Sensitivity and Specificity of QCA in Detecting Coronary Arterial Remodeling After Intracoronary Brachytherapy: A Comparison to

For more than a decade, quantitative coronary angiography (QCA) has been the gold standard for the assessment of coronary stenosis because of its accuracy and objectivity as compared to visual and hand-held caliper measurements.1–3 After the introduction of intracoronary brachytherapy, the QCA methodology for the assessment of irradiated coronaries had to be adjusted to this new mode of therapy because of the existence of new regions of interest: the target segment, injured segment, irradiated segment and vessel segment.4 In a recent report, lumen enlargement (negative late loss) was demonstrated in a subset of vessels receiving 18 Gy of catheter-based ß-radiation after balloon angioplasty alone.5 Previously, we reported vessel enlargement accommodating plaque increase in a volumetric 3-dimensional (3-D) intravascular ultrasound (IVUS) investigation.6 In that report, the lumen remained unchanged at follow-up as an average. In other words, half of the irradiated segments responded to the radiation with an enlargement in the lumen. Thus, intracoronary radiation has the potential to increase lumen diameter.5–7 The aim of this study is to analyze the sensitivity and specificity of QCA parameters to detect a positive vessel remodeling after intracoronary ß-radiation as compared to IVUS. Methods Patients. The study population consisted of 27 consecutive patients who underwent balloon angioplasty (BA) followed by catheter-based intracoronary ß-radiation with a 90Strontium/90Yttrium (90Sr/90Y) source in a single vessel with IVUS imaging and electrocardiogram (ECG)-gated pullback. Patients presented with angina pectoris and/or positive stress test. Patients with myocardial infarction within 72 hours prior to treatment or left ventricular ejection fraction 2.5 mm and Radiation system. The source train of the Beta-Cath™ System consists of a series of 12 independent cylindrical seeds, which contain pure ß-emitting 90Sr/90Y, and is bordered by 2 gold markers (total length of radioactive seeds, 30 mm). The profile of the catheter is 5 French and the source train is not centered. The radiation sources remain at the treatment site for approximately 2–4 minutes to deliver a predetermined dose at 2 mm from the centerline of the axis of the source train. Prescribed radiation doses were 12 Gy (8 vessels), 14 Gy (5 vessels), 16 Gy (9 vessels) and 18 Gy (5 vessels). Procedure. All patients received aspirin (250 mg/day) and intravenous heparin (10,000 IU) during the procedure. Additional heparin was given to maintain activated clotting time > 300 seconds. BA was performed according to standard clinical practice. After successful BA, intracoronary ß-radiation was performed as previously described,8 and repeat angiography and IVUS pullback were carried out. Intracoronary isosorbide dinitrates (200 µg) were administered immediately prior to each IVUS pullback. At follow-up exam (6–8 months), further IVUS analysis of the treated vessel was performed. QCA analysis. QCA analysis was performed offline by an independent analyst. All angiograms were evaluated after intracoronary administration of nitrates. The analysis was performed with the CAAS II analysis system (Pie Medical BV, Maastricht, The Netherlands). Calibration of the system was based on catheter dimensions while unfilled with contrast medium. This method of analysis has been previously validated.2,3 Within a region of interest, the MLD is determined by edge detection and averaged from the 2 orthogonal projections. Reference diameter is automatically calculated by the interpolated method. The analyst is also able to perform a subsegmental analysis, using the software of the CAAS system. The region of interest is automatically divided into subsegments of equidistant length (5.0 ± 0.3 mm). The system also computed mean lumen diameter, maximal lumen diameter and MLD in every subsegment. Late loss was defined as MLD post-treatment minus MLD at follow-up. IVUS image acquisition and quantitative analysis. The coronary segment subject to 3-D reconstruction was examined with a mechanical IVUS system (CVIS, Boston Scientific/Scimed, Inc., Maple Grove, Minnesota) incorporating a 30 MHz, single-element transducer rotating at 1,800 rpm. ECG-gated image acquisition and digitization were performed by a workstation designed for the 3-D reconstruction of echocardiographic images (EchoScan, Tomtec, Munich, Germany). A description of this system has been reported in detail elsewhere.9–11 In brief, the steering logic of the workstation considered the heart rate variability and only acquired images from cycles meeting a predetermined range and coinciding with the peak of the R-wave. A Microsoft Windows™-based contour detection program, developed at the Thoraxcenter, was used for offline volumetric quantification.12 Briefly, this program constructed longitudinal sections from the data set and identified the contours corresponding to the lumen and media boundaries. Volumetric data were calculated by the formula: V = (ni = 1 Ai * H, where V = volume, A = area of external elastic membrane (EEM), lumen or plaque in a given cross-sectional ultrasound image, H = thickness of the coronary artery slice (which was reported by this digitized cross-section) and n = the number of digitized cross-sectional images encompassing the volume to be measured. Checking and editing of the contours of the planar images were performed by 2 independent experienced analysts. Intra-observer variability assessed by analyzing IVUS volumetric studies at least 3 months apart has been reported: -0.4 ± 1.1% in lumen volume, -0.4 ± 0.6% in total vessel (EEM) volume and -0.3 ± 1.0% in vessel wall (plaque + media) volumes using motorized ECG-gated pullback.11 The application of this system has been reported in clinical studies.6,13–15 Definitions. Total vessel volume (TVV), lumen volume (LV) and plaque volume (PV) were calculated from the contours of each cross-section by the software as stated above. In order to assess the volumetric changes of the vessel structures after 6–8 months, the delta (?) value for each measurement was calculated (? = follow-up - post-procedure). To eliminate the influence of the vessel size, percent change (? volume / post-procedure volume) was also calculated. Remodeling of the vessel wall was defined when total vessel (EEM) volume increased or decreased, when compared to post-procedure measurements by at least 2 standard deviations (± 1.2%) of the intra-observer variability. By using this technique, the potential intrinsic error of the method may be avoided.16–18 Selection of the region of interest. The region of interest was the irradiated segment (IRS). It was selected for QCA after reviewing all cinefilms performed during the index procedure. Any angiographic sequence showing the lesion at pre-intervention, post-intervention and follow-up, as well as the position of radiation source, may be displayed simultaneously on the screen using the Rubo DICOM Viewer (Rubo Medical Imaging, Uithoorn, The Netherlands). The ECG tracing is also displayed in any angiographic sequence. By selecting frames in the same part of the cardiac cycle, we were able to define the location of the radiation source relative to the original lesion (30 mm in length). Sidebranches were used as index anatomical landmarks. Distances from these proximal or distal sidebranches to the inner part of the proximal and distal gold markers were computed by the CAAS software. The segment encompassed by the inner part of the 2 radio-opaque markers defined the IRS (Figure 1). All regions of interest were superimposed on the post-procedural and follow-up angiograms. The methodology to angiographically define the segment of interest using this technique has been previously described.19 By applying the same methodology using the Rubo DICOM Viewer and CAAS system, we were able to define the location of the radiation source train with anatomical landmarks on IVUS (Figure 1). IRS was selected based on the anatomical landmarks and the distances from them calculated by the 3-D reconstruction system post-procedure. At follow-up, correct matching of the region of interest was assured by both the use of the same IVUS motorized pull-back system and the comparison of the longitudinal view to that of post-procedure. This methodology for IVUS has previously been described in detail.6,20,21 Statistical analysis. Quantitative data are presented as means ± standard deviations. The comparisons between the volumetric data were performed using a 2-tailed Student’s t-test. Categorical data were compared with Fisher’s exact test. Linear regression analysis was used to investigate the relationship between QCA and IVUS parameters. Sensitivity and specificity were calculated to show the true positive and true negative probabilities of positive remodeling (+ 2.4% increase in TVV). Receiver operator characteristic (ROC) curves were constructed to investigate the diagnostic power of the variable. A p-value of 90° of circumferential arc in > 30% of the cross-sections (n = 10). There were significant increases only in IVUS derived parameters: TVV, PV and any wall thickness parameters (mean, maximum and minimum). Only poor correlation was observed between QCA parameters and change in TVV (Figure 2). According to the ROC curve analysis, change in MLD derived from QCA was not a good indicator of positive remodeling, with a sensitivity of 55% and a specificity of 54% (Table 3). In addition, changes in mean and maximal diameter were not significant parameters to detect positive vessel remodeling. Since the radiation source has an acute dose fall-off, both extreme subsegments received lower doses than the central part of the irradiated segments. When only central subsegments were analyzed, ROC curve area, sensitivity and specificity were better than total subsegments (Table 3). LV quantified by 3-D IVUS correlates with change in TVV (r = 0.562; p Segments where MLD was initially located (n = 27). It has been reported that relocation of MLD is more frequent in brachytherapy than in conventional BA.4 In the current study, relocation of MLD from pre-procedure to post-procedure has occurred in 74% of vessels. Between post-procedure and follow-up, the rate of relocation was 82%. Since MLD is used as a target of the treatment, only the segments where MLD was initially located were examined. Changes in mean diameter, maximal diameter and MLD of those segments were not indicators of positive vessel remodeling as well as the total subsegmental analysis (Table 3). Discussion The aim of this study was to investigate the usefulness of QCA in understanding the mechanism of restenosis prevention by intracoronary brachytherapy. This study demonstrates that parameters derived from QCA are not sufficient to establish the presence of positive vessel remodeling after BA in the setting of catheter-based ß-radiation (90Sr/90Y source). However, mean diameter and MLD were significant indicators for a positive vessel remodeling when only the fully irradiated segments were considered. Mechanisms of restenosis prevention. In animal experimental models, it has been emphasized that intracoronary radiation inhibits neointimal proliferation.22–24 However, experimental data have also suggested that radiation has an effect on vessel remodeling by modifying cell responses in the adventitia.25,26 Whether intracoronary radiation mainly affects positive remodeling or inhibition of tissue proliferation remains to be investigated. It is also a point of debate in human IVUS investigations. Positive vessel remodeling accommodating neointimal ingrowths after 6 months has been demonstrated using 3-D IVUS quantification,6 whereas total vessel area and plaque area remained unchanged in another study.27 In this study, total vessel volume and wall thickness derived from IVUS were partially correlated to the lumen change (Figure 3). Therefore, the contribution of vessel remodeling and tissue proliferation to lumen preservation may have different patterns depending on individual and local elements (local vessel dimensions, delivered dose, plaque morphology and degree of injury). In addition, the central portions of irradiated segments showed higher sensitivity and specificity, with significant ROC curve area in changes in MLD and mean diameter. This finding may demonstrate that higher-dose radiation has the effect of reducing the variability of the response, which leads to vessel enlargement. Dosimetric analysis would be required to address this issue. Value of QCA. QCA has been a standard research tool for more than a decade, providing accurate and reproducible measurements. Most of the clinical trials on restenosis after coronary intervention used angiographic measurements (minimal lumen diameter and percent diameter stenosis) for their endpoints. However, we recently reported that QCA methodology for the assessment of irradiated coronary arteries had to be adjusted to this new mode of therapy, because of the existence of various regions of interest (target segment, injured segment, radiated segment and vessel segment).4 In addition, lumen enlargement with negative late lumen loss has rarely been reported before the introduction of coronary brachytherapy. The dose-finding study using a 90Y source has shown lumen enlargement after catheter-based ß-radiation following BA for the first time.5 Possible mechanisms of lumen enlargement. There are some important aspects to consider when attempting to understand the mechanism of lumen enlargement detected by QCA. First, radiation can potentially induce positive vessel remodeling (i.e., TVV increase) and medial thinning or thickening (i.e., changes in PV). These various changes in vessel wall and morphology may be one of the reasons for the complexity of vessel response after BA followed by intracoronary radiation (Figure 4). Second, we must consider the delivered dose of the ß-emitting source (90Sr/90Y source). Indeed, dose inhomogeneity within the irradiated segments was observed in a previous study.28 In that report, plaque volume at follow-up (comparable to wall thickness at follow-up) was associated with actual dose, which widely ranged over the entire irradiated segments. In the present study, central fully irradiated segments showed better sensitivity and specificity. These findings suggest that actual delivered dose may be a major determinant of vessel remodeling. It is noteworthy that the actual dose delivered to the adventitia cannot be assessed by QCA. Third, frequent relocation of MLD in the present study may also support the variability of the response. This result suggests that the comparison between post-procedure and follow-up is assessed in different positions in most of the cases. Finally, another factor influencing the poor prediction of positive vessel remodeling may be the inaccuracy of the edge-detection method of QCA. Especially immediately after the procedure, a poor correlation between QCA and IVUS results has been reported because of complex lumen morphology after BA.29 Thus, lumen increase detected by QCA may not be fully explained by positive vessel remodeling. Study limitations. Small inaccuracies cannot be completely ruled out because of axial movement of the radiation source during the cardiac cycle. However, the 3-D, reconstructed, volumetric IVUS analysis with ECG-gated pullback used in this study is the most precise method currently available in terms of selection of region of interest by eliminating the artifacts from the cardiac movement. Using this technique, we have demonstrated the behavior of the irradiated vessels, comparing vessel geometry at follow-up with post-procedure.6,30 To investigate the mechanism of action of radiotherapy, this comprehensive technique representing the entire segment of interest may be more relevant than assessing single cross-sectional images or angiographic results, since relocation of MLD frequently occurs after BA followed by intracoronary ß-radiation. However, it is nevertheless important to note that the common angiographic endpoints (i.e., restenosis rate) would be enough to assess the effectiveness of intracoronary brachytherapy in clinical protocols, considering that discrete lesions at follow-up can be well detected by QCA. Conclusion. Lumen enlargement detected by QCA does not always mean a positive vessel remodeling after intracoronary radiation. IVUS analysis may be necessary to investigate the mechanism of restenosis after BA followed by catheter-based radiation.

1. Kuntz RE, Baim DS. Defining coronary restenosis. Newer clinical and angiographic paradigms. Circulation 1993;88:1310–1323. 2. Foley DP, Escaned J, Strauss BH, et al. Quantitative coronary angiography (QCA) in interventional cardiology: Clinical application of QCA measurements. Prog Cardiovasc Dis 1994;36:363–384. 3. Rensing BJ, Hermans WR, Deckers JW, et al. Lumen narrowing after percutaneous transluminal coronary balloon angioplasty follows a near gaussian distribution: A quantitative angiographic study in 1,445 successfully dilated lesions. J Am Coll Cardiol 1992;19:939–945. 4. Sabate M, Costa MA, Kozuma K, et al. Methodological and clinical implications of the relocation of the minimal luminal diameter after intracoronary radiation therapy. Dose Finding Study Group. J Am Coll Cardiol 2000;36:1536–1541. 5. Verin, Popowski Y, de Bruyne B, et al. Endoluminal beta-radiation therapy for the prevention of coronary restenosis after balloon angioplasty. N Engl J Med 2001;344:243–249. 6. Sabate M, Serruys PW, van der Giessen WJ, et al. Geometric vascular remodeling after balloon angioplasty and beta- radiation therapy: A three-dimensional intravascular ultrasound study. Circulation 1999;100:1182–1188. 7. Condado JA, Waksman R, Gurdiel O, et al. Long-term angiographic and clinical outcome after percutaneous transluminal coronary angioplasty and intracoronary radiation therapy in humans. Circulation 1997;96:727–732. 8. King SB, 3rd, Williams DO, Chougule P, et al. Endovascular beta-radiation to reduce restenosis after coronary balloon angioplasty: Results of the beta energy restenosis trial (BERT). Circulation 1998;97:2025–2030. 9. Bruining N, von Birgelen C, de Feyter PJ, et al. Dynamic imaging of coronary stent structures: an ECG-gated three-dimensional intracoronary ultrasound study in humans. Ultrasound Med Biol 1998;24:631–637. 10. Bruining N, von Birgelen C, de Feyter PJ, et al. ECG-gated versus nongated three-dimensional intracoronary ultrasound analysis: Implications for volumetric measurements. Cathet Cardiovasc Diagn 1998;43:254–260. 11. von Birgelen C, de Vrey EA, Mintz GS, et al. ECG-gated three-dimensional intravascular ultrasound: Feasibility and reproducibility of the automated analysis of coronary lumen and atherosclerotic plaque dimensions in humans. Circulation 1997;96:2944–2952. 12. Li W, von Birgelen C, Di Mario C, et al. Semi-automated contour detection for volumetric quantification of intracoronary ultrasound. Comput Cardiol 1994:277–280. 13. Bruining N, Sabate M, de Feyter PJ, et al. Quantitative measurements of in-stent restenosis: A comparison between quantitative coronary ultrasound and quantitative coronary angiography. Cathet Cardiovasc Intervent 1999;48:133–142. 14. Sabate M, Marijnissen JP, Carlier SG, et al. Residual plaque burden, delivered dose, and tissue composition predict 6-month outcome after balloon angioplasty and beta-radiation therapy. Circulation 2000;101:2472–2477. 15. Costa MA, Sabate M, Serrano P, et al. The effect of 32P beta-radiotherapy on both vessel remodeling and neointimal hyperplasia after coronary balloon angioplasty and stenting: A three-dimensional intravascular ultrasound investigation. J Invas Cardiol 2000;12:113–120. 16. Kearney PP, Ramo MP, Shaw TR, et al. Analysis of reproducibility of reference lumen quantitation with intravascular ultrasound in stented coronary arteries. Cathet Cardiovasc Diagn 1997;40:1–7. 17. Sabate M, Kay IP, de Feyter PJ, et al. Remodeling of atherosclerotic coronary arteries varies in relation to location and composition of plaque. Am J Cardiol 1999;84:135–140. 18. Kozuma K, Costa MA, Sabate M, et al. Three-dimensional intravascular ultrasound assessment of noninjured edges of beta-irradiated coronary segments. Circulation 2000;102:1484–1489. 19. Sabate M, Costa MA, Kozuma K, et al. Geographic miss: A cause of treatment failure in radio-oncology applied to intracoronary radiation therapy. Circulation 2000;101:2467–2471. 20. Kozuma K, Costa MA, Sabate M, et al. Relationship between tensile stress and plaque growth after balloon angioplasty treated with and without intracoronary beta-brachytherapy. Eur Heart J 2000;21:2063–2070. 21. Costa MA, Kozuma K, Gaster AL, et al. Three dimensional intravascular ultrasonic assessment of the local mechanism of restenosis after balloon angioplasty. Heart 2001;85:73–79. 22. Waksman R, Robinson KA, Crocker IR, et al. Endovascular low-dose irradiation inhibits neointima formation after coronary artery balloon injury in swine. A possible role for radiation therapy in restenosis prevention. Circulation 1995;91:1533–1539. 23. Weinberger J, Amols H, Ennis RD, et al. Intracoronary irradiation: Dose response for the prevention of restenosis in swine. Int J Radiat Oncol Biol Phys 1996;36:767–775. 24. Wiedermann JG, Marboe C, Amols H, et al. Intracoronary irradiation markedly reduces restenosis after balloon angioplasty in a porcine model. J Am Coll Cardiol 1994;23:1491–1498. 25. Waksman R, Rodriguez JC, Robinson KA, et al. Effect of intravascular irradiation on cell proliferation, apoptosis, and vascular remodeling after balloon overstretch injury of porcine coronary arteries. Circulation 1997;96:1944–1952. 26. Wilcox JN, Waksman R, King SB, Scott NA. The role of the adventitia in the arterial response to angioplasty: the effect of intravascular radiation. Int J Radiat Oncol Biol Phys 1996;36:789–796. 27. Meerkin D, Tardif JC, Crocker IR, et al. Effects of intracoronary beta-radiation therapy after coronary angioplasty: An intravascular ultrasound study. Circulation 1999;99:1660–1665. 28. Sabate M, Marijnissen JP, Carlier SG, et al. Residual plaque burden, delivered dose, and tissue composition predict 6-month outcome after balloon angioplasty and beta-radiation therapy. Circulation 2000;101:2472–2477. 29. Haase J, Ozaki Y, Di Mario C, et al. Can intracoronary ultrasound correctly assess the luminal dimensions of coronary artery lesions? A comparison with quantitative angiography. Eur Heart J 1995;16:112–119. 30. Kay IP, Sabate M, Costa MA, et al. Positive geometric vascular remodeling is seen after catheter-based radiation followed by conventional stent implantation but not after radioactive stent implantation. Circulation 2000;102:1434–1439.