Author + information
- Rishi Puri, MBBS,
- Adam J. Nelson, MBBS,
- Gary Y.H. Liew, MBBS,
- Stephen J. Nicholls, MBBS, PhD,
- Angelo Carbone, MMedSc,
- Dennis T.L. Wong, MBBS,
- James E. Harvey, MD, MSc,
- Kiyoko Uno, MD, PhD,
- Barbara Copus, RN,
- Darryl P. Leong, MBBS, MPH,
- John F. Beltrame, BMBS, PhD,
- Stephen G. Worthley, MBBS, PhD and
- Matthew I. Worthley, MBBS, PhD⁎ ()
- ↵⁎Cardiovascular Investigational Unit, Royal Adelaide Hospital, Level 6 Theatre Block, North Terrace, Adelaide, South Australia 5000, Australia
Multiple invasive coronary imaging modalities are currently available to quantify coronary luminal dimensions. Inherent differences in in vivo quantitative coronary lumen measurements within coronary segments between different imaging modalities remains poorly understood. The chief aim of this study was to evaluate differences between state-of-the-art invasive (intravascular ultrasound [IVUS], Fourier Domain optical coherence tomography [FD-OCT]) and noninvasive (3-dimensional–quantitative coronary angiography [3D-QCA]) coronary imaging techniques in the measurement of in vivo coronary lumen dimensions.
Ten patients (age ≥18 years) electively referred to our cardiac catheterization laboratory for the investigation of chest pain were studied. Informed consent was obtained prior to the procedure. Patients with nonobstructive coronary disease (≤70% stenoses) with no previous coronary revascularization, recent acute coronary syndrome, systolic heart failure, severe valvular heart disease, or renal failure were studied. This study was approved by our local Human Research Ethics Committee.
Coronary angiography was performed via a standard 6-F technique. IVUS was performed using a high-frequency rotational catheter (40 to 45 MHz) at an automated pullback speed of 0.5 mm/s. FD-OCT image acquisition was performed immediately following IVUS acquisition within the same vessel. The C7-XR OCT system (using the 2.7-F C7 Dragonfly Imaging Catheter, LightLab Inc., Westford, Massachusetts) was used, whereby angiographic contrast media was injected through the guiding catheter via an injection pump (settings: 14 ml volume contrast, 4 ml/s, 300 psi) to achieve effective intracoronary clearance of blood for optimal image acquisition. At an automated pullback speed of 20 mm/s, each pullback recorded 54 mm of coronary artery over a 3 to 4 s period. Prior to the commencement of imaging, the Z-offset was adjusted for appropriate image calibration for accurate image measurements offline.
All IVUS and FD-OCT data was analyzed using echoPlaque 3.0.60 (Indec Systems, Santa Clara, California). For each run, the most common distal and proximal fidiciary markers (anatomical side-branches) were chosen from corresponding IVUS and OCT pullbacks that defined the region of vessel to be analyzed. For IVUS imaging, cross-sectional images were selected every 30 frames (0.5 mm) apart, and coronary lumen and external elastic membrane were traced by manual planimetry to calculate the average lumen area. For OCT imaging, cross-sectional images were selected every second frame (0.4 mm apart) with the leading edges of the lumen traced by manual planimetry. Each IVUS and OCT run was precisely divided into 4 mm segments, comprised of 9 cross-sectional IVUS frames (with 8 × 0.5 mm intervals) and 11 cross-sectional OCT frames (with 10 × 0.4 mm intervals). Each segment, therefore, was analyzed separately. EchoPlaque (MIB) 3.0.60 software enabled the corresponding IVUS and OCT pullbacks (with corresponding numbered cross-sectional frames) to be simultaneously played allowing accurate frame-matching of anatomical fiduciary markers, thus ensuring the same segments were analyzed between each IVUS and OCT run per patient (1). Within each segment, the average lumen area (LA) and IVUS-derived plaque burden was calculated.
3D-QCA was performed offline using 3D reconstruction software (Paieon Medical, Rosh Haayin, Israel). The contrast-filled nontapered part of the guiding catheter was used to calibrate pixel size. The 2 best orthogonal views of the entire segment of vessel to be analyzed were used for 3D-QCA reconstructions. The same proximal and distal anatomical fiduciary markers used in the FD-OCT and IVUS analysis were identified from the 3D-QCA reconstructions. Following the identification of the centre of the lumen, the software generated a 3D representation of the arterial lumen. This was evaluated for each pre-defined 4-mm coronary segment studied. Similar to IVUS and OCT measurements, the average 3D-QCA derived LA was calculated per segment.
Continuous variables are expressed as mean ± SD. Paired t test was used to assess for significant differences between each imaging modality. Statistical analysis was performed using GraphPad Prism (version 5.01, GraphPad Software, La Jolla, California). A 2-tailed p value of <0.05 was considered significant.
A total of 80 coronary segments were evaluated in 10 coronary arteries. There was a good agreement between IVUS and OCT measurements of LA (r = 0.90, p < 0.0001). The mean LA of all segments with IVUS was 9.35 ± 4.1 mm2, compared with a corresponding mean LA on OCT as 7.79 ± 3.3 mm2 (p < 0.0001), with a mean relative difference of 13.1%. Plaque burden was found to have no significant influence upon measured differences of lumen diameter and LA between IVUS and FD-OCT. Coronary segments were further stratified according to size defined by tertiles of LA on IVUS [‘small:' mean LA 3.8 ± 0.9 mm2; ‘medium:' mean LA 7.5 ± 1.0 mm2; ‘large:' mean LA 12.9 ± 3.0 mm2). As such, the mean relative differences between IVUS and OCT for LA measurements in small, medium, and large segments were 24%, 9.1%, and 12%, respectively (p = 0.002 for small vs. moderate, p = 0.032 for small vs. large). There was satisfactory agreement between LA derived by 3D-QCA compared with IVUS and OCT (r = 0.66, p < 0.0001). The mean LA of all segments obtained with 3D-QCA was 6.30 ± 3.27 mm2 (p < 0.0001 compared with both IVUS and OCT). The mean relative difference of LA measured with 3D-QCA and IVUS was 41.5%, compared with 28.9% with 3D-QCA and OCT (p = 0.012).
This study, utilizing state-of-the-art invasive coronary imaging technologies, highlights systematic differences in measured coronary lumen areas of corresponding segments. Consistent with previous recently reported ex vivo and in vivo findings (1–3), we highlight consistently larger LA measurements obtained with IVUS compared to FD-OCT across all segments, with a further significant reduction in these luminal measurements when measured in smaller-sized coronary segments, and when measured with 3D-QCA. This also supports recent data which has compared Terumo FD-OCT, IVUS, and conventional QCA measurements in stented coronary segments (3). Until further validation studies are conducted, specific lumen-based cut-off values previously validated with IVUS and currently utilised in contemporary interventional practice should not be arbitrarily translated into the FD-OCT and 3D QCA hemispheres. (Fig. 1)
Please note: Drs. Puri, Leong and Liew are supported by a Postgraduate Medical Research Scholarship from the National Health & Medical Research Council (565579, 519177, and 497809 respectively). Drs. Puri and Leong are jointly funded by the National Heart Foundation of Australia (PC0804045, PC07A3395). Dr. Worthley has a South Australian Early to Mid Career Practitioner Fellowship.
- American College of Cardiology Foundation