Patients' Comorbidities and Overall Survival in Primary Myelofibrosis (PMF)

Cancer patients often experience comorbidities that may affect their therapeutic options, outcome of therapy, and their prognosis. However, none of the prognostic scoring systems for PMF account for an impact, if any, of patients' comorbidities on their prognosis. The aim of this study was to determine the effect of comorbidities on the survival of patients (pts) with PMF.

We reviewed the medical records of 349 consecutive PMF pts who presented to our institution from 2000 to 2008. The Adult Comorbidity Evaluation-27 (ACE-27) form, a validated 27-item comorbidity index for cancer pts, was used to assess the severity of comorbid conditions in these patients at the time of presentation. Data on demographic characteristics including age, performance status (PS), complete blood profile, karyotype, International Prognostic Scoring System (IPSS), stem cell transplant (SCT) and outcomes (leukemic transformation and survival) was also collected. Kaplan-Meier method and log-rank test were used to assess survival. Multivariate analysis was performed using the Cox Proportional Hazards Model.

Of the 349 pts included in this study, 222 (64%) were male and 311 (89%) were white; median age at presentation was 62 years (20–82); median duration of follow-up was 37 months (range 0 – 133). The ACE-27 comorbidity scores were as follows: none, 125 pts (36%); mild, 128 (37%); moderate, 63 (18%); and severe, 33 (9%). Hypertension was the most common comorbidity (31%) followed by diabetes mellitus (14%) and prior solid tumor (11%). Of the 238 pts assessed for JAK2 mutation status, 141 (59%) showed the mutation. One hundred and ninety-three (55%) patients died, 41 (12%) suffered leukemic transformation and 27 (8%) patients underwent SCT. Overall median survival using the Kaplan-Meier method was 39 months. Median survival according to ACE-27 scores was: 52, 38, 36 and 28 months for none, mild, moderate and severe comorbidity scores, respectively (p=0.028). In addition to comorbidities, age >=65 years, male gender, increased performance status (PS) based on the Zubrod Scale, white cell count >25K, platelets <100K, hemoglobin <10 g/dl, transfusion need, albumin <3.5 g/dl, unfavorable karyotype, prior treatment, and increased IPSS risk reached statistical significance for survival (p<0.05). Constitutional symptoms and peripheral blasts did not achieve statistical significance. In the multivariate analyses, contrary to what was expected, comorbidities were not identified as independent predictors of survival. Advanced age (>=65 years), PS of >=2, platelet <100K, transfusion need and IPSS high risk category continued to maintain prognostic significance. The adjusted hazards ratios from multivariate Cox Proportional Hazards Model were 1.02, 2.57, 0.997, 1.85 and 1.75 for age, PS, platelet, transfusion need and IPSS, respectively. A separate multivariate analysis was conducted that included the ACE-27 comorbidity scores and the 5 factors found to be significant. The comorbiditiy scores once again were not identified as independent predictors of survival.

Comorbidities in general are known to impact survival in cancer pts. However, our study failed to show comorbidities to independently effect survival in pts with PMF. Factors like advanced age and poor PS (>=2) were noted to be better predictors of survival. Current prognostic scoring systems however do not include PS as part of risk assessment. Newer prognostic models incorporating PS should be developed to determine survival in pts with PMF.

No relevant conflicts of interest to declare.