Real-World Use of Ruxolitinib in Patients with Myelofibrosis and Anemia or Thrombocytopenia at Diagnosis

Myelofibrosis (MF) is a progressive myeloproliferative neoplasm characterized by megakaryocyte proliferation, constitutional symptoms due to inflammatory cytokine production, bone marrow fibrosis, and splenomegaly [1‒3]. Patients with MF commonly have cytopenias at diagnosis, with approximately one-third having anemia and one-fourth having thrombocytopenia [4, 5]. Cytopenic MF is associated with worse overall survival (OS) compared to MF without cytopenias [6]. The selective oral Janus kinase (JAK) 1/JAK2 inhibitor ruxolitinib (RUX) is approved by the US Food and Drug Administration for the treatment of adult patients with intermediate- or high-risk MF [7]. Regulatory approval was based on results from the randomized, controlled, phase 3 COMFORT-I and COMFORT-II trials [8, 9]. Patients treated with RUX in the COMFORT trials experienced clinical benefit, including improvements in symptoms, spleen size, and OS [8‒11]. Although RUX treatment was associated with dose-dependent anemia and thrombocytopenia, typically occurring in the early part of the treatment period, both conditions were generally manageable and rarely required treatment discontinuation [8, 9]. Whereas the COMFORT trials excluded patients with low platelet counts (PLTs) and had no specific anemia eligibility requirements, additional studies including EXPAND, REALISE, and JUMP evaluated RUX in patients with MF and anemia or thrombocytopenia and found similar efficacy in these cytopenia subgroups as that observed in the COMFORT trials [12‒16]. However, real-world data outside of clinical trials in patients with cytopenias are lacking. One recent real-world study investigated RUX treatment patterns stratified by PLTs; however, patients in that study were required to have discontinued RUX treatment by the time of the analysis, and only 37.5% of patients received RUX as initial MF therapy [17]. Therefore, this retrospective analysis of physician-reported data was conducted to describe the characteristics, RUX treatment patterns, and clinical outcomes of patients with MF and anemia and/or thrombocytopenia at MF diagnosis who received RUX as initial therapy in real-world clinical practice outside of a clinical trial setting.

The study design and patient selection criteria for this study have been previously described [18]. Briefly, adult patients ≥18 years old diagnosed with primary MF, post–polycythemia vera MF, or post–essential thrombocythemia MF between January 1, 2012, and December 31, 2016, who received care for ≥6 months at US community hematology/oncology practices in the Cardinal Health Oncology Provider Extended Network (OPEN) were included. Physicians at the 45 participating sites reviewed medical records and provided data between May and July 2018 for eligible patients.

This study was a secondary analysis of the retrospective medical chart review described above, encompassing all 176 patients who received RUX as initial therapy for MF (other supportive care was permitted before or during RUX treatment, including erythropoiesis-stimulating agents, immunomodulatory agents, steroids, red blood cell [RBC] transfusions, and participation in a clinical trial). Patient subgroups were analyzed based on the presence or absence of anemia (hemoglobin <10 or ≥10 g/dL, respectively). Patients with anemia were further subcategorized as being RBC transfusion dependent (TD) or not dependent, as determined by the treating physician and reflective of individual community practice approaches. Additionally, patient subgroups based on presence or absence of thrombocytopenia (PLT, ≤100 × 10⁹/L or >100 × 10⁹/L, respectively) at MF diagnosis were analyzed. Patients with thrombocytopenia were further subcategorized by PLTs <50 × 10⁹/L or 50–100 × 10⁹/L. Physician-reported patient characteristics and treatment patterns were analyzed for the total population and the subset of patients with or without anemia and with or without thrombocytopenia.

This study was conducted in accordance with the ethical standards of the Declaration of Helsinki. Because of the retrospective nature of this analysis and the use of de-identified patient data, informed consent was not obtained from patients. Western Institutional Review Board confirmed that the study met requirements for the waiver of consent.

Median OS was estimated using Kaplan-Meier analysis and defined as the time from initiation of RUX to death. Any patients still alive at the time of evaluation were censored at the date of the last follow-up. In the Kaplan-Meier survival analysis, survival medians were reported as not reached if survival probability did not reach 50%. Confidence intervals for this median were not estimable (NE) if an event could not be accurately estimated, such as when the median survival time was the last event but not the final observation due to censored observations. All other data were analyzed with descriptive statistics.

Demographics and clinical characteristics at MF diagnosis of the 176 patients who received RUX as initial therapy are shown in Table 1. Mean age at MF diagnosis was 65.1 years, and most patients were male (60.8%) and White (64.2%). The majority of patients had anemia (68.2%), whereas 33.5% had PLTs ≤100 × 10⁹/L. Of the patients who underwent molecular testing for the JAK2V617F mutation, 76.1% were positive. Similarly, 19.2% and 20.7% of patients tested for CALR and MPL mutations, respectively, were positive, whereas 21.2% of patients tested for all 3 driver mutations had triple-negative status. Median (range) length of follow-up after initial MF diagnosis was 21.4 (0–54.9) months.

Of the 176 patients included in this study, 120 had anemia at diagnosis, among whom 42.5% (n = 51) were RBC TD per the treating physician (Table 1). Median (range) time from diagnosis to RUX initiation was 42.5 (0–1,338) days for patients with anemia and 32.5 (0–1,166) days for patients without anemia, with RUX most frequently started at 15 mg twice daily (bid; 44.2%) and 20 mg bid (44.6%), respectively (Table 2). Median (range) length of follow-up for patients with anemia was 21.2 (0–54.9) months, and most patients (69.2%) remained on RUX at the end of the study (Fig. 1a). For patients without anemia, median follow-up was 21.6 (1.1–42.5) months, and 91.1% remained on RUX by end of study. During RUX treatment, 12.5% of patients with anemia had dose increases and 18.3% had dose decreases, with thrombocytopenia the most common reason for dose reduction (86.4%). Despite dose decreases during the study, most patients were able to maintain a RUX dose ≥10 mg bid as evidenced by end-of-study dose, regardless of anemia status (patients with anemia, 88.3%; patients without anemia, 94.6%).

Overall, RUX was discontinued in 30.8% of patients with anemia and 8.9% of patients without anemia, and the majority did not receive dose adjustments prior to discontinuation, regardless of anemia status (Fig. 1a). The most common reason for treatment discontinuation in patients with anemia was disease progression (37.8%; Fig. 2). Discontinuation rarely resulted from adverse events, with only 1 patient each discontinuing for anemia and thrombocytopenia in the anemia subset.

Per the treating physicians, 59.6% of patients with anemia who had a palpable spleen at diagnosis had reductions in spleen size compared with 70.5% of patients without anemia (Fig. 3a). Symptom improvements were observed in 80.3% of patients with anemia and 78.2% of patients without anemia (Fig. 3a). Median OS from initiation of RUX was 45.1 (95% CI: 34.4, NE) months for patients with anemia compared with 42.5 (NE, NE) for patients without anemia.

At MF diagnosis, 59 patients (33.5%) had PLTs ≤100 × 10⁹/L (27 [15.3%] with PLTs <50 × 10⁹/L; Table 3). Nearly all patients with PLTs ≤100 × 10⁹/L at diagnosis also had anemia (56/59 [94.9%]) compared with 54.7% (64/117) of patients with PLTs >100 × 10⁹/L.

Median (range) time from diagnosis to RUX initiation was 50.0 (7–1,338) days for patients with PLTs ≤100 × 10⁹/L at diagnosis and 38.0 (0–1,166) days for patients with PLTs >100 × 10⁹/L. The most common RUX starting dose for patients with PLTs ≤100 × 10⁹/L was 15 mg bid (44.1%) compared with 20 mg bid for patients with PLTs >100 × 10⁹/L (43.6%; Table 4). Median (range) length of follow-up was 21.0 (2.3–54.9) months for patients with PLTs ≤100 × 10⁹/L and 21.5 (0–52.0) for patients with PLTs >100 × 10⁹/L, and most patients remained on RUX at the end of the study regardless of PLT (Fig. 1b). For patients with PLTs ≤100 × 10⁹/L, 22.0% had dose increases and 22.0% had dose decreases, and most were able to maintain a RUX dose ≥10 mg bid at end of study (Table 4).

RUX treatment was discontinued in 32.2% of patients with PLTs ≤100 × 10⁹/L and 19.7% of patients with PLTs >100 × 10⁹/L, and most patients did not receive dose adjustments prior to RUX discontinuation (Fig. 1b). Disease progression and no response were the most common reasons for discontinuation for patients with PLTs ≤100 × 10⁹/L (Fig. 2).

Per the treating physicians, among patients with a palpable spleen at diagnosis, 63.8% of those with PLTs ≤100 × 10⁹/L and 62.1% with PLTs >100 × 10⁹/L had spleen size reductions (Fig. 3b). Symptom improvements were achieved by 74.6% and 82.3%, respectively (Fig. 3b). Median OS (95% CI) from RUX initiation was not reached (25.4, NE) for patients with PLTs ≤100 × 10⁹/L and 42.5 (41.5, NE) months for patients with PLTs >100 × 10⁹/L.

In this real-world analysis of patients with MF who had anemia or thrombocytopenia at MF diagnosis, treatment with RUX was generally well tolerated and provided improvements in spleen size and symptoms. Most patients were initiated on and maintained RUX doses at or above 10 mg bid at the end of the follow-up period (median, 21.4 months), including TD patients with anemia and those with PLTs <50 × 10⁹/L. Additionally, most patients continued treatment with RUX throughout the follow-up period, with rare discontinuations due to anemia or thrombocytopenia (occurring in 1 patient each, both in patients with anemia at MF diagnosis). Physician-determined spleen size reduction (of any measure) was achieved in all subgroups, including majorities of patients who were TD and those with PLTs <50 × 10⁹/L. Physician-determined symptom improvement (of any measure) was also uniformly observed across subgroups, including by more than three-quarters of patients who were TD and those with PLTs <50 × 10⁹/L. It is important to note that because patients with PLTs <50 × 10⁹/L were excluded from prospective interventional trials with RUX, safety data are limited in this patient population, and dose recommendations are not provided in the RUX prescribing information [7].

Results from this real-world analysis are consistent with those from the COMFORT trials, which excluded patients with PLTs <100 × 10⁹/L and had no specific eligibility requirements regarding anemia. Results from COMFORT demonstrated that RUX reduced the size of enlarged spleen, improved symptom burden, and prolonged OS [8‒11]. The current findings are also largely consistent with previous clinical trial reports of RUX use in patients with MF who had anemia or low PLTs. In the phase 1b dose-finding EXPAND study in patients with PLTs <100 × 10⁹/L (split into groups of 50–74 × 10⁹/L and 75–99 × 10⁹/L), patients experienced spleen and symptom improvements, with low levels of discontinuation due to thrombocytopenia in patients with PLTs of 75–99 × 10⁹/L [12]. In EXPAND, RUX 10 mg bid was the recommended maximum safe starting dose, whereas in a separate phase 2 study, patients with PLTs of 50–100 × 10⁹/L received RUX 5 mg bid and were up-titrated based on response and hematologic parameters [12, 13]. In this phase 2 study, which is aligned with dosing recommendations in the RUX label [7], no patients withdrew due to an adverse event, and improvements in spleen volume and symptoms were observed, especially at a dose of 10 mg bid [13]. Clinical data in anemic patients with MF who were treated with RUX were reported in the phase 2 REALISE study, in which patients were initiated on RUX 10 mg bid and up-titrated [14]. Spleen response and symptom improvement were observed for many patients, including 70% with a ≥50% reduction in palpable spleen length at any time; only 1 patient discontinued due to worsening anemia. Finally, results from a retrospective study of patients with cytopenic MF (N = 407), with a median follow-up time of 4.4 years, also indicated tolerance for RUX and improvements in spleen size and symptom burden [19]. The median overall dose of RUX was ≥10 mg bid in 80% of patients with cytopenic MF. At 6 months of treatment, nonpalpable spleen was achieved by 27% of patients, and 60% achieved a ≥50% reduction in the MPN Symptom Assessment Form Total Symptom Score.

Results reported here are also consistent overall with previous findings from the expanded-access JUMP study. Clinically meaningful spleen size reductions and improvements in symptoms were reported in a subgroup of patients with low PLTs (<100 × 10⁹/L), with safety and efficacy consistent with data from the COMFORT trials [15, 16]. Consistent with this study, most (91%) patients in JUMP initiated RUX at doses >10 mg bid, and few patients discontinued due to anemia (2.0%) or thrombocytopenia (3.4%; 10.1% in the low-platelet group). The tolerability and efficacy results from this analysis support the benefits of RUX use in patients with anemia [20].

Several limitations of this study common to retrospective analyses should be noted. The analysis was limited to data available in the chart review database. Some information, such as change in hemoglobin or platelet levels while on RUX treatment, were not collected in the original chart review and were not assessable. Data on symptom and spleen size were qualitative and limited, reflecting how patients were assessed at community practices, which may not equate to the level of benefit observed in prospective clinical trials. Additionally, it should be noted that all data are physician-reported; thus, key aspects including TD status and spleen and symptom response outcomes are reflective of community practice norms at each study center and may not be broadly generalizable. Further, the study data collection occurred before additional JAK inhibitors were approved for the treatment of MF. Thus, the current management of patients with MF and anemia and/or thrombocytopenia is not reflected in this study. Finally, because supportive care treatments were not considered to be first-line treatments for MF, delay before RUX initiation and percentage of patients undergoing watch-and-wait observation may be overstated. Nevertheless, results reported here are largely consistent with previous outcomes reported for RUX from clinical trials in patients with thrombocytopenia or anemia.

In conclusion, this real-world analysis indicates that patients with MF with anemia and/or thrombocytopenia at MF diagnosis derived clinical benefit from RUX, with a generally well-tolerated safety profile consistent with previous clinical trials. Most patients were able to receive RUX doses at or above 10 mg bid and continued RUX treatment during the entire follow-up period. Discontinuations due to anemia or thrombocytopenia were rare, each occurring in <1% of patients with anemia and in no patients with thrombocytopenia. Taken together, these data provide further evidence that RUX is clinically beneficial and well tolerated in patients with MF, including those with anemia or low PLTs.

Writing assistance was provided by Cory Pfeiffenberger, PhD, an employee of ICON (Blue Bell, PA).

As this was a retrospective review of patient data, a waiver of authorization and exemption for ethical approval was granted by Western Institutional Review Board. As this was a retrospective review of patient data, a waiver of authorization and exemption for patient informed consent was granted by Western Institutional Review Board.

Jingbo Yu and Evan Braunstein are employees and shareholders of Incyte Corporation. Thomas Cordaro was a summer intern at Incyte at the time of the study. Thomas Cordaro was a Summer intern at Incyte Corporation during this study. Emily Bland and Tammy Schuler are employees of Cardinal Health.

Incyte Corporation sponsored this study, funded writing assistance provided by ICON (Blue Bell, PA), participated in study design development, and collaborated with authors in the analysis and interpretation of the data and writing the manuscript.

J.Y.: concept of design; analysis and interpretation; drafting of the manuscript; critical revision of the manuscript for important intellectual content; statistical analysis; obtaining funding; administrative, technical, or logistic support; supervision. E.Bla. and T.S.: concept of design; analysis and interpretation; drafting of the manuscript; critical revision of the manuscript for important intellectual content; statistical analysis. T.C.: analysis and interpretation; drafting of the manuscript; critical revision of the manuscript for important intellectual content. E.Bra.: concept of design; analysis and interpretation; drafting of the manuscript; critical revision of the manuscript for important intellectual content; obtaining funding; supervision.

Access to individual patient-level data is not available for this study. Information on Incyte’s clinical trial data sharing policy and instructions for submitting clinical trial data requests are available at: https://www.incyte.com/Portals/0/Assets/Compliance%20and%20Transparency/clinical-trial-data-sharing.pdf?ver=2020-05-21-132838-960. Further inquiries can be directed to [email protected].