Background: Patients (pts) with myelofibrosis (MF) and anemia experience negative impact on both quality of life and prognosis. Pacritinib (PAC), a JAK2/IRAK1/ACVR1 inhibitor, has demonstrated clinically and statistically significant improvement in transfusion requirements in pts with MF who required red blood cell (RBC) transfusion at baseline in the phase 3 study, PERSIST-2. Here, we evaluate the impact of PAC versus best available therapy (BAT), including ruxolitinib (RUX), on symptom burden in pts with baseline RBC transfusions. Methods: This analysis focused on PERSIST-2 pts who randomized ≥12 weeks (for efficacy analyses ≥22 weeks) prior to study termination on PAC 200 mg twice daily or BAT (including RUX) and required RBC transfusions. The proportion of pts who achieved transfusion independence response (TI-R, any 12-week interval with no RBC transfusions) by week 24 was determined for each group. Symptom improvement was measured by Total Symptom Score (TSS; version 2.0, excluding tiredness) and Patient Global Impression of Change (PGIC) at week 24. Statistical testing using Fisher Exact test was performed for PAC vs BAT. Results: The analysis included 3 groups: 35 pts on PAC, 34 on BAT, and 13 of whom were on RUX. Baseline characteristics were similar amongst these groups, including median platelet count (45 vs 45 vs 43 x 10⁹/L) and hemoglobin (8.5 vs 8.7 vs 8.6 g/dL) respectively. More than 50% of pts in each group had prior JAK inhibitor exposure. The median total daily RUX dose was 10 mg at week 24. A significantly greater proportion of pts treated with PAC vs BAT achieved ≥50% TSS response (31% vs 9%, P=0.034), with only 15% of pts achieving TSS response on RUX. Of the 11 PAC pts with TSS response, 6 (54.5%) also achieved TI-R. By contrast, none of the TSS responders on BAT achieved TI-R. A similar trend for ≥50% TSS response was observed in pts with baseline platelets <50 x 10⁹/L with PAC (24%) vs BAT (10%) or RUX (11%). TSS reduction on PAC was primarily due to reductions in both spleen- and cytokine-related symptoms. Physical function-related symptoms (not included in the total score) also showed a higher median % reduction on the PAC arm compared to BAT (tiredness: 37% vs 13%; inactivity: 30% vs 4%). In addition, a higher proportion of pts reported symptoms at week 24 as “very much” or “much” improved (PGIC) with PAC (37%) compared to BAT (9%, P=0.009), including RUX (8%). Pts experiencing ≥1 treatment emergent grade 3 anemia was similar in the PAC (17.1%) and BAT (16.3) groups. No pts in the PAC group reported ≥1 treatment emergent grade 3 fatigue compared to 5% in BAT. Conclusions: PAC has previously demonstrated improvement in symptoms in pts with MF. These findings are consistent in showing that PAC compared to BAT or low dose RUX provides substantial symptom benefit in those pts who require RBC transfusions. PAC could therefore be a potential option to address an unmet need for this patient population. Clinical trial information: NCT02055781.