Author + information
- Received February 12, 2017
- Revision received March 8, 2017
- Accepted March 29, 2017
- Published online June 5, 2017.
- Nazzareno Galiè, MDa,∗ (, )
- Francesco Saia, MDb,
- Massimiliano Palazzini, MDa,
- Alessandra Manes, MDb,
- Vincenzo Russo, MDb,
- Maria Letizia Bacchi Reggiani, PhDa,
- Gianni Dall’Ara, MDa,
- Enrico Monti, MDa,
- Fabio Dardi, MDa,
- Alessandra Albini, MDa,
- Andrea Rinaldi, MDa,
- Enrico Gotti, MDa,
- Nevio Taglieri, MDa,
- Cinzia Marrozzini, MDb,
- Luigi Lovato, MDb,
- Maurizio Zompatori, MDa and
- Antonio Marzocchi, MDb
- aUniversity of Bologna, Department of Experimental, Diagnostic and Specialty Medicine-DIMES, Bologna, Italy
- bS. Orsola-Malpighi University Hospital, Cardiovascular and Thoracic Department, Bologna, Italy
- ↵∗Address for correspondence:
Dr. Nazzareno Galiè, Department of Experimental, Diagnostic and Specialty Medicine-DIMES, University of Bologna, Via Massarenti 9, 40138 Bologna, Italy.
Background Left main coronary artery (LMCA) compression is increasingly recognized as a cause of angina in pulmonary arterial hypertension (PAH).
Objectives This study aimed to evaluate the prevalence of LMCA extrinsic compression from a dilated pulmonary artery (PA) in patients with PAH and angina or angina-like symptoms, determine the usefulness of screening with computed tomography coronary angiography (CTCA), and assess the safety and efficacy of percutaneous coronary interventions (PCIs).
Methods All patients with PAH and angina or angina-like symptoms attending the center between May 1, 2008, and December 31, 2013, underwent CTCA. Patients with confirmed LMCA stenosis on selective coronary angiography had PCI.
Results Of 765 patients with PAH, 121 had angina or angina-like symptoms. Ninety-four patients had abnormal CTCA based on the relationship between the PA and the LMCA and underwent selective coronary angiography. LMCA stenosis ≥50% was detected in 48 of the 94 patients. Forty-five patients underwent PCI with stenting, of whom 41 had sustained angina symptom relief. The 3 other patients had surgical PA reduction plasty. Nine months after PCI, 5 patients had LMCA restenosis and PCI was successfully repeated. The best predictor of LMCA stenosis ≥50% was a PA diameter ≥40 mm. Rates for death or double-lung transplant and the composite rates for death, double-lung transplant, or restenosis at 36 months were 5% and 30%, respectively.
Conclusions The prevalence of LMCA compression in patients with PAH and angina is high. These results suggest that CTCA is indicated in patients with PAH and angina or angina-like symptoms. PCI was well tolerated, improved symptoms, and resulted in favorable long-term outcomes.
Despite recent advances in treatment, prognosis for those with pulmonary arterial hypertension (PAH) is poor, particularly in severe disease (1). Progressive remodeling of the small distal pulmonary arteries leads to increased pulmonary vascular resistance, culminating in right heart failure and death. Early diagnosis can be difficult because the typical symptoms, such as dyspnea, fatigue, and chest pain on exertion, are nonspecific. In patients with PAH, chest pain is common, reported in 7% to 29% of patients (2–5). Some chest pain symptoms resemble those of classic angina pectoris and are described as precordial “discomfort” that is induced by exercise and rapidly regresses with rest.
In patients with PAH, classic angina and angina-like symptoms are usually attributed to the increased and unmatched metabolic demands of the hypertrophied and overloaded right ventricle (6) rather than to coronary artery stenosis; therefore, angiography is not routinely indicated (7,8). However, extrinsic compression of the left main coronary artery (LMCA) by a dilated pulmonary artery (PA) main trunk is increasingly recognized as a cause of angina in PAH. Despite multiple case reports (9,10), the incidence of LMCA compression in PAH is not well established. A small case series reported that, of 26 patients with PAH and angina, 7 (26.9%) had LMCA compression (11).
Along with angina, LMCA compression may be associated with additional complications of severe myocardial ischemia, including myocardial infarction (MI) (12), left ventricle dysfunction (13), arrhythmia, and eventually, sudden death (14–16). Because >25% of deaths in patients with PAH are related to sudden death (17), some of these events might be attributable to LMCA compression, a potentially correctable complication. Various reports have suggested that percutaneous coronary intervention (PCI) may be a therapeutic option for LMCA compression in patients with PAH (9,10,18).
The aims of the current study were to evaluate the prevalence of LMCA compression in patients with PAH and angina or angina-like symptoms, determine the clinical usefulness of computed tomography coronary angiography (CTCA) for identifying those who would benefit from selective coronary angiography (SCA), and assess the safety and efficacy of PCI.
This was a prospective analysis of patients with PAH treated at the Pulmonary Hypertension Center of the S. Orsola-Malpighi University Hospital, Bologna, Italy, between May 1, 2008, and December 31, 2013. All patients with PAH were eligible for inclusion. The diagnosis of PAH was performed in accordance with pulmonary hypertension guidelines (7,8). Patients with PAH and angina or angina-like symptoms (defined as precordial discomfort on physical exertion with fast recovery on rest) were identified by 3 senior cardiologists (N.G., M.P., and A.M.) after individual patient evaluation and collective discussion in cases of discrepancy between evaluations. Patients without angina or angina-like symptoms were excluded. The study procedures complied with the Declaration of Helsinki and were approved by the local ethics committee. Written informed consent was obtained from all patients before the start of the study procedures.
All patients with angina or angina-like symptoms underwent CTCA (Figure 1). The CTCA procedure was performed using the same method as routine CTCA examinations; however, premedication with beta-blockers, as previously reported (19), was avoided.
Based on the CTCA results, patients were categorized into 4 groups according to the relationship between the PA main trunk and the LMCA: 1) normal distance: minimal distance between the PA main trunk and the LMCA >1 mm (Figure 2A); 2) contiguity: distance between the dilated PA main trunk and the LMCA ≤1 mm without dislocation or stenosis (Figure 2B); 3) dislocation: dislocation of the LMCA owing to a dilated PA main trunk with a take-off angle of the LMCA <60° (16) (see Figure 3 for measurement details), associated or not associated with significant lumen stenosis (Figure 2C); and 4) significant stenosis: stenosis of the LMCA ≥50% due to extrinsic compression from a dilated PA main trunk (Figure 2D). We decided to consider a dislocation of the LMCA for angles of <60° according to the data collected in normal individuals by Kajita et al. (16) In this last paper, 75% of normal individuals had an LMCA take-off angle ≥60°, without overlaps with patients with LMCA extrinsic compression.
The main PA trunk diameter was measured at bifurcation level, orthogonal to its long axis. Two cardiovascular radiologists (V.R. and L.L.) performed categorization independently, with evaluation by a third (M.Z.) in cases of discrepancy. The interobserver difference for the measurement of the take-off angle (Figure 3) was within 6° and the correlation was r = 0.94.
Patients who were categorized as having significant stenosis, dislocation, or contiguity underwent SCA to confirm LMCA stenosis and provide a definitive diagnosis of extrinsic LMCA compression. The procedure was performed through the transradial or transfemoral approach by standard technique. A meticulous search of the best view for visualizing the coronary ostium and LMCA stenosis was performed (usually a left anterior oblique view with cranial angulation) with the aim to distinguish the origin of the LMCA from the aortic profile without overlapping with the contiguous portion of the left aortic sinus. At least 1 nonselective contrast medium injection was performed when the catheter tip was in the left aortic sinus to avoid potential interference induced by the guiding catheter inside the LMCA. Stenosis severity was assessed visually by 2 interventional cardiologists (F.S. and A.M.) and confirmed with quantitative coronary analysis (QCA) using dedicated software (CAAS II, Pie Medical, Maastricht, the Netherlands) (20). LMCA-significant stenosis at SCA was defined as a ≥50% reduction of the LMCA luminal diameter (21,22). Right-heart catheterization and a 6-min walk test were also performed within 1 week of CTCA.
Patients with confirmed LMCA-significant stenosis underwent concurrent LMCA PCI with bare-metal stents (BMS) or drug-eluting stents (DES) (Figure 4). PA surgical reduction plasty was performed (23) in cases of planned surgical correction of congenital heart defects or urgent lung transplantation. PCI was performed using a standard technique after administration of an intravenous bolus of unfractionated heparin (60 IU/kg). The choice of stent type (BMS or DES) was at the discretion of the operator, with consideration of the possible background oral anticoagulant therapy for PAH. BMS were preferred in the latter case to reduce the duration of triple antithrombotic therapy, thus limiting bleeding risk. QCA was repeated after stenting. Reference vessel diameter was 4.1 ± 0.6 mm, whereas stent size was 4.0 ± 0.5 mm. Post-dilatation was performed in 13.3% of patients. Maximal inflation pressure was 19.3 ± 2.7 atm, with an average final balloon diameter of 4.6 ± 0.6 mm. Acute stent recoil was not seen in any patient.
Endpoints and follow-up
Patient follow-up was conducted at 1 month and every 3 months thereafter. Clinical endpoints included the occurrence of all-cause and cardiovascular death, MI, stroke, target lesion restenosis, and stent thrombosis, defined according to the Academic Research Consortium (24). Bleeding was classified according to the Bleeding Academic Research Consortium (BARC) (25). SCA follow-up was planned electively at 9 months for all patients undergoing PCI. LMCA restenosis was defined as a diameter stenosis ≥50% at a 9-month elective follow-up period or during an anticipated clinically indicated SCA. QCA was repeated at 9-month follow-up. Angiographic measurements were performed in the same projection at pre- and post-procedure and at follow-up.
The Student t test was used for comparisons of continuous variables and the chi-square test was used for discrete variables. Logistic regression analyses of CTCA scan parameters were performed to assess predictors of LMCA stenosis ≥50%. Parameters that were significant on univariate analysis with a p value <0.10 were included in the multivariate analysis. Significant predictors on multivariate analysis (p < 0.05) were analyzed with receiver-operating characteristic (ROC) curves to identify cutoff values with the best sensitivity and specificity for predicting LMCA stenosis ≥50%. Data are reported as mean ± SD unless otherwise stated. Event-rate curves were calculated and presented as Kaplan-Meier analyses.
The disposition of the 765 patients with PAH who attended the Pulmonary Hypertension Center during the study period is shown in Figure 1. LMCA stenosis ≥50% was confirmed by SCA in 48 patients, representing 40% of the 121 patients with PAH and angina or angina-like symptoms and 6% of the 765 consecutive patients with PAH who were included in the study (Figure 1).
Within the CTCA groups, LMCA stenosis ≥50% was confirmed by SCA in 32 of 35 patients (91.4%) with significant stenosis, in 15 of 49 patients (30.6%) with dislocation, and in 1 of 10 patients (10.0%) with contiguity (Figure 1). Minimal or no atherosclerotic lesions were observed in the LMCA or in other regions of the coronary artery. No patients were lost to follow-up.
The demographic, clinical, exercise, functional, hemodynamic, and PA characteristics of all patients presenting with angina or angina-like symptoms (n = 121) are shown in Table 1. The demographic and clinical characteristics of patients who underwent SCA, stratified according to the presence of LMCA stenosis <50% or ≥50%, are reported in Table 2. No significant differences were observed between the 2 groups except for larger PA diameters in patients with an LMCA stenosis ≥50%.
Predictors of LMCA stenosis
PA diameter, PA diameter index (PA diameter/body surface area), and the ratio of PA and aortic diameters were significantly different among patients with LMCA stenosis ≥50% and patients with stenosis <50%, as confirmed with SCA (Table 2). Additionally, the LMCA take-off angle (16) was 34.1 ± 12.7° in patients with significant LMCA stenosis and 49.6 ± 12.2° in patients without significant stenosis (p < 0.01). On multivariate analysis, the PA diameter assessments correlated with the LMCA degree of stenosis: the odds ratios were 1.17 (95% confidence interval [CI]: 1.10 to 1.25) for PA diameter, 20.5 (95% CI: 5.64 to 74.55) for ratio of PA/aortic diameters, and 1.19 (95% CI: 1.11 to 1.28) for PA diameter/body surface area. Using ROC analysis, the strongest predictor of LMCA stenosis ≥50% at SCA was PA diameter ≥40 mm with a sensitivity of 83% and a specificity of 70% and an ROC area of 0.8512. A PA/aortic ratio ≥1.5 had a sensitivity of 73% and specificity of 70% (ROC area: 0.8007) for predicting stenosis. PA diameter index ≥24 mm/m2 had a sensitivity of 79% and specificity of 68% (ROC area: 0.8472) for predicting stenosis.
PCI and follow-up
LMCA PCI with stenting was performed in 45 patients with LMCA stenosis ≥50% during SCA; 23 patients (51.1%) were treated with BMS and 22 patients (48.9%) were treated with DES. Surgical PA reduction plasty was performed in the other 3 patients (all in the significant stenosis group based on CTCA) (Figure 1) because of concomitant surgical closure of atrial septal defects in 2 patients or urgent lung transplantation in the other patient. Post-PCI antithrombotic therapy included 1 month of double antiplatelet therapy (aspirin plus clopidogrel) in 43 patients, prolonged up to 12 months in those with DES (n = 22). Warfarin treatment was continued in 10 patients (triple antithrombotic therapy) and was temporarily withdrawn in 15 patients during the period of double antiplatelet therapy. Two patients were treated with a single antiplatelet agent plus warfarin.
PCI was successful in all patients. No major complications were observed in the early post-interventional period (in hospital), with no deaths, MI, transient ischemic attack or stroke, repeat PCI, or acute stent thrombosis reported. One DES-treated patient had a vascular complication at the site of access (femoral artery pseudoaneurysm) that necessitated surgical correction and 1 patient treated with a BMS had an acute kidney injury because of contrast-induced nephropathy. PCI significantly increased minimal luminal diameter compared with baseline (1.6 ± 0.6 mm at baseline vs. 3.6 ± 0.5 mm after PCI; p < 0.01) and decreased stenosis (62.8 ± 11.5% at baseline vs. 14.5 ± 8.4% after PCI; p < 0.01) at QCA analysis.
After the PCI procedure, angina or angina-like symptoms had improved or disappeared in 43 of 45 patients (95.5%), with 41 of 45 patients (91.1%) experiencing complete relief from symptoms. Two patients reported no symptom relief: 1 had restenosis due to mechanical recoil and the other had no in-stent restenosis and no new lesions.
At 9 months after PCI, minimal luminal diameter was 3.2 ± 0.7 mm (p < 0.01 vs. after PCI) and stenosis was 20.2 ± 13.9% (p = 0.03 vs. after PCI). At 9 months, 5 patients had LMCA restenosis, in 2 patients due to mechanical recoil by recompression from the PA (1 with BMS and 1 with DES) and in 3 patients due to in-stent restenosis by neointimal proliferation (2 with BMS and 1 with DES); 3 had restenosis with recurrence of symptoms. PCI was successfully repeated, with a single DES implantation in each of these patients.
At a mean follow-up of 22 ± 13 months, among the 45 patients treated with PCI, there were no cardiovascular-related deaths, MIs, strokes, or stent thromboses. Bleeding was reported in 7 patients (BARC type 2 in 6 patients, all managed conservatively, and BARC type 3 in 1 patient). Two patients (1 with a BMS and 1 with a DES) underwent lung transplantation because of progression of PAH; 1 patient died of acute graft rejection after lung transplantation.
Rates for death or double-lung transplant and rates for death, double-lung transplant, or restenosis of all 48 patients with LMCA stenosis ≥50% who underwent PCI or surgical PA remodeling are shown in the Central Illustration. At 36 months, the cumulative rate for deaths or double-lung transplant was 5%, and the cumulative rate for deaths, double-lung transplant, or restenosis was 30%.
This study indicated that the prevalence of LMCA stenosis ≥50%, because of extrinsic compression of an enlarged PA, is at least 6% in the overall PAH population of our referral center, increasing to 40% in patients with angina or angina-like symptoms who have undergone a specific diagnostic procedure. This prevalence was higher than expected, considering previously published observations. Since the first reported autopsy of a patient with LMCA stenosis in 1957 (26), 94 additional clinical cases have been documented to date. Along with 11 small case series describing 46 patients (9,11,18,27–33), reports of LMCA stenosis have mostly been published as individual case studies. In a series of 26 PAH patients with angina, 7 (26.9%) had significant LMCA stenosis by extrinsic compression of the PA (11), supporting our findings. The novelty of our study was based on the systematic adoption of CTCA for deciding on the indication for SCA and on the reporting of the short- and long-term efficacy and safety of PCI with stenting of the LMCA in the setting of extrinsic compression.
In the current study, CTCA identified patients with symptomatic PAH with a high prevalence of LMCA stenosis ≥50%, confirmed by SCA, particularly when dislocation and significant stenosis CTCA patterns were present. CTCA should be included in the diagnostic workup of patients with PAH and angina or angina-like symptoms. Therefore, in accordance with our findings, SCA is indicated if dislocation or significant stenosis patterns are recognized and may be considered in patients with a contiguity pattern, depending on the severity of the symptoms and the PA diameter.
In the current study, patients with PAH with angina and LMCA-significant stenosis, as confirmed by SCA were, on average, younger than the overall population (51 years of age) and younger than patients in registries from developed countries (34,35). The age differences might be explained by a predominance of patients with PAH associated with congenital cardiac shunts (48%) in this study and in other series (11,16). The high prevalence of congenital cardiac shunts may also explain why 70% of the patients developed angina and underwent CTCA more than 5 years after PAH diagnosis. Prolonged survival and sustained high PA pressure or high pulmonary blood flow, which are typical in patients with PAH associated with congenital heart defects, might favor the progressive dilation of the main PA. Therefore, the need for the concomitant presence of different factors might explain the lack of a strict relationship between PA pressure and LMCA extrinsic compression.
The ROC analysis identified a main PA diameter >40 mm as a strong predictor of the presence of LMCA-significant stenosis in this patient population with angina, which is consistent with findings of another study in which PA main trunk diameter ≥40 mm was identified as a predictor of LMCA extrinsic compression (11). Among 19 patients with PA main trunk diameter ≥40 mm in that study, the rate of LMCA extrinsic compression was 37% (11).
The mechanism of LMCA stenosis is strictly linked to PA dilatation, which induces the initial downward displacement of the LMCA and reduction of the take-off angle (Figure 2C). A significant stenosis at the ostium of the LMCA can be observed in this phase in some patients (Figure 2D). If the downward dislocation progresses, a longer segment of LMCA is compressed between the PA and the left aortic sinus, and a segmental stenosis can be observed (Figures 4A and 4B). However, the sensitivity (83%) and specificity (70%) of our ROC analysis showed that not all symptomatic patients with PAH and an LMCA stenosis ≥50% had a PA diameter ≥40 mm. Conversely, such a diameter does not necessarily produce significant LMCA stenosis, which may be explained by individual anatomic variability in the relationships between the PA and the LMCA.
The current study is the largest to report outcomes for patients with PAH treated with PCI and stenting for LMCA stenosis owing to extrinsic compression. PCI was preferred to coronary artery bypass grafts because of the risk related to general anesthesia and cardiopulmonary bypass in patients with severe pre-capillary pulmonary hypertension (7,8). LMCA recanalization was performed with a favorable risk-to-benefit ratio and no major in-hospital complications were observed. BMS were used only in patients with PAH who required long-term oral anticoagulation, to avoid the combination of oral anticoagulants and dual antiplatelet therapy for prolonged periods to limit bleeding risks.
Restenosis, prospectively assessed 9 months after the procedure, was observed in 5 patients (11.1%). Notably, restenosis resulted from in-stent proliferation in 3 patients and extrinsic recompression in 2 patients. PCI was successfully repeated in these cases and required the use of 1 additional DES in each case. In the majority of patients, PCI resulted in the complete and persistent relief of angina or angina-like symptoms and no major cardiovascular events were observed during follow-up (Central Illustration). This supported the long-term efficacy and safety of PCI in preventing potential severe clinical complications, which are expected in patients with LMCA stenosis ≥50%. The alternative procedure to PCI is surgical reduction plasty of PA, which was adopted in 3 patients who underwent planned or urgent thoracic surgery. In these patients, the procedure provided persistent symptomatic and prognostic benefits.
The current study had several limitations, including the use of angina or angina-like symptoms to identify the patient population. These symptoms were present in 15.8% of our patient population, which is lower than the incidence of generic “chest pain” observed in registries (mean incidence 22%) (2–5). Although the use of angina or angina-like symptoms has direct clinical relevance, it is possible that some patients with PAH affected by LMCA compression will be asymptomatic or with aspecific chest pain. We used CTCA as a noninvasive test for LMCA extrinsic compression rather than provocative tests, such as the exercise stress test with electrocardiography or myocardial scintigraphy. However, electrocardiography (7,8) and myocardial scintigraphy (6) can yield inconclusive results because of changes also seen at rest or baseline conditions.
We have not utilized intracoronary imaging to rule out atherosclerosis and stent deformation and this might have reduced our ability to detect the presence and extent of these 2 processes. Additionally, we did not use fractional flow reserve assessment to confirm the angiographic severity of the lesions. However, CTCA holds a strong predictive value for ruling out significant atherosclerotic coronary artery disease in patients with low-to-moderate probability of coronary artery disease, as was the case in our study population (36).
A further limitation of this study was the absence of an untreated control group, which we excluded because of the recognized indication for revascularization in the presence of an LMCA stenosis ≥50% (22,37).
There was a high prevalence of LMCA-significant stenosis because of extrinsic compression by dilated PA in patients with PAH and angina. CTCA is indicated in patients with PAH and angina or angina-like symptoms as an initial investigation and SCA is necessary in selected cases. PA main trunk diameter ≥40 mm predicted LMCA-significant stenosis with acceptable sensitivity and specificity. PCI with stenting provides an optimal risk-to-benefit ratio, both for symptom relief and for long-term outcome.
COMPETENCY IN MEDICAL KNOWLEDGE: LMCA stenosis due to compression by a dilated PA can be identified relatively frequently by CTCA in patients with PAH and angina, and when confirmed by SCA, patients with LMCA compression may benefit from PCI.
TRANSLATIONAL OUTLOOK: Further studies are needed to assess the natural history of LMCA compression in patients with PAH and the safety and long-term efficacy of PCI in this situation.
This work was supported by the Department of Investigational, Diagnostic and Specialty Medicine, University of Bologna (Bologna, Italy) and the National Institute of Biostructures and Biosystems (Rome, Italy). The study sponsor/funder had no role in the design and conduct of the study; collection, analysis, and interpretation of the data; and preparation, review, or approval of the manuscript. Editorial support was provided by Lynda McEvoy, PhD (ApotheCom Ltd., London, United Kingdom) funded by Actelion Pharmaceuticals Ltd. Dr. Saia has received personal fees from Abbott Vascular, Eli Lilly, AstraZeneca, Boston Scientific, Medtronic Inc., The Medicines Company, and St. Jude Medical, outside the submitted work. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- Bleeding Academic Research Consortium
- bare-metal stent(s)
- confidence interval
- computed tomography coronary angiography
- drug-eluting stent(s)
- left main coronary artery
- myocardial infarction
- pulmonary artery
- pulmonary arterial hypertension
- percutaneous coronary intervention
- quantitative coronary analysis
- receiver-operating characteristic
- selective coronary angiography
- Received February 12, 2017.
- Revision received March 8, 2017.
- Accepted March 29, 2017.
- 2017 American College of Cardiology Foundation
- Galie N.,
- Simonneau G.
- Gomez A.,
- Bialostozky D.,
- Zajarias A.,
- et al.
- Galie N.,
- Humbert M.,
- Vachiery J.L.,
- et al.
- Galie N.,
- Humbert M.,
- Vachiery J.L.,
- et al.
- Karrowni W.,
- Sigurdsson G.,
- Horwitz P.A.
- Demerouti E.A.,
- Manginas A.N.,
- Athanassopoulos G.D.,
- Karatasakis G.T.
- Dodd J.D.,
- Maree A.,
- Palacios I.,
- et al.
- Russo V.,
- Gostoli V.,
- Lovato L.,
- et al.
- Reiber J.H.,
- Serruys P.W.,
- Kooijman C.J.,
- et al.
- Windecker S.,
- Kolh P.,
- Alfonso F.,
- et al.
- Cutlip D.E.,
- Windecker S.,
- Mehran R.,
- et al.
- Mehran R.,
- Rao S.V.,
- Bhatt D.L.,
- et al.
- Ngaage D.L.,
- Lapeyre A.C.,
- McGregor C.G.
- Hulten E.,
- Villines T.C.,
- Cheezum M.K.,
- et al.
- Conley M.J.,
- Ely R.L.,
- Kisslo J.,
- Lee K.L.,
- McNeer J.F.,
- Rosati R.A.