The Natural History Of Polycythemia Vera In Two Diagnostic Eras

The 2005 identification of the JAK2 V617F mutation ushered in a new era of discovery for Polycythemia Vera (PV), allowing for its molecular classification, improved diagnostic capabilities, and a potential for targeted therapy. In this era, aspirin has been shown to lower the risk of thrombosis, the hematocrit (hct) target has been validated, and there has been renewed interest in interferon (IFN). These changes in diagnosis (dx) and therapy may impact the natural history of PV in its modern era. To this end, we analyzed 399 PV patients (pts) diagnosed in two eras (pre-2005 and post-2005) from two tertiary centers. (JHH, N=294; NU, N=105).

Pts were seen between 2005 and 2013. Testing to verify PV included an absolute erythrocytosis, JAK2 status, erythropoietin assay, red blood cell mass testing, and bone marrow biopsy. Pts were diagnosed by treating physicians using criteria relevant to the era of dx, between 1968 and 2013. Pts in each era were compared with regard to differences in demographic and clinical factors. The prevalence of myelofibrotic (MF) and leukemic (AML) transformation, as well as deaths was also determined.

Of 105 NU and 294 JHH pts, 45 (43%) and 113 (38%) were diagnosed post-2005 (p=0.43). Pre-2005 and post-2005 pts were similar in age at dx at NU (52 vs. 57 yrs) and JHH (53 vs. 59 yrs). As expected, median disease duration was longer in pre-2005 NU (12 vs. 3 yrs) and JHH pts (9 vs. 2 yrs). 52% and 60% of pre-2005 pts, and 62% and 60% of post-2005 pts were women at NU (p=0.28) and JHH (p=1), respectively. JAK2 V617F testing was positive 33/37 (89%) and 41/44 (93%) of pre-2005 and post-2005 NU patients. At JHH, JAK2 V617F was positive in 286/294 (97%); JAK2 exon 12 was positive in 6/294 (0.02%). RBC mass testing was infrequently done at NU (9/105); only 2 were performed post-2005. At NU, pre-2005 and post-2005 pts had similar median leukocyte counts (8.9 and 8.7 x 10⁹/L) and hct below 45% (40.5 vs. 42.9%). The prevalence of current anti-platelet and hydroxyurea use was similar at NU (pre-2005: 79%, 49% vs. post-2005: 82%, 48%, respectively). 5% and 11% of pre-2005 and post-2005 pts were on IFN.

At NU and JHH, 29% and 27% of pre-2005 and 18% and 21% of post-2005 pts had thrombosis (pre vs. post: p=0.2 for NU; p=0.26 for JHH). Thrombosis in pre-2005 NU pts occurred at a median of 8 yrs from dx compared to post-2005 pts that frequently had events at presentation (median 0 yr; p=0.017). Similarly, fewer pre-2005 JHH pts had events at presentation compared to post-2005 pts (19/49 (39%) vs. 16/24 (67%); p=0.025). At NU, 19/105 (18%) developed MF, at a median of 12 yrs. Only 2/19 (11%) occurred in post-2005 pts, due to shorter disease duration. At JHH, 49/294 (15%) developed MF at a median of 11 yrs, but only in 4/49 (8%) post-2005 pts. At NU and JHH, 5/105 (5%) and 11/294 (4%), developed AML at a median of 15 and 8 yrs from dx. 1 and 4 post-2005 NU and JHH pts developed AML. At NU and JHH, 20/105 (19%) and 42/294 (14%) died. At NU, 16 were pre-2005 pts and 4 were post-2005 pts. At JHH, 35 were pre-2005 pts and 7 were post-2005 pts. AML-deaths were overrepresented in the post-2005 pts (25% NU, 57% JHH).

Remarkably, the epidemiology and natural history of PV was quite similar between two eras of diagnosis and between two large tertiary centers. Age at dx in pre and post-2005 patients was similar, arguing against a notion that pts are diagnosed earlier in adulthood. At NU, hct was typically below 45%, implying adherence to a recently validated target and aspirin use was prevalent regardless of the era of dx. Regardless, thrombosis at presentation was more common in modern era PV patients, whereas pre-2005 patients had more thrombotic events after an official PV diagnosis. As expected, time-dependent complications such as MF and death were more common in pre-2005 pts, and occurred within the 2nd decade of disease. MF was rare in post-2005 pts, who remain in their 1st decade of disease. Overall, AML was rare, but each cohort still had 1^st decade transformations (post-2005 pts), over-representing deaths in this group. While the JAK2 mutation discovery has modernized PV dx, the natural history of PV has not changed. It is in this modern era of PV that we will learn whether or not new (JAK-inhibitors) and renewed (IFN) therapies can prevent or delay onset of feared time-dependent complications.

No relevant conflicts of interest to declare.