Abstract

Background

The objective of this research is to statistically assess the risk of cardiovascular mortality (CVM) between patients with small bowel adenocarcinoma (SBA) and the general population. Additionally, it aims to identify CVM-associated risk factors among individuals with SBA.

Methods

Data obtained between 2000 and 2017 on SBA patients from the Surveillance, Epidemiology, and End Results (SEER) database were examined. Standardized mortality ratios (SMRs) and absolute excess risks (AERs) were obtained to compare CVM between patients and the general US population. To evaluate the cumulative mortality (CM) rate for all causes of death (COD), cumulative hazard curves were constructed. Two multivariate competing risk models were established to determine the independent predictors for CVM.

Results

In the cohort of 5,175 SBA patients observed for 15,068.24 person-years, a total of 205 deaths were attributed to cardiovascular disease (CVD). The overall SMR of CVM was 1.41 (95% confidence interval (CI): 1.23–1.62, P < 0.05), whereas it reached 2.99 during the early stage of latency. Additionally, independent risk factors for CVM included age, marital status, calendar year of diagnosis, disease differentiation degree, SEER stage, and chemotherapy status.

Conclusions

SBA patients exhibited a substantially elevated risk of developing CVM compared to the general US population. During the follow-up period, the CM rate for CVM continued to rise steadily. Timely identification of high-risk groups and effective interventions to safeguard cardiovascular health significantly improve patient prognosis.

Introduction

Clinically, small bowel adenocarcinoma (SBA) is a relatively rare malignancy, considered an orphan disease. Among all types of small bowel cancers, SBA is the most prevalent, accounting for 30-40% of cases [1,2,3]. Although uncommon, the incidence of SBA has been steadily increasing worldwide in recent years, with an estimated 5,300 newly diagnosed cases and 1,100 death cases annually in the United States [4, 5]. A large population-based study in the US revealed a rising trend in the annual percentage change (APC) of its incidence. An APC of 1.47% (P < 0.05) was documented in one study [6], while recent research conducted in the Netherlands reported a higher APC of 3.70% (p < 0.05) [7].

Due to the absence of specific symptoms at the early stage, diagnosis of SBA typically occurs at a relatively older age, ranging from 56 to 76 years [6]. However, the popularization of lymph node assessments, increased safety awareness among physicians, and advanced cancer management have led to improved cancer prognoses in recent years. According to the clinical guidelines provided by the National Comprehensive Cancer Network (NCCN), the current 5-year survival rate is 85% for localized disease and 42% for patients diagnosed with stage IV [8]. As the life expectancy among cancer survivors increases and the mortality rate of primary cancers gradually declines, non-cancer causes of death (COD) have gained prominence. Among these, cardiovascular disease (CVD) stands out as the leading cause of mortality.

Even within the general population, CVD is the prevalent cause of mortality in the world. According to the Global Burden of Disease study, CVD-related deaths have shown a consistent rise, with 12.1Ìýmillion deaths reported in 1990 and 18.6Ìýmillion in 2019. Currently, CVD is linked to one-third of all deaths attributed to all causes [9]. A statistical report encompassing over 2.8Ìýmillion deaths in the US revealed that heart disease accounted for 647,457 (23.0%) deaths, representing a considerable proportion of all COD [10]. CVD patients tend to have lifestyle-associated risk factors similar to cancer patients, such as excessive alcohol intake, obesity, smoking, or/and unhealthy eating habits [11]. The risk of developing CVD in individuals with cancer is higher than that in the general population. Furthermore, certain anticancer therapies, like radiotherapy and chemotherapeutics represented by anthracyclines and cyclophosphamide, as well as targeted therapies including monoclonal antibodies and tyrosine kinase inhibitors (TKIs), have been identified as exerting adverse effects on the cardiovascular system of patients [12,13,14].

Cancer patients face an elevated risk of developing CVD due to specific risk factors, leading to the emergence of a new discipline known as cardio-oncology [15, 16]. Further identification of relevant characteristics of cancer patients susceptible to cardiovascular mortality (CVM) would facilitate the implementation of practical measures aimed at reducing CVM occurrences. Recently, individuals with specific cancer types have demonstrated an elevated CVM risk in comparison with the general population. Moreover, corresponding risk factors have been identified. For instance, endometrial cancer patients exhibited a remarkably greater risk of CVM (standardized mortality ratio (SMR) = 8.80, 95% confidence interval (CI): 8.70-9.00, P < 0.05) [17].

While the epidemiological link between certain types of cancer and CVM has been known, the majority of studies have concentrated on patients with the more prevalent cancer types [18, 19]. In contrast, for SBA, which is experiencing a gradual increase in incidence, no study has generated meaningful results demonstrating the risk of CVM in cancer survivors compared to the general population. Furthermore, research investigating the cumulative incidence of CVD-related time-point-specific mortality in SBA patients, along with the cumulative incidence of CVM stratified by age at diagnosis and primary site, remains insufficient. Despite an array of factors have been recognized as affecting both cancer and CVM, no significant risk factors linked to this cancer have been identified. Moreover, although some studies have recognized potential risk factors for CVM in cancer patients, they have employed the Cox proportional hazards regression model that considers CVM as the sole endpoint event [20, 21]. This approach may produce unreliable results due to the presence of competing events, particularly during long-term follow-up [22]. To accurately identify independent risk factors, it is essential to utilize the competing risk regression for analysis. Therefore, to fill in the research gaps, this study aimed to: (i) determine the notable differences in CVM risk between individuals with SBA, whose incidence is gradually increasing, and the general population, (ii) investigate the cumulative incidence of CVD-related time-point-specific mortality, as well as the cumulative incidence of CVM stratified by age at diagnosis and primary site in SBA patients, and (iii) use the competing risk regression analysis to reveal risk factors with the intent to outline early precautionary measures for cancer patients.

Methods

Data source and patient selection

The SEER*Stat software was utilized to examine the Surveillance, Epidemiology, and End Results (SEER) database (Incidence-SEER 18 Regs excluding AK Research Plus Data, November 2019 Sub (2000–2017) for SMRs). A SEER research data agreement was signed before accessing data files with the accession and approval number 10,095-Nov2021. The requirement for ethical approval and informed consent was exempted due to the de-identification of patient data retrieved from the database and its public accessibility. All individuals diagnosed with SBA as their initial primary tumor between 2000 and 2017 were included in the study. Moreover, this research employed the following histological codes: 8140, 8143–8145, 8210, 8211, 8220, 8255, 8260–8263, 8310, 8480, 8481, and 8490 (ICD-O-3). The primary site codes for the small bowel were C17.0–C17.9 with malignant behavior. Patients diagnosed solely through autopsy or a death certificate and those with incomplete data (marital status, grade, summary stage, etc.) were screened out (Fig.Ìý1). The study focused primarily on CVM as the main event of interest. Utilizing the ICD-10 codes, the primary outcome was grouped into six conditions: heart disease (I00-I09, I11, I13, I20-I51); hypertension without heart disease (I10, I12); cerebrovascular disease (I60-I69); atherosclerosis (I70); aortic aneurysm and dissection (I71); and other conditions affecting arteries, arterioles, capillaries (I72-I78). Furthermore, the competing events encompassed primary cancer mortality (PCM), other cancer mortality (OCM), and other non-cancer mortality (ONCM). This research was conducted following the STROBE guidelines [23].

Fig. 1

Flow chart of the screening process of SBA patients and overall study design. SBA: small bowel adenocarcinoma; SEER: Surveillance, Epidemiology, and End Results; CVM: cardiovascular mortality; SMR: standardized mortality ratio; AER: absolute excess risk; CM: cumulative mortality; COD: causes of death; CIF: cumulative incidence function

Research variables

Information regarded as variables was extracted. These variables encompassed age at diagnosis, race, gender, marital status (married, unmarried), year of diagnosis, primary site (duodenum, jejunum, ileum, unspecific), grade (grade I-IV), summary stage (localized, regional, distant), histologic type (adenocarcinoma; mucinous cell carcinoma (MCC); signet ring cell carcinoma (SRCC), chemotherapy, radiotherapy, surgery, cause of death, and latency (a time-dependent variable that represents the duration from the initial diagnosis of primary tumor to the date of death from any cause [24]).

Statistical analysis

SMRs were utilized to compare CVM between SBA patients and general US residents. These SMRs were calculated as the ratio of the total observed deaths to the expected deaths attributed to CVD [25, 26]. Meanwhile, 95% CIs for all SMRs were computed using the same method [27]. The absolute excess risk (AER), which represents excess deaths per 10,000 person-years within each subgroup, was calculated using the formula: (observed death - expected deaths) / (person-years of observation) [25, 26]. A cumulative incidence function (CIF) curve was generated to illustrate the cumulative incidence of time-point-specific mortality [28]. To assess the association among each factor, univariate competing risk regression analyses were utilized. Moreover, the risk of developing CVM in individuals with SBA was determined through Gray’s test. Significant factors were incorporated into multivariate competing risk analyses for the identification of independent risk factors and the calculation of hazard ratio (HR) as well as 95% CI.

The National Cancer Institute SEER*Stat software (seer.cancer.gov /seerstat) version 8.4.0.1 and R Statistical software version 4.1.3 (R Foundation for Statistical Computing) were employed to conduct all analyses. All the tests were two-sided, with a statistical significance level defined as a P-value < 0.05.

Results

Patient characteristics

A total of 5,175 SBA patients were identified from 2000 to 2017. Among them, 205 individuals died from CVD, and 1,751 died due to SBA. The mean age at diagnosis for patients who experienced CVD-related mortality was 73.9 ± 10.5 years. The majority of these patients were diagnosed between 2000 and 2008 (66.8%), aged > 60 years (86.3%), belonged to the white racial group (76.1%), were male (53.7%), had the duodenum as the primary site (55.1%), developed grade I-II tumors (72.2%), had adenocarcinoma histological type (86.8%) and presented with a localized tumor stage (42.0%). The proportion of married patients (49.3%) was similar to that of unmarried ones (50.7%). Most of the patients underwent surgery (78.5%), followed by chemotherapy (16.1%) and radiotherapy (5.4%).

The mean age at diagnosis for patients who died from SBA was 65.5 ± 13.4 years. Most of them were married (60.5%) and had a distant tumor stage (43.7%). In addition, the proportion of patients who received chemotherapy (50.4%) was similar to patients with no history of chemotherapy (49.6%). TableÌý1 provides a summary of baseline demographics and clinical characteristics of the study population.

Table 1 Demographic, tumor-specific, and treatment characteristics of SBA patients

Changes in all causes of death

The changes in all COD among SBA patients by year of diagnosis, latency, and age at diagnosis are depicted in Fig.Ìý2. Although most SBA patients experienced mortality due to primary cancer between 2000 and 2017, non-cancer COD (including CVM and ONCM) was also taken into account, constituting approximately 10–20% of all recorded deaths (Fig.Ìý2A). A slight decline in the proportions of CVM and ONCM was observed in the initial three years of latency, but these proportions began to rise during the follow-up period. By the fifth year of latency, CVM and ONCM proportions (39.03%) exceeded the proportion of PCM (24.16%) to become the first COD (Fig.Ìý2B). Moreover, with increasing age at diagnosis, the proportions of CVM and ONCM exhibited a gradual rise, whereas the proportion of PCM decreased (Fig.Ìý2C).

Fig. 2

Proportions changes of all causes of death among SBA patients by different time-dependent indicators. (A) Year of diagnosis. (B) Latency. (C) Age at diagnosis. Causes of death were indicated as CVM, ONCM, PCM, and OCM from bottom to top in order. CVM: cardiovascular mortality; PCM: primary cancer mortality; OCM: other cancer mortality; ONCM: other non-cancer mortality

Standardized mortality ratio and absolute excess risk

From 2000 to 2017, the incidence of CVM among the 5,175 SBA patients stood at approximately 3,961 cases per 100,000 person-years. In comparison, the general population exhibited a rate of 2,800 cases per 100,000 person-years, resulting in an SMR of 1.41 (95% CI: 1.23–1.62) (TableÌý2). The AER was calculated to be 39.88 per 10,000 person-years. SMRs and AERs of CVM in patients with SBA, categorized by various factors, are presented in TableÌý2. Significantly raised SMRs and increased AERs were observed among unmarried patients, individuals of both white and black races, those having duodenum or unspecified site as the primary site, grade I / II differentiation degree, localized and distant stages, and those without chemotherapy or radiotherapy, in comparison to the general population. In addition, SBA patients diagnosed at an earlier age were found to be at a higher risk of developing CVM (≤ 60 years: 1.84, 95% CI: 1.23–2.67; >60 years: 1.36, 95% CI: 1.17–1.58). The minority of SBA patients (21.5%) who did not undergo surgery exhibited a substantially higher SMR in comparison to the rest of the patients (surgery: 1.21, 95% CI: 1.03–1.42; no surgery: 3.56, 95% CI: 2.58–4.77).

Table 2 Standardized mortality ratios of cardiovascular mortality among SBA patients

Changes in SMRs of all COD based on the year of diagnosis, latency, and age at diagnosis are depicted in Fig.Ìý3. As the year of diagnosis progressed, the SMRs of all COD, including CVM, exhibited an increase. Although PCM had the highest SMR in earlier diagnoses, the SMRs of OCM and ONCM, which include CVM, surpassed that of PCM in 2019 − 2011 and 2012–2014, respectively (Fig.Ìý3A). The SMRs of all COD in SBA patients gradually declined with increasing latency. However, at the early latency stage, a remarkably higher risk of all COD was observed. During the latency periods of 0–5 months and 6–11 months, the risk of SBA patients developing CVM was remarkably elevated, increasing to 14.41-fold (95% CI, 11.04–18.47) and 8.19-fold (95% CI, 5.45–11.84) compared to the general population, respectively (Fig.Ìý3B). Consistent with the pattern observed in the change of SMRs by latency, the SMR for all COD decreased steadily with age. However, the risk of SBA patients developing CVM ranged from 2.99-26.52-fold compared to the general population (Fig.Ìý3C).

Fig. 3

SMRs changes of all causes of death among SBA patients by different time-dependent indicators. (A) Year of diagnosis. (B) Latency. (C) Age at diagnosis. CVM: cardiovascular mortality; PCM: primary cancer mortality; OCM: other cancer mortality; ONCM: other non-cancer mortality

Cumulative mortality (CM) of all causes of death

The Fine Gray distribution hazard model was employed to generate the cumulative hazard curves for all COD in SBA patients (Supplementary Fig.Ìý1). During the follow-up period, the CM rate for CVM remained lower than any other causes and exhibited a consistent upward trend.

At the 200-month follow-up point, a stratified analysis of SBA patients by age indicated that CM rates of CVM increased progressively with age at diagnosis (TableÌý3). Over the entire follow-up duration, the CM rate for CVM remained the lowest among all COD in patient subgroups aged ≤ 50, 51–60, 61–70, and 71–80 years (Fig.Ìý4A-D). In patients older than 80 years, this rate surpassed that of ONCM and ranked third throughout most of the follow-up period (Fig.Ìý4E). At the 200-month follow-up point, the CM rates for CVM and ONCM were 14.47% and 12.77%, respectively (TableÌý3).

Table 3 Cumulative mortality stratified by age at diagnosis and primary site at 200 months follow-up

Fig. 4

Cumulative mortality for all causes of death among SBA patients stratified by age at diagnosis. (A) Age: ≤ 50 years. (B) Age: 51–60 years. (C) Age: 61–70 years. (D) Age: 71–80 years. (C) Age: > 80 years. CVM: cardiovascular mortality; PCM: primary cancer mortality; OCM: other cancer mortality; ONCM: other non-cancer mortality

In the stratified analysis based on the primary site, it was indicated that patients with duodenal adenocarcinoma exhibited the lowest (5.92%), while those with ileal adenocarcinoma exhibited the highest (7.85%) CM rate for CVM at 200 months of follow-up (TableÌý3). Over that period, the rate of CM for PCM was the highest of all COD for patients with duodenal or jejunal adenocarcinoma (Fig.Ìý5A, B). In contrast, the rate of CM for OCM was elevated among patients with ileal or unspecified adenocarcinoma (Fig.Ìý5C, D). The CM rate of CVM was the lowest in patients with duodenal, jejunal, or unspecified adenocarcinoma throughout the follow-up period (Fig.Ìý5A, B, D). During 62–96 months and 141–180 months of follow-up, the CM rate of CVM surpassed that of ONCM for patients with ileal adenocarcinoma (Fig.Ìý5C).

Fig. 5

Cumulative mortality for all causes of death among SBA patients stratified by primary site. (A) Primary site: Duodenum. (B) Primary site: Jejunum. (C) Primary site: Ileum. (D) Primary site: Unspecific. CVM: cardiovascular mortality; PCM: primary cancer mortality; OCM: other cancer mortality; ONCM: other non-cancer mortality

Univariate analyses of CVM among SBA patients

To evaluate the correlation between each variable and the cumulative incidence of CVM in patients, multiple factors were incorporated into the CIF curves (Supplementary Fig.Ìý2). As indicated by the results of univariate competing risk regression analyses, variables such as age (in years), marital status, year of diagnosis, degree of differentiation, summary stage, chemotherapy, and radiotherapy statuses exhibited significant associations with CVM in SBA patients (Supplementary Fig.Ìý2 and Supplementary Table 1).

Predictors of CVM among SBA patients

The multivariate competing risk regression analysis identified independent predictors of developing CVM and avoided the likelihood of producing false positive findings. The meaningful variables identified in univariate competing risk regression analysis were utilized to make adjustments in the multivariate competing risk of model. The multifactor competitive risk forest plot based on model is presented in Fig.Ìý6.

Fig. 6

Multifactor competitive risk forest plot. Marital status: Unmarried (Single, Separated, Divorced, Widowed, Unmarried or Domestic Partner); I: Well differentiated, II: Moderately differentiated, III: Poorly differentiated, IV: Undifferentiated

The independent variables that were linked to higher risks of CVM are as follows: over 60 years old (HR: 3.103; 95% CI: 2.058–4.679), unmarried (HR: 1.336; 95% CI: 1.014–1.761), and no history of chemotherapy (HR: 2.524; 95% CI: 1.672–3.810). In contrast, the independent variables that were linked to lower risks of CVM are as follows: initial diagnosis between 2009 and 2017 (HR: 0.529; 95% CI: 0.396–0.707), grade III / IV of cancer (HR: 0.716; 95% CI: 0.523–0.981), and either a regional (HR: 0.599; 95% CI: 0.436–0.824) or a distant tumor stage (HR: 0.384; 95% CI: 0.254–0.581).

Discussion

Two substantial contributors to mortality rates are cancer and CVD. Prior research has indicated that CVM risk is rising among individuals with malignancies. This increase can be attributed to decreased mortality related to primary cancer [29]. In a recent retrospective study involving 234,256 patients across 28 primary cancer sites, the data revealed an elevated risk of CVM among cancer patients in comparison to the general population. remarkably, the majority of CVM events were observed in individuals with breast, prostate, or bladder cancers [18].

Due to the rare occurrence of SBA and its high cancer-specific mortality rate, no prior studies have demonstrated a significant difference in CVM risk between SBA patients and the general population. Additionally, the identification of risk factors associated with this particular malignancy remains elusive. Therefore, SEER database analysis was performed on a substantial sample size and authentic cardiovascular mortality data. The analysis revealed a significant difference in the CVM risk between SBA patients and the general population, with an overall SMR of 1.41 (95% CI, 1.23–1.62, P < 0.05). The independent risk factors associated with CVM included age, marital status, calendar year of diagnosis, differentiation degree, SEER stage, and chemotherapy status.

After the assessment of changes in all COD, it was observed that, except for a slight decline in the first three years of latency, the proportions of CVM and ONCM have remarkably increased. Moreover, in the fifth year of latency, these proportions exceeded that of PCM and kept increasing with age. These findings indicate the need for enhanced CVD monitoring and care, particularly among older patients and those with longer survival times.

An analysis of the change of SMR over the years of diagnosis indicated a notable increase in CVM risk with each subsequent year of diagnosis. These findings have been validated by earlier studies and may be explained by advances in cancer treatment as well as reductions in cancer-specific mortality [6]. Patients live longer but are more likely to die from non-cancer diseases like CVD [18]. The SMR reached its highest, particularly during the latency period of 0 to 5 months. This period also corresponded to significantly elevated early-stage risk of CVD compared to the general population. Reports by Fang et al. (2010) and Ye et al. (2019) attribute this trend to psychological and mental stress experienced by patients following a recent cancer diagnosis. This stress is considered a contributing factor to the onset of CVM in these patients [29, 30]. These findings imply that clinicians should monitor and address the psychiatric and psychological well-being of patients following their cancer diagnosis and treatment. Meanwhile, patients exhibited a significantly elevated risk of CVM across all age subgroups, contributing to the observed changes in SMR based on age at diagnosis. Interestingly, there was a notable declining trend in SMR with increasing age. This trend was likely a result of the observed rise in CVM risk with advancing age in the general population.

In this study, the CM rate of CVM was the lowest among all COD but has been steadily increasing. Stratified analysis based on the age of diagnosis revealed that among patients older than 80 years, the CM rate of CVM surpassed that of ONCM and ranked third for the majority of the follow-up period. A previous report on breast cancer has established a similar trend [31].

Following a subgroup analysis based on the primary tumor site, the CM rate of PCM was found to be the highest throughout the follow-up period for patients with duodenal or jejunal adenocarcinoma. In contrast, for patients with ileal or unspecific adenocarcinoma, the CM rate of OCM surpassed that of PCM, emerging as the primary cause of death in this group. These findings emphasize the significance of monitoring the occurrence of other diseases in cancer patients and effectively managing the associated risk factors. This should be done in conjunction with monitoring of cancer treatments and primary cancer care, particularly among patients with ileal or unspecific adenocarcinoma. At 200 months of follow-up, CM rates of CVM stratified by tumor locations were analyzed. The analysis revealed that the rate was lowest in patients with duodenal adenocarcinoma and highest in those with ileal adenocarcinoma. This could be attributed to a potentially poorer prognosis associated with duodenal adenocarcinoma that leads to early death before the onset of CVM [8]. In contrast, patients with ileal adenocarcinoma exhibited a relatively high 5-year survival rate [32] and, thus, lived longer to potentially develop CVM.

Multivariate competing risk model was employed to identify several independent risk factors associated with CVM. Patients over 60 years of age exhibited a higher likelihood of experiencing CVM events. This phenomenon was primarily due to their increased exposure time to certain age-dependent risk factors and changes in the cardiovascular structure and its physiological functions commonly observed in elderly patients [33]. It was also found that patients diagnosed between 2009 and 2017 were less likely to develop CVM, which can be attributed to improvements and advancements in cardio-oncology in recent years [34]. Interestingly, the findings indicated that unmarried patients were inclined to experience CVM events. These findings align with a study on non-Hodgkin’s lymphoma, which reached a similar conclusion [35]. The possible explanation was that marital status had an essential impact on the risk of developing CVM among survivors of certain types of cancer. On the one hand, due to the better financial status of married patients, it was easier to obtain timely and professional medical care [36, 37]. On the other hand, the spouses’ encouragement and support could positively impact the patients both physically and psychologically, and it could reduce the risk of developing CVM [38].

In this study, patients with grade III / IV or a distant tumor stage were less likely to develop CVM. The plausible explanation for this was that patients with advanced tumor stages tend to have shorter lifespans. This limited lifespan reduces their exposure time to CVD risk factors, thereby reducing the chances of developing CVM [39, 40]. In terms of cancer treatment, patients with no history of chemotherapy were inclined to experience CVM events. This finding conflicts with the known cardiovascular damage associated with various therapies. Nevertheless, similar conclusions were also found in other studies [39, 41]. Hence, further research is required to clarify the precise impact of chemotherapy on CVM.

The current study has certain limitations. Firstly, information regarding the confounding factors related to CVD or pre-existing cardiovascular comorbidities was unavailable in the SEER database. Secondly, the database lacked insufficient information on therapy, such as radiation dosage, the chemotherapy agents administered, and molecular-type therapies. Thirdly, a minor selection bias among participants is inevitable due to the retrospective study design.

Conclusions

To conclude, SBA patients exhibited a remarkably elevated CVM risk than the general US population.During the follow-up period, the CM rate for CVM exhibited a consistent upward trend. In patient subgroups aged ≤ 50, 51–60, 61–70, and 71–80 years, CVM had the lowest CM rate of all COD throughout the entire follow-up period. The CM rate among patients over 80 exceeded ONCM and ranked third throughout most of the follow-up period. The lowest CM rate of CVM was in duodenal, jejunal, or unspecified adenocarcinoma patients throughout the entire follow-up period. CVM had a higher CM rate than ONCM for ileal adenocarcinoma patients during part of the follow-up period. Multivariate competing risk model identified age, marital status, calendar year of diagnosis, differentiation degree, SEER stage, and chemotherapy status as independent risk factors for CVM. These findings emphasize the importance of early screening for CVD in patients based on these risk factors. Timely interventions to manage adjustable risk factors could effectively prevent CVM events in these patients.