close

Вход

Забыли?

вход по аккаунту

?

Treatment of lyme arthritis.

код для вставкиСкачать
ARTHRITIS & RHEUMATISM Volume 37
Number 6, June 1994, pp 878-888
0 1994, American College of Rheumatology
878
TREATMENT OF LYME ARTHRITIS
ALLEN C. STEERE, ROBERT E. LEVIN, PHILIP J. MOLLOY, ROBERT A. KALISH,
JAMES H. ABRAHAM, 111, NANCY Y. LIU, and CHRISTOPHER H. SCHMID
Objective. To test treatment regimens for Lyme
arthritis.
Methods. Patients were randomly assigned to
treatment with doxycycline or amoxicillin plus probenecid for 30 days. Patients who had persistent arthritis for
at least 3 months after treatment with oral antibiotics or
parenteral penicillin were given intravenous ceftriaxone
for 2 weeks.
Results. Eighteen of the 20 patients treated with
doxycycline and 16 of the 18 patients who completed the
amoxicillin regimen had resolution of the arthritis
within 1-3 months after study entry. However, neuroborreliosis later developed in 5 patients, 4 of whom
had received the amoxicillin regimen. Of 16 patients (2
from the oral antibiotic study and 14 additional patients)
who had persistent arthritis despite previous oral antibiotics or parenteral penicillin, none had resolution of
the arthritis within 3 months after ceftriaxone therapy.
The HLA-DR4 specificity and OspA reactivity were
associated with a lack of response.
Conclusion. Lyme arthritis can usually be treated
successfully with oral antibiotics, but patients may still
develop neuroborreliosis. Patients with certain genetic
and immune markers may have persistent arthritis
despite treatment with oral or intravenous antibiotics.
Supported by NIH grants AR-20358, AR-40576, and AR07570 and by the Eshe Fund.
Allen C. Steere, MD: New England Medical Center, Tufts
University School of Medicine, Boston, Massachusetts; Robert E.
Levin, MD: East Lyme, Connecticut; Philip J. Molloy, MD: Plymouth, Massachusetts; Robert A. Kalish, MD: New England Medical Center, Tufts University School of Medicine; James H. Abraham 111, MD: New England Medical Center, Tufts University
School of Medicine (currently at Little Rock Diagnostic Clinic,
Little Rock, Arkansas); Nancy Y.Liu, MD: New England Medical
Center, Tufts University School of Medicine (currently at University of Massachusetts Medical Center, Worcester); Christopher H.
Schmid, PhD: New England Medical Center, Tufts University
School of Medicine.
Address reprint requests to Allen C. Steere, MD, New
England Medical Center, NEMC #406, 750 Washington St., Boston, MA 02111.
Submitted for publication October 25, 1993; accepted in
revised form January 5 , 1994.
Lyme disease is caused by the tick-borne spirochete, Borreliu burgdorferi (1). The illness usually
begins with localized infection of the skin, erythema
migrans, followed days to weeks later by spread of the
spirochete to many sites, including the joints (1,2).
Months later, untreated patients often begin to have
intermittent episodes of arthritis in a few large joints,
especially the knee, that recur for a period of several
years (3). A small percentage of these patients have
continuous joint swelling for 1 year or longer; this has
been termed “chronic Lyme arthritis” (3,4). The
synovial tissue in these patients is similar to that seen
in other forms of chronic inflammatory arthritis, including rheumatoid arthritis (5,6).
In an initial antibiotic-treatment study of Lyme
arthritis carried out from 1980 to 1982,7 of 20 patients
(35%) who received intramuscular benzathine penicillin, 7.2 million units, had complete resolution of joint
involvement soon after treatment, compared with
none of 20 patients who were given placebo (P< 0.02)
(7). Of 20 patients treated the following year with
intravenous penicillin G, 20 million units a day for 10
days, 11 (55%) had complete resolution of the arthritis
soon after treatment (7). In 1983 and 1984, Eichenfield
et a1 found that all of the 25 children with Lyme
arthritis who were studied responded to treatment
with either oral phenoxymethyl penicillin, 50 mg/kg/
day, or tetracycline, 30 mglkglday, or with a combination of intravenous penicillin and oral antibiotics (8).
In 1986 and 1987, Dattwyler et a1 randomized 23
patients with Lyme arthritis, chronic neuroborreliosis,
or both to receive either intravenous penicillin, 20
million units a day for 10 days, or intravenous ceftriaxone, 2 gm twice a day for 14 days (9). Only 5 of the
10 patients responded to penicillin (50%) compared
with 12 of the 13 patients (92%) who received ceftriaxone.
In an immunogenetic study of 80 patients with
Lyme arthritis, 57% of those with chronic arthritis had
the HLA-DR4 specificity and 43% had HLA-DR2, as
TREATMENT OF LYME ARTHRITIS
determined by serologic typing methods (10). In several representative patients, the HLA-DR4 association was confirmed using nucleotide sequence typing
(10) and oligonucleotide probes (1l), but the molecular
marker for the HLA-DR2 serologic specificity is not
yet clear. Of the 80 patients, 57 (71%) had IgG antibody reactivity with outer-surface proteins A or B
(OspA or OspB) of the spirochete (12). In 15 patients
who were followed serially, this reactivity developed
near the beginning of prolonged episodes of arthritis,
from 5 months to 7 years after disease onset. The
combination of the HLA-DR4 specificity and Osp
reactivity was associated with chronic arthritis and
lack of response to antibiotic therapy. Thus, it has
been unclear whether the lack of response to therapy
in some patients is due to inadequate antibiotic treatment or to immune phenomena that may persist after
the eradication of viable spirochetes.
Doxycycline and amoxicillin are very active
against B burgdorferi in vitro (13,14), and these 2 oral
antibiotics have largely replaced tetracycline and penicillin V for the treatment of early Lyme disease
(15,16). Serum levels of amoxicillin can be boosted
2-4-fold with probenecid. Ceftriaxone is even more
active against the spirochete in vitro than is either
doxycycline or amoxicillin (13,14), and is often considered to be the drug-of-choice for the treatment of
late Lyme disease (9,17,18).
The goals of the current study were to determine whether oral doxycycline or amoxicillin plus
probenecid could be used effectively for the treatment
of Lyme arthritis, a late manifestation of Lyme disease, and to assess the effect of intravenous ceftriaxone in patients with persistent arthritis despite previous antibiotic therapy.
PATIENTS AND METHODS
Doxycycline versus amoxicillin plus probenecid protocol. In July 1986, we began a randomized study of doxycycline versus amoxicillin plus probenecid for the treatment of
Lyme arthritis. The protocol followed the guidelines of the
US Department of Health and Human Services for the
conduct of clinical research. Patients eligible for the study
were age 13 or older, had the initial attack or intermittent
episodes of arthritis in 1 or a few joints, had at least 1
actively inflamed joint at the time of study entry, and had a
positive antibody response to B burgdorferi, determined by
enzyme-linked immunosorbent assay (ELISA) (19).
Patients were excluded if they were 12 years of age
or younger, if they were pregnant or nursing, if they had
active neuroborreliosis, if they were allergic to the study
medications, or if they had already failed to respond to a
879
30-day course of oral doxycycline or amoxicillin for Lyme
arthritis. Previous treatment with other oral antibiotics more
than 3 months prior to study entry was not an exclusion
criterion.
The patients were recruited from a private practice
site in the Lyme, Connecticut area (REL), a practice location near Cape Cod, Massachusetts (PJM), the Lyme disease
clinic at Yale-New Haven Hospital (1986-1987), and the
Lyme disease clinic at New England Medical Center (19871992). There was no previous information on the efficacy of
these antibiotic regimens to allow calculations of the number
of patients required to show significant differences between
the groups. The enrollment of patients was continued until
40 evaluable patients, 20 in each antibiotic group, completed
the study, including a followup period of at least 1 year.
Twenty-five cards labeled “doxycycline” and 25
labeled “amoxicillin and probenecid” were sealed in envelopes; the cards were shuffled and distributed among the 3
study sites. After informed consent was obtained, a card was
drawn that assigned treatment with either doxycycline, 100
mg twice a day, or amoxicillin and probenecid, 500 mg of
each 4 times a day, for 30 days. Nonsteroidal antiinflammatory drugs, such as aspirin, were also taken by patients with
marked joint inflammation, but intraarticular steroid injections, which might interfere with spirochetal killing, were
not allowed during the course of antibiotic therapy. If a drug
reaction developed during therapy, the other antibiotic regimen was administered for 30 days.
Two and four weeks after starting antibiotic therapy,
patients were examined by the initial study physician. Eight
weeks later, all patients, including those in whom the arthritis had resolved, were examined in the Lyme disease clinic
at Yale-New Haven Hospital or New England Medical
Center. If marked joint swelling was still present, 1 intraarticular steroid injection was allowed. Patients with active
arthritis were seen again in the clinic at 1- to 3-month
intervals. If mild to moderate joint swelling was still present
at the next visit, the course of antibiotic therapy was
repeated with either doxycycline or amoxicillin. If marked
joint swelling was present at that time, they were invited to
participate in the ceftriaxone study.
Ceftriaxone protocol. In July 1986, we also began a
study of intravenous ceftriaxone for patients with Lyme
arthritis who had continuous joint swelling without improvement for at least 3 months after treatment with other
recommended antibiotic regimens (7,17,18), including tetracycline, doxycycline, or amoxicillin for at least 30 days,
intramuscular benzathine penicillin weekly for 3 weeks, or
intravenous penicillin G for at least 2 weeks. Eligibility
criteria were the same as those for the oral antibiotic study,
except that patients were required to have persistent arthritis
and they were not excluded if they had concomitant neuroborreliosis.
Of the 16 patients in this phase of the study, only 2
had participated in the oral antibiotic study; the remaining 14
were referred to the clinic from locations throughout the
northeastern United States after treatment by their physicians with recommended oral antibiotic or parenteral penicillin regimens. No patient who met the eligibility requirements declined participation, and their HLA specificities and
STEERE ET AL
880
Table 1. Pretreatment findings, according to antibiotic treatment group*
Age (years), median (range)
Sex (M/F), no. of patients
Clinical manifestations
History of erythema migrans
Joints affected
One knee
Both knees
Knee and other large joint
One wrist
Duration (months), median (range)
Disease onset to study entry
Cumulative duration of attacks of arthritis
prior to study entry
Prior treatment, no. of patients
Oral antibiotic therapy <30 days
Intraarticular steroids
Laboratory data at study entry
ESR (mm/hour), median (range)
Serum antibody response to all Borrelia
burgdorferi antigens, median (range)
IgM (normal <lo0 units)
IgG (normal <400 units)
IgG antibody response to outer-surface
proteins, median (range)
OspA
OspB
ospc
HLA-DR specificity
HL A-DR4
HLA-DR2
HLA-DR other
Jointfluid WBC ( X 103/mm3),median (range)
Doxyc ycline
(n = 20)
Amoxicillin plus
probenecid
(n = 20)
40 (13-72)
15/5
45.5 (14-67)
13/7
P
0.60
0.37
4
7
0.24
13
3
3
12
4
4
0
0.50
11 (0-110)
3 (0-11)
1.oo
0.46
1
12.5 (1-51)
2 (0.5-21)
3
3
15 (3-65)
5
5
33 (0-72)
0 (WOO)
0 (0-100)
6,400 (800-5 1,200) 6,400 (1,600-25,600)
150 (0-25,600)
300 (0-12,800)
100 (0-25,600)
2
7
10
12 (1040)
50 (0-25,600)
50 (@-6,400)
100 (0-3,200)
8
6
5
30 (7-88)
0.50
0.50
0.50
0.35
0.35
0.40
0.80
0.33
0.39
0.34
0.36
0.03
0.49
0.10
0.32
* HLA typing was not done in 1 patient in each group. The 3 patients who had both HLA-DR4 and
HLA-DR2 were included only in the HLA-DR4 group. Joint fluid analysis was done on 5 patients in
the doxycycline group and 9 patients in the amoxicillin and probenecid group. ESR = erythrocyte
sedimentation rate; WBC = white blood cells.
antibody responses to outer surface proteins were not
known prior to referral.
The 16 patients were treated, usually at home, with
intravenous ceftriaxone, 2 gm once a day, for 2 weeks. The
protocol regarding other antiinflammatory medications and
followup examinations was the same as in the oral antibiotic
study. If the patient still had active synovitis 3 months after
the beginning of treatment with ceftriaxone, a decision was
made about further therapy based upon the severity of joint
swelling, the amount of previous antibiotic therapy, and the
patient’s preference. Options included treatment with antiinflammatory agents alone, ceftriaxone for 1 month, or
arthroscopic synovectomy.
If possible, patients were seen in the clinic 1 year
after study entry. They were asked to call us if their arthritis
flared again or if other manifestations of the illness developed. All patients were contacted by telephone in 1992 to
confirm that they had been well since their last checkup.
Laboratory tests. Blood tests included determination
of the erythrocyte sedimentation rate (Westergren method),
the IgM and IgG antibody titers to B burgdorferi (by ELISA)
(19), and the IgG antibody responses to the Osp A, B, and C
proteins of the spirochete (by ELISA using recombinant
forms of these proteins) (12). Serial samples from a given
patient were tested on a single plate to assess the change in
titer. If possible, joint fluid was analyzed for the white blood
cell count. In the first 13 patients, immunogenetic profiles
were determined by Dr. Robert Winchester, using serologic
methods; these results have been published previously (10).
In the remaining patients, HLA-DR specificities were determined in the Transplant Typing Laboratory at New England
Medical Center, using Terasaki kits.
Statistical analysis. In the randomized trial of oral
antibiotic therapy, an intent-to-treat analysis was done in the
patients in whom the date of resolution of arthritis could be
determined. Categorical factors were compared by Fisher’s
exact test, and continuous factors were compared by
Wilcoxon rank sum test. The frequency of subsequent
neuroborreliosis in each antibiotic group was compared by
log-rank survival test. Predictors of the odds of nonresponse
881
TREATMENT OF LYME ARTHRITIS
addfflonai
-No
sntlbiotka
Retrertmentwith
amoxiciilin after
3 months
1
Doxycyciine
c
18
20
3 more
month
I
2month
4
12 more
month0
1
Enter d r i a x o n e
-study
after
3 monthr
1
,-(iG&Turopathy
No addfflonai
antlbiotlcr
Amoxiciliin/
probenecid
20
3 monthr
2)
9
16
Retreatment with
-moxiciiiin after
C-mkrl-dh"
1
Enter driaxone
-study
after
3 monthr
1
Figure 1. Treatment decisions and outcome of therapy in the 40 patients who participated in the study of
doxycycline versus amoxicillin plus probenecid for the treatment of Lyme arthritis.
were estimated by multiple logistic regression, and predictors of duration of arthritis after treatment were analyzed by
multiple linear regression. The logistic regression model was
evaluated by the area under the receiver operating characteristic (ROC) curve and by the Hosmer-Lemeshow
goodness-of-fit test (20,21). All P values are 2-tailed.
RESULTS
Response to treatment with doxycycline versus
amoxicillin plus probenecid. From July 1986 through
December 1991, 25 patients with Lyme arthritis were
assigned to treatment with doxycycline and 23 were
given amoxicillin and probenecid. Eight patients (5
treated with doxycycline and 3 given amoxicillin) were
excluded from the final analysis. In 4 of these patients,
the presence of degenerative arthritis in the knees
made it difficult to determine when resolution of Lyme
arthritis occurred. The other 4 patients did not complete the followup examinations, but according to
telephone reports, their arthritis resolved within 1 to 3
months after study entry. Of the 40 remaining patients,
4 were entered into the study in 1986, 12 in 1987, 13 in
1988, 4 in 1989, 5 in 1990, and 2 in 1991.
The patients in both groups had similar pretreatment findings (Table 1). They were most commonly young adult men. In most patients, the disease
presented with the sudden onset of swelling in one or
both knees; only a minority had a history of erythema
migrans. Their initial illness had begun a median of 1
year before study entry, and during this period, they
had a median cumulative duration of episodes of
arthritis of 2 or 3 months. Several patients in each
group had received short courses of oral antibiotic
therapy (<30 days) or intraarticular steroid injections
more than 3 months prior to study entry. Eight patients in the amoxicillin plus probenecid group had the
HLA-DR4 specificity compared with 2 in the doxycycline group (P = 0.03), but patients in the doxycycline
group had a higher median level of reactivity with the
OspA and OspB proteins of the spirochete.
Between 4 and 21 days after therapy was begun,
882
STEERE ET AL
n
--------
rl
n n
6
12
18
24
30
36
42
48
No. of Months from Treatment to Resolution of Arthritis
Figure 2. Duration of arthritis after treatment with doxycycline or amoxicillin plus probenecid. The results were
nearly identical with both treatment regimens.
a drug eruption developed in 4 of the patients in the
amoxicillin plus probenecid group, and 3 other patients in this group had gastrointestinal side effects. No
one who was given doxycycline had antibiotic side
effects. Of the 4 patients with a drug eruption, 2
required discontinuation of the antibiotic after 4 or 12
days of therapy. They then received doxycycline for
30 days.
Of the 20 patients in the doxycycline group, 18
(90%) had complete resolution of the arthritis within 1
to 3 months after study entry (Figure 1). Similarly, of
the 18 patients in the amoxicillin plus probenecid
group who completed the course of therapy, 16 (89%)
had complete resolution of joint swelling within this
period. In the 2 patients with drug eruptions whose
therapy was changed to doxycycline, the arthritis
resolved within 1 month.
The remaining 4 patients, 2 in each group, had
persistent arthritis for 3 months or more after treatment. In 2 of the 4 patients, joint swelling had improved by the 3-month followup visit; they were
retreated with oral amoxicillin, but 3-12 more months
Were required for the complete resolution of arthritis.
The other 2 patients, both of whom received intraarticular steroids after oral therapy, still had marked
synovitis at the 3-month followup visit; they were then
entered into the ceftriaxone study (see below). However, despite multiple courses of antibiotic therapy
and, in 1 patient, arthroscopic synovectomy, 6 months
to 4 years were required for the complete resolution of
arthritis. Thus, in most patients, the arthritis resolved
within a few months after treatment, but in a few
patients, many months were required for the resolution of arthritis; the duration of arthritis after treatment was similar in both antibiotic treatment groups
( P = 0.39, by Wilcoxon 2-sample test) (Figure 2).
During the subsequent followup period of 1-6.5
years (mean 3.3 years), 2 of the 40 patients reported 1
brief recurrence of joint swelling 9 and 12 months after
study entry (Figure 1). When they were examined,
joint effusions were not present, and they were not
retreated. Neuroborreliosis developed later in 5 patients, 4 of whom had received the amoxicillin regimen
( P = 0.15, by log-rank survival test). In 3 of these 5
patients, subtle sensory symptoms became apparent
weeks to months after study entry, and electromyograms showed a diffuse polyneuropathy. In the other 2
patients, subtle memory impairment became apparent
2 and 3 years after study entry; cerebrospinal fluid
showed elevated total protein and production of intra-
883
TREATMENT OF LYME ARTHRITIS
thecal antibody to the spirochete. In all 5 patients,
arthritis was no longer present when symptoms of
neuroborreliosis became apparent. Neurologic involvement in these 5 patients was treated successfully
with a 30-day course of ceftriaxone.
In conclusion, 16 of the 20 patients (80%)
treated with doxycycline had resolution of the arthritis
within 1-3 months after study entry and had no later
manifestations of the illness, compared with 11 of the
20 patients (55%) given amoxicillin and probenecid
(relative risk = 1.45, 95% confidence interval 0.922.29).
Response to treatment with ceftriaxone. From
September 1986 through April 1990, 2 patients from
the oral antibiotic study and 14 additional patients who
had had continuous joint swelling without improvement for at least 3 months after treatment with oral
antibiotics or parenteral penicillin were given intravenous ceftriaxone, 2 gm once a day for 2 weeks. In
these 16 patients, the median cumulative duration of
episodes of arthritis prior to study entry ranged from 3
to 28 months (Table 2). During that period, 8 of the
patients had received doxycycline (or tetracycline) for
1 month or more, 3 had been given amoxicillin for this
duration, 3 had been treated with intravenous penicillin for 2 or 3 weeks, and 2 had received one of these
oral regimens and parenteral penicillin. Five had been
treated with intraarticular steroids from 3 months to 3
years prior to study entry. The majority of the 16
patients had the HLA-DR4 specificity and a high level
of IgG reactivity with the OspA protein of the spirochete (Table 2). Ceftriaxone was well tolerated, except
for diarrhea in 3 patients and mildly elevated liver
function test results in 1 patient.
None of the 16 patients had resolution of arthritis within 1 to 3 months after treatment with ceftriaxone (Figure 3). In 7 patients, 1 of whom was given an
intraarticular steroid injection 2 months after ceftriaxone therapy, joint swelling resolved within 4-12
months without further antibiotic therapy. Three
months or more after study entry, another 7 patients
were given intravenous ceftriaxone, 2 gm once a day,
this time for 1 month, followed months later in 3 cases
by intraarticular steroid injections. In 3 of these 7
patients, arthritis resolved within the next 3-36
months without further antibiotic therapy. The remaining 4 patients, who received 1-month courses of
ceftriaxone, and 2 patients, who were given 2-week
courses of this medication, subsequently had arthroscopic synovectomies. Of these 6 patients, 5 had
resolution of joint swelling within months after the
Table 2. Pretreatment
ceftriaxone*
findings
in
patients
treated
with
Ceftriaxone
(n = 16)
Age (years), median (range)
Sex (M/F), no. of patients
Clinical manifestations
History of erythema migrans
One knee affected
Both knees affected
Duration (months), median (range)
Disease onset to study entry
Disease onset to first antibiotic treatment
Cumulative duration of attacks of arthritis
prior to study entry
Prior treatment, no. of patients
Antibiotic therapy for arthritis
Intraarticular steroids
Laboratory data at study entry
ESR (mndhour), median (range)
Serum antibody response to all Borrelia
burgdorferi antigens, median (range)
IgM (normal < 100 units)
IgG (normal <400 units)
IgG antibody response to outer-surface
proteins, median (range)
OspA
OspB
ospc
HLA-DR specificity
HLA-DR4
HLA-DR2
HLA-DR other
Joint fluid WBC ( X 103/mm3),median
(range)
34 (1 1-59)
11/5
6
13
3
14 (4-96)
9.5 (3-86)
5.5 (3-28)
16
5
22 (2-65)
0 (M00)
6,400 (800-51,200)
800 (04,400)
400 (M,400)
100 (1004,400)
10
3
L
14 (8-31)
* HLA typing was not done in 1 patient. Joint fluid analysis was
done on 6 patients. ESR = erythrocyte sedimentation rate; WBC =
white blood cells.
procedure, but 1 had recurrent synovitis for another 2
years.
During the subsequent followup period of 1.86.4 years (mean 3.5 years), 2 of the 16 patients
reported brief recurrences of joint swelling after persistent episodes of arthritis had resolved. In both
cases, the recurrent effusions were not observed by a
physician, and they resolved without further treatment.
Lyme encephalopathy accompanied by symptoms of fibromyalgia developed later in 1 patient who
received a 2-week course of ceftriaxone (see case
report in ref. 22). After retreatment with a 1-month
course of ceftriaxone, results of his cerebrospinal fluid
analysis were normal, but the symptoms of fibromyalgia persisted.
Comparison of responders and nonresponders to
antibiotic therapy. In an effort to identify prognostic
factors, the pretreatment findings in the 36 patients
STEERE ET AL
884
Study
Entry
I
Additional
Therapy
cl(bluCOn,
Ibr 2 waokr
18
Resolution Subsequent
of Arthritis Complications
t
Figure 3. Treatment decisions and outcome of therapy in the 16 patients who participated in the study of
intravenous ceftriaxone for the treatment of persistent Lyme arthritis.
who responded to treatment within 3 months, “responders,” were compared with those of the 18 patients who had persistent arthritis for more than 3
months after treatment, “nonresponders.” The responder group included 18 patients who received
doxycycline, 16 who were given amoxicillin and probenecid, and 2 whose therapy was changed from
amoxicillin to doxycycline because of a drug reaction.
The nonresponder group consisted of the 16 patients
who received intravenous ceftriaxone, 2 of whom
were initially in the oral antibiotic study, and 2 who
were only treated with the oral regimens.
Compared with the responders, the nonresponders had a greater cumulative duration of episodes of arthritis (but not duration of disease)prior to
treatment (P = 0.002). They also had a greater frequency of the HLA-DR4 specificity (P= 0.003) and a
higher median level of IgG reactivity with the OspA
protein of B burgdorferi (P = 0.02) (Table 3). None of
the other pretreatment findings was associated with
persistent arthritis after antibiotic therapy. Furthermore, there was no association between subsequent
neuroborreliosis and HLA-DR specificity or Osp reactivity.
To determine the combined effect of the significant risk factors, a multiple logistic regression model
was developed using data from the 51 patients from
whom HLA-DR data were available. The odds of
nonresponse were 1.3 times greater for each month of
arthritis prior to treatment, and 22 and 23 times
greater, respectively, for patients who had the HLADR4 specificity or OspA reactivity. The ROC area for
this model is 0.92, and model calibration shows good
agreement between the observed and predicted risk
(P = 0.79, by Hosmer-Lemeshow test).
Patients with both the HLA-DR4 and OspA
markers had a predicted risk of nonresponse of 2276% as the pretreatment duration of arthritis increased
from 1 to 10 months. Those who had either HLA-DR4
or OspA reactivity had a risk of 1-12%, and those with
neither of these markers had a risk of <1% (Table 4).
When these factors were analyzed by multiple linear
regression, they were also significant predictors of the
duration of arthritis after treatment, but 66% of the
variation in the posttreatment duration of arthritis was
not explained by these factors (model R2 = 34%).
Thus, the 3 significant factors distinguished responders from nonresponders, but they did not predict the
duration of nonresponse.
By the 1-year followup visit, the median IgG
antibody response to B burgdorferi was 4-fold lower in
both responders and nonresponders. Of the 3 patients
TREATMENT OF LYME ARTHRITIS
885
Table 3. Pretreatment findings, according to treatment outcome*
Responders
(n = 36)
Age (years), median (range)
Sex (M/F), no. of patients
Clinical manifestations (%)
History of erythema migrans
One joint affected
More than one joint affected
Duration (months), median (range)
Disease onset to study entry
Cumulative duration of attacks of arthritis
prior to study entry
Prior treatment, no. of patients (%)
Intraarticular steroids
Laboratory data at study entry
ESR (mm/hour), median (range)
Serum antibody response to all Borrelia
burgdorferi antigens, median (range)
IgM (normal <I00 units)
IgG (normal <400 units)
IgG antibody response to outer-surface
proteins, median (range)
OspA
OspB
ospc
HLA-DR specificity
HLA-DR4
HLA-DR2
HLA-DR other
Joint fluid WBC ( X 103/mm3),median
(range)
IgG antibody response to all B
burgdorferi antigens 1 year after
treatment, median (range)
Nonresponders
(n = 18)
P
43 (13-72)
26/10
33 (1 1-59)
1216
0.16
0.45
11 (31)
24 (67)
12 (33)
6 (33)
14 (77)
4 (23)
0.53
0.30
0.30
12 (0.25-110)
3 (0-12)
4
20 ( 0 4 7 )
0 (0-8W
6,400 (800-51,200)
50 (0-25,600)
50 (04,400)
100 (0-25,600)
7 (19)
12 (33)
I5 (42)
27 (6.6-52)
1,600 (400-12,800)
12.5 (4-96)
5 (1.5-28)
0.78
0.002
5 (27)
0.12
22 (2-65)
1 .o
0 (0-100)
12,800 (3,200-5 1,200)
1,200 (0-25,600)
400 (0-12,800)
100 (04,400)
12 (67)
4 (22)
I (5)
22.5 (8-88)
3,200 (800-12,800)
0.40
0.05
0.02
0.13
0.92
0.003
0.30
0.005
0.39
0.01
* The total number of patients in this analysis was 54 rather than 56 because 2 patients participated in
both the oral and intravenous antibiotic studies. HLA-DR typing was not done on 3 of the 54 patients.
ESR = erythrocyte sedimentation rate; WBC = white blood cells.
who subsequently developed Lyme encephalopathy, 2
had a 4-fold decline in their specific IgG antibody
response by the I-year followup, and this response
Table 4. Predicted percentage of patients not responding to treatment, according to different combinations of 3 significant pretreatment variables*
Cumulative duration of
attacks of arthritis before
treatment (months)
HLA-DR4 and OspA
HLA-DR4 only
OspA only
Neither HLA-DR4 nor OspA
DISCUSSION
I
3
5
10
1
0.05
32
2
2
0.1
45
3
4
0.2
76
12
12
0.6
Marker
22
1
remained low when they later had symptoms of neuroborreliosis. The other patient with this complication
had a 2-fold decline in his IgG antibody response by
the 3-month followup visit and a 2-fold increase by the
1-year followup. No other study patient had a decline
followed by an increase in the specific antibody response.
~~
* To calibrate these probabilities with the risks expected in the
community, the regression constant was adjusted so that the average predicted rate of not responding to treatment was lo%, which
was the frequency of nonresponse in the oral antibiotic study in
which patients were recruited primarily from the community.
In the current study, most of the patients with
Lyme arthritis had gradual resolution of joint swelling
within 1 to 3 months after therapy with either doxycycline or amoxicillin plus probenecid. Because Lyme
arthritis may resolve slowly, no additional antibiotic
therapy was allowed for 3 months before or after study
entry. If patients who still had joint swelling after 1
month of oral therapy had been given intravenous
therapy at that time, it might have been concluded
886
incorrectly that the arthritis resolved because of the
intravenous treatment.
Of the 2 oral regimens tested, therapy with
amoxicillin and probenecid was more problematic.
First, both antibiotic regimens were equally effective
in the treatment of the arthritis, but subsequent neuroborreliosis occurred primarily in the amoxicillin
group. Only small amounts of both drugs cross the
blood-brain barrier, but doxycycline penetrates cells,
is more lipid-soluble, and has been used successfully
to treat early neuroborreliosis (23,24). Second, amoxicillin and probenecid must be taken more frequently,
which makes compliance more difficult. Although this
may have been a factor in the greater frequency of
neuroborreliosis in the amoxicillin group, the fact that
both drugs were equally effective in the treatment of
the arthritis suggests that patients in the amoxicillin
group did take the medication regularly. Finally, the
combination of these 2 drugs seems to increase the
frequency of drug eruption. Twenty percent of the
current patients had this complication, as did 3% and
37% of patients given this regimen in 2 other studies
(14,15). Since the number of patients in this study was
too small to show a significant difference between the
doxycycline and amoxicillin groups, it is unclear
whether this difference is real. It is unlikely that a
larger study could be done to answer this question.
Lyme arthritis is becoming an uncommon illness due
to the successful treatment of early manifestations of
the disorder. The important point is that Lyme arthritis can usually be treated successfully with either of
these 2 commonly used and inexpensive oral antibiotic
agents.
The major drawback of oral antibiotic therapy
is that patients with Lyme arthritis may have latent or
active neuroborreliosis, which may be inadequately
treated with oral antibiotics. For that matter, this
complication also occurred in 1 patient in the current
study who was treated with intravenous ceftriaxone
for 2 weeks. Should all patients with Lyme arthritis
have neurologic testing, including lumbar puncture, or
would it be advisable to treat all patients with Lyrne
arthritis with 30-day courses of intravenous ceftriaxone? We think that neither lumbar puncture nor ceftriaxone therapy is routinely indicated since neurologic abnormalities develop later in only a small
percentage of patients with Lyme arthritis. Furthermore, compared with oral treatment, intravenous therapy is more difficult, the frequency of side effects is
greater, and the cost is considerably more. A 1-month
course of doxycycline or amoxicillin costs about 60
STEERE ET AL
dollars compared with approximately 3,000 dollars or
more for a 1-month course of intravenous ceftriaxone.
However, the physician must be alert to the
presence of subtle neurologic symptoms. In the oral
antibiotic study, all 5 patients in whom neurologic
involvement subsequently became apparent were
treated in the early years of the study before the late
neurologic manifestations of Lyme disease were well
described. In retrospect, all 5 told of subtle peripheral
dysesthesias or memory impairment at the time of
study entry. We now recognize dysesthesia as an
important clue to the diagnosis of polyneuropathy
(25,26), and electromyography should be performed
on such patients. Memory impairment is the most
common symptom of encephalopathy (27,28), and
patients with this symptom should have cerebrospinal
fluid analyses. The physician faces a similar problem
in patients with early Lyme disease. In a previous
study, 2 of 30 patients (7%) with early disseminated
infection who were treated with doxycycline or amoxicillin plus probenecid subsequently developed neurologic involvement (16) compared with 5 of 40 patients
( 13%) in the current study. The neurologic abnormalities in all of these patients were effectively treated
with 2- to 4-week courses of intravenous ceftriaxone.
A major question has been whether ceftriaxone
therapy would solve the problem of apparent treatment failure in Lyme arthritis. In this study, the 16
patients who had no response to oral antibiotics or
parenteral penicillin still had persistent arthritis for at
least 3 months after treatment with intravenous ceftriaxone. Was this outcome influenced by the fact that
some of these patients were treated with intraarticular
steroids prior to study entry or after antibiotic therapy? We doubt that this would be the sole explanation
since only a minority of the patients received intraarticular steroids, none received them during the
course of antibiotic therapy, and the frequency of prior
treatment with intraarticular steroids in the responders
and nonresponders did not differ significantly.
In our opinion, the lack of response during
multiple courses of antibiotics suggests that Lyme
arthritis may sometimes persist for months or even
several years after the eradication of viable spirochetes. This concept is supported by the results of a
recent study in which synovial fluid samples collected
over a 17-year period were tested for the presence of
B burgdorferi DNA by polymerase chain reaction
(PCR) (29). Of 73 untreated patients, 70 had positive
test results. However, of the 10 patients tested who
had chronic arthritis despite multiple courses of anti-
TREATMENT OF LYME ARTHRITIS
biotic therapy, 7 had negative test results in all posttreatment samples and no patient had a positive result
after more than 2 months of oral antibiotics or 3 weeks
of intravenous treatment. The course of the arthritis in
these patients may not be different from the natural
history of the illness, since persistent Lyme arthritis
usually resolves within several years, even in untreated patients (3).
This study further supports the concept that
chronic or treatment-resistant Lyme arthritis has an
immunogenetic basis. The presence of either the
HLA-DR4 specificity or OspA reactivity increased the
odds of nonresponse, and the presence of both of these
markers was associated with the greatest risk of persistent arthritis. In inbred strains of mice, an association has also been noted between chronic arthritis and
certain H-2 haplotypes, the equivalent of the human
HLA-DR allele (31). However, in the current study,
the duration of arthritis after treatment was only
partially explained by these risk factors. Other unidentified factors are surely important. For example, in a
recent study of 83 patients with Lyme arthritis, those
with high concentrations of interleukin-1 (IL-1) receptor antagonist and low concentrations of IL-Ip in
synovial fluid had rapid resolution of attacks of arthritis, whereas patients with the reverse pattern of cytokine concentrations had long intervals to recovery
(32). Too few joint fluid samples were available from
patients in the current study to correlate cytokine
levels with duration of arthritis.
RECOMMENDATIONS
We recommend treatment with oral antibiotics
in patients with Lyme arthritis who do not have
concomitant neurologic involvement. In patients who
do not respond, it is perhaps reasonable to treat once
again with oral antibiotics or with a 2- to 4-week
course of ceftriaxone since a 1-month course of oral
therapy may not always eradicate viable spirochetes.
After treatment, the antibody response to B burgdorferi declines slowly and is rarely helpful in assessing
the adequacy of treatment. Even an apparent increase
in titer must be interpreted with caution unless the
samples are tested together on the same plate. During
the study, decisions about retreatment were made
empirically, but PCR testing of synovial fluid, if done
in a reliable laboratory, may help to guide these
decisions.
If patients have persistent arthritis despite this
therapy and if the results of PCR testing are negative,
we prescribe antiinflammatory agents. Intraarticular
steroids seem to be the medication that is most likely
to provide relief of knee swelling, but we do not
continue this approach if after 1 or 2 injections patients
still develop recurrent effusions. Nonsteroidal antiinflammatory agents, including hydroxychloroquine
(33), may be of some benefit. If persistent joint swelling is quite painful or if it limits function, arthroscopic
synovectomy may reduce the period of joint swelling
(34). Lyme arthritis seems to be particularly amenable
to this approach since the arthritis is often localized to
the knees and since the disease itself may remit by the
time the synovium grows back.
ACKNOWLEDGMENTS
We are indebted to Drs. Hal Dinerman, Robert
Schoen, and Elena Massarotti for help with patient care; to
Ms Elise Taylor for long-term followup of patients; to Ms
Gail McHugh, Ms Leslie Barber, Ms Jennifer Whelan, and
Mr. Bruce Reinhardt for Lyme disease serologic testing of
patient samples; and to Dr. Robert Winchester and Ms Joan
Feld, at that time at the Hospital for Joint Diseases in New
York, and to Dr. Arthur Rabson and Ms Rose Tavares at the
Tissue Typing Laboratory at New England Medical Center,
for HLA-DR typing of patients.
REFERENCES
1. Steere AC: Lyme disease. N Engl J Med 321:586596, 1989
2. Steere AC, Bartenhagen NH, Craft JE, Hutchinson GJ, Newman JH, Rahn DW, Sigal LH, Spieler PH, S t e m KS, Malawista
SE: The early clinical manifestations of Lyme disease. Ann
Intern Med 99:76-82, 1983
3. Steere AC, Schoen RT, Taylor E: The clinical evolution of
Lyme arthritis. Ann Intern Med 107:725-731, 1987
4. Steere AC, Gibofsky A, Patarroyo ME, Winchester RJ, Hardin
JA, Malawista SE: Chronic Lyme arthritis: clinical and immunogenetic differentiation from rheumatoid arthritis. Ann Intern
Med 90:89&901, 1979
5. Johnston YE, Duray PH, Steere AC, Kashgarian M, Buza J,
Malawista SE, Askenase PW: Lyme arthritis: spirochetes found
in synovial microangiopathic lesions. Am J Pathol 118:26-34,
1985
6. Steere AC, Duray PH, Butcher EC: Spirochetal antigens and
lymphoid cell surface markers in Lyme synovitis: comparison
with rheumatoid synovium and tonsillar lymphoid tissue. Arthritis Rheum 31:487495, 1988
7. Steere AC, Green J, Schoen RT, Taylor E, Hutchinson GJ,
Rahn DW, Malawista SE: Successful parenteral penicillin therapy of established Lyme arthritis. N Engl J Med 3122369-874,
1985
8. Eichenfield AH, Goldsmith DP, Benach JL, Ross AH, Loeb
FX, Doughty RA, Athreya BH: Childhood Lyme arthritis:
experience in an endemic area. J Pediatr 109:753-758, 1986
9. Dattwyler RJ,Halperin JJ, Volkman DJ, Luft BJ: Treatment of
late Lyme borreliosis: randomized comparison of ceftriaxone
and penicillin. Lancet 1:1191-1194, 1988
10. Steere AC, Dwyer E, Winchester R: Association of chronic
STEERE ET AL
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
Lyme arthritis with HLA-DR4 and HLA-DR2 alleles. N Engl J
Med 323:219-223, 1990
Ruberti G, Begovich AB, Steere AC, Klitz W, Erlich HA,
Fathman GC: Molecular analysis of the role of HLA class I1
genes DRBI, DQBI, and DPBl in susceptibility to Lyme
arthritis. Hum Immunol 31:20-27, 1 9 9 1
Kalish RA, Leong JM, Steere AC: Association of treatmentresistant chronic Lyme arthritis with HLA-DR4 and antibody
reactivity to OspA and OspB of Borrelia burgdorferi. Infect
Immun 61:277&2279, 1993
Johnson RC: Isolation techniques for spirochetes and their
sensitivities to antibiotics in vitro and in vivo. Rev Infect Dis 11
(SUPPI 6):1505-1510, 1989
Mursic VP, Wilske B, Schierz G, Holmburger M, Sub E: In
vitro and in vivo susceptibility of Borrelia burgdorferi. Eur J
Clin Microbiol 6:424426, 1987
Dattwyler RJ, Volkman DJ, Conaty SM, Platkin SP, Luft BJ:
Amoxycillin plus probenecid versus doxycycline for treatment
of erythema migrans borreliosis. Lancet 2: 1404-1406, 1990
Massarotti EM, Luger SW, Rahn DW, Messner RP, Wong JB,
Johnson RC, Steere AC: Treatment of early Lyme disease. Am
J Med 92:39#03, 1992
Rahn DW, Malawista SE: Lyme disease: recommendations for
diagnosis and treatment. Ann Intern Med 114:472481, 1991
Treatment of Lyme disease. Med Lett 34:95-97, 1992
Dressler F, Whalen JA, Reinhardt BN, Steere AC: Western
blotting in the serodiagnosis of Lyme disease. J Infect Dis
167:392400, 1993
Hanley JA, McNeil BJ: The meaning and use of the area under
a receiver operating characteristic (ROC) curve. Radiology
143~29-36, 1982
Hosmer DW, Lemeshow S. Applied Logistic Regression. New
York, John Wiley, 1989
Dinerman H, Steere AC: Lyme disease associated with fibrornyalgia. Ann Intern Med 117:281-285, 1992
Dotevall L, Alestig K, Hanner P, Norkrans G, Hagberg L: The
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
use of doxycycline in nervous system Borrelia burgdorferi
infection. Scand J Infect Dis Suppl 53:7479, 1988
Dotevall L, Hagberg L: Penetration of doxycycline into cerebrospinal fluid in patients treated for suspected Lyme neuroborreliosis. Antimicrob Agents Chemother 16:1078-1080, 1989
Logigian EL, Steere AC: Clinical and electrophysiological
findings in chronic neuropathy of Lyme disease. Neurology
42:303-311, 1992
Halperin JJ, Little BW, Coyle PK, Dattwyler RJ: Lyme disease:
cause of a treatable peripheral neuropathy. Neurology 37: 1 7 W
1706, 1987
Logigian EL, Kaplan RF, Steere AC: Chronic neurologic manifestations of Lyme disease. N Engl J Med 323:1438-1444, 1990
Halperin JJ, Luft BJ, Anand AK, Roque MD, Alverez 0,
Volkman DJ, Dattwyler RJ: Lyme neuroborreliosis: central
nervous system manifestations. Neurology 39:753-759, 1989
Nocton JJ, Dressler F , Rutledge BJ, Rys PN, Persing DH,
Steere AC: Detection of Borrelia burgdorferi DNA by polymerase chain reaction in synovial fluid from patients with Lyme
arthritis. N Engl J Med 330:229-234, 1994
Kalish RA, Leong JM, Steere AC: The humoral immune response to the OspA protein of Borrelia burgdorferi in Lyme
disease (abstract). Arthritis Rheum 36 (suppl 9):S41, 1993
Schaible UE, Kramer MD, Wallich R, Tran T, Simon MM:
Experimental Borrelia burgdorferi infection in inbred mouse
strains: antibody response and association of H-2 genes with
resistance and susceptibility to development of arthritis. Eur J
Immunol21:2397-2405, 1991
Miller LC, Lynch EA, Isa S, Logan JW, Dinarello CA, Steere
AC: Balance of synovial fluid IL-Ip and IL-1 receptor antagonist and recovery from Lyme arthritis. Lancet 341:146-148,
1993
Coblyn JS, Taylor P: Treatment of chronic Lyme arthritis with
hydroxychloroquine. Arthritis Rheum 24:1567-1569, 1981
Schoen RT, Aversa JM, Rahn DW, Steere AC: Treatment of
refractory chronic Lyme arthritis with arthroscopic synovectomy. Arthritis Rheum 34:1056-1060, 1991
Документ
Категория
Без категории
Просмотров
0
Размер файла
954 Кб
Теги
treatment, lyme, arthritis
1/--страниц
Пожаловаться на содержимое документа