SHIP: Red cell distribution width as a predictor of thromboembolic complications

Thromboembolic complications are a high risk for patients with myeloproliferative neoplasms (MPN), yet prognostic parameters for increased risk are currently lacking.¹ Previously, a study identified risk factors for patients with polycythemia vera for clinical prediction of thromboembolic events.²

The MPN Hub has also reported on the risk of thrombosis in patients with secondary myelofibrosis.³ Below, the MPN Hub summarizes a new analysis investigating the predictive value of these parameters in a normal  (individuals without myeloproliferation or hematopoietic cancers) control population in order to validate these findings.¹

Study design

This analysis was conducted in two independent cohorts of patients from the Study of Health in Pomerania (SHIP). In total, 2,491 patients were assessed between 1997 and 2001, and 4,358 patients were assessed between 2008 and 2011.¹ Baseline characteristics of established risk factors for thromboembolic events are summarized in Table 1. Receiver operating characteristic curves were calculated to assess the predictive value of the model.¹

Table 1. Baseline characteristics*

 Results

In both cohorts, age, hypercholesterolemia and red cell distribution width (RDW) were identified as predictors of thromboembolic events.¹ Thromboembolic events had occurred in 12.9% of individuals in the SHIP-START cohort and 21.4% of the SHIP-TREND cohort, and confirmed risk factors were as follows:¹

• In the SHIP-TREND cohort, increased age, hypercholesterolemia, and elevated body mass index scored as predictive factors for risk of thromboembolism (Figure 1A).
• In the SHIP-START cohort, male sex and diabetes mellitus were also scored as predictive factors for risk of thromboembolism (Figure 1B).
• In both cohorts, age, hypercholesterolemia, and RDW were confirmed as predictors of thromboembolic events.
• RDW with an odds ratio of 1.28 (95% confidence interval [CI], 1.11–1.47) in the SHIP-START cohort and 1.25 (95% CI, 1.12–1.38) in the SHIP-TREND cohort.

Figure 1. A. SHIP-START-0 final model odds ratios. B. SHIP-TREND-0 final model odds ratios* 

*Adapted from Manz, et al.¹
BMI, body mass index; CI, confidence interval; HCL, hypercholesterolemia; OR, odds ratio; PLT, platelet; RDW, red cell distribution width.

Conclusion

These results confirm that RDW could be used as an independent predictive parameter for thromboembolic evets in the general population.¹ Further validation of predictive scoring systems combining predictive laboratory parameters are required, but this study confirms this would be worth investigating; with model accuracy rates of 89.2% and 86.8% for the SHIP-START and SHIP-TREND cohorts respectively.¹