIV era, and no study has actually shown this not to be the case in HIV-infected persons. Our study found that 10 to 14 1-g doses of ceftriaxone were moderately effective in treating HIV-infected patients for latent syphilis or asymptomatic neurosyphilis. The finding of a 23% relapse rate within I to 2 years supports the concept [6,24] that antimicrobial therapy is not likely to bring about a biologic cure of syphilis in a person whose immune system is compromised by HIV infection. We do not envision a higher rate of long-lasting responses if intravenous penicillin or higher doses of ceftriaxone were used, nor have any data shown this to be the case. Although we do not have data on the efficacy of treatment with 10 to 14 days of procaine penicillin, we believe that this far less costly regimen is likely to be as effective as ceftriaxone since the basic problem is the immunologic status of the host, not antibiotic dosage; in any case, a prospective study comparing these regimens is needed.

ACKNOWLEDGMENT We are deeply grateful to Inez Campbell for secretarial assistance.

REFERENCES 1. Centers for DiseaseControl. Primary and secondary syphilis--United States, 1981-1990. MMWR 1991; 40: 314-23. 2. Tramont EC. Syphilis in the AIDS era. N Engl J Med 1987; 316: 16(7)-1. 3. Stamm WE, Handsfield HH, RompaloAM, Ashley RL, Roberts PL, Corey L. The association between genital ulcer disease and acquisition of HIV infection in homosexual men. JAMA 1988; 260: 1429-33. 4. Quinn TC, Glasser D, Cannon RO, et aL Human immunodeficiency virus infection among patients attending clinics for sexually transmitted diseases.N Engl J Med 1988; 318: 197-202.

5. Hook EW III. Syphilis and HIV infection. J Infect Dis 1989; 160: 530-4. 6. Musher DM, Hamill RJ, Baughn RE. Effect of human immunodeflciency virus (HIV) infection on the course of syphilis and on the response to treatment. Ann Intern Med 1990; 113: 872-81. 7. Chesney AM, Kemp JE. Incidence of Spirochaeta pal/ida in cerebrospinal fluid during early stage of syphilis. JAMA 1924; 83: 1725-8. 8. Stokes JH. Modern clinical syphilology: diagnosis-treatment-case study. Philadelphia: WB Saunders, 1934. 9. Kampmeier RH. Essentials of syphilology. Philadelphia: Lippincott, 1943. 10. Lukehart SA, Hook EW III, Baker-Zander SA, Collier AC, Critchlow CW, Handsfield HH. Invasion of the central nervous system by Treponema pallidum: implications for diagnosis and treatment. Ann Intern Med 1988; 109: 855-62. 11. Mohr JA, Griffiths W, Jackson R, Saadah H, Bird P, Riddle J. Neurosyphilis and penicillin levels in cerebrospinal fluid. JAMA 1976; 236: 2208-9. 12. Polnikorn N, Witoonpanich R, Vorachit M, Vejjajiva S, Vejjajiva A. Penicillin concentrations in cerebrospinal fluid after different treatment regimens for syphilis. Br J Vener Dis 1980; 56: 363-7. 13. Perdrup A, Jorgensen BB, Pederson NS. The profile of neurosyphilis in Denmark. A clinical and serological study of all patients in Denmark with neurosyphilis disclosed in the years 1971-1979 inclusive by Wasserman reaction (CWRM) in the cerebrospinal fluid. Acta Derma Venereol Suppl (Stockh) 1981; 96: 3-14. 14. Jaffe HW, Kabins SA. Examination of cerebrospinal fluid in patients with syphilis. Rev Infect Dis 1982; 4 Suppl: $842-7. 15. Harris RL, Rutecki PA, Donovan DT, Bradshaw MW, Williams TW Jr. Fever, headache, and hearing loss in a young homosexual man. Hosp Pract 1985; 20: 167-72. 16. Bayne LL, Schmidley JW, Goodin DS. Acute syphilitic meningitis: its occurrence after clinical and serologic cure of secondary syphilis with penicillin G. Arch Neurol 1986; 43: 137-8. 17. Johns DR, Tierney M, Felsenstein D. Alteration in the natural history of neurosyphilis by concurrent infection with the human immunodeficiency virus. N Engl J Med 1987; 316: 1569-72. 18. Berry CD, Hooton TM, Collier AC, Lukehart SA. Neurologic relapse after benzathine penicillin therapy for secondary syphilis in a patient with HIV infection. N Engl J Med 1987; 316: 1587-9. 19. Clark R, Carlisle JT. Neurosyphilis and HIV infection. South Med J 1988; 81: 1204-5. 20. LanskaMJ, LanskaDJ, SchmidleyJW. Syphiliticpolyradiculopathyin an HlVpositive man. Neurology1988; 38: 1297-301. 21. KaseCS, LevitzSM, WolinskyJS, SuBsCA. Pontinepure motor hemiparesis due to meningovascularsyphilisin human immunodeficiencyvirus-positivepatients [letter]. Arch Neurol 1988; 45: 832. 22. KatzDA, BergerJR. Neurosyphilisin acquiredimmunodeficiencysyndrome. Arch Neurol 1989; 46: 895-8. 23. Smith ME, CanalisRF. Otologicmanifestationsof AIDS:the otosyphilisconnection. Laryngoscope1989; 99: 365-72. 24. Musher DM. Syphilis, neurosyphilis,penicillin and AIDS. J Infect Dis 1991; 163: 1201--6. 2,5.Tramont EC. Treponemapallidum(syphilis). In: MandellGL, DouglasRGJr., Bennett JE, editors. Principlesand practiceof infectiousdiseases.3rd ed. New York: Churchill Livingstone, 1990: 1794-809. 26. SanfordJ. Guideto antimicrobialtherapy. West Bethesda, MD: Antimicrobial Therapy Inc., 1991. 27. Le SauxN, RonaldAR. Roleof ceftriaxonein sexuallytransmitted diseases. Rev Infect Dis 1989; 11: 299-309. 28. Zenker PN, Rolfs RT. Treatment of syphilis, ].989. Rev Infect Dis 1990; 12 Suppl: $591-609. 29. Scheld WM, Rocha H, Sande MA, Bryan JP. Rationale for clinical trials in evaluating ceftriaxone in the therapy of bacterial meningitis. Am J Med 1984; 77 Suppl 4C: 42-9. 30. Johnson RC, Bey RF, Wolgamot SJ. Comparison of the activities of ceftriaxone and penicillin G against experimentally induced syphilis in rabbits. Antimicrob Agents Chemother 1982; 21: 984-9. 31. Korting HC, Walther D, Riethmuller U, Meurer M. Comparative in vitro susceptibility of Treponema pallidum to ceftizoxime, ceftriaxone and penicillin G. Chemotherapy 1986; 32: 352-5. 32. Hook EW III, Roddy RE, Handsfield HH. Ceftriaxone therapy for incubating and early syphilis. J Infect Dis 1988; 158: 881-4. 33. Moorthy TI, Lee CT, Lira KB, Tan T. Ceftriaxone for treatment of primary

November 1992 The American Journal of Medicine Volume 93

487

SYPHILISTREATMENT IN HIV INFECTION / IX)WELLET AL syphilis in men; a preliminary study. Sex Transm Dis 1987; 14: 116-9. 34. V'~glale R, Bruno J, Gilbert P. Ceftriaxone in the treatment of primary and secondary syphilis. A comparative study with benzathine penicillin [abstract 75]. In: Proceedingsof the 15th International Congresson Chemotherapy, Istanbul, 1987. 35. Schofer H, Vogt HJ, Milbrodt R. Ceftriaxone for the treatment of primary and secondary syphilis. Chemotherapy 1989; 35: 140-5. 36. Hook EW Ill, Baker-Zander SA, Moskovitz BL, Lukehart SA, Handsfield HH. Ceftriaxone therapy for asymptomatic syphilis. Case report and Western blot analysis of serum and cerebrospinal fluid IgG response to therapy. Sex Transm Dis 1986; 13: 185-8. 37. Manual of tests for syphilis. Atlanta, U.S. Public Health Service, Centers for Disease Control, 1969: 15-22, 33-42. 38. Frederick WR, Delapenha R, Barnes S, Olopenia L, Saxinger C, Greaves W. Secondary syphilis and HIV infection [abstract 1175]. In: Abstracts of the 281h Interscience Conference on Antimicrobial Agents and Chemotherapy. Washington, DC: American Society for Microbiology, 1988: 320. 39. Gregory N, Sanchez M, Buchness MR. The spectrum of syphilis in patients with HIV infection. J Am Acad Dermatol 1990; 23: 1061-7. 40. Goh BT, Smith GW, Samarasinghe L, Singh V, Lim KS. Penicillin concentrations in serum and cerebrospinal fluid after intramuscular injection of aqueous procaine penicillin 0.6 MU with and without probenecid. Br J Vener Dis 1984; 60: 371-3. 41. van der Valk PG, Kraal EJ, van Voorst Vader PC, Haaxma-Reiche H, Snijder JA. Penicillin concentrations in cerebrospinal fluid (CSF) during repository treatment regimen for syphilis. Genitourin Med 1988; 64: 223-5. 42. Stoumbos VD, Klein ML. Syphilitic retinitis in a patient with acquired immunodeficiency syndrome-related complex [letter]. Am J Ophthalmo11987; 103: 104. 43. McMillan A, Young H, Peutherer JF. Influence of human immunodeficiency

488

virus infection on treponemal serology in patients who have been treated for syphilis. J Infect 1990; 21: 95-103. 44. Romanowski B, Sutherland R, Fick GH, Mooney D, Love EJ. Serologic response to treatment of infectious syphilis. Ann Intern Med 1991; 114: 1005-9. 45. Musher DM. Evaluation and management of an asymptomatic patient with a positive VDRL reaction. Curr Clin Top Infect Dis 1988; 9: 147-57. 46. Marshall DW, Brey RL, Butzin CA, Lucey DR, Abbadessa SM, Boswell RN. CSF changes in a longitudinal study of 124 neurologically normal HIV-l-infected U.S. Air Force personnel. J AIDS 1991; 4: 777-81. 47. Matlow AG, Rachlis AR. Syphilis serology in human immunodeficiency virusinfected patients with symptomatic neurosyphilis: case report and review. Rev Infect Dis 1990; 12: 703-7. 48. Hutchinson CM, Rompalo AM, Reichart CA, Hook EW III. Characteristics of patients with syphilis attending Baltimore STD clinics. Multiple high-risk subgroups and interactions with human immunodeficiency virus. Arch Intern Med 1991; 151: 511-6. 49. Fiumara NJ. Serologic responses to treatment of 128 patients with late latent syphilis. Sex Transm Dis 1979; 6: 243-6. 50. Burke JM, Schaberg DR. Neurosyphilis in the antibiotic era. Neurology 1985; 35: 1368-71. 51. Hicks CB, Benson PM, Lupton GP, Tramont EC. Seronegative secondary syphilis in a patient infected with the human immunodeflciency virus (HIV) with Kaposi sarcoma: a diagnostic dilemma. Ann Intern Med 1987; 107: 492-5. 52. Haas JS, Bolen 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. 53. Johnson PDR, Graves SR, Stewart L, Warren R, Dwyer B, Lucas CR. Specific syphilis serological tests may become negative in HIV infection. AIDS 1991; 5: 419-24.

November 1992 The American Journal of Medicine Volume 93