Abstract
Objective:
Patients with ANCA-associated vasculitis (AAV) are at elevated risk for cardiovascular disease (CVD). A clearer understanding of the association between changes in disease activity and lipid levels in AAV would inform CVD risk stratification in this population.
Methods:
Lipid levels were assessed using baseline and month 6 stored serum samples from the Rituximab for ANCA-associated Vasculitis (RAVE) trial which randomized subjects to rituximab or cyclophosphamide followed by azathioprine. Paired t tests and multivariable linear regression were used to assess changes in lipid levels.
Results:
Of the 142 subjects with samples available, the mean age was 52.3 (±14.7) years, 72 (51%) were male, 95 (67%) were proteinase (PR3)-ANCA+, 72 (51%) had a new diagnosis of AAV, and 75 (53%) were treated with rituximab. Levels of several lipid levels increased between baseline and month 6, including total cholesterol (+12.4mg/dL, [+7.1, +21.0]), low-density lipoprotein (+10.3 mg/dL, [+6.1, +17.1]), and apolipoprotein B (+3.5 mg/dL, [+1.0, +8.3]). These changes were observed among newly-diagnosed and PR3-ANCA+ subjects but not among those with relapsing disease or myeloperoxidase (MPO)-ANCA+ subjects. There was no difference in changes in lipid levels between rituximab- and cyclophosphamide-treated patients. Changes in lipids correlated with changes in ESR, but not with other inflammatory markers or glucocorticoid exposure.
Conclusion:
Lipid levels increase during remission induction among patients with AAV with newly-diagnosed disease and those who are PR3-ANCA+. Disease activity and ANCA type should be considered when assessing lipid profiles to stratify CVD risk in patients with AAV.
Introduction:
ANCA-associated vasculitis (AAV) causes intense systemic inflammation and injury to predominantly small- and medium-sized vessels. As short-term survival in AAV has improved, cardiovascular disease (CVD) is increasingly recognized as a common cause of morbidity and mortality among patients with this disease (1–4). AAV is associated with a 2-fold higher risk of CVD when compared to age- and sex-matched controls (1, 4). Thus, periodic assessment of CVD risk, including lipid screening, is recommended for patients.
Lipid levels are known to be affected by inflammatory states in patients with other immune-mediated conditions (5–7), especially rheumatoid arthritis (RA). AAV is somewhat unique compared with RA because the endothelium is a primary target of the pathologic process (8, 9). In light of associations between lipids, endothelial cell dysfunction and damage (10), and CVD risk, understanding lipid level variation in AAV is important. However, data pertaining to lipid levels in AAV are sparse, especially regarding changes in lipid levels during the course of treatment.
We evaluated lipid parameters across clinically relevant disease subsets during remission induction in the RAVE trial.
Materials and Methods:
RAVE Trial:
Details of the RAVE trial design have been previously reported (11). Proteinase 3 (PR3)- or myeloperoxidase (MPO)-ANCA-positive patients with GPA or MPA and severe disease (Birmingham Vasculitis Activity Score for Wegener’s Granulomatosis [BVAS/WG] of > 3, or one major item) were assigned to either: 1) cyclophosphamide (CYC, 2mg/kg, adjusted for renal insufficiency) for 3–6 months, followed by azathioprine (AZA) (2mg/kg) for a total of 18 months; or, 2) rituximab (RTX, 4 weekly infusions of 375mg/m2) followed by placebo. Patients in both groups received the same glucocorticoid protocol, which included 1–3 days of IV methylprednisolone followed by 1mg per kilogram per day of prednisone. The prednisone dose was then tapered until discontinuation by 6 months if the patient had achieved and maintained remission. Patients who received glucocorticoids for longer than 14 days before screening, oral or intravenous cyclophosphamide within 4 months prior to enrollment, or previous therapy with RTX or alemtuzumab were excluded. Data for analysis of the RAVE trial was accessed from the Immune Tolerance Network (https://www.itntrialshare.org/, on October 17, 2018).
Covariates:
Age, sex, ANCA type (PR3- or MPO-ANCA+), BVAS/WG scores, inflammatory markers (C-reactive protein [CRP] and erythrocyte sedimentation rate [ESR]), body mass index (BMI), smoking status (never, former, current), baseline comorbidities, and disease status (new or relapsing at baseline) were collected during the trial. Methylprednisolone doses were converted to prednisone doses to calculate total glucocorticoid exposure by assuming that 4mg of methylprednisolone is equivalent to 5mg of prednisone. Interleukin (IL)-6 and soluble IL-6 receptors (sIL-6R) concentrations were measured separately, as reported elsewhere (12, 13). Statin use at baseline or between baseline and month 6 was assessed during the trial.
Research Specimens:
Stored serum samples available from the ITN021AI RAVE trial participants at baseline and/or month 6 were used. We excluded patients at baseline and/or month 6 if they had received statins prior to blood collection (N=7). Research specimens were provided by the Immune Tolerance Network and supported by the National Institute of Allergy and Infectious Diseases of the National Institutes of Health (UM1AI109565). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Lipid Measurement:
Total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), apolipoprotein (apo) A1, and apoB were measured according to standardized techniques in clinical laboratories (14, 15). Atherogenic indices (TC:HDL and apoB:A1 ratios) were calculated.
Statistical Analysis:
Continuous variables are reported as mean ± standard deviation (SD) or median and interquartile range (IQR), where appropriate. Baseline routine lipid parameters (TC, LDL-C, and HDL-C) were categorized according to levels recommended in lipid management guidelines.(16) We examined cross-sectional and longitudinal changes in lipid parameters in the entire study group as well as in subgroups representing distinct disease states (disease status at baseline [new vs relapsing] (17) and ANCA type [PR3- vs MPO-ANCA+]) (18–20) and treatment strategies (RTX vs CYC/AZA).
Paired t tests were used to assess the change in lipid levels between baseline and month 6. Paired t tests and multivariable linear regression models were used to compare baseline lipid levels between subgroups as well as the differences in changes between baseline and month 6 in lipid levels between subgroups. In multivariable models, covariates of interest included age, sex, baseline disease status, ANCA type, randomization arm, and glucocorticoid exposure during the screening period. The primary analyses did not account for glucocorticoid administration that occurred after baseline to avoid adjusting for causal intermediates of the exposures of interest (e.g., ANCA type, baseline disease status). However, glucocorticoid use between baseline and six months was included in secondary analyses to assess its impact on our results.
To determine the cross-sectional association between inflammation and lipid levels, we evaluated the correlation between each lipid and ESR, CRP, IL-6, and sIL-6R at baseline. We then added baseline ESR and IL-6, individually, to multivariable models comparing newly diagnosed and relapsing patients because they correlated with baseline lipid levels.
To determine the association between changes in inflammation and changes in lipid levels, we assessed the association of the difference between baseline and month 6 in each lipid level with the difference in each inflammatory marker using univariate linear regression. We also individually incorporated the difference in each inflammatory marker between baseline and month 6 into each multivariable model used in our primary analysis to evaluate whether differences across subgroups were associated with differences in changes in inflammatory markers.
A two-sided P value of < 0.05 was considered significant in all analyses. We used SAS, version 9.4 (SAS Institute, Cary, North Carolina, USA) for all statistical analyses.
Results:
There were 142 participants with available specimens (Table 1). At baseline, the mean (SD) age was 52.3 (±14.7) years and the majority were male (N=72, 51%), PR3-ANCA+ (N=95, 67%), and newly diagnosed (N=72, 51%). The median (IQR) baseline BVAS/WG was 8 (6, 10). There were 75 participants (53%) randomized to RTX and 67 (47%) randomized to CYC/AZA.
Table 1:
Baseline Demographics of Cohort
| Overall | Disease Status | ANCA Type | Treatment | ||||
|---|---|---|---|---|---|---|---|
| New | Relapsing | PR3+ | MPO+ | RTX | CYC/AZA | ||
| N | 142 | 72 (51%) |
70 (49%) |
95 (67%) |
47 (33%) |
75 (55%) |
67 (47%) |
| Age* | 52.3 (14.7) |
55.5 (14.7) |
49.0 (14.0) |
49.2 (14.2) |
58.5 (13.8) |
53.5 (15.6) |
51.0 (13.6) |
| Male | 72 (51%) |
38 (53%) |
34 (49%) |
55 (58%) |
17 (36%) |
33 (44%) |
39 (58%) |
| White | 130 (92%) |
65 (90%) |
65 (93%) |
88 (93%) |
42 (89%) |
67 (89%) |
63 (94%) |
| BMI | 28.8 (6.0) |
27.9 (5.3) |
29.8 (6.6) |
29.6 (6.2) |
27.3 (5.4) |
28.2 (6.4) |
29.5 (5.6) |
| AAV Type | |||||||
| GPA | 105 (74%) |
42 (58%) |
63 (90%) |
92 (97%) |
13 (28%) |
55 (73%) |
50 (75%) |
| MPA | 36 (25%) |
29 (40%) |
7 (10%) |
3 (3%) |
33 (70%) |
19 (25%) |
17 (25%) |
| Indeterminant | 1 (1%) |
1 (1%) |
0 (0%) |
0 (0%) |
1 (2%) |
1 (1%) |
0 (0%) |
| ANCA Type | |||||||
| PR3-ANCA+ | 95 (67%) |
38 (53%) |
57 (81%) |
95 (100%) |
0 (0%) |
52 (69%) |
43 (64%) |
| MPO-ANCA+ | 47 (33%) |
34 (47%) |
13 (19%) |
0 (0%) |
47 (100%) |
23 (31%) |
24 (36%) |
| Disease Characteristics | |||||||
| New Diagnosis | 72 (51%) |
72 (100%) |
0 (0%) |
38 (40%) |
34 (72%) |
35 (47%) |
37 (55%) |
| BVAS/WG | 8 (6, 10) |
8 (6, 11) |
7 (5, 10) |
8 (6, 10) |
8 (6, 10) |
8 (6, 10) |
8 (6, 10) |
| Renal Involvement | 95 (67%) |
55 (76%) |
40 (57%) |
58 (61%) |
37 (79%) |
48 (64%) |
47 (70%) |
| Creatinine | 1.41 (0.77) |
1.46 (0.73) |
1.33 (0.80) |
1.26 (0.69) |
1.66 (0.85) |
1.45 (0.85) |
1.34 (0.66) |
| Alveolar Hemorrhage | 30 (21%) |
15 (21%) |
15 (21%) |
22 (23%) |
8 (17%) |
15 (20%) |
15 (22%) |
| ESR (mm/Hr) | 41 (19, 61) |
53 (27, 83) |
36 (13, 53) |
34 (17, 59) |
51 (20, 65) |
34 (19, 58) |
51 (14, 82) |
| CRP (mg/dL) | 1.2 (0.5, 3.7) |
2.0 (0.8, 6.5) |
0.8 (0.4, 2.1) |
1.1 (0.4, 4.1) |
1.2 (0.6, 3.0) |
1.1 (0.5, 2.8) |
1.7 (0.4, 5.1) |
| IL-6 (pg/mL) | 2.9 (0.7, 20.9) |
2.6 (0.8, 21.9) |
3.1 (0.7, 20.4) |
4.9 (0.8, 21.2) |
1.2 (0.6, 10.8) |
3.4 (0.8, 29.9) |
2.2 (0.7, 17.6) |
| sIL-6R (ng/mL) | 25,910 (19,856, 38,300) |
28,636 (22,435, 45,975) |
24,859 (18,563, 29,991) |
25,197 (19,203, 33,306) |
28,069 (20,710, 43,206) |
25,317 (19,359, 41,036) |
26,464 (20,710, 34,291) |
| Treatment | |||||||
| Rituximab | 75 (53%) |
35 (49%) |
40 (57%) |
52 (55%) |
23 (49%) |
75 (100%) |
0 (0%) |
| CYC/AZA | 67 (47%) |
37 (51%) |
30 (43%) |
43 (45%) |
24 (51%) |
0 (0%) |
67 (100%) |
| Glucocorticoids (mg) at baseline | 1,127 (791) |
1,407 (234) |
918 (1,048) |
790 (575) |
1,380 (911) |
1,376 (789) |
505 (417) |
| Glucocorticoids (mg) at month 6 | 3,743 (640) |
3,938 (721) |
3,355 (173) |
3,789 (704) |
3,510 (1,259) |
4,017 (617) |
3,195 (49) |
| Comorbidities | |||||||
| Smoking | |||||||
| Current | 9 (6%) |
1 (1%) |
8 (11%) |
8 (8%) |
1 (2%) |
6 (8%) |
3 (4%) |
| Former | 43 (30%) |
26 (36%) |
17 (24%) |
30 (32%) |
13 (28%) |
20 (27%) |
23 (34%) |
| Never | 90 (63%) |
45 (63%) |
45 (64%) |
57 (60%) |
33 (70%) |
49 (65%) |
41 (61%) |
| Diabetes | 12 (8%) |
8 (11%) |
4 (6%) |
8 (8%) |
4 (9%) |
8 (11%) |
4 (6%) |
| Hypertension | 49 (35%) |
30 (42%) |
19 (27%) |
27 (28%) |
22 (47%) |
27 (36%) |
22 (33%) |
| Hyperlipidemia | 7 (5%) |
5 (7%) |
2 (3%) |
7 (7%) |
0 (0%) |
4 (5%) |
3 (4%) |
RTX: Rituximab; CYC/AZA: Cyclophosphamide followed by azathioprine
Continuous variables reported as mean (SD) except for ESR, CRP, IL-6, and sIL-6R which are reported as median (IQR)
Table 2 includes the distribution of baseline lipid levels in the entire study group. The baseline TC, HDL-C, and LDL-C were 166.1 (±37.2) mg/dL, 50.0 (±21.2) mg/dL, and 95.6 (±30.8) mg/dL, respectively. Most participants had a baseline TC (N=112, 79%), HDL-C (N=95, 67%), and LDL-C (N=139, 98%) in the normal range. At baseline, those with a new diagnosis of AAV had significantly lower TC (−20.5mg/dL, P=0.002), LDL-C (−16.0mg/dL, P=0.006), HDL-C (−12.5mg/dL, P<0.001), and apoA1 (−21.6mg/dL, P<0.001) compared to those enrolled with relapsing disease in adjusted analyses (Table 3). Relapsing patients had significantly lower apoB:apoA1 (−0.1, P=0.02) and TC:HDL-C (−0.6, P=0.02) at baseline. Adjustment for ESR attenuated the results such that differences in TC and LDL-C were no longer statistically significant but trends remained similar (−14mg/dL, P=0.06 and −10.9mg/dL, P=0.09, respectively); in contrast, adjusting for IL-6 did not affect the results. There were no differences in lipid levels at baseline between MPO-ANCA+ and PR3-ANCA+ patients or between those randomized to RTX and CYC/AZA (Table 3).
Table 2:
Baseline and Month 6 Lipid Parameters
| Lipid Parameter |
Baseline N=142 |
Month 6 N=142 |
Difference (95% CI) |
|---|---|---|---|
| TC | 166.1 (37.2) | 178.5 (43.8) | +12.4 (+7.1, +21.0)*** |
| HDL-C | 50.0 (21.2) | 49.4 (16.6) | −0.6 (−5.0, +2.1) |
| LDL-C | 95.6 (30.8) | 106.0 (35.6) | +10.3 (+6.1, +17.1)*** |
| ApoA1 | 121.4 (30.8) | 126.5 (26.7) | +5.1 (−1.4, +10.1) |
| ApoB | 89.6 (22.6) | 93.1 (25.8) | +3.5 (+1.0, +8.3)* |
| TC:HDL | 3.7 (1.4) | 4.0 (1.5) | +0.2 (+0.1, +0.6)* |
| B:A1 | 0.8 (0.3) | 0.8 (0.3) | −0.02 (−0.05, +0.04) |
Bold font indicates statistical significance
<0.05
<0.01
<0.001
Table 3:
Cross-sectional lipid parameters at baseline stratified by disease subsets
| Disease Status | ANCA Type | Treatment Group | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| New N=72 |
Relapse N=70 |
Unadjusted Difference (95% CI) |
Adjusted Difference† (95% CI) |
PR3+ | MPO+ | Unadjusted Difference (95% CI) |
Adjusted Difference† (95% CI) |
RTX | CYC/ AZA |
Unadjusted Difference (95% CI) |
Adjusted Difference† (95% CI) |
|
| TC | 157.3 (39.3) |
175.2 (32.9) |
−17.9 (−29.9, −5.8) |
−20.5** (−33.7, −7.4) |
165.3 (36.9) |
167.8 (38.3) |
−2.6 (−15.7, +10.6) |
−3.5 (−17.8, +10.9) |
167.8 (39.6) |
164.2 (34.6) |
+3.6 (−8.8, +16.0) |
−2.7 (−15.3, +9.9) |
| HDL-C | 46.2 (20.8) |
53.8 (21.1) |
−7.6 (−14.5, −0.6) |
−12.5*** (−19.3, −5.7) |
47.7 (17.8) |
54.5 (26.6) |
−6.8 (−14.2, +0.6) |
−5.8 (−13.3, +1.6) |
52.3 (21.8) |
47.3 (20.5) |
+5.0 (−2.0, +12.1) |
+1.8 (−4.8, +8.3) |
| LDL-C | 87.5 (33.6) |
103.9 (25.3) |
−16.4 (−26.2, −6.5) |
−16.0** (−27.3, −4.7) |
96.0 (29.0) |
94.8 (34.5) |
+1.3 (−9.6, +12.2) |
−2.3 (−14.6, +10.1) |
96.0 (32.2) |
95.2 (29.4) |
+0.7 (−9.5, +11.0) |
−2.9 (−13.8, +8.0) |
| ApoA1 | 115.1 (33.2) |
127.7 (28.9) |
−12.6 (−22.9, −2.3) |
−21.6*** (−31.4, −11.9) |
117.7 (30.3) |
128.7 (33.4) |
−10.9 (−22.0, +0.1) |
−8.5 (−19.2, +2.2) |
125.2 (33.1) |
117.1 (29.7) |
+8.1 (−2.4, +18.6) |
+2.3 (−7.1, +11.7) |
| ApoB | 86.3 (24.4) |
93.1 (20.2) |
−6.8 (−14.2, +0.6) |
−5.9 (−14.3, +2.5) |
90.6 (22.3) |
87.7 (23.4) |
+2.9 (−5.1, +10.9) |
+2.7 (−6.5, +11.9) |
89.5 (23.6) |
89.8 (21.6) |
−0.4 (−7.9, +7.2) |
−3.2 (−11.3, +4.9) |
| TC:HDL | 3.9 (1.6) |
3.6 (1.1) |
+0.3 (−0.1, +0.8) |
+0.6* (+0.1, +1.1) |
3.8 (1.3) |
3.6 (1.5) |
+0.2 (−0.2, +0.7) |
+0.2 (−0.4, +0.7) |
3.6 (1.3) |
3.9 (1.4) |
−0.3 (−0.7, +0.2) |
−0.2 (−0.7, +0.2) |
| B:A1 | 0.8 (0.3) |
0.8 (0.2) |
+0.03 (−0.1, +0.1) |
+0.1* (+0.02, +0.2) |
0.8 (0.3) |
0.7 (0.3) |
+0.1 (−0.01, +0.2) |
+0.1 (−0.02, +0.2) |
0.8 (0.3) |
0.8 (0.3) |
−0.1 (−0.1, +0.04) |
−0.05 (−0.1, +0.04) |
Adjusted for age, sex, ANCA-type, disease status, glucocorticoid exposure prior to baseline
Bold font indicates statistical significance
<0.05
<0.01
<0.001
Over the course of the study, there was a significant increase in the total cholesterol (+12.4 mg/dL, 95% CI: +7.1 to +21.0), LDL-C (+10.3 mg/dL, 95% CI: +6.1 to +17.1), apoB (+3.5 mg/dL, 95% CI: +1.0 to +8.3), and TC:HDL-C ratio (+0.2, 95% CI: +0.1 to +0.6) but the HDL-C, apoA1, and apoB:A1 ratio remained stable during the study period (Table 2). These changes in lipid levels differed between subgroups. In adjusted analyses, those with newly-diagnosed disease had significantly greater increases in TC (+22.3 mg/dL, 95% CI: +7.6 to +37.0), HDL-C (+11.1 mg/dL, 95% CI: +4.0 to +18.2), LDL-C (+17.6 mg/dL, 95% CI: +5.8 to +29.5), and apoA1 (+23.1 mg/dL, 95% CI: +11.5 to +34.7) than those with relapsing disease during remission induction (Table 4). At month six, there were no significant differences in lipid levels between the group with newly-diagnosed disease compared to that with relapsing disease at baseline. In adjusted analyses, PR3-ANCA+ patients had significantly greater increases in TC (+20.4 mg/dL, 95% CI: +4.4 to +36.4) and LDL-C (+14.2 mg/dL, 95% CI: +1.3 to +27.1) compared to MPO-ANCA+ (Table 5). There was no difference in the change in lipid levels during remission induction when comparing patients classified as having MPA versus GPA (data not shown). Rituximab- and cyclophosphamide/azathioprine-treated participants, had comparable changes in lipid levels during remission induction (Table 6).
Table 4:
Change in lipid parameters during remission induction stratified by disease status at baseline
| New Diagnosis | Relapsing Disease | New Diagnosis vs Relapsing Disease | ||||||
|---|---|---|---|---|---|---|---|---|
| Baseline | Month 6 | Difference (95% CI) |
Baseline | Month 6 | Difference (95% CI) |
Unadjusted Difference of Differences (95% CI) |
Adjusted† Difference of Differences (95% CI) |
|
| TC | 157.3 (39.3) | 178.5 (47.6) |
+21.1*** (+11.2, +33.4) |
175.2 (32.9) | 178.5 (40.0) | +3.3 (−2.4, +14.0) |
+17.8 (+0.4, +35.3) |
+22.3** (+7.6, +37.0) |
| HDL-C | 46.2 (20.8) | 48.9 (15.7) | +2.7 (−3.6, +7.9) |
53.8 (21.1) | 49.9 (17.5) |
−3.9** (−9.3, −0.8) |
+6.6 (+0.1, +13.1) |
+11.1** (+4.0, +18.2 ) |
| LDL-C | 87.5 (33.6) | 105.5 (37.3) |
+18.0*** (+10.9, +27.0) |
103.9 (25.3) | 106.4 (34.0) | +2.5 (−3.1, +11.7) |
+15.5 (+0.3, +30.7) |
+17.6** (+5.8, +29.5) |
| ApoA1 | 115.1 (33.2) | 128.0 (26.0) |
+12.9* (+2.9, +21.8) |
127.7 (28.9) | 124.9 (25.4) | −2.8 (−9.9, +2.6) |
+15.7 (+0.3, +31.1) |
+23.1*** (+11.5, +34.7) |
| ApoB | 86.3 (24.4) | 93.1 (27.0) |
+6.8** (+2.5, +12.8) |
93.1 (20.2) | 93.1 (24.8) | +0.02 (−3.7, +6.8) |
+6.8 (+0.1, +13.5) |
+7.3 (−0.6, +15.2) |
| TC:HDL | 3.9 (1.6) | 3.9 (1.3) | −0.01 (−0.3, +0.4) |
3.6 (1.1) | 4.0 (1.6) |
+0.4*** (+0.3, +0.9) |
−0.5 (−0.9, −0.009) |
−0.6* (−1.1, −0.1) |
| B:A1 | 0.8 (0.3) | 0.7 (0.2) | 0.1 (−0.1, +0.02) |
0.8 (0.2) | 0.8 (0.3) | +0.01 (−0.03, +0.09) |
−0.1 (−0.1, −0.001) |
−0.1* (−0.2, −0.03) |
Adjusted for age, sex, ANCA-type, glucocorticoid exposure prior to baseline, and randomization arm
Bold font indicates statistical significance
<0.05
<0.01
<0.001
Table 5:
Change in lipid parameters during remission induction stratified by ANCA type
| PR3-ANCA | MPO-ANCA | PR3- vs MPO-ANCA | ||||||
|---|---|---|---|---|---|---|---|---|
| Baseline | Month 6 | Difference (95% CI) |
Baseline | Month 6 | Difference (95% CI) |
Unadjusted Difference of Differences (95% CI) |
Adjusted† Difference of Differences (95% CI) |
|
| TC | 165.3 (36.9) | 181.2 (44.6) |
+15.9*** (+9.1, +26.4) |
167.8 (38.3) | 173.5 (42.1) | +5.7 (−4.6, +18.9) |
+10.2 (+0.2, +20.1) |
+20.4* (+4.4, +36.4) |
| HDL-C | 47.7 (17.8) | 48.2 (15.4) | +0.5 (−4.2, +4.0) |
54.5 (26.6) | 51.6 (18.5) | −2.9 (−11.0, +3.0) |
+3.4 (+0.1, +6.7) |
+7.6 (−0.1, +15.4) |
| LDL-C | 96.0 (29.0) | 108.5 (37.1) |
+12.4*** (+6.1, +20.7) |
94.8 (34.5) | 101.3 (32.4) | +6.6 (−0.1, +16.7) |
+5.9 (+0.1, +11.6) |
+14.2* (+1.3, +27.1) |
| ApoA1 | 117.7 (30.3) | 124.2 (23.3) | +6.5 (−0.8, +12.6) |
128.7 (33.4) | 130.6 (29.4) | +1.9 (−9.5, +12.5) |
+4.6 (+0.1, +9.1) |
+10.2 (−2.4, +22.8) |
| ApoB | 90.6 (22.3) | 95.2 (26.5) |
+4.6* (+0.9, +10.5) |
87.7 (23.4) | 89.3 (24.2) | +1.6 (−3.1, +8.3) |
+3.0 (+0.1, +5.9) |
+6.5 (−2.1, +15.1) |
| TC:HDL | 3.8 (1.3) | 4.1 (1.6) |
+0.3* (+0.1, +0.7) |
3.6 (1.5) | 3.7 (1.2) | +0.1 (−0.2, +0.6) |
+0.2 (+0.004, +0.4) |
+0.04 (−0.5, +0.6) |
| B:A1 | 0.8 (0.3) | 0.8 (0.3) | −0.02 (−0.1, +0.05) |
0.7 (0.3) | 0.7 (0.3) | −0.01 (−0.1, +0.1) |
−0.02 (−0.03, −0.0003) |
−0.04 (−0.1, +0.06) |
Adjusted for age, sex, disease status, glucocorticoid exposure prior to baseline, and randomization arm
Bold font indicates statistical significance
<0.05
<0.01
<0.001
Table 6:
Change in lipid parameters during remission induction stratified by treatment group
| RTX | CYC | RTX vs CYC | ||||||
|---|---|---|---|---|---|---|---|---|
| Baseline | Month 6 | Difference (95% CI) |
Baseline | Month 6 | Difference (95% CI) |
Unadjusted Difference of Differences (95% CI) |
Adjusted† Difference of Differences (95% CI) |
|
| TC | 167.8 (39.6) | 176.8 (47.4) |
+9.0* (+1.1, +20.7) |
164.2 (34.6) | 180.3 (39.5) |
+16.1*** (+7.6, +27.6) |
−7.1 (−14.0, −0.1) |
−3.8 (−18.0, +10.4) |
| HDL-C | 52.3 (21.8) | 49.4 (15.7) | −2.9 (−9.1, +0.2) |
47.3 (20.5) | 49.5 (17.7) | +2.1 (−3.7, +7.4) |
−5.1 (−10.0, −0.1) |
−5.1 (−11.9, +1.8) |
| LDL-C | 96.0 (32.2) | 104.4 (39.4) |
+8.5* (+2.4, +17.3) |
95.2 (29.4) | 107.7 (31.0) |
+12.5** (+5.2, +21.9) |
−4.0 (−7.9, −0.1) |
−1.4 (−12.8, +10.1) |
| ApoA1 | 125.2 (33.1) | 126.8 (23.4) | +1.6 (−6.6, +8.2) |
117.1 (29.7) | 126.1 (28.2) | +9.0 (−0.8, +17.4) |
−7.4 (−14.7, −0.1) |
−5.4 (−16.6, +5.8) |
| ApoB | 89.5 (23.6) | 91.8 (27.6) | +2.4 (−1.8, +8.8) |
89.8 (21.6) | 94.6 (23.8) |
+4.7* (+0.7, +11.0) |
−2.4 (−4.7, −0.05) |
−0.3 (−8.0, +7.3) |
| TC:HDL | 3.6 (1.3) | 3.9 (1.4) |
+0.3** (+0.1, +0.7) |
3.9 (1.4) | 4.1 (1.6) | +0.2 (−0.2, +0.7) |
+0.1 (+0.002, +0.1) |
+0.2 (−0.3, +0.7) |
| B:A1 | 0.8 (0.3) | 0.7 (0.2) | −0.02 (−0.1, +0.04) |
0.8 (0.3) | 0.8 (0.3) | −0.02 (−0.1, +0.1) |
+0.004 (+0.0001, +0.008) |
+0.01 (−0.08, +0.1) |
Adjusted for age, sex, ANCA-type, disease status, and glucocorticoid exposure prior to baseline
Bold font indicates statistical significance
<0.05
<0.01
<0.001
In univariate analyses, we found associations between changes in ESR (per 1mm/hr increase) with change in TC (β= − 0.4 mg/dL, P=0.003), LDL-C (β= − 0.3 mg/dL, P=0.01), HDL-C (β= − 0.3 mg/dL, P<0.001), apoA1 (β= − 0.6 mg/dL, P<0.001), apoB:A1 ratio (β= + 0.004 mg/dL, P<0.001) and TC:HDL-C ratio (β= + 0.02 mg/dL, P<0.001). We also observed associations between change in IL-6 (per 1pg/mL increase) with changes in the apoB:A1 ratio (β = +0.001, P=0.03) as well as the TC:HDL-C ratio (β= + 0.002, P=0.01) but not in other lipid levels. There was a trend toward a statistically significant association between change in CRP (per 1mg/dL increase) with change in LDL-C (β= − 1.93 mg/dL, P=0.06) but not other lipid levels. There were no associations between change in sIL-6R with changes in lipid levels.
We observed associations between change in BVAS/WG (per 1 unit increase) with changes in TC (β= − 3.6 mg/dL, P=0.001), LDL-C (β= - 3.02 mg/dL, P<0.001), HDL-C (β= − 1.8 mg/dL, P=0.001), apoA1 (β= − 2.6 mg/dL, P=0.004), and apoB:A1 ratio (β= + 0.02, P=0.02). A similar trend was observed in the TC:HDL-C ratio, though it did not achieve statistical significance (β= + 0.08, P=0.05).
Between baseline and month 6, newly-diagnosed patients had a significantly greater decrease in ESR and CRP than patients with relapsing disease (−38.6 [32.4] mm/hr vs. −12.9 [21.3] mm/hr, P<0.001, and −8.5 [28.2] mg/dL vs. −1.0 [3.4] mg/dL, P=0.037, respectively); there were no statistically significant differences in the change in IL-6 or sIL-6R between the two groups. Including change in ESR in the adjusted models (but not other inflammatory markers) fully explained the differences observed in lipid parameter changes when comparing newly diagnosed and relapsing patients as well as PR3- and MPO-ANCA+ patients.
Results from the primary analysis remained unchanged when glucocorticoid exposure between baseline and month 6 and BMI change between this same interval were included in the multivariable regression models.
Discussion:
We observed significant changes in lipid levels during the remission induction phase of AAV treatment, a period characterized by marked changes in disease activity and intensive immunosuppression. Newly-diagnosed patients with AAV and those with PR3-ANCA positivity showed the greatest changes in lipid levels between baseline and six months. The changes observed were particularly prominent in the serum concentrations of TC, LDL-C, and apoB. Our findings suggest that lipid profiles in AAV vary according to disease phase (e.g., new vs relapsing) as well as according to ANCA type. Changes in lipid levels were generally independent of treatment with either rituximab (RTX) or cyclophosphamide/azathioprine (CYC/AZA) and of changes in inflammatory markers. The exception to this observation was the relationship between ESR and lipid levels, highlighting an association between inflammation and lipid metabolism as well as the complexities of acute phase reactant changes.
To our knowledge, this is the first study to evaluate temporal changes in lipid levels in AAV and to compare the effects of RTX and CYC/AZA on lipid levels. Three prior cross-sectional studies with small sample sizes reported lipid levels in AAV. One found that total cholesterol and high-density lipoprotein were lower among incident cases of AAV compared to controls (21). In the other two studies, the timing of lipid assessment in relation to disease activity was not reported (22, 23). Smits et al. reported lipid parameters in an AAV cohort which were higher when compared to matched controls as well as those we report in this study. Their observations contrast with those reported in other inflammatory conditions (5, 24, 25). While these differences may reflect differences between clinical trial and clinic-based cohorts, our findings align well with the preponderance of data in other inflammatory conditions (5, 24, 25) as well as in a prior cross-sectional study evaluating CVD risk in a small AAV cohort (21).
Our findings add to a growing body of literature describing an inverse relationship between lipid levels and inflammation (5, 24, 25). Though lower lipid levels are often thought to reflect a lower risk of CVD, previous studies suggest that the association between LDL-C with CVD risk is actually U-shaped, such that those with very low LDL-C are also at increased CVD risk (e.g., the “lipid paradox”) (26, 27). As such, clarifying the association between lipid levels and inflammation in AAV is an important step towards understanding CVD risk and risk stratification in AAV.
This is the first study to report that changes in lipid parameters are restricted to certain disease subsets, specifically PR3-ANCA+ patients and those who are newly diagnosed. The differences in regard to ANCA type appear to reflect differences in inflammatory states (especially as indicated by the impact of adjustment for the ESR) between PR3- and MPO-ANCA+ patients and in the context of the differences already recognized between these two types of AAV, our findings augment a growing body of literature describing important differences between these disease subsets. The differences in lipid trajectories between newly diagnosed and relapsing patients resemble our previous findings regarding BMI changes in the RAVE trial (17). In that study, newly diagnosed patients had greater increases in their BMI during the first 6 months of the trial when compared to those with relapsing disease.
The findings regarding BMI and lipid parameter changes suggest important differences in the inflammatory state of patients with newly-diagnosed AAV compared to those with relapsing disease. These differences may pertain to the duration of active disease in newly diagnosed patients, possibly longer on average than those with relapsing disease, who are typically watched carefully for disease recurrence. They may also relate to the severity of disease activity at baseline; critical illness, for instance, is known to be associated with lower LDL. Patients with relapsing disease at baseline typically have disease flares that are less severe than those whose diagnoses have just been established. Indeed, this is reflected by the lower inflammatory markers and BVAS/WG scores at baseline among the relapsing disease patients. The effect of this more intense inflammation at baseline among patients with newly-diagnosed disease appears to have a substantial effect on lipid metabolism in these patients.
While details regarding the duration of symptoms before the baseline visit were not available in this study (28), the differences in associations between ESR and other acute phase reactants with lipid parameters at baseline and over the course of the study support this hypothesis. In contrast to CRP, IL-6, and sIL-6R, ESR is an indirect measure of inflammation that changes more slowly in response to changes in the inflammatory state (29). Since many patients were exposed to up to two weeks of glucocorticoids at baseline, CRP, IL-6, and sIL-6R at that visit may not have reflected the recent inflammatory state as closely as the ESR. Of note, IL-6 was recently found to play a key role in lipid metabolism in the setting of inflammation (30), but we were unable to detect associations between IL-6 and lipid levels in our study, likely because of the quick effects of glucocorticoids on acute phase reactants.
Our findings have implications for the assessment of CVD risk in AAV given that AAV patients are increased risk for CVD (1–4) and routine CVD risk assessments are recommended (1). Inflammatory states, including AAV, are associated with changes in lipid levels, but little is known about how to interpret the results of lipid screening in such patients. Postponing lipid screening as part of cardiovascular risk stratification efforts until after periods of disease activity have subsided is one consideration but it is unclear if lipid levels at this later timepoint are associated with CVD risk. For instance, following an acute myocardial infarction (31), it is recommended that lipids not be measured more than 48 hours after presentation because of differences in steady state and the effects of acute inflammation. Regardless, periodic cardiovascular risk assessment should be considered in AAV given the increased risk of CVD in this population (1–4).
Our study has certain limitations. First, this was a post-hoc analysis of a clinical trial, which may limit its generalizability. However, RAVE trial participants were representative of patients with severe AAV and our observations resemble those of other studies of inflammatory conditions using non-clinical trial populations. Using a clinical trial population permitted us to account for covariates of interest in a rigorous manner, including statin use, glucocorticoid exposure, and acute phase reactants. Second, we did not have knowledge of lipid profiles prior to the onset of AAV so we cannot comment on whether the increase in lipids represents a return to pre-disease levels. Third, while we did not find significant differences in the association between lipid level changes and treatment with RTX or CYC/AZA, the significant glucocorticoid exposure during the first six months of the trial may have had a greater impact on lipids than either RTX or CYC/AZA. Fourth, fasting status was not recorded at study visits but fasting status is not thought to significantly affect these lipid level measurements (32). Moreover, fasting status was unlikely to vary across disease subsets or timepoints so our conclusions regarding differences and changes in lipid levels would not be affected.
In conclusion, we found that lipid levels significantly increase during remission induction, especially among those with newly diagnosed disease and those who are PR3-ANCA+. These differences suggest that lipid metabolism differs across AAV disease subsets. Disease activity and timing during the period of treatment should be considered when screening for lipid disorders in AAV patients and assessing CVD risk.
Acknowledgments
Funding Support: Dr. Wallace receives funding from NIH/NIAMS (K23AR073334 and L30 AR070520) and the Rheumatology Research Foundation (Scientist Development Award). Dr. Liao is supported by NIH R01HL127118. Aspects of the data presented in this manuscript were collected using support from the Vasculitis Clinical Research Consortium, which has received support from the NIH/NIAMS (U54-AR-057319, RC1-AR-058303, and P60-AR-047785), the National Center for Research Resources (U54-RR-019497), the National Institute of Neurological Disorders and Stroke (NS-064808), and the Office of Rare Diseases Research.
Footnotes
Financial Disclosures: None
References:
- 1.Yates M, Watts RA, Bajema IM, Cid MC, Crestani B, Hauser T, et al. EULAR/ERA-EDTA recommendations for the management of ANCA-associated vasculitis. Ann Rheum Dis. 2016;75(9):1583–94. [DOI] [PubMed] [Google Scholar]
- 2.Morgan MD, Turnbull J, Selamet U, Kaur-Hayer M, Nightingale P, Ferro CJ, et al. Increased incidence of cardiovascular events in patients with antineutrophil cytoplasmic antibody-associated vasculitides: a matched-pair cohort study. Arthritis Rheum. 2009;60(11):3493–500. [DOI] [PubMed] [Google Scholar]
- 3.Faurschou M, Mellemkjaer L, Sorensen IJ, Svalgaard Thomsen B, Dreyer L, Baslund B. Increased morbidity from ischemic heart disease in patients with Wegener’s granulomatosis. Arthritis Rheum. 2009;60(4):1187–92. [DOI] [PubMed] [Google Scholar]
- 4.Houben E, Penne EL, Voskuyl AE, van der Heijden JW, Otten RHJ, Boers M, et al. Cardiovascular events in anti-neutrophil cytoplasmic antibody-associated vasculitis: a meta-analysis of observational studies. Rheumatology (Oxford). 2018;57(3):555–62. [DOI] [PubMed] [Google Scholar]
- 5.Navarro-Millan I, Charles-Schoeman C, Yang S, Bathon JM, Bridges SL Jr., Chen L, et al. Changes in lipoproteins associated with methotrexate or combination therapy in early rheumatoid arthritis: results from the treatment of early rheumatoid arthritis trial. Arthritis Rheum. 2013;65(6):1430–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Kang EH, Liao KP, Kim SC. Cardiovascular Safety of Biologics and JAK Inhibitors in Patients with Rheumatoid Arthritis. Curr Rheumatol Rep. 2018;20(7):42. [DOI] [PubMed] [Google Scholar]
- 7.Liao KP, Cai T, Gainer V, Cagan A, Murphy SN, Liu C, et al. Lipid and lipoprotein levels and trend in rheumatoid arthritis compared to the general population. Arthritis Care Res (Hoboken). 2013;65(12):2046–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Bosch X, Guilabert A, Font J. Antineutrophil cytoplasmic antibodies. Lancet. 2006;368(9533):404–18. [DOI] [PubMed] [Google Scholar]
- 9.Halbwachs L, Lesavre P. Endothelium-neutrophil interactions in ANCA-associated diseases. J Am Soc Nephrol. 2012;23(9):1449–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Booth AD, Almond MK, Burns A, Ellis P, Gaskin G, Neild GH, et al. Outcome of ANCA-associated renal vasculitis: a 5-year retrospective study. Am J Kidney Dis. 2003;41(4):776–84. [DOI] [PubMed] [Google Scholar]
- 11.Stone JH, Merkel PA, Spiera R, Seo P, Langford CA, Hoffman GS, et al. Rituximab versus cyclophosphamide for ANCA-associated vasculitis. N Engl J Med. 2010;363(3):221–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Monach PA, Warner RL, Tomasson G, Specks U, Stone JH, Ding L, et al. Serum proteins reflecting inflammation, injury and repair as biomarkers of disease activity in ANCA-associated vasculitis. Ann Rheum Dis. 2013;72(8):1342–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Berti A, Warner R, Johnson K, Cornec D, Schroeder D, Kabat B, et al. Brief Report: Circulating Cytokine Profiles and Antineutrophil Cytoplasmic Antibody Specificity in Patients With Antineutrophil Cytoplasmic Antibody-Associated Vasculitis. Arthritis Rheumatol. 2018;70(7):1114–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Rifai N, Cole TG, Iannotti E, Law T, Macke M, Miller R, et al. Assessment of interlaboratory performance in external proficiency testing programs with a direct HDL-cholesterol assay. Clin Chem. 1998;44(7):1452–8. [PubMed] [Google Scholar]
- 15.Rifai N, Iannotti E, DeAngelis K, Law T. Analytical and clinical performance of a homogeneous enzymatic LDL-cholesterol assay compared with the ultracentrifugation-dextran sulfate-Mg2+ method. Clin Chem. 1998;44(6 Pt 1):1242–50. [PubMed] [Google Scholar]
- 16.Program NCE. ATP III Guidelines At-A-Glance: Quick Desk Reference. In: SERVICES USDOHAH, Service PH, Health NIo, National Heart L, and Blood Institute, eds; 2001. [Google Scholar]
- 17.Wallace ZS, Miloslavsky EM, Cascino M, Unizony SH, Lu N, Hoffman GS, et al. Effect of Disease Activity, Glucocorticoid Exposure, and Rituximab on Body Composition During Induction Treatment of Antineutrophil Cytoplasmic Antibody-Associated Vasculitis. Arthritis Care Res (Hoboken). 2017;69(7):1004–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Lyons PA, Rayner TF, Trivedi S, Holle JU, Watts RA, Jayne DR, et al. Genetically distinct subsets within ANCA-associated vasculitis. N Engl J Med. 2012;367(3):214–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Hilhorst M, van Paassen P, Tervaert JW, Limburg Renal R. Proteinase 3-ANCA Vasculitis versus Myeloperoxidase-ANCA Vasculitis. J Am Soc Nephrol. 2015;26(10):2314–27. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Unizony S, Villarreal M, Miloslavsky EM, Lu N, Merkel PA, Spiera R, et al. Clinical outcomes of treatment of anti-neutrophil cytoplasmic antibody (ANCA)-associated vasculitis based on ANCA type. Ann Rheum Dis. 2015;75(6):1166–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Berti A, Matteson EL, Crowson CS, Specks U, Cornec D. Risk of Cardiovascular Disease and Venous Thromboembolism Among Patients With Incident ANCA-Associated Vasculitis: A 20-Year Population-Based Cohort Study. Mayo Clin Proc. 2018;93(5):597–606. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Bramlage CP, Kroplin J, Wallbach M, Minguet J, Smith KH, Luders S, et al. Management of cardiovascular risk factors in patients with ANCA-associated vasculitis. J Eval Clin Pract. 2017;23(4):747–54. [DOI] [PubMed] [Google Scholar]
- 23.Petermann Smits DR, Wilde B, Kianersi Adegani M, de Jongh H, Van Paassen P, Cohen Tervaert JW. Metabolic syndrome in ANCA-associated vasculitis. Rheumatology (Oxford). 2013;52(1):197–203. [DOI] [PubMed] [Google Scholar]
- 24.Kirkham BW, Wasko MC, Hsia EC, Fleischmann R, Genovese MC, Matteson EL, et al. Effects of golimumab, an anti-tumour necrosis factor-α human monoclonal antibody, on lipids and markers of inflammation. Ann Rheum Dis. 2014;73(1):161–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Gabay C, McInnes IB, Kavanaugh A, Tuckwell K, Klearman M, Pulley J, et al. Comparison of lipid and lipid-associated cardiovascular risk marker changes after treatment with tocilizumab or adalimumab in patients with rheumatoid arthritis. Ann Rheum Dis. 2016;75(10):1806–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Liao KP, Liu J, Lu B, Solomon DH, Kim SC. Association between lipid levels and major adverse cardiovascular events in rheumatoid arthritis compared to non-rheumatoid arthritis patients. Arthritis Rheumatol. 2015;67(8):2004–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Myasoedova E, Crowson CS, Kremers HM, Roger VL, Fitz-Gibbon PD, Therneau TM, et al. Lipid paradox in rheumatoid arthritis: the impact of serum lipid measures and systemic inflammation on the risk of cardiovascular disease. Ann Rheum Dis. 2011;70(3):482–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Green P, Theilla M, Singer P. Lipid metabolism in critical illness. Curr Opin Clin Nutr Metab Care. 2016;19(2):11105. [DOI] [PubMed] [Google Scholar]
- 29.Poulton CJ, Nachman P, Hu Y, McGregor JA, Jennette CE, Falk RJ, et al. Pathways to renal biopsy and diagnosis among patients with ANCA small-vessel vasculitis. Clin Exp Rheumatol. 2013;31(Suppl. 75):S32–S7. [PMC free article] [PubMed] [Google Scholar]
- 30.Gabay C, Kushner I. Acute-phase proteins and other systemic responses to inflammation. N Engl J Med. 1999;340(6):448–54. [DOI] [PubMed] [Google Scholar]
- 31.Robertson J, Porter D, Sattar N, Packard CJ, Caslake M, McInnes IB, et al. Interleukin-6 blockade raises LDL via reduced catabolism rather than via increased synthesis: a cytokine-specific mechanism for cholesterol changes in rheumatoid arthritis. Ann Rheum Dis. 2017;76(11):1949–52. [DOI] [PubMed] [Google Scholar]
- 32.Rosenson RS. Myocardial injury: the acute phase response and lipoprotein metabolism. J Am Coll Cardiol. 1993;22(3):933–40. [DOI] [PubMed] [Google Scholar]
- 33.Sidhu D, Naugler C. Fasting time and lipid levels in a community-based population: a cross-sectional study. Arch Intern Med. 2012;172(22):1707–10. [DOI] [PubMed] [Google Scholar]