Introduction: Myeloproliferative neoplasms (MPN) are clonal disorders of the hematopoietic system characterized by an excessive proliferation of myeloid cells and progressive bone marrow fibrosis. MPNs are result of an abnormal constitutive activation of the JAK/STAT signaling pathway. Ruxolitinib is the first JAK1/2 inhibitor approved for the treatment of myelofibrosis (MF), the most aggressive of MPNs. However, the beneficial effects of ruxolitinib cannot be attributed to its anticlonal activity but rather to its reduction in inflammatory cytokine production, spleen size and symptom burden. Helper (CD4+) and cytotoxic (CD8+) T-cells are key mediator elements in the adaptive immune system. Disrupted homeostasis of functionally diverse T-cell subpopulations (naïve, memory and effector) can result in abnormal cytokine production. The aim of this study was to determine the baseline T-cell subset composition in patients with MF and to monitor the immunomodulatory effects of JAK1/2 inhibition.

Methods: CD4+ and CD8+T-cell subpopulations were measured in PB samples from healthy controls (n=16) and PB and BM samples from patients (n=47) with MF treated on a phase I/II clinical trial of ruxolitinib, using multiparametric flow cytometry. Subsets were immunophenotypically defined based on the cell-surface expression of CD45RO and CD62L as follows: naïve T-cells, CD45RO-CD62L+; central memory (CM), CD45RO+CD62L+; effector memory (EM), CD45RO+CD62L-; and terminally differentiated effector memory T-cells (TEM) CD45RO-CD62L-.

Results: Our results showed no significant difference in the distribution of helper vs. cytotoxic T-cells between untreated MF patients and healthy subjects. Nevertheless, profound alterations in both the CD4+ and CD8+ compartments were found in MF patients. Patients with MF had significantly fewer antigen inexperienced naïve (38.7±3.1% vs. 12.9±2.0%, p<0.0001; 26.5±2.6% vs. 7.0±1.2%, p<0.0001) and central memory (23.9±1.7% vs. 6.9±0.9%, p<0.0001; 10.7±1.1% vs. 2.7±0.4%, p<0.0001) T-cells than control subjects. At the same time, terminally differentiated effector memory T-cells were significantly increased in MF patients (13.0±1.0% vs. 44.0±2.5%, p<0.0001, 26.6±2.4% vs. 58.3±2.2%, p<0.0001).

To determine the effects of JAK1/2 inhibition on the T-cell subset distribution, we compared baseline (n=47) T-cell subsets with on-treatment (n=49) patient samples. Median follow-up time was 2.8 years (range: 0.2-8.0 years). We found that ruxolitinib administration increased the naïve and CM T-cells in both the CD4+ (12.9±2.0% vs. 20.1±1.4%, p=0.011; 6.9±0.9% vs. 18.3±1.2%, p<0.0001) and CD8+ populations (7.0±1.2% vs. 11.4±1.1%, p=0.02; 2.7±0.4% vs. 5.4±0.5%, p=0.0001), whereas it decreased TEM (44.0±2.5% vs. 25.5±1.7%, p<0.0001; 58.3±2.2% vs. 48.8±2.4%, p=0.0072).

Remarkably, only patients who achieved a ≥50% spleen size reduction (SR) had a significant increase in naïve CD4+ (11.0±2.5% vs. 24.2±1.8%, p=0.0002, compared with 17.0±5.2% vs. 16.0±2.4%, p=0.98 for SR <50%) and CD8+ T-cells (5.4±1.3% vs. 12.3±1.5%, p=0.0036, compared with 10.4±3.2% vs. 9.3±2.3%, p=0.96 for SR <50%) during ruxolitinib treatment.

Conclusions: In patientswith MF, T-cell subsets are skewed towards the effector phenotype. It has been shown that terminally differentiated effector memory T-cells are generated as a result of cytokine-driven rather than antigen-driven proliferation and differentiation stimuli. These highly efficient effector cells produce vast amounts of pro-inflammatory cytokines that account for and/or contribute to the chronic inflammatory milieu commonly found in MF. The JAK1/2 inhibitor reverses the equilibrium towards naïve T-cell phenotype to some extent, possibly contributing to the diminished cytokine production seen after JAK1/2 inhibition. This effect is more pronounced in patients with a better response, as measured by the degree of spleen size reduction.

In conclusion, even though MF is a disease of the myeloid lineage, we show evidence of severe immune derangements in T-cell subpopulations. Ruxolitinib might exert its benefit for MF patients due to its modulating effect on T-cells known to produce high cytokine levels.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.

This icon denotes a clinically relevant abstract