Abstract
Objective
The APS ACTION Registry studies long-term outcomes in persistently antiphospholipid antibody (aPL)-positive patients. Our primary objective was to determine whether clinically meaningful aPL profiles at baseline remain stable over time. Our secondary objectives were to determine (1) whether baseline characteristics differ between patients with stable and unstable aPL profiles, and (2) predictors of unstable aPL profiles over time.
Methods
A clinically meaningful aPL profile was defined as positive lupus anticoagulant (LAC) test and/or anticardiolipin (aCL)/anti-β2 glycoprotein-I (anti–β2-GPI) IgG/M ≥ 40 U. Stable aPL profile was defined as a clinically meaningful aPL profile in at least two-thirds of follow-up measurements. Generalized linear mixed models with logit link were used for primary objective analysis.
Results
Of 472 patients with clinically meaningful aPL profile at baseline (median follow-up 5.1 yrs), 366/472 (78%) patients had stable aPL profiles over time, 54 (11%) unstable, and 52 (11%) inconclusive. Time did not significantly affect odds of maintaining a clinically meaningful aPL profile at follow-up in univariate (P = 0.906) and multivariable analysis (P = 0.790). Baseline triple aPL positivity decreased (OR 0.25, 95% CI 0.10–0.64, P = 0.004) and isolated LAC test positivity increased (OR 3.3, 95% CI 1.53–7.13, P = 0.002) the odds of an unstable aPL profile over time.
Conclusion
Approximately 80% of our international cohort patients with clinically meaningful aPL profiles at baseline remain stable at a median follow-up of 5 years; triple aPL-positivity increase the odds of a stable aPL profile. These results will guide future validation studies of stored blood samples through APS ACTION Core Laboratories.
Keywords: Anticardiolipin Antibodies, Antiphospholipid Antibodies, Antiphospholipid Syndrome
Antiphospholipid syndrome (APS) is an autoimmune disorder characterized by thrombosis and pregnancy morbidity in patients with persistently positive antiphospholipid antibodies (aPL). aPL that are used for APS classification include lupus anticoagulant (LAC) test, anticardiolipin antibodies (aCL), and anti-β2 glycoprotein-I antibodies (anti–β2-GPI)1.
The assessment of aPL profiles upon evaluation of aPL-positive patients is critical. Persistently positive aPL are more likely to have important clinical implications, while transiently positive aPL, especially of low titer, may be a result of infections or medications. Certain aPL profiles, such as LAC positivity, high titer aCL/anti–β2-GPI, or triple aPL positivity, are more strongly associated with aPL-related clinical events, although traditional risk factors also need to be taken into account while evaluating aPL-positive patients2. The course of aPL positivity over time is also important in the risk stratification and management of patients; however, there are limited prospective data on the course of aPL tests over time.
Antiphospholipid Syndrome Alliance for Clinical Trials and International Networking (APS ACTION) is an international network created to design and conduct large-scale multicenter studies and clinical trials in persistently aPL-positive patients. The APS ACTION clinical database and repository (“Registry”) was created to study the natural course of persistently aPL-positive patients with or without autoimmune disorders over at least 10 years; the Registry allows us to perform large-scale cross-sectional and prospective analyses, which will eventually help us better understand the clinical characteristics of patients with APS.
In this analysis of the APS ACTION Registry, our primary objective was to determine whether clinically meaningful aPL profiles (defined as a positive LAC test and/or aCL/anti–β2-GPI IgG/M ≥ 40 U) at baseline remain stable over time in persistently (on 2 or more occasions at least 12 weeks apart) aPL-positive patients. Our secondary objectives were to determine (1) whether demographic, clinical, and laboratory characteristics at baseline differ between patients with stable and unstable aPL profiles over time, and (2) predictors of unstable aPL profiles over time.
MATERIALS AND METHODS
APS-ACTION Registry. The APS-ACTION Registry is a Web-based data capture system developed in Research Electronic Data Capture (REDCap) that includes patients with persistently positive aPL with or without other systemic autoimmune disease. Inclusion criteria are positive aPL, based on the Updated Sapporo APS Classification Criteria1, tested at least twice within 1 year prior to enrollment. Patients are followed every 12 ± 3 months with clinical and laboratory data, and blood collection. Each participating center has ethics board approval (Hospital for Special Surgery Institutional Review Board ID 2014–252, lead coordinating center), and all patients have provided written informed consent that allows publication of this material. Study cohort. As of January 2019, there were 796 patients enrolled in APS ACTION Registry from 26 centers worldwide; 472 patients with baseline clinically meaningful aPL profiles and follow-up visits with available aPL tests were included in this analysis (Figure 1).
Figure 1 –
aPL profiles over time (N = 482).
* Reasons for inconclusive follow-up aPL profile: (1) missing determinant aPL test(s) (those used to determine the baseline clinically meaningful aPL profile) with no other positive aPL tests (n = 28), and (2) negative determinant aPL test result(s) with missing other aPL test result(s) (n = 24). aPL: antiphospholipid antibody.
Data collection.
For this retrospective and prospective Registry analysis, we retrieved clinical and laboratory data at baseline and follow-up. The clinical data included information on demographics (age, sex, race, and ethnicity), concomitant systemic autoimmune diseases, aPL-related history (thrombotic and obstetric), noncriteria aPL manifestations (i.e., thrombocytopenia, autoimmune hemolytic anemia, cardiac valve disease, livedo reticularis/racemosa, skin ulcers, aPL nephropathy, and cognitive dysfunction), and medications. All available standard-of-care measurements (retrospective and prospective) for LAC, aCL IgG/M, and anti–β2-GPI IgG/M from the Registry were utilized. For the baseline visit, we used the most recent aPL profile (LAC, aCL/anti–β2-GPI IgG/M). At each annual follow-up visit, we used the first available aPL profile that was reported for that time period. High aCL/anti–β2-GPI titers reported as “greater than x” units (e.g., > 80 U) were converted to “x” units (e.g., 80 U) to facilitate the statistical analysis.
Definitions.
We defined a clinically meaningful profile as a positive LAC test and/or aCL/anti–β2-GPI IgG/M ≥ 40 U, and a stable clinically meaningful aPL profile if it remained stable in at least two-thirds of follow-up aPL measurements. We defined aCL/anti–β2-GPI IgG/M as positive when the reported titer was ≥ 40 U. We selected a cutoff positivity of 40 U for aCL/anti–β2-GPI, given that (1) low aPL titers are generally transient and clinically less significant, and (2) there is limited literature demonstrating that titers > 40 U are more likely to be associated with aPL-related manifestations1,3,4 and therefore are considered more effective in clinical care of aPL-positive patients. Inconclusive aPL profile during the follow-up was defined as (1) missing determinant aPL test result(s) (those used to determine the baseline clinically meaningful aPL profile) with no other positive aPL tests, or (2) negative determinant aPL test result(s) with missing other aPL test result(s).
Statistical analysis.
We hypothesized that a clinically meaningful aPL profile at baseline remains stable over time. Univariate and multivariable generalized linear mixed models (GLMM) with logit link were used to assess the effect of time and other variables of interest on odds of clinically meaningful aPL profiles over time. A GLMM framework allowed us to introduce random effects to account for within-subject correlation due to repeated measures of aPL profiles across follow-up. T-test (for normally distributed variables), Wilcoxon rank-sum (for abnormally distributed variables), and Fisher exact tests (for categorical variables) were employed to compare clinical characteristics of patients with stable versus unstable aPL profiles. Univariate and multivariable logistic regression were used to examine predictors of unstable aPL profiles (negative LAC and aCL/anti–β2-GPI IgG/M < 40 U) over time.
RESULTS
At baseline, 472 patients had a clinically meaningful aPL profile. The median age of these patients was 49 years (IQR 39–59) and 349 (74%) were female. The median follow-up was 5.1 years (IQR 4.3–5.8), and the median number of follow-up visits with aPL tests was 2 (IQR 1–3). Based on the different number of available aPL tests at each year of follow-up, 254 (73%) had clinically meaningful aPL profiles at 1-year follow-up, 216 (72%) at 2-year, 177 (72%) at 3-year, 135 (73%) at 4-year, and 61 (70%) at 5-year (Figure 1, Table 1).
Table 1-.
aPL profiles over time (N = 482)
| Baseline | 0–12 M | 12–24 M | 24–36 M | 36–48 M | 48–60 M | |
|---|---|---|---|---|---|---|
| No. patients with follow-up | N/A | 452 | 398 | 357 | 282 | 138 |
| No. patients with aPL results | 482 | 348 | 302 | 245 | 184 | 87 |
| Meaningful aPL profilea | 472 (98) | 254 (73) | 216 (72) | 177 (72) | 135 (73) | 61 (70) |
| Not meaningful aPL profileb | 3 (1) | 31 (9) | 29 (10) | 24 (10) | 14 (8) | 7 (8) |
| Inconclusive aPL profilec | 7 (1) | 63 (18) | 57 (19) | 44 (18) | 35 (19) | 19 (22) |
Values are n (%) unless otherwise indicated.
Positive LAC test and/or aCL/anti-β2-GPI IgG/M ≥ 40 U.
Negative LAC test and aCL/anti-β2-GPI IgG/M < 40 U.
Missing determinant aPL test result(s) (those used to determine the baseline clinically meaningful aPL profile) with no other positive aPL tests, or negative determinant aPL test result(s) with missing other aPL test result(s). aCL: anticardiolipin antibody; anti-β2-GPI: anti-β glycoprotein I antibody; aPL: antiphospholipid antibody; LAC; lupus anticoagulant; M: months.
aPL profile stability over time.
Three hundred and sixty-six of 472 (78%) patients had stable and 54/472 (11%) had unstable aPL profiles over a median follow-up of 5 years. One hundred and fifty-one (32%) patients contributed to the stability analysis with 1 follow-up visit, 99 (21%) with 2, 105 (22%) with 3, 87 (18%) with 4, and 27 (6%) with 5. In 52/472 (11%) patients, the assessment was inconclusive; thus, these patients were excluded from further analysis (Figure 1). A univariate GLMM demonstrated that time across follow-up did not significantly affect odds of maintaining a stable clinically meaningful aPL profile over time (P = 0.906). Similar results were observed when the model was adjusted for age, active smoking, concomitant autoimmune disease, and hydroxychloroquine (HCQ) use at baseline (P = 0.790).
Demographic, clinical, and laboratory characteristics differences between stable versus unstable aPL profile status.
Table 2 describes baseline demographic, clinical, and laboratory characteristics of the 420 patients who had stable and unstable clinically meaningful aPL profiles at follow-up. LAC, aCL IgM, anti–β2-GPI IgG positivity, and positivity on ≥ 2 aPL tests at baseline were associated with a stable aPL profile (P < 0.001, P = 0.004, P = 0.005, and P < 0.001, respectively). While aCL IgG or anti–β2-GPI IgM positivity was not associated with a stable aPL profile (P = 0.06 for both), a larger proportion of patients with a stable aPL profile were aCL IgG (50% vs 35%) and anti–β2-GPI IgM positive (21% vs 9%) at baseline. In addition, patients with stable clinically meaningful aPL profiles, compared to those with unstable aPL profiles, were more likely to have higher aCL IgG (median 46 U vs 16 U) and anti–β2-GPI IgG (median 22 U vs 3 U) titers at baseline, and triple aPL positivity (46% vs 13%), while they were less likely to have isolated LAC test positivity (17% vs 41%) or isolated anti–β2-GPI IgG/M positivity (2% vs 9%). No differences were noted between patients with or without concomitant autoimmune disease at baseline.
Table 2-.
Baseline clinical and laboratory characteristics of patients (N = 420) with stable or unstable clinically meaningful aPL profiles at follow-up.
| Total n=420 | Clinically Meaningful aPL Profile | P-value | ||
|---|---|---|---|---|
| Stable n = 366 | Unstable n = 54 | |||
| Female | 305 (73) | 267 (73) | 38 (70) | 0.74 |
| Age, yrs, median (IQR) | 48.9 (48.1–50.4) | 48.6 (47.9–49.4) | 48.6 (48–50) | 0.09 |
| White | 279 (78) | 238 (77) | 41 (87) | 0.30 |
| Non-Latin American | 165 (39) | 137 (37) | 28 (52) | 0.46 |
| Autoimmune disease | 0.76 | |||
| aPL/APS only | 278 (66) | 244 (67) | 34 (63) | |
| Other SAID | 148 (35) | 128 (35) | 20 (37) | |
| aPL-related history | ||||
| Vascular eventa (any) | 285 (68) | 245 (67) | 40 (74) | 0.35 |
| Venous event (any) | 183 (64) | 153 (62) | 30 (75) | 0.16 |
| Arterial event (any) | 125 (44) | 115 (47) | 10 (25) | 0.01 |
| TIA (any) | 38 (9) | 37 (10) | 1 (2) | 0.04 |
| Pregnancy morbidityb, n | 136 | 119 | 17 | 0.83 |
| Spontaneous abortionsc, n | 13 | 10 | 3 | 0.21 |
| Premature birthd, n | 37 | 34 | 3 | 0.56 |
| Unexplained fetal deathe, n | 76 | 67 | 9 | 0.80 |
| aPL tests | ||||
| LACf (+) | 319 (80) | 288 (83) | 31 (58) | < 0.001 |
| aCL IgG ≥ 40 U | 202 (48) | 183 (50) | 19 (35) | 0.06 |
| aCL IgM ≥ 40 U | 93 (22) | 89 (24) | 4 (7) | 0.004 |
| anti-β2-GPI IgG ≥ 40 U | 139 (33) | 130 (36) | 9 (17) | 0.005 |
| anti-β2-GPI IgM ≥ 40 U | 81 (19) | 76 (21) | 5 (9) | 0.06 |
| ≥ 2 positive aPL tests | 244 (58) | 226 (62) | 18 (33) | < 0.001 |
| aPL titers, U, median (IQR) | ||||
| aCL IgG | 36(10–93) | 46 (13–100) | 16 (4–56) | < 0.001 |
| aCL IgM | 12 (5–39) | 13 (5–42) | 8.5 (2–15.5) | 0.006 |
| anti-β2-GPI IgG | 19 (3–74) | 22 (3–83) | 3 (1–30) | < 0.001 |
| anti-β2-GPI IgM | 9 (2–33) | 10 (2–39) | 4 (1–20) | 0.04 |
| aPL profiles | ||||
| Triple aPL positivity | 174 (41) | 167 (46) | 7 (13)g | < 0.0001 |
| Double aPL positivityh | 120 (29) | 106 (29) | 13 (26) | 0.75 |
| Isolated LAC test positivity | 84 (20) | 62 (17) | 22 (41) | 0.0002 |
| Isolated aCL IgG/M positivity | 29 (7) | 23 (6) | 6 (11) | 0.24 |
| solated anti-β2-GPI IgG/M positivity | 13 (3) | 8 (2) | 5 (9) | 0.02 |
| Medications | ||||
| Aspirin | 201 (48) | 187 (51) | 14 (26) | < 0.001 |
| Warfarin | 223 (53) | 192 (52) | 31 (57) | 0.68 |
| Hydroxychloroquine | 194 (46) | 168 (46) | 26 (48) | 0.82 |
Values are n (%) unless otherwise indicated.
During an average follow-up of 5 years, new thrombosis occurred in 30 (7%) of420 patients (24 with history of baseline thrombosis, and 6 patients without), 29/30 had a stable clinically meaningful aPL profile at follow-up.
Out of 207 patients with history of pregnancy (with or without morbidity).
3 consecutive unexplained spontaneous abortions before 10th week.
Premature birth before 34th week due to eclampsia, preeclampsia, or placental insufficiency.
Unexplained fetal death at or beyond 10th week.
LAC test was reported by each center as positive or negative (screening by dilute Russell Viper Venom Time [dRVVT] and activated Partial Thromboplastin Time [aPTT] in 55% of patients, dRVVT in 28%, aPTT in 9%, other methods in 5%, and nonreported methods in 2%).
Seven triple aPL-positive patients with unstable aPL profile were mostly on warfarin with fluctuating LAC test status, and relatively low level aPL ELISA at baseline (2 patients only had IgM isotype).
Any combination of 2 positive aPL tests based on the laboratory criteria of the Updated Sapporo APS Classification Criteria. aCL: anticardiolipin antibody; anti-β GPI: anti-β glycoprotein I antibody; aPL: antiphospholipid antibody; APS: antiphospholipid syndrome; LAC: lupus anticoagulant; SAID: systemic autoimmune diseases; TIA: transient ischemic attack.
Predictors of an unstable aPL profile over time.
In a univariate unadjusted logistic model with unstable aPL profile as the outcome, triple aPL positivity at baseline was associated with a 75% decreased likelihood for unstable aPL profiles at follow-up (OR 0.25, 95% CI 0.1–0.6, P = 0.004; Table 3). Further, (1) patients with isolated LAC test positivity at baseline had 3.3-times higher odds for unstable aPL profiles (OR 3.3, 95% CI 1.5–7.1, P = 0.002), and (2) aCL or anti–β2-GPI IgG/M ≥ 40 U, but without any aCL/anti–β2-GPI positivity, was associated with lower odds of unstable aPL profiles over time. In a multivariable logistic model adjusted for age, sex, active smoking, concomitant autoimmune disease, and HCQ use at baseline: (1) triple aPL positivity at baseline was associated with lower odds of unstable aPL profiles (OR 0.17, 95% CI 0.1–0.4, P < 0.0001), and (2) isolated LAC test positivity and isolated anti–β2-GPI positivity at baseline were associated with higher odds of unstable aPL profiles (OR 3.65, 95% CI 1.9–6.8, P < 0.0001; OR 4.17, 95% CI 1.2–14.1, P = 0.02, respectively). When the multivariate model was further adjusted for individual aPL tests, (1) baseline LAC test positivity (OR 0.26, 95% CI 0.1–0.7, P = 0.005) and (2) baseline aCL IgG ≥ 40 U (OR 0.24, 95% CI 0.1–0.7, P = 0.006) were associated with lower odds of unstable aPL profiles over time.
Table 3-.
Baseline predictors of unstable aPL profile at follow-up.
| Univariate (Unadjusted) | OR (95% CI) | P |
|---|---|---|
| Baseline aPL profile | ||
| Triple aPL positive | 0.25 (0.10–0.64) | 0.004 |
| Double aPL positivea | 0.67 (0.31–1.46) | 0.32 |
| Isolated LAC test positivity | 3.30 (1.53–7.13) | 0.002 |
| Isolated aCL IgG/M positivityb,c | 2.13 (0.71–6.37) | 0.18 |
| Isolated anti-β2-GPI IgG/M positivityb,c | 2.31 (0.49–10.75) | 0.29 |
| Baseline individual aPL tests | ||
| LAC test positivity | 0.21 (0.11–0.41) | < 0.001 |
| aCL IgG ≥ 40 U | 0.21 (0.10–0.45) | < 0.001 |
| aCL IgM ≥ 40 U | 0.26 (0.09–0.73) | 0.01 |
| anti-β2-GPI IgG ≥ 40 U | 0.26 (0.11–0.65) | 0.004 |
| anti-β2-GPI IgM ≥ 40 U | 0.28 (0.08–0.93) | 0.04 |
| Age | 1.27 (0.98–1.63) | 0.06 |
| Sex, male | 0.70 (0.33–1.48) | 0.35 |
| HCQ use | 0.94 (0.47–1.87) | 0.85 |
| Autoimmune disease | 1.27 (0.60–2.67) | 0.54 |
| Active smoking | 1.56 (0.32–7.53) | 0.58 |
| Multivariable (Adjusted) | OR (95% CI) | P |
| Baseline aPL profile | ||
| Triple aPL positived | 0.17 (0.07–0.39) | < 0.0001 |
| Isolated LAC test positivityd | 3.65 (1.94–6.84) | < 0.0001 |
| Isolated anti-β2-GPI positivityd | 4.17 (1.24–14.1) | 0.02 |
| Baseline individual aPL tests | ||
| LAC test positivitye | 0.26 (0.10–0.66) | 0.005 |
| aCL IgG ≥ 40 Uf | 0.24 (0.09–0.66) | 0.006 |
Values in bold are statistically significant.
Any combination of 2 positive aPL tests based on the laboratory criteria of the Updated Sapporo APS Classification Criteria.
Any titer above normal range.
The small number of patients in isolated aCL and anti-β2-GPI aPL profile categories precluded from further analysis of isolated IgG or IgM isotype for these groups.
Adjusted for age, sex, active smoking, concomitant autoimmune disease, and HCQ use at baseline.
Adjusted for age, sex, active smoking, concomitant autoimmune disease, HCQ use at baseline, clinically meaningful (≥ 40 U) aCL IgG and IgM at baseline, and clinically meaningful (≥ 40 U) anti-β2-GPI IgG and IgM at baseline.
Adjusted for age, sex, active smoking, concomitant autoimmune disease, HCQ use at baseline, LAC test result at baseline, clinically meaningful (≥ 40 U) aCL IgM at baseline, and clinically meaningful (≥ 40 U) anti-β2-GPI IgG and IgM at baseline. aCL: anticardiolipin antibody; anti-β2-GPI: anti-β glycoprotein I antibody; aPL: antiphospholipid antibody; APS: antiphospholipid syndrome; HCQ: hydroxychloroquine; LAC: lupus anticoagulant.
Individual aPL result stability over time.
Table 4 describes the course of aCL and anti–β2-GPI IgG/M titers over time based on their assignments to 1 of the following categories at baseline and follow-up: 0–19 U, 20–39 U, 40–79 U, and ≥ 80 U. Approximately 90% and 60–80% of follow-up tests in patients with a baseline titer of 0–19 U and ≥ 80 U, respectively, remained in the same category. For baseline titers of 20–39 U and 40–79 U, during the follow-up, 23–30% and 19–33% remained in the same range, 36–60% and 41–65% decreased to a lower category, and 17–36% and 16–28% increased to a higher category, respectively. With respect to the LAC tests, 88% of patients with baseline isolated LAC positivity receiving no anticoagulation had a stable clinically meaningful profile at follow-up, compared to 52% on anticoagulation (OR 6.9, P = 0.009).
Table 4-.
Individual aPL course over time based on titers at baseline.
| aPL Titer at Follow-up, U, % | |||||||
|---|---|---|---|---|---|---|---|
| Baseline Titer, U | No. Patients at Baseline | No. Patients Follow-up aPL | 0–19 | 20–39 | 40–79 | ≥ 80 | |
| aCL IgG | 0–19 | 195 | 420 | 89 | 9 | 2 | 1 |
| aCL IgM | 281 | 652 | 91 | 5 | 4 | 1 | |
| Anti-β2-GPI IgG | 159 | 375 | 90 | 5 | 3 | 2 | |
| Anti-β2-GPI IgM | 206 | 477 | 94 | 3 | 1 | 1 | |
| aCL IgG | 20–39 | 53 | 145 | 41 | 23 | 24 | 12 |
| aCL IgM | 54 | 140 | 51 | 24 | 18 | 7 | |
| Anti-β2-GPI IgG | 34 | 83 | 36 | 30 | 19 | 14 | |
| Anti-β2-GPI IgM | 31 | 72 | 60 | 24 | 7 | 10 | |
| aCL IgG | 40–79 | 74 | 199 | 25 | 22 | 33 | 21 |
| aCL IgM | 49 | 113 | 29 | 14 | 33 | 24 | |
| Anti-β2-GPI IgG | 41 | 90 | 20 | 21 | 31 | 28 | |
| Anti-β2-GPI IgM | 17 | 37 | 51 | 14 | 19 | 16 | |
| aCL IgG | > 80 | 111 | 255 | 10 | 7 | 18 | 65 |
| aCL IgM | 41 | 104 | 6 | 4 | 29 | 62 | |
| Anti-β2-GPI IgG | 68 | 139 | 6 | 5 | 9 | 79 | |
| Anti-β2-GPI IgM | 40 | 90 | 11 | 8 | 11 | 70 | |
aPL: antiphospholipid antibody, aCL: anticardiolipin antibody; anti-β2-GPI: anti-β2 glycoprotein I antibody.
DISCUSSION
Our large-scale analysis of persistently positive aPL patients demonstrated that a clinically meaningful aPL profile, defined as a positive LAC test and/or aCL/anti–β2-GPI IgG/M ≥ 40 U, remains stable during a median follow-up of 5 years independent of age, active smoking, concomitant systemic autoimmune disease, and HCQ use at baseline. Triple aPL positivity increases and isolated LAC positivity decreases the odds of a stable aPL profile.
In patients with triple aPL positivity (LAC, aCL, and anti–β2-GPI of the same isotype) at initial aPL testing, 98% have been shown to have confirmed persistence of titers at 12 weeks, compared to 84% for double aPL positivity (aCL and anti–β2-GPI of the same isotype) and 40% for isolated aPL test positivity (LAC, aCL, or anti–β2-GPI)5. Based on a limited number of studies, 70–90% of patients with persistently positive aPL profiles remain positive during follow-up, ranging from 2 to 10 years6,7,8. In contrast, 1 study of 105 women with persistently positive aPL tests (49 with primary APS) found that in 59% of patients, the aPL profile become negative within approximately 10 years of follow-up9. The limitations of these studies include retrospective study designs with varying follow-up times and frequency of aPL tests, different cutoff levels used to define aPL positivity (≥ 20 or 40 U, or > 99th percentile of controls), and incomplete analysis of aPL profiles. Using a large, multicenter, international database of patients with persistently positive aPL profiles, we demonstrated that clinically meaningful aPL profiles remain stable over time at a median follow-up of 5 years. Our results are based on explicit and clinically relevant definitions of aPL profile positivity, and prospectively collected clinical and laboratory data.
Interpretation of aPL tests should be done cautiously since not every positive test is clinically important. Triple aPL positivity10,11 or LAC positivity12 is known to confer a higher risk for aPL-related clinical events compared to aCL and anti–β2-GPI positivity. Additionally, IgG aCL and anti–β2-GPI are more likely to be associated with clinical events compared to IgM13. The clinical significance of low titer aPL (20–39 U) should be interpreted carefully since it may be transient and associated with infectious triggers. Persistence of aPL positivity (when tested at least 12 weeks apart) and medium to high titers of aCL and anti–β2-GPI, as defined by the Updated Sapporo APS Classification Criteria1, are more likely to be associated with APS. To that point, our present study shows that patients who maintain a stable clinically meaningful aPL profile at 5 years of follow-up are more likely to have, at baseline, LAC test positivity, 2 or more positive aPL tests (including triple aPL positivity), and higher ELISA titers for aCL IgG that are clinically meaningful.
In our analysis, patients with a stable clinically meaningful aPL profile over time had more frequent history of arterial events (P = 0.01) or transient ischemic attacks (P = 0.04) at baseline and were more frequently taking aspirin (P < 0.001). One potential explanation could be the higher frequency of a triple positive aPL profile in patients with a stable aPL profile over time compared to those with unstable profiles (46% and 13%, respectively). Yet, venous events at baseline did not show a similar trend despite the triple positive aPL profile. This finding, if not occurring due to a relatively small number of patients in the unstable group, is hypothesis-generating and should be explored in future studies.
When determining predictors for an unstable aPL profile over time, we adjusted for various factors that have been implicated in maintenance of aPL test positivity. First, the use of HCQ was considered a potentially contributing factor, as a retrospective study has demonstrated that patients with systemic lupus erythematosus (SLE) and persistently positive aPL profiles (positive LAC and/or an aCL/anti–β2-GPI ≥ 40 U) were less likely to be taking HCQ compared to patients with transiently positive or negative profiles14; HCQ may also decrease aCL IgG/M levels and dRVVT (dilute Russell Viper Venom Time) prolongation15. Second, smoking was implicated in triggering aPL production, yet interpretation of relevant studies is difficult since smoking is a risk factor for thrombosis along with aPL16. Finally, we speculated that presence of concomitant autoimmune disease (such as SLE) may be associated with stable aPL tests since SLE is characterized by aberrant autoantibody production; a small study has supported that SLE activity was higher in patients with persistently positive aPL tests (LAC and aCL)17. Therefore, even after adjusting for age, sex, active smoking, concomitant autoimmune disease (mainly SLE), and HCQ use at baseline, triple aPL positivity was still 83% less likely to be associated with an unstable aPL profile.
The LAC test, when persistently positive, is highly associated with obstetric and thrombotic events. Despite guidelines, LAC results among laboratories may be discrepant due to lack of standardization and use of different screening tests. In addition, LAC results may be unreliable when tested on anticoagulation including direct oral anticoagulants (DOAC)18,19. An exercise among 4 different laboratories demonstrated that discordant or inconclusive LAC test results occur in 45% of patients with history of thrombosis or suspected APS, which increases to 75% when only patients on vitamin K antagonists are examined20. In our cohort, isolated LAC test positivity had significantly higher odds of being associated with an unstable aPL profile; we speculate that this finding was due to a relatively high number of anticoagulated patients. For more accurate assessment, future APS-ACTION studies will be completed using Core Laboratory LAC test results, which have been performed using methods with minimal interference with anticoagulation.
Our study has several limitations. First, we have missing follow-up data, as aPL testing was based on the discretion of the treating physician; however, we plan to reassess the aPL profiles in future studies using stored blood samples from each patient visit. Second, we could not assess the aPL profile stability in 11% of patients who had inconclusive aPL profiles in our cohort. A portion of these patients (24/52) could potentially have been added to the unstable aPL group; however, we wanted to avoid basing our results on the assumption that the rest of the aPL tests remained negative when no data were available. Third, median aCL/anti–β2-GPI titers may have been underestimated, as (1) for titers reported as “greater than x units” we used the upper limit, and (2) we used all available titers irrespective of positivity. Fourth, the association between stable aPL profile over time and aPL-related clinical events at follow-up was not formally explored as this was beyond the primary objective of the study, yet the available descriptive data show that among patients with thrombotic events at follow-up, 97% (29/30) had a stable, clinically meaningful aPL profile. Future analyses of the Registry will specifically address predictors of first and recurrent events in aPL-positive patients. Finally, referral bias may influence the generalizability of our findings.
Despite these limitations, the APS ACTION Registry comprises patients from tertiary referral centers across the world, and we believe that the large number of patient data provide a better understanding of aPL profile changes over time. The findings of this study are expected to inform and serve as a comparator for future validation studies of aPL profiles in stored blood samples of patients in the APS ACTION Registry, bypassing issues of assay and protocol heterogeneity among different laboratories across the world, interference of anticoagulation use at time of testing, and missing data.
In conclusion, using a large, multicenter, international database of patients with persistently positive aPL profiles, we demonstrated that the majority, approximately 80%, of clinically meaningful aPL profiles remain stable over time at a median follow-up of 5 years. These results will help guide future validation studies of stored blood samples through APS ACTION Core Laboratories.
ACKNOWLEDGEMENTS
We would like to thank JoAnn Vega for her research and administrative assistance including patient recruitment and data monitoring, and all members of APS ACTION for the valuable help with data acquisition. We would also like to thank Dr. Michael Lockshin and Dr. Medha Barbhaiya for their critical review of the study.
APPENDIX –. List of Study Collaborators
AntiPhospholipid Syndrome Alliance for Clinical Trials and InternatiOnal Networking (APS ACTION). Argentina: Santa Fe (Guillermo Pons-Estel); Australia: Sydney (Bill Giannakopoulos, Steve Krilis); Brazil: Rio de Janeiro (Guilherme de Jesus, Roger Levy), São Paulo (Danieli Andrade); Canada: Quebec (Paul F. Fortin); China: Beijing (Lanlan Ji, Zhouli Zhang); France: Nancy (Stephane Zuily, Denis Wahl); Italy: Brescia (Cecilia Nalli, Laura Andreoli, Angela Tincani), Milan (Cecilia B. Chighizola, Maria Gerosa, Pierluigi Meroni), Padova (Vittorio Pengo), Turin (Massimo Radin, Savino Sciascia); Jamaica: Kingston (Stacy Davis); Japan: Sapporo (Olga Amengual, Tatsuya Atsumi); Lebanon: Beirut (Imad Uthman); Netherlands: Utrecht (Maarten Limper, Philip de Groot); Spain: Barakaldo (Guillermo Ruiz Irastorza, Amaia Ugarte), Barcelona (Ignasi Rodríguez-Pinto, Roberto Ríos-Garcés, Ricard Cervera), Madrid (Esther Rodriguez, Maria Jose Cuadrado), Cordoba (Maria Angeles Aguirre Zamorano, Rosario Lopez-Pedrera); Turkey: Istanbul (Bahar Artim-Esen, Murat Inanc); United Kingdom: London (Maria Laura Bertolaccini, Hannah Cohen, Maria Efthymiou, Munther Khamashta, Ian Mackie, Giovanni Sanna); USA: Ann Arbor (Jason S. Knight), Baltimore (Michelle Petri), Chapel Hill (Robert Roubey), Durham (Tom Ortel), Galveston (Emilio Gonzalez, Rohan Willis), New York City (H. Michael Belmont, Steven Levine, Jacob Rand, Medha Barbhaiya, Doruk Erkan, Jane Salmon, Michael Lockshin), Salt Lake City (D. Ware Branch).
Footnotes
• The APS ACTION registry was created using REDCAP provided by the Clinical and Translational Science Center at Weill Cornell Medical College (CTSC grant UL1 TR000457). For this work, EG was supported by Clinical and Translational Science Center at Weill Cornell Medicine (UL1 TR002384).
• MP (Hopkins Lupus Cohort) is supported by a grant from the NIH RO1 AR069572. PF is recipient of a tier 1 Canada Research Chair on Systemic Autoimmune Rheumatic Diseases. MLB is supported by the King’s British Heart Foundation Centre for Research Excellence Award RE/18/2/34213. The rest of the authors report no conflicts of interest.
REFERENCES
- 1.↵.Miyakis S, Lockshin MD, Atsumi T, Branch DW, Brey RL, Cervera R, et al. International consensus statement on an update of the classification criteria for definite antiphospholipid syndrome (APS). J Thromb Haemost 2006;4:295–306. [DOI] [PubMed] [Google Scholar]
- 2.↵.Garcia D, Erkan D. Diagnosis and management of the antiphospholipid syndrome. N Engl J Med 2018;379:1290. [DOI] [PubMed] [Google Scholar]
- 3.↵.Levine SR, Salowich-Palm L, Sawaya KL, Perry M, Spencer HJ, Winkler HJ, et al. IgG anticardiolipin antibody titer > 40 GPL and the risk of subsequent thrombo-occlusive events and death. A prospective cohort study. Stroke 1997;28:1660–5. [DOI] [PubMed] [Google Scholar]
- 4.↵.Udry S, Latino JO, Belizna C, Perés Wingeyer S, Fernández Romero DS, de Larrañaga G. A high-risk laboratory profile of antiphospholipid antibodies and thrombosis is associated with a large number of extra-criteria manifestations in obstetric antiphospholipid syndrome. Immunol Res 2019;67:478–85. [DOI] [PubMed] [Google Scholar]
- 5.↵.Pengo V, Ruffatti A, Del Ross T, Tonello M, Cuffaro S, Hoxha A, et al. Confirmation of initial antiphospholipid antibody positivity depends on the antiphospholipid antibody profile. J Thromb Haemost 2013;11:1527–31. [DOI] [PubMed] [Google Scholar]
- 6.↵.Devignes J, Smaïl-Tabbone M, Hervé A, Cagninacci G, Devignes MD, Lecompte T, et al. Extended persistence of antiphospholipid antibodies beyond the 12-week time interval: association with baseline antiphospholipid antibodies titres. Int J Lab Hematol 2019;41:726–30. [DOI] [PubMed] [Google Scholar]
- 7.↵.Erkan D, Derksen WJM, Kaplan V, Sammaritano L, Pierangeli SS, Roubey R, et al. Real world experience with antiphospholipid antibody tests: how stable are results over time? Ann Rheum Dis 2005;64:1321–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.↵.Martinez-Berriotxoa A, Ruiz-Irastorza G, Egurbide MV, Garmendia M, Gabriel Erdozain J, Villar I, et al. Transiently positive anticardiolipin antibodies and risk of thrombosis in patients with systemic lupus erythematosus. Lupus 2007;16:810–6. [DOI] [PubMed] [Google Scholar]
- 9.↵.Riancho-Zarrabeitia L, Daroca G, Muñóz P, López-Hoyos M, Haya A, Martinez-Taboada VM. Serological evolution in women with positive antiphospholipid antibodies. Semin Arthritis Rheum 2017;47:397–402. [DOI] [PubMed] [Google Scholar]
- 10.↵.Pengo V, Ruffatti A, Legnani C, Gresele P, Barcellona D, Erba N, et al. Clinical course of high-risk patients diagnosed with antiphospholipid syndrome. J Thromb Haemost 2010;8:237–42. [DOI] [PubMed] [Google Scholar]
- 11.↵.Pengo V, Ruffatti A, Legnani C, Testa S, Fierro T, Marongiu F, et al. Incidence of a first thromboembolic event in asymptomatic carriers of high-risk antiphospholipid antibody profile: a multicenter prospective study. Blood 2011;118:4714–8. [DOI] [PubMed] [Google Scholar]
- 12.↵.Galli M, Luciani D, Bertolini G, Barbui T. Lupus anticoagulants are stronger risk factors for thrombosis than anticardiolipin antibodies in the antiphospholipid syndrome: a systematic review of the literature. Blood 2003;101:1827–32. [DOI] [PubMed] [Google Scholar]
- 13.↵.Kelchtermans H, Pelkmans L, de Laat B, Devreese KM. IgG/IgM antiphospholipid antibodies present in the classification criteria for the antiphospholipid syndrome: a critical review of their association with thrombosis. J Thromb Haemost 2016;14:1530–48. [DOI] [PubMed] [Google Scholar]
- 14.↵.Broder A, Putterman C. Hydroxychloroquine use is associated with lower odds of persistently positive antiphospholipid antibodies and/or lupus anticoagulant in systemic lupus erythematosus. J Rheumatol 2013;40:30–3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.↵.Petri M, Avci M, Magder LS. Hydroxychloroquine and prednisone have different effects on antiphospholipid antibodies in SLE, with hydroxychloroquine not reducing IgA anticardiolipin. Arthritis Rheumatol 2018;70. [Google Scholar]
- 16.↵.Binder SR, Litwin CM. Anti-phospholipid antibodies and smoking: an overview. Clin Rev Allergy Immunol 2017;53:1–13. [DOI] [PubMed] [Google Scholar]
- 17.↵.Sarabi ZS, Sahebari M, Rezaie AE, Norouzi MT, Hashemzadeh K, Mirfeizi Z. The relationship between systemic lupus erythematosus activity and persistent positive antiphospholipid antibodies. Curr Rheumatol Rev 2018;14:145–52. [DOI] [PubMed] [Google Scholar]
- 18.↵.Pengo V ISTH guidelines on lupus anticoagulant testing. Thromb Res 2012;130 Suppl 1:S76–7. [DOI] [PubMed] [Google Scholar]
- 19.↵.Favaloro EJ, Mohammed S, Curnow J, Pasalic L. Laboratory testing for lupus anticoagulant (LA) in patients taking direct oral anticoagulants (DOACs): potential for false positives and false negatives. Pathology 2019;51:292–300. [DOI] [PubMed] [Google Scholar]
- 20.↵.Sciascia S, Radin M, Cecchi I, Rubini E, Scotta A, Rolla R, et al. Reliability of lupus anticoagulant and anti-phosphatidylserine/prothrombin autoantibodies in antiphospholipid syndrome: a multicenter study. Front Immunol 2019;10:376. [DOI] [PMC free article] [PubMed] [Google Scholar]