Background
Tobacco use increases the risk of contracting a broad spectrum of diseases and is one of the leading causes of mortality [
1‐
3]. Cigarette smoke consists of a highly complex aerosol of more than 4000 compounds and induces chronic inflammation with elevated levels of pro-inflammatory cytokines, endothelial dysfunction, activation of leukocytes and platelets, and not least, sustained systemic oxidative stress [
4‐
9]. Interestingly, the Philadelphia-negative chronic myeloproliferative neoplasms (MPN) are characterized by raised levels of several of these cells and inflammatory biomarkers [
10,
11]. In recent years, evidence of tobacco use as a contributing factor in the development of MPN has increased [
7,
12‐
15].
MPN are acquired clonal hematologic cancers originating from the hematopoietic stem cell encompassing the three subtypes: essential thrombocythemia (ET), polycythemia vera (PV), and myelofibrosis (MF) [
16]. One hypothesis linking tobacco use with MPN includes the possibility that tobacco triggers a low-grade inflammatory state precipitating the development of myeloid mutations and thereby giving rise to MPN with the most critical phenotypic driver mutations being the JAK2V617-, CALR-, and MPL-mutations [
11‐
13,
16]. Patients with MPN are at risk of a wide array of complications, including thrombosis, hemorrhage, bone marrow fibrosis, and progression to acute leukemia. Furthermore, MPN patients often suffer from numerous symptoms such as fatigue, pruritus, bone pain, night sweats, abdominal pain, and reduced quality of life due to splenomegaly, extramedullary hematopoiesis, treatment side effects and also inflammation [
17‐
19]. Recent research indicates an association between individual MPN symptoms and specific markers of inflammation [
18]. In line with these findings on inflammation, research in other types of cancer shows that tobacco users have a higher symptom burden than non-tobacco users, a finding that has only been sparsely investigated in patients with MPN [
20,
21]. Hence, the aim of this study was to analyze the association between tobacco use and MPN-related symptom burden and, furthermore, explore patient perspectives on tobacco use in an MPN context.
Discussion
With the development of questionnaires specific to MPN symptoms, it has become evident that MPN patients often suffer from a severe symptom burden leading to both affected functionality and reduced quality of life [
17,
23,
24]. Targeted treatment with the JAK1/2-inhibitor, ruxolitinib, reduces splenomegaly and constitutional symptoms significantly for some MF and PV patients, but the majority of MPN patients do not benefit from ruxolitinib treatment and discontinuation rates are high due to treatment-related anemia and thrombocytopenia [
25,
26]. Hence, the need for interventions alleviating symptom burden is still urgent. However, the multifactorial nature of the disease contributing to the different symptoms complicates matters and stresses the need for a multidirectional approach. A newly published paper on body mass index indicates that underweight and obesity are associated with an increased total symptom score among MPN patients [
27]. Accordingly, we hypothesize that smoking represents another, perhaps even more important, contributor to the accumulate symptom burden of MPN. Supporting our hypothesis are two previous MPN studies that found current smokers were more fatigued compared with never smokers [
17,
28]. Likewise, recent research demonstrates increased odds of depression and anxiety among MPN patients who currently smoke [
29]. However, to our knowledge, the present study is the first to investigate whether total symptom burden (TSS), individual MPN-10 symptoms, and QoL are associated with tobacco use.
In this cross-sectional study measuring symptoms and quality of life with validated questionnaires, we found various interesting results. Overall, our findings indicate that current/former smokers have a higher symptom burden than never smokers, which is in line with research in other malignancies. Studies of patients with prostate cancer have found that smoking has a negative impact on both general and disease-specific symptom burden and QoL. Further support comes from research in both head and neck cancers and lung cancers, showing a negative influence from smoking on symptoms, side effects of cancer treatment, and quality of life [
20,
21,
30,
31].
We speculate that the higher symptomatology found among current/former smokers diagnosed with MPN is a result of an accumulated level of chronic inflammation. In healthy, non-smoking individuals, the balance between pro-inflammatory and anti-inflammatory responses is carefully regulated. Conversely, as described comprehensively in two previous published reviews [
9,
18], dysregulation of this balance is a hallmark feature of MPN with the Janus kinase cascade playing a key role in the signaling of inflammatory cytokines. The JAK2V617F mutation induces constitutive activation of the Janus kinase generating a measurable elevation of pro-inflammatory cytokines in MPN patients [
9,
10,
18,
27,
32,
33]. Likewise, smoking triggers systemic inflammation by releasing and activating acute phase proteins and pro-inflammatory cytokines; thus an additive effect of MPN and smoking seems plausible as many of the elevated inflammatory markers (CRP, IL-1β, IL-6, IL-8, TNFα, GM-CSF and MCP-1) are overlapping [
4,
5,
8‐
10,
18,
27,
34].
In line with evidence of a range of MPN symptoms being associated with specific markers of inflammation, our results show that, on average, current/former smokers reported more severe fatigue than never smokers. This finding is consistent with the results of earlier studies on fatigue in MPN patients [
17,
28]. Cytokines, particularly TNFα, have been associated with fatigue in other types of cancer and are, as aforementioned, elevated in MPN [
18,
35]. Another known contributor to fatigue is anemia. The association between anemia and smoking is complex with opposing mechanisms. Tobacco use is a risk factor for several diseases (e.g., cancers and gastric ulcers) that are associated with anemia, but the low hemoglobin may be counterbalanced by an increase in the formation of red blood cell due to chronic carbon monoxide exposure. However, even though some stabilization can be achieved, tobacco use, in general, has a negative influence on hematopoiesis [
36] and accordingly, current/former smokers more often reported having anemia than their never-smoking counterparts. An established reason for anemia in MPN patients is bone marrow fibrosis. Nevertheless, inflammation, particularly from IL1, IL6, and TNF, might be another contributing cause, promoting deregulation of erythropoietin leading to reduced production of red blood cells [
37]. Repeatedly, quality of life studies have highlighted fatigue as the most severe of the MPN symptoms and insufficient treatments are available [
17,
23,
28]. Hence, the strengthened evidence of smoking as an aggravating factor of fatigue suggests that smoking cessation counseling may be a beneficial therapeutic intervention.
Besides fatigue, this study identified several symptoms associated negatively with tobacco use; on average, current/former smokers reported significantly higher inactivity scores relative to never smokers. Many reasons may account for this including fatigue, and a significantly higher prevalence of anemia and chronic obstructive pulmonary disease (COPD) among current/former smokers. Moreover, previous research has demonstrated associations between sedentary lifestyles and pro-inflammatory states [
18,
38]. Furthermore, current/former smokers were more likely to struggle with early satiety and abdominal discomfort than never smokers. Common causes of these symptoms include splenomegaly, which was also more prominent in current/former smokers. Abdominal thrombosis may also account for these symptoms and interestingly, has been previously associated with elevated levels of CRP, an inflammatory marker elevated in both MPN and smokers [
8,
39]. Notably, this study did not find any difference in history of thrombosis between current/former smokers and never smokers but did not examine the possibility of microthrombosis as a contributor. A higher proportion of current/former smokers than never smokers in our study experienced concentration problems which is in line with substantial evidence of smoking’s harmful effects to the brain, The role of oxidative stress and pro-inflammatory cytokines as a link between smoking and cognitive impairment is yet to be fully elucidated, however, it is thought-provoking to note that research in hematologic diseases (and the general population) have shown that increases in circulating inflammatory cytokines, including IL-1, TNFα, IL-6, and CRP, are associated with cognitive deficiencies [
18,
40].
When stratifying by MPN subtype we found that for PV patients, current/former smokers experienced worse bone pain than never smokers. This is consistent with previous studies showing increased pain among current smokers with cancer [
41]. In contrast, neither symptom severity nor symptom prevalence differed significantly between MF current/former smokers and MF never smokers. We speculate that a contributing reason could be that in the group of current/former smokers with MF, six patients reported a history of bone marrow transplantation compared with none of the patients in the group of never smokers with MF. Assuming that transplanted patients, in general, have a lower symptom score, this would contribute to an overall lower average in the current/former smokers group masking a potential difference due to smoking status.
Of particular interest was the finding that, on average, never smokers reported a significantly better overall quality life compared to current/former smokers. Considering QoL by MPN subtype, we found no significant differences according to smoking status for PV and MF patients while the better QoL among never smokers persisted for ET patients. Whether the association of QoL and tobacco use truly differs between the MPN subtypes or this is an incidental finding due to the relatively small number of patients in some of the strata remains to be explored in future studies. Nevertheless, the better overall quality of life among never smokers with MPN is of great importance and substantiates the majority of existing literature; several studies investigating quality of life in the general population report lower QoL among current smokers and furthermore, looking at cancer patients, findings confirm a worse health-related quality of life among smokers [
20,
31,
42,
43]. Compared with other studies, the mean difference QoL between never smokers and current/former smokers was quite small. The small mean difference is probably, at least partly, due to the merge of current smokers and former smokers into one group. A Finnish study [
42] demonstrates that former smokers report higher QoL than current smokers, approaching that of never smokers. Another contributing reason could be the use of different questionnaires across studies. Many of the studies reporting a larger difference in quality of life by smoking status use a QoL score calculated from different symptom scores, which is in contrast to the MPN-SAF single item with a self-reported rating of overall quality of life, which is influenced by additional non-health-related factors. From a biochemical standpoint, nicotine ingestion results in activation of nicotinic acetylcholine receptors, which can subsequently modulate the functional pathways related to anxiety, depression and stress response [
44]. Many patients experience sedating and calming effects from smoking which could possibly serve as a contributor to improved quality of life.
Evidence states that smoking is associated with other substance abuse and alcohol [
30]. Accordingly, we found current/former smokers more often reported higher alcohol consumption, at least two alcohol problems, and an increased sense of guilt related to drinking behavior. As it is widely recognized that alcohol is hepatotoxic, neurotoxic and carcinogenic, and furthermore associated with increased symptom burden and poor QoL, our findings underscore the need to comprehensively screen current and former MPN tobacco users for other harmful substance use behaviors [
45,
46].
Despite the irrefutable evidence for the multiple harmful effects from tobacco use, the prevalence of current smokers in the US in 2000 was estimated to be as high as 22.2%, while the corresponding estimates for former smokers and never smokers were 24.4 and 53.4%, respectively [
1]. Bearing in mind that our findings are of more recent date (2018), it nevertheless represents a striking difference in usage patterns. A plausible explanation of the lower prevalence of current smokers in our study might be that being diagnosed with cancer is a crucial event in a person’s life, giving rise to new motivation to attempt to quit smoking, a theory with conflicting evidence [
30,
47]. Comparing our study population with other MPN cohorts, the prevalence of current smoker is similar to the findings in The Fatigue Study (current smokers, 6%) but markedly lower than in the MPNhealthSurvey (current smokers, 18%), likely reflective of regional variations in usage [
28,
48].
The National Comprehensive Cancer Network (NCCN) Clinical Practice Guidelines in Oncology recommends
“Smoking cessation should be offered as an integral part of oncology treatment” and
“Providers should inform patients of the potential benefits of smoking cessation, including improved survival, treatment outcomes, and health-related quality of life” [
49] Thus, an important finding in this study was the rather high percentage of current smokers reporting having discussed tobacco use with a physician, however as 15% of the current smokers do not recall ever having received tobacco cessation counseling from a physician there is still room for improvement. This is especially true given that almost one-fifth of the MPN patients were unaware of the increased thrombotic risk due to smoking. Patients in our study reported nicotine replacement as the most frequent medication attempted to aid cessation. However, e-cigarettes were also reported as a smoking cessation aid and approximately one-third of the current/former smokers stated that they consider vaping safer than smoking. Interestingly, all of the patients (
n = 4) who had switched from smoking to vaping reported improvements in their health. Although emerging research suggests some medical disorders are less prevalent with vaping than smoking, this is based on limited data [
2,
50]. Hence, much remains unknown about thrombotic predisposition and inflammatory state induced through vaping. Even more elusive are the consequences of vaping for MPN patients; no studies have investigated either acute or long-term complications from vaping in an MPN population. However, a recent study in mice indicates that two months of e-cigarette vapor exposure induces suppression of bone marrow hematopoietic stem and progenitor cells [
51]. Furthermore, research in human keratinocytes and epithelial lung cells has demonstrated that excessive vaping induces inflammatory responses activating the JAK-STAT pathway and increasing production of pro-inflammatory cytokines and chemokines (e.g., IL-1, IL-8, and TNF-α) [
50]. Bearing in mind the significant role of JAK2 and inflammation in the etiology of the MPN disease, future studies of possible adverse effects of vaping in MPN patients are crucial.
Limitations in our study include the self-reported design with a lack of validation of MPN diagnosis, MPN treatment and smoking status. In addition, MPN mutation status was not included in the survey. Moreover, the patients’ reports of historical data may be affected by recall bias. Another limitation is that
current smokers and
former smokers were combined due to a low number of current smokers; this might mask a much larger difference between current smokers and never smokers. Moreover, the cross-sectional design restrains us from drawing conclusions on the causality between smoking status and symptom burden. Also, we cannot rule out that the observed negative association of tobacco use and quality of life, at least partially, could be confounded by other differences between current/former smokers and never smokers, e.g., alcohol abuse and comorbidity. COPD and anemia were more prevalent among current/former smokers than among never smokers, which (along with other comorbidities associated with smoking) might contribute to the observed difference in symptom burden and quality of life. Likewise, symptom scores were not adjusted for the potential impact of current MPN treatment. However, no significant differences in frequency of ruxolitinib or hydroxyurea treatment between current/former smokers and never smokers were present. Finally, when interpreting questionnaire data, it is always relevant to consider if the results are above the minimal important difference, i.e., are of clinical relevance [
52]. The mean differences, found in the current study, are quite small, which was expected due to the multiple factors contributing to the severity of the MPN symptoms. However, with no treatment eliminating MPN symptoms available, it is vital to take action against all modifiable causes in order to achieve an overall substantial symptom burden relief. Therefore, despite limitations, we consider many of the findings in this study to be of clinical significance, especially as they relate to potential worsening of the pro-inflammatory MPN environment and opportunities to educate MPN patients on smoking-related risks.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.