Abstract
Background
Antithrombotic agents to treat patients with acute myocardial infarction can cause bleeding, which may reveal undiagnosed cancer. However, the relationship between bleeding and new cancer diagnosis and the prognostic impact is still unclear.
Methods and Results
We analyzed the new cancer diagnosis, Bleeding Academic Research Consortium 2, 3, or 5 bleeding, and all‐cause death of 10 364 patients with acute myocardial infarction without a history of previous cancer in a multicenter acute myocardial infarction registry. During a median of 4.9 years, 1109 patients (10.7%) experienced Bleeding Academic Research Consortium 2, 3, or 5 bleeding, and 338 patients (3.3%) were newly diagnosed with cancer. Bleeding Academic Research Consortium 2, 3, or 5 bleeding was associated with an increased risk of new cancer diagnosis (subdistribution hazard ratio [sHR] 3.29 [95% CI, 2.50–4.32]). In particular, there were robust associations between gastrointestinal bleeding and new gastrointestinal cancer diagnosis (sHR, 19.96 [95% CI, 11.30–29.94]) and between genitourinary bleeding and new genitourinary cancer diagnosis (sHR, 28.95 [95% CI, 14.69–57.07]). The risk of all‐cause death was not lower in patients diagnosed with new gastrointestinal cancer after gastrointestinal bleeding (hazard ratio [HR], 4.05 [95% CI, 2.04–8.02]) and diagnosed with new genitourinary cancer after genitourinary bleeding (HR, 2.79 [95% CI, 0.81–9.56]) than in patients newly diagnosed with cancer without previous bleeding.
Conclusions
Clinically significant bleeding, especially gastrointestinal and genitourinary bleeding, in patients with AMI was associated with an increased risk of new cancer diagnoses. However, the bleeding preceding new cancer detection was not associated with better survival.
Registration
URL: https://www.clinicaltrials.gov; Unique identifier: NCT02385682 and NCT02806102.
Keywords: bleeding, cancer, gastrointestinal cancer, myocardial infarction, urogenital cancer
Subject Categories: Percutaneous Coronary Intervention, Clinical Studies, Cardiovascular Disease, Mortality/Survival
Nonstandard Abbreviations and Acronyms
- BARC
Bleeding Academic Research Consortium
- COREA‐AMI
Cardiovascular Risk and Identification of Potential High‐Risk Population in Acute Myocardial Infarction
Clinical Perspective.
What Is New?
Clinically relevant bleeding, especially gastrointestinal and genitourinary bleeding, after percutaneous coronary intervention for acute myocardial infarction was associated with an increase in new cancer diagnosis.
The mortality of patients with precedent bleeding and new cancer diagnosis was not lower than those newly diagnosed with cancer without bleeding.
What Are the Clinical Implications?
Prompt evaluation of the bleeding focus is required not only to manage the bleeding but also to identify the bleeding focus and rule out the possibility of cancer‐related bleeding.
A larger study is necessary to reevaluate this finding and search the treatment strategy to improve the prognosis of patients suffering concomitantly from myocardial infarction and cancer.
Treatment with antiplatelet agents after percutaneous coronary intervention (PCI) is related to a higher risk of bleeding. Clinically significant bleeding after PCI is not uncommon and occurs in the range of 1% to 8% per year, depending on the patient cohort and the definition of the bleeding. 1 , 2 , 3 , 4 Post‐PCI bleeding may not only result in the premature discontinuation of antiplatelet agents for bleeding cessation but may also reveal a previously unrecognized malignant tumor during the evaluation of the bleeding focus. It is well known that the most common locations of the bleeding are the gastrointestinal tract and the genitourinary tract. However, the incidence and prognostic impact of cancer accompanying those bleedings are still unknown. 5 Antiplatelet agents may need to be discontinued for a period to manage the bleeding and investigate the bleeding focus, leading to thrombotic complications. 6 It may be more dangerous in patients with acute myocardial infarction (AMI) who have higher thrombogenicity and are recommended to maintain dual antiplatelet therapy longer than those with stable coronary artery disease. 7 , 8 Demonstrating the relevance and extent of the subsequent cancer diagnosis following bleeding may allow physicians to properly focus on a diagnostic workup for possible underlying conditions.
Thus, we investigated the association between bleeding and a new cancer diagnosis and the association between mortality and a new cancer diagnosis in a large, real‐world, PCI‐treated AMI population. We also investigated the association between gastrointestinal bleeding and gastrointestinal cancer and the association between genitourinary bleeding and genitourinary cancer. We compared the mortality of patients with AMI diagnosed with new gastrointestinal or genitourinary cancer with or without gastrointestinal or genitourinary bleeding to identify the mortality benefit of the bleeding for cancer detection in patients with AMI.
METHODS
The data that support the findings of this study are available from the corresponding author upon reasonable request.
Study Population
The COREA‐AMI (Cardiovascular Risk and Identification of Potential High‐Risk Population in Acute Myocardial Infarction) registry was designed to evaluate the real‐world, long‐term clinical outcomes of all patients with AMI who underwent PCI from January 2004 to August 2014. 9 This registry contained clinical parameters at baseline and long‐term clinical follow‐up data for as long as possible by 2019. Participating university hospitals already had web‐based registries that prospectively enrolled all consecutive patients undergoing PCI and all patients with AMI. All the included patients were aged >20 years and underwent PCI. Patients managed with a conservative strategy were excluded. The study was conducted in compliance with the Declaration of Helsinki for investigations in humans. The study protocol was approved by the institutional review board of the participating centers, and all participants provided written informed consent. The registry was registered on ClinicalTrials.gov (study numbers: NCT02385682 and NCT02806102).
AMI was diagnosed by detecting elevated cardiac biomarkers at least 1 value above the 99th percentile of the upper reference limit, with temporal increases and decreases, and at least 1 of the following indications: symptoms of ischemia, new or presumed new significant ST‐segment T‐wave changes or new left bundle‐branch block, development of pathological Q waves on electrocardiography, imaging evidence of new loss of viable myocardium or new abnormality in regional wall motion or intracoronary thrombus on angiography. 10
All data were collected on a web‐based system after deleting personal information. Independent research personnel collected baseline clinical, laboratory, and medication data, and independent trained reviewers and interventional cardiologists assessed the angiographic and procedural data. The medication, including antiplatelet therapy, was checked every 6 months until 3 years after index PCI. Clinical events and outcomes were obtained from electronic medical records and telephone conversations. All the adverse clinical events of interest were determined by source documents. They were confirmed centrally by the committee of the Cardiovascular Center of Seoul St. Mary's Hospital, Seoul, Republic of Korea. Mortality was verified by determining whether the patient had been removed from the National Health Insurance Service, the single government‐managed insurance program in Korea that includes almost the entire population of the country. The final data set was managed by independent statisticians of the Clinical Research Coordinating Center and sealed with a code by a clinical research associate.
Among 10 719 patients with AMI who underwent PCI, we enrolled 10 364 patients, excluding 355 patients who had already been diagnosed with cancer before undergoing PCI. We analyzed the first occurrence of the Bleeding Academic Research Consortium (BARC) 2, 3, or 5 bleeding, new diagnoses of cancer, and mortality of patients.
Assessment of Clinically Relevant Bleeding
We analyzed the first occurrence of BARC 2, 3, or 5 bleeding, new diagnoses of cancer, and all‐cause death of all patients in the COREA‐AMI registry. Bleeding events according to BARC classification were collected using a dedicated assessment form (Table S1). BARC type 2 is any overt, actionable sign of hemorrhage that does not fit the criteria for type 3, 4, or 5 but does meet at least 1 of the following criteria: (1) requiring nonsurgical medical intervention by a health care professional, (2) leading to hospitalization or increased level of care, or (3) prompting evaluation. BARC type 3 is related to any transfusion with overt bleeding (type 3a), cardiac tamponade, bleeding requiring surgical intervention for control or intravenous vasoactive agents (type 3b), or intracranial hemorrhage, subcategories confirmed by autopsy or imaging or lumbar puncture, or intraocular bleed compromising vision (type 3c). BARC type 5 is any fatal bleeding. The location of the bleeding was also obtained from the bleeding assessment form. We defined gastrointestinal bleeding as hematemesis, melena, or hematochezia and genitourinary bleeding as hematuria or vaginal bleeding.
Diagnosis of Cancer
All cancer diagnoses were assessed and validated by a medical record review by a panel of physicians. The diagnosis of cancer, including nonmelanoma skin and blood cancer, was confirmed primarily by the pathology or cytology reports. Otherwise, the clinical diagnosis of cancer was established based on robust clinical and radiological or laboratory marker evidence according to the current guidelines. The cancer diagnosis date was assigned based on the date of pathologic diagnosis or the clinical diagnosis date. Cancer types were classified by the anatomic and primary systems involved. We defined gastrointestinal cancer as cancer involving the esophagus, stomach, duodenum, jejunum, ileum, colon, or rectum. Genitourinary cancers were defined as cancer involving the prostate, spermatic cord, uterus, cervix, vagina, kidney, ureter, bladder, or urethra. Information on the cancer stages of each cancer was not available.
Mortality
We also assessed the mortality after AMI according to clinically relevant bleeding and new cancer diagnosis. All‐cause death was considered cardiovascular death unless a noncardiovascular death could be clearly identified.
Statistical Analysis
Continuous variables were summarized as mean±SD or median (first quartile–third quartile), and categorical variables were expressed by percentages. We examined the number and proportion of new cancer diagnoses with or without prior bleeding. Aalen‐Johansen estimates were used to determine the cumulative incidence of new cancer after AMI because of competing risks of death. Fine‐Gray regression analysis was conducted to calculate the subdistribution hazard ratios in the presence of competing risks. Furthermore, cause‐specific hazard ratios were analyzed using a Cox proportional hazard model in which those who experienced death are treated as censored for cancer. Bleeding events were treated as a time‐varying covariate in all analyses. Age and sex were included as covariates in Cox proportional hazard models.
We compared the mortality of patients according to new cancer diagnosis and prior bleeding. We also compared the mortality of patients diagnosed with new gastrointestinal or genitourinary cancer with or without gastrointestinal or genitourinary bleeding. According to the experience of bleeding, the risk of mortality after PCI was assessed by Cox proportional hazard model in patients with or without a new cancer diagnosis. Prior bleeding and new cancer diagnosis were treated as time‐dependent variables. The mortality risk was also adjusted with age, sex, Killip class, hypertension, diabetes, current smoking, previous stroke, atrial fibrillation on admission, left ventricular ejection fraction, estimated glomerular filtration rate, and multivessel disease. Extended Kaplan‐Meier estimates were used to determine the cumulative incidence of death according to prior bleeding and cancer occurrence. Death and the loss of follow‐up are treated as censorings. The P values were 2‐sided, and <0.05 was considered statistically significant in all analyses.
RESULTS
Baseline Characteristics of Patients According to the Occurrence of Clinically Relevant Bleeding
A total of 1109 patients experienced BARC 2, 3, or 5 bleeding during a median follow‐up period of 4.9 years (interquartile range, 3.0–7.3). Of all BARC 2, 3, or 5 bleeding, bleeding after a new cancer diagnosis occurred in 29 patients (2.6%) who were allocated to patients without BARC 2, 3, or 5. Table 1 presents the baseline characteristics of patients who experienced BARC 2, 3, or 5 bleeding compared with patients without BARC 2, 3, or 5 bleeding. The median follow‐up period was shorter in patients with bleeding than in those without bleeding (4.4 versus 5.0 years, P<0.001). Patients with BARC 2, 3, or 5 bleeding more frequently had bleeding risk factors such as old age, low body mass index, history of stroke, history of bleeding, anemia, thrombocytopenia, and taking anticoagulants or analgesics. In addition, patients with bleeding were more likely to have a higher Killip class, hypertension, diabetes, and lower left ventricular systolic fraction. Antiplatelet agents before PCI were more frequent in patients with BARC 2, 3, or 5 bleeding. The mean duration of dual antiplatelet therapy was shorter in patients with BARC 2, 3, or 5 bleeding than in those without bleeding. Only about 20% of patients were taking dual antiplatelet therapy, and 3% were taking anticoagulation at the time of bleeding. No patient was taking triple therapy (dual antiplatelet therapy plus anticoagulation) at the time of bleeding.
Table 1.
Baseline Patient Characteristics According to the Occurrence of Bleeding
| Characteristic | Whole cohort, N=10 364 | No bleeding, N=9284 | Bleeding, N=1080 | P value |
|---|---|---|---|---|
| Age, y | 63.5±12.8 | 63.1±12.8 | 66.7±12.4 | <0.001 |
| Age ≥75 y, % | 2264 (21.8) | 1943 (20.9) | 321 (29.7) | <0.001 |
| Men, % | 7413 (71.5) | 6741 (72.6) | 672 (62.2) | <0.001 |
| BMI, kg/m2 | 24.2±3.3 | 24.2±3.2 | 23.6±3.4 | <0.001 |
| Killip class ≥3 | 2437 (23.5) | 2099 (22.6) | 338 (31.3) | <0.001 |
| Diagnosis of MI (%) | ||||
| NSTEMI | 4699 (45.3) | 4189 (45.1) | 510 (47.2) | 0.2 |
| STEMI | 5665 (54.7) | 5095 (54.9) | 570 (52.8) | |
| Hypertension (%) | 5420 (52.3) | 4772 (51.4) | 648 (60.0) | <0.001 |
| Diabetes (%) | 3277 (31.6) | 2891 (31.1) | 386 (35.7) | 0.002 |
| Current smoker (%) | 4219 (40.7) | 3872 (41.7) | 347 (32.1) | <0.001 |
| Prior MI (%) | 437 (4.2) | 380 (4.1) | 57 (5.3) | 0.079 |
| Prior stroke (%) | 611 (5.9) | 609 (6.6) | 124 (11.5) | <0.001 |
| ICH and moderate or severe stroke (%) | 122 (1.2) | 101 (1.1) | 21 (1.9) | 0.02 |
| Liver cirrhosis (%) | 16 (0.2) | 14 (0.2) | 2 (0.2) | 0.68 |
| Previous bleeding (%) | 38 (0.4) | 28 (0.3) | 10 (0.9) | 0.004 |
| Atrial fibrillation | 339 (3.3) | 283 (3.0) | 56 (5.2) | <0.001 |
| LVEF, % | 53.1±11.3 | 53.3±10.8 | 51.6±11.6 | <0.001 |
| eGFR, mL/min per 1.73 m2 | 75.5±25.9 | 76.6±25.3 | 65.7±28.5 | <0.001 |
| eGFR <30 (%) | 689 (6.6) | 538 (5.8) | 151 (14.0) | <0.001 |
| eGFR 30–59 (%) | 1909 (18.4) | 1634 (17.6) | 275 (25.5) | <0.001 |
| Hemoglobin <11 g/dL (%) | 1092 (10.5) | 861 (9.3) | 231 (21.4) | <0.001 |
| Moderate anemia (%) | 1972 (19.0) | 1791 (19.3) | 181 (16.8) | 0.049 |
| Platelet <100 × 109/L (%) | 106 (1.0) | 85 (0.9) | 21 (1.9) | 0.003 |
| Antithrombotic treatment | ||||
| SAPT before AMI (%) | 1179 (11.4) | 998 (10.7) | 181 (16.8) | <0.001 |
| Aspirin before AMI (%) | 1095 (10.6) | 933 (10.0) | 162 (15.0) | <0.001 |
| Clopidogrel before AMI (%) | 518 (5.0) | 439 (4.7) | 79 (7.3) | <0.001 |
| DAPT before AMI (%) | 449 (4.3) | 389 (4.2) | 60 (5.6) | 0.045 |
| Discharge medication | ||||
| Aspirin (%) | 9681 (98.2) | 8697 (98.3) | 984 (97.0) | 0.004 |
| Clopidogrel (%) | 8482 (86.1) | 7614 (86.1) | 868 (85.6) | 0.698 |
| Prasugrel or ticagrelor (%) | 1303 (13.2) | 1174 (13.2) | 129 (12.6) | 0.613 |
| Anticoagulation (%) | 328 (3.2) | 266 (2.9) | 62 (5.7) | <0.001 |
| DAPT duration | 19.9±14.1 | 20.1±14.1 | 18.8±14.2 | 0.007 |
| DAPT at the time of bleeding (%) | 352 (32.6) | |||
| Anticoagulation at the time of bleeding (%) | 28 (2.6) | |||
| Other medications at discharge | ||||
| Statin (%) | 8920 (96.7) | 8025 (96.9) | 895 (95.1) | 0.006 |
| β‐Blocker (%) | 8130 (78.4) | 7318 (78.8) | 812 (75.2) | 0.022 |
| RAS inhibitor (%) | 7702 (74.3) | 6923 (74.6) | 779 (72.1) | 0.173 |
| NSAID (%) | 59 (0.6) | 46 (0.5) | 13 (1.3) | 0.006 |
| HBR by ARC‐HBR criteria (%) | 3138 (30.3) | 2617 (28.2) | 521 (48.2) | <0.001 |
| Angiographic characteristics | ||||
| LM disease (%) | 670 (6.5) | 583 (6.3) | 87 (8.1) | 0.029 |
| Multivessel disease (%) | 5569(53.7) | 4945 (53.3) | 624 (57.8) | 0.005 |
| Culprit (%) | 0.415 | |||
| LM/LAD | 5257(50.7) | 4696 (50.6) | 561 (51.9) | |
| LCX/RCA | 5107(49.3) | 4588 (49.4) | 519 (48.1) | |
| Total stent no. | 1.6±0.9 | 1.6±0.9 | 1.6±0.9 | 0.01 |
Values are n (%) or mean±SD. Moderate anemia, hemoglobin 11–12.9 g/dL for men and 11–11.9 g/dL for women. AMI indicates acute myocardial infarction; ARC‐HBR, Academic Research Consortium for High Bleeding Risk; BMI, body mass index; DAPT, dual antiplatelet therapy; eGFR, estimated glomerular filtration rate; HBR, high bleeding risk; ICH, intracranial hemorrhage; LAD, left anterior descending artery; LCX, left circumflex artery; LM, left main; LVEF, left ventricular ejection fraction; MI, myocardial infarction; NSTEMI, non–ST‐segment–elevation myocardial infarction; RAS, renin‐angiotensin system; RCA, right coronary artery; SAPT, single antiplatelet therapy; and STEMI, ST‐segment–elevation myocardial infarction.
New Cancer Diagnosis After Clinically Relevant Bleeding
Among 10 364 patients not diagnosed with cancer previously, a total of 338 patients (6.5 per 1000 patient‐years) were diagnosed with new cancers (Table 2). The median interval between index PCI for AMI and new cancer diagnosis in patients with BARC 2, 3, or 5 bleeding were shorter and more skewed to the early period (skewness, 1.2 versus 0.6; Figure S1) than in those without bleeding. Major cancer types were gastrointestinal cancer, lung cancer, and genitourinary cancer. The most common gastrointestinal cancer was stomach cancer (53.8%), and the most common genitourinary cancer was prostate cancer (37.5%). The cancer characteristics are summarized in Table S1.
Table 2.
New Cancer Diagnosis According to the Occurrence of Bleeding
| Cancer diagnosis | Whole cohort, N=10 364 | No bleeding, N=9284 | Bleeding, N=1080 | P value |
|---|---|---|---|---|
| New cancer diagnosis | 338 (6.4) | 250 (4.7) | 88 (17.2) | <0.001 |
| Type of cancer | <0.001 | |||
| Gastrointestinal cancer | 132 (2.6) | 86 (1.6) | 46 (9.0) | |
| Genitourinary cancer | 48 (0.9) | 32 (0.6) | 16 (3.4) | |
| Lung cancer | 73 (1.4) | 63 (1.2) | 10 (3.4) | |
| Hepatobiliary cancer | 44 (0.9) | 34 (0.6) | 10 (3.0) | |
| Breast cancer | 11 (0.2) | 9 (0.2) | 2 (0.4) | |
| Other | 30 (0.6) | 26 (0.5) | 4 (1.2) | |
| Interval from index AMI to cancer diagnosis in patients with new cancer, y | 2.5 [0.9–4.5] | 2.6 [1.0–4.7] | 2.0 [0.8–3.4] | 0.029 |
Values are the number of events (the incidence per 1000 patient‐years) or median [interquartile range]. Bleeding confers the first occurrence of Bleeding Academic Research Consortium 2, 3, or 5 bleeding. Specific subtypes of each type of cancer are summarized in Table S1. AMI indicates acute myocardial infarction.
Cumulative incidence of new cancer by competing risk analyses according to clinically relevant bleeding are depicted in Figure S2. BARC 2, 3, or 5 bleeding was associated with an >3‐fold increased risk of new cancer diagnosis in multivariate competing risk analysis (Figure 1). When comparing BARC 2 bleeding and BARC 3 or 5 bleeding, BARC 2 bleeding was associated with higher hazard ratio (HR) for new cancer diagnosis than BARC 3 or 5 bleeding. (Table S2).
Figure 1. Incidence and hazard ratio of new cancer diagnosis by multivariate competing risk analysis according to clinically relevant bleeding.
Values in the figure are HR (95% CI) according to subdistribution and cause‐specific hazard model. Bleeding confers the first occurrence of Bleeding Academic Research Consortium 2, 3, or 5 bleeding. GI indicates gastrointestinal; GU, genitourinary; and HR, hazard ratio.
Gastrointestinal bleeding occurred in 434 patients (4.2%), and genitourinary bleeding occurred in 157 patients (1.5%) before new cancer detection. Gastrointestinal bleeding was associated with a 15‐ to 19‐fold higher hazard of new gastrointestinal cancer diagnosis (Figure 1). Genitourinary bleeding was associated with a 29‐ to 30‐fold higher risk of a new genitourinary cancer diagnosis (Figure 1). The median intervals from index AMI to gastrointestinal bleeding (254 days) and genitourinary bleeding (354 days) were shorter than other kinds of BARC 2, 3, or 5 bleeding (1689 days). Most gastrointestinal and genitourinary cancers were diagnosed within 1 year after gastrointestinal bleeding and genitourinary bleeding (Figure 2).
Figure 2. Cumulative incidence of new cancer diagnosis after bleeding.
Values in the figure are percent (95% CI). GI indicates gastrointestinal; and GU, genitourinary.
Mortality According to Previous Clinically Relevant Bleeding and New Cancer Diagnosis
Mortality after AMI according to previous clinically relevant bleeding and new cancer diagnosis is demonstrated in Figure 3 and Table S3. The patients with a new cancer diagnosis had higher noncardiovascular mortality but similar cardiovascular mortality compared with those without a new cancer diagnosis. In comparison, both cardiovascular and noncardiovascular mortality were higher in patients with BARC 2, 3, or 5 bleeding than in those without bleeding. BARC 2, 3, or 5 bleeding before a new cancer diagnosis was associated with increased all‐cause death, cardiovascular death, and noncardiovascular death compared with a new cancer diagnosis without previous bleeding. The patients diagnosed with new cancer after the bleeding had the highest rate of all‐cause death after PCI compared with other groups of patients stratified by a new cancer diagnosis and BARC 2, 3, or 5 bleeding. The HRs of the patients diagnosed with new cancer after bleeding in comparison with each group are as follows: against those with bleeding but without cancer (adjusted HR, 2.15 [95% CI, 1.71–2.71]; P<0.001), those with new cancer diagnosis but without bleeding (adjusted HR, 1.04 [95% CI, 0.94–1.15]; P=0.66), and those with neither bleeding nor cancer (adjusted HR, 5.21 [95% CI, 3.64–7.46]; P<0.001). The extended Kaplan‐Meier curve of all‐cause death in 4 groups is depicted in Figure S3.
Figure 3. Comparison of incidence of death according to previous clinically relevant bleeding.
Bleeding confers the first occurrence of Bleeding Academic Research Consortium 2, 3, or 5 bleeding. Bleeding before a new cancer diagnosis was associated with increases in all‐cause death, cardiovascular death, and noncardiovascular death. The incidence rate per 1000 person‐years of all‐cause death was highest in patients diagnosed with new cancer after clinically relevant bleeding. The increased rate of all‐cause mortality in patients diagnosed with new cancer after clinically relevant bleeding was driven by increased noncardiovascular death. CV indicates cardiovascular.
Patients diagnosed with cancer within 6 months after AMI had a similar mortality risk to those diagnosed after 6 months (adjusted HR, 1.43 [95% CI, 0.96–2.13]; P=0.076). Precedent bleeding before a new cancer diagnosis was not associated with better survival, irrespective of the timing of the diagnosis of cancer (Figure S4, Table S4).
The incidence of all‐cause death during a median of 2.8 years after a new gastrointestinal and genitourinary cancer diagnosis according to previous gastrointestinal and genitourinary bleeding was also evaluated (Figure S5). The mortality risk after gastrointestinal cancer diagnosis with former gastrointestinal bleeding was higher compared with that without former gastrointestinal bleeding (21/44 [47.7%] versus 31/88 [35.2%]; adjusted HR, 4.05 [95% CI, 2.04–8.02]; P<0.001). The mortality rate after genitourinary cancer diagnosis was also numerically higher in patients with previous genitourinary bleeding than in those without previous genitourinary bleeding (7/13 [53.8%] versus 8/35 [22.9%]; adjusted HR, 2.79 [95% CI, 0.81–9.56]; P=0.103). Therefore, gastrointestinal bleeding or genitourinary bleeding preceding the detection of gastrointestinal cancer or genitourinary cancer was not associated with lower mortality risk.
DISCUSSION
We demonstrated that clinically significant bleeding after PCI for AMI was related to a new cancer diagnosis. More specifically, there was a strong association between gastrointestinal and genitourinary bleeding and gastrointestinal and genitourinary cancer, diagnosed primarily within 1 year after bleeding. Gastrointestinal or genitourinary bleeding preceding gastrointestinal or genitourinary cancer diagnosis had similar mortality compared with those diagnosed with gastrointestinal or genitourinary cancer but without precedent gastrointestinal or genitourinary bleeding.
Bleeding is frequent in patients with AMI undergoing dual antiplatelet therapy after PCI and may be the first manifestation of underlying cancer. Visible bleeding itself has been a strong indicator of bleeding‐related organ cancer, for which a prompt investigation is recommended. A previous study reported that a hospital‐based diagnosis of lower gastrointestinal bleeding was a robust clinical marker of prevalent gastrointestinal cancer, particularly colorectal cancer. 11 Hematuria is also associated with an increased risk of urinary tract cancer. It is recommended to evaluate hematuria even if the patient is receiving antithrombotic therapy. 12 , 13 Postmenopausal bleeding might capture uterine cancer with high specificity. 14
The relationship between clinically relevant bleeding after antithrombotic treatment and a new cancer diagnosis was also examined in patients with stable atherosclerosis, 15 patients with acute coronary syndrome, 16 and patients with atrial fibrillation. 17 , 18 Clinically relevant bleeding was quite frequent in these studies, occurring in 10% to 30% of the patients depending on the follow‐up period and antithrombotic treatment regimen. The association between bleeding and cancer detection was consistently strong, with HRs of >3.5 in these studies. Two studies 15 , 18 also assessed the relationship between organ‐specific bleeding after antithrombotic treatment and respective organ cancer. The studies consistently reported that both gastrointestinal and genitourinary tract bleeding were strongly and explicitly associated with a new cancer diagnosis of the individual organ. The HRs of gastrointestinal bleeding to gastrointestinal cancer were ≈13, and the HRs of genitourinary bleeding to genitourinary cancer were 18 and 32, similar to those in our study. The present study demonstrated the corresponding result in gastrointestinal and genitourinary bleeding after PCI in patients with AMI undergoing antiplatelet therapy for more extended periods. Considering the high mortality rates of patients with AMI, we performed competing risk analyses to adjust the patients who died before cancer diagnosis. In this study, most gastrointestinal cancer or genitourinary cancer diagnoses were made within the first year after gastrointestinal bleeding and genitourinary bleeding, which means there was also a solid correlation between gastrointestinal and genitourinary bleeding and the diagnosis of new gastrointestinal and genitourinary cancer.
Patients newly diagnosed with cancer in our cohort were older but more likely to have higher left ventricular function. Previous studies reported that patients with new cancer diagnoses following cardiovascular disease were older and more likely to have comorbidities. 16 , 17 , 19 Meanwhile, the patient's age did not affect the risk of new cancer after gastrointestinal and genitourinary bleeding in our subgroup analysis. Bleeding that led to a cancer diagnosis after myocardial infarction may be the only sign induced by the development of the cancer. It would be hard to predict the occurrence of cancer with baseline characteristics at AMI, and it would be crucial to examine actively if organ bleeding occurs.
Although undesirable, bleeding may unmask cancers that would otherwise potentially remain undiagnosed for a longer period. A new cancer diagnosis after an AMI was associated with mortality, severe bleeding, and adverse cardiovascular outcome. 20 Meanwhile, clinicians have long believed that an earlier diagnosis of cancer in patients with bleeding might lead to improved outcomes, depending on the site and stage of cancer at the time of diagnosis. 21 , 22 However, there have been no data demonstrating a survival benefit among patients diagnosed with cancer with prior bleeding compared with those diagnosed with cancer without prior bleeding. 23 Our study showed that cancer detection in patients with bleeding was not associated with an increased survival rate in either gastrointestinal or genitourinary cancer groups. Each bleeding and cancer might have independently affected the survival rate. Discontinuation of the antiplatelet agent, blood extravasation, and transfusion after bleeding can lead to mortality in patients with AMI. 2 Even if a cancer is diagnosed at an early stage by bleeding, the mortality rate may increase because of bleeding compared with cancer diagnosed without previous bleeding. The strength of our study is that we demonstrated that the long‐term mortality after gastrointestinal or genitourinary cancer detected by gastrointestinal bleeding or genitourinary bleeding was not different from those not detected by bleeding.
This study had the inherent limitation of its observational nature with registry data. The investigation of cancer in these patients was left to the attending physician's discretion. The present analysis did not adjudicate the specific causes of death according to bleeding‐related death with or without cancer and cancer‐related death without bleeding. The reported cancer was not diagnosed by systematic cancer research for all bleeding patients with a standard protocol. We reported a retrospective correlation between cancer and the first occurrence of BARC‐classified bleeding. Information on spontaneous bleeding or traumatic/iatrogenic bleeding was not collected. Patients without BARC 2, 3, or 5 bleeding had a short period (<3 years) to cancer diagnosis and will likely have less chance to observe potential bleeding than patients with a longer follow‐up time. However, we stress the importance of complete clinical examination for patients with clinically significant bleeding after AMI. We categorized the cause of death as cardiovascular and noncardiovascular deaths. Our study demonstrated that noncardiovascular death was higher in patients with new cancer diagnoses than those without cancer, suggesting that mortality might be related to cancer or bleeding. This study also lacked detailed information on cancer stages at diagnosis and cancer treatment, symptoms except bleeding before cancer diagnosis, and confounding mortality factors. Further investigation should involve the association between the early detection of other types of cancer without organ‐specific bleeding after PCI compared with cancer detected by organ‐specific bleeding (ie, gastrointestinal bleeding, genitourinary bleeding, and bronchopulmonary bleeding).
CONCLUSIONS
In patients with myocardial infarction, clinically significant bleeding was associated with an increase in new cancer diagnoses. Gastrointestinal and genitourinary tract bleeding were strongly related to new gastrointestinal or genitourinary cancer diagnoses. Gastrointestinal or genitourinary bleeding preceding gastrointestinal or genitourinary cancer diagnosis had similar mortality compared with those diagnosed with gastrointestinal or genitourinary cancer but without precedent bleeding. The detection of cancer by bleeding may not be associated with better survival in patients with AMI who underwent PCI.
Sources of Funding
None.
Disclosures
None.
Supporting information
Tables S1–S4
Figures S1–S5
Acknowledgments
The authors thank M. Park (Department of Biostatistics, Clinical Research Coordinating Center, Catholic Medical Center, The Catholic University of Korea) for statistical advice.
Supplemental Material is available at https://www.ahajournals.org/doi/suppl/10.1161/JAHA.122.026588
For Sources of Funding and Disclosures, see page 9.
References
- 1. Costa F, van Klaveren D, James S, Heg D, Räber L, Feres F, Pilgrim T, Hong MK, Kim HS, Colombo A, et al. Derivation and validation of the predicting bleeding complications in patients undergoing stent implantation and subsequent dual antiplatelet therapy (PRECISE‐DAPT) score: a pooled analysis of individual‐patient datasets from clinical trials. Lancet. 2017;389:1025–1034. doi: 10.1016/s0140-6736(17)30397-5 [DOI] [PubMed] [Google Scholar]
- 2. Genereux P, Giustino G, Witzenbichler B, Weisz G, Stuckey TD, Rinaldi MJ, Neumann FJ, Metzger DC, Henry TD, Cox DA, et al. Incidence, predictors, and impact of post‐discharge bleeding after percutaneous coronary intervention. J Am Coll Cardiol. 2015;66:1036–1045. doi: 10.1016/j.jacc.2015.06.1323 [DOI] [PubMed] [Google Scholar]
- 3. Baber U, Mehran R, Giustino G, Cohen DJ, Henry TD, Sartori S, Ariti C, Litherland C, Dangas G, Gibson CM, et al. Coronary thrombosis and major bleeding after PCI with drug‐eluting stents: risk scores from PARIS. J Am Coll Cardiol. 2016;67:2224–2234. doi: 10.1016/j.jacc.2016.02.064 [DOI] [PubMed] [Google Scholar]
- 4. Raposeiras‐Roubín S, Faxén J, Íñiguez‐Romo A, Henriques JPS, D'Ascenzo F, Saucedo J, Szummer K, Jernberg T, James SK, Juanatey JRG, et al. Development and external validation of a post‐discharge bleeding risk score in patients with acute coronary syndrome: the BleeMACS score. Int J Cardiol. 2018;254:10–15. doi: 10.1016/j.ijcard.2017.10.103 [DOI] [PubMed] [Google Scholar]
- 5. Serebruany VL, Cherepanov V, Cabrera‐Fuentes HA, Kim MH. Solid cancers after antiplatelet therapy: confirmations, controversies, and challenges. Thromb Haemost. 2015;114:1104–1112. doi: 10.1160/TH15-01-0077 [DOI] [PubMed] [Google Scholar]
- 6. Mehran R, Baber U, Steg PG, Ariti C, Weisz G, Witzenbichler B, Henry TD, Kini AS, Stuckey T, Cohen DJ, et al. Cessation of dual antiplatelet treatment and cardiac events after percutaneous coronary intervention (PARIS): 2 year results from a prospective observational study. Lancet. 2013;382:1714–1722. doi: 10.1016/s0140-6736(13)61720-1 [DOI] [PubMed] [Google Scholar]
- 7. Ibanez B, James S, Agewall S, Antunes MJ, Bucciarelli‐Ducci C, Bueno H, Caforio ALP, Crea F, Goudevenos JA, Halvorsen S, et al. 2017 ESC guidelines for the management of acute myocardial infarction in patients presenting with ST‐segment elevation: the task force for the management of acute myocardial infarction in patients presenting with ST‐segment elevation of the European Society of Cardiology (ESC). Eur Heart J. 2018;39:119–177. doi: 10.1093/eurheartj/ehx393 [DOI] [PubMed] [Google Scholar]
- 8. Collet JP, Thiele H, Barbato E, Barthélémy O, Bauersachs J, Bhatt DL, Dendale P, Dorobantu M, Edvardsen T, Folliguet T, et al. 2020 ESC guidelines for the management of acute coronary syndromes in patients presenting without persistent ST‐segment elevation. Eur Heart J. 2021;42:1289–1367. doi: 10.1093/eurheartj/ehaa575 [DOI] [PubMed] [Google Scholar]
- 9. Choi IJ, Lim S, Choo EH, Hwang BH, Kim CJ, Park MW, Lee JM, Park CS, Kim HY, Yoo KD, et al. Impact of intravascular ultrasound on long‐term clinical outcomes in patients with acute myocardial infarction. JACC: Cardiovasc Interv. 2021;14:2431–2443. doi: 10.1016/j.jcin.2021.08.021 [DOI] [PubMed] [Google Scholar]
- 10. Thygesen K, Alpert JS, Jaffe AS, Simoons ML, Chaitman BR, White HD, Thygesen K, Alpert JS, White HD, Jaffe AS, et al. Third universal definition of myocardial infarction. J Am Coll Cardiol. 2012;60:1581–1598. doi: 10.1016/j.jacc.2012.08.001 [DOI] [PubMed] [Google Scholar]
- 11. Viborg S, Sogaard KK, Farkas DK, Norrelund H, Pedersen L, Sorensen HT. Lower gastrointestinal bleeding and risk of gastrointestinal cancer. Clin Transl Gastroenterol. 2016;7:e162. doi: 10.1038/ctg.2016.16 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Nielsen M, Qaseem A. High value care task force of the American College of P. hematuria as a marker of occult urinary tract cancer: advice for high‐value care from the American College of Physicians. Ann Intern Med. 2016;164:488–497. doi: 10.7326/M15-1496 [DOI] [PubMed] [Google Scholar]
- 13. Norgaard M, Veres K, Ording AG, Djurhuus JC, Jensen JB, Sorensen HT. Evaluation of hospital‐based hematuria diagnosis and subsequent cancer risk among adults in Denmark. JAMA Netw Open. 2018;1:e184909. doi: 10.1001/jamanetworkopen.2018.4909 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Clarke MA, Long BJ, Del Mar MA, Arbyn M, Bakkum‐Gamez JN, Wentzensen N. Association of endometrial cancer risk with postmenopausal bleeding in women: a systematic review and meta‐analysis. JAMA Intern Med. 2018;178:1210–1222. doi: 10.1001/jamainternmed.2018.2820 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Eikelboom JW, Connolly SJ, Bosch J, Shestakovska O, Aboyans V, Alings M, Anand SS, Avezum A, Berkowitz SD, Bhatt DL, et al. Bleeding and new cancer diagnosis in patients with atherosclerosis. Circulation. 2019;140:1451–1459. doi: 10.1161/CIRCULATIONAHA.119.041949 [DOI] [PubMed] [Google Scholar]
- 16. Raposeiras‐Roubin S, Abu‐Assi E, Munoz‐Pousa I, Rossello X, Cespon‐Fernandez M, Melendo Viu M, Caneiro‐Queija B, Cobas‐Paz R, Bastos G, Iniguez‐Romo A. Usefulness of bleeding after acute coronary syndromes for unmasking silent cancer. Am J Cardiol. 2020;125:1801–1808. doi: 10.1016/j.amjcard.2020.03.023 [DOI] [PubMed] [Google Scholar]
- 17. Chang TY, Chan YH, Chiang CE, Lin YJ, Chang SL, Lo LW, Hu YF, Tuan TC, Liao JN, Chung FP, et al. Risks and outcomes of gastrointestinal malignancies in anticoagulated atrial fibrillation patients experiencing gastrointestinal bleeding: a nationwide cohort study. Heart Rhythm. 2020;17:1745–1751. doi: 10.1016/j.hrthm.2020.05.026 [DOI] [PubMed] [Google Scholar]
- 18. Raposeiras Roubin S, Abu Assi E, Barreiro Pardal C, Cespon Fernandez M, Munoz Pousa I, Cobas Paz R, Parada JA, Represa Montenegro M, Melendo Miu M, Blanco Prieto S, et al. New cancer diagnosis after bleeding in anticoagulated patients with atrial fibrillation. J Am Heart Assoc. 2020;9:e016836. doi: 10.1161/JAHA.120.016836 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Conen D, Wong JA, Sandhu RK, Cook NR, Lee IM, Buring JE, Albert CM. Risk of malignant cancer among women with new‐onset atrial fibrillation. JAMA Cardiol. 2016;1:389–396. doi: 10.1001/jamacardio.2016.0280 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Velders MA, Hagstrom E, James SK. Temporal trends in the prevalence of cancer and its impact on outcome in patients with first myocardial infarction: a Nationwide study. J Am Heart Assoc. 2020;9:e014383. doi: 10.1161/JAHA.119.014383 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Roe MT, Strickler J. Addressing the conundrum of bleeding and cancer detection with antithrombotic therapies for chronic atherosclerotic cardiovascular disease. Circulation. 2019;140:1460–1462. doi: 10.1161/CIRCULATIONAHA.119.042875 [DOI] [PubMed] [Google Scholar]
- 22. Andreotti F, Maggioni AP. Cancer unmasked by bleeding during anticoagulant therapy: when a problem may become an opportunity. Eur Heart J. 2020;43:e45–e47. doi: 10.1093/eurheartj/ehaa164 [DOI] [PubMed] [Google Scholar]
- 23. Mauri L, Elmariah S, Yeh RW, Cutlip DE, Steg PG, Windecker S, Wiviott SD, Cohen DJ, Massaro JM, D'Agostino RB Sr, et al. Causes of late mortality with dual antiplatelet therapy after coronary stents. Eur Heart J. 2016;37:378–385. doi: 10.1093/eurheartj/ehv614 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Tables S1–S4
Figures S1–S5