Mycoplasma hominis meningitis in a neonate: Case report and review

Journal of Infection (2008) 57, 338e343

www.elsevierhealth.com/journals/jinf

CASE REPORT

Mycoplasma hominis meningitis in a neonate: Case report and review Atsuko Hata a,*,d, Yuiko Honda a,d, Kaoru Asada b, Yuko Sasaki c, Tsuyoshi Kenri c, Daisuke Hata a a

Department of Pediatrics, Kitano Hospital, The Tazuke Kofukai Medical Research Institute, Osaka, Japan Department of Laboratory Medicine, Kitano Hospital, The Tazuke Kofukai Medical Research Institute, Osaka, Japan c Department of Bacterial Pathogenesis and Infection Control, National Institute of Infectious Diseases, Tokyo, Japan b

Accepted 5 August 2008 Available online 14 September 2008

KEYWORDS Mycoplasma hominis; Meningitis; Minocycline; Moxifloxacin; Neonate

Summary Despite frequent colonization with Mycoplasma hominis, the invasive disease is rare in neonates. This study describes a neonatal case with meningitis in which M. hominis was isolated from a cerebrospinal fluid sample by culture and detected by PCR. The M. hominis infection was confirmed by elevated metabolic inhibition titers against the isolated M. hominis strain and anti-M. hominis antibodies in serum samples. Minocycline and moxifloxacin were effective against M. hominis, which caused meningitis in the patient. However, the patient exhibited left hemiplegia because of massive brain infarction. Based on data of the previously reported 28 cases in addition to our case, the high morbidity and mortality of the M. hominis central nervous system infection were confirmed; it was assumed to result from delayed diagnosis and ineffective initial therapy. Early diagnosis and prompt initiation of appropriate antimicrobial treatment are necessary for a favorable prognosis. Fourth-generation fluoroquinolones, especially moxifloxacin, deserve wider use in such cases. Crown Copyright ª 2008 Published by Elsevier Ltd on behalf of The British Infection Society. All rights reserved.

Introduction

* Corresponding author. Department of Pediatrics, Kitano Hospital, The Tazuke Kofukai Medical Research Institute, 2-4-20 Ohgimachi Kitaku Osaka 530-8480, Japan. Tel.: þ81 6 6312 1221; fax: þ81 6 6312 8867. E-mail address: [email protected] (A. Hata). d They equally contributed to this work.

Mycoplasma hominis frequently colonizes the human genitourinary tract. It is generally associated with localized genitourinary tract infections including pelvic inflammatory diseases, pyelonephritis, and chorioamnionitis.1 Neonates can acquire the microbe in a colonized birth canal and develop an infection. Fetal infection might also occur in utero. However, invasive infection of young infants with

0163-4453/$34 Crown Copyright ª 2008 Published by Elsevier Ltd on behalf of The British Infection Society. All rights reserved. doi:10.1016/j.jinf.2008.08.002

Neonatal M. hominis meningitis M. hominis has been described very rarely. In both full term and preterm infants, M. hominis is a rare cause of neonatal meningitis, with or without neurological deficits.2e14 We present a neonatal case of meningitis that was admitted with fever and focal seizures and which was diagnosed as M. hominis infection. Although empiric antimicrobial treatment using b-lactams for neonatal meningitis failed, the meningitis appeared to respond to treatment with minocycline and moxifloxacin, to which the isolated microbe was susceptible.

Case report A previously healthy 25-day-old female neonate was admitted to our hospital because of sudden onset of fever, vomiting, and focal seizures in the left arm. She was born weighing 2850 g after a 38-week uneventful gestation by spontaneous vaginal delivery. On admission she weighed 3200 g and her temperature was 38.1  C. She was alert but focal seizures were visible occasionally in the left arm. Her anterior fontanel was flat and nuchal rigidity was absent. Laboratory findings revealed an elevated WBC count of 23,400/mm3; CRP 1.43 mg/dl; and blood glucose 101 mg/dl. Cerebrospinal fluid (CSF) obtained by lumbar puncture contained 1880 WBC/mm3, with 68% polynuclear cells and 32% mononuclear cells, 306 mg/dl protein, and 2 mg/dl glucose. No microorganism was detectable by Gram staining. Treatment with intravenous ampicillin and cefotaxime was started as empirical antimicrobial therapy for neonatal bacterial meningitis. Intravenous immunoglobulin, dexamethasone, and mannitol were administered for 3 days. The next day (on day 2) the fever subsided and seizures ceased. A second lumbar puncture performed on day 3 yielded CSF containing 557 WBC/mm3, 315 mg/dl protein, and 18 mg/dl glucose. At that time, blood and CSF samples obtained on admission cultured for common bacterial pathogens were negative. On day 6, a third lumbar puncture was performed because of her recurrent fever of 38.0  C. The CSF contained 494 WBC/mm3, 200 mg/dl protein, and 4 mg/dl glucose. The blood WBC was 14,600/mm3 and CRP was 3.3 mg/dl. At this time, the CSF sample obtained on day 3 grew pinpoint colonies on the bloodeagar plates. Gram staining of these colonies was negative, suggestive of Mycoplasma species. Subsequently, the CSF isolates were identified by PCR as M. hominis using 16S ribosomal primers and sequence analysis.15 Because M. hominis was considered as a pathogen causing meningitis, intravenous minocycline (4 mg/kg/day) and ciprofloxacin (22 mg/kg/day) were added to the treatment regimen. Brain CT revealed a massive low-density lesion in the right middle cerebral artery territory with diffuse meningeal enhancement. Because herpes simplex virus (HSV) encephalitis was also considered, acyclovir was added. Two days later (on day 8) the fever subsided. A fourth lumbar puncture was performed; the abnormal CSF findings persisted, with 494 WBC/mm3, 267 mg/dl protein and 13 mg/dl glucose. Ampicillin and cefotaxime were replaced with chloramphenicol (200 mg/kg/ day). Acyclovir was stopped because IgM antibody against HSV was reportedly negative. On day 9, brain MR imaging

339 confirmed the CT findings and showed a stenotic lesion at the root of the right middle cerebral artery. Subsequently, the CSF findings improved: 174 WBC/mm3, 211 mg/dl protein, 16 mg/dl glucose on day 14; and 114 WBC/mm3, 172 mg/dl protein, 20 mg/dl glucose on day 17. The blood WBC was 10,000/mm3 and CRP was 0.10 mg/dl on day 17. Based on the MIC and MBC values tested for the isolated M. hominis (as described below), chloramphenicol was stopped and intravenous ciprofloxacin was replaced with oral moxifloxacin (13 mg/kg/day) on day 17. The CSF parameters were 106 WBC/mm3, 150 mg/dl protein, 27 mg/dl glucose on day 20; they were 65 WBC/ mm3, 110 mg/dl protein, 30 mg/dl glucose on day 27. The blood WBC was 4700/mm3 and CRP was <0.10 mg/dl on day 27. Minocycline and moxifloxacin were stopped on day 34. The total durations of treatments of minocycline and moxifloxacin were, respectively, 28 days and 17 days. She was discharged on day 40 of illness. After discharge, she started rehabilitative treatment because she exhibited hemiplegia in the left extremities. No adverse effects of the above antimicrobials were observed during the course. After her discharge, nested PCR analyses of the CSF samples on days 1, 8, 14, 17, 20 and 27 (PCR Mycoplasma Detection Set; Takara Bio Inc., Tokyo, Japan) revealed the presence of M. hominis DNA, except for the sample on day 27. These CSF samples cultured for the other pathogens remained negative. Serum metabolic inhibition titers against the isolated M. hominis (from the CSF sample on day 3) were <1:4 on day 1 and 1:8 on day 27. Serum titers of anti-M. hominis antibodies (enzyme immunoassay) were IgG 160, IgM 320 on day 1; they were IgG > 2560, IgM > 2560 on day 27. Recent infections with HSV, cytomegalovirus or Toxoplasma were excluded by serological examinations. These findings indicate that M. hominis was the most likely pathogen causing meningitis in this patient. The MIC values for the isolated M. hominis were <0.063 mg/ml for minocycline, 4.0 mg/dl for chloramphenicol, 0.50 mg/dl for ciprofloxacin, and <0.063 mg/ml for moxifloxacin. The MBC values were 0.25 mg/ml, 32 mg/dl, 1.0 mg/dl and <0.063 mg/ml for the respective agents. The concentrations in the serum and CSF samples were, respectively, 2.12 mg/ml, 0.57 mg/ml for minocycline (day 8); 20.8 mg/ml, 9.3 mg/ml for chloramphenicol (day 10); 4.42 mg/dl, 0.53 mg/ml for ciprofloxacin (day 8); and 0.30 mg/ml, 0.16 mg/ml for moxifloxacin (day 20). Therefore, the CSF concentrations tested were above both MIC and MBC only for minocycline and moxifloxacin. These results suggested that minocycline and moxifloxacin were effective against M. hominis, which caused meningitis in this patient.

Discussion In fact, M. hominis is a frequent colonizer of the genital tract of pregnant women. Neonates usually acquire the microbe by passage through the colonized birth canal.1 The sites of M. hominis colonization are the pharynx, respiratory tract and external auditory canals in both sexes and the external genitalia in females. Presumably, M. hominis spreads to the bloodstream, central nervous system (CNS) or both from these sites. Despite the frequent colonization with M. hominis, invasive disease is rare in neonates.

340

Table 1

Profiles of previously reported 28 neonatal patients and our patient with Mycoplasma hominis CNS infection

Ref. number/ patient number

Age at onset/ sex (M/F)

GA/BW/ delivery mode (V/C)

Blood: WBC (/mm3)/ CRP (mg/dl)

CSF: WBC (/mm3)/ protein (mg/dl)/ glucose (mg/dl)

Empiric therapy

Definite therapy

Day of illness when definite therapy initiated

Duration of antimicrobial treatment

CNS complications

Outcome

Ref. 2a/ nZ7

Day 6b/ n.d.

n.d./n.d.

233b <547>c/ 115b<700>c/ 33b<17>c

CP

CP GM

10b

31b

‘‘Spina bifida’’

Death [4]e Neurological deficits [3]e

Ref. 3/ nZ1

Day 8/M

39 (32e42) wkb/ 3100 (1800e4050) gb/ n.d. 35 wk/ 1530 g/V

19,500/ n.d.

0/316/17

PCG, KM

ABPC CP

10 10

n.d. n.d.

Right hemiplegia

Ref. 4/ nZ2

Day 8/F

33 wk/ 1740 g/V

36,000/ n.d.

19,500/70/9

CP TC

Day 15/F

‘‘Term’’/ 4020 g/V

26,100/ n.d.

4000/360/n.d.

Ref. 5/ nZ1

Day 10/M

35 wk/ 2200 g/V

n.d./n.d.

1020/175/23

Ref. 6/ nZ1

Day 12/M

30 wk/ 1080 g/V

21,000/ n.d.

160/480/7

Ref. 7/ nZ1

Day 20/M

26 wk/ 920 g/V

n.d./n.d.

73/358/16

Ref. 8/ nZ1

Day 6/M

37 wk/ 3560 g/C

n.d./n.d.

7100/270/n.d.

ABPC, CP, GM GM, MCIPC, ABPC ABPC, CBPC, GM CP, PC, CXM PC, GM, ABPC, VCM None

Brain abscess Periventricular hemorrhage None

Ref. 9/ nZ3

Day 32/M

35wk/ 2750 g/V 32wk/ 2440 g/C 37 wk/ 2950 g/C 37 wk/ 4020 g/V

n.d./n.d.

500/170/n.d.

n.d./n.d.

Day 1/M Day 3/M Day 20/M

21 21

CP OTC

32 35

3d 21

None

Normal development

DOXY LCM

19 35

n.d. n.d.

Hydrocephalus

Normal development

CP DOXY

12 15

14 10

None

Normal development

CP DOXY

22 40

11 16

Intraventricular hemorrhage Hydrocephalus

Death

DOXY

n.d.

14

Spastic paralysis

None

None

None

None

White matter atrophy Brain calcification None

6/140/n.d.

None

None

None

None

n.d./n.d.

2/100/n.d.

None

CLDM

n.d.

n.d.

24,000/ 7.4

650/175/9

ABPC, AMK, CTX

DOXY

8

25

Haemophilus influenzae sepsis and pneumonia None Left hemisphere infarction Hydrocephalus

Right hemiplegia

Normal development Death Normal development Death

A. Hata et al.

Ref. 10/ nZ1

8 9

10e100b/ n.d./n.d.

Ceph3, AMK

None

None

None

None

Death [1]e

17,500/ n.d.

32,000/ 1156/ ‘‘Undetectable’’

DOXY CLDM

14 14

42 n.d.

41 wk/ 4340 g/V 37 wk/ 2990 g/V 38 wk/ 2850 g/V

35,000/ n.d. 29,000/ 3.7/ 23,400/ 1.4

216/520/0

Brain abscess, left hemisphere infarction, hydrocephalus Ventriculitis, encephalitis Hydrocephalus

Developmental delay

Day 18/ M Day 7/ M Day 25/ F

ABPC, GM, VCM, CTX ABPC, CTRX MEPM, AMPC ABPC, CTX

Normal development ‘‘Alive’’

Right hemisphere infarction

Left hemiplegia

Day 7 (1e16)b/ n.d.

Ref. 12/ nZ1

Ref. 13/ nZ1 Ref. 14/ nZ1 Our patient

1322/240/20 1880/306/2

CP CPFX CLDM MINO CPFX CP MXF

n.d.

21

14 27 6 6 8 17

21 9 28 11 9 18

Neonatal M. hominis meningitis

n.d./n.d.

Day 21/ M

33 (31e36) wkb/1500 (1100e2300) gb/V [3]e; C [4]e ‘‘Term’’/ 3070 g/V

Ref. 11/ nZ7

CNS, central nervous system; GA, gestational age; BW, birth weight; V, vaginal delivery; C, cesarean section; CSF, cerebrospinal fluid; n.d., not described; CP, chloramphenicol; GM, gentamicin; PCG, penicillin G; KM, kanamycin; ABPC, ampicillin; TC, tetracycline; MCIPC, cloxacillin; OTC, oxytetracycline; CBPC, carbenicillin; DOXY, doxycycline; LCM, lyncomycin; CXM, cefuroxine; VCM, vancomycin; CLDM, clindamycin; AMK, amikacin; CTX, cefotaxime; Ceph3, third-generation cephalosporines; EM, erythromycin; CTRX, ceftriaxone; MEPM, meropenem; AMPC, amoxicillin; CPFX, cipfloxacin; MINO, minocycline; MXF, moxifloxacin. a All these patients had spina bifida. b Average and/or range are shown. c < >, Average of the highest WBC count and protein level and of the lowest glucose level during the course in each case. d CP was replaced with OTC after the 3-day treatment because the isolated pathogen was found to be resistant to CP. e [ ], Number of patient(s).

341

342 Here, we present the case of a female neonate with meningitis where M. hominis was isolated from the CSF by culture and PCR. The M. hominis infection was confirmed by elevation of metabolic inhibition titers against the isolated M. hominis and anti-M. hominis antibodies in the serum samples. Difficulty of obtaining positive cultures on routine bacteriological media implies the possibility that mycoplasmal CNS infection in neonates is under-diagnosed. In clinical practice, the infection is to be suspected when neonates have clinical and laboratory signs of meningitis but negative results of routine CSF cultures. The ability of this microbe to form tiny colonies on blood agar is expected to allow its isolation in patients, as was the case with our patient. Microbiological diagnostic procedures targeting mycoplasmal infection must be set up accordingly. A review of the relevant literature revealed 28 neonatal cases with M. hominis CNS infection (Table 1). We analyzed the clinical and laboratory features of the 29 patients, including ours. The age of disease onset ranges from day 1 to day 32. Twelve were males and three were females (14 unknown). Eight were born at term and seven were born preterm (14 unknown). Fifteen were born by vaginal delivery and seven were born by cesarean section (seven unknown). Of note, three patients (including ours) manifested CNS infection late in the neonatal period (later than three weeks)9,12; all were born at term by vaginal delivery. Of the total of 29 patients, 10 (34%) showed CNS complications that were associated with M. hominis CNS infection.3,5,7e10,12e14 Eight (28%) left neurological sequelae2e4,8,12 and eight (28%) died.2,7,9e11 Our patient exhibited left hemiplegia because of massive brain infarction. Such high morbidity and mortality of M. hominis CNS infection are assumed to result from delayed diagnosis and ineffective initial therapy. By contrast, two patients recovered with no treatment or with apparently inappropriate treatment.9,11 At present, therapy for neonatal M. hominis infections is indicated in infections associated with pure growth of the microbe and evidence that the disease manifestations are compatible with an infectious process rather than mere colonization. Early diagnosis and prompt initiation of appropriate treatment are necessary for favorable prognosis. Given its rarity, however, a lack of consensus pertains for the optimal treatment for neonatal M. hominis meningitis. Current recommendations are based on local clinical experience and in vitro susceptibility testing. Mycoplasma species are not susceptible to antibiotics that inhibit cell wall synthesis, such as b-lactams, because they do not contain cell walls. Therefore, penicillins and cephalosporins used as empiric therapy of neonatal meningitis are not effective against M. hominis. Actually, M. hominis is naturally resistant to macrolide antibiotics: it has been shown as susceptible to tetracyclines, chloramphenicol, lincosamides, and fluoroquinolones (FQs) from results of in vitro studies.1,16 Tetracyclines have been the first-line agents. However, tetracyclines are bacteriostatic in action and their use might be contra-indicated in young infants because of its adverse effects. FQs are newer antimicrobials that exert their effects through inhibition of DNA gyrase and have excellent broad-spectrum bactericidal activity. Although FQs are not approved for use in children, some evidence exists

A. Hata et al. that they are safe, well tolerated and effective against various bacterial infections in children.17,18 These drugs might be reserved for situations in which no other antimicrobial drugs can be used feasibly. Of the 29 patients described in Table 1, eight were treated with tetracyclines4e8,10,12; the other two (Wolthers et al.’s case14 and our case) were treated with FQs; they received these drugs after initial empiric therapy was found to be ineffective. In our patient, addition of minocycline and moxifloxacin appeared to decrease the WBC count and protein and increase in glucose in CSF, as well as decrease in WBC count and CRP in blood. These observations are consistent with the results of susceptibility testing of the M. hominis strain isolated from the patient’s CSF. To the best of our knowledge, this is the first case treated with moxifloxacin in a neonate with M. hominis meningitis without concomitant adverse effects. Moxifloxacin and gatifloxacin are fourth-generation FQs that are effective against systemic infections. They are absorbed rapidly on oral administration. Actually, M. hominis appears more susceptible to moxifloxacin and gatifloxacin than the second- (ciprofloxacin) and third-generation FQs (levofloxacin) based on in vitro susceptibility testing results. Regarding the isolated M. hominis (from the CSF sample on day 3), the MIC and MBC values for moxifloxacin, gatifloxacin, ciprofloxacin, and levofloxacin were, respectively, <0.063 mg/ml, <0.063 mg/ml; 0.25 mg/ml, <0.063 mg/ml; 0.50 mg/dl and 1.0 mg/dl; and 0.13 mg/ml, 0.5 mg/ml. Whether treatment with fourth-generation FQs is preferable to tetracyclines, chloramphenicol, lincosamides or the preceding FQs in neonatal patients with M. hominis meningitis remains to be clarified. Possible adverse effects, including gastrointestinal symptoms, hepatic and renal dysfunction, arthropathies and hematologic abnormalities, are of special concern in young subjects. It was reported very recently that M. hominis and Ureaplasma urealyticum cord blood infections are far more common than ever realized, with incidence of 23% in preterm born neonates.19 Positive cultures are associated with maternal placental inflammation, neonatal systemic inflammatory syndrome, and bronchopulmonary dysplasia. Whether preterm infants would benefit from routine culture for either or both pathogens and appropriate antimicrobial therapy is a question to be resolved through further studies. Our case illustrates the importance of suspecting M. hominis as a cause of neonatal meningitis when cultures for common bacterial pathogens remain negative and when empiric antibiotic treatments using b-lactams are ineffective. Minocycline and moxifloxacin were apparently effective against M. hominis, which caused meningitis in the patient. In light of the high morbidity and mortality associated with neonatal M. hominis CNS infection, we consider that fourth-generation FQs, especially moxifloxacin, deserve wider use in such cases.

References 1. Taylor-Robinson D. The role of mycoplasmas in pregnancy outcome. Best Pract Res Clin Obstet Gynaecol 2007;21: 425e38.

Neonatal M. hominis meningitis 2. Wealthall SR. Mycoplasma meningitis in infants with spina bifida. Dev Med Child Neurol Suppl 1975;35:117e22. 3. Siber GR, Alpert S, Smith AL, Lin JS, McCormack WM. Neonatal central nervous system infection due to Mycoplasma hominis. J Pediatr 1977;90:625e7. 4. Gewitz M, Dinwiddie R, Rees L, Volikas O, Yuille T, O’Connell B, et al. Mycoplasma hominis. A cause of neonatal meningitis. Arch Dis Child 1979;54:231e3. 5. Hjelm E, Jonsell G, Linglo ¨f T, Ma ˚rdh PA, Møller B, Sedin G. Meningitis in a newborn infant caused by Mycoplasma hominis. Acta Paediatr Scand 1980;69:415e8. 6. Kirk N, Kovar I. Mycoplasma hominis meningitis in a preterm infant. J Infect 1987;15:109e10. 7. McDonald JC, Moore DL. Mycoplasma hominis meningitis in a premature infant. Pediatr Infect Dis J 1988;7:795e8. 8. Waites KB, Rudd PT, Crouse DT, Canupp KC, Nelson KG, Ramsey C, et al. Chronic Ureaplasma urealyticum and Mycoplasma hominis infections of central nervous system in preterm infants. Lancet 1988;1:17e21. 9. Waites KB, Duffy LB, Crouse DT, Dworsky ME, Strange MJ, Nelson KG, et al. Mycoplasmal infections of cerebrospinal fluid in newborn infants from a community hospital population. Pediatr Infect Dis J 1990;9:241e5. 10. Alonso-Vega C, Wauters N, Vermeylen D, Muller MF, Serruys E. A fatal case of Mycoplasma hominis meningoencephalitis in a full-term newborn. J Clin Microbiol 1997;35:286e7. 11. Sethi S, Sharma M, Narang A, Aggrawal PB. Isolation pattern and clinical outcome of genital mycoplasma in neonates from a tertiary care neonatal unit. J Trop Pediatr 1999;45:143e5.

343 12. Rao RP, Ghanayem NS, Kaufman BA, Kehl KS, Gregg DC, Chusid MJ. Mycoplasma hominis and Ureaplasma species brain abscess in a neonate. Pediatr Infect Dis J 2002;21:1083e5. 13. Knausz M, Niederland T, Dosa E, Rozgonyi F. Meningo-encephalitis in a neonate caused by maternal Mycoplasma hominis treated successfully with chloramphenicol. J Med Microbiol 2002;51:187e8. 14. Wolthers KC, Kornelisse RF, Platenkamp GJ, Schuurman-Van Der Lem MI, van der Schee C, Hartwig NG, et al. A case of Mycoplasma hominis meningo-encephalitis in a full-term infant: rapid recovery after start of treatment with ciprofloxacin. Eur J Pediatr 2003; 162:514e6. 15. Sasaki T, Nishiyama T, Shintani M, Kenri T. Evaluation of a new method for identification of bacteria based on sequence homology of 16S rRNA gene. PDA J Pharm Sci Technol 1997;51:242e7. 16. Ngan CC, Lim T, Choo CM, Toh GL, Lim YS. Susceptibility testing of Singapore strains of Mycoplasma hominis to tetracycline, gatifloxacin, moxifloxacin, ciprofloxacin, clindamycin, and azithromycin by the Etest method. Diagn Microbiol Infect Dis 2004;48:207e10. 17. Murray TS, Bartimore RS. Pediatric uses of fluoroquinolone antibiotics. Pediatr Ann 2007;36:336e42. 18. Committee on Infectious Diseases. The use of systemic fluoroquinolones. Pediatrics 2006;118:1287e92. 19. Goldenberg RL, Andrews WW, Goepfert AR, Faye-Petersen O, Cliver SP, Carlo WA, et al. The Alabama Preterm Birth Study: umbilical cord blood Ureaplasma urealyticum and Mycoplasma hominis cultures in very preterm newborn infants. Am J Obstet Gynecol 2008;198:43e5.