Comorbidity Scales in Patients with Myeloproliferative Neoplasms: Which One Fits Best?

Background. As most patients with myeloproliferative neoplasms (MPN) suffer from indolent disease, comorbidities come more into focus. When using new therapeutic regimen like tyrosine kinase inhibitors (TKI) or IMIDs this might become even more important as certain comorbidities such as cardiovascular or metabolic diseases tend to aggravate during such treatments. Several comorbidity indices are used among variant chronic diseases, e.g. showing the strong influence of comorbidities on the survival of CML patients in the TKI era using the Charlson comorbidity score (Saussele et al ASH 2013). All available scores focus on different key aspects. In the context of MPN, it remains unclear which index best reflects comorbidities that impact on complications, therapy and overall survival.

Aims. The aim of our analysis was to compare the two most frequently used comorbidity indices, i.e. Charlson and CIRS with respect to their usefulness in patients with MPN.

Methods. Clinical data of 125 patients (pts) were evaluated from the MPN center at Aachen University hospital in Germany. All pts carried a MPN diagnosis as defined by WHO criteria (2008) and were aged 18 years or older. Among this cohort two different comorbidity indices, the so called CIRS (comorbidity illness rating scale (Linn BS et al., 1968, J Am Geriatr Soc)) and the Charlson score (Charlson ME et al., 1987, J Chron Dis) were evaluated. Both scores were calculated twice, once including and once excluding the MPN diagnosis in order to evaluate the influence of the underlying disease separately from the comorbidities. In addition, data about MPN disease complications, patient age and MPN subtype were included.

Results. 125 MPN pts were included in this analysis. 57 % were male. Median age was 62 years (range 26 – 88). Analysis according to MPN subgroups were performed (see table 1). Median CIRS (excluding MPN) was 4 (range 0 to 22), median Charlson score (excluding MPN) was 0 (range 0 to 8). Pts with secondary myelofibrosis showed highest median CIRS and Charlson scores (median CIRS excl. MPN of 9, median Charlson score excl. MPN of 2). Pts who suffered from complications (n = 42) (defined as thromboembolism, severe bleeding, or death) had a median CIRS of 7,5 and a median Charlson score of 1. In contrast, pts without complications demonstrated a median CIRS of 3 and a median Charlson score of 0 (scores without taking diagnosis “leukemia” into account). Median age of pts in the complications-group was slightly higher with 68.5 years in comparison to the “non-complication-group” (58 years). In the complications-group a median of 58% of CIRS points derived from categories such as cardiac, hypertension, vascular, renal or metabolic disease (5 categories out of 14), indicating that these diseases might have a higher impact on MPN patient outcome. In the non-complications-group 57% of the CIRS point derived from the above mentioned categories.

Summary and Conclusions.

The CIRS score, although created as a geriatric score, proved to be useful in creating certain risk profiles (namely cardiovascular and metabolic comorbidities) that seem to be important for prediction of MPN complications. Nevertheless, for younger pts, the CIRS score appears to overemphasize comorbidities in relation to the underlying hematopoietic neoplasm (e.g. a 36 year old patient with hypothyreosis (well substituted) and psoriasis (treated with local therapy) harboring a significant CIRS score of 4). The Charlson score was created in a younger cohort and may reflect comorbidities of younger MPN pts better. However, a lot of comorbidities such as hypertension and metabolic diseases (eg. hyperlipidemia) are not included in this score. Together, this suggests the need for a more MPN-specific score to assess comorbidities in this patient population.