close

Вход

Забыли?

вход по аккаунту

?

Risk factors for serious infection during treatment with cyclophosphamide and high-dose corticosteroids for systemic lupus erythematosus.

код для вставкиСкачать
ARTHRITIS & RHEUMATISM
Vol. 39. No. 9, Sepremher 1996, pp 1475-1482
0 1996. American College of Rheurnatology
1475
RISK FACTORS FOR SERIOUS INFECTION DURING
TREATMENT WITH CYCLOPHOSPHAMIDE AND
HIGH-DOSE CORTICOSTEROIDS FOR
SYSTEMIC LUPUS ERYTHEMATOSUS
BE'lTY D. PRYOR, SABA G. BOLOGNA, and LESLIE E. KAHL
Objective. To determine risk factors for serious
infection during treatment with cyclophosphamide
(CYC) and high-dose corticosteroids in systemic lupus
erythematosus (SLE).
Methods. Records of 100 SLE patients who had
received CYC were examined for documentation of
serious infections that occurred during CYC therapy
and the subsequent 3 months.
Results. Infection occurred in 45 of 100 patients
during CYC therapy. Patients with infection were more
likely to have multiple organ disease (4Wo versus 2Wo;
P = OM), a lower nadir in the white blood cell (WBC)
count (2,818 versus 3,558 celldpl; P = 0.02), and a
higher maximum corticosteroid dose (195 versus 73 mg;
P S 0.01) than patients without infection. Infection
occurred with equal prevalence in those who received
intravenous (IV) (39%) or oral (40%) CYC, but was
more common with use of sequential IV and oral
therapy (68%). By multivariate analysis, the strongest
association with infection was a WBC nadir $3,000
cells/pl (odds ratio [OR] 2.8, 95% confidence interval
[95% CI] 1.4-5.5) and use of sequential IV and oral
CYC (OR 23, 95% CI 12-43), Infection occurred in
more CYC-treated patients taking concomitant steroids
than in those treated with high-dose steroids alone (45%
versus 12%;P = 0.001). Fatal and opportunistic infections during CYC therapy were associated with a low
WBC nadir and a high ma)rimum corticosteroid dose.
Conelusion. The risk of serious infection in patients with SLE is influenced by the inclusion of CYC in
Supported in part by NIH grant RR-00036to the Washington
University Clinical Research Center.
Betty D. Pryor, MD, Saba G. Bologna, MSN,Leslie E. Kahl,
MD: Washington University School of Medicine, St. Louis, Missouri.
Address reprint requests to k s l i e E. Kahl, MD, 660 South
Euclid, Box 8045, St. Louis, MO 63110.
Submitted for publication July 12, 1995; accepted in revised
form March 11, 1996.
the treatment regimen. The likelihood of infection in
this setting is enhanced by CYC-induced reductions in
the total WBC count C3,OOO cells/pl and by sequential
IV and oral therapy. Both of these factors may be
indicators of aggressive cytotoxic treatment, underscoring the need for close observation during treatment to
minimize the risk of serious infection.
Systemic lupus erythematosus (SLE) is a systemic
autoimmune disease with clinical manifestations that
range from troublesome rashes and arthralgias to lifethreatening nephritis and cerebritis. Treatment of severe
flares of disease has traditionally involved high-dose
corticosteroids. More recently, immunosuppressive
medications, particularly cyclophosphamide (CYC),
have often been added to the treatment regimen, both
for their steroid-sparing effect and for their potentially
superior therapeutic efficacy (1-4).
While such cytotoxic therapy may partially account for improved survival in SLE (9it may also
contribute to the substantial risk of serious infections in
these patients. Infection is a primary or secondary factor
in 28-52% of SLE deaths (6-9) and is also a common
nonfatal event in SLE. The pathogenesis of infection in
this setting is complex, such that the specific risk for
infection attributable to immunosuppressive therapy
with cytotoxic agents remains unclear. SLE itself (lO,ll),
the presence of active nephritis (12,13), and overall
disease activity (14-16) have all been linked to an
increased risk of infection. Moreover, the role of corticosteroid therapy in infection risk is a subject of controversy (10-15).
Few studies have provided detailed information
regarding infection rates and risk factors during treatment with the cytotoxic agent CYC (1). In one study,
cytotoxic therapy of unspecified type was significantly
correlated with death from infection, particularly oppor-
PRYOR ET AL
1476
tunistic infection (17). The cytotoxic agent azathioprine
has been implicated as an infection risk factor in several
studies (12,15), but exonerated in others (7,13). With
respect to CYC, several investigators have indicated that
it does not generally increase the risk for nonfatal
bacterial infections in SLE, although it does increase the
likelihood of fungal and viral infections (1,18,19). It has
also been suggested that the risk of infection during
CYC treatment of active SLE is closely dependent on
the dose of concomitantly administered corticosteroids (20).
To clarify these issues, we retrospectively determined infection rates and defined infection risk factors
in a large series of SLE patients treated with CYC. We
specifically sought to delineate the relative contributions
of SLE disease activity, corticosteroid dose, route of
CYC therapy, and leukopenia as risk factors for
infection.
PATIENTS AND METHODS
Patients and clinical data. Patients who had received
CYC for the treatment of SLE were identified from our
computerized SLE database. At the time of this study, the
database contained information on approximately 480 patients, 99% of whom met the American College of Rheumatology (ACR) classification criteria for SLE (21).
All medical records concerning the period of CYC
therapy and the 3 months following therapy were reviewed for
indication for treatment, route of administration, corticosteroid dose, laboratory data, and occurrence of infection as
defined below. The maximum corticosteroid dose during a
course was defined as the prednisone milligram-equivalent
dose received for more than 3 consecutive days.
Lupus disease activity at the start of CYC therapy and
at the time of infection was estimated using the Systemic Lupus
Activity Measure (SLAM), a graded assessment of clinical and
laboratov features of SLE (range of scores 0-86) (22). For
this retrospective chart review, a modified SLAM score was
calculated using the total number of points divided by the total
points with information available, multiplied by 100. Disease
activity was also addressed by classlfying patients as having
single-organ versus multiple-organ disease, with the latter
defined as the presence of two or more manifestations of SLE
for which CYC was thought to be indicated.
Infections occurring during CYC therapy and the
following 3 months were considered serious clinical events and
were included in the analysis if they met the following criteria:
required hospitalization and antibiotics (except viral infections); and yielded positive cultures or abnormal radiographic
findings. Herpes zoster infections were included if typical
dermatomal vesicular cutaneous eruptions were present. Viral
gastroenteritis, localized and uncomplicated genital or oral
Candida, and mucocutaneous herpes simplex viral infections
were excluded because of potential underreporting of these
conditions. Infections diagnosed at autopsy were excluded
from analysis because autopsies were not routinely done in
every case of death.
To clarify the relative contribution of corticosteroids to
the risk of serious infection among CYC-treated patients, we
also reviewed records of a comparison group of 43 patients
with SLE, according to the ACR criteria (21), who had
received high-dose steroids without CYC. Patients in this
group received 260 mg of prednisone equivalent for 214 days.
Infections were recorded according to the criteria noted above
for the period of prednisone treatment beginning with 2 6 0 mg
and continuing until the dosage had been tapered to 20
mg/day, and for 1 month following that treatment.
Statistical analysis. The first course of CYC for each
patient and the first infection during a course were used for
analysis in order to avoid overrepresenting individual risk
factors for each patient more than once. For 1 patient, the
second course was used in analysis because records of the first
course were unavailable. Chi-square, Fisher’s exact test, or
Wilcoxon’s rank sum test were used where appropriate. Cox
proportional-hazard test was used for the time-dependent
multivariate analysis.
RESULTS
One hundred ten patients who received cyclophosphamide for treatment of disease exacerbations
were identified from the SLE database. Sufficient data
were available for analysis of 100 of these patients
(91%), who constituted the final study group. This group
was composed of 82 women and 18 men, of whom 59
were Caucasian, 36 were African-American, and 5 were
of other races. Patients were treated between 1973 and
1992. The ACR criteria for SLE (21) were met by 99 of
the patients. Patients were followed up for a mean of
14.4 months.
Characteristics of patients and CYC courses.
The demographic and clinical characteristics of the 100
study patients, which were obtained at the start of CYC
treatment, along with data regarding the CYC therapy,
are summarized in Table 1. No uniform treatment
protocol was used by the clinicians who treated the
patients. Renal disease was the sole indication for
treatment in 40 patients, and was present in 32 of the 38
patients (84%) with multiple organ disease. As expected,
patients with multiple organ involvement had a higher
mean SLAM score than those with any single organ
involved (mean score 26.7 versus 14.6; P < 0.01).
Among the 19 patients who received both intravenous (IV) and oral CYC sequentially, 16 received IV
therapy initially and oral therapy subsequently. The
mean duration of therapy for all patients was 11.5
months (range 1 IV infusion to 96.5 months of therapy).
Three patients received concomitant cytotoxic therapy
with other agents: 1 took vinblastine and 1 took vincris-
1477
RISK OF INFECTION DURING CYC THERAPY FOR SLE
tine for refractory thrombocytopenia; 1 took cyclosporin
A for central nervous system vasculitis, nephritis, and
pulmonary hemorrhage.
Infections during CYC therapy. Using the criteria described in Patients and Methods, 45 of the 100
patients developed infection during their first course of
CYC. The mean duration of CYC therapy prior to
infection was 4.1 months. The mean cumulative dose of
CYC prior to infection was 9.3 gm (129.1 mglkg). The
proportion of CYC courses during which infection occurred was similar throughout the 20-year time period
encompassed in this study.
Several potential risk factors for infection were
examined among these 45 patients (Table 2). Of note,
infection occurred when disease was less active than at
the beginning of CYC therapy, as indicated by the
S L A M scores. Moreover, the mean steroid dose was
lower at the time of infection than the maximum dose
that was received during treatment, and both the total
white blood cell (WBC) and absolute neutrophil counts
were higher at the time of infection than at their nadir
during treatment. Race and sex did not alter infection risk.
Treatment was halted within 2 weeks of infection
in 16 patients, 6 of whom developed 1 or more additional
infections over the ensuing 3 months. The remaining
Table 1. Characteristics of 100 SLE patients and their cyclophosphamide therapy'
Variable
Demographics
Age, mean 2 SD years
Female sex, %
Race, YO
Caucasian
African-American
Other
SLE duration, mean (range) years
Disease activity, mean 2 SD SLAM score
Indication for cyclophosphamide, %
Renal
Other single organ
CSNPNS
Hematologic
Vasculitis
Other
Multiple organ
Cyclophosphamide route, %
Oral
Intravenous
Sequential IV and oral
Value
34.7 2 13.9
86
53
42
5
7.1 (0.002-38.3)
19.1 % 12.7
40
22
10
5
4
3
38
40
41
19
* SLE = systemic lupus erythematosus; S L A M = Systemic Lupus
Activity Measure; CNS = central nervous system; PNS = peripheral
nervous system; IV = intravenous.
Table 2. Characteristics of 45 SLE patients in whom infection occurred during CYC therapy'
Variable
SLE disease activity, SLAM swre
At CYC initiation
At infection
Corticosteroid dose, mglday
Maximum
At infection
WBC, celldpl
Nadir
At infection
ANC, celldpl
Nadir
At infection
Serum creatinine, mddl
At CYC initiation
At infection
Value
P
19.6 2 13.6
16.6 2 13.4
<0.05
195 (30-2300)
114 (5-2300)
cO.01
2,818 2 1,556
8,450 2 5,878
CO.01
1,489 2 705
7,149 2 5,555
<0.01
2.4 t 2.1
2.4 2 2.1
NS
* Values are the mean -t SD, except for the corticosteroid dose, which
is the mean (range) because of skewed distributions. The highest
corticosteroid dose of 2,500 is indicative of daily bolus intravenous
therapy. P values are for comparisons within categories. SLE =
systemic lupus erythematosus; CYC = cyclophosphamide;
SLAM = Systemic Lupus Activity Measure; WBC = white blood cells;
ANC = absolute neutrophil count; NS = not significant.
patients received a mean of 13.2 months of additional
treatment, during which time 9 had 1 or more additional
infections. Fifteen patients received 2 or more courses of
CYC. Among the 7 patients who had an infection during
the first course, 2 had an infection in subsequent courses,
compared with 4 of 8 patients who had no infection
during the first course.
Qpe of infection. Bacterial infections were the
most common, occumng in 26 of 100 courses of CYC,
with opportunistic infections in 11 and herpes zoster in 8
courses. For each of these classes of pathogens, we
examined risk factors both at the time of infection and at
any time during the treatment course. No specific risk
factor, including maximum corticosteroid dose and leukopenia, correlated with the risk of developing a bacterial infection. Bacterial organisms that were identified
during cyclophosphamide therapy are listed in Table 3.
Bacterial superinfection occurred in 2 of 8 patients with herpes zoster infection, with subsequent
bacteremia and septic pulmonary emboli in 1 patient.
Herpes zoster infection correlated with sequential oral
and IV CYC administration (P = 0.05) and a higher
maximum steroid dose during the CYC course (P =
0.03).
The pathogens identified in opportunistic infections which occurred during CYC therapy are listed in
PRYOR ET AL
1478
Table 3. Bacterial infections during cyclophosphamide therapy*
Organism
Total
Site (n)
Escherichia coli
5
Pseudomonos
4
Urine (Z), blood/urine (2), blood (1)
Urine (Z), blood/lung (l),blood/
lung/urine/sinus (1)
Urine (2). pancreas (l), skin (1)
Blood/skin (1). blood/urine (l), skin
Mebsielln
Staphylococcus aureiu
4
Citrobacter
Enterobacter
Proteeus
Streptococcus pneicmoniae
Haemophilus
Lirteria
Morgariella
Staphylococcus species
Streptococcus group A
Unidentified
2 Pancreas (I), urine (1)
2 Urine (2)
2 Skin (l), urine (1)
2 Blood (l), bloodflung (1)
1 Lung
1 Blood/CNS/urine
1 Pancreas
3
40
(1)
UJ
w
cn
30
K
3
I
I rn
t
BACTERIAL
0
t;
0
iiii
20
I-
z
OPPORTUNISTIC
W
0
K
w
n
10
1 Lung
1 Blood/CNS
4 Lung (3). sinus (1)
* Five patients had multiple simultaneous bacterial infections. CNS
0
=
central nervous system.
SINQLE ORGAN
DISEASE
(n
-
021
MULTIPLE ORQAN
DISEASE
(0
-
3.1
INDICATION FOR TREATMENT
Table 4. All opportunistic infections are listed, including
those diagnosed at autopsy (when available), in order to
illustrate the spectrum and relative frequency of organisms involved, although as noted above, only the first
infection that occurred during the first course of CYC in
each patient was used for data analysis.
Opportunistic infections were more common
among patients treated for multiple organ disease than
for any single organ disease (P = 0.02) (Figure 1). In
Table 4. Nonzoster opportunistic infections during cyclophosphamide
therapy
Organism
Site (n)
Total
Candiah albicans
41
Pneumocystis carinii
crypt0c0cc1cs
neofomns
Tomlopsis glabrata
Presumed viral
Aspergillus fumigalus
Cytomegalovirus
Herpes simplex virus
Influenza A
Mycobacterium karlsasii
Nocardio asteroides
Trichopliyton rubmrn
3
2
2
2t
1
1*
1
1
1
1
1§
Lung/urine, bloodllung,
blood, esophagushrachea
Lung
CNS/urine, lung
Blood/lung, urine
Esophagus, lung
Lung
Adrenal
Lung
Lung
Skin
CNS/lung
Skin
* One case of Candida esophagitis/tracheitis was diagnosed at autopsy
and not included in the data analysis.
t Included were 1 case of pulmonary infiltrates on chest radiograph,
which resolved without antibiotics or change in steroid dose, and 1case
of csophagitis with a biopsy consistent with herpes simplex or cytomegalovirus (CMV) infection.
$ O n e case of CMV adrenalitis was diagnosed at autopsy and not
included in the data analysis.
Q Invasive cellulitis requiring intravenous amphotericin.
Figure 1. Distribution of types of infection by treatment indication.
Opportunistic infection occurred more frequently in patients who
received cyclophosphamide for multiple organ disease than for any
single organ disease (P = 0.02). N o significant inter-group differences
in the bacterial and herpes zoster infections were found.
contrast to the findings for bacterial infection, a higher
maximum steroid dose (P = 0.04) and a lower WBC
nadir at some time during the treatment course were
associated with opportunistic infection (P = 0.02). However, when we examined risk factors present at the time
of diagnosis of infection among the 3 classes of pathogens, no association was found between the WBC count,
absolute neutrophil count, current steroid dose, or disease activity (SLAM score) and any type of infection.
Comparison of CYC courses with and without
infection. Potential risk factors for infection were next
compared between CYC courses with and without the
occurrence of infection (Table 5). Among patients who
became infected, multiple organ disease was significantly
more frequent than any single organ involvement (49%
versus 29%; P = 0.04). The likelihood of infection was
similar when either oral (40%) or intravenous (39%)
cyclophosphamide was given alone, compared with
courses in which both regimens were given sequentially
(68%; P = 0.003). Among this latter group, infection
occurred equally as often in the oral phase as in the IV
phase. The majority of patients who received both routes
of CYC sequentially (68%) had multiple organ disease;
the WBC nadir and maximum steroid dose did not differ
from those who received CYC by either route alone.
RISK OF INFECTION DURING CYC THERAPY FOR SLE
Table 5. Comparison of cyclophosphamide courses with and without
infection*
Variable
Infection
(n = 45)
35.5 ? 13.8
Age at initiation of CYC,
mean 2 SD years
SLE duration, mean
7.3 (0.0-38.3)
(range) years
Disease activity at
19.6 ? 13.6
initiation of CYC,
mean ? SD SLAM
score
Indication for therapy,
n (%)
Renal
15 (33)
Other single organ
8 (18)
Multiple organ
22 (49)
Route, n (%)
Oral
16 (36)
IV
16 (36)
Sequential IV and oral
13 (28)
Duration of therapy,
13.3 (0.03-2,500)
mean (range)
months
Steroids
Maximum, mean
195 (30-2,500)
(range) mg/day
Median
100
Lowest WBC, mean %
2,818 t 1,556
SD cells/pl
Lowest ANC, mean 2
1,489 2 705
SD cells/pl
Serum creatinine at
2.2 (0.5-12.3)
initiation of CYC,
mean (range) mddl
No Infection
(n = 55)
P
34.0 t 14.0
NS
6.7 (0.0-27.7)
NS
18.6 ? 12.0
NS
25 (46)
14(25)
16 (29)
NS
NS
0.04
24(44)
25 (45)
6 (11)
10.1 (0.03-35.0)
NS
NS
0.003
NS
73 (20-150)
80
3,558 2 1,339
50.01
0.02
1,759 2 495
NS
2.1 (0.5-9.1)
NS
* P values are for comparisons between courses with and without
infection. See Table 2 for definitions.
The maximum corticosteroid dose during the
CYC treatment course was higher in patients who
developed an infection than in those who did not (mean
195 mg versus 73 mg; P 5 0.01). However, half of the
infections occurred at a steroid dose of 540 mg/day, and
24% occurred at 525 mglday. Among patients who did
not become infected, the majority (80%) had received a
maximum steroid dose of >40 mg/day at some point
during their CYC course. Patients who received 3 days
of bolus IV steroids at initiation of CYC had the same
overall frequency of infection as those who did not, but
had an increased frequency of herpes zoster (8 of 40
versus 2 of 60; P < 0.01).
The lowest total WBC count at any time during
the CYC course was significantly lower in patients who
developed an infection than in those who did not (mean
2,818 versus 3,558 cells/pl; P = 0.02). However, the
difference in lowest absolute neutrophil count between
the groups was not significant. Among patients with a
1479
WBC nadir of 53,000 cells/pl, 55% developed an infection at some point during treatment, compared with
36% of patients with a WBC nadir of >3,000 cells/pl
(P = 0.001). The WBC nadir occurred within the 3
weeks prior to the infection in only 5 patients (11%).All
patients with a WBC nadir of 53,000 cells/pl who had
received sequential IV and oral CYC became infected,
compared with 58% who had received oral and 52% who
had received IV treatment alone (P = 0.007).
The presence of active renal disease alone, or as
part of multiple organ disease, did not correlate with the
development of infection. All 4 of the patients receiving
hemodialysis during the period of CYC treatment developed infections. Similarly, diabetes did not appear to
be a risk factor for infection. Two of the 3 patients who
received other concomitant cytotoxic agents (cyclosporin
A and vincristine, respectively) developed an infection.
Factors associated with infection in univariate
analysis, including route of CYC, maximum steroid
dose, lowest WBC count, and presence of multiple organ
disease, were entered stepwise into a proportionate
hazards regression model using the time to first infection
as the dependent variable. Use of sequential oral and IV
CYC (odds ratio [OR] 2.3, 95% confidence interval
[95% CI] 1.4-4.3) and the WBC nadir of 53,000 cells/pl
(OR 2.8, 95% CI 1.4-5.5) were the only factors significantly associated with risk of infection in this model.
Deaths during CYC therapy. Infection was the
primary cause of death in 7 of 11 patients who died
during CYC therapy. Opportunistic infections accounted for 3 deaths, with fungal sepsis in 2 patients (1
of whom had also received vincristine) and Pneumocysfis
carinii pneumonia plus cytomegalovirus adrenalitis in
1 patient. Gram-negative sepsis (2 patients), gramnegative peritonitis following abdominal surgery (1 patient), and bacterial pneumonia (1 patient) were responsible for the other 4 infection-related deaths. One
patient, who was also receiving cyclosporin A, died of an
intracranial hemorrhage with concomitant Aspergillus
pneumonia and gram-negative sepsis.
Patients who developed fatal infections were
older (51.7 versus 32.5; P = 0.007) and had a lower WBC
nadir during their therapy (1,571 versus 3,047 cells/pl;
P = 0.02) than those with nonfatal infections. Six of the
7 patients with fatal infections had multiple organ involvement with SLE. The mean duration of CYC was
shorter in patients with fatal, compared with nonfatal,
infections (4.9 months versus 14.8 months; P = 0.03), but
the steroid dose, WBC count, and SLAM score at the
time of infection were similar.
PRYOR ET AL
1480
Table 6. Comparison of CYC-treated and high-dose corticosteroidtreated patients*
Steroids
(n = 43)
Variable
Demographics
Age, mean t SD years
34.5 t 14.1
SLE duration, mean
4.3 (0.0-20.8)
(range) years
SLE activity, mean 2 SD 14.9 Z 8.0
SLAM score
Indication for therapy, n (%)
Renal
Other single organ
Multiple organ
Duration of steroids 2 2 0
mg/day, mean 5 SD
months
Infection during therapy,
n @J)
Death due to infection,
n (%I
Cyclophosphamide
( n = 100)
P
34.7 -t 12.9
7.3 (0.0-38.3)
0.003
19.1 2 12.7
NS
40 (40)
NS
8 (19)
17 (39)
18 (42)
5.8 ? 4.2
22(22)
38 (38)
6.3 2 6.2
0.01
0.03
NS
NS
5 (12)
45 (45)
0.001
1 (2)
7 (7)
NS
* i' values are for comparisons between steroid-treated and CYCtreated groups. See Table 2 for definitions.
Comparison with steroid-treated patients. All of
our CYC-treated patients also received corticosteroid
therapy. To examine the relative contribution of corticosteroids to the risk of infection, we studied 43 additional
SLE patients who were treated with high-dose steroids
alone (260 mg for at least 14 days). The mean period of
followup was 6.8 months. As in the CYC-treated patients, only the first infection during a treatment course
was included in this analysis.
The demographic features of the 2 treatment
groups were similar (Table 6), except for the duration of
SLE. The proportion of patients with multiple organ
disease was similar. Infection occurred in significantly
more patients treated with CYC than with high-dose
steroids alone (45% versus 12%; P = 0.001). There were
0.018 patients infected per month of treatment in the
steroid-only group, compared with 0.031 in the CYConly group. Among the patients treated with high-dose
steroids alone, there were 4 herpes zoster infections and
1 bacterial infection (pyelonephritis). The mean steroid
dose at the time of infection was 42.5 mg/day (range
15-60 mg/day). A similar proportion of steroid-treated
patients had fatal infections (2%) compared with the
€YC-treated patients (7%).
DISCUSSION
In contrast to several previous reports, this study
demonstrates that serious bacterial and nonbacterial
infections are common among SLE patients during
cyclophosphamide therapy, resulting in significant morbidity and mortality. The rate of serious infections
remains high (37%), even if herpes zoster infections are
excluded from analysis. Although the data presented
here represent the first infection during the first course
of CYC in order to minimize confounding variables,
similar results were obtained when the analysis was
performed using all courses and all infections.
A principal finding reported here is the high
frequency of serious bacterial infections during CYC
treatment. Moreover, the risk of bacterial infection,
unlike that of zoster or opportunistic infections, did not
correlate with the corticosteroid dose. Rather, bacterial
infections developed far more often in our patients who
received CYC with concomitant steroids than in those
who received high-dose steroids alone (26% versus 2%;
P = 0.002). Thus, cyclophosphamide itself appears to
substantially enhance the risk of bacterial infection in
this setting.
Opportunistic infections were strikingly common
and involved a wide spectrum of pathogens among the
CYC-treated patients, accounting for 24% of all infections, and half of the infection-related deaths. Because
the presence of multiple organ disease, lower WBC
nadir, and a higher maximum steroid dose all significantly correlated with opportunistic infection, we cannot
identify a single mechanism for this finding. Other
studies have similarly found an association between
opportunistic infection and corticosteroid dose (13,17).
Hellmann et al additionally reported an association
between the use of cytotoxic agents and death secondary
to opportunistic infection in SLE patients (17).
We found that a WBC nadir of 53,000 cells/pl at
some point during treatment was a significant risk factor
for infection, as shown in our multiple regression model.
Equally important, a low WBC nadir was also associated
with death due to infection. We interpret the degree of
drug-induced leukopenia to be an indicator of the
intensity of the CYC treatment. Accordingly, factors
related to CYC rather than the other variables we
examined appear to account for much of the risk of
infection. Although the majority of patients (88%) had a
WBC count >3,000 cells/pl at the time of presentation
with infection, preexisting leukopenia might have been
obscured by patients' responses to the stress accompanying infection. Morgan et a1 (20), in contrast, found no
relationship between infection and WBC nadir. Overall,
our data suggest that avoidance of a WBC nadir of
53,000 cells/pl during CYC therapy would reduce the
risk of serious infection.
1481
RISK OF INFECTION DURING CYC THERAPY FOR SLE
The frequency of infection among CYC-treated
patients in our series is higher than that in several recent
reports (1,3). This may be due, in part, to the predominance of patients with nephritis alone in those series, as
compared with our patient population, in which multiple
organ involvement was present in 38%. Although Balow
et al(l6) reported a lower frequency of major infections
in their SLE nephritis patients treated with CYC, the
death rate due to infection in their series (8%) was the
same as the rate in ours (7%). Patients in our retrospective series were also not treated according to any predefined protocol, and may have received daily high-dose
steroids for a longer period of time than those in other
series. However, the relatively low frequency of infection
among our comparison group who received only highdose steroids makes this explanation less likely.
Our finding that oral and IV routes of CYC
therapy posed the same infection risk is similar to that
reported by Austin et al (1). Surprisingly, the use of
sequential IV and oral therapy was strongly associated
with infection. However, sequential therapy did not
seem to produce more severe immunosuppression, as
reflected by the degree of leukopenia, which was similar
to that in patients who received either route alone. The
most likely explanation is that most of these patients had
multiple organ disease, which suggests that sequential
therapy may be a marker for patients with more severe
disease.
Corticosteroids have previously been implicated
as a risk factor for infection in SLE (7,ll-13,15). Our
study confirms and extends the findings by showing that
the maximum steroid dose during a CYC course was
significantly higher in those patients who later developed
infection. However, this factor was not important in the
multiple logistic regression model when adjustment was
made for disease activity, WBC nadir, and route of
therapy. Additionally, 82% of our patients were taking a
lower dose of steroid than their course maximum at the
time of infection, with half taking 5 4 0 mg/day of prednisone and one-fourth taking <25 mg/day. This is in
contrast to the report by Morgan et al (20), where
infections were very rare at a dose of 5 2 5 mg. Moreover,
when compared with patients taking high-dose steroids
alone, our CYC-treated patients (all of whom were
taking concomitant steroids) had a significantly higher
frequency of infection (45% versus 12%). Thus, although corticosteroids may contribute to the risk of
infection in CYC-treated patients, they are clearly not
the sole risk factor.
There are several practical implications of our
findings which may help minimize the risk of infection
during CYC therapy for SLE. Close monitoring of the
WBC count during therapy is indicated, to avoid a nadir
WBC count of <3,000 cells/pl. Since the potential for
infection is as substantial during IV CYC as it is for oral
therapy, both groups of patients must be monitored
carefully. Patients with very active SLE involving multiple organ systems warrant particularly close observation.
A vigilant search for opportunistic organisms is indicated early in the course of a treatment-resistant infection that occurs during CYC therapy. Finally, aggressive
and early treatment of infection in older SLE patients
with multiple organ involvement may help to decrease
the high mortality rate in this subset of patients.
ACKNOWLEDGMENTS
The authors gratefully acknowledge Clare Conway for
preparation of the manuscript, and the rheumatologists of
Barnes Hospital and Washington University School of Medicine for allowing their patients to be included in this series.
REFERENCES
1. Austin HA 111, Klippel JH, Balow JE, LeRichie NGH, Stein AD,
Plotz PH: Therapy of lupus nephritis: controlled trial of prednisone and cytotoxic drugs. N Engl J Med 314:614-619, 1986
2. Felson DT, Anderson J: Evidence for the superiority of immunosuppressive drugs and prednisone over prednisone alone in lupus
nephritis. N Engl J Med 311:1528-1533, 1984
3. McCune WJ, Golbus J, Zeldes W, Bohlke P, Dunne R, Fox D:
Clinical and immunologic effects of monthly administration of
intravenous cyclophosphamide in severe systemic lupus erythematosus. N Engl J Med 318:1423-1431, 1988
4. Boumpas DT, Barez S, Klippel JN, Balow JE: Intermittent cyclophosphamide for the treatment of autoimmune thrombocytopenia
in SLE. A n n Intern Med 112674-677, 1990
5. Klippel JH: Systemic lupus erythematosus: treatment related
complications superimposed on chronic disease. JAMA 263:18121815, 1990
6. Karsh J, Klippel JH, Balow JE, Decker J L Mortality in lupus
nephritis. Arthritis Rheum 22764-769, 1979
7. Rosner S, Ginzler EM, Diamond HS, Weiner M, Schlesinger M,
Fries JF, Wasner C, Medsger TA Jr, Ziegler G, Klippel JH, Hadler
NM, Albert DA, Hess EV, Spencer-Green G, Grayze.1 A, Worth
D, Hahn BH, Barnett EV: A multicenter study of outcome in
systemic lupus erythematosus: 11. Causes of death. Arthritis
Rheum 252612-617, 1982
8. Jonsson H, Nived 0, Sturfelt G: Outcome of systemic lupus
erythematosus: a prospective study of patients from a defined
population. Medicine (Baltimore) 68:141-150, 1989
9. Harisdangkul V, Nilganuwonge S, Rockhold L Cause of death in
systemic lupus erythematosus. South Med J 80:1249-1253, 1987
10. Staples PJ, Gerding DN, Decker JL, Gordon RS Jr: Incidence of
infection in systemic lupus erythematosus. Arthritis Rheum 17:l10, 1974
11. Nived 0, Sturfelt G, Wollheim F: Systemic lupus erythematosus
and infection: a controlled and prospective study including and
epidemiological group. QJM 55:271-287, 1985
12. Lee P. Urowitz MB, Bookman AAM, Koehler BE, Smythe HA,
Gordon DA. Ogryzlo MA: Systemic lupus erythernatosus: a review
1482
13.
14.
15.
16.
17.
of 110 cases with reference to nephritis, the nervous system,
aseptic necrosis and prognosis. QJM 46:l-32, 1977
Ginzler E, Diamond H, Kaplan D, Weiner M, Schlesinger M,
Seleznick M: Computer analysis of factors influencing frequency of
infection in systemic lupus erythematosus. Arthritis Rheum 21:3744, 1978
Dull) KNW, DuEy C, Gladman DD: Infection and disease activity
in systemic lupus erythematosus: a review of hospitalized patients.
J Rheumatol 18:1180-1184, 1991
Petri M, Genovese M: Incidence of and risk factors for hospitalizations in systemic lupus erythematosus: a prospective study of
the Hopkins lupus cohort. J Rheumatol 19:1559-1565, 1992
Balow JE, Austin HA 11, Tsokos GC, Antonovych 'IT, Steinberg
AD, Klippel JH: Lupus nephritis. Ann Intern Med 106:79-94,
1987
Hellmann DB, Petri M, Whiting-O'Keefe Q: Fatal infections in
systemic lupus erythematosus: the role of opportunistic organisms.
Medicine (Baltimore) 66341-348, 1987
PRYOR ET AL
18. Steinberg AD, Plotz PH, Wolff SM, Wong VG, Agus SG, Decker
JL: Cytotoxic drugs in treatment of nonmalignant diseases. Ann
Intern Med 76:619-642, 1972
19. Nashel DJ: Mechanism of action and clinical applications of
cytotoxic drugs in rheumatic disorders. Med Clin North Am
695317-840, 1985
20. Morgan MC, Matteson E, Dunne R, Brown M, Fox DA, McCune
WJ: Complications of intravenous cyclophosphamide: association
of infections with high daily doses of prednisone in lupus patients
(abstract). Arthritis Rheum 33 (suppl 5):R27, 1990
21. Tan EM, Cohen AS, Fries JF, Masi AT, McShane DJ, Rothfield
NF, Schaller JG, Tala1 N, Winchester RJ:The 1982 revised criteria
for the classification of systemic lupus erythematosus. Arthritis
Rheum 25:1271-1277, 1982
22. Liang MH, Socher SA, Larson MG, Schur PH: Reliability and
validity of six systems for the clinical assessment of disease activity
in systemic lupus erythematosus. Arthritis Rheum 32:1107-1118,
1989
Документ
Категория
Без категории
Просмотров
0
Размер файла
723 Кб
Теги
factors, lupus, treatment, serious, dose, high, systemic, erythematosus, corticosterone, infectious, risk, cyclophosphamide
1/--страниц
Пожаловаться на содержимое документа