Practical management of patients with myelofibrosis receiving ruxolitinib

Abstract

Myelofibrosis (MF) is characterized by bone marrow fibrosis, progressive anemia and extramedullary hematopoiesis, primarily manifested as splenomegaly. Patients also experience debilitating constitutional symptoms, including sequelae of splenomegaly, night sweats and fatigue. Ruxolitinib (INC424, INCB18424, Jakafi, Jakavi), a JAK1 and JAK2 inhibitor, was approved in November 2011 by the US FDA for the treatment of intermediate-or high-risk MF, and more recently in Europe and Canada for the treatment of MF-related splenomegaly or symptoms. These approvals were based on data from two randomized Phase III studies: COMFORT-I randomized against placebo, and COMFORT-II randomized against best available therapy. In these studies, ruxolitinib rapidly improved multiple disease manifestations of MF, reducing splenomegaly and improving quality of life of patients and potentially prolonging survival. However, as with other chemotherapies, ruxolitinib therapy is associated with some adverse events, such as anemia and thrombocytopenia. The aims of this article are to provide a brief overview of ruxolitinib therapy, to discuss some common adverse events associated with ruxolitinib therapy and to provide clinical management recommendations to maximize patients’ benefit from ruxolitinib.

Myelofibrosis (MF) is a life-threatening and debilitating disease, and has the worst prognosis of all of the classic Philadelphia chromosomenegative myeloproliferative neoplasms (MPNs), with a median survival of 69 months [1,2]. MF can present as a primary (PMF) disorder or evolve from two other MPNs: polycythemia vera (post-PV) MF or essential thrombocythemia (post-ET) MF [3]. Although these three MPNs are clinically distinguished from one another, there do not appear to be any major prognostic or therapeutic differences between post-PV/ET MF and PMF, and most clinical trials include all three types of MF patients [4]. Regardless of its origin, MF is characterized by bone marrow fibrosis, progressive anemia and extramedullary hematopoiesis, primarily manifested as splenomegaly. Patients also experience debilitating symptoms, including sequelae of splenomegaly (e.g., pain, early satiety, limited mobility), and constitutional symptoms, such as pruritus, night sweats and fatigue [5]. The single most common cause of death for patients with MF is transformation to acute myeloid leukemia (AML; 20%); however, most patients die from other problems related to MF, such as progression without transformation, and thrombotic or cardiovascular events [2].

A breakthrough in the understanding of the pathogenesis of MPNs occurred in 2005 when several researchers identified the JAK2 V617F mutation as the most frequent molecular abnormality in Philadelphia-negative MPNs [6–9], occurring in approximately 50% of ET and PMF patients, and around 95% of PV patients [10]. Although the most common molecular mechanism for the activation of the JAK-STAT pathway is the JAK2 V617F mutation, additional mutations that activate this pathway have also been observed in patients, including mutations in JAK2 exon 12, MPL and LNK [11–13]. However, it is now widely recognized that dysregulation of the JAK signaling pathway is associated with the pathogenesis of MF, independent of the JAK2 mutation status [14].

Ruxolitinib (INC424, INCB18424, Jakafi, Jakavi), a potent inhibitor binding to the ATP-kinase domain of JAK1 and JAK2, was approved in November 2011 by the US FDA for the treatment of intermediate-or high-risk MF [15], and more recently by the European Medicines Agency [16] and Health Canada [17] for the treatment of MF-related splenomegaly or symptoms. These approvals were based on data from two pivotal randomized Phase III trials. The COMFORT trials compared ruxolitinib with placebo or physician’s choice of best available therapy (BAT) [18,19]. Ruxolitinib rapidly improved multiple disease manifestations of MF, including splenomegaly and constitutional symptoms, resulting in improved quality of life (QoL) of patients, and potentially prolonged survival [18,19]. Ruxolitinib has been generally well tolerated [18–20], but is associated with some adverse events (AEs) that require management, such as anemia and thrombocytopenia. The aims of this article are to provide a brief overview of ruxolitinib therapy, discuss some common AEs associated with ruxolitinib therapy and to provide management recommendations to optimize patients’ outcome on ruxolitinib.

Overview of ruxolitinib efficacy

Phase III COMFORT studies

Ruxolitinib monotherapy has been evaluated in patients with primary and post-PV/ET MF in the intermediate-2-or high-risk prognostic groups, according to the International Prognostic Scoring System (IPSS) [18,19]. COMFORT-I was double-blinded and placebo-controlled, whereas COMFORT-II was an open-label study that compared ruxolitinib with BAT (the physician’s choice of commercially available agents or no therapy at all). The primary end point for both studies was a reduction of ≥35% in spleen volume from baseline as assessed by MRI/computed tomography at week 24 [19] in COMFORT-I and week 48 in COMFORT-II [18]. This primary end point was chosen based on the results of the Phase I/II Study 251 [20], in which a 50% reduction in palpable spleen size was associated with a median 35% reduction in spleen volume (based on three-dimensional MRI reconstruction). In both COMFORT studies, the primary end point was met with significantly more patients in the ruxolitinib arms than in the control arms achieving a ≥35% reduction in splenomegaly.

Nearly all patients (97% for each study) treated with ruxoli-tinib had some decrease in spleen size at some time versus 24% with placebo and 56% with BAT. The median duration of response with ruxolitinib, defined as the length of time that a reduction of at least 35% in spleen volume was maintained, has not been reached in either study after 2 years, demonstrating that clinically meaningful responses achieved with ruxolitinib are durable. Additionally, ruxolitinib resulted in improvements in global health status, physical role and social functioning, as well as the individual symptoms studied (night sweats, itching, abdominal discomfort, pain under the ribs on left, early satiety and muscle/bone pain) [18,19,21]. These improvements in MF symptoms were rapid, with the majority of responses occurring within the first 4 weeks of treatment. By contrast, symptoms worsened from baseline in control-treated patients.

Furthermore, data from both COMFORT studies have suggested that ruxolitinib may prolong survival [22,23]. In the most recent update of COMFORT-I (median follow-up, 102 weeks), a survival advantage continued to be observed for patients randomized to the ruxolitinib arm (hazard ratio [HR]: 0.58; 95% CI: 0.36–0.95; p = 0.028) [22]. In COMFORT-II (median follow-up of 112 weeks), patients randomized to ruxolitinib also showed longer survival than those randomized to BAT (HR: 0.51; 95% CI: 0.26–0.99; p = 0.041; the p-value from a log-rank test is provided for descriptive purposes and was not adjusted for multiple comparisons) [23].

Which patients will benefit most from ruxolitinib therapy?

It is important to accurately identify prognostic and disease-related factors that will aid in making treatment decisions that are appropriate for individual patients diagnosed with MF [2,24]. A common misconception is that JAK inhibition is effective primarily in patients who have the JAK2 V617F mutation. However, ruxolitinib demonstrated comparable efficacy in patients with or without the V617F mutation [25,26]. Both COMFORT studies and the 251 study enrolled patients with IPSS intermediate-2- or high-risk disease. Nevertheless, MF-related symptoms and splenomegaly have been observed in patients across all IPSS risk groups [2]. Patients with intermediate-1-risk disease were included in a global compassionate access program, and accounted for 17% of patients enrolled. The symptom responses reported in this group of patients was comparable with that of higher-risk patients in the COMFORT studies. Furthermore, the IPSS and Dynamic International Prognostic Scoring System (DIPSS) risk scores for MF are based on a retrospective set of chart reviews in which the only symptoms that were believed to be relevant in measuring prognosis were weight loss >10%, night sweats and unexplained fevers, and, unfortunately, several key symptoms such abdominal pain, fatigue and pruritus, were not assessed. Thus, MF patients, even if low risk, who are symptomatic in these latter respects, may likely have a worse prognosis than would be predicted by IPSS or DIPSS [2,24]. Given that the majority of patients with MF are symptomatic and have splenomegaly [27], ruxolitinib might benefit most of the 97% of patients for whom allogeneic transplantation is not feasible [28]. Along these lines, the European Commission has identified two broad categories of patients with MF who are appropriate candidates for treatment with ruxolitinib: patients with splenomegaly and/or patients with symptoms [15–17].

Patients with symptoms

Most patients with MF have debilitating symptoms that severely impair their QoL [2,27]. Although splenomegaly is often a contributing cause of symptoms [27], symptoms can also be present in patients without splenomegaly. Indeed, there is published evidence of patients without splenomegaly who benefit from ruxolitinib [29]. Constitutional symptoms (fever, night sweats and weight loss) are prognostic factors for shortened survival [2] and are included as variables in IPSS [2] and DIPSS risk stratification [30]. Other constitutional symptoms of MF include fatigue, pruritus and bone and muscle pain [27]. Ruxolitinib has been shown to improve all of these symptoms, although some symptoms show greater response than others. For instance, the reduction in bone pain was less impressive than the responses in fatigue and pruritus. Abnormal levels of inflammatory cytokines have been linked to the presence of constitutional symptoms in patients with MF [31,32], and changes in these cytokine levels associated with ruxolitinib therapy were consistent with the observed rapid reductions in symptoms [19,20].

Patients with splenomegaly/hepatomegaly

Enlargement of the liver and spleen in MF is most frequently due to extramedullary hematopoiesis. Splenomegaly is present in approximately 90% of patients at diagnosis [2], with marked splenomegaly present in 64% of patients across all of the IPSS-defined risk groups (low, 54%; intermediate-1, 62%; intermediate-2, 70%; high, 60% [Cervantes F, Unpublished Data]). As spleen size increases, symptoms, such as early satiety, diarrhea and pain in the left upper quadrant, are generally more severe [33]. In addition, splenic sequestration is a substantial contributor to cytopenia [2,32,34] and, massive splenomegaly can sometimes cause portal vein hypertension, leading to complications such as variceal bleeding and ascites [35].

Other patients

Patients with other manifestations of MF may also benefit from ruxolitinib therapy including those with portal vein hypertension, pulmonary hypertension and extramedullary hematopoiesis. Resolution of bleeding esophageal varices has been reported after treatment with ruxolitinib [36], and in a Phase II study of patients with AML (n = 38; post-MPN AML, n = 18), ruxolitinib has shown modest antileukemic activity as a single agent, particularly in the post-MPN AML subgroup [37].

Managing common AEs associated with ruxolitinib therapy

Thrombocytopenia was identified as the dose-limiting toxicity in the Phase I study of ruxolitinib [20]. While ruxolitinib therapy has been generally well tolerated in the two Phase III studies [18,19], it is important to be familiar with the AE profile of ruxolitinib in order to maintain dose intensity and maximize response. The following sections will review the AE profile of ruxolitinib and provide practical advice for managing common AEs.

Hematologic toxicity

JAK-STAT signaling is critical for normal hematopoiesis, so reversible suppression of myelopoiesis is an expected consequence of inhibiting this pathway and would be anticipated with any effective JAK2-inhibitor treatment. Cytopenia is a common AE, but rarely resulted in therapy discontinuation in the COMFORT studies (Table 1). Cytopenia due to ruxolitinib treatment can be managed with dose reduction and/or interruptions, the authors would encourage clinicians to modify dose rather than interrupt therapy since interruption results in recurrence of symptoms. Indeed in clinical practice sometimes a lower dose is used than that recommended. For example, in patients with platelets between 100 and 200 × 10⁹/l, a starting dose of 15 mg two-times a day (b.i.d.) could be moderated to 10 mg b.i.d.

Table 1.

Hematologic laboratory abnormalities.

Laboratory parameter	COMFORT-I				COMFORT-II
	Ruxolitinib (n = 155)		Placebo (n = 151)		Ruxolitinib (n = 146)		Best available therapy (n = 73)
	All grades (%)	≥grade 3 (%)	All grades (%)	≥grade 3 (%)	All grades (%)	≥grade 3 (%)	All grades (%)	≥grade 3 (%)
	Anemia	83	45	44	16	82	40	49	21
Thrombocytopenia	71	14	21	2	69	9	29	7
Neutropenia	19	7	4	3	12	6	8	1

Data taken from [17].

Anemia

Disease-related anemia is a well-established risk factor for shorter survival [2,24,38]. Most patients with MF are anemic at presentation, with 35–54% of patients having marked anemia (<10 g/dl) at diagnosis [2,24,38], and approximately 25% of patients are transfusion dependent [24]. Furthermore, anemic patients on placebo had worse symptoms than those without anemia (COMFORT-I). The pathogenesis of anemia in MF is poorly understood, but contributing factors may include marrow failure, hypersplenism and blood loss. While anemia was one of the most frequently reported AEs in the ruxolitinib arms of both COMFORT studies, it was also frequently reported at baseline for patients in each arm of these studies and, in total, anemia led to discontinuation in only one ruxolitinib-treated patient (<1%). Decreases in hemoglobin levels in patients treated with ruxolitinib were generally transient, reaching a nadir after 8–12 weeks of therapy, and gradually recovering to near-baseline levels and indistinguishable from levels in patients treated with controls (placebo and BAT). This pattern was observed in patients regardless of whether they had received transfusions during therapy [15]. Additionally, anemia did not affect ruxolitinib efficacy. Patients treated with ruxolitinib who experienced anemia (Hb <10 g/dl) achieved reductions in spleen volume and symptoms similar to those patients with higher hemoglobin levels [25].

The emergence of anemia and thrombocytopenia in the placebo arm of COMFORT-I highlights the problem of cytopenia due to disease progression. In this situation, a falling hemoglobin level is usually accompanied by an increase in MF symptoms or spleen size, whereas anemia in patients on ruxolitinib may occur despite an improvement in symptoms. It is always important to investigate new instances of anemia.

All patients should be educated that ruxolitinib is likely to increase their risk of anemia and need for blood transfusions. The total number of units of red cells transfused to ruxolitinib-treated patients was similar to placebo-treated patients, indicating that the need for transfusions is generally transient. The authors would suggest avoiding dose modification for anemia until after the 12th–16th week. However, in an already transfusion-dependent patient, the authors consider it advisable to start at a lower ruxolitinib dose and slowly dose escalate, remembering that the anemia may improve over time.

There are emerging data from McMullin et al. suggesting that ruxolitinib can be safely co-administered with erythropoietin-stimulating agents (ESAs) [39]. Concomitant use of an ESA was reported in 13 of 146 ruxolitinib-treated patients and, while the sample size was small, concomitant ESA use did not appear to affect the efficacy of ruxolitinib and the combination was generally well tolerated. Ruxolitinib-treated patients who received an ESA had a similar median dose intensity to those who did not receive an ESA. JAK2 is downstream of the erythropoietin receptor, so there is the potential for ESAs to counter the JAK2 inhibitory effects of ruxolitinib (and other JAK inhibitors), and further studies of combination treatment should be performed. Anecdotally, some investigators have observed that the use of danazol in combination with ruxolitinib improves anemia. However, while there is an ongoing study assessing this combination, there are as yet no published results, and standard precautions with regard to monitoring danazol therapy should be observed [40].

Example case

Patient:	65-year-old woman with polycythemia vera experiencing night sweats, fatigue, 15 kg weight loss and splenomegaly.
Diagnosis:	Post-polycythemia vera MF, high risk by IPSS.
Treatment:	Initially treated with pegylated interferon, but her spleen size continued to increase, and symptoms persisted. Patient was enrolled in the COMFORT-II study, and treated with ruxolitinib 20 mg b.i.d. based on her baseline platelet count of 235 × 109/l. Her baseline spleen volume was 9600 cm3 (Figure 1a). The patient experienced a rapid improvement in sense of well-being and pruritus, and within 12 weeks, her spleen volume was reduced by 47%.
Management:	After 12 weeks of ruxolitinib therapy, her hemoglobin levels decreased from a baseline value of 12.1–8.3 g/dl, and her platelet levels decreased to 137 × 109/I. The dose of ruxolitinib was reduced to 15 mg b.i.d. for a 12-week period, and during that time, her hemoglobin and platelet levels recovered to 10 g/dl and 152 × 109/I, respectively. The dose of ruxolitinib 20 mg b.i.d. was resumed after 12 weeks, her hemoglobin levels remained between 8.5 and 9.3 g/dl and her platelet levels remained between 135 and 144 × 109/l. At the most recent follow-up, this patient’s hemoglobin was 11.5 g/dl and platelets were 175 × 109/l.
Outcome:	The patient’s spleen size continued to decrease during the period of ruxolitinib-dose reduction. After 15 months of ruxolitinib therapy, her spleen was no longer palpable, and by 24 months, the volume decreased by approximately 90% (Figure 1B). The patient remains asymptomatic and reports an improved QoL.
Commentary:	This patient had an extremely gratifying result with ruxolitinib; the images show reduction in hepatomegaly and regain of body weight. She was originally deemed unfit for an allogeneic transplant, and was recently offered this option again, but she declined as her QoL has much improved. She will require careful monitoring of her blood film and other parameters to detect early disease progression.

Figure 1. Spleen volume before and after ruxolitinib therapy.

(A) Spleen volume at baseline. (B) Spleen volume after 24 months of ruxolitinib therapy.

Thrombocytopenia

The COMFORT studies required a minimum platelet count of 100 × 10⁹/l at study entry due to the known effect of ruxolitinib on platelet count. The emergence of grade 3/4 thrombocytopenia was seen in 7% of ruxolitinib-treated patients, and in 2% of those receiving placebo. In most instances, thrombocytopenia was reversible: the median time to recovery of platelet counts above 50 × 10⁹/l was 14 days after drug cessation. Since symptoms tend to recur with cessation, it is preferable to manage platelet count by dose adjustment and to avoid stopping ruxolitinib due to thrombocytopenia.

As MF progresses, patients tend to develop disease-related thrombocytopenia because of ineffective hematopoiesis and increased splenic sequestration [41]. Disease-related thrombocytopenia has been shown to be an adverse prognostic factor for overall survival [28,38,42] and for leukemic transformation [43]. At diagnosis, approximately 20% of patients with MF have platelet counts <100 × 10⁹/l [28]. There has been limited experience with patients who have baseline platelet counts <100 × 10⁹/l. Preliminary results of two ongoing studies investigating the safety and efficacy of ruxolitinib in MF patients with baseline platelet counts of 50–100 × 10⁹/l have shown that treatment with lower doses of ruxolitinib is effective and well tolerated [44,45]. Platelet count should be used to guide the starting dose for ruxolitinib therapy (Table 2). Since thrombocytopenia is likely to occur early in treatment, monitoring should be performed every 1–2 weeks until doses are stabilized, and then as clinically indicated. Discontinuation because of thrombocytopenia with ruxolitinib or control treatments was rare (0.7% with ruxolitinib and 0.9% with control). In patients with decreasing platelet counts on treatment, dose reductions and/or interruptions of ruxolitinib should be considered, with the goal of avoiding interruptions in order to ensure successful therapy.

Table 2.

Recommended ruxolitinib starting dose by baseline platelet counts.

Baseline platelet count (×109/l)	Recommended ruxolitinib starting dose
>200	20 mg b.i.d.
100 to 200	15 mg b.i.d.
50 to <100	5 mg b.i.d.

b.i.d.: Twice daily.

Bleeding in MF is multifactorial, and may be due to thrombocytopenia, impaired platelet function, coagulation defects or portal vein hypertension. Bleeding events were more common in the ruxolitinib arms than in the control arms (33 vs 23%), with most of the excess events due to bruising or superficial bleeding. The percentage of grade 3/4 bleeding events was similar (5 vs 3%). Patients with any increased signs of bleeding should be carefully monitored, and dose reductions or interruptions of ruxolitinib should be considered.

Non-hematologic toxicities & biochemical abnormalities

Overall, grade 3/4 AEs and biochemical abnormalities were infrequent, and were rarely reported more often in the ruxo-litinib arms of the COMFORT studies than in the control arms (Tables 3–5). The three most frequent non-hematologic adverse reactions with ruxolitinib were bruising, dizziness and headache (Table 3). Whereas the most common grade 3/4 non-hematologic AEs were fatigue, diarrhea and peripheral edema (Table 4). A summary of other non-hematologic AEs and management advice is provided in Table 6.

Table 3.

Percentage of patients with adverse drug reactions ‡1% in the COMFORT studies.

Adverse event	COMFORT-I				COMFORT-II
	Ruxolitinib (n = 155)		Placebo (n = 151)		Ruxolitinib (n = 146)		Best available therapy (n = 73)
	All grades (%)	≥grade 3 (%)	All grades (%)	≥grade 3 (%)	All grades (%)	≥grade 3 (%)	All grades (%)	≥grade 3 (%)
	Any bleeding	37	5	26	3	27	5	18	3
Bruising	27	<1	15	0	15	0	6	0
Other bleeding	13	3	9	<1	14	2	14	3
Gastrointestinal bleeding	4	1	4	2	6	1	1	0
Intracranial bleeding	<1	<1	1	1	1	1	0	0
Flatulence	5	0	1	0	1	0	0	0
Pyrexia	12	<1	8	<1	15	2	10	0
Urinary tract infections	10	0	5	1	15	2	7	0
Herpes zoster	2	0	1	<1	7	<1	0	0
Weight gain	9	<1	1	<1	11	2	1	<1
Dizziness	19	<1	8	0	10	3	10	3
Headache	16	0	6	0	12	1	6	0
Angina pectoris/unstable angina	0	0	0	0	4	0	1	0
Bradycardia/sinus bradycardia	3	0	1	0	3	0	0	0
Palpitation	3	0	<1	0	5	0	1	0

Data taken from [17].

Table 5.

Biochemical laboratory abnormalities in the COMFORT studies.

Laboratory parameter	COMFORT-I				COMFORT-II
	Ruxolitinib (n = 155)		Placebo (n = 151)		Ruxolitinib (n = 146)		Best available (n = 73)
	All grades (%)	≥grade 3 (%)	All grades (%)	≥grade 3 (%)	All grades (%)	≥grade 3 (%)	All grades (%)	≥grade 3 (%)
	Elevated alanine aminotransferase	28	1	9	0	25	1	7	0
Elevated aspartate aminotransferase	19	0	7	0	20	0	4	0
Hypercholesterolemia	17	0	<1	0	16	0	7	0

Data taken from [17].

Table 4.

Non-hematologic adverse events observed in ≥10% of patients in the COMFORT studies, regardless of study drug relationship.

Adverse event	COMFORT-I				COMFORT-II
	Ruxolitinib (n = 155)		Placebo (n = 151)		Ruxolitinib (n = 146)		Best available therapy (n = 73)
	All grades (%)	≥grade 3 (%)	All grades (%)	≥grade 3 (%)	All grades (%)	≥grade 3 (%)	All grades (%)	≥grade 3 (%)
	Fatigue	25	5	34	7	12	<1	8	0
Diarrhea	23	2	21	0	23	1	12	0
Peripheral edema	19	0	23	1	22	0	26	0
Ecchymosis	19	0	9	0	NR	NR	NR	NR
Dyspnea	17	1	17	4	16	<1	18	4
Dizziness	15	<1	7	0	NR	NR	NR	NR
Nausea	15	0	19	<1	13	<1	7	0
Headache	15	0	5	0	10	1	4	0
Constipation	13	0	12	0	NR	NR	NR	NR
Vomiting	12	<1	10	<1	NR	NR	NR	NR
Pain in extremity	12	1	10	0	12	<1	4	0
Insomnia	12	0	10	0	NR	NR	NR	NR
Arthralgia	11	2	9	<1	12	<1	7	0
Pyrexia	11	<1	7	<1	14	2	10	0
Abdominal pain	10	3	41	11	11	3	14	3
Asthenia	NR	NR	NR	NR	18	1	10	<1
Nasopharyngitis	NR	NR	NR	NR	16	0	14	0
Cough	NR	NR	NR	NR	14	0	15	<1
Back pain	NR	NR	NR	NR	10	2	11	0

NR: Not reported.

Reproduced with permission from [18,19]. With permission from Massachusetts Medical Society.

Table 6.

Summary of non-hematologic adverse events and their management.

Event	Summary	Management recommendations
Fatigue	• Although this was one of the more frequently reported AEs with ruxolitinib (COMFORT-I: 25%; COMFORT-I: 18%), more patients (34%) in the placebo arm of COMFORT-I reported experiencing fatigue than those treated with ruxolitinib, suggesting that the fatigue reported on study was likely MF associated rather than treatment related	• Beyond attempting to improve fatigue through the management of anemia, non-pharmacologic interventions, such as exercise, should be encouraged [26]
Gastrointestinal events	Mild diarrhea, nausea and vomiting were common AEs reported in patients with ruxolitinib in both of the COMFORT studies (Tables 3 & 4) [18,19] • However, in the control arms of these studies, gastrointestinal events were generally reported at a similar or higher frequency than in ruxolitinib-treated patients • Flatulence was also reported as an adverse reaction (COMFORT-I: 5%; COMFORT-II: 1%) [16,17]	• Generally, these events can be controlled with antiemetic and antidiarrheal (e.g., loperamide) medications without dosage interruptions or reductions
Dizziness	Dizziness was reported in some patients enrolled in the COMFORT studies (Tables 3 & 4 [18,19]) • Most of the cases of dizziness among patients receiving ruxolitinib were determined to be unrelated to study medication	• Dizziness did not lead to treatment discontinuation for any patient in the COMFORT studies, and patients did not require concomitant medication to treat dizziness • The authors suggest investigation of other causes of dizziness
Headache	• Overall, mild headache was reported in 14% of patients receiving ruxolitinib (vs 6% of patients receiving controls), with no patient experiencing a grade 3/4 AE	• Supportive drugs are rarely necessary; however, if the duration is ≥3–4 days, short courses of nonsteroidal anti-inflammatory drugs may be helpful if the patient does not have a history of gastrointestinal bleeding and has platelet counts >100 × 109/l • Patients presenting with persistent and severe headache should be evaluated to exclude the extremely rare chance of intracranial involvement of extramedullary hematopoiesis [53–56]
Infection	• Patients who received ruxolitinib in the COMFORT studies experienced more urinary tract infections and reported more cases of herpes zoster than did patients who received controls; however, the majority of these cases were not considered to be serious, and none led to dose reductions or discontinuation of study medication (Table 3) [16,17]	• Patient should be assessed for underlying bacterial, mycobacterial, fungal or viral infections prior to the start of ruxolitinib therapy, and any active, serious infections should be resolved; in addition, patients should be warned about the possibility of reactivating these infections • Throughout the course of treatment, patients should be carefully monitored for the signs and symptoms of infection, and if an infection is noted, the appropriate treatment should be promptly initiated
Liver function abnormalities	• Asymptomatic elevations in alanine aminotransferase and aspartate aminotransferase, not associated with relevant liver toxicity, were frequently observed with ruxolitinib treatment (Table 5) [16,17]	• These elevations did not lead to discontinuations from study and did not require dose modifications • Prior to starting ruxolitinib, liver function tests should be performed and then periodically during treatment • If patients experience ≥grade 3 elevations, brief dose reductions or interruptions may be necessary until grade 1, and treatment can be resumed at the prior dose
Cardiac	• Grade 3/4 cardiac AEs were rare in the COMFORT studies (2.7% for both ruxolitinib and BAT) • Statistically significant QT prolongation was not observed in the placebo-controlled COMFORT-I study; however, at weeks 4 and 24 in COMFORT-II, significant QTc increases from baseline of mean 4–5 s were observed • Among subjects with normal PR values at baseline, the proportion who developed PR values >200 ms during treatment was 12% for ruxolitinib and 5% with placebo/BAT	• Patients with certain cardiac histories (low heart rate, history of syncope or arrhythmia, sick sinus syndrome, sinoatrial or atrioventricular block, ischemic heart disease or congestive heart failure) should be carefully observed for the occurrence of cardiac abnormalities • To the extent possible, physicians should avoid prescribing or using concomitant medications that could result in a decrease in heart rate and/or PR interval prolongation in patients treated with ruxolitinib • Prior to initiating ruxolitinib and periodically during treatment, electrocardiograms should be performed, in addition to regular monitoring of the pulse rate and blood pressure
Weight gain	• Ruxolitinib-treated patients in the COMFORT studies had substantially more weight gain than patients who received placebo or BAT • However, with patients with MF often present with cachexia and are underweight, thus the weight gain observed with ruxolitinib may be a benefit	• Like any other patient, those receiving ruxolitinib who experience undesirable weight gain should be advised about dietary and lifestyle changes

AE: Adverse events; BAT: Best available therapy; MF: Myelofibrosis.

Ruxolitinib in other special patient populations

Patients receiving concomitant CYP3A4 inhibitors

Ruxolitinib is metabolized in the liver, mainly by cytochrome P450 3A4 (CYP3A4), and is excreted primary via urine [46,47]. Thus, the dose of ruxolitinib should be reduced by approximately 50% with concomitant use of strong CYP3A4 inhibitors (such as clarithromycin, grapefruit juice, azole antifungals and antiretroviral protease inhibitors). Patients receiving any CYP3A4 inhibitor or inducer should be carefully monitored and may require dose adjustments. There are currently no data on the monitoring of plasma drug levels for ruxolitinib.

Renal & hepatic impairment

The starting dose of ruxolitinib in patients with hepatic impairment (1.5 × ULN) should be reduced by approximately 50% to be administered twice daily, with subsequent dosing titration based on monitoring of safety and efficacy. No specific dose adjustment is needed in patients with mild or moderate renal (creatinine clearance [CrCl]: 30–50 ml/min) impairment, but in patients with severe renal impairment, the starting dose should be reduced by approximately 50% to be administered twice daily. However, ruxolitinib should be avoided in patients with moderate or severe renal impairment with platelet counts <100 × 10⁹/l. In patients with end-stage renal disease (chronic kidney failure), not requiring dialysis (CrCl less than 15 ml/min), ruxolitinib should be avoided. In patients with end-stage renal disease on dialysis, the starting dose of ruxolitinib should be 15 or 20 mg once daily based on platelet counts, and should be administered only on the day of dialysis, after the procedure. Because of increased metabolite exposure and a lack of knowledge about the potential consequences of these exposures, dose modifications in this patient population should be guided by careful monitoring of safety and efficacy on an individual basis.

Pregnancy & breastfeeding

Symptomatic MF in women of childbearing potential is rare. The potential risk of ruxolitinib in humans during pregnancy is unknown, and preclinical animal studies demonstrated that ruxolitinib was embryo/fetal toxic. Accordingly, women of childbearing age should use effective contraception, and ruxolitinib should not be used during breastfeeding. The authors also suggest that effective contraception should be used by males whose partners are of childbearing potential.

Patients with secondary leukemia

Regarding leukemia progressing from MF, the usual approach is to either treat with aggressive chemotherapy with a plan to transplant in remission or to use more palliative chemotherapy [48,49]. With regard to the latter, a number of regimens are used and published data indicate that azacitidine has limited utility [50,51]. Regarding the use of ruxolitinib in patients progressing to leukemia, Eghtedar et al. recently used ruxolitinib as salvage therapy in a group of 38 heavily pretreated patients, with refractory leukemias, to target the inhibition of JAK-STAT signaling [37]. There were three responders (two complete responders and one responder with incomplete blood count recovery), and patients experienced some improvement in symptoms. There was no correlation between response and the presence of the JAK2 V617F mutation. It is possible that, in the future, combination chemotherapies including ruxolitinib may be useful for these patients; however, at present, the authors would not suggest using ruxolitinib in this setting.

Does ruxolitinib alter the natural history of MF?

Results to date have not shown an effect on leukemia-free survival and no regression of fibrosis on follow-up bone marrow biopsies. JAK2 V617F allelic burden has shown only a median reduction of around 15%. Nevertheless, there is preliminary evidence of an improvement in overall survival that appears to be due to improved performance status rather than due to arresting disease progression.

When should a patient discontinue ruxolitinib?

For responding patients, the authors recommend that ruxolitinib should be continued as long as the symptoms of disease are better than at baseline. Responses to ruxolitinib are usually seen within the first 3–6 months of treatment. For those patients in whom there is no reduction in spleen size or symptoms after this time, alternative therapies should be sought. Earlier discontinuation should be considered in case of persistent toxicity that does not resolve with dose modification.

How to discontinue ruxolitinib therapy?

For patients who require treatment interruptions, or for those who completely discontinue therapy, it should be anticipated that patients will experience a gradual return of MF-associated symptoms to levels similar to those prior to initiating ruxolitinib. For this reason, proactive dose reduction to manage AEs is preferable to drug interruption. In patients in COMFORT-I, the return of symptoms occurred within approximately 1 week [19]. For those patients who have had treatment interruptions because of toxicity, upon resolution of the events and reinitiation of ruxolitinib treatment, symptoms will likely improve similar to levels experienced prior to dose interruption. Although there have been isolated cases of patients discontinuing ruxolitinib during acute intercurrent illnesses, after which the patients’ clinical course continued to worsen [52], some of these patients were receiving higher doses of ruxolitinib than are now used in clinical practice, and it has not been established whether abrupt discontinuation of ruxolitinib contributed to these events. Similar instances have not been seen in the larger, controlled COMFORT studies, and in these studies there has been no pattern of AEs to suggest that ruxolitinib cessation is associated with any acute effects [18,19]. However, unless abrupt discontinuation is required, dose tapering of ruxolitinib may be considered, particularly in patients with an intercurrent systemic illness.

Conclusions

JAK inhibitors, such as ruxolitinib, have begun to significantly improve the treatment of MF. In addition to proven efficacy, the overall clinical experience with ruxolitinib has shown that it is generally well tolerated. Because of the newness of ruxolitinib in standard practice, the authors recognize that their recommendations are based on the current data and may need future refinement. However, with diligent monitoring and dose adjustment in response to AEs, they expect that successful treatment with ruxolitinib will help many patients to regain QoL, and have prolonged survival. In addition, there are a number of alternative JAK inhibitors at varying phases of clinical development which may yet prove to have a different efficacy or safety profile than ruxolitinib.

Expert commentary

JAK inhibitors, such as ruxolitinib, have begun to revolutionize the treatment of MF. In addition to proven efficacy, the overall clinical experience with ruxolitinib has shown that it is generally well tolerated. However, as with all drugs, some AEs occur and require management to ensure optimal patient outcomes. With careful monitoring and dose adjustment, the authors expect that successful treatment with ruxolitinib will help many patients to regain QoL, and have prolonged survival. As clinicians gain more experience treating patients with MF, the authors anticipate that the management of patients will evolve and improve over time and as other JAK inhibitors with unique AE profiles become available, it may be possible to select a compound that is best suited to a particular patient’s requirements.

Example case

Patient:	68-year-old man with PMF experiencing night sweats, fatigue, anemia and splenomegaly.
Diagnosis:	PMF, high risk by IPPS.
Treatment:	Initially treated with hydroxycarbamide, but his spleen size continued to increase, symptoms persisted and anemia worsened, such that he required a transfusion. His spleen was palpable 26 cm below the costal margin.
	Patient was enrolled in the COMFORT-II study, and treated with ruxolitinib 15 mg b.i.d. based on his baseline platelet count of 110 × 109/l.
	The patient experienced an improvement in sense of well-being, and his spleen length was reduced 2 cm on palpation.
Management:	After 12 weeks of ruxolitinib therapy, hemoglobin levels remained stable, but his platelet levels decreased to 75 × 109/l.
	The dose of ruxolitinib was reduced to 10 mg b.i.d., and thereafter, to 5 mg b.i.d. as his platelet count continued to decrease.
	The patient’s platelet count remained between 45 and 55 × 109/l (Figure 2), his spleen enlarged beyond baseline and symptoms returned.
Outcome:	After discussion with the patient, he was withdrawn from ruxolitinib therapy since frequent hospital visits to monitor thrombocytopenia were burdensome, and the patient seemed to have little benefit because of continued dose reductions and interruptions. Symptoms worsened once ruxolitinib was finally stopped, and he was offered hydroxycarbamide. He then transferred to another clinical trial of a different agent.
Commentary:	Due to thrombocytopenia, this patient was not achieving target response to ruxolitinib. He was subsequently enrolled into a clinical trial with a different JAK inhibitor and is tolerating this agent. This, perhaps, reflects a parallel to chronic myelogenous leukemia therapy in which some tyrosine kinase inhibitors seem to better suit some patients in terms of tolerance or comorbidities.

Figure 2.

Platelet levels over time.

Five-year view

As with any new treatment, it will be important to continue to assess the long-term benefits and safety profile of ruxolitinib. Currently, improvement in bone marrow fibrosis observed in the COMFORT studies has been modest, but longer follow-up may be required to see improvements in patients treated with JAK inhibitors. Additionally, several studies are underway that will explore the combination of ruxolitinib with other drugs such as deacetylase, mTOR and smoothened inhibitors to determine if combination therapies can further improve patient outcomes. Finally, results from studies evaluating ruxolitinib in patients with PV and ET are anticipated and it will be important to assess the impact of early JAK inhibitor intervention in these diseases and if this has an impact on reducing the transformation rate to PPV-MF and PET-MF.

Key issues.

• Myelofibrosis (MF) is a relatively rare condition and with new MF therapies becoming commercially available, it is important to review the clinical challenges of treating MF and to become familiar with how to manage patients receiving JAK inhibitor therapy.

• Ruxolitinib is a JAK1 and JAK2 inhibitor that rapidly improved multiple disease manifestations of MF, reducing splenomegaly and improving quality of life (QoL) of patients, and potentially prolonging survival.

• Ruxolitinib was approved based on data from two pivotal randomized Phase III trials: The COMFORT trials compared ruxolitinib with placebo or physician’s choice of best available therapy (BAT).

• JAK-STAT signaling is critical for normal hematopoiesis, so reversible suppression of myelopoiesis is an expected consequence of inhibiting this pathway.

• Cytopenia due to ruxolitinib treatment can be managed with dose reduction and/or interruptions.

• Overall, grade 3/4 AEs and biochemical abnormalities were infrequent, and were rarely reported more often in the ruxolitinib arms of the COMFORT studies than in the control.

• With diligent monitoring and dose adjustment in response to AEs, the authors expect that successful treatment with ruxolitinib will help many patients to regain QoL, and have prolonged survival.