Ageing populations, improved diagnostic accuracy, disease awareness and therapy effectiveness have led to an increase in cancer incidence and prevalence.1 Cancer and cardiovascular (CV) disease share common risk factors, and some cancer therapies are known to increase the likelihood of coronary artery disease (CAD).2 As a result, it is no surprise that patients with cancer comprise a significant proportion of patients undergoing percutaneous coronary intervention (PCI), ranging from 2% to 9% in multiple observational studies.3–6 Thus, the issue of revascularisation in patients with cancer is increasingly relevant.
This review on revascularisation in patients with cancer focuses primarily on three areas: what evidence is available; whether patients with cancer should undergo revascularisation; and what specific factors should be considered when performing revascularisation in patients with cancer. Details pertaining to the pathophysiology of CAD in patients with cancer, diagnosis, stratification and medical therapy (other than revascularisation-specific topics such as antithrombotic therapy) fall outside the scope of this paper and are either not covered or are touched upon only briefly.
We conducted our initial search using PubMed/MEDLINE, Embase, Cochrane and Scopus. Every possible combination of the following two groups of words was used: cancer, oncology, or oncologic; and coronary revascularisation, PCI, coronary artery bypass surgery. Original research or review papers that addressed the issue of revascularisation in patients with cancer were selected for analysis. Final inclusion in this review was based on the authors’ critical appraisal of quality and pertinence.
Available Evidence for Revascularisation in Patients with Cancer
Currently available evidence remains limited. Almost all studies are observational, and only one study was derived from clinical trials, albeit a non-prespecified post hoc subanalysis. Studies are often retrospective and do not always differentiate between past or active cancer. Technical details regarding PCI are lacking or may not represent contemporary practices because of the historical nature of the reports. Figure 1 summarises this section, showing current evidence-based factors to consider when contemplating revascularisation in patients with cancer.
General Impact of Revascularisation on Outcomes
Coronary Artery Bypass Graft Surgery
Evidence for coronary artery bypass graft (CABG) in patients with cancer is observational and particularly limited. The most extensive data source is the US National Inpatient Sample (NIS). A propensity score analysis of more than 100,000 patients up to 2015 showed that patients with cancer had nearly threefold higher rates of major bleeding and an increased incidence of stroke, underscoring their greater frailty.7 However, contrary to common concerns, the study did not confirm an elevated risk of mortality, given that in-hospital mortality rates were comparable between groups.7
The SWEDEHEART registry data, comprising all CABG patients in Sweden from 1997 to 2015, demonstrated that the prevalence of cancer nearly quadrupled over time, from 4% to almost 16%. Patients with cancer were older and had more frequent CV and non-CV comorbidities. Cancer was an independent predictor of long-term mortality, driven only by cancer-related mortality. CV mortality was similar between groups, suggesting that the CV mortality effect of CABG in patients with cancer may be preserved.8 These findings suggest that CABG may be a viable option for these patients and should not be routinely disregarded.
Percutaneous Coronary Intervention
A number of studies have addressed the outcomes of PCI in patients with cancer, in general. Many do not specifically address cancer details or the cardiovascular clinical context. Nevertheless, these studies offer a broad overview of the clinical course of these patients and the potential prognostic impact of PCI.
Regarding ischaemic outcomes, in a single-centre retrospective study in Japan, Takeuchi et al. reported an increased rate of MI in patients with cancer after PCI, but cancer was not an independent predictor of such events.9 Another single-centre retrospective study from Japan found cancer to be an independent predictor of target lesion revascularisation at 1 year.10 A large Kyoto registry reported higher rates of stent thrombosis.11 Analysis of a large propensity-score-matched 10-year cohort from Mayo Clinic also found higher rates of MI and repeat revascularisation in patients with cancer, together with increased rates of stent thrombosis.12 However, when stratified by dual antiplatelet therapy (DAPT) score, the impact of cancer on ischaemic outcomes was attenuated, albeit not entirely eliminated. Importantly, this score was predictive of ischaemic events in both patients with and without cancer, suggesting that it may be of use in this population.12 In contrast, two long-term observational studies, from Duke University Medical Center and the Bern PCI registry, found no increased rates of MI, repeat revascularisation or stent thrombosis in patients with cancer, nor have they found cancer to be a predictor of such events.13,14 A recent patient-level subanalysis of four high-bleeding risk drug-eluting stent (DES) PCI trials found no increased risk of MI, stroke, stent thrombosis or repeat revascularisation.15 Taken together, these studies suggest that cancer may not substantially increase the risk of ischaemic outcomes in patients with cancer, although the well-known higher CV morbidity burden, and perhaps the direct effect of cancer or related therapies, may explain the higher rates of ischaemic events observed in some of the studies.
In contrast to ischaemic outcomes, studies have consistently reported a generally higher risk of bleeding events following PCI in patients with cancer. Whether nationwide, local or international, western or eastern, increased bleeding seems to be a ubiquitous finding in observational studies reporting this outcome.4,11,12,14,16,17 Importantly, when severity of bleeding is reported, patients with cancer also have greater rates of significant bleeding (Bleeding Academic Research Consortium type 2 or greater).12 Last, in a patient-level subanalysis of four high-risk DES PCI trials, an increased risk of patient bleeding, including major bleeding, was found in patients with cancer.15 The same finding was observed in a validation study of the Academic Research Consortium high bleeding risk (ARC-HBR) criteria, focusing specifically on patients with cancer.18 Bleeding risk is strongly influenced by cancer-specific factors and is further discussed in the following section.
The most widely studied outcome in patients with cancer undergoing PCI has been mortality. Most studies report higher mortality rates after PCI in patients with cancer as well as CV death and mortality in hospital or at any time.4,6,11–14,17,19–21 A recent meta-analysis reported similar findings.2 Importantly, cancer diagnosed after PCI also has a negative impact on mortality.13 The aforementioned patient-level subanalysis of four high bleeding risk (HBR) DES PCI clinical trials also found an increased risk of mortality in this subset.15
This does not mean that PCI should be withheld in patients with cancer. Indeed, in a large retrospective study of more than 30 million US patients, with machine learning-augmented propensity score analysis, Monlezun et al. found a reduced mortality risk (OR 0.82; 95% CI [0.75–0.89]; p<0.001) for any inpatients receiving PCI, regardless of oncological or cardiovascular context. Furthermore, while hospital costs for patients with cancer were higher than in the general population, PCI reduced the overall cost of hospital stay.4,22 This could only be explained by a favourable impact on clinical outcomes and its consequent reduction in future health expenditure, otherwise PCI would have simply added to the treatment cost in patients with cancer.
Last, a few studies address other outcomes in brief, having also reported increased risks of hospital readmission and heart failure hospitalisation following PCI in patients with cancer.9,11,16 This is not likely to be related to PCI directly (nor do studies establish any causality, either positive or negative), but instead to be related to the baseline comorbidity burden of patients with cancer, as well as the direct effect of cancer and oncology treatments.
In summary, with regards to outcomes of patients with cancer undergoing PCI in general, currently available evidence is broadly supportive of revascularisation, in the sense that an improvement of outcomes can be expected. PCI should thus not be generally withheld in patients with cancer, but rather routinely considered. However, clinicians should be wary of an increased risk of adverse events in this subpopulation and therefore be particularly careful regarding patient selection and follow-up. Management strategies, both interventional and medical, should be meticulous and highly individualised.
Revascularisation in Acute Coronary Syndrome
Numerous studies report that patients with cancer with acute coronary syndrome (ACS) less often received treatment from a cardiologist, invasive management or PCI, and more often received bare metal stents (BMS). They also often received less intensive medical therapy, be it neurohormonal blockade drugs, P2Y12 inhibitors or statins. Interestingly, these findings have been observed in studies from all over the world, such as the US, Canada, Switzerland, the UK, Sweden, the Netherlands, Portugal and international studies including countries from both North and South America, Europe and Asia.3,6,17,21,23–26
Higher rates of mortality after ACS have also been reported in patients with cancer.3,5,6,17,21,23,24,26–28 In-hospital mortality and early mortality (<30 days) are often cardiovascular, whereas later mortality is more likely to be cancer-related.24,26,28
Higher rates of reinfarction, heart failure and bleeding have also been reported in patients with cancer with ACS.6,16,17,21,23,25
The aforementioned findings raise a fundamental question: was the worse prognosis that was observed in the aforementioned studies the result of suboptimal treatment offered to patients with cancer, the result of cancer itself, or both? A number of studies have attempted to isolate the effect of therapy in ACS patients, as well as the effect of cancer, and offer some insight.
Several studies have addressed the impact of PCI in ACS. Monlezun et al. reported a reduced mortality rate in patients receiving PCI for every subset of ACS, including unstable angina.22 The magnitude of benefit was greater for non-ST-elevation MI (NSTEMI) and especially high for ST-elevation MI (STEMI). Importantly, patients with active malignancy derived the most benefit. A retrospective study of NSTEMI patients at MD Anderson Cancer Center reported improved survival in patients undergoing invasive management, especially if undertaken in the first 72 hours after presentation.27 A large US NIS propensity-matching retrospective study of STEMI patients with active cancer reported that patients with cancer derive similar benefits of primary PCI to those in patients without cancer, including lower rates of in-hospital mortality, cardiac complications and stroke, across multiple cancer types (lung, breast, prostate, colon and haematological).3 The authors also did not report an increased risk of bleeding in patients receiving primary PCI.3 Similar findings were reported in the Portuguese registry of Acute Coronary Syndromes, regardless of ACS type.6
Some groups have also reported on the impact of optimal medical therapy in patients with cancer and ACS. A multicentre international study from Europe, North and South America and Asia involving patients undergoing PCI in the context of ACS, and consisting predominantly of NSTEMI patients, reported that all elements of medical therapy (antithrombotic therapy, statins and neurohormonal blockade heart failure drugs) were associated with reduced rates of death and MI. There was also a trend towards reduced rates of death or MI in patients treated with DES, but not in patients treated with plain old balloon angioplasty (POBA).17
Some studies have also attempted to assess the prognostic effect of cancer in ACS patients. Yousif et al. conducted a multicentre prospective registry (in Switzerland, UK and Bahrain) with independent event adjudication, comprising the full spectrum of ACS (55% STEMI, 6.6% unstable angina) patients.29 Importantly, patients were treated according to guidelines and therefore the suboptimal treatment pattern observed in the aforementioned studies was not observed.3,6,17,21,23–26 Furthermore, only patients with active cancer were eligible. They reported higher 1-year mortality, cardiovascular mortality, revascularisation and bleeding events even after adjustment for covariates, suggesting that active cancer is independently associated with a worse prognosis, even if optimal therapy is pursued.29 The SWEDEHEART registry identified cancer as an independent predictor of adverse mortality, recurrent MI, heart failure and major bleeding.21
Last, one more factor is worth considering. Risk stratification scores are often not validated for patients with cancer. In the setting of non-ST-segment elevation-ACS, the Global Registry of Acute Coronary Events (GRACE) score is often used for risk stratification and selection of the timing of intervention. Importantly, although the GRACE score is higher in patients with cancer, suggesting that an earlier invasive strategy should be routinely used, it is worth noting that several cancer-specific factors should be considered prior to intervention, given the higher risk of bleeding and complications in patients with cancer.29,30 This often requires a thorough patient history and collection of data, which is not always feasible in <24 hours. Furthermore, some authors have suggested that the GRACE score performs poorly in patients with cancer, underestimating mortality.31
Taken together, three major findings may be drawn from currently available evidence in the setting of ACS: patients with cancer receive suboptimal treatment, with lower rates of invasive management and optimal medical therapy; both revascularisation and medical therapy seem to retain the ability to reduce adverse outcomes in patients with cancer (including mortality) and should thus not be routinely withheld, but instead considered; and even with optimal treatment, cancer is an independent predictor of adverse prognosis, be it cardiovascular events, bleeding or mortality (early mortality following ACS is often cardiovascular, late mortality is often oncological).
Consequently, multiple expert opinion papers and international guidelines advocate for an invasive management and PCI in patients with cancer, limiting a conservative approach for low-risk cases or life expectancy <6 months.32–35
Revascularisation in Chronic Coronary Syndrome
To the best of our knowledge, no dedicated studies of revascularisation in chronic coronary syndrome (CCS) in patients with cancer have been published. However, it is worth noting that many of the studies mentioned here do not always differentiate outcomes according to clinical presentation. Currently available evidence for the general population in major clinical trials suggests that although PCI has the potential to reduce the rate of spontaneous MI, there is no impact on overall mortality.36,37
As a result, a routinely conservative strategy may be better suited for CCS in most patients with cancer, as has been suggested in an expert position paper.33 However, a number of specific scenarios may apply in cardio-oncology in which major surgery may be required, such as when revascularisation is deemed necessary. Likewise, in the case of severe or refractory angina symptoms, patients with cancer should not be denied revascularisation, given the well-established effect of PCI in symptom relief.36,38 Current international guidelines suggest that patients with cancer should be managed similarly to the general population, taking into consideration cancer-specific factors.34,39
Impact of Cancer-related Factors
Although it is common to refer to patients with cancer as one group, the expression is inherently flawed from a clinical and scientific perspective, given the significant heterogeneity with regards to primary tumour location, stage and active versus past cancer. These are all relevant factors that significantly alter patient outcomes and which should be taken into consideration when planning revascularisation. The issue of radiotherapy-induced CAD also merits particular consideration.
Published studies have focused more specifically on lung, colon, prostate, breast and haematological cancers.
Lung cancer has repeatedly been associated with worse outcomes, with increased in-hospital, 30-day and 1-year mortality.4,5,22,28 Furthermore, bleeding, cardiovascular or cerebrovascular complications and readmissions have also been higher in lung patients with cancer.4,5,16 Colon cancer carries the highest risk of bleeding.4,5,22,28 Increased rates of complications and hospital readmission due to acute MI or heart failure have also been reported.4,16,25
Prostate cancer has a somewhat better profile, with an increased risk of bleeding but with a relatively good prognosis.4 Breast cancer has the best prognosis of common solid tumours, with the lowest rates of mortality, bleeding, complications or readmissions.4,5,22,28
Haematological malignancies have been associated with a higher risk of in-hospital death, bleeding and procedure-related complications.19,28 Importantly, however, a prospective observational study with independent event adjudication reported that solid tumours have a higher rate of major adverse cardiovascular or cerebrovascular events, as well as increased all-cause and CV mortality, when compared with haematological malignancies.29 The exact outcome of revascularisation in patients with haematological malignancies is not yet clear, especially when compared with solid tumours as a whole.
Active cancer carries a worse prognosis, regardless of primary tumour location.4,5,16,25 This may be both a consequence of the disease itself and also of the effect of cancer treatment. In contrast, historical cancer (i.e. no active malignancy and cancer diagnosis at least 2 years previously) or cancer diagnosed more than 6 months previously is far less impactful.26,40 With regards to staging, the presence of metastases is the central feature. Whatever the primary tumour location, metastasis is associated with increased mortality, complications, major cardiovascular or cerebrovascular events and bleeding.4,5,16
Radiotherapy-induced CAD often manifests in patients who have been treated for (left) breast cancer, Hodgkin lymphoma or other mediastinal cancers. A period >5–10 years often elapses between therapy and disease onset. Often, CAD may be diffuse and especially severe in proximal segments, including ostia.41 Patients undergoing PCI with DES have outcomes similar to the general population, although those treated with BMS or POBA have higher mortality rates.42,43 In the case of CABG, radiotherapy-related atherosclerosis of the mammary arteries and impaired tissue healing after sternotomy are critical considerations.41
Technical Considerations for Revascularisation in Patients with Cancer
Antiplatelet Regimen and Bleeding
Arguably, the choice of antiplatelet regimen is the most pressing medical therapy issue in patients undergoing revascularisation, especially PCI, given that bleeding is of particular concern in patients with cancer. Anaemia and/or thrombocytopenia, and coagulopathy are very common. Furthermore, patients with primary tumour or metastases in the gastrointestinal or genitourinary tract constitute a subset particularly prone to bleeding.
Shortening of DAPT duration is of particular relevance in patients with cancer. Major trials have demonstrated the safety and efficacy of a minimum of 1 month of DAPT in HBR patients, many of whom had active cancer as an inclusion criterion.44–47 Of these, the MASTER DAPT trial was the largest to date, encompassing 275 patients (5.3% of the included patients) with active cancer and 802 patients (15.4% of the consented patients) with a history of cancer. No differences were noted with regards to ischaemic events, and major bleeding was reduced by 36%.47
Recently, a patient-level subanalysis of four HBR DES PCI trials was published.15 The authors selected patients with a cancer diagnosis up to 3 years following intervention, and 1-month DAPT was used. No increased rates of any ischaemic outcomes were noted in patients with cancer, who nonetheless had higher mortality and bleeding. Importantly, the included trials had a very high proportion of ACS patients, meaning that this strategy is probably appropriate for this subset of patients as well, for whom shortening DAPT duration would theoretically be more concerning.15
As a result, while trials directly assessing DAPT regimens solely in patients with cancer are lacking, currently available evidence supports DAPT shortening to 1 or 3 months. This decision should, however, be informed by the patient’s bleeding risk, which is commonly assessed by risk scores.
The ARC-HBR criteria may be used for estimating bleeding risk in patients undergoing PCI.32,48 These criteria consider cancer diagnosed in the last 12 months as one of the major bleeding criteria. Very recently, an adapted version of these criteria was validated in a large cohort of patients with cancer (active or in remission) spanning a decade (2012–2022) at Mount Sinai Hospital.18 The authors removed cancer as a criterion and retained the remainder of the criteria. A total of 56.9% were still classified as HBR. These patients had a risk of bleeding more than twice as high as the remainder, driven by periprocedural bleeding. Furthermore, such patients also had an increased risk of major adverse cardiac and cerebrovascular events and death.18 This suggests that the modified ARC-HBR criteria may be used in patients with cancer.
The ARC-HBR criteria may also be used as a trade-off model and have recently been validated in cancer patients.49 Using this approach, patients are categorised into three groups: higher risk of bleeding, higher risk of ischaemic events, or comparable bleeding/ischaemic risk. As expected, cancer patients were most often classified as higher bleeding risk.15
The PRECISE-DAPT score has also been adapted to patients with cancer. Using a very large cohort of UK STEMI patients, Dafaalla et al. added cancer as a binary factor to this score.50 The authors found that cancer patients were often reclassified as high-bleeding risk (from 65.5% to 94%), but the score’s discrimination ability was retained (C-statistic from 0.60 to 0.64).
Considering that currently available evidence is generally favourable for shortening of DAPT duration (given the resulting improvement in bleeding without a signal of significantly increased thrombotic risk) and that cancer patients are most often classified as HBR with multiple scoring methods, a DAPT duration of 1–3 months is probably appropriate for most patients with cancer. Indeed, multiple scientific societies and International Cardio-Oncology Society/European Society of Cardiology (ICOS/ESC) guidelines suggest the shortening of DAPT following PCI in patients with cancer to 1–3 months, even in patients with ACS.32,33,34
With regards to the choice of P2Y12 inhibitor, clopidogrel has a more favourable bleeding profile than ticagrelor or prasugrel and has a lower potential for interactions. This was the preferred P2Y12 inhibitor in HBR trials, in which more potent agents were used infrequently.13–16,44–47 As a result, clopidogrel should be the preferred agent for PCI in cancer patients in most cases, as also indicated in current guidelines.32–34
With regards to anaemia and thrombocytopenia, while specific evidence is lacking, a number of expert opinion papers from national or international scientific societies and the ICOS/ESC cardio-oncology guidelines suggest a minimum platelet count of 50,000 cells/µl for CABG and 30,000 cells/µl for PCI, considering platelet transfusion at a threshold of 20,000 cells/µl.32,33,34,51 For anaemia, a restrictive threshold of 7 g/dl is often the standard for red blood cell transfusion. However, as discussed above, revascularisation often carries a higher risk of bleeding in patients with cancer. Expert opinion papers from interventional scientific societies suggest that a higher level of 9–10 g/dl is advised for revascularisation, particularly if CABG, non-radial PCI or complex PCI is expected.33,51 They highlight the importance of identifying and correcting the predisposing factors to anaemia as much as possible, including iron, B12 and/or folate supplementation. Antiplatelet agents should be withheld when platelet count is critically low: below 10,000 cells/µl for acetylsalicylic acid, below 30,000 cells/µl for clopidogrel, and below 50,000 cells/µl for prasugrel or ticagrelor.32–34
Other secondary or tertiary measures, such as statins, are of course of great importance in patients with cancer and every effort should be made to optimise cardiovascular risk factor modification. These patients are more often at higher cardiovascular risk than the general population and more frequently have an indication for statin use.52 However, a detailed review of such measures falls well outside the scope of this review paper.
Technical Factors during Percutaneous Coronary Intervention
All scientific societies stress the importance of meticulous technical factors during PCI, using every resource at the operator’s disposal to reduce the risk of adverse events.32–34 As a result, radial access is strongly recommended given its strong evidence base.53,54 Intra-coronary imaging has been shown to reduce the risk of major adverse cardiovascular events, as evidenced by a recent large meta-analysis of 22 trials and almost 16,000 patients, and is also routinely recommended.33–35,39,55 While, as mentioned, BMS have historically often been used more often in patients with cancer (and even recommended for such cases), currently available recommendations support DES as standard for all patients.33–35,39,51
Figure 2 summarises the technical factors to be considered with regard to revascularisation in patients with cancer.
Conclusion
The issue of revascularisation in patients with cancer has grown steadily over the last few decades. Its relevance is only expected to increase. Currently available evidence and recommendations are increasingly favouring revascularisation in patients with cancer, especially for ACS, taking into consideration cancer-specific factors. The old practice of commonly withholding invasive management of these patients no longer seems appropriate. Notwithstanding, dedicated randomised evidence regarding this topic is very much needed, given both the scarcity and observational nature of currently available data. Scientific societies and industry should prioritise this topic for research.