Five-year retrospective analysis of invasive vs. conservative strategies for non-ST elevation myocardial infarction in critically ill cancer patients

Introduction

Background

Cancer and coronary artery disease are two of the leading causes of mortality worldwide (1). Patients with cancer are at increased risk for coronary artery disease because of shared risk factors and cancer-induced proinflammatory and prothrombotic states (2). Additionally, chemotherapy, targeted therapies, and chest radiation lead to acute and chronic coronary damage (3). The incidence of acute coronary syndrome (ACS) in cancer patients is twice as high as that in the general population and is significantly greater in the first 6 months after diagnosis and in advanced cancer stages (4,5). Among all cardiovascular comorbidities, non-ST segment elevation myocardial infarction (NSTE-MI) is recognized as one of the most critical conditions in cancer patients (6).

In some studies, cancer patients who develop ACS are associated with greater complications, higher in-hospital mortality rates, and lower percutaneous coronary intervention (PCI) rates than their non-cancer counterparts (7). Retrospective data have shown that interventional treatment strategies have better outcomes for patients with cancer who have ST-segment elevation myocardial infarction (STEMI) (2). However, in patients with advanced cancer and NSTE-MI, PCI has not been shown to have a significant mortality benefit over pharmacological treatment alone. The 2022 European Society of Cardiology guidelines in Cardio-Oncology suggest an invasive approach for the management of STEMI in cancer patients and high-risk NSTE-MI patients with a survival period of more than 6 months. A noninvasive strategy can be acceptable, however, for low-risk NSTE-MI patients who lack persistent ischemia or hemodynamic instability, especially when their personal cancer prognosis is less than 6 months (2).

Critically ill cancer patients admitted to the intensive care unit (ICU) represent a distinct subgroup because they often present with hemodynamic instability, multiorgan dysfunction, frequent thrombocytopenia, and complex comorbidities. These factors increase ischemic and bleeding risks, complicate pharmacologic and interventional management, and make decision-making particularly challenging compared with non-ICU or non-cancer NSTE-MI populations.

Despite these guidelines, the management of ACS in cancer patients can be particularly challenging because of factors such as vulnerability, increased bleeding and thrombosis risk, thrombocytopenia, polypharmacy, associated comorbidities and the complexity of the overall clinical situation, and the requirement for surgical procedures, introduce additional levels of complexity in treatment planning (2). The complexity of making treatment decisions in such cases is heightened by the absence of a clear consensus on whether a conservative or invasive approach is superior (4). On the other hand, in ICU patients, the clinical features are more complex, and decision-making is more difficult.

Rationale and knowledge gap

To date, no study has specifically compared interventional versus pharmacological management of NSTE-MI in oncology patients admitted to the ICU. This lack of evidence for this particular group of patients represents an important gap in clinical practice, especially in ICU settings. While research has previously investigated ACS outcomes in cancer patients, studies exclusive to ICU cancer patients are limited.

Objective

The primary objective of this study was to describe the demographic, clinical, and laboratory characteristics of critically ill cancer patients with NSTE-MI and to identify factors associated with the selection of interventional versus pharmacological treatment strategies in the ICU. We hypothesized that, in this critically ill population, interventional treatment might be associated with lower in-hospital mortality compared with pharmacological management, while acknowledging that baseline clinical differences could influence outcomes. In-hospital mortality was evaluated as a secondary, exploratory outcome to provide additional insight into the potential prognostic implications of each strategy. By focusing on this less well-studied subgroup, our aim was to generate clinically relevant descriptive data that could support the development of more individualized treatment approaches for this high-risk population. We present this article in accordance with the STROBE reporting checklist (available at https://cdt.amegroups.com/article/view/10.21037/cdt-2025-525/rc).

Methods

This retrospective cohort study evaluated the interventional and pharmacological treatment preferences in cancer patients who developed NSTE-MI during their ICU stay. The study assessed patient characteristics, treatment selection patterns, and in-hospital mortality.

This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of Dr. Abdurrahman Yurtaslan Ankara Oncology Training and Research Hospital (No. 2024–10/151). As this was a retrospective study, the requirement for informed consent was formally waived by the Ethics Committee. The study was registered at ClinicalTrials.gov (NCT06702436) on November 19, 2024.

Medical records were retrospectively reviewed for patients admitted to the ICU of the Dr. Abdurrahman Yurtaslan Ankara Oncology Training and Research Hospital between January 1, 2020, and December 31, 2024. Patients aged ≥18 years, diagnosed with NSTE-MI, and with either an active cancer diagnosis or a history of cancer were eligible for inclusion. The diagnosis of NSTE-MI was identified based on International Classification of Diseases, 10th Revision (ICD-10) codes and confirmed by electrocardiography, cardiac biomarker analysis, and clinical evaluation, in accordance with contemporary European Society of Cardiology guidelines for the management of non-ST-elevation ACS (8). In-hospital mortality was defined as death occurring during the index hospitalization for NSTE-MI, regardless of the underlying cause.

Among the patients included in the study, those whose clinical data were incomplete and those diagnosed with STEMI were excluded from the study. Patients were excluded if they had: (I) STEMI; (II) incomplete clinical, laboratory, or diagnostic data required for primary analyses; (III) myocarditis; or (IV) type 2 myocardial infarction (MI) secondary to non-coronary causes (e.g., sepsis-induced hypoperfusion). The patient selection process and exclusion criteria are summarized in Figure 1.

Figure 1 Flowchart of the study. Patients were categorized into two groups based on initial treatment decisions: pharmacological treatment (managed with medical therapy alone, without undergoing coronary angiography) and interventional treatment (referred for coronary angiography after evaluation by the invasive cardiology team). Among patients in the interventional group, coronary angiography was performed in 25 cases, and PCI was carried out in 12 of these patients. Twelve patients were evaluated by the interventional cardiology team but did not undergo coronary angiography: these patients were analyzed within the interventional treatment group, in line with the initial treatment decision. CAG, coronary angiography; MI, myocardial infarction; NSTE-MI, non-ST segment elevation myocardial infarction; PCI, percutaneous coronary intervention.

All patients were managed in a general medical-surgical ICU at Dr. Abdurrahman Yurtaslan Ankara Oncology Training and Research Hospital. The hospital does not have dedicated cardiac ICU. However, 24-hour cardiology consultation was available for all patients. Decisions regarding interventional therapy were made by the cardiology team at the same institution. Patients selected for PCI were transferred to an external tertiary care center with a full interventional cardiology team for the procedure and further management. The tertiary interventional cardiology team performed a secondary evaluation and proceeded with PCI if deemed appropriate.

Patients were divided into two groups according to the initial treatment decision. The pharmacological treatment group consisted of patients for whom a medical therapy strategy was initially selected, meaning that no coronary angiography (CAG) was planned as part of their initial management. The interventional treatment group consisted of patients for whom an initial interventional strategy was selected, meaning that they were referred to for CAG and evaluated by the invasive cardiology team. Treatment allocation was based on clinician judgment (cardiology + ICU team) and not on any predefined study protocol. Decisions incorporated hemodynamic status, bleeding/thrombosis risk, overall cancer prognosis, and patient-specific clinical factors. Some patients initially assigned to the interventional treatment strategy ultimately received pharmacological therapy after reassessment at the tertiary center; however, group allocation remained based on the initial treatment decision. In our study, CAG was performed promptly after the diagnosis of ACS. The invasive cardiology team assessed patients immediately following the ACS diagnosis. Patients deemed suitable for invasive evaluation were transferred to a specialized center for CAG and potential PCI, while those considered inappropriate candidates continued with medical therapy. Angiography was generally conducted without unnecessary delay to enable timely interventional treatment. Not all patients scheduled for interventional treatment underwent PCI; within this group, CAG was performed in 25 patients, and PCI was carried out in 12 of these patients. Twelve patients were evaluated by the interventional cardiology team but did not undergo CAG; they were not included in the pharmacological treatment group, as analyses were based on initial treatment decisions.

Data collection and evaluation

Patient data were obtained retrospectively from the electronic medical records system. Demographic data, medical history, laboratory results, hemodynamic parameters, and treatment data were recorded. Hemodynamic instability was defined as mean arterial blood pressure <65 mmHg, the need for vasopressors or clinical signs of end-organ hypoperfusion. Metastatic disease was defined as radiologically or histopathological confirmed distant metastasis documented in oncology records. Age, sex, body mass index (BMI), existing comorbidities [including hypertension, diabetes mellitus, chronic renal failure, coronary artery disease, stroke, chronic obstructive pulmonary disease (COPD), and prior coronary artery bypass graft (CABG) surgery], presence of metastasis, cancer types, and oncology-related treatments—including chemotherapy, radiotherapy, and immunotherapy—were evaluated based on medical records.

Among the included patients, 35 were receiving active chemotherapy and 7 were undergoing radiotherapy at the time of ICU admission. Missing data were handled by excluding patients from specific analyses when relevant information was not available. For example, two patients lacked complete data following referral for CAG; these patients were excluded from analyses that required the missing information. Exclusion of these patients did not materially affect the overall results. Data on dyslipidemia were not available in the patient records and therefore were not included in the analysis.

Surgical history within 1 week prior to ICU admission and NSTE-MI diagnosis was specifically recorded. Clinical parameters, including heart rate and systolic blood pressure, as well as laboratory parameters—serum creatinine, troponin I levels measured at the time of ACS diagnosis and serially thereafter to determine the peak value, C-reactive protein (CRP), white blood cell count, hemoglobin, and hematocrit values were collected from blood samples obtained at the time of ACS diagnosis. Ejection fraction and wall motion abnormalities were obtained from echocardiography reports performed at the time of ACS diagnosis. Vasopressor and inotropic drug use data were retrieved from ICU records.

Medication data, including use of oral antidiabetics, insulin, antihypertensives, bronchodilators, antiplatelet agents, anticoagulants, beta blockers, angiotensin converting enzyme inhibitors/angiotensin receptor blockers (ACE-I/ARBs), statins, and other medications not otherwise specified (categorized as “Others”), were recorded. Antiplatelet and anticoagulant drug use specifically after NSTE-MI diagnosis was also evaluated. Anemia was defined according to standard clinical thresholds: hemoglobin <13 g/dL for males and <12 g/dL for females.

In this study, factors influencing the preference for interventional versus pharmacological treatment were considered the primary endpoint. In-hospital mortality was evaluated as the secondary endpoint, and mortality rates were compared between treatment groups.

Global Registry of Acute Coronary Events (GRACE) score and risk assessment

The GRACE score was calculated to estimate the mortality risk of the patients included in the study. This scoring system is commonly used to predict in-hospital and 6-month mortality in patients with ACS. However, previous studies have indicated that its predictive power is limited in cancer patients. In our study, only in-hospital all-cause mortality was evaluated, and the long-term predictive power of the GRACE score was not analyzed.

Statistical analysis

The Statistical Package for the Social Sciences (SPSS) statistical software was used for data analysis. Continuous variables were assessed for normality using the Kolmogorov-Smirnov or Shapiro-Wilk tests and analyzed via appropriate statistical methods. Normally distributed variables are presented as mean ± standard deviation (mean ± SD) and were compared using the Student’s t-test. Variables that were not normally distributed are expressed as median with interquartile range (IQR) and were compared using the Mann-Whitney U test. Categorical variables are reported as frequencies and percentages (%) and were compared using the Pearson Chi-squared test or Fisher’s exact test, where appropriate.

Clinical outcomes (in-hospital mortality) were compared between the pharmacological and interventional treatment groups using univariate methods (e.g., Chi-squared or Fisher’s exact tests).

Additionally, Kaplan-Meier survival analysis was performed to assess time-to-event data for in-hospital mortality, and the difference between treatment groups was evaluated using the log-rank test. Median survival times and 95% confidence intervals (CIs) were reported. All patients had complete in-hospital outcome data; however, exact time-to-event data were not available for 3 patients. Sensitivity analyses were limited to the exclusion of patients with incomplete angiography-related data (n=2), and these exclusions did not materially change the results. No additional sensitivity analyses were performed due to the small sample size. Cox proportional hazards modeling was not applied due to the absence of long-term time-to-event data; Kaplan-Meier curves are presented for descriptive purposes only.

To adjust for potential confounding variables, a multivariable logistic regression analysis was performed. Given the retrospective design, no formal sample size or power calculation was conducted; the study sample reflects all eligible patients within the defined period. The relatively small sample size may limit statistical power, particularly for subgroup analyses, and this limitation is acknowledged.

Cox proportional hazards modeling was not applied because the primary outcome, in-hospital mortality, was assessed within a fixed and relatively short hospital stay, and exact time-to-event data were not available or relevant for this cohort. To maintain an appropriate events-per-variable ratio given the number of outcome events, a maximum of five covariates were included in the model. Covariates were selected based on clinical relevance and previous literature and consisted of metastatic disease, GRACE score, serum creatinine level and vasopressor use. A hierarchical (block-wise) approach was applied: Block 1 included treatment type (pharmacological vs. interventional) as the sole predictor, and Block 2 included the remaining covariates. The results are reported as odds ratios (ORs) with 95% CIs. Model fit was evaluated using −2Log Likelihood and Nagelkerke R2, and statistical assumptions were checked prior to analysis. Additionally, the relationship between the GRACE score and mortality was analyzed using Spearman correlation analysis. Statistical significance was defined as P<0.05.

Results

In this study, interventional and pharmacological approaches for the treatment of MI with non-ST segment elevation in oncological patients were evaluated retrospectively. A total of 99 patients were included in the study, with 39 patients (39.4%) referred for interventional treatment and 60 patients (60.6%) managed with pharmacological treatment. The mean age was 66.7±9.6 years, and the mean BMI was 25.9±5.4 kg/m². A total of 70.7% (70/99) of the patients were male, and 29.3% (29/99) were female. Among cancer types, the rate of solid malignancy was 86.9% (86/99), and the rate of hematological malignancy was 13.1% (13/99). The most common types of solid malignancies were lung (13/86, 15.2%), bowel (13/86, 15.2%) and hepatobiliary (12/86, 14.1%) cancers. The demographic data and clinical parameters of the patients included in the study are shown in Table 1.

Patients were categorized into two groups based on the initial treatment decisions: pharmacological treatment (managed with medical therapy alone, without undergoing CAG) and interventional treatment (referred for CAG after evaluation by the invasive cardiology team). Among patients in the interventional group, CAG was performed in 25 cases, and PCI was carried out in 12 of these patients. Twelve patients were evaluated by the interventional cardiology team but did not undergo CAG; these patients were analyzed within the interventional treatment group, in line with the initial treatment decision. When the comorbidities of the patients were examined, hypertension was detected in 31.9% (30/94), diabetes mellitus in 23.4% (22/94), coronary artery disease in 31.9% (30/94) and COPD in 3.2% (3/94). When the interventional and pharmacological treatment groups were compared, although the rates of hypertension and diabetes were higher in the interventional treatment group there was no significant difference between the groups (hypertension: P=0.16, 95% CI: 0.964–5.681, diabetes: P=0.09, 95% CI: 0.856–5.947). When the metastasis status was evaluated, the rate of metastatic disease was 35.9% (n=14/39) in the interventional treatment group and 38.3% (n=23/60) in the pharmacological treatment group, and this difference was not statistically significant (P=0.97, 95% CI: 0.390–2.078). However, wall movement abnormalities were more common in the interventional treatment group [14/31 (45.2%) vs. 13/52, (25.0%), P=0.10, 95% CI: 0.960–6.361]. Interventional treatment was initially considered for 39 patients, who were evaluated by the interventional cardiology team. Following the assessment, 12 patients were redirected to pharmacological treatment, while 27 patients underwent CAG. Data could not be retrieved for 2 of these patients; thus, analyses were performed on 25 patients among whom 12 patients received PCI. The mortality rate was 33.3% (13/39) among patients in whom interventional treatment was preferred. Specifically, the mortality rate was 20.0% (5/25) in patients who underwent CAG and 16.7% (2/12) in those who underwent PCI. In contrast, 60 patients received pharmacological treatment, with a mortality rate reaching 68.3%, and this difference was statistically significant (P<0.001, 95% CI: 0.098–0.547) (Table 2).

There were significant differences between the interventional treatment group and the pharmacological treatment group in terms of heart rate (P=0.01, 95% CI: 2.815–24.562), serum creatinine (P=0.03, 95% CI: −0.348 to 0.745), troponin (P=0.02, 95% CI: −1,092 to 678), cancer surgery one week before the diagnosis of ACS (P=0.003, 95% CI: 1.679–9.820), vasopressor therapy (P<0.001, 95% CI: 0.044–0.442) and ejection fraction (P=0.007, 95% CI: 2.036–10.601). However, there was no significant difference between the groups in terms of comorbidities, GRACE score, CRP level, or white blood cell count (Tables 3,4).

Mortality rates were significantly lower in the interventional treatment preferred group than in the pharmacological treatment preferred group (OR =0.23, 95% CI: 0.10–0.55). Additionally, the 12 patients who underwent PCI were compared to all other patients who did not receive PCI (n=83), including both those who received pharmacological treatment and those in the interventional group who did not undergo PCI. In this comparison, the probability of in-hospital mortality was significantly lower in patients who underwent PCI (OR =0.13, 95% CI: 0.03–0.61), suggesting a substantial reduction in risk, though this should be interpreted cautiously given the small sample size and retrospective design.

The hospital mortality rate was calculated as 54.5% (54/99). Among the factors affecting mortality, the presence of metastases, the use of vasopressors, and surgical history were prominent. The mortality rate was 75.7% (28/37) in patients with metastases and 41.9% (26/62) in non-metastatic patients (P=0.002, 95% CI: 1.743–10.645). Similarly, the mortality rate was 80.6% (25/31) in patients who used vasopressors and 42.6% (29/68) in patients who did not use vasopressors (P<0.001, 95% CI: 2.036–15.421). While the mortality rate was 37.0% (n=20/54) in patients with a history of surgery, it was 75.6% (34/45) in patients without a history of surgery, and this difference was statistically significant (P<0.001, 95% CI: 0.079–0.457) (Table 5). Antiplatelet drug use after NSTE-MI diagnosis is shown in Table 6. Overall antiplatelet use was 82.5% (80/97), with higher rates in the interventional treatment group (86.5%) versus pharmacological treatment (80.0%). Use was lower in patients with hematologic malignancy (61.5%) compared to solid malignancy (85.7%). Antiplatelet use was universal in patients who underwent PCI (100%) and higher in those with CAG (92.0%). In this study, the relationship between the GRACE score and in-hospital mortality was examined, and a statistically significant positive correlation was found (Spearman’s rho =0.253, P=0.01). The median follow-up duration was 14.5 days (IQR, 7.0–28.0 days). Time-to-event data were available for 96 of 99 patients; 3 patients had missing time-to-event information but complete outcome status.

Kaplan-Meier analysis showed a median survival time of 36 days (95% CI: 0–89.3) in the interventional treatment group and 20 days (95% CI: 11.5–28.5) in the pharmacological treatment group. The difference in survival between the groups was not statistically significant (log-rank test, P=0.09) (Figure 2). To assess the robustness of these findings, sensitivity analyses were performed by excluding patients with missing or incomplete follow-up data; these exclusions did not materially alter the survival estimates.

Figure 2 Kaplan-Meier survival curves comparing ICU survival times between pharmacological (blue line) and interventional treatment (green line) groups. ICU survival times were analyzed using the log-rank test, which showed no statistically significant difference between the groups (P=0.09). Cox proportional hazards modeling was not applied, as the analysis focused on in-hospital mortality within a fixed observation period; Kaplan-Meier curves are presented for descriptive purposes only. Censored cases are indicated by tick marks. Time-to-event data were available for 96 of 99 patients; 3 patients had missing time-to-event information but complete outcome status. Sensitivity analyses were limited to the exclusion of patients with incomplete angiography-related data (n=2), and these exclusions did not materially change the results. No additional sensitivity analyses were performed due to the small sample size. ICU, intensive care unit.

Anemia was observed in 86.9% of patients (86 out of 99). The median hemoglobin level was 9.7 g/dL (IQR, 8.8–11.2 g/dL), and the mean hemoglobin value was 10.0±2.0 g/dL. Hospital mortality differed according to anemia status and treatment modality. Among patients receiving pharmacological treatment, mortality was 66.7% in anemic patients and 83.3% in non-anemic patients, with no statistically significant difference (P=0.65). In the interventional group, mortality was 40.6% in anemic patients, whereas no deaths occurred among non-anemic patients; however, this difference did not reach statistical significance (P=0.07). Overall, anemia status was not significantly associated with hospital mortality (P=0.24).

Hierarchical logistic regression analysis showed that interventional treatment was significantly associated with lower in-hospital mortality in Block 1 (OR =0.25, 95% CI: 0.09–0.70, P=0.008). After adjusting for metastasis, serum creatinine, vasopressor use, and GRACE score in Block 2, the presence of metastatic disease remained a significant risk factor for mortality (OR =4.46, 95% CI: 1.55–12.86, P=0.006), while the other covariates were not statistically significant predictors (Table 7).

Discussion

Key findings

In this study, the treatment strategies applied in patients with NSTE-MI diagnoses who were diagnosed with cancer and followed up in ICU for any reason were evaluated. Pharmacological treatment was more frequently preferred than interventional treatment. Additionally, mortality rates were significantly higher in patients receiving pharmacological treatment than in those receiving interventional procedures. Studies evaluating the coexistence of cancer and ACS have shown that the PCI rate in cancer patients is lower than that in non-cancer individuals (7,9). Similarly, a large retrospective cohort study demonstrated that cancer patients presenting with myocardial injury had significantly higher long-term mortality compared to non-cancer patients, emphasizing the elevated cardiovascular risk in this population (10).

Our findings are generally consistent with previous studies evaluating ACS in patients with malignancy. For example, Balanescu et al. similarly reported lower mortality with invasive management in selected cancer patients with ACS. However, unlike the broader ACS population, oncology patients represent a more heterogeneous group, with considerable variation in tumor type, disease stage, treatment exposure, and physiological reserve. This heterogeneity may influence both treatment selection and outcomes. Therefore, our results add clinically relevant insight by focusing specifically on critically ill oncology patients, a population that remains underrepresented in major ACS trials (11).

There are several reasons why pharmacological treatment is preferred for the treatment of NSTE-MI in cancer patients. The fact that thrombocytopenia is more common in cancer patients is one of these reasons. The inability to use antiplatelet drugs that should be used after interventional treatment and possible stent application may make pharmacological treatment preferable. In one study, the lack of antiplatelet therapy was identified as a major cause of stent thrombosis [hazard ratio (HR) =36.5, 95% CI: 8.0–167.8] (12). The same study highlighted that cancer diagnosis itself is a significant risk factor for stent thrombosis. In addition to thrombosis, undesirable events related to bleeding are more common in cancer patients (13). In the BleeMACS project, which is a multicenter observational study, the presence of cancer was shown to be the strongest risk factor for bleeding in patients undergoing PCI (HR =1.5, 95% CI: 1.1–2.1, P=0.02) (14). Readmission rates due to post-PCI bleeding have also been reported to be high (15,16). Several mechanisms may explain the observed association between interventional treatment and lower mortality. Inflammatory and immune-related mechanisms in cancer patients may exacerbate cardiovascular risk, suggesting that both the tumor biology and cancer therapies contribute to adverse outcomes, which could partially account for the higher mortality observed in medically treated patients (17). Early revascularization may reduce ongoing ischemia, support hemodynamic stability, and prevent additional myocardial damage. Intervention may also reduce recurrent ischemic events in patients with cancer-related prothrombotic states. Conversely, patients treated pharmacologically may remain at risk of reinfarction or persistent myocardial ischemia. Moreover, bleeding and thrombosis risks—which are frequently altered in malignancy—may differentially influence outcomes depending on the treatment strategy. These mechanisms remain hypothetical and require prospective confirmation. Our findings suggest that bleeding risk could be an important consideration influencing the selection of interventional treatment in this population. The possibility of increased renal damage after contrast media in patients with acute kidney injury is another reason for the preference for pharmacological treatment. In our study, patients receiving pharmacological treatment had significantly higher serum creatinine levels than did those receiving interventional treatment. In the SWEDEHEART study, the rate of CAG decreased as the estimated glomerular filtration rate decreased (18). However, less mortality has been reported in patients with mildly to moderately impaired renal function and early revascularization than in those who receive medical treatment (18). Elevated serum creatinine appears to be a concern in interventional treatment; however, early revascularization may reduce mortality.

Transporting hemodynamically unstable ICU patients to the angiography unit poses a significant risk. The incidence of adverse cardiovascular events during critically ill patient transfer ranges from 6% to 24% (19). The incidence of cardiopulmonary arrest during transfer was found to be greater in patients receiving vasopressor therapy (75–39.1%, P=0.043) (20). In our study, pharmacological treatment was preferred for the majority of patients receiving vasopressor therapy, likely because of concerns regarding transport safety.

The use of recommended ACS pharmacotherapy in cancer patients may be limited due to factors such as coagulopathies or thrombocytopenia. In our study, the rate of antiplatelet drug use was 80% in the pharmacological treatment group and 100% in the PCI-treated group, while it was lower at 61.5% among patients with hematologic malignancies. Although our data indicates variation in antiplatelet use, the study does not directly assess the impact of thrombocytopenia or bleeding complications, and thus conclusions about causality should be drawn cautiously. Previous literature suggests that thrombocytopenia and high bleeding risk can limit antiplatelet therapy in cancer patients with NSTE-MI, with previous studies reporting that antiplatelet use in this population may be lower than 45%. These findings highlight the importance of individualized treatment decisions to balance ischemic benefit and bleeding risk in this high-risk group (21). Bleeding risk stratification remains a critical aspect of decision-making in NSTE-MI patients with active cancer, particularly when considering invasive treatment options. Malignancy is an established risk factor for both ischemic and bleeding complications, posing unique therapeutic challenges. Recently, the Predicting Risk of Clinical Ischemic/Hemorrhagic and Severity Evaluation-High Bleeding Risk (PRECISE-HBR) score was introduced as a simplified yet comprehensive bleeding risk model that combines key laboratory parameters (e.g., hemoglobin, white blood cell count, renal function) with established The Academic Research Consortium for High Bleeding Risk (ARC-HBR) criteria, including active cancer, to better identify high bleeding risk patients after PCI (22). This 7-item score has demonstrated superior discrimination for major bleeding [Bleeding Academic Research Consortium (BARC) 3 or 5] at 1 year compared to earlier tools, including Predicting Bleeding Complications in Patients Undergoing Stent Implantation and Subsequent Dual Antiplatelet Therapy (PRECISE-DAPT). Although our retrospective dataset did not permit the calculation of a formal PRECISE-HBR score, our findings underscore the importance of individualized treatment decisions in this high-risk population. Future studies should prospectively integrate validated bleeding risk tools like PRECISE-HBR to help optimize therapeutic strategies for NSTE-MI in cancer patients, where the balance between ischemic benefit and bleeding hazard remains particularly delicate.

The GRACE score is widely used to predict in-hospital and 6-month mortality in ACS patients (23). However, this score focuses primarily on cardiovascular mortality and may not be sufficient to predict all-cause mortality in cancer patients. Several studies have suggested that the GRACE score has limited predictive power in oncology populations (24). In our study, although patients receiving pharmacological treatment had higher GRACE scores, no significant difference was found between the interventional and pharmacological treatment groups. However, these findings emphasize the potential need for oncology-specific risk models to improve prognostication. We observed a positive correlation between GRACE scores and mortality. The median GRACE scores for survivors and non-survivors were 116 and 126, respectively (P=0.006). The expected mortality rate for these scores should not exceed 10% according to GRACE risk models (25). However, the observed in-hospital mortality rates were 68.3% in the pharmacological treatment group and 33.3% in the interventional treatment group, far exceeding GRACE predictions. These findings suggest that the predictive ability of the GRACE score in cancer patients may be limited. Although hematologic malignancies may differ from solid tumors in terms of thrombocytopenia risk and ICU outcomes, a sensitivity analysis excluding these patients did not change the primary results, indicating that the overall findings are robust despite this heterogeneity.

Given the high burden of both cardiovascular and cancer-related mortality, there is a clear need for new prediction models tailored to oncology populations, incorporating cancer-specific variables such as tumor type, metastasis, and ongoing chemotherapy regimens, in order to more accurately guide treatment decisions. In our study, multivariable logistic regression analysis demonstrated that the presence of metastasis and the type of treatment received (interventional vs. pharmacological) were independently associated with in-hospital mortality. Specifically, metastasis significantly increased the risk of mortality (OR =5.56, 95% CI: 1.59–19.44, P=0.007), while interventional treatment was associated with lower mortality (OR =0.26, 95% CI: 0.08–0.85, P=0.03). Other variables showed no statistically significant or borderline effects. The model’s explanatory power was moderate, with a Nagelkerke R2 of 0.39, indicating that the selected variables moderately explained mortality outcomes. These findings highlight the importance of metastasis status and early interventional treatment as key factors influencing mortality in critically ill cancer patients and underscore their relevance in clinical decision-making.

In our study, the primary outcome was in-hospital mortality. However, we acknowledge that other clinically relevant endpoints—such as non-fatal ischemic events (e.g., reinfarction) and bleeding complications—are particularly important in the management of cancer patients with ACS. These outcomes could further inform the balance between invasive and conservative strategies. Although our data did not allow for the systematic collection of such events, we believe their inclusion in future research would be crucial for a more comprehensive assessment of safety and efficacy.

Future directions and clinical implications

The demand for individualized risk models in cancer patients is becoming more pertinent. Given the high risk of mortality from both cardiovascular and cancer-related etiologies, there is a need to create new predictive models with cancer-specific variables, including tumor type, metastasis, and current chemotherapy regimens that are not based on conventional cardiovascular risk factors alone.

This research offers important clinical information by contrasting the results of interventional and pharmacological treatment approaches in cancer patients with NSTE-MI who are admitted to the ICU. Research clearly reveals that there is a much higher mortality rate among patients undergoing pharmacological treatment, whereas interventional treatment leads to a reduced mortality rate. Several factors, including thrombocytopenia, high risk of bleeding, renal failure, hemodynamic instability, and complications of patient transfer, appear to influence treatment choice, especially pharmacologic treatment versus invasive treatment.

The management of ACS in cancer patients remains a challenging clinical problem. In particular, widely used risk stratification instruments such as the GRACE score have modest predictive utility in cancer patients. Recent statements in cardio-oncology highlight that integrating multidisciplinary approaches and tailored risk stratification, including the consideration of cancer therapy-related cardiovascular toxicity, can improve patient outcomes (26). Although our study revealed a positive correlation between the GRACE score and mortality, the much higher observed mortality rates in cancer patients than the GRACE predictions underscore the need for more robust, cancer-specific risk stratification models. Emerging evidence suggests that inflammatory biomarkers, such as the neutrophil-lymphocyte ratio (NLR), may provide additional prognostic value in predicting adverse cardiovascular events, including the mechanical complications of STEMI (27). Given the limitations of traditional scoring systems, integrating the NLR into risk assessment models could increase predictive accuracy, particularly in critically ill cancer patients with NSTE-MI. Moreover, artificial intelligence (AI)-based algorithms could further refine MI risk prediction and aid in treatment decision-making, particularly in complex cases where conventional risk scores may fall short (28). In this context, further research is warranted to evaluate the integration of AI driven risk models and inflammatory biomarkers such as the NLR into clinical practice.

It is important to acknowledge that patients selected for pharmacological treatment might have had a poorer baseline prognosis, including more advanced cancer stages or greater clinical instability, which could contribute to their higher observed mortality rates independently of treatment modality. Moreover, differences in baseline clinical status between the treatment groups indicate clear evidence of confounding by indication. Patients managed pharmacologically more often had hemodynamic instability, higher vasopressor requirements, worse renal function, and lower left ventricular ejection fraction at the time of NSTE-MI diagnosis—characteristics that likely influenced clinicians to favor medical management over invasive evaluation. This real-world selection bias likely contributed to the higher observed mortality in the pharmacological group. Although we adjusted for key covariates in multivariable models, residual confounding cannot be excluded; therefore, the observed association between treatment strategy and mortality should be interpreted as hypothesis-generating rather than causal. Given the retrospective nature of this study, unmeasured confounding factors could have influenced treatment allocation and outcomes, limiting causal inference. Therefore, the association between treatment modality and mortality should be interpreted with caution.

On the basis of these findings, a multidisciplinary and individualized strategy is essential for the management of ACS in patients with cancer. The development of refined risk stratification models incorporating cancer stage, metastatic status, comorbidities, and current oncological therapies may enable more precise treatment decision-making. Large prospective studies investigating the long-term impact of early revascularization are necessary to develop stronger evidence for the benefit of interventional therapies. In addition, the complexity of clinical decision-making in this high-risk population underscores the importance of structured multidisciplinary collaboration. Integrating oncologists, cardiologists, and critical care specialists in treatment planning may improve the accuracy of risk-benefit assessments and facilitate the use of individualized management algorithms. Such collaborative frameworks may prove especially valuable when navigating therapeutic dilemmas such as bleeding risk, thrombocytopenia, or treatment delays due to cancer-related factors. Finally, enhanced collaboration between critical care specialists and interventional cardiology teams regarding the management of critically ill patients with cancer can optimize patient outcomes and improve overall clinical decision-making.

Limitations

There are several limitations in this study. First, as a retrospective study, it is subject to information bias, missing data, and documentation errors, which may affect the accuracy of the dataset. Several clinically important variables—such as cancer stage, frailty, functional status, and detailed oncologic treatment history—were not uniformly available and therefore could not be included in the multivariable models. Moreover, the retrospective design inherently limits the ability to control unmeasured confounding variables and precludes any causal inference. Additionally, relatively small sample size and single-center ICU setting limit both the statistical power and the external validity of the findings. Several subgroup analyses, including those involving treatment modalities, were constrained by limited case numbers, increasing the risk of type II error. Furthermore, baseline differences between treatment groups—such as higher vasopressor use, elevated serum creatinine, and lower ejection fraction in the pharmacological group—likely reflect selection bias, with sicker patients less likely to undergo interventional therapy. Although multivariable adjustment was applied, residual confounding cannot be fully excluded, and these factors should be considered when interpreting the results and generalizing findings to other patient populations. Multicenter studies with larger cohorts are needed to validate these findings.

Second, the study focused exclusively on in-hospital mortality. Long-term outcomes such as post-discharge survival, readmissions, non-fatal ischemic events, and bleeding complications—particularly relevant in oncology patients—were not available. This restricts a more comprehensive safety and efficacy evaluation of treatment strategies. Future prospective, multicenter studies with standardized outcome definitions and follow-up protocols are warranted to address these gaps and better inform individualized treatment decisions in this high-risk population.

Finally, the limitations of the GRACE score in cancer patients require further investigation. Additionally, twelve patients initially assigned to the interventional strategy did not undergo CAG due to clinical considerations. While these patients were analyzed according to the initial treatment decision, this approach may introduce a minor misclassification bias. Nonetheless, this classification reflects the intended clinical strategy and the real-world decision-making process in critically ill cancer patients. While our study suggests that it may not be an adequate predictor of mortality, we did not directly compare it with alternative risk models. Given the unique physiological characteristics of critically ill cancer patients, our findings suggest that the GRACE score—or a modified version incorporating oncology-specific variables such as tumor stage, treatment history, or inflammation markers—may inform future risk assessment strategies. Development of oncology-specific ACS risk models could be an important area for further research. Despite these limitations, our study provides important information about the effectiveness of NSTE-MI treatment strategies in cancer patients and the validity of the GRACE score in this patient population. Future multidisciplinary and prospective studies should evaluate these findings in larger patient groups and contribute to the creation of clinical guidelines specific to cancer patients.

Conclusions

This 5-year retrospective study highlights the challenges involved in managing NSTE-MI in critically ill cancer patients. Although pharmacological therapy is often chosen due to concerns such as thrombocytopenia, bleeding risk, renal impairment, and hemodynamic instability, our findings suggest that interventional treatment may be associated with lower in-hospital mortality when clinically feasible. The notably higher mortality observed among patients receiving conservative management may reflect baseline differences and highlights the need for individualized treatment considerations. However, due to the retrospective design, unmeasured confounding, and baseline differences between groups, these results should be interpreted cautiously. The retrospective nature of this analysis limits causal inference, and the observed associations may be influenced by selection bias. Larger, prospective, multicenter studies are essential to validate these findings and to support the development of cancer-specific decision-making frameworks for ACS management. Given the importance of bleeding risk in this population, incorporation of validated risk scores—such as the PRECISE-HBR score—may enhance safety and help tailor treatment strategies more effectively. Enhanced multidisciplinary collaboration between cardiology, oncology, critical care, and palliative care teams will remain crucial for optimizing outcomes in this vulnerable population.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://cdt.amegroups.com/article/view/10.21037/cdt-2025-525/rc

Data Sharing Statement: Available at https://cdt.amegroups.com/article/view/10.21037/cdt-2025-525/dss

Peer Review File: Available at https://cdt.amegroups.com/article/view/10.21037/cdt-2025-525/prf

Funding: None.

Conflicts of Interest: Both authors have completed the ICMJE uniform disclosure form (available at https://cdt.amegroups.com/article/view/10.21037/cdt-2025-525/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of Dr. Abdurrahman Yurtaslan Ankara Oncology Training and Research Hospital (No. 2024–10/151). As this was a retrospective study, the requirement for informed consent was formally waived by the Ethics Committee.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Alam N, Wright AK, Ashcroft DM, et al. Cancer and cardiovascular disease. Lancet 2020;395:1903-4. [Crossref] [PubMed]
2. Lyon AR, López-Fernández T, Couch LS, et al. 2022 ESC Guidelines on cardio-oncology developed in collaboration with the European Hematology Association (EHA), the European Society for Therapeutic Radiology and Oncology (ESTRO) and the International Cardio-Oncology Society (IC-OS). Eur Heart J 2022;43:4229-361. [Crossref] [PubMed]
3. Gevaert SA, Halvorsen S, Sinnaeve PR, et al. Evaluation and management of cancer patients presenting with acute cardiovascular disease: a Consensus Document of the Acute CardioVascular Care (ACVC) association and the ESC council of Cardio-Oncology-Part 1: acute coronary syndromes and acute pericardial diseases. Eur Heart J Acute Cardiovasc Care 2021;10:947-59. [Crossref] [PubMed]
4. Navi BB, Reiner AS, Kamel H, et al. Risk of Arterial Thromboembolism in Patients With Cancer. J Am Coll Cardiol 2017;70:926-38. [Crossref] [PubMed]
5. Navi BB, Reiner AS, Kamel H, et al. Arterial thromboembolic events preceding the diagnosis of cancer in older persons. Blood 2019;133:781-9. [Crossref] [PubMed]
6. Lucà F, Parrini I, Abrignani MG, et al. Management of Acute Coronary Syndrome in Cancer Patients: It's High Time We Dealt with It. J Clin Med 2022;11:1792. [Crossref] [PubMed]
7. Rohrmann S, Witassek F, Erne P, et al. Treatment of patients with myocardial infarction depends on history of cancer. Eur Heart J Acute Cardiovasc Care 2018;7:639-45. [Crossref] [PubMed]
8. Collet JP, Thiele H, Barbato E, et al. 2020 ESC Guidelines for the management of acute coronary syndromes in patients presenting without persistent ST-segment elevation. Eur Heart J 2021;42:1289-367. [Crossref] [PubMed]
9. Mohamed MO, Lopez-Mattei JC, Parwani P, et al. Management strategies and clinical outcomes of acute myocardial infarction in leukaemia patients: Nationwide insights from United States hospitalisations. Int J Clin Pract 2020;74:e13476. [Crossref] [PubMed]
10. Sabaté-Tormos M, Bardají A, Peiró OM, et al. Cancer and myocardial injury in patients with suspected acute coronary syndrome. Cardiooncology 2025;11:21. [Crossref] [PubMed]
11. Balanescu DV, Donisan T, Deswal A, et al. Acute myocardial infarction in a high-risk cancer population: Outcomes following conservative versus invasive management. Int J Cardiol 2020;313:1-8. [Crossref] [PubMed]
12. van Werkum JW, Heestermans AA, Zomer AC, et al. Predictors of coronary stent thrombosis: the Dutch Stent Thrombosis Registry. J Am Coll Cardiol 2009;53:1399-409. [Crossref] [PubMed]
13. Bharadwaj AS, Swamy PM, Mamas MA. Outcomes of percutaneous coronary interventions in cancer patients. Expert Rev Cardiovasc Ther 2020;18:25-32. [Crossref] [PubMed]
14. Iannaccone M, D'Ascenzo F, Vadalà P, et al. Prevalence and outcome of patients with cancer and acute coronary syndrome undergoing percutaneous coronary intervention: a BleeMACS substudy. Eur Heart J Acute Cardiovasc Care 2018;7:631-8. [Crossref] [PubMed]
15. Kwok CS, Wong CW, Kontopantelis E, et al. Percutaneous coronary intervention in patients with cancer and readmissions within 90 days for acute myocardial infarction and bleeding in the USA. Eur Heart J 2021;42:1019-34. [Crossref] [PubMed]
16. Potts JE, Iliescu CA, Lopez Mattei JC, et al. Percutaneous coronary intervention in cancer patients: a report of the prevalence and outcomes in the United States. Eur Heart J 2019;40:1790-800. [Crossref] [PubMed]
17. Miao M, Liu X, Zhang H, et al. Immuno-inflammatory mechanisms in cardio-oncology: new hopes for immunotargeted therapies. Front Oncol 2025;15:1516977. [Crossref] [PubMed]
18. Szummer K, Lundman P, Jacobson SH, et al. Influence of renal function on the effects of early revascularization in non-ST-elevation myocardial infarction: data from the Swedish Web-System for Enhancement and Development of Evidence-Based Care in Heart Disease Evaluated According to Recommended Therapies (SWEDEHEART). Circulation 2009;120:851-8. [Crossref] [PubMed]
19. Droogh JM, Smit M, Absalom AR, et al. Transferring the critically ill patient: are we there yet? Crit Care 2015;19:62. [Crossref] [PubMed]
20. Min HJ, Kim HJ, Lee DS, et al. Intra-hospital transport of critically ill patients with rapid response team and risk factors for cardiopulmonary arrest: A retrospective cohort study. PLoS One 2019;14:e0213146. [Crossref] [PubMed]
21. Yusuf SW, Daraban N, Abbasi N, et al. Treatment and outcomes of acute coronary syndrome in the cancer population. Clin Cardiol 2012;35:443-50. [Crossref] [PubMed]
22. Gragnano F, van Klaveren D, Heg D, et al. Derivation and Validation of the PRECISE-HBR Score to Predict Bleeding After Percutaneous Coronary Intervention. Circulation 2025;151:343-55. [Crossref] [PubMed]
23. Fox KA, Dabbous OH, Goldberg RJ, et al. Prediction of risk of death and myocardial infarction in the six months after presentation with acute coronary syndrome: prospective multinational observational study (GRACE). BMJ 2006;333:1091. [Crossref] [PubMed]
24. Koo CY, Zheng H, Tan LL, et al. Global Registry of Acute Coronary Events Score Underestimates Post-Acute Coronary Syndrome Mortality among Cancer Patients. Cancers (Basel) 2023;15:5222. [Crossref] [PubMed]
25. GRACE ACS Risk and Mortality Calculator. 2025. Available online: https://www.mdcalc.com/calc/1099/grace-acs-risk-mortality-calculator
26. Bloom MW, Vo JB, Rodgers JE, et al. Cardio-Oncology and Heart Failure: a Scientific Statement From the Heart Failure Society of America. J Card Fail 2025;31:415-55. [Crossref] [PubMed]
27. Yoon GS, Choi SH, Woo SI, et al. Neutrophil-to-Lymphocyte Ratio at Emergency Room Predicts Mechanical Complications of ST-segment Elevation Myocardial Infarction. J Korean Med Sci 2021;36:e131. [Crossref] [PubMed]
28. Kim D, Hwang JE, Cho Y, et al. A Retrospective Clinical Evaluation of an Artificial Intelligence Screening Method for Early Detection of STEMI in the Emergency Department. J Korean Med Sci 2022;37:e81. [Crossref] [PubMed]