麻豆传媒

ATLANTA -- Maintaining patients with polycythemia vera at hematocrit levels below 45% reduced thrombotic complications without increasing rates of other serious adverse events compared with a higher hematocrit target, researchers said here.

Relative to patients managed with a target hematocrit level of 45% to 50% in a randomized trial, those with lower targets had significantly lower rates of a composite outcome that included cardiovascular death and other major thrombotic events (hazard ratio for high versus low hematocrit 3.91, 95% CI 1.45 to 10.53, P=0.007), reported Roberto Marchioli, MD, of the Consorzio Mario Negri Sud in Santa Maria Imbaro, Italy, and colleagues.

Rates of progression to leukemia, myelofibrosis, or myelodysplasia were low in both groups in the 365-patient trial and did not differ significantly, the researchers reported at the American Society of Hematology's annual meeting. The findings were published simultaneously online in the New England Journal of Medicine.

In an editorial accompanying the NEJM report, Jerry Spivak, MD, of Johns Hopkins University, said it was notable that, "in the genomic era, the lowly hematocrit still has a role."

Moreover, the even humbler treatment method of phlebotomy appears to be the therapy of choice for polycythemia vera, Spivak suggested, insofar as the study results indicated that hydroxyurea and aspirin did not make a difference in outcomes.

In polycythemia vera, patients produce an excessive number of otherwise normal red blood cells, raising blood volume and viscosity which, in turn, creates a range of clinical problems. These include increased rates of major thrombotic events. The condition can also progress to leukemia, myelofibrosis, or other disorders that add to the patient's burden.

The current standard of care is to use phlebotomy and/or drugs such as hydroxyurea to keep hematocrit below 45%, Marchioli and colleagues explained, but this approach had never been tested in a randomized trial.

They randomized 365 patients recruited from 26 Italian clinics to be treated to a target of either less than 45% or 45% to 50%.

Originally, the study was planned to enroll 1,000 patients, said study chairman Tiziano Barbui, MD, of Ospedali Riuniti di Bergamo in Italy, speaking at a press briefing here. But partway into the study, the investigators ran into difficulties recruiting patients because of competition with other trials testing so-called JAK2 inhibitors for polycythemia vera.

Consequently, study leaders decided to stop enrollment at 365, Barbui said.

The exact treatment regimen was left to the judgment of individual treating physicians and could include phlebotomy, drug therapy, or both. However, clinicians were encouraged to use hydroxyurea as "the drug of choice" for patients with age- or history-related risk factors for thrombosis and in those with progressive thrombocytosis or splenomegaly.

The primary study outcome measure was time until cardiovascular death or major thrombotic event including stroke, acute coronary syndrome, transient ischemic event, deep vein thrombosis, peripheral arterial thrombosis, abdominal thrombosis, or pulmonary embolism. Other types of adverse events were tracked as secondary endpoints.

At study entry, phlebotomy was employed in about two-thirds of patients in both treatment arms, and about half received hydroxyurea, also with no difference between groups. Small numbers of patients were receiving other drugs including pipobroman, busulfan, and interferon.

After six months, Marchioli and colleagues reviewed the treatments chosen by clinicians in the study.

Among patients assigned to the low hematocrit target, 7.7% were started on phlebotomy and 5.5% on hydroxyurea. Upward dosage adjustments in patients in this group already on hydroxyurea were made in 5.5% and the number of phlebotomies increased by 36.4%.

In the group with the higher hematocrit target, phlebotomies were stopped in 28.4% of patients and hydroxyurea was withdrawn in 4.4%. Among patients remaining on phlebotomy and/or hydroxyurea, the number of phlebotomies increased 15.8% and the average drug dose declined by 9.8%.

With median follow-up of 31 months, five patients in the low-hematocrit group and 18 in the high-hematocrit arm had primary-endpoint events.

Marchioli and colleagues calculated incidence rates for the primary outcome of 1.1 per 100 patient-years for the low-hematocrit group versus 4.4 per 100 patient-years with the higher target.

Rates of all cardiovascular events, including those not counted in the primary endpoint, were 4.4% with low hematocrit versus 10.9% with high hematocrit (HR 2.69, 95% CI 1.19 to 6.12, P=0.02).

About three-quarters of patients in both groups were successfully maintained in their target hematocrit ranges. The mean levels were 44.4% in the low-hematocrit group and 47.5% in the high-hematocrit arm.

In the low- versus high-hematocrit groups, six and two patients progressed to myelofibrosis, respectively. Two and one patient, respectively, developed myelodysplasia; two and five patients, respectively, developed leukemia.

Differences in nonhematologic adverse events were not statistically significant. Of 39 such events seen in 35 patients, 25 of the events occurred in the low-hematocrit group.

In their NEJM report, the researchers called the benefit in cardiovascular events with the low target "largely unanticipated on the basis of previous reports." Those included analyses from two previous studies that had found no increase in thrombotic events with a hematocrit target of 45% to 50%.

But they were nonprespecified, post-hoc analyses marred by substantial limitations and potential sources of error, the researchers noted.

Marchioli and colleagues also found that the benefits were maintained across subgroups defined by type of therapy, sex, age, platelet and white cell counts, thrombosis history and risk, splenomegaly, aspirin use, and anticoagulant use.

Spivak, however, wished for more detailed analyses of possible gender differences.

"Women normally have a lower red-cell mass and hematocrit than men, and women with polycythemia vera are at risk for intra-abdominal venous thrombosis with an apparently normal hematocrit," he wrote in the editorial.

"Therefore, the hematocrit target described by Marchioli et al. is adequate for men but inadequate for women, who on the basis of other studies should have a target hematocrit of less than 42%."

For their part, the study authors noted that their investigation did not address the possibility of even greater benefits with a maximum hematocrit target lower than 45%.

Comment