Approach to MPN Symptom Assessment

Abstract

Purpose of Review

Myeloproliferative neoplasms (MPNs) are recognized for their debilitating symptom burdens. The purpose of this review is to understand the complexity of the MPN symptom burden and identify how validated MPN Patient Reported Outcome (PRO) tools may be integrated into clinical practice to assess the MPN symptom burden.

Recent Findings

Significant heterogeneity exists both within and between MPN subtypes. Surrogates of disease burden such as risk scores and MPN chronicity often fail to correlate with symptomatic burden. Validated MPN PROs allow for precise and rapid assessment of the MPN symptom burden in clinical and trial settings. Their growing use among investigators has resulted in improved understanding of how the MPN disease burden and overall patient experience is impacted by novel and traditional therapies.

Summary

PRO tools are an integral part of National Comprehensive Cancer Center (NCCN) guidelines for MPN treatment and should be regularly employed in disease burden assessment.

Introduction

For the past century, clinicians have struggled to understand the genotypic and phenotypic complexities that collectively define the myeloproliferative neoplasms (MPNs). As early as the 1950s, and with little more than basic histochemistry and a clinical exam at their disposal, clinicians were quick to acknowledge the uniqueness of the MPN profile as one of intense symptomatic burden, marked splenomegaly and profound myeloid hyperplasia or paradoxical cytopenias [1•]. From these early beginnings, our understanding of the MPNs including polycythemia vera (PV), essential thrombocythemia (ET), and myelofibrosis (MF), as distinct but interrelated hemopathies has advanced, prodding us to address the uncertainty of how often seemingly benign disease states can induce such profound symptomatic responses. Research into this conundrum has led to important understandings of the interdependence of gene mutations and cytokine signaling cascade disruptions on clinical presentation. In this article, we examine the origins of MPN symptomatology, the role of symptoms as independent metrics of disease burden and integration of symptom assessment into clinical practice.

Overview of the MPNs

The MPNs are recognized clonal stem cell hemopathies that arise from dysregulation of myeloid precursors. Each of the major subtypes (ET, PV, MF) are typified by their own unique genotypic abnormalities, symptom profiles, complications, and prognosis. In the case of PV, the disorder is characterized by erythrocytosis, presence of the JAK2V617F mutation (96%), predilection for thrombotic complications, severe constitutional symptoms, mild to moderate splenomegaly, and risk of transformation to myelofibrosis (Post-PV MF; 12–21%) or acute myeloid leukemia (AML; 7%) [2–4]. Prognosis is determined by using the 2013 Leukemia Criteria which includes the risk factors of age ≥ 67 or 57–66 years, venous thrombosis history, and leukocytes ≥ 15 × 10(9)/L [5••].

In contrast, ET manifests with thrombocytosis, debilitating vasomotor symptoms, recurrent venous or arterial thrombotic and hemorrhagic complications, mild to moderate splenomegaly, JAK2 (60%), CALR (25%), MPL (5%) or “triple-negative” (15%) gene mutation status, and a low risk of transformation to post-ET MF (2%) or AML (4%) [3, 6]. Median survival is similar to age-matched controls. Prognostic scoring is performed using the International Prognostic Score for Essential Thrombocythemia (IPSET) and includes the features of age ≥ 60 years, white blood cell count ≥ 11 × 10(9)/L, and history of thrombosis [7].

The most debilitating MPN subtype, MF may be primary (PMF) or result secondarily from ET or PV transformation (post-ET MF and post-PV MF). The condition is characterized by significant hypercatabolic symptoms; severe splenomegaly; profound cytopenias; marrow fibrosis with resultant extramedullary hematopoiesis; gene mutations in JAK2 (65%), CALR (25%), MPL (5%), or absence of all three (“triple negative”; 10%); risk of transformation to AML, and a reduced life expectancy ranging from 15 years (low risk) to 1.3 years (high risk) [8–11]. Prognosis is estimated using the Dynamic International Prognostic Scoring System Plus (DIPSS Plus) which includes the variables of age > 65 years, leukocyte count > 25,000 × 10(9)/L, hemoglobin < 10 g/dL, circulating blast cells ≥ 1%, constitutional symptoms, platelet count < 100,000 × 10(9)/L, anemia requiring transfusions, and unfavorable karyotype [11].

Despite the variable profiles of MPN subtypes, all seem to have a similar constellation of symptoms. The most commonly recognized symptoms include fatigue, early satiety, abdominal discomfort, inactivity, concentration problems, night sweats, itching, bone pain, fevers, and weight loss (Fig. 1) [12]. In 2017, the National Comprehensive Cancer Network (NCCN) published revised guidelines incorporating symptomatic assessment into the response criteria for MF [13••]. This momentous event was one of many advancements signifying societal and institutional recognition of the important role symptoms play within the disease. NCCN guidelines are currently in process for ET and PV. In tandem, numerous investigations have focused on unearthing the intricacies of symptom development and propagation. The task has proven to be complex. As stated, significant heterogeneity and overlap exists among the MPN subtypes, particularly as they relate to symptoms. In 2013, our team performed an investigation into the MPN symptom burden with the aim of identifying associations between patient symptoms and disease features including prognostic risk scores [14]. The study utilized a cluster analysis approach from prospectively gathered symptom burden data on 1470 PV, ET, and MF patients and identified that individual clusters exist within each MPN subgroup (5 clusters in ET, 5 clusters in PV, 4 clusters in MF). The study furthermore demonstrated that many low-risk MPN patients struggle with significant symptomatic burden, independent of disease features, length of disease, or prognosis. It should be noted that a few MPN symptoms have indeed been shown to impact patient prognosis (constitutional symptoms in MF, pruritus in PV) and are included as components to MF risk scoring [5•, 11].

Fig. 1.

MPN-SAF TSS symptoms by frequency (a) and severity (b)

Associations between MPN symptoms and disease features have been an area of intense study. Recent investigations have identified that not only do symptoms overlap with other symptoms, but that they also closely correlate with other disease features such as laboratory derangements, the presence of vascular complications, mutational status, and risk scores [15, 16•, 17–19]. For instance, we performed a study evaluating the impact of thrombocytopenia (platelets < 100 × 10(9)/L) on patient symptomatology in MF. Significantly worse symptom burden was noted for most symptoms in patients with thrombocytopenia despite similar prognostic scores to patients without thrombocytopenia [20]. Patients with severe thrombocytopenia (< 50 × 10(9)/L) showed similar symptom scores to patients with moderate thrombocytopenia (51–100 × 10(9)/L) suggesting that the presence of thrombocytopenia is more important than degree of severity. Cytokines, including interleukins, interferons, and soluble growth factors, also play a critical role as regulators of cellular processes as they relate to cell growth, migration, and apoptosis. Dysregulation of cytokine signaling processes are recognized drivers of the chronic inflammation that contribute to many MPN symptoms. In a 2013 analysis of the blinded phase of COMFORT-1 trial, MF patients had altered levels of five cytokines (VCAM1, LEPTIN, TNFRII, TIMP1, B2MICG) which correlated with changes in the MPN-SAF TSS when controlled for other variables such as age, sex visit by trial arm interaction, arm, and body mass index (BMI) [21•]. Evaluations such as these and others clearly require objective data on MPN symptom burden via validated scoring tools. Below, we discuss the development of these tools and their use in clinical practice.

Assessment of the MPN Symptoms

The burdensome MPN symptom profile has long been regarded a perplexing disease feature that stems from a diverse interplay between inflammatory dysregulation, progressive anatomic changes (hepatomegaly or splenomegaly), and confounding psychosocial stressors [14]. The historic variability with which most patients symptomatically present has posed challenges to clinicians aiming to perform an adequate appraisal of the symptom burden. Objective information as gleaned from the physical examination, laboratory data, cytogenetic analysis and radiographic imaging has historically served as the benchmark by which disease progression or improvement was monitored. Attempts to characterize symptomatology were performed either subjectively through verbal office interviews or with non-dedicated patient-reported outcome (PRO) tools such as the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire Core model (EORTC QLQ-C30), Brief Fatigue Inventory (BFI), Functional Assessment of Cancer Therapy-Anemia (FACT-AN), and Godin Leisure Time Activity Score (LAS). Each of these tools failed to capture the breadth, severity and frequency of classical MPN symptoms. Complicating this issue has been evidence demonstrating that many MPN symptoms progress longitudinally thru time while others plateau [22]. Fatigue, concentration difficulties, insomnia, sexual concerns, cough, night sweats, and pruritus all appear to worsen in severity and early satiety, concentration difficulties, insomnia, sexual difficulties, cough, and night sweats become more prevalent. Abdominal pain/discomfort, inactivity, headaches, dizziness, numbness, sad mood, bone pain, fevers, and weight loss all appear to remain stable throughout the disease course. Quality of life (QOL), an important metric of MPN disease status, also diminishes over time.

In 2007, the first attempt to understand the MPN symptom profile was performed through an internet-based survey of 1179 MPN patients [23]. Patients were queried on symptoms as they relate to splenomegaly, pruritus (52.2%), fatigue (80.7%), fevers (13.7%), night sweats (49.2%), bone pain (43.9%), and undesired weight loss (13.1%). The results demonstrated a profound deficit in symptom management for each of these complaints, despite use of available treatment regimens such as immunomodulatory agents, steroids, hydroxyurea (HU), androgen analogs, and erythropoiesis-stimulating agents (EPOs). Patients furthermore had impairments in their ability to perform work outside of the home, and many were medically disabled because of the disorder. Of particular interest was the finding that multiple sources may account for any one MPN symptom. In the case of fatigue, anemia, transfusion dependency, other symptoms (fever, weight loss, pruritus), and prior thrombohemorrhagic complications were all associated with increasing fatigue. A noteworthy finding was the observation that “severe” symptom burden may be observed in otherwise early stage or “asymptomatic” disease states and that this symptom burden frequently outscores the burden for other prominent malignancies such as non-Hodgkin lymphoma (NHL).

With this information, our team undertook the first effort to design a dedicated MPN PRO. The 20-item Myelofibrosis Symptom Assessment Form (MF-SAF) became the first validated MPN-PRO for use in MF and incorporated items as related to abdominal discomfort, fatigue, cough, night-sweats, bone pain, weight loss, early satiety, abdominal pain, inactivity, pruritus, fever, and quality of life (Table 1) [24]. Items were scored on a “no,” “yes,” or 0 (absent) to 10 (worst imaginable) rating scale. The PRO was co-administered with the Brief Fatigue Inventory (BFI), Brief Pain Inventory (BPI), and memorial Symptom Assessment Scale (MSAS).

Table 1.

MPN patient-reported outcome tools

	MF-SAF	MPN-SAF	MPN-SAF TSS
MPN subtype assessed	MF	ET, PV, MF	ET, PV, MF
Items in score	20	27	10
Comparison surveys	Memorial Symptom Assessment Scale (MSAS), Brief Fatigue Inventory (BFI)	EORTC QLQ-C30, BFI	EORTC QLQ-C30, BFI

EORTC QLQ-C30 European Organization for Research and Treatment of Cancer Quality of Life Questionnaire Core Model, ET essential thrombocythemia, MF myelofibrosis, MPN myeloproliferative neoplasm, PV polycythemia vera, SAF symptom assessment form, TSS total symptom score

The scoring tool proved sensitive in detecting MF symptoms and was quickly expanded to include other MPN disorders including ET and PV. Entitled the Myeloproliferative Neoplasm Symptom Assessment Form (MPN-SAF), this 27-item tool was enhanced with microvascular-related symptoms including insomnia, sexual dysfunction, vertigo, lightheadedness, dizziness, numbness/tingling, headaches, and concentration difficulties (Table 1) [12]. An updated version of the MPN-SAF reduced the question length to the top 10 most frequent and relevant MPN symptoms (Table 1). Entitled the MPN-SAF Total Symptom Score (MPN-SAF TSS; MPN-10), this tool has since been validated and made available in 21 languages for 26 countries [25, 26]. Each of these tools has proven modifiable for use in clinical trials (Table 2). For instance, the MF-SAF 2.0 was used as an abbreviated version of the MF-SAF in COMFORT-1 and focused primarily on abdominal discomfort, pain under the left ribs, early satiety, night sweats, itching, bone or muscle pain, and inactivity. A modified version of the MPN-SAF (Modified Myelofibrosis Symptom Assessment Form) incorporated only six key symptom items and was used in the evaluation of the safety and efficacy of fedratinib in MF populations [27]. Additionally, the MPN-SAF has been modified to focus primarily on cytokine effects as the MPN-SAF TSS Cytokine Symptom Cluster (TSS-C) which evaluated tiredness, itching, muscle aches, night sweats, and sweats while awake [28••]. The tool proved useful in the randomized, double blind, double-dummy symptom study (RELIEF) in PV populations.

Table 2.

MPN PRO use in clinical trials by therapy

	Trial Name	MPN	PRO
Ruxolitinib	COMFORT 1	Myelofibrosis	MF-SAF 2.0
	COMFORT 2	Myelofibrosis	FACT-LYM
	RETHINK	Myelofibrosis	MPN-10
	RESPONSE	Polycythemia vera	MPN-SAF
	RELIEF	Polycythemia vera	MPN-SAF
	MAGIC	Essential thrombocythemia	MPN-SAF
Fedratinib	JAKARTA	Myelofibrosis	MF-SAF
Pacritinib	PERSIST 1&2	Myelofibrosis	MPN-SAF
Momelotinib	SIMPLIFY 1&2	Myelofibrosis	MPN-SAF
Pomalidomide	RESUME	Myelofibrosis	FACT-AN
PEG INFa2a	MPD-RC112	Myelofibrosis	MPN-SAF
	MPD-RC112	Essential thrombocythemia	MPN-SAF

FACT-LYM Functional Assessment of Cancer Therapy in Lymphoma, MF-SAF Myelofibrosis Symptom Assessment Form, MPN-SAF Myeloproliferative Neoplasm Symptom Assessment Form, PEG INFa2a pegylated-interferon alpha-2a

Using MPN PROs in the Clinical Setting

As MPN symptoms are integrally associated with perceived patient QOL, disease progression, and the ultimate selection of therapies, their assessment becomes an important part of routine clinical evaluation. Serial use of available MPN scoring tools allows for direct assessment of the patient experience and has been proven in clinical trial settings to be sensitive to symptomatic changes for therapies or disease complications. Results may furthermore corroborate other more objective parameters such as weight loss and performance status [29]. In 2017, the National Comprehensive Cancer Network (NCCN) developed guidelines for PRO use in MF [13••]. Guidelines for ET and PV are currently in progress. Per NCCN guidelines, all MF patients regardless of risk status should undergo assessment with the MPN-SAF TSS during the initial evaluation, with subsequent repeat PRO assessment every 3 to 6 months to monitor for symptom progression or response to therapy. Objective symptomatic assessment offers the added benefit of highlighting populations that may benefit from additional imaging or laboratory evaluation. For instance, worsening abdominal pain in the setting of stable splenomegaly may signify other pathologies such as splenic, hepatic, or portal vein thrombosis. Similarly, progressive fatigue may suggest worsening anemia or transition to MF in the cases of ET or PV.

To what extent symptomatic changes constitute “improvement” or “worsening” of disease burden remains a topic of ongoing study. Numerous MPN-SAF scoring tools have been developed with the purpose of defining “cutoff values” for symptomatic improvement. The original COMFORT-1 study design utilized a 50% improvement from baseline TSS as response criteria [30]. Another recent investigation divided MPN SAF TSS scores into quartiles by the percentage of scores between 0 and 24% (Quartile 1; TSS 0–7), 25–49% (Quartile 2; TSS 8–17), 50–74% (Quartile 3; TSS 18–31), and 75–100% (Quartile 4; TSS > 32) [31]. Another study of ET and PV patients identified that MPN-10 total scores greater than or equal to 20, a worst-individual item score of greater than 5, or a combination of both may signal a patient that may benefit from symptom-based treatment independent of prognostic score [32]. Regardless of the method chosen, it remains important to include the patient’s reflections on their disease status and treatment goals to ensure symptomatic needs are met.

Management of Individual MPN Symptoms

The refinement of the symptom assessment process has allowed for greater individualization of the treatment plan. Research into how individual symptoms impact each other, disease complications, and prognosis have demonstrated that symptoms are more likely manifestations of concomitant interdependent pathological processes than expressions of singular drivers. A number of the more burdensome symptoms have been specifically analyzed and are discussed in detail below.

Fatigue

Fatigue is persistently rated as the most severe and frequent MPN-related complaint. From the patient’s perspective, fatigue is often vocalized as being a subjective sense of “tiredness” that exists independent of activities performed or time of day. The disorder is recognized to be multifactorial, correlating with baseline clinical features, disease stage, complications, and therapies. A recent online survey of 879 MPN patients identified that chronic fatigue (35.3%), sporadic fatigue (29.9%), or both (28.8%) are regularly expressed (> 80% of MPN patients) with the average symptom burden being rated “severe” (TSS 5.9/10) [33••]. Most prominent in MF, the symptom correlates closely with functional capacity. Indeed, the average MPN patient is only capable of performing an average of 25.1 metabolic equivalents (METS)—a value significantly less than healthy controls and akin to values observed in Parkinson’s disease patients [23].

In a similar investigation of 813 MPN patients, it was determined that work hours, number of sick days taken, activities of daily living, and the need for medical disability or early retirement were all negatively impacted by the presence of an MPN [34]. In another study of 1788 MPN patients, higher body mass index (BMI) and concurrent use of alcohol and tobacco use were significantly associated with more fatigue [33••]. Patients with other medical comorbidities such as diabetes mellitus, restless leg syndrome, fibromyalgia, chronic fatigue syndrome, and renal disease also appear to be at higher risk of experiencing this complaint. Not unexpectedly, use of antidepressants, antihistamines, and anti-anxiety medications have all shown close correlations. A recent study of ET and PV patients resistant or intolerant to hydroxyurea utilized the validated EQ-5D to evaluate five dimensions of mobility, self-care, usual activities, pain/discomfort, and anxiety/depression and reported health state [35]. Severe to extreme difficulties in self-care, mobility, and performance of usual activities were noted in both populations.

In clinical practice, assessment of fatigue requires adequate appraisal of known contributors such as anemia, nutritional deficiencies, uncontrolled pain, deconditioning, and neurocognitive disorders such as depression. Both pharmacological and non-pharmacological treatments have proven efficacious in reducing MPN-related fatigue. Patients that regularly participated in exercise strategies reported significant improvement in the symptom [33••]. Likewise, pharmacological interventions such as caffeine and stimulant use have been applied with some degree of efficacy. In a more recent study performed by our group, 38 MPN patients were asked to participate in a 12-week online streamed yoga intervention [36]. Dramatic improvements were noted in TSS, depression, and sleep patterns in normal weight MPN patients suggesting this to be a viable therapy for consideration. From a pharmacological standpoint, JAK2 inhibitors (ruxolitinib) demonstrate the most robust impact on fatigue. The COMFORT-1 study comparing ruxolitinib to placebo in MF patients showed statistically significant improvements in fatigue/inactivity with ruxolitinib with a 60% reduction from baseline in treated patients vs. 32.1% worsening from baseline in placebo populations [30].

Depression

Depression and mood disturbances are common MPN symptoms and have been shown to correlate with all individual MPN-10 symptom items. In a recent 70-item internet survey of 1788 MPN patients, mental health assessments were obtained including the PHQ-2, MHI-5, POMS-B, and MPN-SAF. Twenty-three percent of respondents scored values that indicated high probability of depression. Patients endorsed having been seen for or diagnosed with depression (32.0%), anxiety (20.5%), stress (26.2%), and grief (15.0%) with 22% receiving treatment for their mood disorder within the previous 6 months [37]. Depressive symptoms fail to correlate with prognostic risk suggesting low-risk patients may be at high risk for depression development. Higher MPN-SAF scores (individual and TSS) were significantly associated with higher likelihood of depressive symptoms. It is important to recognize that MPN patients may relay depression “equivalents” such as fatigue, loss of appetite, lack of interest in sexual activity, or simply “sad mood.” Assessment of MPN-related depression requires further exploration of any of the symptoms above if moderate to severe MPN-SAF scores are observed. Co-administration of the PHQ-9 may assist in ruling in or out the disorder. No standard treatments for MPN-related depression have been developed and pharmacological intervention and counseling or group therapies have historically been employed. Further referral to licensed professionals is recommended as undermanaged depression could theoretically impact patient choice of therapies and potentially prognosis.

Abdominal Symptoms

Abdominal-related complaints are common among MPN populations and stem from a variety of causes including mechanical sources (splenomegaly/hepatomegaly), ischemia (splenic infarcts), organ hypertension (portal vein, hepatic vein, splenic vein thrombosis), and extramedullary hematopoiesis [38]. Abdominal pain (63%), abdominal discomfort (72%), early satiety (76%), and weight loss (48%) represent the most common disturbances and independently compromise patient QOL [12]. A recent investigation of MF patients with non-palpable spleens, moderate spleens (1–10 cm), and large spleens (> 10 cm) identified that patients with large spleen were more likely to have primary myelofibrosis and that the presence of splenomegaly correlated with higher symptom scores and increased frequency of early satiety and abdominal discomfort [39]. Interestingly, splenomegaly did not appear to correlate with other expected symptoms such as QOL or weight loss.

Interpretation of MPN-SAF results for this symptom is best performed in conjunction with a thorough clinical examination. Clinical features such as jaundice, organomegaly, and masses from EMH may be detected and suggest the need for further dedicated imaging, either with abdominal computed tomography (CT) or ultrasound. Laboratory testing is also frequently useful, and clinicians should consider testing basic complete blood counts, hepatic panels, and serum lactate levels in the setting of this complaint. A variety of treatment modalities are available for splenomegaly-related symptoms, including hydroxyurea (HU), JAK2-inhibitors, and immunomodulators. Non-pharmacological therapies such as splenic irradiation or splenectomy may be considered should the benefits outweigh the risks of intervention.

Sexuality and Gender-Specific Complaints

Gender and sexuality-related symptoms are common but under-evaluated complaints within MPN populations. A recent study we performed on MPN-related sexuality identified that up to 64% of MPN patients struggle with this symptom (43% severe symptoms) and that the symptom correlates closely with all domains of patient functionality and quality of life [15]. The symptom furthermore correlates with other symptoms including depression and nocturnal/microvascular-related symptoms. As such, it remains important that the issue is addressed as potential treatments are available.

Patients over age 65 or those struggling with cytopenias are most likely to experience this symptom, and some MPN therapies, such as steroids, may exacerbate the symptom. In our MPN Genders investigation, 2006 MPN patients were prospectively studied to determine how gender related to patient characteristics, disease complications, and symptom expression [17]. The study found that female patients were more likely to have ET and male patients were more likely to have PV. In general, females had more frequent and severe symptoms for almost all individual symptoms reviewed, particularly as they related to intra-abdominal-related symptoms and microvascular symptoms. However, QOL remained similar between genders. The results of this study suggested that either males undervocalize their symptom burden, or females have simply learned to compensate for more severe symptoms in order to maintain their QOL. Treatments are directed towards the suspected source. Interpersonal relationship and intimacy-related issues should be referred to licensed counselors. Restrictive abdominal symptoms should be managed as per above, and referral to gynecology may be beneficial in younger MPN populations as hormonal dysregulation is common. Importantly, a thorough review of the patient’s medications is essential given recognized side effects of drugs such as steroids and danazol.

Conclusion

MPN symptoms directly influence patient QOL, selection of therapies and survival. The complex interrelationship between disease stage, clinical features, and symptoms has historically complicated efforts to understand how symptoms truly impact the physical, emotional, and social aspects of the disease. Validated MPN-specific PRO tools including the MF-SAF, MPN-SAF, and MPN-10 allow for rapid and precise assessment of patient symptoms. Their ongoing integration into NCCN guidelines for disease burden assessment and management serves a pivotal role in detecting undermanaged symptoms, monitoring for disease advancement, and evaluating potential side effects from therapies. Continued research will undoubtedly offer additional insight into how each symptom dimension contributes to the overall burden and ultimately impacts patient QOL.