Patient characteristics
A total of 7954 patients with a diagnosis of SLE were identified, with 879 males (11.1%) and 7075 females. We excluded 266 patients who had ever been diagnosed with other rheumatoid diseases. 19 patients received a diagnosis of HM posterior to or synchronously with SLE. In order to control confounding factors, 136 patients ever diagnosed with other kinds of malignancies were excluded, so there were 7533 cancer-free patients eligible for analysis. With three controls matching one case, we enrolled 19 patients as the case group and 57 as the control group (Group B) (Fig 1). The data of 4 patients with HM when first diagnosed with SLE were incomplete, so only 15 were enrolled in analysis (Group A) at the baseline (Fig 1, Table 1-2). Demographic and clinical characteristics are shown in Table 1. Twelve (80%) were female in Group A, with no difference from Group B (86%). Age for patients in Group A were higher than that in Group B (52 (42-63) vs 31 (25-47), P = 0.002). None was reported a previous history of HM, while 10 (66.7%) patients developed with HM synchronously with and 5 (33.3%) posterior to SLE. Though only 6 patients reported family history of tumor, the frequency was higher in Group A than that in Group B (13.3% vs 7.0%, P = 0.034). For clinical features, patients in Group A had a higher frequency of infection (60.0% vs 22.8%, P = 0.014) and splenomegaly (46.7% vs 15.8%, P = 0.027) than Group B. Infection always involved respiratory system except for one suffering from panniculitis in the control group (data not shown).
Table 1.
Clinical features of SLE patients with/without hematological malignancies
Parameter
|
Group A
(n=15)
|
Group B
(n=57)
|
P -value
|
Female (n, %)
|
12 (80.0)
|
49(86.0)
|
0.867
|
Age at SLE diagnosis (years), M (Q1-Q3)
|
52 (42-63)
|
31 (25-47)
|
0.002*
|
Age at HM diagnosis (years, mean ± SD)
|
52±15
|
-
|
-
|
Lymphadenopathy (n, %)
|
9 (60.0)
|
26(45.6)
|
0.321
|
Fever (n, %)
|
9 (60.0)
|
21(36.8)
|
0.106
|
Infection (n, %)
|
9 (60.0)
|
13(22.8)
|
0.014*
|
Hypertension (n, %)
|
1 (6.7)
|
9(15.8)
|
0.625
|
Diabetes mellitus (n, %)
|
0 (0)
|
1(1.8)
|
1.000
|
Dyslipidemia (n,%)
|
13 (13/14, 92.9)
|
40 (40/56, 71.4)
|
0.186
|
Smoking (n, %)
|
1 (6.7)
|
5(8.8)
|
1.000
|
Drinking (n, %)
|
1 (6.7)
|
2(3.5)
|
1.000
|
Family history of tumor (n, %)
|
2 (13.3)
|
4(7.0)
|
0.034*
|
Rash (n, %)
|
4 (26.7)
|
20(35.1)
|
0.538
|
Arthralgia (n, %)
|
5 (33.3)
|
14(24.6)
|
0.721
|
Pleural effusion (n, %)
|
5 (33.3)
|
15(26.3)
|
0.829
|
Pericardial effusion (n, %)
|
5 (33.3)
|
16(29.1)
|
0.936
|
Splenomegaly (n, %)
|
7 (46.7)
|
9(15.8)
|
0.027*
|
* Statistical significance (P<0.05)
Group A: SLE patients with hematological malignancies
Group B: SLE patients without hematological malignancies
Abbreviation: HM, hematological malignancies.
Table 2.
Laboratory parameters of SLE patients with/without HM at the baseline
Parameters
|
Group A
(n=15)
|
Group B
(n=57)
|
P -value
|
Hematological abnormality (n, %)
|
13 (86.7)
|
39 (68.4)
|
0.280
|
WBC (ⅹ109), M (Q1-Q3)
|
3.40 (1.80-5.68)
|
3.90 (2.85-6.25)
|
0.072
|
RBC (ⅹ1012, mean ± SD)
|
3.08±0.88
|
3.59±0.64
|
0.016*
|
Hb (g/L, mean ± SD )
|
89.97±23.18
|
103.43±20.41
|
0.031*
|
Platelet (ⅹ109, mean ± SD)
|
118.80±64.68
|
160.54±82.66
|
0.074
|
Urine protein positivity (n, %)
|
5 (33.3)
|
27 (47.4)
|
0.330
|
24hTP(g), M (Q1-Q3) †
|
0.49 (0.29-1.13)
|
1.40 (0.36-4.92)
|
0.156
|
Scr (umol/L), M (Q1-Q3)
|
62.00 (47.50-73.50)
|
55.00 (47.50-73.50)
|
0.856
|
GFR (ml/min/1.73m2), M (Q1-Q3)
|
85.57 (77.04-109.31)
|
108.76 (91.70-121.71)
|
0.099
|
TP (g/L, mean ± SD)
|
66.88±11.74
|
64.72±13.16
|
0.567
|
Albumin (g/L, mean ± SD)
|
29.70±6.69
|
31.72±7.58
|
0.351
|
Globin (g/L, mean ± SD)
|
37.18±13.39
|
33.33±9.64
|
0.310
|
Lupus nephritis (n, %)
|
5 (33.3)
|
28 (49.1)
|
0.275
|
TC (mmol/L), M (Q1-Q3)
|
3.05 (2.54-3.45)
|
3.91 (3.09-4.56)
|
0.030*
|
TG (mmol/L), M (Q1-Q3)
|
1.33 (0.75-2.02)
|
1.64 (1.20-2.37)
|
0.157
|
HDL (U/L), M (Q1-Q3)
|
0.61 (0.42-0.69)
|
0.96 (0.77-1.16)
|
0.003*
|
LDL (U/L), M (Q1-Q3)
|
1.74 (1.11-2.34)
|
2.39 (1.67-2.95)
|
0.058
|
ALT (U/L), M (Q1-Q3)
|
24.00 (10.00-35.00)
|
23.00 (13.00-49.00)
|
0.682
|
AST (U/L), M (Q1-Q3)
|
29.00 (17.00-58.00)
|
28.00 (18.00-56.50)
|
0.950
|
LDH(U/L), M (Q1-Q3)
|
262.00 (197.00-453.00)
|
273.00 (206.00-614.00)
|
0.767
|
EBV-IgM+(n, %) ††
|
0 (0)
|
2 (10)
|
0.975
|
EBV-IgG+(n, %)
|
7 (100)
|
20 (100)
|
-
|
C3 (g/L, mean ± SD)
|
0.78±0.35
|
0.67±0.36
|
0.314
|
C4 (g/L, mean ± SD)
|
0.16±0.12
|
0.13±0.08
|
0.401
|
ESR (mm/h), M (Q1-Q3)
|
48.00 (34.50-100.25)
|
41.50 (19.75-74.00)
|
0.533
|
CRP (mg/L), M (Q1-Q3)
|
16.31 (2.82-34.71)
|
3.88 (1.50-7.17)
|
0.018*
|
Ferritin (ng/ml), M (Q1-Q3) †††
|
692.45(475.20-1248.28)
|
188.85 (71.92-369.53)
|
0.007*
|
SLEDAI-2K, M (Q1-Q3)
|
9.00 (4.00-14.00)
|
12.00 (8.00-18.00)
|
0.184
|
ANA (n, %)
|
15 (100)
|
57 (100)
|
-
|
Anti-dsDNA antibody (n, %)
|
4 (26.7)
|
38 (66.7)
|
0.005*
|
Anti-Smith antibody (n, %)
|
5 (5/11, 45.5)
|
19(19/50, 38.0)
|
0.907
|
Anti-Rib P (n, %)
|
4(4/14, 28.6)
|
26(26/54, 48.1)
|
0.189
|
Anti-Nuc (n, %)
|
5(5/12, 41.7)
|
31 (31/55, 56.4)
|
0.355
|
Anti-His (n, %)
|
3(3/13, 23.1)
|
25(25/54, 46.3)
|
0.128
|
Anti-SSA/Ro52 (n, %)
|
9(9/13, 69.2)
|
24(24/52, 46.2)
|
0.137
|
Anti-SSA/Ro60 (n, %)
|
8(8/12, 66.7)
|
26(26/49, 53.1)
|
0.395
|
Anti-SSB (n, %)
|
4(4/12, 33.3)
|
9(9/53, 17.0)
|
0.379
|
ACA (n, %)
|
2(2/8, 25.0)
|
2(2/49, 4.1)
|
0.161
|
APL (n, %)
|
4(4/7, 57.1)
|
15(15/27, 55.6)
|
1.000
|
* Statistical significance (P<0.05)
Group A: SLE patients with hematological malignancies
Group B: SLE patients without hematological malignancies
Hematological abnormality: Patients present with leukopenia (WBC<3.0 ⅹ109), anemia or thrombocytopenia.
Lupus nephritis: 24-hour total urinary protein ≥0.5g or the confirmation of renal biopsy before the end of follow-up.
† The data of Group A (7 patients) and Group B (29 patients).
†† EBV-IgM and EBV-IgG were both the data of Group A (7 patients) and Group B (20 patients).
††† The data of Group A (6 patients) and Group B (20 patients).
Abbreviation: HM, hematological malignancies; WBC, white blood cell; RBC, red blood cell; Hb, hemoglobin; 24hTP, 24-hour total urinary protein; Scr, serum creatinine; GFR, glomerular filtration rate; TP, total protein; TC, total cholesterol; TG, triglyceride; HDL, high-density lipoprotein; LDL, low-density lipoprotein; ALT, alanine transaminase; AST, aspartate transaminase; LDH, lactic dehydrogenase; EBV, Epstein-Barr virus; CMV, cytomegalovirus; ESR, erythrocyte sedimentation rate; CRP, C-reactive protein; SLEDAI-2K, systemic lupus erythematosus disease activity index-2000; ANA, anti-nuclear antibody; Anti-Rib P, Anti-ribosomal P-protein antibody; Anti-Nuc, Anti-nucleosome antibody; Anti-His, Anti-histone antibody; ACA, Anti-centromere antibody; APL, Anti-phospholipid antibody.
Laboratory findings
Baseline laboratory parameters of SLE patients with or without HM were shown in Table 2. Majority of patients in both groups suffered from hematological abnormality (86.7% vs 68.4%, P = 0.280), including leukocytopenia, thrombocytopenia or anemia. Both groups tended to suffer from anemia while patients in Group A had a lower level of hemoglobin. Dyslipidemia was found more frequently in Group A, especially high-density lipoprotein, while C-reactive protein and ferritin were significantly higher comparing with Group B. As for the expression of autoantibodies, anti-nuclear antibody presented positively in all the patients at the baseline. Lower percentage of patients in Group A expressed anti-dsDNA antibody (26.7% vs 66.7%, P = 0.005), while percentage of patients with positive anti-ribosomal P-protein antibody, anti-nucleosome antibody and anti-histone antibody were comparable between the two groups. The ratio of CD4+/CD8+ T cell (0.54 (0.48-0.79) vs 0.92 (0.57-1.34), P = 0.080) and percentage of B cell (9.06±7.43 vs 19.49±14.45, P = 0.137, Table 3) seemed to be lower in Group A while still need further confirmation. Disease activity estimated by SLEDAI-2K at the diagnosis of SLE did not show significant difference between the two groups (9.00 (4.00-14.00) vs 12.00 (8.00-18.00), P = 0.184).
Table 3.
The lymphocyte subsets of SLE with/without hematological malignancies
Lymphocyte subsets (%)
|
Group A
(n=15)
|
Group B
(n=57)
|
P -value
|
CD3+ T cell
|
68.90±17.56
|
70.48±13.42
|
0.735
|
CD3+CD8+ T cell
|
39.26±7.97
|
34.87±12.27
|
0.546
|
CD3+CD4+ T cell
|
24.48±11.04
|
34.72±15.30
|
0.164
|
CD4+/CD8+ T cell (M, Q1-Q3)
|
0.54 (0.48-0.79)
|
0.92 (0.57-1.34)
|
0.080
|
NK cell (CD16+CD56+)
|
21.38±21.72
|
12.62±9.75
|
0.425
|
B cell (CD19+)
|
9.06±7.43
|
19.49±14.45
|
0.137
|
Group A: SLE patients with hematological malignancies
Group B: SLE patients without hematological malignancies
† The data of lymphocyte subsets were from Group A (5 patients) and Group B (22 patients)
Distribution of hematological malignancies and SIR
Of fifteen patients with HM (Additional file1, Supplementary Table 1), ten patients developed HM synchronously with SLE while five posterior to SLE. Four patients (26.7%) were diagnosed with acute myeloid leukemia (AML), seven (46.7%) NHL, with diffuse large B-cell lymphoma (DLBCL) accounting for more than a half, two (13.3%) HL and two multiple myeloma (MM).
In the total of 7954 SLE patients in our institution from 2013 to 2020, nineteen patients were diagnosed with HM (Additional file1, Supplementary Table 1)), with the incidence rate of approximately 0.24%. There was a significantly increased SIR of lymphoma (27.559, 95% CI 10.437-72.766) and leukemia (12.708, 95% CI 4.086-39.524). Among them, three were male, accounting for 0.34% (3/879) while female for 0.23% (16/7075) when it comes to gender difference of incidence in SLE patients with HM.
Risk factors of hematological malignancies in patients with SLE
Based on the baseline comparison above, odds ratio (OR) was calculated by univariate and multivariate logistic analysis to explore risk or protective factors of SLE complicated with HM (Table 4). In univariate logistic analysis, older age at SLE diagnosis (OR 1.075, 95%CI 1.028-1.125, P = 0.002), infection (OR 5.077, 95%CI 1.523-16.925, P = 0.008) and high levels of ferritin (OR 1.003, 95%CI 1.000-1.005, P = 0.033) were risk factors of developing with HM, while high levels of hemoglobin (OR 0.970, 95%CI 0.945-0.998, P = 0.038), anti-dsDNA antibody (OR 0.182, 95%CI 0.051-0.647, P = 0.009) and hydroxychloroquine application (OR 0.143, 95%CI 0.041-0.504, P = 0.002) might play a protective role. In multivariate analysis, older age at SLE diagnosis (OR 1.098, 95%CI 1.031-1.170, P = 0.004) and infection (OR 5.759, 95%CI 1.077-30.804, P = 0.041) were risk factors while high levels of hemoglobin (OR 0.954, 95%CI 0.917-0.993, P = 0.021) was potentially a protective factor.
Table 4.
Risk factors for hematological malignancies development in patients with SLE
Parameter
|
Univariate logistic analysis
|
Multivariate logistic analysis
|
OR(95% CI)
|
P -value
|
OR(95% CI)
|
P -value
|
Age at SLE diagnosis
|
1.075(1.028-1.125)
|
0.002*
|
1.098(1.031-1.170)
|
0.004*
|
Infection
|
5.077(1.523-16.925)
|
0.008*
|
5.759(1.077-30.804)
|
0.041*
|
Hemoglobin
|
0.970(0.945-0.998)
|
0.038*
|
0.954(0.917-0.993)
|
0.021*
|
Ferritin
|
1.003 (1.000-1.005)
|
0.033*
|
-
|
-
|
Anti-dsDNA antibody
|
0.182(0.051-0.647)
|
0.009*
|
0.230(0.044-1.219)
|
0.084
|
SLEDAI-2K at SLE diagnosis
|
0.952(0.873-1.038)
|
0.264
|
-
|
-
|
Glucocorticoids
|
0.361(0.055-2.388)
|
0.291
|
-
|
-
|
Mycophenolate mofetil
|
0.308(0.063-1.505)
|
0.146
|
-
|
-
|
Hydroxychloroquine
|
0.143(0.041-0.504)
|
0.002*
|
0.221(0.043-1.135)
|
0.071
|
* Statistical significance (P<0.05)
Abbreviation: SLEDAI-2K, systemic lupus erythematosus disease activity index-2000.
Treatment and prognosis
Table 5.
Treatment for SLE with/without hematological malignancies
Medication
|
Group A
(n=15)
|
Group B
(n=57)
|
P -value
|
Glucocorticoids (n, %)
|
13 (86.7)
|
54 (94.7)
|
0.601
|
Pulse therapy of glucocorticoids (n, %)
|
2 (13.3)
|
7 (12.3)
|
1.000
|
IVIG (n, %)
|
3 (20.0)
|
17 (29.8)
|
0.666
|
Cyclophosphamide (n, %)
|
0 (0)
|
9 (15.8)
|
0.228
|
Mycophenolate mofetil (n, %)
|
2 (13.3)
|
19 (33.3)
|
0.231
|
Methotrexate (n, %)
|
0 (0)
|
2 (3.5)
|
1.000
|
Leflunomide (n, %)
|
0 (0)
|
5 (8.8)
|
0.536
|
Azathioprine (n, %)
|
1 (6.7)
|
2 (3.5)
|
1.000
|
Cyclosporin (n, %)
|
0 (0)
|
2 (3.5)
|
1.000
|
Tacrolimus (n, %)
|
1 (6.7)
|
6 (10.5)
|
1.000
|
Thalidomide (n, %)
|
2 (10.5)
|
1 (1.8)
|
0.204
|
Hydroxychloroquine (n, %)
|
6 (40.0)
|
49 (86.0)
|
0.001*
|
Rituximab (n, %)
|
0 (0)
|
3 (5.3)
|
0.856
|
Belizumab (n, %)
|
0 (0)
|
1 (1.8)
|
1.000
|
Table 5 shows all the medications since the day of SLE diagnosis to hematological malignancies occurrence (Group A) or the end of follow-up (Group B).
Group A: SLE with hematological malignancies
Group B: SLE without hematological malignancies
* Statistical significance (P<0.05)
Abbreviation: IVIG, intravenous immunoglobin.
Treatment strategies were displayed in Table 5. Thirteen (86.7%) patients in Group A and fifty-four (94.7%) in Group B received glucocorticoids, with no statistical significance. For patients in Group A, none had ever been exposed to cyclophosphamide (CYC) prior to the diagnosis of HM while nine (15.8%) in Group B were ever given CYC treatment. Lower percentage of patients in Group A were treated with hydroxychloroquine (HCQ) than that in Group B (40.0% vs 86.0%, P = 0.001). The cumulative probability of survival analyzed by Kaplan-Meier methods was shown in Fig. 2. Up to January, 2021, the median follow-up periods were both 22.50 months in the two groups and the median survival time for Group A was 30 months from SLE diagnosis and 15 months from HM diagnosis. Patients in Group B had a significantly better prognosis than that in Group A (P = 0.0037). Male tended to have a worse prognosis no matter if they were complicated with HM. The risk factors of mortality were estimated by Cox regression (Table 6) showing that female (RR 0.219, P = 0.009) and hydroxychloroquine (RR 0.281, P = 0.024) might be protective factors. Main cause of death as we collected was multiple organ failure due to malignancy itself or pulmonary infection.
Table 6.
Multivariate proportional hazards Cox regression on risk factors of mortality in patients with SLE
Parameter
|
RR (95% CI)
|
P -value
|
Gender (female)
|
0.219 (0.070-0.681)
|
0.009*
|
Age at SLE diagnosis
|
1.024 (0.984-1.065)
|
0.242
|
Hematological malignancy
|
1.471 (0.354-6.107)
|
0.595
|
Hydroxychloroquine
|
0.281 (0.094-0.845)
|
0.024*
|
* Statistical significance (P<0.05)
Abbreviation: SLE, systemic lupus erythematosus.