Article Text

PDF

Cardiac imaging and non-invasive testing
Multislice computed tomography coronary angiography for risk stratification in patients with an intermediate pretest likelihood
  1. J M van Werkhoven1,2,
  2. O Gaemperli3,4,
  3. J D Schuijf1,
  4. J W Jukema1,2,
  5. L J Kroft5,
  6. S Leschka6,
  7. H Alkadhi6,
  8. I Valenta4,
  9. G Pundziute1,
  10. A de Roos5,
  11. E E van der Wall1,2,
  12. P A Kaufmann4,7,
  13. J J Bax1
  1. 1
    Department of Cardiology, Leiden University Medical Centre, Leiden, The Netherlands
  2. 2
    The Interuniversity Cardiology Institute of the Netherlands, Utrecht, The Netherlands
  3. 3
    Department of Cardiology, University Hospital Zurich, Zurich, Switzerland
  4. 4
    Department of Nuclear Medicine, University Hospital Zurich, Zurich, Switzerland
  5. 5
    Department of Radiology, Leiden University Medical Centre, Leiden, The Netherlands
  6. 6
    Institute of Diagnostic Radiology, University Hospital Zurich, Zurich, Switzerland
  7. 7
    Zurich Integrative Human Physiology, University of Zurich, Zurich, Switzerland
  1. Correspondence to Dr J van Werkhoven, Department of Cardiology, Leiden University Medical Centre, Albinusdreef 2, 2333 ZA Leiden, The Netherlands; j.m.van_werkhoven{at}lumc.nl

Abstract

Objectives: To assess whether multislice computed tomography coronary angiography (MSCTA) may be useful for risk stratification of patients with suspected coronary artery disease (CAD) at intermediate pretest likelihood according to Diamond and Forrester.

Design and patients: MSCTA images were evaluated for the presence of significant CAD in 316 patients with suspected CAD (60% male, average (SD) age 57 (11) years) and an intermediate pretest likelihood according to Diamond and Forrester. Patients were followed up to determine the occurrence of an event.

Main outcome measures: A combined end point of all-cause mortality, non-fatal infarction and unstable angina requiring revascularisation.

Results: Significant CAD was seen in 89 patients (28%), whereas normal MSCTA or non-significant CAD was seen in the remaining 227 (72%) patients. During follow-up (median 621 days (25–75th centile 408–835) an event occurred in 13 patients (4.8%). The annualised event rate was 0.8% in patients with normal MSCT, 2.2% in patients with non-significant CAD and 6.5% in patients with significant CAD. Moreover, MSCTA remained a significant predictor (p<0.05) of events after multivariate correction (hazard ratio  =  3.460 (95% CI 1.142 to 10.480).

Conclusions: The results suggest that in patients with an intermediate pretest likelihood, MSCTA is highly effective in re-stratifying patients into either a low or high post-test risk group. These results further emphasise the usefulness of non-invasive imaging with MSCTA in this patient population.

Statistics from Altmetric.com

Coronary artery disease (CAD) is a leading cause of mortality and morbidity world wide. The diagnosis and management of this disease is increasingly dependent on non-invasive imaging strategies. With the introduction of multislice computed tomography coronary angiography (MSCTA) non-invasive assessment of coronary anatomy has become possible, allowing early identification of atherosclerosis. MSCTA has a high diagnostic accuracy for the detection of significant CAD (⩾50% luminal narrowing) on conventional coronary angiography,1 2 3 4 5 6 7 and may be particularly useful for diagnosis in patients with an intermediate pretest likelihood for significant CAD.8 Although the prognostic value of MSCTA has been evaluated in previous studies,7 9 10 11 12 13 14 15 16 no studies have specifically examined the target population for MSCTA. The purpose of this study, therefore, was to assess if MSCTA may be useful for risk stratification in patients with suspected CAD and an intermediate pretest likelihood.

Methods

The study population comprised 331 patients with suspected CAD and an intermediate pretest likelihood. Patients were clinically referred for further cardiac assessment as part of a current study protocol examining the prognostic value of MSCTA. From this prospective registry, results dealing with the incremental prognostic value of MSCTA over myocardial perfusion imaging have been recently published.13

Baseline clinical demographic values were recorded from the electronic patient file based on a doctor’s documented history. Symptoms were classified as typical angina, atypical angina, non-anginal chest pain and asymptomatic. Typical anginal chest pain was defined as a combination of (a) discomfort in the anterior chest, neck, shoulders, jaw, or arms; (b) precipitated by physical exertion or psychic stress; and (c) relieved by rest or nitroglycerin within minutes. Atypical chest pain was defined as chest pain with two of these three factors and non-anginal chest pain was defined as chest pain with fewer than two of these three factors.17 Pretest likelihood was defined according to Diamond and Forrester criteria, which are based on age, gender and symptomatic status.18 Intermediate likelihood was defined as a pretest likelihood between 13.4% and 87.2%. In addition, asymptomatic diabetic patients were also classified as having an intermediate pretest likelihood according to the increased prevalence of CAD and increased risk of events in this population.19 20 21 Exclusion criteria were cardiac arrhythmias, renal insufficiency (defined as a glomerular filtration rate <30 ml/min), known hypersensitivity to iodine contrast media and pregnancy. In addition, patients with an uninterpretable MSCTA examination were excluded from further analysis. The study was approved by the local ethics committee in both participating centres.

Patients were scanned using a 64-slice CT scanner (Aquillion64, Toshiba Medical Systems, Tokyo, Japan; General Electrics LightSpeed VCT, Milwaukee, Wisconsin, USA; or Sensation64, Siemens, Forchheim, Germany). For each patient, the heart rate and blood pressure were monitored before the scan. In the absence of contraindications, patients with a heart rate exceeding the threshold of 65 bpm were given β-blocking drugs (50–100 mg metoprolol, oral or 5–10 mg metoprolol, intravenous). The contrast-enhanced helical scan was performed using 80–140 ml of non-ionic contrast agent (Xenetix 300; Iomeron 400; or Iodixanol 370) administered at a flow rate of 3.5–5 ml/s followed by a bolus of saline flush (30–50 ml at 3.5–5 ml/s). All scan parameters have been previously published.22 23 24

Datasets were reconstructed from the retrospectively gated raw data. Images were reconstructed with an effective slice thickness of 0.3, 0.5 or 0.625 mm with the Toshiba, Siemens and GE systems, respectively. Coronary arteries were evaluated using the reconstruction dataset with the least motion artefacts, typically an end-diastolic phase. Post-processing was performed on dedicated workstations (Vitrea2, Vital Images, USA; Advantage, GE Healthcare, USA; Leonardo, Siemens, Germany). The interpretation of MSCTA angiograms was performed in a standardised manner using the axial slices, curved multiplanar reconstructions and maximum intensity projections. MSCTA examinations were scored for each patient by two experienced observers in each centre, based on the maximum luminal diameter stenosis. Discrepancies in interpretation were resolved by consensus. Normal MSCTA was defined as completely normal anatomy, non-significant stenosis was defined as the presence of luminal narrowing with a maximal luminal diameter stenosis <50% and significant stenosis was defined as the presence of an atherosclerotic lesion exceeding the threshold of ⩾50% luminal narrowing. The effective dose of the calcium score (CS) and MSCTA scans was estimated from the product of the dose–length product and an organ weighing factor (k = 0.014 mSv × (mGy × cm)−1) for the chest as the investigated anatomical region.25

Patient follow-up data were gathered using clinical visits or standardised telephone interviews. The following events were regarded as clinical end points: all-cause mortality, non-fatal myocardial infarction and unstable angina requiring revascularisation. Two combined end points were used: a combined end point of all events; and a combined hard end point including only all-cause mortality and non-fatal myocardial infarction. Non-fatal infarction was defined based on criteria of typical chest pain, elevated cardiac enzyme levels and typical changes on the ECG. Unstable angina was defined according to the European Society of Cardiology guidelines as acute chest pain with or without the presence of ECG abnormalities, and negative cardiac enzyme levels.26 Patients with stable complaints undergoing an early elective revascularisation within 60 days after MSCTA were excluded from the survival analysis.

Continuous variables were expressed as mean and standard deviation, and categorical baseline data were expressed in numbers and percentages. Cox regression analysis was used to determine the prognostic value of MSCTA. First, univariate analysis of baseline clinical variables and MSCTA variables was performed using a composite end point of all-cause mortality, non-fatal infarction and unstable angina requiring revascularisation. For each variable a hazard ratio with a 95% confidence interval (95% CI) was calculated. Finally, a multivariate model was created using backward stepwise selection to assess the independent predictive value of MSCTA corrected for baseline clinical variables. At each step, the least significant variable was discarded from the model, until all variables in the model reached a p value <0.25.

Cumulative event rates for MSCTA were obtained by the Kaplan–Meier method, and the survival curves were compared using the log-rank test. Statistical analyses were performed using SPSS software (version 12.0, SPSS, Chicago, Illinois, USA). A p value <0.05 was considered statistically significant.

Results

Patient characteristics

The study population, derived from our prospective registry,13 consisted of 331 patients with suspected CAD and an intermediate pretest likelihood. The MSCTA examination was uninterpretable in 15 patients (4.5%). Reasons for uninterpretability were the presence of motion artefacts, increased noise due to high body mass index, and breathing. After exclusion of these patients, a total of 316 patients remained for analysis. The average (SD) age of the study cohort was 57 (11) years and 60% of patients were men. Patients presented with asymptomatic diabetes (25%) non-anginal chest pain (22%), atypical chest pain (48%) and typical chest pain (5%). Table 1 presents a complete overview of the baseline characteristics of the study population.

Table 1

Patient characteristics

MSCT results

During MSCTA image acquisition, an average (SD) heart rate of 63 (10) bpm was recorded. MSCTA was classified as normal in 85 (27%) of patients. Atherosclerosis was detected in the remaining 231 (73%) patients, classified as non-significant CAD (<50% luminal narrowing) in 142 (45%) and significant CAD (⩾50% luminal narrowing) in 89 (28%) patients (fig 1). The estimated average radiation dose for the coronary angiography protocol was 18.1 (5.9) mSv and the estimated radiation dose for the CS protocol was 1.4 (0.6) mSv.

Figure 1

Pie chart illustrating the distribution of multislice computed tomography coronary angiography findings. CAD, coronary artery disease.

Follow-up results

In total, 26 (8%) patients were lost to follow-up and 21 (7%) patients underwent early revascularisation. In the remaining 269 patients the median follow-up time achieved was 621 days (25–75th centile 408–835). During the follow-up period an event occurred in 13 patients (5%). All-cause mortality was reported in five patients (2%), whereas non-fatal myocardial infarction occurred in three patients (1%), and five patients (2%) were revascularised owing to unstable angina pectoris. In the excluded patients with an early revascularisation, myocardial infarction occurred in one of 21 patients.

Event rates

In patients with a normal MSCTA an annualised event rate (annualised event rate for hard events between parentheses) of 0.8% (0%) was seen, while in patients with atherosclerosis (non-significant and/or significant CAD) the annualised event rate was 3.5% (2.6%). Significant CAD (⩾50% luminal narrowing) on MSCTA resulted in an event rate of 6.5% (4.6%). Figure 2 shows the total event rates (all-cause mortality, non-fatal myocardial infarction and unstable angina requiring revascularisation) and the hard event rates (all-cause mortality and non-fatal myocardial infarction) in all patients and stratified according to MSCTA.

Figure 2

The annualised event rate in all patients and stratified according to MSCTA results. (A) Bar graph for all events (all-cause mortality, non-fatal myocardial infarction and unstable angina requiring revascularisation); (B) bar graph for hard events (all-cause mortality and non-fatal myocardial infarction). CAD, coronary artery disease; MSCTA, multislice computed tomography coronary angiography.

Survival analysis

Table 2 lists the baseline univariate predictors of events. Significant CAD (⩾50% luminal narrowing) was the best univariate predictor of events with a hazard ratio of 3.9 (95% CI 1.3 to 11.7). After correcting for smoking in a multivariate model this variable remained an independent predictor of events (table 2). The Kaplan–Meier survival analysis in fig 3 further illustrates the usefulness of ⩾50% luminal narrowing on MSCTA as a cut-off point for risk stratification. A significant difference (log-rank p value = 0.008) in survival was observed between patients with normal or non-significant CAD and patients with significant CAD (⩾50% luminal narrowing) on MSCTA. Similar results were obtained when using a combined hard end point of all-cause mortality and non-fatal myocardial infarction (log-rank p value = 0.008; fig 3).

Figure 3

Kaplan–Meier survival curves in patients with normal or non-significant CAD (<50% luminal narrowing) compared with patients with significant CAD (⩾50% luminal narrowing). (A) Kaplan–Meier curves for all events (all-cause mortality, non-fatal myocardial infarction and unstable angina requiring revascularisation); (B) Kaplan–Meier curves for hard events (all-cause mortality and non-fatal myocardial infarction). CAD, coronary artery disease.

Table 2

Univariate and multivariate predictors of events

Discussion

The main finding of this study is that in an intermediate pretest likelihood population, MSCTA has a good prognostic value and may effectively identify patients at higher or lower risk for coronary events. These results further emphasise the usefulness of non-invasive imaging with MSCTA in this patient population.

MSCTA is increasingly used in the diagnosis of CAD, and may be specifically useful for ruling out CAD. The diagnostic accuracy of MSCTA has been studied extensively, and in early single-centre studies an average weighted sensitivity of 97.5 (95% CI 96 to 99) and specificity of 91 (95% CI 87.5 to 95) has been observed.5 More recently, several prospective multicentre studies have been published showing similar sensitivities and specificities.1 3 4 Importantly, however, in the majority of these diagnostic accuracy studies patients were included who were referred for conventional diagnostic coronary angiography and thus with high pretest probability. Accordingly, when interpreting these observations, it is important to take into account the pretest probability, as according to Bayes’ theorem, pretest probability may have a major influence on the positive and negative predictive value. With increasingly higher pretest probability, the prevalence of disease is higher and as a consequence the positive predictive value will increase, with a subsequent decrease in the negative predictive value. Conversely, in populations with a lower pretest probability, with lower disease prevalence, positive predictive value will decrease, but negative predictive value will improve.

The relationship between pretest probability and the diagnostic accuracy of MSCTA was recently studied by Meijboom et al among patients with a high, intermediate, or low pretest likelihood. In patients with a high pretest likelihood for CAD, the post-test probability for significant CAD after MSCT was not substantially different from the pretest probability. As a result, a normal examination did not result in sufficient reduction of post-test probability to reliably rule out the presence of significant CAD. These observations indicate that the clinical value of MSCTA is limited in this patient group. In contrast, in patients with a low and intermediate pretest likelihood, a negative MSCTA scan reduced the post-test probability of CAD to 0%.8 The effectiveness of MSCTA in patients with an intermediate pretest likelihood was also confirmed by Henneman et al.27 The authors assessed the prevalence of completely normal MSCTA, and thus the efficacy to rule out CAD, in patients with suspected CAD and related these observations to the pretest likelihood of CAD. The authors showed that normal MSCTA was seen in only 17% of patients with a high pretest likelihood, further underlining the limited clinical value in those patients. Conversely, MSCTA ruled out the presence of atherosclerosis in 33% of patients with an intermediate pretest likelihood.

However, in addition to diagnosis, prognostication is an important component of imaging tests and determines subsequent management. Thus far, several prognostic studies have examined the potential prognostic value of MSCTA.7 9 10 11 13 However, most of these observations have been derived from heterogeneous patient populations including patients with known CAD. In several more recent publications the prognostic value of MSCT has been determined specifically in patients with suspected CAD.14 15 16 In the study by Carrigan et al 227 patients without documented CAD were included.16 The absence of obstructive CAD was associated with a 99% freedom from cardiac death, myocardial infarction and revascularisation during an average of 2.3 years of follow-up. In patients with one or more vessels with obstructive CAD a significantly increased event rate was observed (log-rank p value = 0.01). Similar findings were recently reported by Hadamitzky et al.14

Although these observations underline the usefulness of MSCT for prognosis in patients with suspected CAD, dedicated data in the target population for MSCT, patients with an intermediate pretest likelihood, are lacking. Importantly, the results of this study confirm the prognostic value of MSCT in this particular patient population as MSCTA was shown to be highly effective for risk stratification. Indeed, hard event rates were <1% for patients without significant stenosis on MSCTA, indicating that these patients may be safely reassured. In contrast, the presence of a significant stenosis implied a substantially increased risk approaching 5% of coronary events.

Several limitations need to be acknowledged. Even though the diagnostic accuracy of MSCTA is high, uninterpretable images are still being encountered in a small percentage of patients owing to motion because of high or irregular heart rates or breathing during the examination. It is expected, however, that the number of uninterpretable studies will continue to decrease with newer-generation scanners. Currently 64-slice MSCTA is still associated with a high radiation exposure, although the radiation dose of MSCTA will decrease with the use of dedicated dose-reduction techniques that have recently become available.28 29 30 31 Importantly, low-dose computed tomography with prospective ECG triggering has recently been shown to reduce the radiation burden while maintaining diagnostic image quality and a high diagnostic accuracy.32 33

In conclusion, our results suggest that in an intermediate pretest likelihood population, MSCTA is highly effective in re-stratifying patients into either a low or high post-test risk group. These results further emphasise the usefulness of non-invasive imaging with MSCTA in this patient population.

REFERENCES

View Abstract

Footnotes

  • Funding JMvW is financially supported by a research grant from the Netherlands Society of Cardiology (Utrecht, The Netherlands). GP is financially supported by a training fellowship grant of the European Society of Cardiology (Sophia Antipolis, France) Huygens scholarship. HA is supported by the by the National Centre of Competence in Research, Computer Aided and Image Guided Medical Interventions of the Swiss National Science Foundation (Zurich, Switzerland) and has research grants from Siemens Medical Solutions (Forchheim, Germany). PAK is supported by a grant from the Swiss National Science Foundation (Berne, Switzerland) (SNSF-professorship grant No PPOOA-114706), and has research grants from GE Healthcare (Milwaukee, Wisconsin, USA). JJB has research grants from Medtronic (Tolochenaz, Switzerland), Boston Scientific (Maastricht, The Netherlands), BMS Medical Imaging (N Billerica, Massachusetts, USA), St Jude Medical (Veenendaal, The Netherlands), Biotronik (Berlin, Germany), GE Healthcare (St Giles, UK) and Edwards Lifesciences (Saint-Prex, Switzerland).

  • Competing interests None declared.

  • Ethics approval Approval from Leiden University Medical Centre and University Hospital Zurich.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.