Page 176 - 2019_07 resto del Mondo-web

T. Dittrich et al.
(Table 1). Accordingly, eGFR showed a highly significant negative correlation with all cardiac biomarkers (Online Supplementary Table S3). Serum NT-proBNP levels were sig- nificantly determined by both heart involvement status and renal function (Online Supplementary Figure S3). As renal function decreased, more patients both with and without heart involvement exceeded the NT-proBNP thresholds of any of the staging systems. As a conse- quence, patients with eGFR<50 and with AF were attrib- uted higher scores than their respective controls in all the staging systems analyzed (Table 1, Figure 2B,C). As shown
above, eGFR<50 or AF are by themselves prognostically adverse factors. Thereafter we addressed the question of whether, within the high-risk cardiac biomarker cate- gories, the patients with eGFR<50 or AF actually fared bet- ter, worse or equally (Online Supplementary Figure S6). Apparently both described effects counterbalance each other in the MAYO3b staging system. However, patients with eGFR<50 classified as highest risk by the MAYO2004 and MAYO2012 staging systems still had a worse progno- sis than patients with eGFR ≥50 mL/min/1.73 m2 (6.4 vs. 14.9 months and 4.7 vs. 9.6 months, both P<0.001).
Figure 1. Survival of the whole cohort and the subgroups divided according to renal function and the presence or absence of atrial arrhythmias. Kaplan-Meier plots depicting overall survival (OS) from diagnosis. The shaded areas represent the 95% confidence interval estimates. The median OS (with corresponding 95% confi- dence intervals) were as follows: all patients, 38.3 (31.9 - 44.5) months; patients without an atrial arrhythmia or pacemaker rhythm (AF), 45.5 (38.9 - 54.2) months; patients with AF, 11.9 (7.4 - 23.0) months; patients with an estimated glomerular filtration rate (eGFR) ≥50 mL/min/1.73 m2, 52.9 (43.0 - 68.8) months, and patients with an eGFR <50 mL/min/1.73 m2, 18.3 (12.9 - 24.7) months.
Table 2. Results of univariable and multivariable analyses including different cardiac scoring systems.
Prognostic factor Comparison
Age, years per 10 years
Univariable HR 95% CI
1.27 1.17 - 1.38
1.13 0.96-1.33 0.96 0.80-1.15 2.10 1.79 - 2.47 1.85 1.58 - 2.17 1.96 1.61 - 2.40 2.34 1.71 - 3.19 6.36 4.71 - 8.58 2.36 1.73-3.22 4.19 3.04 - 5.77 11.26 8.22 - 15.42 1.82 1.35 - 2.44 3.74 2.84 - 4.92
7.08 5.44 - 9.21
P
<0.001
MAYO2004
HR 95% CI P
1.18 1.09 - 1.29 <0.001
Multivariable
MAYO3b MAYO2012
HR 95%CI P HR 95%CI P
1.15 1.05-1.25 0.002 1.16
1.07 - 1.27 <0.001 - -
- -
- -
Sex malevs.female Lightchaintype lambdavs.kappa
0.133 -
0.654 - <0.001* 1.60 <0.001 1.33 <0.001 1.28 <0.001 1.92 <0.001 4.55
<0.001 - <0.001 - <0.001 - <0.001 - <0.001 -
<0.001 -
-
-
- 1.89 - 1.57 - 1.57 - 2.64 - 6.24 -
-
-
-
-
-
- -
- - <0.001 1.53 <0.001 1.15
0.019 1.24 <0.001 - <0.001 -
- 2.05 - 3.39 - 8.10
- - - - - - - 1.81 <0.001 - - 1.37 0.105 1.32 - 1.53 0.041 1.24 - - - - - -
dFLC, mg/L
eGFR, mL/min1.73m2 Atrial arrhythmia MAYO 2004 stage II MAYO 2004 stage III MAYO3bstageII MAYO 3b stage IIIa MAYO 3b stage IIIb MAYO 2012 stage II MAYO 2012 stage III
MAYO 2012 stage IV
<180 vs. ≥180 <50 vs. ≥50 any vs. no stage I stage I stageI stage I stage I stage I stage I
stage I
1.35 1.13 1.04 1.40 3.31
1.29 0.97 1.01
1.12 1.00
1.32
2.61
4.77
- 1.56
- 1.52
- - - - - - - - - -
<0.001 0.045
1.49–2.81<0.001 - 2.43 – 4.73 <0.001 - 5.75 - 11.40 <0.001 -
-
-
-
-- - 1.77
-- - 3.44
-- - 6.26
- 2.38
- 4.55
- 8.21
<0.001
<0.001
<0.001
Summary of one univariable and three multivariable survival analyses using Cox proportional hazards regression.Statistically significant differences are represented in bold,italics.*The thresh- old of 180 mg/L for the difference between involved and uninvolved free light chains (dFLC) was chosen to allow for comparison with the MAYO2012 system, which already includes dFLC at this threshold. HR: hazard ratio; 95% CI: 95% confidence interval; eGFR: estimated glomerular filtration rate.
1454
haematologica | 2019; 104(7)