The incidence of myocardial infarction and stroke in head and neck cancer patients

Various treatment modalities are used for head and neck cancer (HNC). This study analyzed the incidence and risks of myocardial infarction (MI) and stroke by cancer site and treatment modality in 22,737 patients newly diagnosed with HNC registered in the Korean National Health Insurance Service database in 2007–2013. An additional 68,211 patients without HNC, stroke, or MI were identified as the control group. The risks for MI (hazard ratio [HR] = 1.38, 95% confidence interval [CI] 1.24–1.53), stroke (HR = 1.48, 95% CI 1.37–1.60), and mortality (HR = 5.30, 95% CI 5.14–5.47) were significantly higher in the HNC group. Analysis by cancer site showed the risk of MI and mortality was highest in hypopharynx cancer, while the risk of stroke was highest in nasopharynx and paranasal sinus cancer. Analysis by treatment modality showed the highest risks for MI (HR = 1.88, 95% CI 1.31–2.69) and mortality (HR = 2.95, 95% CI 2.75–3.17) in HNC patients receiving chemotherapy (CT) alone, while HNC patients receiving CT with surgery had the highest risk for stroke (HR = 1.81, 95% CI 1.14–2.88). Careful attention to MI and stroke risks in HNC patients is suggested, especially those who received both CT and radiotherapy.

www.nature.com/scientificreports/ lence of hypertension and DM was higher in the HNC group than that in the control group (33.14% vs. 30.63% and 14.77% vs. 12.06% respectively; all P < 0.001) s (Table 1).
MI, stroke, and mortality risks in the HNC group compared to those in the control group. After 5-10 years of follow-up, we identified 467 cases of MI, 887 cases of stroke, and 9206 cases of mortality in the HNC group and 1422 cases of MI, 2567 cases of stoke, and 7281 cases of mortality in the control group. Compared to the control group, the overall incidence rates of MI ( 14-5.47) were also higher in the HNC group ( Table 2). The incidence probabilities of MI and stroke and the survival probabilities in the HNC and control groups are shown in Fig. 1. In the HNC group, the incidence of MI, stroke, and mortality appeared to occur immediately within 2 years, and as the follow-up period was prolonged, an increase in imbalance between groups was observed (P < 0.0001).
MI, stroke, and mortality risks in the HNC group by cancer site. We investigated the risks of MI, stroke, and mortality at each HNC site. Compared to the control group, patients with hypopharynx cancer had higher risks of MI (HR = 2.40, 95% CI 1.85-3.11) and mortality (HR = 9.01, 95% CI 8.48-9.58), whereas those with nasopharynx and paranasal sinus cancer had a higher risk of stroke (HR = 2.22, 95% CI 1.91-2.58) ( Table 3). The incidence probabilities of MI and stroke and survival probabilities by cancer site in the HNC www.nature.com/scientificreports/ and control groups are shown in Fig. 2. In the HNC group excluding the salivary gland and others group, the incidence probability of MI was similar or higher than that of the control group (P < 0.0001). In the HNC group excluding the tongue and oral cavity, salivary gland and others group, the incidence probability of stroke was higher than that of the control group (P < 0.0001). All HNC subgroups had a lower survival probability compared to the control group (P < 0.0001).
MI, stroke, and mortality risks in the HNC group by age and sex. .63) were highest in people over 80 years of age. When analyzed by sex, the incidence rates and risks of MI, stroke, and mortality were higher in male patients (Table 4).

Discussion
To our knowledge, this is the first nationwide study to investigate the relative risks of MI, stroke, and mortality incidences among HNC patients who underwent surgery, CT, and RT compared to a general non-cancer control population matched for age and sex. After adjusting for age, sex, and comorbidities such as hypertension (HTN), diabetes mellitus (DM), and dyslipidemia, the risks of MI, stroke, and mortality were significantly higher in the HNC group compared to those in the control group. Analysis by cancer site showed the highest risk of MI and mortality in patients with hypopharynx cancer, while the risk of stroke was highest in patients with nasopharynx and paranasal sinus cancer. Analysis by treatment modality showed high incidences of MI, stroke, and mortality for treatment with CT alone or with surgery or RT. After adjusting for age, sex, and comorbidities, the risks of MI and stroke were highest in patients who received CT alone or with surgery. HTN, DM, and dyslipidemia are well-known risk factors for vascular events such as stroke and MI 19,20 . In addition, studies on the effects of HTN, DM, and dyslipidemia on the development of cancer are ongoing. A previous large, population-based case-control study reported inverse associations between type II DM, HTN, dyslipidemia, and HNC 21 . However, in the present study, the incidence of HTN and DM were higher in HNC patients than in the controls. Even after adjusting for the risk factors of vascular events such as DM, HTN, and dyslipidemia, the risks of stroke and MI remained higher in HNC patients.
Previous studies reported the highest and lowest 5-year HNC survival rates for lip cancer and hypopharyngeal cancer, respectively 22,23 . The 8th edition of the American Joint Committee on Cancer (AJCC) divides lip cancer into mucosal and cutaneous lip. Mucosal lip is included in the oral cavity while cutaneous lip is included in cutaneous carcinoma of the head and neck 24 . Thus, this study included mucosal lip cancer as an oral cavity cancer. When lip cancer was excluded from previous studies, salivary gland cancer had the highest 5-year survival rate in HNC 22,23 . In the present study, larynx cancer showed the lowest mortality risk and hypopharynx cancer showed the highest mortality risk in HNC patients.
Several HNC cohort studies have investigated the risk of stroke following treatment. A population-based study similar to the present study reported that HNC patients are at increased risk of developing stroke, especially the young age group and those who received both RT and CT 25 . In our study, HNC increased the risk of stroke by 48.3% compared to control. However, unlike the previous study, the stroke risk increased with age and the stroke risk was highest in patients in the present study who received both surgery and CT. www.nature.com/scientificreports/ Many studies have shown that RT is associated with increased risks of cerebrovascular disease in HNC patients. The exact mechanism of radiation-induced carotid or cerebral artery injury is not clear but is thought to be due to endothelial dysfunction, injury and occlusion of the vasa vasorum, and accelerated atherosclerosis 26,27 . Several population-based cohort studies reported increased cerebrovascular risk in HNC patients who underwent RT alone compared to that in patients who underwent surgery with or without radiation 12,28 . A recent populationbased cohort study reported that the risk of stroke increased with RT after adjusting for other socioeconomic and clinical risk factors. The authors suggested that attention should be paid to advanced-age patients with low socioeconomic status and known clinical risk factors of stroke 13 . However, in the present study, patients who received RT alone or surgery with or without radiation had lower risks of stroke than those who received no treatment and the stroke risk was highest in patients who underwent both surgery and CT.
The risk of thromboembolic events increases after chemotherapy in patients with various cancers. Ischemic cerebrovascular accidents caused by chemotherapy have been reported in various cancers, including HNC, breast cancer, and lymphoma 17,[29][30][31] . The mechanism of chemotherapy-induced thromboembolism is also not clear, although tumor embolization, vasculitis, nonbacterial thrombotic endocarditis, consumption coagulopathy, or complications related to chemotherapeutic agents have been suggested as causes of thromboembolism 18,32,33 . Cisplatin is the main anti-cancer drug in CT of HNC and commonly causes cerebrovascular events 18 . After CT with cisplatin, cerebrovascular accidents occur due to vasospasm, hyperreninemia or hyperaldosteronemia, platelet hyperaggregation, decreased tissue activators, endothelial dysfunction, and elevated serum cholesterol levels 18,[34][35][36][37] .
There have also been reports of chemotherapy-induced cardiotoxicity (CIC). CIC occurs due to the direct effect of the drug on the cardiovascular system or indirect effects due to thrombogenic status or hemodynamic flow alterations 38 . The risk of CIC development is affected by several factors, including the type of chemotherapy agent, the administered dose, and the drug administration rate 39,40 . However, to our knowledge, there have been no reports of increased MI after treatment in HNC patients. Although the mechanism is not clear, the results of this study showed an increased incidence of MI after treatment in HNC patients, especially in those who received CT alone.
Contrary to previous studies, where the risk of stroke increases due to RT, the group receiving both surgery and CT in the present study showed the highest stroke incidence rate or HR. According to the results of this study, CT appeared to increase the risks of MI, stroke, and mortality more than RT. Patients who received CT alone are likely to have received palliative CT due to advanced stage or underlying diseases. Therefore, the highest mortality rate among patients who received CT alone may have been affected by complications such as MI and stroke.
This study has several limitations. First, alcohol consumption, smoking, and body mass index (BMI) are important risk factors for stroke and MI as well as HNC. However, the KNHIS data did not provide information on alcohol consumption, smoking and BMI. Second, the type and administered dose of CT agent and the intensity of RT used in treatment could not be confirmed. Finally, our study was based on an Asian population; thus, the results may be different in Western populations. Therefore, our results need to be analyzed in other ethnic groups.
In conclusion, the results of our study showed increased MI and stroke risks among HNC patients compared to those in the general non-cancer population. After adjusting for age, sex, and comorbidities, the risks of MI and mortality were highest in hypopharynx cancer, while the risk of stroke was highest in nasopharynx and paranasal sinus cancer. Analysis by treatment modality showed that the risks of MI, stroke, and mortality were highest in the group that received CT alone or with surgery. Therefore, careful attention to MI and stroke risk for HNC patients is suggested, especially in those who receive both CT and RT.  43 . This program aims to detect cardiovascular risk factors, including age, sex, smoking status, alcohol consumption, hypertension, DM, and dyslipidemia, for subsequent educational counseling.

Methods
The KNHIS database has been widely used in numerous epidemiological and health services-related research studies 44 . Its detailed profile and configuration have been previously described 45,46 . The research protocol was approved by the Institutional Review Board of the Pusan National University Hospital and informed consent was waived. www.nature.com/scientificreports/ to the ICD-10-CM codes. We included hypertension, DM, and dyslipidemia. Income level was categorized into quartiles based on individual insurance contribution; the medical aid population (~ 3% of Korean population) were merged with the lowest income quartile group for the analyses. The index dates were randomly selected as the control group. The design of this study was as follows; (1) analyze the incidence and risks of MI, stroke, and mortality in the total HNC and control groups; (2) analyze the incidence and risk of MI, stroke, and mortality by HNC site (tongue and oral cavity, oropharynx, nasopharynx and paranasal sinus, hypopharynx, larynx, salivary gland, and others); and (3) analyze the incidence and risks of MI, stroke, and mortality by age, sex, and treatment modality (no treatment, surgery alone, RT alone, CT alone, surgery and CT, surgery and RT, concurrent chemoradiation therapy (CCRT), surgery and CCRT) in the HNC group.
Statistical analysis. We compared the distributions of demographic status and comorbidity between the HNC and control groups using chi-square tests for categorical variables and t-test for continuous variables. We calculated the incidence rates of MI, stroke, and mortality by dividing the number of incident cases by the total follow-up period. The incidence rates of MI, stroke, and mortality are presented as 1000 person-years. The hazard ratios (HRs) and 95% confidence intervals (95% CIs) describing the risks of stroke, MI and mortality were calculated using Cox regression models adjusted for age, sex, and other comorbidities, including hypertension, DM, dyslipidemia, and income. The incidence and survival probability was calculated by using the Kaplan-Meier curves, and the log-rank test was performed to analyze differences among the groups.
The HNC group was compared to the control group. To assess the risk of MI, stroke, and mortality by HNC site, the study population was divided into six subgroups based on ICD-10-CM code (tongue and oral cavity, oropharynx, nasopharynx and paranasal sinus, hypopharynx, larynx, salivary gland, and others). To assess the effect of therapy, the study population was divided into seven subgroups based on treatment modality (no treatment, surgery alone, RT alone, CT alone, surgery and CT, surgery and RT, concurrent chemoradiation therapy (CCRT), surgery and CCRT). Statistical analyses were performed using SAS version 9.2 (SAS Institute, Cary, NC, USA). A two-sided P-value < 0.05 was considered to indicate statistical significance.

Data availability
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request. www.nature.com/scientificreports/ Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat iveco mmons .org/licen ses/by/4.0/.