Marko Lucijanić, MD, PhD
DUBROVNIK, CROATIA—The Glasgow Prognostic Score (GPS) may predict survival in patients with myelofibrosis (MF), according to research presented at Leukemia and Lymphoma: Europe and the USA, Linking Knowledge and Practice.
In a retrospective study, MF patients who were considered intermediate-risk according to the GPS had roughly twice the risk of death as good-risk patients.
High-risk patients had a nearly 24-fold greater risk of death than good-risk patients.
Marko Lucijanić, MD, PhD, of the University Hospital Dubrava in Zagreb, Croatia, presented these findings at the meeting. Results were also published in a recent issue of Blood Cells, Molecules, and Diseases.1
Dr. Lucijanić and his colleagues analyzed 88 patients—67 with primary MF and 21 with secondary MF—who were treated at the University Hospital Dubrava from 2004 to 2018.
Patients were divided into GPS risk categories:
- Good-risk patients had C reactive protein (CRP) ≤10 mg/L and albumin ≥35 g/L
- Intermediate-risk patients had either CRP >10 mg/L or albumin <35 g/L
- Poor-risk patients had both CRP >10 mg/L and albumin <35 g/L.
Results
The researchers found that CRP and albumin were independent predictors of overall survival (OS) in MF.
“What we saw is that both CRP and albumin were univariately associated with inferior overall survival when the patients were divided at the proposed cutoff levels,” Dr. Lucijanić said.
“And both CRP and albumin remained statistically significant when analyzed together in a Cox regression model additionally adjusted for DIPPS [Dynamic International Prognostic Scoring System].”
In the univariate analysis, the hazard ratio (HR) for death was 3.42 (P<0.001) for patients with CRP >10 mg/L and 4.68 (P<0.001) for patients with albumin <35 g/L.
In a multivariate analysis, the HR for death was 2.49 (P=0.013) for patients with CRP >10 mg/L and 2.74 (P=0.031) for patients with albumin <35 g/L.
“So no surprise that, when [CRP and albumin were] combined into the Glasgow Prognostic Score, the GPS could identify three subgroups of patients with distinct prognosis,” Dr. Lucijanić said.
When the researchers assessed OS according to GPS, they found the HR for death was:
- 2.77 for intermediate-risk patients compared to good-risk patients (P<0.001).
- 15.78 for poor-risk patients compared to good-risk patients (P<0.001).
- 5.82 for poor-risk patients compared to intermediate-risk patients (P<0.001).
In a Cox-regression model for OS that was adjusted for DIPPS, age, and gender, the HR was:
- 2.08 for intermediate-risk patients compared to good-risk patients (P=0.040)
- 23.52 for poor-risk patients compared to good-risk patients (P<0.001).
Dr. Lucijanić noted that patients who were intermediate- or poor-risk according to the GPS were more likely to have symptoms of aggressive disease that are associated with non-response to JAK inhibitors.
The intermediate- or poor-risk patients were more likely to have constitutional symptoms, massive splenomegaly, blast phase disease, circulatory blasts, higher absolute monocyte counts, lower hemoglobin, lower platelets, higher lactate dehydrogenase, higher red cell distribution width, transfusion dependency, and higher ferritin. The patients also had higher DIPSS scores.
“So, to summarize, higher GPS recognizes patients with more aggressive disease features at a higher risk of death,” Dr. Lucijanić said. “And not only does it discriminate survival of myelofibrosis patients, it seems to do so in a DIPPS-independent manner. However, you should be aware that this is a small, retrospective study from a single center with a very long follow-up period during which patients were exposed to different types of therapies.”
Dr. Lucijanić did not declare any conflicts of interest.
1. Lucijanić M et al. Blood Cells Mol Dis. 2018 Sep;72:14-16. doi: 10.1016/j.bcmd.2018.06.001.