KILL KINETICS OF BORRELIA BURGDORFERI
AND BACTERIAL FINDINGS IN RELATION
TO THE TREATMENT OF LYME BORRELIOSIS
V. Preac Mursic, W. Marget, U. Busch, D. Pleterski Rigler, S. Hagl
Infection. 24 (1996) 9 -16
SUMMARY: For a better understanding of the persistence of Borrelia
burgdorferi sensu lato (s.l.) after antibiotic therapy, the kinetics of
killing B. burgdorferi s.l. under amoxicillin, doxycycline, cefotaxime,
ceftriaxone, azithromycin and pencillin G were determined. The killing
effect was investigated in MKP medium and human serum during a 72 h
exposure to antibiotics. Twenty clinical isolates were used, including
ten strains of Borrelia afzelii and ten strains of Borellia garinii. The
results show that the kinetics of killing borreliae differ from antibiotic
to antibiotic. The killing rate of a given antibiotic is less dependent
on the concentration of the antibiotic than on the reaction time.
Furthermore, the data show that the strains of B. afzelii and B. garinii
have a diffenent reaction to antibiotics used in the treatment of Lyme
borreliosis and that different reactions to given antibiotics also exist
within one species. The B. garinii strains appear to be more sensitive to
antibiotics used in therapy. Furthermore, the persistence of B
burgdorferi s.l. and clinical recurrences in patients despite seemingly
adequate antibiotic treatment is described. The patients had clinical
disease with or without diagnositc antibody titers to B. burgdorferi.
INTRODUCTION: The diagnosis of Lyme borreliosis is based on clinical
asessment and the detection of specific antibodies. However,
interpretation of serological tests and results may not be
straightforward, because such antibodies may persist for months and can
indicate prior exposure to Borrelia as well as active disease. Negative
serologic results do not necessarily exclude Borrelia infection (1 - 4).
Cultivation of Borrelia from CSF, organ and joint biopsies is difficult
and relatively seldom done; the PCR method (polymerase chain reaction) has
not yet been sufficiently evaluated for routine diagnosis of human
specimens. Especially the diagnosis of late manifestations and persistence
of the disease can be problematic.
Also, the treatment of Lyme borreliosis is still associated with several
unanswered questions. One of the main problems is the establishment of
the optimal antibiotic therapy. The results of the comparative in vitro
and in vivo studies of the susceptibility of Borrelia burgdorferi sensu
lato (s.l.) demonstrate the sensitivity of borreliae to a number of
antibiotics (5 - 9). However, some pateints developed symptoms of the
disease later despite antibiotic treatment (10, 1-4, 11, 12).
In the present report the results of the kill kinetics of B. burgdorferi
s.l. as well as the persistence of borreliae after antibiotic treatment
KILL KINETICS OF BORRELIA BURGDORFERI S.L. UNDER VARIOUS ANTIBIOTICS
MATERIALS AND METHODS: Antibiotic susceptibility of B. burgdorferi s.l.
and minimal inhibitory concentration (MIC90) of 30 antibiotics have been
reported previously (7, 8).
The kill kinetics of borreliae under various antibiotics was determined in
MKP medium (13) and in undiluted volunteer serum with negative Lyme
borreliosis serological tests. The Borrelia garinii (ten strains) and
Borrelia afzelii ( ten strains) used in this study were isolated from
human spinal fluid and skin biopsy. The tubes of the MKP medium or serum
containing antibiotic in concentrations of 1, 2 and 4 mg/l were inoculated
with 10,000,000 cells/ml. control tubes with 10,000,000 cells/ml but no
antibiotics were included in every run. Incubation took place at 33
degrees C. The cultures were examined for motile and non-motile borreliae
by dark-field microscopy after 6, 12, 24, 48 and 72 h of incubation and by
subcultures. subcultures in MKP medium without antibiotics were made and
observed for growth for a further 4 weeks. The agents tested included
amoxicillin, azithromycin, cefotaxime, ceftriaxone, doxycycline and
*******See original for figures*****
Figures 1-5 illustrate the data on killing rates in B. afzelii species;
Figures 6-10 on ten isolates of B. garinii species in MKP medium. The
results show that the kill kinetics of the borreliae differs from
antibiotic to antibiotic. The killing rate of a given antibiotic for
borreliae is less dependent on the concentration of the antibiotic than on
the reaction time. Furthermore, the data show that the killing effect of
isolates of B. garinii differs from that in B. afzelii species. Very
interesting and unexpected is the different effect of antibiotics on the
isolates within one species. Also the different reaciton of one strain to
tested antibiotics is surprising.
To find out the antiborrelial activity of serum and simultaneously the
effect of serum and antibiotics against Borrleia isolates, we examined the
effect of antibiotics in human serum. As shown (Figures 11, 12), the
efficacy of antibiotics in volunteer serum is not significantly better
than in MKP medium. Therefore, serum did not significantly enhance the
bacteriocidal activity of the tested antibiotics and the activity was not
reduced in the presence of serum.
In summary, the results of killing kinetics suggest that:
1. The strains of B. afzelii and B. garinii spp. react differently against
antibiotics used in the treatment of Lyme disease.
2. The different reactions of strains to antibiotics also exist within
3. There exist different effects of one antibiotic against strains tested
as well as different reactions of the strain to antibiotics tested.
4. The killing rate of a given antibiotic is dependent on reaction time
5. B. garinii strains seem to be more sensitive to antibiotic tested than
B. afzelii strains.
6. The antibiotics take a long time to become effective.
7. The different killing kinetics of B. burgdorferi sensu lato strains
can be of importance in a treatment regimen.
PERSISTENCE OF BORRELIA BURGDORFERI S.L. IN PATIENTS AFTER ANTIBIOTIC
PATIENTS AND METHODS
PATIENTS: Patients were examined between October 1992 and August 1994.
Clinical data are listed in table 1. (NOT INCLUDED - SEE ORIGINAL)
SEROLOGICAL TESTS: Antibodies to B burgdorferi s.l. in blood and CSF were
determined by indirect immunofluorescence test (IFT) and ELISA as
previously described (14).
BACTERIOLOGICAL EXAMINATION: Mitral valve excisate, joint and skin
biopsies were examined for Borrelia by darkfield microscopy and by culture
in MKP medium as previously described (13, 15). The cultures were
incubated at 33 degrees C for 5-15 weeks.
Isolates were identified with monoclonal antibodies L321F11 and L221F8 by
Western blot (16) and by pulsed-field gel elctrophoresis (PFGE) as
described by Casjens and Huang (17) and Busch et al. (18).
Susceptibility of B. burgdorferi s.l. isolates to doxycycline, cefotaxime
and ceftriaxone was evaluated with MIC90 in the MKP medium. The
suitability of MKP medium for in vitro antimicrobial susceptibility
testing for B. burgdorferi s.l. has been established previously (7,8).
RESULTS: Borrelia isolation was successful in all cases after prolonged
incubation of cultures, between 5-15 weeks in MKP medium. Using MAB
L321F11 and L221F8 the isolates were grouped according to serotype (1, 2,
3, 4, 6, 7), with PFGE they were divided into genospecies B. afzelii and
B. garinii (see cases).
A 51-year-old man was admitted to the hospital in the autumn of 1987
because of plexus neuritis. the inflammatory parameters CAP and ESR were
increased; the IgG antibody titers against B. burgdorferi s.l. in serum
were positive (1:256), while the IgM titer was negative, Western blot
showed typical B. burgdorferi s.l. bands. The patient was treated with
cefotaxime 3X2 g per day i.v. for 12 days. In February 1988, the effect
of the antibiotic therapy was controlled; antibody titers to B.
burgdorferi s.l. and inflammatory parameters were negative. In January
1993, the patient had a new attack. He claimed to have had a headache,
attacks of perspiration and pseudoradicular pain located in the region of
the right arm plexus. The antibody titers against B. burgdorferi s.l. and
Western blot were negative. A tick bite in the summer of 1992 is dubious.
Erytherma migrans was not seen. In September 1993, the patient had a
relapse with progressive cardiac pain and dyspnoea on exertion.
Angiograpy and echocardiography revealed a 3rd degree mitral
insufficiency. Furthermore, a break of tendon filament of the papillary
muscle was diagnosed. The patient had a history of 7 years of
cardiopathy. In November 1993, he was admitted to university hospital
where a replacement of the mitral valve was carried out. Intraoperatively,
a heavy degenerative alteration of an already primary dysplastic mitral
valve with defective substance and multiple tendon filament breaks on the
anterior and posterior valve velum were found. The mitral valve excisate
was investigated for culture isolation of borreliae due to his Lyme
history (1987 Lyme IgG +, WB +). Borrelia was cultured from the excisate
after prolonged incubation (9 weeks) in MKp medium. IgM and IgG antibody
titers against B. burgdorferis.l. (ELISA, IFT, Immunoblot) were negative.
The patient was then treated with ceftriaxone.
A 13-year-old otherwise healthy boy was admitted to a department of
infectious disease because of intensive pain and swelling of the right
knee. He had a 1-year history of attacks of pain. A tick bite was not
remembered; an erythema migrans had never been seen. On admission,
cultures ofthe effusion revealed no growth of bacteria but increased B.
burgdorferi s.l. specific IgG and IgM(1:512 w.o. absorption) levels in
serum. After puncture he developed fever and swelling in the left knee.
The patient received ceftriaxone 2g/day for 14 days. the swelling
decreased but several days after treatment, the swelling of the right knee
and right ankle recurred; there was no pain. At that point ibuprofen was
given. The knee swelling continued for 4 months, serum Lyme IgG was 1:256
(w.o. absorption). The patient received ibuprofen for several weeks.
During a 4-month period, the intensity of his symptoms always increased
after a longer walk. When he was readmitted, 6 months after the first
admission, a severe swelling of the right knee, limited mobility and a
hydarthrosis were diagnosed. The symptoms in the left knee were less
intensive. Complete hospital records including all clinical and
laboratory investigations were obtained. The biopsy samples from the
synovia and the effusion were examined for mycobacteria and Borrelia.
There was no serological evidence of a bacterial or viral infection except
that of Lyme borreliosis (IgG 1:128 w. absorption). B. afzelii was
cultured from the synovia as well as from the effusion. At the follow-up
examination, 1 month after the synovectomy and aspiration, the swelling
and the effusion in the right knee were diagnosed. After a course of 14
days of ceftriaxone the swelling and effusion disappeared.
A second synovia examination, 6 months after the first synovectomy, showed
an exudative synovitis. Lyme serology and B. burgdorferi s.l. cultures
were negative. Within a few weeks of conservative therapy, the synovitis
was cured, whereas the limitation of flexibility persisted. The last
follow-up examination in 1993 confirms this finding.
A 17-year-old man developed recurrent episodes of a painful knee and was
treated with a corticosteroid. In early 1993, he was admitted to a
department of rheumatology because of painful swelling in the left and
right knees. The physical and laboratory examinations indicated arthritis
in both knees as well as in the right elbow. Upon physical examination
the patient was febrile. Serologic examinations for rheumatoid factor and
syphilis were negative, whereas serum Lyme IFT-IgG was borderline. The
patient was treated with ceftriaxone 2g/day for 14 days. In October
1993, he was suffering from recurrent arthritis in both knees. The IgG
antibody titers against B. burgdorferi s.l. in the serum and in synovial
fluid were positive (IFT 1:512, ELISA >24) and IgM was negative.
B. afzelii was isolated from the effusion after 5 weeks of incubation in
MKP medium. subsequently cefotaxime was given 6g/day for 14 days.
Erythema migrans was not noted, although a tick bite was recalled.
Antibiotic treatment resulted in marked reduction of complaints.
A 35-year-old man was admitted to a dermatologic clinic with a 1-year
history of headaches, intensive back pain, skin eruption and arthralgias.
On admission, a lymphocytoma benignum was suspected. The neurological
examination was normal. Serum Lyme IgG and IgM were negative. However,
nonmotile borreliae were isolated from skin biopsy. After four
subcultures in MKP medium enriched with 10 p.c. of a 35 p.c. albumin
bovine solution (Sigma, Germany), the borreliae were motile. With MAB and
PFGE, the strains were classified as B. burgdorferi serogroup 2 and
genospecies B. garinii. A tick bite was not noted. The patient was
treated with ceftriaxone 2g/day for 14 days. The back pain gradually
diminished, whereas the other symptoms persisted. The patient received
doxycycline for 10 days in February 1993. In March 1993, the patient was
readmited due to persistent arthralgias. Antibody titers against B.
burgdorferi s.l. in serum were negative, but Borrelia was isolated from a
subsequent biopsy (taken in the immediate vicinity of the prior biopsy).
This isolate remained impassive and oral genicillin G was given for 14
days. At a follow-up examination in June 1993, B. burgdorferi s.l.
serology and cultures were negative. At the last examination in December
1993 the Lyme serology was negative, and culture was not done. The
antibiotic treatment resulted in a reduction of arthralgias.
A 28-year-old woman developed recurrent episodes of pain in her knee,
hand, shoulder and the talus and was treated with corticosteroids and
doxycycline. After a 2-year history of pain and an increase of
inflammaiton in the knee and hand joints, a synovectomy was performed.
Culture from biopsies for borreliae and Lyme serological tests were done.
Lyme igM and IgG were negative; nevertheless, B. afzelii was isolated from
hand synovia, and the patient was treated with ceftriaxone.