Treatment of Q Fever Endocarditis: Comparison of 2 Regimens

ORIGINAL INVESTIGATION Treatment of Q Fever Endocarditis Comparison of 2 Regimens Containing Doxycycline and Ofloxacin or Hydroxychloroquine

Didier Raoult, MD, PhD; Pierre Houpikian, MD; Herve´ Tissot Dupont, MD; Jean Marc Riss, MD; J. Arditi-Djiane, MD; Philippe Brouqui, MD

Background: Q fever endocarditis, caused by Coxiella ing this regimen only. The mean duration of therapy for burnetii, is fatal in 25% to 60% of patients. Currently, treat- cure in this group was 55 months (median, 60 months). ment with a long-term tetracycline and quinolone regi- Twenty-one patients received the doxycycline and hy- men for at least 4 years is recommended, although relapses droxychloroquine regimen: 1 patient died of a surgical are frequent. complication, 2 are still being treated, 17 were cured, and 1 is currently being evaluated. Two patients treated for Methods: Between January 1987 and December 1997, 12 months but none of the patients treated for longer than the reference treatment of Q fever endocarditis was com- 18 months relapsed. The mean duration of treatment in pared with one of doxycycline and hydroxychloroquine this group was 31 months (median, 26 months). No sig- sulfate. Patients were treated by conventional therapy un- nificant differences were observed between the 2 regi- til May 1991 and then by the new regimen. Microimmu- mens in terms of death, valve surgery, or tolerance. The nofluorescence was used for antibody-level determina- mortality rate for both regimens in this study was 5%. tion for diagnosis and follow-up. Conclusion: Prescription of the doxycycline and hy- Results: Thirty-five patients were included in the study, droxychloroquine combination for at least 18 months al- 26 males and 9 females. Of 14 patients treated with a doxy- lows shortening of the duration of therapy and reduc- cycline and quinolone combination, 1 died, 7 relapsed tion in the number of relapses. (3 were re-treated and 4 switched to the new regimen), 1 is still being treated, and 5 were considered cured us- Arch Intern Med. 1999;159:167-173

FEVER, a rickettsiosis response to phases I and II. However, caused by Coxiella bur- because of its intracellular position, C netii, was originally burnetii is difficult to eradicate, explain- described by Derrick in ing why—when the disease was 1937.1 This zoonosis is described during the 1960s and 1970s— Q widespread throughout relapses and even death occurred fre- the world and can present as either acute quently. In some series,4 two thirds of or chronic disease. Endocarditis is the the patients died. Since that time, the most serious and often fatal form of role of combined and prolonged antibi- chronic Q fever, and presents unspecifi- otic drug therapy for these patients has cally as a blood culture–negative endo- been emphasized.5-7 The current recom- From the Unite´ des Rickettsies, carditis. Diagnosis relies on nonspecific mended treatment duration is from 3 Faculte´deMe´decine, Universite´ cardiac findings, evidence of an infec- years to lifetime, even when quinolone is de la Mediterrane´e CNRS tious or inflammatory process, and added to the reference tetracycline (Drs Raoult, Houpikian, Tissot detection of specific antibodies. Coxiella therapy. This combination treatment Dupont, and Brouqui); burnetii is a strict intracellular gram- enhances patient survival,6 but the treat- Infectious Disease Units, negative bacterium. It lives and multi- ment of Q fever endocarditis is now the Hoˆpital F. Houphoue¨t Boigny plies in the phagolysosomes of infected longest of any bacterial disease, includ- (Drs Houpikian and Brouqui) cells at pH 4.8.2,3 Coxiella burnetii exhib- ing leprosy and other mycobacterial dis- and Hoˆpital Conception its a phase variation (phases I and II), eases.8 By testing C burnetii antibiotic (Dr Tissot Dupont); Ophthalmology Department, which is extremely valuable for diagno- susceptibility in a cellular model of infec- Hoˆpital Timone (Dr Riss); and sis. The serologic response of patients tion, it was previously shown that intra- Toxicology Laboratory, Hopital with acute Q fever is mainly directed cellular pH may significantly diminish Salvator (Dr Arditi-Djiane), against phase II, whereas patients with the bactericidal potency of antibiotic Marseille, France. chronic Q fever exhibit a serologic drugs.9,10 This bactericidal effect of doxy-

ARCH INTERN MED/ VOL 159, JAN 25, 1999 167

Downloaded From: https://jamanetwork.com/ on 09/27/2021 PATIENTS AND METHODS SURVEILLANCE AND FOLLOW-UP All patients were monitored monthly with a physical ex- PATIENTS amination, blood cell count, transaminase dosage, serologic examination, and hydroxychloroquine plasma level Patients were treated from January 1987 to May 1991 with determination. When treatment was stopped, blood cul- doxycycline and ofloxacin and from May 1991 to Decem- tures were obtained for C burnetii and other organisms. A ber 1997 with doxycycline and hydroxychloroquine regi- cardiologist examined each patient 2 to 4 times a year, and mens. Diagnosis was performed by confirmation of C bur- an ophthalmologist (J.M.R.) performed a flash electroreti- netii infection in patients with chronic endocarditis based nogram and a fundus examination twice a year to detect on previously published criteria.12 An IgG titer of 1/1600 chloroquine accumulation.14 Patients were considered cured and an IgA titer of 1/100 directed against phase I of C bur- when IgG antibodies to C burnetii phase I were lower than netii were required for diagnosis. By using the Duke crite- 1/800 and IgM and IgA antibodies were lower than 1/50.12 ria modified for C burnetii, all these patients had a definite When these results were obtained, antibiotic drug treat- diagnosis of endocarditis.13 Treatment was prescribed ment was stopped and the patients were further moni- by 1 of us (D.R.) after receiving patient information and tored by clinical examination and blood sampling, includ- consent. ing blood cell count, serologic examination, and blood culture performed for C burnetii and other organisms.15 Pa- TREATMENT REGIMENS tients were examined monthly for the first 6 months, every 3 months for a further year, and subsequently every 6 Each patient initially received 1 tablet of doxycycline (100 months. The patients had an echocardiographic examina- mg) twice daily and 1 tablet of ofloxacin (200 mg) 3 times tion twice a year for 2 years. daily. In the study’s second stage, patients received doxycycline at the same dose and 1 tablet of hydroxychloro- SERUM HYDROXYCHLOROQUINE quine (200 mg) 3 times a day. Doses of hydroxychloro- CONCENTRATION DETERMINATION quine were adapted after monthly plasmatic level evaluation to maintain a concentration between 0.8 and 1.2 µg/mL. Measurement of hydroxychloroquine blood levels was per- Patients were made aware of the risks associated with over- formed by liquid chromatography as previously described16 exposure to the sun and that they should wear gloves and using a 25-cm Si 60,5 lichrospher column (Merck, Darmstadt, a hat and use protective sunscreens. Patients with an evo- Germany). The mobile phase consisted of a mixture of 20% lutive immunocompromised disease such as cancer or ac- acetonitrite and ammoniac ethanol (92/8 by volume). Sepa- quired immunodeficiency syndrome were excluded from ration was performed by means of a gradient elution. Hydroxy- the study. chloroquine and quinine, the internal standard, were extracted

cycline was restored by incubating cells with chloro- diac incompetency. The 7 patients who relapsed had an quine, 1 mg/mL, which acted as a lysosomotropic alka- increase of antibody titers associated with clinical signs linizing agent and raised the pH of the phagolysosome within 6 months of treatment interruption. These fail- from 4.8 to 5.7,11 the optimal doxycycline activity being ures were observed only during the early stages of the obtained at pH 6.6. The purpose of the present study regimen. Subsequently, stable low titer of anti–phase I was to assess the clinical efficacy of the combination of antibodies had to be observed before therapy was stopped doxycyline and hydroxychloroquine on the treatment (IgG, Ͻ800; IgA, Ͻ50). The mean treatment duration for of Q fever endocarditis compared with a previously ini- patients considered cured was 55 months (median, 60 tiated regimen of doxycyline and ofloxacin.1,6 months). No patient relapsed when treated for longer than 4 years, and only 1 patient relapsed after 3 years of treat- RESULTS ment. However, 8 patients are still being treated or were switched to a doxycycline and hydroxychloroquine regi- Twenty-six men and 9 women were included in the study. men (Table 1). Fourteen patients were treated with a combination of Twenty-one patients were treated with a combina- doxycycline and ofloxacin (Table 1). One patient died tion of doxycycline and hydroxychloroquine and 7 relapsed and were subsequently treated with the (Table 2). One patient died 30 months after complet- same regimen (3 patients) or were switched to the new ing treatment: he underwent mitral valve replacement regimen (4 patients). Five patients were cured using this and died postoperatively of cardiac failure, although the regimen only. A total of 21 treatments using this regi- valve was sterile. All patients experienced photosensi- men were prescribed. All the patients experienced pho- tivity, mainly on the hands and nose. One patient had tosensitivity during the summer after beginning treat- an irreversible cutaneous pigmentation caused by doxy- ment. One patient had an irreversible cutaneous cyline treatment, and 1 patient had an infraclinical reti- pigmentation attributable to doxycycline use.18 No diar- nal accumulation of hydroxychloroquine that led to rhea or joint pain was reported. The patient who died withdrawal of this drug (patient 3). However, because was still receiving antibiotic drug treatment when he had the treatment was completed, it did not change the a valvular replacement and died postoperatively of car- course of the therapy. Two patients had decreased

ARCH INTERN MED/ VOL 159, JAN 25, 1999 168

Downloaded From: https://jamanetwork.com/ on 09/27/2021 in an alkaline medium. After mixing for 15 minutes, the shell-vial assay with human embryonic lung fibroblasts. organic phase was recovered and evaporated. The residue Three confluent shell-vials were inoculated for each blood was removed by acetonitrile, 50 µL, and 20 µL was in- sample and were incubated for 6 and 14 days at 37°C un- jected into the chromatograph. The standard range was 0.5 der a 5% carbon dioxide atmosphere. Detection of C bur- to 10.0 µg/mL. The limit of detection was 0.2 µg/mL. netii was carried out directly with the shell-vial assay by direct immunofluorescence labeling with a rabbit anti–C SEROLOGIC EXAMINATION burnetii polyclonal antibody. If immunofluorescence test results were negative after 6 and 14 days of incubation, the An indirect fluorescent antibody assay was used in this study. culture was negative. If positive, the culture was then sub- Antigens of C burnetii phase I and phase II (nine mile strain) cultured to yield the isolate. were placed on 30-spot slides (Dynex, Cel-Line Associates Inc, New Field, NJ) and fixed in cold acetone for 20 minutes. The ASSESSMENTS OF CLINICAL OUTCOME serum samples were diluted in phosphate-buffered saline so- lution (PBS) from 1:25 to 1:1600. After the serum samples were A patient was considered to be definitely cured only after overlayed onto the antigen dots, slides were incubated for 30 an incident-free 6 months of follow-up. A patient was con- minutes at 37°C in a moist chamber, then washed 3 times in sidered as relapsing if an unexplained cardiac abnormality PBS. Fluorescein-conjugated antiserum IgG (␥-chain specific), (valvular dysfunction or echocardiographic change), an in- IgM (µ-chain specific), or IgA (␣-chain specific) (Institut Pas- flammatory syndrome, hepatitis, or a 4-fold increase of IgG teur Production, Paris, France) diluted 1:200 in PBS was then or IgA antibodies to C burnetii was observed. When valvu- overlayed onto dried slides and were again incubated for 30 lar surgery was performed, samples of valve tissue were in- minutes in a moist chamber at 37°C. After washing in PBS, the oculated on cell tissue culture in an attempt to isolate C slideswereairdriedandmountedwithglycerol.Theslideswere burnetii15 and were studied by immunohistochemistry.17 examined under an epifluorescence microscope (magnifica- tion ϫ400; Nikon, Tokyo, Japan). Before determining the lev- STATISTICAL TESTS els of IgM and IgA, the serum IgG was removed using a rheumatoid factor adsorbent (Calbiochem-Behring, Marburg, Ger- Data were computerized using database and spreadsheet many). This factor was diluted in water (1:5), mixed with the programs (Microsoft Access and Excel, Microsoft Corp, Red- serum diluted in PBS (1:5), and centrifuged. mond, Wash). Statistical tests were performed using a soft- ware program (Epi-Info, Centers for Disease Control and CULTURE PROCEDURE Prevention, Atlanta, Ga). Proportions were compared using 1-tailed Fisher exact tests, whereas means were com- Isolation of C burnetii by cell culture was performed as pre- pared using Student t tests. A difference was considered sig- viously reported.15 Briefly, culturing was performed by the nificant at PϽ.05.

serum IgG (anti–phase I) antibody titers to less than cantly different in terms of relapse, which is much more 400 within 10 months of therapy despite high initial frequent with doxycycline and quinolone combination antibody levels and consequently discontinued treat- treatment. The mean hydroxychloroquine regimen du- ment after 1 year. They both relapsed within the first 6 ration was also significantly lower by 2 years (Table 3). months of follow-up. After these cases, we decided to Moreover, patients treated with doxycycline plus ofloxa- treat all patients for 18 months or longer, and none of cin for more than 48 months are still undergoing treat- the 15 patients treated subsequently relapsed. The 2 ment, and the regimens of 4 patients have been altered patients who relapsed were treated for a further 2 years. to doxycycline plus hydroxychloroquine, and 3 of them Among 21 patients, they were the only 2 who had a sig- are now cured. nificantly lower CD4+/CD8+ cell ratio.19 Hydroxychloro- We tried to compare the evolution of anti–phase I quine plasma levels were determined to maintain a antibody levels, but the initial levels varied from 1600 steady concentration of 1.0 ± 0.2 µg/mL. Six patients did to 256 000 for IgG and IgA. With our relatively small not require any change and received 600 mg for the number of patients, the curves were hardly interpret- duration. One patient, a man weighing 115 kg (patient able because of the size of the SD. However, the anti– 5), had low levels of hydroxychloroquine; after 3 phase I IgM antibodies were the first to reach a titer of months, the dosage was increased to 800 mg/d. In the less than 1:50, and this significantly more rapidly (6 16 remaining patients, the dosage was decreased to 400 months) with the doxycycline plus hydroxychloroquine mg/d within a mean of 6 months (range, 2-13 months). regimen than with the doxycyxline plus ofloxacin regi- Nine patients required a second dosage reduction to men (30 months). As for IgG and IgA, no significant 150 mg within a mean of 3.4 months (range, 2-7 differences were noted. months). The mean treatment duration for cured patients was 31 months, with a median of 26 months. COMMENT Two patients are still undergoing treatment. Only 2 patients have been treated for longer than 4 years. Chronic endocarditis is the most serious complication The 2 regimens are identical in terms of mortality, of infection with C burnetii, often leading to life- valve surgery, and tolerance (Table 3) but signifi- threatening conditions.1,20 Most often, Q fever endocar-

ARCH INTERN MED/ VOL 159, JAN 25, 1999 169

Downloaded From: https://jamanetwork.com/ on 09/27/2021 Table 1. Summary of Patients Treated With Combination Doxycycline and Ofloxacin*

Patient No./ Valve Valve Adverse Sex/Age, y Valve Prosthesis Surgery Delay, mo† Treatment, mo Reaction Follow-up, mo Outcome 101/M/27 A Am Am 18 12 PS 9 Relapse 70‡ PS 32 Cure 102/M/39 MV, A Am MVb 1 18 PS, CP 9 Relapse 70‡ PS, CP 32 Cure 103/M/47 MV, A MVm, Am Am 9 38 PS 27 Relapse 20 PS . . . Still being treated 104/M/46 A . . . Am 48 30 PS 9 Relapse 46‡ PS . . . Still being treated 105/F/62 MV MVb ...... 18 PS 9 Relapse 70 PS . . . Still being treated 106/M/52 A Ab Am 35 30 PS 9 Relapse 63 PS . . . Still being treated 107/M/30 A . . . Am 1 36 PS 9 Relapse 31‡ PS 38 Cure 108/M/36 A Am ...... 26 PS 23 Cure 109/M/57 A Am 3 30 PS . . . Dead 110/F/47 MV MVb MVb 1 and 54 72 PS 23 Cure 111/M/37 A . . . Am 1 55 PS 28 Cure 112/F/38 MV ...... 62 PS 15 Cure 113/F/29 MV, A MVm, Am MVm, Am 1 68 PS 5 Currently being evaluated 114/F/49 MV, A MVb ...... 60 PS 32 Cure

*A indicates aortic valve; MV, mitral valve; b, bioprosthesis; m, mechanical prosthesis; PS, photosensibilization; CP, cutaneous pigmentation; and ellipses, data not applicable. †Delay between onset of treatment and surgery. ‡Shifted to doxycycline and hydroxychloroquine treatment.

Table 2. Summary of Patients Treated With Doxycycline and Hydroxychloroquine Combination*

Patient No./ Valve Valve Adverse Sex/Age, y Valve Prosthesis Surgery Delay, mo† Treatment, mo Reaction Follow-up, mo Outcome 1/F/69 MV MVb MVb 1 and 7 18 PS 30 Cure 2/M/66 MV, A MVb ...... 12 PS 9 Relapse 24 PS 26 Cure 3/M/42 MV MVb ...... 12 PS 9 Relapse 24 PS, RAC 23 Cure 4/M/44 MV . . . MVm 2 18 PS 38 Cure 5/M/44 MV ...... 12 PS 32 Cure 6/M/50 MV, A Ab . . . 3 18 PS 12 Cure 7/F/59 MV MVb ...... 18 PS 48 Cure 8/M/51 A ...... 24 PS 12 Cure 9/M/14 A ...... 41 PS 17 Cure 10/M/32 MV, A . . . MVb 16 36 PS 13 Cure 11/M/63 MV MVm MVm 3 24 PS 18 Cure 12/F/53 A . . . Am 0 55 PS 40 Cure 13/M/53 MV, A ...... 58 PS, CP 9 Cure 14/M/56 MV MVm MVm . . . 30 PS 80 Dead 15/M/43 A . . . Am 0 48 PS 26 Cure 16/F/54 MV MVb MVb 0 48 PS 14 Cure 17/M/48 A Am ...... 27 PS 9 Cure 18/M/70 A Ab ...... 36 PS 12 Cure 19/M/40 MV . . . MVb 5 32 PS . . . Still being treated 20/M/18 Fallot Tb ...... 36 PS . . . Still being treated 21/M/72 MV . . . MVm 0 34 PS 4 Currently being evaluated

*A indicates aortic valve; M, mitral valve; T, tricuspid valve; b, bioprosthesis; m, mechanical prosthesis; PS, photosensibilization; CP, cutaneous pigmentation; RAC, retina accumulation of chloroquine; and ellipses, data not applicable. †Delay between onset of treatment and surgery.

ditis develops in patients with abnormal valves or blood therapy, and mortality rates can exceed 60%.22,23 The vessels21 and involves mitral or aortic valves. Relapses antibiotic drug of choice for treatment is still to be are frequent despite prolonged antimicrobial chemo- determined. The ␤-lactams, used either alone or in

ARCH INTERN MED/ VOL 159, JAN 25, 1999 170

Downloaded From: https://jamanetwork.com/ on 09/27/2021 Table 3. Comparison of Outcomes Between the 2 Therapeutic Regimens

Doxycycline and Doxycycline and Ofloxacin Hydroxychloroquine (n = 14) (n = 21) P * Treatment duration among cured patients, mean ± SD, mo 56 ± 18 31 ± 14 Ͻ.001 Valve surgeries, No. 10 10 .16 Deaths, No. 1 1 .77 Relapses/completed treatments, No. 7/15 2/23 .01 Relapses after 18 mo of treatment, No. 6/11 0/16 .001 Decreasing slope of anti–phase I Coxiella burnetii antibodies IgG −0.08 −0.10 r = 0.92 (NS) IgM −0.09 −0.09 r = 0.71 (Significant) IgA −0.08 −0.09 r = 0.76 (NS) Photosensibilization, No. 14 21 NA Irreversible cutaneous pigmentation, No. 1 1 .77

*NS indicates not significant; NA, not applicable.

Table 4. Mortality and Failure Associated With Use of Various Antibiotic Regimens During Q Fever Endocarditis*

2, Tetracycline or 1, Doxycycline 3, 5, Tetracycline or and Co-trimoxazole 4, Doxycycline and Patients, Doxycycline Lincomycin or Alone or Doxycycline Hydroxychloroquine Source No. Alone Clindamycin Associated and Quinolone† Sulfate† Tobin et al29 10 1/2 2/4 1/4 ...... Varma et al23 8 1/3 ... 0/5 ...... Turcketal28 16 . . . 4/14 ...... Haldane et al36 5 ...... 2/5 ...... Wilson et al22 16 9/16 ...... Ellis et al37 8 2/5 1/3 ...... Levy et al6 32 6/9 . . . 1/1 1/16 . . . Subramanya et al34 1 ...... 1/1 ...... Present study ...... 0/5‡ 1/21 Total ... 19/35 (54%) 7/21 (33%) 5/16 (31%) 1/21 (5%) 1/21 (5%)

*Data are given as number of deaths or failures/number of treated patients. Ellipses indicate data not applicable. †Columns 4 and 5: comparison with column 1, PϽ.002 (␹2); comparison with column 2, PϽ.03 (1-tailed, Fisher exact test); comparison with column 3, PϽ.05 (1-tailed, Fisher exact test). ‡Some of the patients were partially described in Levy et al.6

combination with aminoglycosides, are not effective.6,24 used alone. The antibiotic drug regimen used was tetra- Since the observation of in vitro susceptibility of cycline, 2 g, combined with co-trimoxazole (trimetho- C burnetii to oxytetracycline (Terramycin) treat- prim, 320 mg, and sulfamethoxazole, 1600 mg) daily. ment,25,26 tetracycline compounds have been the main- Some researchers32,33 reported a favorable response to stay in chronic Q fever treatment. Tetracycline, used treatment with chloramphenicol when combined with either alone or in conjunction with valve replacement, tetracycline, whereas others28,32 reported chlorampheni- has often been effective in controlling Q fever infec- col not to be effective when used either alone or com- tions.22,27,28 The antibiotic drug regimen currently used bined with tetracycline. Rifampin use had promising is tetracycline, 2 g/d for several months then 1 g/d for at results when combined with tetracycline.4,33 Subra- least 1 year. Surgery is undertaken when heart failure manya et al34 reported the complete failure of the com- occurs. However, patients often relapse, and some die bination of rifampin and co-trimoxazole to treat C bur- owing to irreversible damage of the heart valves. There- netii infection in a patient who had adverse toxic effects fore, combination therapy with tetracycline and other with tetracycline treatment. A 6-month course of doxy- antibiotic drugs has been proposed. Turck et al28 cycline plus rifampin, followed by doxycycline treat- described 16 patients with Q fever endocarditis treated ment alone, seemed to be more effective than doxycy- with tetracycline alone, lincomycin alone, or a combi- cline treatment alone.24 However, rifampin use may be nation of the 2 compounds. The regimen combining contraindicated in patients with mechanical valves who lincomycin, 2 g, and tetracycline, 1 g, daily was sug- are being treated with anticoagulant agents. gested to be most effective. Combinations of tetracy- More recently, the demonstration of in vitro sus- cline and co-trimoxazole were evaluated because co- ceptibility of C burnetii to the newly available fluoro- trimoxazole was reported23,29-31 to be effective when quinolones prompted the use of these compounds to

ARCH INTERN MED/ VOL 159, JAN 25, 1999 171

Downloaded From: https://jamanetwork.com/ on 09/27/2021 treat chronic Q fever.35 Haldane et al36 and Ellis et al37 Reprints: Didier Raoult, MD, PhD, Unite´ des Rick- carried out a compilation of the 7 reports on the anti- ettsies, CNRS-UPRESA 6020 Universite´delaMe´diterra- biotic drug efficacy in treating the disease; however, ne´e, Faculte´deMe´decine, 27 Boulevard Jean Moulin, because of the lack of follow-up and the small size of 13385 Marseille Cedex 05, France (e-mail: Didier.Raoult each series, we chose only to compare the failure and @medecine.univ-mrs.fr). death ratios within the different regimens (Table 4). Monotherapy using tetracycline or doxycycline leads REFERENCES to a mortality rate higher than 50%. Taking a regimen

associating tetracycline compounds to lincomycin or 1. Raoult D, Marrie T. Q fever. Clin Infect Dis. 1995;20:489-496. clindamycin, one third of the patients died. The asso- 2. Hackstadt T, Williams JC. Stability of the adenosine 5’-triphosphate pool in Coxi- ciation of co-trimoxazole is controversial, but it is sig- ella burnetii: influence of pH and substrate. J Bacteriol. 1981;148:419-425. nificantly less efficient than the association of either 3. Maurin M, Benoliel AM, Bongrand P, Raoult D. Phagolysosomes of Coxiella bur- quinolones or hydroxychloroquine to doxycycline netii–infected cell lines maintain an acidic pH during persistent infection. Infect Immun. 1992;60:5013-5016. (Table 4). These 2 last regimens had a low mortality 4. Raoult D, Etienne J, Massip P, et al. Q fever endocarditis in the south of France. rate of approximately 5%, which is lower than that J Infect Dis. 1987;155:570-573. observed in other prosthetic valve endocarditis. The 5. Stein A, Raoult D. Q fever endocarditis. Eur Heart J. 1995;16:19-23. mean death ratio in prosthetic valve endocarditis 6. Levy PY, Drancourt M, Etienne J, et al. Comparison of different antibiotic regi- 38 mens for therapy of 32 cases of Q fever endocarditis. Antimicrob Agents Che- caused by other bacteria is 50%. These observations mother. 1991;35:533-537. show the dramatic improvement in the prognosis of Q 7. Brouqui P, Tissot-Dupont H, Drancourt M, et al. Chronic Q fever: ninety-two cases fever endocarditis, with the death ratio being low for from France, including 27 cases without endocarditis. Arch Intern Med. 1993; the past 10 years since these new regimens were initi- 153:642-648. ated compared with the series reported by Raoult et 8. Marrie TJ. Coxiella burnetii (Q fever). In: Mandell GL, Bennett JE, Dolin R, eds. 4 Principles and Practice of Infectious Diseases. New York, NY: Churchill Living- al, in which 6 (40%) of 15 patients died. This stone Inc; 1995:1727-1734. 6 progress has previously been reported, although the 9. Maurin M, Benoliel AM, Bongrand P, Raoult D. Phagolysosomal alkalinization treatment duration, when quinolones were used, and the bactericidal effect of antibiotics: the Coxiella burnetii paradigm. J Infect remained long (between 3 years and lifetime). This Dis. 1992;166:1097-1102. protracted regimen is required because of the inhibi- 10. Raoult D, Drancourt M, Vestris G. Bactericidal effect of doxycycline associated with lysosomotropic agents on Coxiella burnetii in P388D1 cells. Antimicrob Agents tion of the bactericidal effect of tetracycline and qui- Chemother. 1990;34:1512-1514. nolone by the low pH of the host cell phagolysosome 11. De Duve C, De Barsy T, Poole B, Trouet A, Tulkens P, Van Hoof F. Lysosomo- in which C burnetii resides.9 However, it has now been tropic agents. Biochem Pharmacol. 1974;23:2495-2531. shown that the addition of chloroquine to the regimen 12. Tissot-Dupont H, Thirion X, Raoult D. Q fever serology: cutoff determination for microimmunofluorescence. Clin Diagn Lab Immunol. 1994;1:189-196. restores the bactericidal effect of doxycycline at a 13. Fournier PE, Casalta JP, Habib G, Messana T, Raoult D. Modification of the di- 9,10 1-µg/mL concentration. Hydroxychloroquine is agnostic criteria proposed by the Duke Endocarditis Service to permit improved widely used in rheumatic practice at a dose of 600 mg/ diagnosis of Q fever endocarditis. Am J Med. 1996;100:629-633. d.39 The patient’s daily dose was adjusted to maintain 14. Hache JC, Defoort S, Puech B, Bale F, Mancel E. Les proble`mes e´lectrore´ti- the hydroxychloroquine plasmatic concentration at 1 nographiques des affections des coˆnes et de la macula. Ophthalmologie. 1989; 97:279-282. µg/mL. This was achieved by administration of a daily 15. Musso D, Raoult D. Coxiella burnetii blood cultures from acute and chronic Q- dose of 150 to 800 mg. The adverse effects of this fever patients. J Clin Microbiol. 1995;33:3129-3132. compound consist of photosensitivity and retinal 16. Chaulet JF, Robet Y, Prevosto JM, Soares O. Simultaneous determination of chlo- accumulation. We assessed the condition of the retina roquine and quinine in human biological fluids by high-performance liquid chro- in all patients throughout the trial but were obliged to matography. J Chromatogr. 1993;613:303-310. 17. Brouqui P, Dumler JS, Raoult D. Immunohistologic demonstration of Coxiella discontinue treatment in only 1 patient. It is likely that burnetii in the valves of patients with Q fever endocarditis. Am J Med. 1994;97: even 5 years of treatment will be well tolerated. Doxy- 451-458. cycline, hydroxychloroquine, and ofloxacin use all 18. White SW, Besanceney C. Systemic pigmentation from tetracycline and mono- have a photohypersensitivity risk, which was experi- cycline therapy. Arch Dermatol. 1983;119:1-2. 19. Sabatier F, Dignat-George F, Mege JL, Brunet C, Raoult D, Sampol J. CD4+ T-cell enced by all our patients. Despite repeated reminders, lymphopenia in Q fever endocarditis. Clin Diagn Lab Immunol. 1997;4: the first summer was difficult for all patients; however, 89-92. most were subsequently more careful. Hypersensitivity 20. Maurin M, Raoult D. Current concepts and perspectives in Q fever treatment. In: to sunlight was therefore the major complication with Raoult D, ed. Antimicrobial Agents and Intracellular Pathogens. Boca Raton, Fla: both regimens, although 1 patient with irreversible CRC Press; 1993:153-179. 21. Raoult D. Host factors in the severity of Q fever. Ann N Y Acad Sci. 1990;590: skin pigmentation was observed in both groups. 33-38. Comparing the 2 regimens, treatment with doxy- 22. Wilson HG, Neilson GH, Galea EG, Stafford G, O’Brien MF. Q fever endocarditis cycline and hydroxychloroquine was more effective in in Queensland. Circulation. 1976;53:680-684. terms of the percentage of relapses and the treatment du- 23. Varma MPS, Adgey AAJ, Connolly JH. Chronic Q fever endocarditis. Br Heart J. ration. We think that doxycycline and ofloxacin should 1980;43:695-699. 24. Raoult D, Levy PY, Harle JR, et al. Chronic Q fever: diagnosis and follow up. Ann be prescribed for at least 4 years, whereas doxycycline N Y Acad Sci. 1990;590:51-60. and hydroxychloroquine should not be prescribed for 25. Jackson EB. Comparative efficacy of several antibiotics on experimental shorter than 11⁄2 years or longer than 4 years. rickettsial infections in embryonated eggs. Antibiot Chemother. 1951;1:231- 241. 26. Ormsbee RA, Pickens EG. Comparison by means of the complement-fixation test Accepted for publication April 23, 1998. of the relative potencies of chloramphenicol, aureomycin, and terramycin in ex- We thank Richard Birtles, PhD, for reviewing the perimental Q fever infections in embryonated eggs. J Immunol. 1951;67: manuscript. 437-448.

ARCH INTERN MED/ VOL 159, JAN 25, 1999 172

Downloaded From: https://jamanetwork.com/ on 09/27/2021 27. Hart RJ, Shaw DB. Q-fever endocarditis. BMJ. 1973;3:233. 35. Raoult D. Treatment of Q fever. Antimicrob Agents Chemother. 1993;37: 28. Turck WP, Howitt G, Turnberg LA, et al. Chronic Q fever. QJM Q J Med. 1976; 1733-1736. 45:193-217. 36. Haldane EV, Marrie TJ, Faulkner RS, et al. Endocarditis due to Q fever in Nova 29. Tobin MJ, Cahill N, Gearty G, et al. Q fever endocarditis. Am J Med. 1982;72:396-400. Scotia: experience with five patients in 1981-1982. J Infect Dis. 1983;148: 30. Freeman R, Hodson ME. Q fever endocarditis treated with trimethoprim and sul- 978-985. famethoxazole. BMJ. 1972;1:419-420. 37. Ellis ME, Smith CC, Moffat MA. Chronic or fatal Q-fever infection: a review of 31. Mattheis MS, Silvermann M, Paretsky D. Studies on the physiology of rickett- 16 patients seen in North-East Scotland (1967-80). QJM Q J Med. 1983;52: siae, IV: folic acids of Coxiella burnetii. J Bacteriol. 1963;85:37-41. 54-66. 32. Kristinsson A, Bentall HH. Medical and surgical treatment of Q fever endocardi- 38. Karchmer AW, Gibbons GW. Infections of prosthetic heart valves and vascular tis. Lancet. 1967;30:693-697. grafts. In: Bisno AL, Waldvogel FA, eds. Infections Associated With Indwelling 33. Kimbrough RC III, Ormsbee RA, Peacock M, et al. Q fever endocarditis in the Medical Devices. Washington, DC: ASM Press; 1997:213-249. United States. Ann Intern Med. 1979;91:400-402. 39. Lipsky PE. Rheumatoid arthritis. In: Isselbacher KJ, Braunwald E, Wilson JD, Mar- 34. Subramanya NI, Wright JS, Khan MA. Failure of rifampicin and co-trimoxazole tin JB, Fauci AS, Kasper DL, eds. Harrisson’s Principles of Internal Medicine. New in Q fever endocarditis. BMJ. 1982;285:343-344. York, NY: McGraw-Hill Book Co; 1994:1648-1653.

Consult the Archives to Stay Ahead in Your Specialty

Peer-reviewed, primary source journals are a physician’s best information resource. And the Archives journals, from the world’s leading publisher of medical information, are the best choice available to gain fresh insights and keep up with the latest advances. Stay at the forefront of medicine. Call toll-free 800-AMA-2350 or fax 312-464-5831 for subscription information. E-mail: [email protected] Subscribe today! P8FA3

ARCH INTERN MED/ VOL 159, JAN 25, 1999 173

Downloaded From: https://jamanetwork.com/ on 09/27/2021