Background
Rituximab in patients with antiphospholipid syndrome (APS) is mostly used for treating
noncriteria APS manifestations such as thrombocytopenia, and is not a first- or second-line
treatment for management of thrombosis. However, at times rituximab is used empirically
as adjunct treatment in patients with recurrent thrombosis occurring in spite of anticoagulant
therapy (i.e., anticoagulation failure) when combinations of increased anticoagulation
intensity, addition of antiplatelet therapy, and reduction of all thrombotic risk
factors fail to prevent recurrent thrombosis.[1]
[2] Knowledge of its effect on the naturally fluctuating antiphospholipid antibody (APLA)
titers is, however, limited. We sought to determine the effect of rituximab on decreasing
the magnitude and type of APLA in patients with thrombotic APS.
Methods
This is a retrospective case series of adult patients (age ≥ 18) diagnosed with APS
at the University of North Carolina Hospital inpatient and outpatient systems between
August 2014 and August 2020. The electronic medical record encounter data were queried
using the keyword “rituximab” and International Classification of Diseases 9th edition
(ICD9/ICD10) codes for arterial and venous thromboses, APS, and catastrophic APS (CAPS).
Venous thromboembolic events were defined as pulmonary embolism, deep vein thromboses
of the upper and lower extremities, inferior vena cava, and renal veins. Arterial
thromboses were considered if they affected major vessels. All thrombotic events were
confirmed by chart review of radiology imaging reports. APLA titers were identified
by current procedure terminology codes and considered positive if the lupus anticoagulant
(LA) was positive (dilute Russell's viper venom time [DRVVT] ratio ≥1.20) in a DRVVT-based
LA assay and/or hexagonal phospholipid neutralization (HPN) difference in an LA-aPTT-based
assay >8 seconds, anticardiolipin (aCL) immunoglobulin G (IgG) >23 GPL, aCL IgM >10,
and anti-β2-glycoprotein (anti- β2GPI) IgG or IgM >20 G or M units. The aCL and anti-β2GPI positivity were arbitrarily categorized before data collection as low, medium,
and high (< 40; 40–100; and >100 GPL, MPL, G units or M units, respectively).
“Time zero” was the start and stop dates of the first treatment course of rituximab.
Negative or positive numbers indicated time before the start or after the stop dates,
respectively. Laboratory observations 12 months before and after time zero were collected
but censored after the start of the second treatment, if applicable. Titer changes
are described.
Results
Thirty-seven patients with APS were identified and 28 excluded for the following reasons:
no available APLA titers (n = 10); only one set of titers (n = 10); no titers obtained within 12 months of treatment (n = 5); only posttreatment titers available (n = 3). Nine patients with thrombotic APS met the inclusion criteria for the primary
analysis. Here are the associated diagnoses: APS only (n = 4, including 1 patient with livedoid vasculopathy), CAPS (n = 2), APS and immune thrombocytopenia (n = 2), and APS and systemic lupus erythematosus (n = 1). The medical records were reviewed for the indication for rituximab and use
of concomitant treatment. At the time of treatment with rituximab, all patients were
on therapeutic anticoagulation including antiplatelet therapy if they had arterial
thrombotic events. All were exclusively treated with rituximab for thrombotic manifestations
of APS, including the two patients with CAPS (ID#3 and 4) who also received therapeutic
plasma exchange. The sample comprised women (n = 8), median age of 47 years (40–79 years), Caucasian (n = 6), black (n = 2), and Hispanic/Latino (n = 1). The following rituximab regimens were used: 375 mg/m2 weekly ×4 doses (n = 4), 1 g rituximab every 2 weeks ×2 doses (n = 3), and 1 patient treated with both doses at two independent encounters.
Prior to treatment, 3 patients were triple APLA-positive, 2 were double, and 4 were
single-positive based on the presence of a moderate or high positive aCL, anti-β2GPI IgG titers, and/or positive LA. All 3 triple-positive patients remained triple-positive
after treatment, 1 of 2 double-positive patients became single-positive, and 3 of
4 single-positive patients developed negative APLA tests. Pretreatment, 6 patients
had a positive LA (5/6 positive HPN difference and 4/6 positive DRVVT ratio) ([Fig. 1]).
Fig. 1 The changes in the HPN difference and the DRVTT ratio pre- and posttreatment with
rituximab for each patient. Time zero is noted as “0” and marks the initiation of
treatment. To the left of time zero is the pretreatment period and to the right is
the posttreatment. Patients 3 and 4 were treated with therapeutic plasma exchange
at the time of rituximab. DRVTT, dilute Russell viper venom time; HPN, hexagonal phospholipid
neutralization.
In 2/4 patients, both the HPN and DRVVT ratio decreased, and in 2/4 both the HPN and
DRVVT ratio increased. The mean decreases in HPN and DRVVT ratio were 58 seconds and
0.75, respectively. The mean increases in HPN and DRVVT ratio were 80 seconds and
0.625, respectively. The LA remained positive after treatment in all in whom the LA
had been positive before treatment.
The changes in aCL and anti-β2GPI titers per patient are displayed in [Fig. 2]. Among patients with titers in the medium range (5/6) and high range (1/6) and positive
pretreatment aCL tests (6/9), all remained positive posttreatment: 3/6 remained in
their pretreatment category, 1/6 increased from medium to high range, 1/6 decreased
from medium to low, and 1/6 decreased from high to a medium titer range. Prior to
treatment, all (3/9) aCL IgM-positive pretreatment titers were medium-range; posttreatment
2/3 became negative and 1/3 remained unchanged.
Fig. 2 The changes in anticardiolipin and β-2-glycoprotein IgM and IgG titers pre- and posttreatment
with rituximab. Time zero is noted as “0” and marks the initiation of treatment. To
the left of time zero is the pretreatment period and to the right is the posttreatment.
Patient 5 had a negative aCL and aβ2GPI titers measured 1,917 days before treatment—these
pretreatment data points are not represented. Patients 3 and 4 were treated with therapeutic
plasma exchange at the time of rituximab. aCL, anticardiolipin; IgG, immunoglobulin;
IgM, immunoglobulin M.
The aβ2GPI IgG titers were abnormal in 6/9 patients. Pretreatment, 3/6 were low, 1/6 medium,
and 2/6 in high-titer ranges. After treatment 3/6 remained unchanged, 2/6 increased
from low to medium, 1/6 from medium to high, and 1/6 that had been negative pre-rituximab
became positive. There were 2/9 pre- and posttreatment aβ2GPI IgM titers and these
remained unchanged.
One of nine (1/9) patients had a recurrent thrombotic event within 1 year of rituximab
treatment and died 2 years later from a spontaneous intracranial hemorrhage while
on therapeutic anticoagulation and while a platelet count was 47,000 × 109/L.
Discussion/Conclusion
We found that the use of rituximab in patients with APS did not meaningfully or consistently
lower APLA titers. This finding is in keeping with the previous single-center prospective
Erkan et al's study which showed no substantial change in the APLA profile of 19 APS
patients—all patients who had positive APLA results at baseline had positive results
at 24 and 52 weeks on follow-up testing.[3] In contrast, the retrospective study by Agmon-Levin et al showed that in 23 patients
with APS who had pre- and post-rituximab APLA titer testing, there was a statistically
significant decrease in APLA titers posttreatment in the group of 13 patients who
had a clinical treatment response, but not in the group of 10 patients who did not
have a clinical response.[4] Furthermore, our review of the literature of case reports and small case series
shows that the effect of rituximab on APLA titers is inconsistent and equivocal—increasing,
decreasing, or remaining unchanged in some patients.[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
While our study is limited by its small sample size due to the significant number
of patients who had to be excluded because of nonavailability of APLA test results
in the electronic medical record system, its retrospective nature, the inclusion of
results obtained with various APL assays with the potential for inter-assay result
variations, and the heterogeneity of the clinical phenotypes, it nonetheless serves
as useful confirmatory data to the existing limited published literature that rituximab
use in APS may not be a useful approach to lower APLA titers. Whether rituximab lowers
future thrombosis risk in APS patients, particularly in those who have had an anticoagulation
failure, has not been studied and our study does not provide an answer to this clinically
important question.[8] However, the assumption that rituximab therapy as an adjunct to anticoagulation
therapy might lead to a lower future thrombosis risk by lowering APLA titers is not
supported by our study and may be better answered by prospective clinical trials.
In summary, the results of our study, the 2013 Erkan et al study, and the limited
case report and small case series literature dampen our enthusiasm to use empiric
rituximab in patients with APS and recurrent venous thromboembolism to potentially
lower APLA titers and, through that, the risk of recurrent thrombosis.
