Common CVD Biomarkers May Hold Prognostic Value in MDS, Study Finds

Commonly used blood-based markers of cardiovascular disease (CVD) independently predicted outcomes of myelodysplastic syndromes (MDS) in a multicenter retrospective cohort study published in Blood Cancer Journal.

The findings suggest that circulating N-terminal pro-B-type natriuretic peptide (NT-proBNP) and growth differentiation factor-15 (GDF-15), both soluble biomarkers of CVD, may be indicators of the course of MDS and outcomes.

While there are several well-validated prognostic systems for MDS, prognostication for lower-risk MDS can be problematic, the authors explained. Strategies for this disease setting vary substantially, ranging from watchful waiting to early intervention with allogeneic stem cell transplantation. Additionally, current biomarkers in MDS prognosis are limited to cytogenetics and somatic mutations.

In previous research, soluble CVD biomarkers, including NT-proBNP and GDF-15, have been linked to cancer incidence and may drive tumor growth, the authors noted. A decrease in high-sensitivity C-reactive protein levels after statin treatment, however, was found to reduce cancer mortality significantly in past research.

“On top of epidemiological evidence, experimental data further support a pathophysiological link between clonal hematopoiesis of indeterminate potential (CHIP), a precursor of MDS, with CVD development through inflammation-mediated accelerated atherosclerosis,” the authors wrote. “Despite the evidence of a bidirectional interplay between CVD and malignancies and the reported association of CVD biomarkers with all-cause mortality in cancer patients no study to date has addressed the value of CVD biomarkers in the prognostic assessment of MDS patients.”

The retrospective, multicenter cohort study included 105 patients with MDS, all of whom had serum levels of troponin T, proBNP, GDF-15, and CRP measured. Kaplan–Meier estimates and multivariate analysis using Cox regression were used to analyze overall survival (OS), Leukemia-free survival (LFS), and time to progression (TTP). OS was defined as “the time from sampling to last follow-up or death from any cause” and LFS as “the time from sampling to leukemic progression or death.” TTP was the time from sampling to disease progression.

At a median follow-up of 23.9 (95% CI, 11-36.8) months in the overall cohort, the median OS was 37 (95% CI, 12.6-61.4) months, and the median TTP was 25 (95% CI, 11.938.1) months. A total of 36 patients (34.3%) experienced progression to acute myeloid leukemia, and CVD was attributed as the primary cause of death in 3 out of 30 reported deaths.

Measures that showed significant prognostic value for OS were baseline hemoglobin, bone marrow blast (BMB) percentage and CRP, GDF-15, and NT-proBNP levels. In univariate analysis, hemoglobin, absolute neutrophil count, BMB cytogenetic category, CRP, NT-proBNP, and GDF-15 were significantly associated with TTP.

Only BMB, GDF-15, and NT-proBNP were independently correlated with OS in multivariate analysis (HR, 1.054; 95% CI, 1.027-1.081, HR, 1.080; 95% CI, 1.015-1.149, and HR, 1.270; 95% CI, 1.107-1.457, respectively. BMB (HR = 1.041, 95% CI = 1.017-1.066), GDF-15 (HR = 1.063, 95% CI = 1.011-1.118), and NT-proBNP (HR = 1.207, 95% CI = 1.052-1.384) were also independently associated with TTP.

When assessed as standalone prognostic values, GDF-15 and NT-proBNP, the best cutoff level for GDF-15 was a value above 3727 ng/L, which was significantly associated with worse OS (P = 0.010) and TTP (P = 0.006). The value selected for NT-proBNP, 175 ng/L, could not function as a prognosticator for either OS or TTP.

When combined as a composite score called CardioScore, the tested values for GDF-15 and NT-proBNP were able to improve the prognostic power of the Revised International Prognostic Scoring System (IPSS-R), which is currently the most widely used system for prognostication, the authors noted. The addition of CardioScore to IPSS-R resulted in upstaging for 12.5% (n = 12) patients and performed better for OS and TTP.

Because molecular analysis is also used for MDS prognostication with the Molecular International Prognostic Scoring System (IPSS-M), the authors analyzed the CVD biomarkers’ prognostic values when known somatic mutations were added to the algorithm in a subset of patients n = 29). NT-proBNP levels were independently associated with TTP (HR, 1.586; 95% CI, 1.139-2.207; P = 0.006), OS (HR, 2.023; 95% CI, 1.109-3.690; P = 0.022), number of mutations, and BMB. IPSS-M scores adjusted to include NT-proBNP redistributed 20.7% of patients (n = 6) and outperformed the IPSS-M without NT-proBNP adjustment in terms of OS and TTP.

Study limitations included the analysis of baseline values only, which does not account for changes in hematological markers and CVD markers that cannot be captured retrospectively in a reliable manner. The low number of CVD deaths also precludes definite conclusions, the authors noted. Still, the findings suggest further investigation into CVD biomarkers and MDS pathobiology, including studies in larger cohorts that are prospectively examined.

“Our findings reveal a previously unrecognized association between circulating NT-proBNP and GDF-15 with MDS course and outcome and are in line with numerous reports showing a robust association between cardiac and cancer incidence and mortality,” the authors concluded. “Of note, risk factors for CVD and pre-existing CVD, as captured by the MDS-[specific comorbidity index] in our analyses, did not affect OS and TTP, indicating that the correlation of survival with CVD biomarkers is rather linked to non-CVD-related or at least all-cause death.”

Reference

Mitroulis I, Papadopoulos V, Lamprianidou E, et al.Common cardiovascular biomarkers can independently predict outcome of patients with myelodysplastic syndromes. Blood Cancer J. Published online May 3, 2023. doi:10.1038/s41408-023-00844-4