Многомерные детерминанты смертности в результате сердечно- сосудистых заболеваний у пациентов, переживших рак языка: популяционное когортное исследование на базе данных SEER

Введение. Рак языка (РЯ) является распространенной злокачественной опухолью головы и шеи с высокой заболеваемостью и смертностью. Эпидемиологические исследования показывают, что структура причин смерти у пациентов с РЯ изменяется со временем: в ранние сроки преобладает смертность по причине онкологического заболевания, тогда как у долгожителей возрастает доля причин не связанных с онкологическим заболеванием, в первую очередь сердечно-сосудистых заболеваний (ССЗ). Лечебные подходы, такие как лучевая и химиотерапия, могут дополнительно увеличивать сердечно-сосудистые риски.
Цель. Определить риск смертности от сердечно-сосудистых заболеваний у пациентов, переживших рак языка, выявить группы высокого риска и проанализировать многомерные детерминанты на популяционном уровне.
Материалы и методы. Проведено ретроспективное когортное исследование на основе базы SEER (2000–2021 гг.). В анализ включено 7691 пациент с РЯ после применения критериев исключения. Стандартизованные коэффициенты смертности (SMR) и избыточный риск определялись путем сравнения наблюдаемой смерт- ности от ССЗ с ожидаемой в сопоставимых по возрасту, полу и расе группах населения. Для стратификации риска использовали кривые накопленного риска Нельсона—Алена и пуассоновскую регрессию.
Результаты. Риск смертности от ССЗ у пациентов с РЯ был существенно выше (SMR = 3,16; 95% ДИ: 3,15–3,17), с выраженной гетерогенностью между подгруппами. Наивысший риск зафиксирован у пациен- тов, подвергшихся лучевой терапии (SMR = 11,48), с отдалёнными метастазами (SMR = 7,05), а также среди социально уязвимых групп — афроамериканцев (SMR = 5,76) и вдов/разведённых (SMR = 15,92–16,18). Основной причиной ССЗ стала ишемическая болезнь сердца (42,9%, SMR = 15,61), далее следовали церебро- васкулярные заболевания (17,3%, SMR = 13,69). У пациентов, не получавших химиотерапию, также отмечен повышенный риск (SMR = 7,33), что может быть связано с сопутствующей патологией либо не описанными методами лечения.
Заключение. Выжившие после рака языка имеют значительно повышенное бремя смертности от сердечно- сосудистых заболеваний, обусловленное токсичностью терапии, особенностями опухоли и социально-демографическими неравенствами. Необходима интеграция сердечно-сосудистого мониторинга в наблюдение за такими пациентами, особенно после лучевой терапии и в группах с социальной уязвимостью. Дальнейшие исследования должны быть направлены на механистические аспекты, создание прогностических моделей риска и разработку интервенций, ориентированных на уменьшение двойного бремени онкологических и сердечно-сосудистых заболеваний.
Ключевые слова: рак языка; SEER (Surveillance, Epidemiology, and End Results); сердечно-сосудистые заболевания (ССЗ)
Конфликт интересов. Авторы заявляют об отсутствии конфликта интересов.
Финансирование. Работа выполнена без спонсорской поддержки.

Background. Tongue cancer (TC) is a prevalent head and neck malignancy with significant morbidity and mortality. Epidemiological studies indicate that mortality causes in TC patients shift over time: cancer-related deaths dominate early stages, while long-term survivors face increasing non-cancer mortality, primarily cardiovascular disease (CVD). Treatment modalities such as radiotherapy (RT) and chemotherapy may exacerbate CVD risks, compounded by sociodemographic and tumor-related heterogeneity.
Objective. This study aimed to quantify CVD-related mortality risk in TC survivors, identify high-risk subgroups, and explore multidimensional determinants using population-level data.
Methods. A retrospective cohort study was conducted using the Surveillance, Epidemiology, and End Results (SEER) database (2000–2021). A total of 7,691 TC patients were included after applying exclusion criteria.Standardized mortality ratios (SMRs) and excess risks were calculated by comparing observed CVD deaths to expected rates in age-, sex-, and race-matched general populations. Nelson-Aalen cumulative hazard curves and Poisson regression were employed for risk stratification.
Results. CVD-related mortality in TC survivors was significantly elevated (SMR = 3.16, 95% CI: 3.15–3.17), with pronounced heterogeneity across subgroups. Radiotherapy-exposed patients (SMR = 11.48), those with distant metastases (SMR = 7.05), and socially vulnerable populations – Black individuals (SMR =5.76) and widowed/separated patients (SMR = 15.92–16.18) – faced the highest risks. Ischemic heart disease dominated CVD deaths (42.9%, SMR =15.61), followed by cerebrovascular diseases (17.3%, SMR = 13.69). Non-chemotherapy patients exhibited elevated risks (SMR = 7.33), potentially due to comorbidities or unmeasured treatment interactions.
Conclusion. TC survivors face a substantially increased CVD mortality burden, driven by treatment toxicity, tumor biology, and sociodemographic disparities. Integrating cardiovascular surveillance into survivorship care is critical, particularly for irradiated patients and socially vulnerable groups. Future research should prioritize mechanistic studies, risk prediction models, and equity-focused interventions to mitigate dual cancer-CVD burdens. Key words: TC SEER Cardiovascular Disease
Conflict of interest. The authors declare that there is no conflict of interest.
Funding. The study has not received any funding.

Background
TC (TC) is one of the most common head and neck malignancies globally, with significant morbidity and mortality rates [1, 2]. Epidemiological studies indicate that causes of death in TC patients exhibit distinct time-dependent patterns: cancer-related deaths dominate during early follow-up periods, while long-term survivors face a gradual shift in mortality risk toward non-cancer factors, particularly cardiovascular disease (CVD) [3,4]. Treatment modalities, including surgery, radiotherapy, and chemotherapy, may substantially elevate non-cancer mortality risks due to treatmentrelated toxicities. Additionally, TC mortality risk demonstrates marked heterogeneity influenced by multidimensional factors such as age, race, marital status, and tumor stage [5]. Despite advancements in therapeutic strategies and improved prognoses, TC patients remain at high risk of mortality, especially from CVD, which significantly exceeds rates observed in the general population. This study aimed to comprehensively analyze mortality patterns and associated risk factors in TC patients, with a focus on the incidence and determinants of CVD-related deaths [6,7,8]. The Surveillance, Epidemiology, and End Results (SEER) database, a publicly accessible cancer registry encompassing over 8 million cancer cases across multiple U.S. states, was utilized for this study. Renowned for its large sample size and comprehensive followup information, the SEER database has been widely employed to investigate cancer incidence, prognosis, and mortality risks, including assessments for colorectal and breast cancers [9,10]. Leveraging this population-based cohort, we systematically evaluated mortality risks in TC patients, emphasizing the characteristics of CVD-related deaths and high-risk subgroups to provide robust evidence for clinical practice. Materials and methods This study extracted data from the Surveillance, Epidemiology, and End Results (SEER) database for all TC patients diagnosed between January 1, 2000, and December 31, 2021. For comparison, general population mortality data from the same period (2000–2021) were obtained from the National Center for Health Statistics (NCHS) [11]. A total of 10,728 patients diagnosed with TC were initially identified. Inclusion criteria required histologically confirmed tongue malignancies, while exclusion criteria removed patients who received no antitumor therapy, those with secondary or multiple primary tumors, and cases with incomplete diagnostic/survival information or tumors identified solely via death certificates. After applying these criteria, 7,691 eligible patients were included, of whom 492 died from cardiovascular disease (CVD) [12]. CVD-related deaths, defined by SEER codes for six conditions – (1) cerebrovascular diseases, (2) diseases of arteries/arterioles/capillaries, (3) hypertension, (4) ischemic heart disease, (5) pulmonary heart disease, and (6) other/ unspecified circulatory disorders – served as the primary endpoint. Competing risks included TC-related deaths, deaths from metastatic complications, and non-CVD/non-cancer deaths. This retrospective study utilized de-identified clinical data and adhered to SEER ethical guidelines, exempting informed consent. This multicenter retrospective cohort study analyzed demographic and clinical variables: age at diagnosis, race, sex, diagnosis year, histological type, tumor stage, radiotherapy/chemotherapy status, marital status, median household income, and follow-up information (survival time and cause of death). Causes of death were categorized as: (1) TC-related, (2) metastatic/complication-related, (3) CVDrelated, or (4) other non-cancer causes. CVD-related deaths were specifically coded using SEER-defined terms: hypertensive disease, ischemic heart disease, cerebrovascular diseases, pulmonary heart disease, diseases of arteries/arterioles/capillaries, and other/ unspecified circulatory system disorders [13]. The primary outcome was CVD-related death. Patients not deceased due to CVD at the last follow-up were censored. Followup duration was calculated from the date of diagnosis to death or censoring. Survival times recorded as 0 months were converted to 0.5 months per epidemiological conventions. Statistical Analysis CVD mortality rates were calculated by dividing CVD-related deaths by the total number of TC patients. Standardized mortality ratios (SMRs) were derived by comparing observed CVD deaths in TC subgroups to expected deaths in age-, sex-, and race-matched general populations (reference data from the National Cancer Institute). Poisson regression was used to estimate 95% confidence intervals (CIs) for SMRs. Excess risk was defined as: Nelson-Aalen cumulative hazard curves were plotted to assess temporal trends in CVD-related mortality across subgroups. Statistical significance was set at p<0.05p<0.05. Analyses were performed using SEER*STAT, Python, and Microsoft Excel 2019. Results The study included 7,691 eligible TC patients with a mean age of 62.83 years and a median follow-up duration of 38 months. The majority of patients were aged 40–59 years (35.3%), White (n=6,252, 81.2%), male (n=4,333, 56.3%), married (n=4,249, 55.24%), and had localized tumor invasion (n=4,866, 63.26%). Median household income exceeded $70,000 in 79.88% of cases. Histologically, squamous cell carcinoma predominated (n=7,561, 98.3%), while other pathological types accounted for 1.69% (n=130). Radiotherapy and chemotherapy were administered to 27.42% (n=2,109) and 17.38% (n=1,331) of patients, respectively. Overall survival analysis revealed significant time-dependent patterns in all-cause mortality. Among 3,761 deaths during followup, mortality peaked in early follow-up periods: 34.69% (n=1,305) occurred within <1 year, followed by 33.98% (n=1,278) in the 1–3 year interval. A dynamic shift in causes of death was observed: cancer-related deaths declined from 56.8% in the <1-year period to 15.3% in the >10-year period (χ2 trend test, p<0.001), while noncancer- related deaths increased from 41.9% to 80.2%, indicating a transition toward non-oncological mortality risks in long-term survivors (Figure 1). Among non-cancer deaths, cardiovascular disease (CVD) was the leading cause (n=491, 13.06%), with significant heterogeneity in CVD subtypes (p<0.05): ischemic heart disease predominated (42.9%), followed by other/unspecified circulatory disorders (23.6%), cerebrovascular diseases (17.3%), hypertensive disease (10.7%), and pulmonary heart disease/arterial disorders (2.6% each). A total of 491 TC patients died from cardiovascular disease (CVD). Compared to an age-, sex-, and race-matched general population cohort (expected deaths = 155.2), the standardized mortality ratio (SMR) was 3.16 (95% CI: 3.15–3.17, p<0.0001), with an excess risk of 335.8 per 10,000 person-years. Baseline characteristics and subgroup-specific SMRs are detailed in Table 1. This study demonstrates that CVD-related mortality risk in TC patients is significantly elevated compared to the general population (overall SMR = 3.16, 95% CI: 3.15–3.17), with marked heterogeneity across demographic and clinical subgroups. Age exhibited a nonlinear inverse association with SMR: middle-aged to older patients (60–69 years: SMR = 3.55; 70–79 years: SMR = 3.59) had the highest risks, whereas younger (15–39 years: SMR = 1.08) and elderly (≥80 years: SMR = 2.95) patients showed lower risks (all p<0.05). Racial disparities were pronounced, with Black patients facing the highest risk (SMR = 5.76, 95% CI: 4.00–8.10), followed by Asian or Pacific Islander patients (SMR = 5.51, 95% CI: 4.19–7.25). Social support deficits further amplified risk, as separated patients exhibited an exceptionally high SMR of 16.18 (p<0.05). Tumor-related factors also influenced outcomes: distant metastases (SMR = 7.05, 95% CI: 4.61–10.36) and radiotherapy exposure (SMR = 11.48 vs. 2.03 in non-irradiated patients) were associated with substantially elevated CVD mortality. Ischemic heart disease was the leading cause of cardiovascular disease (CVD)-related deaths, accounting for 42.9% of CVD mortality (Standardized Mortality Ratio [SMR] = 15.61, 95% CI: 13.46–18.00). Other CVD subtypes also exhibited significantly elevated risks: cerebrovascular diseases (17.3% of CVD deaths, SMR = 13.69), other/unspecified circulatory disorders (23.6%, SMR = 13.52), and hypertensive disease (10.7%, SMR = 11.25) (all P<0.05). Notably, despite fewer deaths from pulmonary heart disease (SMR = 13.68) and diseases of arteries/capillaries (SMR = 12.50), their risks remained significantly higher than those in the general population (P<0.05). Figure 2 displays Nelson-Aalen cumulative hazard curves for cardiovascular disease (CVD)-related mortality in TC patients, stratified by subgroups including age, race, sex, marital status, household income, tumor stage, radiotherapy, and chemotherapy. The analysis demonstrates that cumulative CVD mortality risks significantly increased with prolonged follow-up in high-risk subgroups, such as older patients (e.g., ≥70 years), those with distant metastases, individuals receiving radiotherapy, Black and Asian/Pacific Islander populations, and widowed or separated patients. Key determinants of elevated CVD mortality included advanced tumor stage (e.g., distant metastasis: SMR = 7.05), treatment modalities (e.g., radiotherapy: SMR = 11.48), race (e.g., Black patients: SMR = 5.76), marital status (e.g., separated individuals: SMR = 16.18), and older age (60–79 years: SMR = 3.55–3.59). These factors exhibited strong associations with CVDrelated mortality (P<0.05 for all), underscoring the multifactorial nature of cardiovascular risk in this population. Discussion TC is one of the most common head and neck malignancies worldwide, with significant prognostic disparities across disease stages. Advanced-stage patients exhibit substantially lower survival rates compared to early-stage counterparts. While advancements in multimodal therapies – including surgery, radiotherapy, and chemotherapy – have modestly improved outcomes for advanced TC, overall prognosis remains suboptimal. Despite extensive research on treatment strategies and long-term survival in advanced disease, studies focusing on mortality risk determinants, particularly noncancer causes such as cardiovascular disease (CVD), remain limited. This study systematically evaluated CVD-related mortality risk in TC patients using the SEER database, revealing a significantly elevated standardized mortality ratio (SMR = 3.16, 95% CI: 3.15–3.17) compared to the general population, with marked multidimensional heterogeneity. First, age emerged as a critical determinant: middle-aged to older patients (60–79 years) faced the highest risk (SMR = 3.55–3.59), whereas younger (15–39 years, SMR = 1.08) and elderly (≥80 years, SMR = 2.95) patients exhibited lower risks. This pattern may reflect age-related declines in baseline cardiovascular health and cumulative treatment toxicities. Second, tumor stage strongly influenced outcomes: distant metastases conferred the highest cumulative risk (SMR = 7.05), followed by locally advanced disease with nodal involvement, while localized tumors carried the lowest risk. These findings suggest that tumor burden exacerbates cardiovascular injury, potentially through systemic inflammation or metabolic dysregulation [14]. The impact of treatment modalities on cardiovascular disease (CVD) mortality risk is significant. Patients who underwent radiotherapy exhibited a substantially higher risk of CVD-related mortality compared to those who did not receive radiotherapy (SMR = 11.48 vs. 2.03). Similarly, patients who did not undergo chemotherapy also demonstrated a higher risk (SMR = 7.33). This finding may appear to deviate from general medical logic and could be attributed to several factors. Prior to initiating antitumor treatments, patients’ cardiovascular health is thoroughly evaluated, and only those with relatively better cardiovascular conditions are deemed eligible for chemotherapy. Consequently, chemotherapy patients tend to benefit from longer survival times, which are closely associated with an increased likelihood of CVD-related mortality [15,16]. The chemotherapy regimens commonly used for tongue cancer patients include platinum-based agents and taxanes, both of which exert multifaceted effects on the cardiovascular system. Platinum-based chemotherapeutic agents (e.g., cisplatin, carboplatin, oxaliplatin) are widely utilized for their broad-spectrum antitumor efficacy, including the treatment of head and neck cancers, lung cancer, and breast cancer [17]. These agents interact with DNA, inducing cross-linking, disrupting DNA function, and ultimately leading to apoptosis. In a study of testicular cancer survivors, 53% of patients treated with cisplatin developed hypertension, with an incidence rate 2.3 times higher than that of the control group [18]. Sociodemographic disparities further stratified risk. Black (SMR = 5.76, 95% CI: 4.00–8.10) and Asian/Pacific Islander (SMR = 5.51, 95% CI: 4.19–7.25) patients faced disproportionately higher CVD mortality than White patients (SMR = 4.87), likely due to socioeconomic inequities, healthcare access barriers, or genetic predispositions. Marital status also significantly impacted outcomes: widowed (SMR = 15.92) or separated (SMR = 16.18) patients had exceptionally high risks compared to married counterparts (SMR = 5.51), highlighting the role of social support in mitigating cardiovascular morbidity [19,20]. Limitations This study has several limitations. First, reliance on SEER data introduces potential biases from incomplete records (e.g., missing diagnostic/survival dates) and unmeasured confounders (e.g., lifestyle factors, comorbidities). Second, competing risks from complex mortality causes (e.g., cancer progression, non-CVD comorbidities) may overestimate CVD-specific mortality. Third, insufficient treatment details (e.g., radiotherapy doses, chemotherapy regimens) limit mechanistic insights into therapy-related cardiovascular toxicity. Finally, the 21-year study span introduces temporal heterogeneity, as evolving treatment protocols and CVD management guidelines may affect longitudinal risk patterns. Future Directions Prospective cohorts integrating granular treatment data, cardiovascular biomarkers (e.g., high-sensitivity troponin, inflammatory cytokines), and social determinants of health are needed to clarify mechanisms – such as radiation-induced oxidative stress or tumor-derived endothelial damage – and refine risk prediction models. Interventions targeting high-risk subgroups (e.g., irradiated patients, socioeconomically disadvantaged populations) should prioritize cardioprotective strategies (e.g., statins, lifestyle modifications). Community-level initiatives to address healthcare access disparities, particularly for Black and uninsured groups, are equally critical to mitigate biological and structural CVD risks [21, 22]. By bridging oncology and cardiology disciplines, future efforts may reduce the dual burden of cancer and CVD, ultimately improving long-term outcomes for TC survivors. Conclusion This study underscores the imperative to integrate cardiovascular risk surveillance into survivorship care for TC patients, particularly those exposed to radiotherapy or experiencing social vulnerability. The multidimensional determinants of CVD-related mortality – spanning treatment toxicity, tumor biology, and sociodemographic disparities – highlight the need for interdisciplinary collaboration between oncology and cardiology. Future efforts should prioritize mechanistic research and equity-focused interventions to mitigate the dual burden of