Clinical characteristics and symptom burden of Thai myeloproliferative neoplasm patients

ABSTRACT Objective Patients living with myeloproliferative neoplasms (MPNs) suffer from symptom burden that affect quality of life. Due to the differences in cultures, climates, and genetic background, we aimed to investigate the symptom burden of Thai MPN patients Methods A comprehensive survey using the MPN-10 questionnaire was carried out between September 1, 2014, and September 30, 2017. The scores obtained were then correlated with clinical outcomes.. Results A total of 145 patients were enrolled. Nearly 90% of patients reported being symptomatic. The mean MPN-10 score was 13.6 (SD = 11). The mean MPN-10 score was highest in PMF, whereas the mean score and intensity of individual items were surprisingly low in ET and PV. Notably, the mean MPN-10 score was significantly higher in patients with documented splenomegaly compared to those with a normal-sized spleen. However, there were no correlations between MPN-10 scores and the mutation status, disease complications such as thrombosis and hemorrhage, progression to myelofibrosis or leukemia, and mortality. Patients who needed regular transfusions reported a higher MPN-10 score compared to those who did not. Conclusion The MPN-10 score did not predict survival outcomes among Thai MPN patients. Higher MPN-10 was associated with more transfusion. Thai MPN patients reported lower MPN-10 compared to western population especially PV and ET.


Introduction
Myeloproliferative neoplasms (MPNs) are clonal myeloid malignancies characterized by cellular proliferation of one or more hematopoietic lineages without dysplastic features.The classic MPNs or BCR-ABL1-negative MPN include polycythemia vera (PV), essential thrombocythemia (ET), and primary myelofibrosis (PMF) [1].Since 2005, a gain-of-function mutation of JAK2V617F in committed myeloid progenitors was discovered as a major driver mutation in the three classic MPNs [2][3][4][5].Several research papers have demonstrated that JAK2V617F is the most prevalent driver mutation in MPNs which is detectable in more than 90% of patients with polycythemia vera (PV) and approximately 50-60% of those with essential thrombocythemia (ET) and primary myelofibrosis (PMF).The lack of JAK2 mutations in approximately half of patients has led to the discovery of MPL [6] and CALR mutations [7].Other molecular pathogenesis of MPNs including MPL and CALR mutations have also been comprehensively studied [8].These three driver mutations cover almost all the classic MPNs, whereas 10-15% of ET and PMF which do not harbor these driver mutations are termed triple negative MPNs.
Major causes of morbidity and mortality in MPN are thrombohemorrhagic complications and progression to acute myeloid leukemia [9][10][11].Nonetheless, MPNs are now considered to have a favorable prognosis.Although people with MPNs have improved lifeexpectancy in this era, a substantial proportion of patients suffer from significant constitutional symptoms.MPN-related symptoms have been well-documented by several international surveys [12][13][14].MPN symptoms include fatigue, itchiness, night sweats, bone pain, fever, and weight loss.The prevalence and severity of individual symptoms differ among MPN subtypes.Hematopoietic stem cell transplantation is the only curative treatment for MPN.Unfortunately, this option is only reserved for young patients and those with advanced-stage disease.The goal of MPN treatment is to alleviate a patient's symptom burden especially in those with myelofibrosis.Higher symptom burden is associated with decreased daily activity and poor quality of life.Several studies of JAK inhibitors marked the symptom burden as a surrogate outcome of treatment response in the patient's point of view in addition to the clinical response [15][16][17][18][19][20].Different symptom assessment tools have been used to find links between disease outcomes and MPN symptoms.Several questionnaires have been developed to standardize the evaluation of symptom response in clinical trials, ranging from verbal interviews to patient-reported outcomes.The Myeloproliferative Neoplasm Symptom Assessment Form total symptom score (MPN-SAF TSS) or MPN-10 score [14] consists of an original 17-item Myeloproliferative Neoplasm Symptom Assessment Form (MPN-SAF) and the 10-item Brief Fatigue Inventory (BFI).This instrument has been validated in more than 20 languages across nearly 30 countries [21].MPN-10 score has been translated into Thai language to assess patient quality of life.However, the score has not been evaluated in Thai population.Therefore, we aimed to investigate the clinical characteristics and disease-related complications in Thai MPN patients and related patient symptom burden as assessed by MPN-10 score.

Study participants
A cross-sectional survey was conducted at King Chulalongkorn Memorial Hospital (KCMH) from 1 September 2014, to 30 September 2017.Data were collected from PV, ET and MF patients including post PV and post ET-MF (diagnosis according to WHO 2008 criteria [22]) who were 18 years or older.Patients who did not give informed consent were excluded.Baseline demographic data (age, sex, and comorbid diseases), clinical characteristics (presenting sign and symptoms including splenomegaly), complete blood count (CBC) at diagnosis, and driver genetic mutations were retrospectively reviewed and retrieved from electronic medical records at diagnosis or at the time of referral to KCMH.Bone marrow aspiration and biopsy were reviewed by an experienced hematopathologist.The risk categories of patients were classified by age and prior history of thrombosis in PV and ET [23].The Dynamic International Prognostic Scoring System (DIPSS) [24] or DIPSS-plus [25] was used to classify patients with PMF.Transfusion dependency was defined as patients who needed more than 1 unit of red blood cells per month.Disease-related thrombohemorrhagic complications included thrombosis or bleeding events that occurred at diagnosis or at any time during follow-up at KCMH.All patients were treated under institutional review-approved retrospective protocols and in accordance with the declaration of Helsinki.
MPN related symptom assessment MPN-SAF TSS and MPN-10 were translated into Thai language by a Thai MPN working group and assessed as the sum of 10 items: worst fatigue, early satiety, abdominal discomfort, inactivity, concentration problems, night sweats, itching, bone pain, fever, and weight loss.The intensity of scores had a range from 0 (absent) to 10 (worst imaginable) for each symptom.The total score had a range from 0 to 100.MPN-10 questionnaire was conducted as an observational cross-sectional study.The participating patients were asked to complete the survey and to provide answers as completely and accurately as possible.If the patients were not able to read but were able to answer the questionnaire, the patients were interviewed using a translated MPN-10 questionnaire by the investigators.

Statistical analyses
Continuous data are described by means (± standard deviations, SD) or medians (± interquartile range, IQR) as appropriate.Qualitative data are presented as frequency and percentage.The Kruskal-Wallis test was used to compare differences between the three groups.Chi-square test or Fisher's exact test examined relationships between quantitative variables.Cox proportional hazard regression model was used for the multivariable analysis.Overall survival was calculated from date of diagnosis to death or last visit.Event-free survival was calculated from date of diagnosis to thrombotic or hemorrhagic events.Survival analysis according to MPN-10 score was calculated from the date of MPN-10 survey to death or last visit.Survival analysis was done using the Kaplan-Meier survival curve and the comparison was performed using log-rank test.All tests were two-tailed.Statistical significance was met if the p value less than 0.05.All data were analyzed using SPSS version 22 (SPSS Inc., Chicago, IL).

Survival outcomes
With a median follow-up time of 5.9 years (0-20 years), the median overall survival (OS) was not reached in PV, ET and PMF.During the study period, 8% (12/145) of the cohort died.The survival analysis revealed that PMF had inferior OS compared to ET and PV (P < 0.00) (Figure 2) with 100% survival among low-risk PV, ET, and PMF patients.Five-year OS among high-risk PV, ET, and PMF was 88.7%, 86.5%, and 56%, respectively.Univariate analysis found that age > 70 years, diagnosis of PMF, bleeding complications, disease progression, and transfusion dependence were risks for inferior survival.Multivariate analysis showed only age >70 years and transfusion-dependent status predicted poorer outcomes with odds ratios of 2.66 (1.05-6.72;P = 0.04) and 6.23 (2.62-14.82;P < 0.00), respectively (Table 3).

MPN related symptoms assessed by MPN-10 score and clinical outcomes
The median time from diagnosis to the score evaluation was 2.3 (0-18.5)years.The MPN-10 scores were calculable in 79 patients (54%).The majority (68%) of patients were evaluated within the first 3 years of diagnosis.Additionally, the median time from the score evaluation to disease progression was 12 months (1-37 months).The proportion of cytoreductive therapy received at the time of survey is described in Table 1.The scores were not evaluated in 45% (66/145) due to disabilities (11), death (12), or inability to contact (43).Nearly 90% of evaluable patients were symptomatic (score > 0).The mean MPN-10 score was similar across ages.Female patients with PV reported a nonstatistically significant difference in MPN-10-TSS compared to male patients (15 vs. 10, P = 0.24).There was also no gender effect on MPN-10-TSS in ET and PMF.The mean MPN-10 score of all MPN patients was 13.6 (SD,11).The mean MPN-10 score of PMF was 25 (SD,12), which was significantly higher than those of PV (11.7;SD,11), and ET (10.7;SD,9.7) (P < 0.00).The mean intensity of the individual items of MPN-10 was 1.1 (SD, 1.9) in PV, 1.0 (DS,1.7) in ET, and 1.3 (SD, 2.5) in PMF.The most common symptom was fatigue, which had the highest prevalence in PV, ET, and PMF.The other common symptoms were inactivity, itching, early satiety, concentration problems, and abdominal discomfort.In PV, ET, and PMF patients, the two most common symptoms were fatigue and inactivity, while PMF had the highest symptom intensity (mean 5.5; SD 2.2 and mean 4.3; SD 2.8, respectively) The third most common symptom was different among MPNs with itching in PV, concentration problems in ET, and early satiety in PMF.Bone pain was the least reported item among MPN patients.Among ET who had variable mutations, total symptom score (TSS) was not different according to the mutation subtypes.The mean MPN-10 scores of triple negative-ET vs. JAK2V617F-ET vs. CALR-mutated ET were 8.3 (0-19) vs. 10.2 (0-25) (P = 0.99) vs. 12.5 (0-36) (P = 0.42) The prevalence and symptom severity are presented in Figure 3 and Table 4.
Of the patients with an evaluable MPN-10 score, 97% (77 out of 79) had documented spleen size evaluated by ultrasonography or physical examination.It was found that the mean MPN-10 score was significantly higher in patients with documented splenomegaly (21, with a range of 7-30) compared to those with a normal-sized spleen (7.5, with a range of 2-15) (P < 0.001).However, the MPN-10 TSS did not correlate with driver mutations and disease complications including thrombosis, hemorrhage, disease progression to secondary myelofibrosis, AML transformation, and death (Supplementary table S1-S4, Supplementary figure 1).Early satiety was associated with inferior survival in the univariate analysis with an odds ratio of 9.35 (95%CI, 1.16-75.16;P = 0.04), while bone pain did not meet statistical significance (OR 3.63, 95%CI, 0.94-14.07;P = 0.06).None of the  S5).Transfusion-dependent patients had significantly worse fatigue, early satiety, abdominal discomfort, and inactivity.Patients with bleeding complications experienced a higher intensity of inactivity and greater weight loss compared to those without bleeding events.

Discussion
MPNs are chronic diseases that affect the quality of life of patients.A large study from the Mayo Clinic demonstrated that the median survival of ET and PV are 15 and 18 years, respectively.Among low-risk ET patients, the median overall survival reaches 30 years [26].Despite the fact that patients with MPN can have long-term survival, the mortality of MPN patients exceeds the age-matched general population [27][28][29].Disease-related complications such as inflammation-related symptom burden, cardiovascular disease, and MF/AML progression are serious concerns.Additionally, MPN is pro-inflammatory and can cause debilitating symptoms that result in considerable morbidity.The major treatment goals include prevention of thrombotic complications, decreasing symptom burden, and minimizing disease progression.Understanding disease characteristics, symptom burden, and how the disease impacts patients living with MPN is crucial.
This study investigated clinical characteristics, complications, and symptom burden in a Thai MPN population.The baseline characteristics of patients were consistent with previous reports with the majority diagnosed with ET followed by PV and PMF and a median age of 61 years.Gender-specific distributions were also similar to other MPN studies [21].PV and PMF were male predominant, while the majority of ET patients were female.The frequency of driver mutations (JAK2, CALR, and MPL) was distributed as expected.More than half of the patients harbored at least one cardiovascular risk factor and approximately 10% had a history of thrombosis prior to MPN diagnosis.Over 65% of patients were high risk and receiving treatment, primarily cytoreductive therapies.Due to limitations in healthcare coverage, only 5% of patients received JAK inhibitor treatment.During the study period, PV had the highest prevalence of thrombosis while bleeding complications were rare among all MPN subtypes.A high disease risk score was a significant risk factor for thrombosis whereas CKD was the only risk factor for bleeding complications in our cohort, contradicting previous studies.With a median follow-up time of 5.9 years, 11% of PV and ET developed secondary myelofibrosis.More than 50% of patients had survived by the end of the study.Mortality was found to be strongly associated with age and transfusion status, consistent with previous findings.
Recognition of symptom burden in MPN patients allows healthcare professionals to identify patients who may benefit from symptom-modifying therapy.This study explores the symptom burden by MPN-10 score, then correlates scores with patient outcomes.PMF had a higher symptom burden than PV and ET as reported in several studies.Not only was the mean MPN-10 score of Thai patients higher in PMF than PV and ET, but the intensity of individual items including fatigue, early satiety, abdominal discomfort, and inactivity was comparable to western populations.Additionally, it is important to consider that the limited number of patients receiving JAK inhibitor treatment could also have an impact on the results of the high MPN-10 score among PMF patients.Of note, the frequency of concentration problems in this study was low.One of the explanations was that MPN-10 scores have been developed from common symptoms in western populations which may be different than Asian populations.Thai people, like many Asian people, have a tendency to not convey their negative emotions openly to other people.People from western cultures are often more comfortable expressing their own feelings and needs.These cultural differences may have affected the results.
There are also differences in occupational structure.For example, concentration problems would be more problematic for executive workers than for laborers.Another reason might be differences in genetic/epigenetic backgrounds.Various mechanisms are responsible for systemic symptoms in MPN patients [30].Several factors are shown to contribute to symptom burden, interplaying between genetic diversity, inflammatory cytokines, disease phenotypes such as splenomegaly and disease complications.Despite the similar prevalence of driver mutations, the frequency and intensity of individual MPN-10 items of PV and ET patients were lower compared to other international surveys in Caucasians [12,31] or Asian populations [32,33].In the MERGE study [33], which included Asian and South Asian populations, 10% were from Southeast Asia.The mean MPN-10-TSS of PMF was comparable between the MERGE study and our study (23.5 [17.47] vs. 24.4[12]).Of note, the mean MPN-10-TSS was lower in PV (16.6 [14.84] vs. 11.7 [11.4])Our study also demonstrates that the MPN-10 score was correlated with splenomegaly but was not associated with other patient characteristics such as age or gender, nor did it predict clinical outcomes including thrombosis, hemorrhage, secondary myelofibrosis, AML transformation, or death.Contrary to the study from Egypt [32], ElNahass YH et al. showed that MPN-10 was associated with driver mutations and predicts survival in MPN patients.A higher proportion of PMF patients were included in the study compared to our study.MPN-10-TSS scores were also twice as high suggesting that the difference in baseline characteristics of patients may have impacted outcomes.Our study also showed that patients with transfusion dependence reported significantly higher MPN-10-TSS compared to those without transfusions.This may represent an advance phase of the disease.Moreover, transfusion dependence predicted worse survival independent of the diagnosis, disease risk score, and mutational status which was consistent with previous reports [34,35].This is the first comprehensive long-term follow-up MPN study in Thailand focusing on clinical characteristics and symptom burden related to disease outcomes.The symptom scores may be different in Thai  patients due to differences in language, culture, climate, and genetic background.In our study, Thai MPN patients, especially those with PMF, experienced a significant symptom burden.Although MPN-10-TSS did not predict disease-related outcomes such as thrombosis or survival in our sample, a large number of patients struggled with high symptom burden.None of these burdensome symptoms are included in disease prognosis scores.In Thailand, which is a limited-resource country, identifying patients who would benefit most from symptom-alleviating therapy is even more important.From our data, JAK inhibitors would be most beneficial for transfusiondependent PMF patients.This study has certain limitations that need to be addressed.First, this is a cross-sectional, singlecenter retrospective cohort study with a limited sample size with only half of the patients reporting MPN-10 scores.Second, this study did not demonstrate how symptom burden had impacted patient quality of life.Lastly, there is no longitudinal follow-up of MPN-10 score to understand the dynamics of MPN-10 that could provide insights into disease outcomes and early recognition of disease-related complications.
In the conclusion, our study suggested that the MPN-10 score was not found to be predictive of disease-related outcomes in Thai MPN populations.Thai MPN patients reported lower MPN-10 scores compared to what has been previously reported in western populations, especially in the case of PV and ET.Patients who were transfusion-dependent had higher MPN-10 scores compared to those who did not require transfusions.Future research should be conducted on a larger scale to capture the undermanaged symptoms more effectively.Furthermore, the study of cytokines and molecular mechanisms can enhance our understanding of the association between the heterogeneous clinical presentation of MPN symptoms and the underlying pathogenesis of the disease.

Figure 1 .
Figure 1.Distribution of driver mutations in myeloproliferative neoplasms based on disease phenotype.

Table 1 .
Characteristics of patient at diagnosis.Compared categorical data using Chi-square or fisher's exact test; Compared continuous data using Wilcoxon rank sum test.

Table 2 .
Characteristics of patients with thrombotic complications.

Table 3 .
Multivariate analysis for risk factors associated with mortality.
HR: Hazard ratio, aHR: adjusted hazard ratio, P-value was evaluated by cox regression and stepwise back ward to select final model, multivariate models were developed by adjusting for covariates with p < 0.1 in univariate models.andET(14.6[14.26]vs.10.7[9.7]).The common symptoms displayed a similar pattern on individual items, but with less intensity among PV and ET patients.

Table 4 .
Assessment of patient symptom severity and symptom prevalence by disease type.