Prevalence of Comorbidities at Diagnosis in Myelodysplastic Syndromes: A Systematic Review and Meta-Analysis

Emerging evidence suggests that clonal hematopoiesis (CH), a precursor state of myelodysplastic syndromes (MDS) is strongly associated with increased prevalence of common chronic disorders with an inflammatory background. However, data in MDS are scarce and emanate only from diverse retrospective databases and registries suffering from lack of critical information. For this reason we assessed the prevalence of comorbidities in MDS patients at the time of diagnosis in relation to the relevant global indices through a systematic review and meta-analysis. We searched PubMed, Scopus, and ClinicalTrials.gov databases for the terms “myelodysplastic syndromes AND comorbidities”. Additional search using Google Scholar and personal communication was performed. All available data regarding the prevalence of cardiovascular, pulmonary, renal, and liver disease, type II diabetes mellitus, dementia, autoimmunity, and second malignancy (excluding non-melanoma skin cancer) at the time of MDS diagnosis was retrieved from eligible publications. Data synthesis was performed using the Freeman-Tuckey double-arcsine transformation. The prevalence of comorbidities in MDS patients were compared to the ones in the general population as reported in any available official source. Adjustment for age and sex was performed to match the pool of MDS patients. Age, gender, IPSSR risk category, and data source were evaluated as potential sources of heterogeneity. The GRADE assessment tool was used to rate evidence certainty for every endpoint assessing risk of bias, imprecision, inconsistency, indirectness, publication bias, and effect size. The STATA 18 software was used for computational purposes.

Twenty-eight out of 2,017 retrieved sources of evidence, including 69,784 MDS patients, were qualitatively and quantitatively analyzed. MDS patients had a mean age of 71.6 ± 5.2 years and were 58% males (95% CI: 56%-61%). CCI ≥1 was observed in 55% (47%-62%) of patients and MDS-CI ≥1 in 44% (36%-53%) of them.

The estimated overall prevalence of comorbidities among MDS patients was for cardiac: 27% (22-31%) [coronary heart disease 15% (9-21%); heart failure 14% (10-19%)]; cerebrovascular: 8% (5-10%); pulmonary: 10% (7-13%); renal: 6% (4-9%); liver: 4% (2%-6%); diabetes: 18% (15-22%); autoimmune disorders: 4% (3-6%); dementia: 5% (2-9%); cancer: 12% (9%-15%).

There is high level of evidence certainty that cardiac comorbidity (P<0.0001) and particularly heart failure (P<0.0001), cerebrovascular disease (P=0.0001), renal comorbidity (P<0.0001), autoimmunity (P=0.002), and cancer (P<0.0001) were more prevalent among MDS patients than in sex- and age-matched global population. On the contrary, diabetes (P=0.053) and dementia (P=0.241) were comparable, while pulmonary comorbidity (P=0.033), and liver comorbidity (P<0.001) appear to be less prevalent in MDS patients compared to the general population.

Heterogeneity regarding prevalence of comorbidities was considerable; excess heterogeneity was mainly attributed to data source rather than age, gender, and IPSSR risk category. Of note, publications based on national healthcare databases reported almost twice as high prevalence for cerebrovascular disease, pulmonary comorbidities, and cancer when compared to MDS registries, raising concerns about potential misdiagnosis of MDS in the case of databases and/or the quality of annotated data in registries. Our findings reinforce the notion of an increased association of MDS with chronic, inflammation-related disorders. However, prospective comparisons with non-CH patient cohorts will ultimately be needed to properly address the effect of major confounders, such as common predisposing factors for both CH and inflammaging conditions.