Phase 3, Trial of Interferon-α versus Hydroxyurea in Polycythemia Vera and Essential Thrombocythemia

[Manouche]

Key Points

Rates of thrombosis and progression were low in patients with ET/PV treated with either HU or IFN in this randomized study.

PEG was more effective in normalizing counts and reducing JAK2V617F VAF in PV while HU induced more histopathologic responses in ET.

ashpublications.org/blood/a...

[charl17]

I wonder what they mean by "histopathological responses?"

[Manouche in reply to charl17]

Basically it means that HU was able to induce some positive changes in the bone marrow tissue.

[EPguy]

Thanks for the fresh studies as you are so good at.

The study goes to only 12 or 24 months. This seems to be a common limitation of many INF studies, while INF needs longer to really kick in.

--

Below are notes on Jak2 allele. But some other items first:

I can't find any actual progression data in the materials, but I would think it needs to go out for at least 5 years. Still wonder why there is no prog data in the 5+ year Conti PV study.

The bone marrow responses on HU are neat, (Fig. S5 in link below) I've not heard of that ability before. But this part of the study stops at 12 months so the INF has no time to do its thing. We know INF can also do this in some studies. With a longer trial maybe there could be support for HU/INF combo.

In Fig. S8A the allele response on INF for CALR is very impressive, entirely gone in just two years. I've seen elsewhere this can depend which type of CALR is present; such an extreme result needs more inspection. HU is not much effect.

The primary endpoint was CR, <<CR for HU was 37% and 35% for PEG (p=0.80) at 12 months>> That P value is not good, it's supposed to be under 0.03 to be significant as I understand it. Makes one wonder what P values are in the other numbers.

--

I made plots from the Jak2 allele vs time. These posts can have only one image per post, so I will follow with two more replies/plots.

--

In this one for PV from the supplemental data there are some surprisingly familiar trends. Compare to my post #3 below here, which shows the old ContiPV plot I made. The plot here is for PEG, while the ContiPV in post #3 is for Besremi, but they remain similar. See the blue INF allele going down while HU starts its familiar up trend, similar to Conti PV (Besremi) . The down trend is less pronounced than the Besremi study, maybe Bes works better?

Supplemental data is from here:

ash.silverchair-cdn.com/ash...

PV Jak2 Jan10Report

[OZland in reply to EPguy]

Hi, what does it mean exactly “entirely gone”??In Fig. S8A the allele response on INF for CALR is very impressive, entirely gone in just two years. I've seen elsewhere this can depend which type of CALR is present; such an extreme result needs more inspection. HU is not much effect.

[EPguy in reply to OZland]

I added the trend line to Fig. S8A, see in light blue below. See also the thin black line I added at Zero allele. On INF CALR allele is right at zero best one can tell from the plot. This is unexpected and they should expand on this finding, it being not even mentioned in the their Abstract.

CALR allele Jan 10

[EPguy]

Post #2

This is a plot I did for Jak2 in ET from Fig. S7 in the supplemental data. Plots are flatter naturally because the alleles are smaller. Trends are similar to PV, except the initial rise for year 1 on INF. This rise includes two outlier data points. Really needs more time to see. From start to 2 years INF is down from ~17 to 9% allele, below the possibly important 10% range. HU trend is very similar to Conti PV findings in post #3.

ET Jak2 Jan 10

[EPguy]

Post #3

Here is the old Conti PV plot for comparison to the two new ones above. Compare the ones labeled "allele" and note the present study stops at year 2.

Besremi Allele

[Bluetop in reply to EPguy]

Mnay thanks for taking the time to do these plots and for your analysis

[Bluetop]

Very interesting -good to see the research continuing -Hydroxy has clearly been the workhorse for many years andstill the details of how it works don't seem to be well understood.

[EPguy]

I got the link for the full report to work. I'm paying a lot of attention as I should soon be switching to Bes. This is very dense and making my head hot, I'm sort of thinking out loud here but all are welcome to comment.

Some summary points:

-The jak2 allele reduction for INF seems real with two Ph 3 studies showing quite similar results.

-The advantage for HU in histological response seems real at 12 months.

-Is either of these relevant to progression? The big question the authors also ask.

-is 12 months too short for a quality study of this sort?

--

It's actually 1st presented in 2018. More items of interest, in no particular order:

--

<<JAK2V617F VAF decreased through month 48 with PEG, but increased with HU after month 12.>> / This matches exactly the ContiPV chart in my post #3 below. I'm starting to believe it with two good studies matching here.

<<Since PEG is thought to act at the level of the mutated MPN stem cell11-14, it was unanticipated that histological responses (HPR) were more frequent in HU patients>>

<<The

discordance between the greater reduction of JAK2V617F VAF due to PEG

and the improvement in marrow morphology (HPR) with HU is difficult to

reconcile>> / *No kidding*.

HPR is for a complete response, rather than the partial one also listed in the figures.

<<HU had a histopathologic response (HPR) of 12/52 (23%) versus 3/57 (5%) for PEG, p=0.01 (Supplemental Figure S5). I'm think this includes fibrosis but not sure.

Best HPR response was seen in 18/54 (33%) of HU patients and 10/59 (17%) for PEG, p=0.05.>>

Best response I believe means response at any time of the study vs at the specific 12 month endpoint. If that is correct, both agents do not increase HPR smoothly, rather the response dips at 12 months from its best result.

--

In Conti PV<<A difference (presuming HU vs INF) in the incidence of thrombotic events or evidence of progression was not observed.>>

--

<<PROUD/CONTINUATION-PV study did not include ET patients, bone marrow histopathology, cytogenetics or monitor the effects of these treatments on patient symptoms when evaluating CR, which may account for their higher CR rates. (compared to the present MPD-RC 112 study) >>

ContiPV would actually look better if they had included an endpoint of allele reduction (cytogenetics).

--

Good summary of the clashing info:

<<Given the divergence of molecular response and bone marrow histopathologic

remission, we raise caution of using changes in JAK2V617F VAF, marrow histopathology or

clearance of cytogenetic abnormalities as measures of disease course modification.>>

--

Here is a different report with reported HPR benefit for INF:

ashpublications.org/blood/a...

<<BM response (in this different study) -

Seventy-four patients were evaluated for BM response. Histopathologic remission (HPR) was observed in 9 patients (12.2%; ET, 5; PV, 4). Histopathologic remissions were associated with a hematologic PR/CR in 8 of the 9 patients. In 7 patients, BM fibrosis progressed to grade 2+ (scale, 0-3+) while they were receiving PEG therapy, but only 1 patient met clinical criteria for transformation to MF. The remaining patients had stable degrees of BM fibrosis.>> This report does not compare HU so no HU fibrosis info here.

In this study HPR is much better for INF 12.2 vs 5% above. But HU is still better and this is only a Phase 2 and they share many authors.

INF for BM response in this study was similar between ET and PV.