Long-term cardiovascular risk in survivors of hematologic malignancies: a meta-analysis Authors: Siqing Liu, Shuran Liu, Yining Hu, Jiawen Deng, Guoli Sun, Xuexian Fang Published: 2026 ABSTRACT AbstractBackgroundThe survival rate for hematologic malignancies is steadily improving. With the encouraging trend and the aging of these survivors, there is an ever-increasing responsibility for identifying adverse cardiovascular outcomes associated with carcinogenesis and/or anti-cancer therapies across the span of their lives. However, prospective studies have yielded inconsistent results.MethodsWe performed a meta-analysis to summarize the evidence regarding the association between leukemia, Hodgkin lymphoma (HL), and non-Hodgkin lymphoma (NHL) in relation to later risk of cardiovascular disease (CVD). Relevant literatures were identified from PubMed and Web of Science databases through May 2024 and reviewing reference lists of retrieved articles.ResultsFrom the 2412 studies screened, we included 50 studies with 550,385 participants, spanning 1940–2020 for leukemia and 1940–2016 for both Hodgkin and non-Hodgkin lymphoma. The pooled fully adjusted RRs of total CVD, ischemic heart disease (IHD), heart failure (HF), and stroke associated with leukemia were 2.68 (95% confidence interval: 2.22–3.25), 1.46 (1.25–1.70), 4.13 (2.41–7.06), and 3.05 (2.16–4.32), respectively. For HL or NHL survivors, similar positive associations were observed in our study. In addition, compared with survivors with chemotherapy (2.12; 1.89–2.38), radiotherapy-treated survivors had a greater risk of CVD (4.06; 2.84–5.81).ConclusionsIn summary, survivors of hematologic malignancies, especially those who receive radiotherapy, are at highest risk of developing CVD, emphasizing the need for targeted prevention strategies.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1186/s12885-026-16294-7.Keywords: Leukemia, Hodgkin lymphoma, Non-Hodgkin lymphoma, Cardiovascular disease, Meta-analysis Background Hematologic malignancies are classified into multiple biologic subtypes, consisting of leukemia, Hodgkin lymphoma (HL), and non-Hodgkin lymphoma (NHL). These myeloid and lymphatic tumors make up a significant portion of the global disease burden, accounting for approximately 6.6% of all cancer around the world [1, 2]. Over the past decades, advances in early diagnosis and therapy have substantially improved the cure rate of hematologic malignancies, resulting in a growing population of survivors [3–5]. However, a significant proportion of these long-term survivors may develop a second malignancy and/or non-neoplastic complications, among which cardiovascular disease (CVD) is the most common and serious one [6]. CVD and hematologic malignancies share many environmental and lifestyle factors, such as smoking, alcohol abuse, low physical activity, obesity, and chronic inflammation, suggesting that patients with hematologic malignancy may develop a high-risk population for CVD [7]. In addition, cancer therapies, such as chemotherapy (CT) and radiotherapy (RT), are associated with greater CVD risk during treatment and/or in survivorship [8]. For example, anthracyclines, which comprise a sort of drugs including doxorubicin, daunorubicin, epirubicin, and idarubicin, are the cornerstone of CT for leukemia and other cancers; however, their clinical use is largely limited by a cumulative and irreversible cardiotoxicity [9, 10]. RT-related cardiovascular adverse events include atherosclerosis, thromboembolism, myocardial infarction (MI), and stroke [11]. In recent years, increasing evidence has suggested a greater risk of different CVD subtypes, such as ischemic heart disease (IHD), heart failure (HF), and stroke, in survivors of hematologic malignancies. There are, however, inconsistencies within the literature due to differences in sample size, ethnicity, and other characteristics. Here, we conducted a comprehensive meta-analysis of population-based cohort studies to thoroughly investigate the associations of CVD risk with different hematologic malignancies. In addition, we also examined and compared the distinct effects of therapeutic regimes on cardiovascular outcomes, in order to help patients optimize the selection of primary treatment. Methods The present meta-analysis was designed, executed, and analyzed in accordance with the Meta-analyses of Observational Studies in Epidemiology (MOOSE) protocol and was reported following Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) guidelines [12, 13]. The study was prospectively registered at the Prospective Register of Systematic Reviews (PROSPERO registration number: CRD42024581433). Search strategy We systematically searched the databases PubMed and Web of Science for cohort studies published through May 31, 2025. The following keywords were used in the literature search: (leukemia OR lymphoma OR Hodgkin OR non-Hodgkin OR hematologic malignancies) AND ("cardiovascular disease" OR "heart failure" OR "stroke" OR "coronary heart disease" OR "myocardial infarction") AND (cohort OR follow-up OR longitudinal OR population). It is important to note that we found due to limited research data on multiple myeloma after searching, we excluded it from our study design and did not include it in our final analysis. Furthermore, all individual plasma cell disorders have been excluded from this study. We restricted our search to studies conducted in humans without language restriction. For any additional studies, a review of the references cited in the relevant reports was also conducted. Study selection Our current meta-analysis included articles that met the following five criteria: (1) they used prospective or retrospective population-based cohort study design; (2) participants were diagnosed with hematologic malignancies, including leukemia, HL and NHL, without CVD history or diagnosis. We categorized the disease populations as follows: the leukemia group included acute lymphoblastic leukemia (ALL), acute myeloid leukemia (AML), chronic myeloid leukemia (CML), and other subtypes; the non-Hodgkin lymphoma (NHL) group comprised diffuse large B-cell lymphoma (DLBCL), follicular lymphoma (FL), chronic lymphocytic leukemia (CLL), and other variants. We explicitly excluded myelodysplastic syndromes (MDS) and plasma cell disorders. Control populations were selected from three sources: the general population, matched healthy controls, and patients with non-hematological malignancies. The outcome was the risk of CVD, including but not limited to IHD, MI, HF and stroke; (5) the authors presented risk estimates with 95% confidence intervals (95% CIs). We defined a follow-up duration of ≥ 5 years as "long-term". Our primary effect variable was the type of hematological malignancy, and the primary outcome was the occurrence of cardiovascular disease. During the literature screening process, we excluded biological studies based on animals or cells, non-original research (e.g. editorials, commentaries, or reviews), abstracts without adequate data, and duplicated studies. Additionally, we did not impose any restrictions on the study time range or whether the study population was derived from clinical trials. To ensure that all eligible studies were correctly identified, we employed a two-step selection process. Two independent investigators (S.L. and Y.H.) performed the literature search and initial screening of all titles and/or abstracts. Then the full text of every potentially relevant article was examined. A third investigator (X.F.) was consulted to resolve discrepancies. Data extraction Data extraction was conducted independently by four investigators (S.L., S.L., Y.H., and J.D.) using a standardized data extraction form. A third investigators (X.F.) participated in group discussions to resolve disagreements. From each included article, we collected the following information: the first author’s name, year of publication, country where the study was conducted, name of the study (if applicable), study design (prospective or retrospective), participant age at baseline, follow-up duration, gender distribution, sample size (the number of cases and total participants), classification of hematological malignancies, and corresponding risk estimates of CVD with 95% CIs. The risk estimates with the most adjustments were recorded for further analysis. Quality assessment was conducted by the Newcastle–Ottawa scale [14]. Each study is awarded nine points on this scale: four for participant selection and measurement of exposure, two for comparability, and three for outcome evaluation and follow-up adequacy. The aggregate scores 0–3, 4–6, and 7–9 represent low, moderate, and high quality, respectively. Statistical analysis In this meta-analysis, relative risk (RR) with 95% CIs was adopted as the common measure of associations in all included studies; where necessary, hazard ratio (HR) and standardized incidence/mortality ratio (SIR/SMR) were used in lieu of RR. For studies with stratified risk estimates, we used a fixed-effects model to pool the estimates to obtain a study-specific estimate, assuming an internal consistency within a single study. The combined RRs and 95% CIs in the meta-analysis were computed by an inverse-variance-weighted random-effects model. Heterogeneity across studies was assessed using the I2 metric, with values of 25%, 50%, and 75% representing low, moderate, and high heterogeneity, respectively [15]. For sensitivity analysis, we conducted subgroup analyses on potential sources of heterogeneity, including gender, geographic location, follow-up duration, leukemia subtype, and different therapeutic regimes. Publication bias was determined by Egger’s linear regression test and Begg’s rank correlation test [16, 17]. All statistical analyses in the study were conducted by Stata version 17 (StataCorp LLC). Differences with a P value of less than 0.05 were considered significant unless otherwise noted. Results Literature search and study characteristics In total, 2412 articles were retrieved from the databases after duplicates were removed. We excluded 2262 articles based on titles and abstracts and checked the full texts of remaining 114 studies. Following the inclusion and exclusion criteria, 71 articles were excluded, and 7 additional articles were added after reference reviews (Supplementary Fig. 1). Finally, 50 unique studies, containing 550,385 participants in total, were selected for the present meta-analysis, which we found are all retrospective cohort study. Supplementary Tables 1 and 2 summarized the detailed characteristics regarding all included studies of our meta-analysis. The year of publication ranged from 1982 to 2024. For study periods, 6 studies reported treatment years, 3 reported data collection years, 1 reported inpatient years, and the others reported the years of diagnosis. In total, 44 studies covered periods longer than 10 years. Overall, the study time span was 1940–2020 for leukemia and 1940–2016 for both HL and NHL. In terms of study regions, 20 studies were conducted in the Americas (19 in US, 1 in Canada), 26 studies were conducted in Europe, 1 in Asia, 1 in Oceania, and 2 cross-regional collaborative studies (one between the United States and Qatar, and the other between the US and the UK). In total, 23 articles involved CVD; 20 articles involved Coronary heart disease (CHD), Ischemic heart disease (IHD)); 24 articles involved Myocardial infarction (MI); 33 articles involved Heart failure (HF); and 38 articles involved stroke and cerebrovascular diseases. 24 studies included a comparison of leukemia survivors versus healthy populations; 32 and 27 investigated the cardiac risk in survivors of HL and NHL, respectively. The mean or median follow-up durations ranged from 3.1 to 24 years. Among these studies, all had a follow-up duration of > 1 year, and 39 (78%) had a median or mean follow-up duration of ≥ 5 years. Five studies (10%) either did not report the follow-up duration or had an undetermined specific follow-up duration even after we carefully reviewed the literature information. The quality scores for the studies ranged from 6 to 9, with an average score of 8 (Supplementary Table 3). Regarding age distribution, patients with ALL tend to be relatively young, primarily consisting of children, adolescents, and young adults, with a diagnosis age of less than 16 years (Supplementary Table 4). In contrast, patients with DLBCL are predominantly elderly, with a median diagnosis age of 75 years. Although age definitions reported in different studies varied (Supplementary Table 5), we can conclude that the three types of hematological malignancies exhibit differences in their age distribution data, which may affect the assessment of cardiovascular disease risk. Associations between leukemia and CVD Risk Twenty-three studies investigated the risk of total CVD, IHD, HF, and stroke among leukemia survivors (Fig. 1). The pooled results suggest significantly elevated CVD risk in leukemia survivors (RR: 2.68; 95%CI: 2.22–3.25); however, we found evidence of between-study heterogeneity (I2 = 80.0%). For subtypes of CVD, leukemia survivors had a 46% higher risk of IHD (RR: 1.46; 95%CI: 1.25–1.70) with moderate heterogeneity (I2 = 60.0%). The pooled RRs comparing leukemia survivors versus healthy populations were 4.13 (95%CI: 2.41–7.06; I2 = 96.6%) for HF risk and 3.05 (95%CI: 2.16–4.32; I2 = 95.9%) for stroke risk. Forest plots for risk of total CVD, IHD, heart failure, and stroke in leukemia survivors. *Risk estimate by fixed-effects analysis Associations between HL and CVD risk Twenty-nine studies examined the risk of total CVD, IHD, HF, and stroke among HL survivors (Fig. 2). On average, HL survivors experience CVD at a rate 4.04 times higher than that of the general population (95% CI: 3.11–5.23). In addition, we found considerable degree of heterogeneity in the result (I2 = 97.8%). With respect to specific types of CVD, the pooled RR of HL survivors for IHD risk was 3.84 (95% CI: 2.92–5.06), also with high heterogeneity (I2 = 97.8%). Similarly, significantly elevated HF risk was observed in HL survivors (RR: 5.87; 95%CI: 3.69–9.32; I2 = 97.4%). Compared to others, stroke poses the least health threat to HL survivors (RR: 1.99; 95%CI: 1.57–2.52; I2 = 87.9%). Forest plots for risk of total CVD, IHD, heart failure, and stroke in HL survivors. *Risk estimate by fixed-effects analysis Associations between NHL and CVD risk Twenty-six studies investigated the risk of total CVD, IHD, HF, and stroke among NHL survivors (Fig. 3). The pooled RR of total CVD in these survivors was found to be 2.39 (95%CI: 1.05–5.44) and there was a high level of heterogeneity among the studies (I2 = 99.8%). The cardiovascular risks for NHL survivors varied by specific types, with HF presenting the most significant threat. The pooled RR for HF was 2.84 (95%CI: 1.90–4.24) and it also exhibited high heterogeneity (I2 = 95.8%). In contract, NHL survivors had a 18% higher risk for IHD (RR: 1.18; 95%CI: 1.01–1.38; I2 = 81.4%) and a 35% higher risk for stroke (RR: 1.35; 95%CI: 1.11–1.64; I2 = 83.8%). Forest plots for risk of total CVD, IHD, heart failure, and stroke in NHL survivors. *Risk estimate by fixed-effects analysis Subgroup analyses Current study revealed high heterogeneity for most associations; thus we performed subgroup analyses stratified by gender, geographic location, follow-up duration, and leukemia subtype to assess the consistency of the primary results (Table 1). Our analysis revealed that the association between leukemia and the risk of CVD was not substantially changed by the classification of ALL and AML (P = 0.754). Notably, leukemia survivors from North America (RR: 2.98; 95%CI: 2.30–3.87; I2 = 84.2%) or Europe (RR: 2.71; 95%CI: 2.13–3.45; I2 = 95.9%) showed a significant correlation to elevated CVD risk, whereas such an association was not observed in other regions (RR: 1.22; 95%CI: 0.89–1.66; I2 = 25.9%). The associations were more evident in studies with > 10 years follow-up compared with studies with ≤ 10 years follow-up among all included survivors of HL. Furthermore, the association was more pronounced in studies with follow-up of 10 years or longer (RR: 3.80; 95%CI: 2.98–4.86; I2 = 98.4%) compared to those with follow-up periods less than 10 years (RR: 2.33; 95%CI: 1.68–3.22; I2 = 93.3%) among HL survivors (P = 0.018). Subgroup analysis stratified by gender was not performed on leukemia because of limited data. But in NHL survivors, the association for CVD risk was more robust among female (RR: 2.17; 95%CI: 1.83–2.57; I2 = 6.50%) than male (RR: 1.67; 95%CI: 1.46–1.91; I2 = 15.9%) population (P = 0.018). Table 1Subgroup analysesNRR (95%CI)I2 (%)P-valueLeukemiaDisease types ALL53.46 (2.11–5.68)73.90.754 AML73.91 (2.20–6.95)80.3Location North America92.98 (2.30–3.87)84.2< 0.0001 Europe122.71 (2.13–3.45)95.9 Other areas31.22 (0.89–1.66)25.9Follow up ≤ 10 years52.29 (1.35, 3.88)97.10.437 > 10 years132.85 (2.37, 3.45)92.8Hodgkin lymphomaSex Male83.62 (2.70–4.84)93.80.685 Female84.04 (2.59–6.28)92.9Location North America114.57 (3.28–6.36)97.10.063 Europe183.05 (2.35–3.97)98.7Follow up ≤ 10 years52.33 (1.68, 3.22)93.30.018 > 10 years213.80 (2.98, 4.86)98.4Non-Hodgkin lymphomaSex Male31.67 (1.46–1.91)15.90.018 Female32.17 (1.83–2.57)6.50Location North America102.52 (0.95–6.65)99.80.422 Europe151.68 (1.46–1.95)95.3Follow up ≤ 10 years121.89 (0.77, 4.62)99.80.957 > 10 years101.94 (1.60, 2.34)95.8 Subgroup analyses Comparison of cardiotoxicity between CT and RT Next, we examined the cardiovascular risks associated with different treatment regimens for patients with hematologic malignancies (Fig. 4, Supplementary Tables 6,7). Our findings revealed that patients undergoing RT alone faced a risk of cardiotoxicity more than four times higher than those without RT (RR: 4.06; 95%CI: 2.84–5.81; I2 = 93.1%). Patients treated with CT alone also had an increased risk of developing CVD (RR: 2.12; 95%CI: 1.89–2.38; I2 = 28.6%), but the risk was relatively lower than that associated with RT. However, when CT was combined with RT, the risk of CVD significantly increased (RR: 4.05; 95%CI: 2.99–5.49; I2 = 93.9%). These results indicate that cardiotoxicity from RT plays a more significant role in the increased CVD risk among survivors of hematologic malignancies. Cardiovascular risk by treatment regimens in survivors of hematologic malignancies. *Risk estimate by fixed-effects analysis Publication bias With regard to leukemia or HL and CVD and its subtypes, either Egger's linear regression test or Begg's rank correlation test did not reveal any evidence of publication bias. However, there was evidence of publication bias for studies reporting on the correlation between NHL and total CVD risk (Supplementary Table 8). Discussion This systemic meta-analysis, based on 50 cohort studies from 14 countries, provides the most recent epidemiological evidence demonstrating that survivors of hematologic malignancies, including leukemia, HL, and NHL, face increased CVD risks in the years following their diagnosis. These findings align with previous research indicating that tumor and CVD share many risk factors, such as inflammation and oxidative stress [18, 19]. Our analysis of treatment regimens uncovered that survivors who received RT alone had a greater risk of CVD compared with those who received CT alone. Existing literature highlights the prevalence of cardiovascular complications following high-dose mediastinal radiation therapy [20–24]. This increased risk is likely due to radiation's impact on the pericardium, myocardium, or coronary vessels, potentially causing fibrosis or damage to small vessels [21]. In addition, Vakilpour et al. reported that cardiovascular toxicity from cancer treatments may be influenced by a combination of patient characteristics, treatment type, dosage, and duration. And HL patients who undergo mediastinal radiotherapy have a heightened risk of developing IHD. This risk is further exacerbated by factors such as radiation dose, age at radiotherapy (particularly < 25 years), treatment duration, type of radiation source, and concurrent risk factors like smoking [25]. Furthermore, radiation therapy can lead to a condition known as radiation-related heart disease (RAHD), which invades all layers of the heart and presents differently from traditional cardiovascular diseases. RAHD includes IHD, pericardial disease, valve disease, and cardiomyopathy, collectively termed 'pan-cardiac' involvement [26]. This complexity partly stems from the use of outdated radiation therapy equipment and technology [21]. Advancements in radiotherapy technology, such as intensity-modulated radiation therapy, deep inspiratory breath holding, respiratory gating techniques, and proton therapy, show promise in reducing cardiac radiation exposure [27]. These modern techniques may lower both the incidence and mortality rates of heart disease, thus helping to mitigate future CVD complications [22, 28]. Existing literature suggests that the shift from traditional radiotherapy techniques (e.g., IFRT) to modern approaches (e.g., IMRT, ISRT) may reduce cardiac toxicity in hematological malignancy patients by improving targeting precision and lowering radiation doses to cardiac substructures. Future studies reporting specific XRT modalities are needed for robust inferences. Beyond technical advances, minimizing valvular risk requires strict adherence to dose-heart-volume constraints, as mean doses > 25–30 Gy to the aortic or mitral valves significantly predict dysfunction [29]. Furthermore, for advanced HL patients with large masses achieving a complete metabolic response after ABVD, the GITIL/FIL HD0607 trial supports safely omitting consolidation radiotherapy to eliminate this source of cardiac exposure [30]. Together, these complementary strategies—precise dosimetric limits and response-adapted omission—are essential to mitigate the long-term cardiovascular burden of mediastinal irradiation. Numerous studies have demonstrated that patients with leukemia and lymphoma who receive CT have an increased risk of CVD [31, 32]. Anthracyclines are currently among the chemotherapy drugs with the highest risk of cardiac toxicity [33]. HF is one of the most serious late-stage complications of anthracycline drug therapy for cancer [34–37]. Although the precise mechanisms of cardiac toxicity from anthracyclines are not fully understood, oxidative stress is believed to play a significant role. Research indicates that dexrazoxane is effective and safe in significantly reducing the risk of cardiac toxicity [38–40]. The degree of toxicity varies among different chemotherapy drugs, and radiation-induced cardiac toxicity is dose-dependent. Thus, grouping all chemotherapy drugs and radiation treatments together, without considering their type and dosage, may lead to an underestimation of their specific risks. Additionally, the stress, anxiety, and depression associated with cancer diagnosis and treatment may further increase the risk of CVD in survivors [41]. This aspect requires further investigation to better understand its impact. Furthermore, hematopoietic stem cell transplantation (HSCT), while a potentially curative treatment for aggressive hematological malignancies, is significantly limited by its association with elevated cardiovascular risk. Survivors face a heightened burden of cardiovascular diseases—such as cardiomyopathy, heart failure, and arrhythmias—and increased CV-related mortality [42–45]. These complications are influenced by cumulative cardiotoxic exposures, including prior anthracyclines, conditioning regimens, and comorbidities [46]. Most evidence remains retrospective and adult-focused, underscoring the need for prospective studies across all ages to clarify HSCT-specific cardiovascular risks and guide preventive care. Given the substantial long-term cardiovascular risks identified in our meta-analysis, the establishment of dedicated cardio-oncology units is imperative for the optimal management of survivors of hematologic malignancies. Sensitive echocardiographic techniques, particularly the assessment of left ventricular global longitudinal strain (GLS), can detect subclinical dysfunction [47]. As emphasized by recent guidelines and real-world evidence, a multidisciplinary approach centered on these units significantly improves outcomes through strict surveillance and early intervention [48, 49]. Crucially, primary prevention is enhanced by adopting less cardiotoxic frontline regimens. Evidence supports substituting conventional doxorubicin with non-pegylated liposomal doxorubicin in the MBVD schedule, which maintains robust antineoplastic activity while significantly reducing cardiac events in high-risk and elderly patients [50, 51]. Thus, integrating these advanced monitoring techniques, early pharmacological interventions, and optimized treatment protocols within specialized cardio-oncology frameworks is essential to safeguard the long-term cardiovascular health of hematologic malignancy survivors. Due to limitations in the available research data, we did not perform specific subgroup analyses based on race. However, a previous study found that Black children and adolescents/young adults with HL and NHL faced an increased risk of CVD mortality compared with white individuals. Racial differences in CVD risk were not observed in leukemia. When analyzing by age, the study revealed no racial difference in CVD mortality risk among Black leukemia and NHL survivors compared to white individuals aged 0–14 at diagnosis. However, the data for HL survivors were insufficient for analysis. Conversely, for individuals diagnosed with any of the three types of cancer between the ages of 15–34, racial differences in CVD risk were evident. In summary, Black survivors of blood cancers, particularly children and young adults, are at a higher risk of CVD-related mortality compared to their white counterparts. This increased risk may be attributed to differences in treatment-related factors, socioeconomic status, behavioral risk factors such as obesity and inactivity, and follow-up care [52]. While our findings reinforce established evidence on cardiovascular risk among hematologic malignancy survivors from recent studies [53] and other cohort studies [54–56], they also provide new insights through. Consistent with the findings of Yong et al. and Paterson et al. our results underscore the potential cardiotoxic effects of cancer treatments such as anthracyclines and chest radiotherapy. However, while we observed significant associations of total cardiovascular diseases, ischemic heart disease, heart failure, and stroke with leukemia, HL, and NHL, Yong et al. found no significant increase in stroke risk among NHL patients, and Florido et al. reported no significant rise in coronary heart disease among cancer survivors. Such discrepancies may reflect differences in study populations, data definitions, treatment protocols and eras, as well as statistical approaches. Our research is a comprehensive effort in integrating diverse populations, including various ethnicities, age groups, and demographics, to examine the association between hematologic malignancies and cardiovascular risk, which differentiates it from previous single-cohort studies. To our knowledge, this is the first meta-analysis offering a comprehensive view of this relationship through an extensive literature review, with results showing a consistent trend across most studies. Although prior studies, as noted in Yong et al., have already explored this domain, our study benefits from the following aspects of its design: it included a larger number of studies, analyzed a greater cohort of cancer patients, utilized a more recent search cutoff date, and conducted more comprehensive specific subgroup analyses. We employed advanced statistical methods, such as the random-effects model, to account for variability among studies. Beyond identifying cardiovascular risks, our study tested tailored treatment protocols in subgroup analyses, revealing significant differences in responses and outcomes. This stratified approach aimed to minimize potential adverse effects. Overall, our findings provide new insights into managing cardiovascular disease risk in patients with hematologic malignancies, with significant clinical implications. Despite these strengths, several limitations should be noted. Most included studies were retrospective cohorts, which may introduce residual confounding and bias, and reliance on registry or electronic health record data could further lead to inaccuracies. Many studies spanned more than 20 years and lacked detailed information on treatment modalities, particularly radiotherapy techniques (e.g., IFRT, ISRT, IMRT), limiting our ability to assess the impact of evolving therapies. Heterogeneity across studies also remained unexplained despite subgroup analyses, likely reflecting differences in age distributions, disease stage, prior HSCT, and treatment protocols. Publication bias, especially among NHL survivors, and limited follow-up in some studies may have restricted the evaluation of long-term cardiovascular outcomes. Finally, the approximation of effect estimates to RRs and the absence of absolute risk measures in most studies may have minor effects on data interpretation. Conclusions In conclusion, our study provides substantial evidence linking hematologic malignancies to an increased risk of CVD. We also found that treatment regimens, gender, regional differences, and follow-up age may influence CVD risk in specific hematologic tumors. However, the potential for publication bias and significant variability in the data underscore the need for careful interpretation of these findings. Further research is required to confirm these associations. Our research raises awareness among cancer survivors about their risk of heart disease and encourages healthcare providers to emphasize preventive measures. Supplementary Information Supplementary Material 1. Authors’ contributions S.L. and X.F. designed the study; S.L., Y.H., and J.D. conducted literature search and data extraction. S.L. performed statistics. S.L. and X.F. wrote the manuscript. G.S. revised the manuscript. All authors read and approved the final version of the manuscript. Funding This study was supported by research grants from the National Natural Science Foundation of China (82170644 to X.F.) and the Natural Science Foundation of Zhejiang Province (LR23H020001 to X.F.). Data availability Not applicable to the meta-analysis of published studies. Declarations Ethics approval and consent to participate Not applicable. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Contributor Information Guoli Sun, Email: sunguoli2019@outlook.com. Xuexian Fang, Email: xfang@hznu.edu.cn. References Associated Data Supplementary Materials Supplementary Material 1. Data Availability Statement Not applicable to the meta-analysis of published studies.