Author + information
- Received July 27, 2017
- Revision received September 28, 2017
- Accepted October 23, 2017
- Published online January 1, 2018.
- Rebecca Dann, BSca,
- Tarik Hadi, PhDb,
- Emilie Montenont, PhDa,
- Ludovic Boytard, PhDb,
- Dornaszadat Alebrahim, MDb,
- Jordyn Feinstein, MAb,
- Nicole Allen, BSca,
- Russell Simon, MDb,
- Krista Barone, BScb,
- Kunihiro Uryu, PhDc,
- Yu Guo, MAa,
- Caron Rockman, MDb,
- Bhama Ramkhelawon, PhDb,d,∗ ( and )
- Jeffrey S. Berger, MD, MSa,b,∗∗ ()
- aDivisions of Cardiology and Hematology, Department Medicine, New York University School of Medicine, New York, New York
- bDivision of Vascular Surgery, Department of Surgery, New York University School of Medicine, New York, New York
- cElectron Microscopy Resource Center, The Rockefeller University, New York, New York
- dDepartment of Cell Biology, New York University School of Medicine, New York, New York
- ↵∗Address for correspondence:
Dr. Bhama Ramkhelawon, Division of Vascular Surgery, Department of Surgery, New York University School of Medicine, 530 First Avenue, New York, New York 10016.
- ↵∗∗Dr. Jeffrey S. Berger, Divisions of Cardiology and Hematology, Department of Medicine, New York University School of Medicine, 530 First Avenue, New York, New York 10016.
Background Peripheral artery disease (PAD), a diffuse manifestation of atherothrombosis, is a major cardiovascular threat. Although platelets are primary mediators of atherothrombosis, their role in the pathogenesis of PAD remains unclear.
Objectives The authors sought to investigate the role of platelets in a cohort of symptomatic PAD.
Methods The authors profiled platelet activity, mRNA, and effector roles in patients with symptomatic PAD and in healthy controls. Patients with PAD and carotid artery stenosis were recruited into ongoing studies (NCT02106429 and NCT01897103) investigating platelet activity, platelet RNA, and cardiovascular disease.
Results Platelet RNA sequence profiling mapped a robust up-regulation of myeloid-related protein (MRP)-14 mRNA, a potent calcium binding protein heterodimer, in PAD. Circulating activated platelets were enriched with MRP-14 protein, which augmented the expression of the adhesion mediator, P-selectin, thereby promoting monocyte–platelet aggregates. Electron microscopy confirmed the firm interaction of platelets with monocytes in vitro and colocalization of macrophages with MRP-14 confirmed their cross talk in atherosclerotic manifestations of PAD in vivo. Platelet-derived MRP-14 was channeled to monocytes, thereby fueling their expression of key PAD lesional hallmarks and increasing their directed locomotion, which were both suppressed in the presence of antibody-mediated blockade. Circulating MRP-14 was heightened in the setting of PAD, significantly correlated with PAD severity, and was associated with incident limb events.
Conclusions The authors identified a heightened platelet activity profile and unraveled a novel immunomodulatory effector role of platelet-derived MRP-14 in reprograming monocyte activation in symptomatic PAD. (Platelet Activity in Vascular Surgery and Cardiovascular Events [PACE]; NCT02106429; and Platelet Activity in Vascular Surgery for Thrombosis and Bleeding [PIVOTAL]; NCT01897103)
Peripheral artery disease (PAD) is a clinical manifestation of systemic atherosclerosis provoking stenosis of the arteries supplying the lower limbs. It is estimated that more than 200 million people have PAD worldwide (1), and individuals with symptomatic PAD are at heightened risk for cardiovascular morbidity and mortality accompanied by impairment of quality of life (2). Although the pathogenesis of atherosclerosis affecting major coronary arteries is well characterized, the physiopathology of PAD, which manifests preferentially at lower extremity territories, is still unclear, and the mechanisms that regulate this complex disorder are not well understood.
Cumulative clinical and experimental studies have well established that platelets directly contribute to the development of atherothrombosis (3–5). Notably, antiplatelet therapies targeting thrombus formation are the mainstay for vascular occlusive diseases such as acute myocardial infarction and stroke. Although increased circulating platelet activity was previously described to correlate with the occurrence of PAD, knowledge of the mechanisms pertaining to this clinical observation is still lacking. Besides, thrombus-free atherosclerotic lesions causing stenosis also manifest in asymptomatic individuals with PAD, suggesting that in addition to their thrombotic potential, platelets can perform alternative sentinel functions in PAD. Coincidentally, the immunomodulatory role of platelets has been illuminated in early atherosclerosis (6–9). Because neither thrombosis nor rupture is allocated to lesions in their infancy, we can speculate that at this stage, when inflammation culminates, platelets can interact with key immune players to drive atherosclerosis. Indeed, being uniquely positioned in the peripheral blood, platelets can perform effector functions and act at the interface of the major inflammatory cell population by engaging different receptor–ligand bridges to direct the inflammatory response in tissues. As such, increased circulating platelet–leukocyte clusters have been demonstrated in patients with myocardial infarction, with angina, and during cardiopulmonary bypass (10). Recently, a deleterious role of platelet and neutrophil extracellular trap interaction was also described in sepsis (11). Pharmacological attempts have proven promising in disrupting such platelet-immune networks in different pathological settings (12). Collectively, these studies reinforce the role of activated platelet interactions in diseases, supplementary to their role in thrombus formation. However, so far, clinical evidence and functional evaluation of such platelet-immune complexes are completely undefined in the context of PAD.
Given that activated platelets are a major culprit in the pathogenesis of atherothrombosis and are prone to form interactions with circulating cellular subsets, we aimed to determine the effector role of platelets in a cohort of PAD. Here, we show that monocyte–platelet aggregates (MPA) were robustly increased in PAD and that myeloid-related protein (MRP)-14 harbored in activated platelets was instrumental in this process. Platelet-derived MRP-14 further activated monocytes by inducing the expression of proatherogenic markers and their directed migration. Moreover, MRP-14 was higher in PAD patients with critical limb ischemia (CLI) and was predictive of incident cardiovascular and limb events. Together, our present results identify a novel and central effector role for platelet-derived MRP-14 to synergize physically with activated monocytes and foster inflammation in PAD.
Details of the experimental methods are available in the Online Appendix. Briefly, informed consent and recruitment were performed under the New York University Langone Medical Center Institutional Review Board. Patients with PAD and carotid artery stenosis (CAS) were recruited into a biorepository and ongoing studies (NCT02106429 and NCT01897103) investigating platelet activity.
Platelets were isolated by negative CD45 selection as previously described (13). To ensure purity of platelet population, only samples with low CD45 transcript levels (greater than 35 cycles) were used for RNA sequencing. In addition, only samples with platelet-leukocyte ratio of 1×107 were used for platelet transcriptomics (14,15). MPA were measured by double CD61+CD14+ staining and imaged using scanning electron microscopy.
Data are reported as mean ± SEM or SD where appropriate. Clinical studies in which results were not normally distributed, median and interquartile range (first to third quartiles) were presented. Statistical significance between 2 experimental groups was performed using a parametric Student’s t-test, nonparametric Mann-Whitney U test, or Spearman test for correlations datasets.
Conditional logistic regression was used to estimate relative risk and 95% confidence interval after the population was divided into groups on the basis of the median cutpoint for MRP-14. Adjusted risk estimates were obtained from regression models that adjusted for cardiovascular risk factors including age, sex, race/ethnicity, smoking status, and history of diabetes, hyperlipidemia, and coronary artery disease. All probability values were 2-tailed, and p values of <0.05 were considered statistically significant.
Platelet function is augmented in symptomatic PAD
Platelet counts of PAD patients were not different from age-matched healthy controls (Figure 1A). Mean platelet volume, a standardized platelet size measurement, and the percentage of reticulated platelets, a subtype of nascent RNA-enriched platelets, were similar in PAD versus controls (Figures 1B and 1C). Interestingly, platelet activity characterized by surface expression of the adhesion protein, P-selectin, was significantly enhanced both at baseline and following thrombin activation (Figure 1D). PAC-1 expression was higher in stimulated platelets in PAD (Figure 1E). Consistently, platelet aggregation in response to submaximal agonist stimulation was increased in PAD (Figures 1F to 1I). Collectively, these findings demonstrated that platelets exhibit increased activation capacity, suggesting a causative role in PAD.
PAD platelets are enriched with MPR-14
Because platelet transcriptional mapping can provide key mechanistic insights into the role of platelets in diseases, we performed RNA sequencing of platelets isolated from PAD and healthy subjects matched for age, sex, race/ethnicity, and aspirin use. Platelet RNA profiling of 3 healthy and 3 PAD subjects generated a library of distinct expression patterns (Figure 2A). Utilizing a fold-change cut of >1.5 with a nominal p value <0.05, we found a unique signature for MRP-14, a potent chemoattractant and myeloid cell function regulator. Pathway analysis using DAVID software version 6.8 (National Institute of Allergy and Infectious Diseases, Bethesda, Maryland) confirmed that our screening was purified platelet profiling and free from leukocyte contamination (data not shown) as revealed by gene patterns enriched in blood coagulation and hemostasis (Figure 2B). Notably, a panel of inflammatory genes was up-regulated in PAD, suggesting that platelets could act at the interface of immune cells in PAD. Reverse-transcription polymerase chain reaction analysis of a larger cohort confirmed the robust expression of MRP-14 transcripts in PAD platelets (Figure 2C). Enzyme-linked immunosorbent assays performed using platelet releasates (PR) demonstrated that MRP-14 concentration was significantly higher in the diseased versus the healthy cohort. Thrombin-activated platelets induced a further increase in detectable levels of MRP-14 (Figure 2D). PAD platelets seeded on fibrinogen-coated slides stained positive for both the platelet-specific marker CD61 (red) and MRP-14 (green) (Figure 2E). These findings suggested that consistent with the enhanced MRP-14 mRNA concentrations, elevated MRP-14 protein was stored in platelets in patients with PAD. Because MRP-14 is secreted, we next measured its circulating concentrations. MRP-14 was elevated 2-fold in the plasma of PAD compared with controls. A significantly higher concentration of MRP-14 was detected in serum and was further increased in PAD (Figure 2F). To establish a causative association between unbound circulating levels of MRP-14 and platelet-derived MRP-14, we performed a correlation analysis, which revealed a positive association between platelet-derived and circulating MRP-14 (Figure 2G).
Platelet-derived MRP-14 promotes MPA via P-selectin
To gain insight into the role of MRP-14 in PAD, we assessed its effect on platelet aggregation, given its reported role in thrombus formation (16). Stimulation of platelets with increasing doses of recombinant MRP-14 failed to modulate their aggregation induced by epinephrine (Figure 3A), adenosine diphosphate (ADP), and collagen (data not shown). Consistent with these findings, flow cytometry assays demonstrated that PAC-1 surface expression remained unchanged when platelets were stimulated with recombinant MRP-14 (Figure 3B). Interestingly, stimulation of platelets with MRP-14 dose-dependently enhanced the surface expression of P-selectin under basal conditions and following thrombin activation (Figure 3C). Because P-selectin is a key cell adhesion mediator, we reasoned that activated platelets expressing P-selectin could readily interact with other immune cells. To verify this, we assessed leukocyte–platelet aggregates (LPA) characterized by double-positive CD45+CD61+ gating of subcellular populations. LPA was significantly increased in PAD patients compared with healthy controls (Figure 3D). Although the percentage of LPA positively associated with platelet MPR-14 mRNA, no correlation with circulating protein levels of MRP-14 was detected (Online Figures 1A and 1B). To further dissect which specific cellular subtypes within the CD45+ leukocyte population was interacting with platelets, we analyzed MPA characterized by double-positive CD14+CD61+ clusters since monocytes are drivers of atherosclerosis. MPA were significantly increased in PAD (Figure 3E), suggesting that activated platelets readily liaise with monocytes in PAD. Notably, CD14+CD61+ significantly correlated with both MRP-14 transcript (Online Figure 1C) and protein levels in this setting (Figure 3F), consistent with the increased display of P-selectin in activated platelets. Indeed, activated platelets bound more to primary CD14+ monocytes in a P-selectin–dependent manner (Figure 3G). Consistently, MRP-14–stimulated platelets interacted more with THP-1 cells, which was abrogated in the presence of P-selectin blocking antibody (Figure 3H). This was paralleled by the increased surface expression of P-selectin glycoprotein ligand-1 (PSGL-1), the primary ligand for P-selectin (Figure 3I) and mRNA expression in THP-1 monocyte-like cells (Figure 3J).
Platelet-derived MRP-14 activates the inflammatory profile of monocytes
As illustrated in Figure 4A, incubation of PAD platelets (stained in green) with THP-1 cells revealed that they could indeed build solid connections with activated monocytic cells (Figure 4A). Meticulous analysis using zoomed images showed that platelets either adhered to the surface of monocytes (box 1), fused with their cytoplasmic constituents (box 2), or in some cases, established tubular connections with monocytes (far right, box 3) (Figure 4B). Immunofluorescence of human PAD lesions revealed a colocalized expression of CD61 platelets (green) and CD68 macrophages (red) confirming the relevance of MPA in vivo (Figure 4C).
We then reasoned that within the MPA, platelets and monocytes being in close proximity, MRP-14 could be transferred to monocytes to further activate them. To test this hypothesis, we stained for MRP-14 in monocytes following incubation with PR isolated from PAD patients. We detected a robust MRP-14 staining harbored in monocytes stimulated with PAD PR as opposed to healthy PR. Quantification of the mean fluorescence confirmed a significant increase of MRP-14 in PAD (Figure 4D). Prior depletion of MRP-14 in PAD PR using anti–MRP-14 antibody blunted the expression of MRP-14 in primary monocytes as opposed to immunoglobulin G (IgG) control antibody (Figure 4E). Similar results were obtained with monocytic THP-1 cells (Online Figure 2). Stimulation of monocytes with recombinant MRP-14 significantly increased the mRNA expression of atherosclerosis hallmarks including interleukin-1β, tumor necrosis factor α, and chemokine (C-C motif) ligand 2 (Figure 4F).
Notably, incubation of monocytic cells with PAD PR, but not with healthy PR, increased the mRNA expression of these 3 proinflammatory markers (Figure 4G). This was dependent of MRP-14, because pre-incubation of the PR with a capture antibody directed against MRP-14 abrogated their induction (Figure 4H). Collectively, these data reinforced the associative role of MPA and MRP-14 in driving inflammation in PAD.
Platelet-derived MRP-14 drives monocyte migration
THP-1 cells were next challenged to migrate to PR collected from either healthy or PAD individuals through a real-time migration detection assay. Increased migration to PAD PR compared with healthy PR (Figure 5A) was observed. PAD increased migration, which was reversed in the presence of MRP-14 blocking antibody, but not the control IgG antibody (Figure 5B). Notably, the chemotaxic profile of THP-1 cells to recombinant MRP-14 was faster than when challenged to migrate to vehicle (Figure 5C).
To delve into the mechanism by which MRP-14 promotes monocyte migration, we assessed its effect on actin cytoskeleton reorganization, a key process that precedes directed migration. PAD-derived PR induced characteristic membrane ruffles, lamellipodia, and filapodia, indicative of actin reorganization and rapid cell spreading and movement. PR concentrates from healthy subjects maintained a quasi spherical shape, consistent with the reduced-motility morphology. Quantification of actin filaments by immunofluorescence confirmed these findings (Figure 5D). In accordance, staining of serial PAD lesions revealed that MRP-14 (green) was highly expressed in CD68-positive (red) cells that accumulated not only at the interface of the intima and the media, but also in the deeper medial layers as indicated by the arrows (Figures 5F and 5G). MRP-14 was also detected in enucleated extracellular regions within the thrombotic areas, consistent with our data demonstrating the colocalized expression of MRP-14 with CD61+ platelets in lesions (Figure 5E). Collectively, these results demonstrate that MRP-14 originating from platelets coordinates macrophage activation that might contribute to the complications associated with PAD.
MRP-14 is higher in PAD than in carotid artery disease and correlates with PAD severity
To determine whether MRP-14 levels are specific to PAD, we measured serum MRP-14 in patients with lower extremity PAD and CAS. As shown in Figure 6A, there was a significant increase in MRP-14 in PAD versus CAS (276.6 ± 11.62 mg/l vs. 222.3 ± 21.55 mg/l; p = 0.04). Among the 68 PAD patients, 43 (63.2%) were found to have a more severe phenotype of PAD of CLI, defined by rest pain, ulceration, or gangrene. Patients with CLI had significant elevated levels of MRP-14 compared with a less severe PAD phenotype (Figure 6B) (298.5 ± 98.9 mg/l vs. 234.8 ± 74.6 mg/l; p = 0.004).
Increased MRP-14 in PAD patients with incident cardiovascular and limb events
Patients with PAD are at significant risk for major adverse cardiovascular and limb events (MACLE) (death, myocardial infarction, stroke, or amputation). For this reason, we sought to further validate MRP-14 gene target in 68 PAD patients undergoing lower extremity revascularization and followed for a median 878 days (interquartile range: 605 to 949 days). Baseline characteristics of study participants are shown in Online Table 1. PAD patients with versus without a MACLE had significantly higher MRP-14 levels at baseline (303.1 ± 113.6 mg/l vs. 247.0 ± 62.42 mg/l; p = 0.01). As noted in Figure 6C, the cumulative incidence of MACLE differed according to the median MRP-14 levels. After adjusting for age, sex, race/ethnicity, smoking status, and prior diabetes, hypertension, and coronary artery disease, values above the median were associated with a 2.5-fold increase in MACLE (hazard ratio: 2.5, 95% confidence interval: 1.1 to 5.9; p = 0.03).
Despite the high incidence of limb events, PAD is commonly overlooked and underdiagnosed by the medical community (2,17). An important impediment to the awareness and care of these patients is the limited understanding of the mechanisms that contribute to the pathogenesis of PAD (18). Although inflammation is important for the initiation and progression of PAD (19), a fundamental question relating to the factors responsible for the focal increase of these inflammatory agents remain poorly defined. The current study presents information to expand our current knowledge of PAD by providing novel insights into the characteristic platelet phenotype, transcriptome, and the role of platelet as an effector cell in patients with symptomatic PAD (Central Illustration).
The platelet phenotype observed in this population is consistent with a pioneer study that noted increased platelet activity in obstructive arterial disease more than 3 decades ago (20). Follow-up studies in PAD cohorts demonstrated increased platelet surface expression markers in stable PAD (21) and increased platelet activity across the spectrum of PAD severity (22). However, other attempts to characterize platelet activation in PAD have yielded conflicting results (23,24), most likely attributable to technical issues in sample handling. Although these studies demonstrated platelet activation in PAD, nonetheless, they did not delve into the mechanism underlying these observations.
We used a high-throughput genetic RNA sequencing approach to screen for potential candidates responsible for the increased activation of platelets observed in PAD. We identified MRP-14 among the most differentially expressed transcripts harbored in platelets from PAD patients compared with age-, sex-, and race/ethnicity-matched controls. To our knowledge, this is the first time that such a screening has been performed in platelets in PAD. A microarray profiling strategy was performed by Healy et al. (25), where they revealed the up-regulation of MRP-14 transcript levels in platelets isolated from patients presenting with myocardial infarction. The consistency with our findings is not surprising because patients with lower extremity PAD experience a high risk of incident atherothrombosis, including myocardial infarction, stroke, and acute limb ischemia (26,27). Here, we suggest that patients with PAD may be at particular risk for elevated MRP-14 levels. In fact, CLI (a severe phenotype of PAD) patients were noted to have significantly higher levels of MRP-14. In a prospective cohort of PAD patients undergoing lower extremity revascularization, higher MRP-14 was associated with adverse cardiovascular and limb events.
Attempts to infuse activated platelets in atherosclerosis-prone apolipoprotein-E–deficient (Apoe−/−) mice accelerated lesion formation partly via the increased interaction of platelet P-selectin with monocytes (28). Here, we demonstrate that activated platelets can directly interact with CD14+ monocytes. This firm interaction engaged the transfer of MRP-14 to activated circulating monocytes and tissue macrophages. This is consistent with the observation that activated platelets can deliver chemokines such as RANTES and PF4 to vascular and immune cells (28). These authors observed an effector role for MPA in mouse models that reshaped the inflammatory response by increasing the surface expression of adhesion molecules on the endothelium. Remarkably, we demonstrate that the depletion of platelet-derived MRP-14 using directed antibodies repressed monocytic migration and blunted their expression of proinflammatory markers.
Platelets contain approximately 60 granules that store key molecules with effects on wound healing, immune activity, and inflammation (29). Over the last decade, growing research has enabled the discovery of novel effector roles for platelets by demonstrating that they can carry chemokine and proinflammatory proteins. Here, we demonstrate that activated platelets may contribute to the pool of MRP-14 in PAD. Although other cell types, including neutrophils (30), monocytes, and tissue macrophages (31–33), express MRP-14, platelets and megakaryocytes may serve as an additional source (16,25). Immunofluorescence staining of purified platelets and aortoiliac thrombi demonstrate that MRP-14 is expressed in platelets. MRP-14 was elevated in PR and in serum and plasma of PAD patients. In such a pathological scenario, we could imagine that the display of unbound MRP-14 could also be sequestered by activated platelets, thereby increasing their intracellular pool, although this was not assessed in our present study. Our results indicate that MRP-14 is increased in platelets in the basal state and following thrombin activation, suggesting that under favorable conditions, platelets are able to synthetize MRP-14 protein de novo. As such, enucleated platelets emanate from extensions of megakaryocytes, resident in the bone marrow, carry an imprint of RNA, and are equipped with transcriptional machinery capable of protein biosynthesis. Consistently, we observed an increased MRP-14 mRNA expression in platelets of PAD patients.
To understand the functional role of MRP-14 in platelets, we stimulated platelets with recombinant MRP-14. Surprisingly, MRP-14 induced exclusively the adhesive capacity of platelets characterized by augmented P-selectin surface expression but did not have an effect on PAC-1 expression or platelet–platelet aggregation. These data suggest that the signal transduction underlying platelet adhesion differs from those involved in aggregation and that MRP-14 seems to selectively participate in the former. Consistently, the genetic deletion of MRP-14 revealed a reduction in platelet expression of P-selectin in response to collagen and arachidonic acid, independent of platelet aggregation (16). P-selectin plays an essential role in platelet–leukocyte interaction and atherosclerotic lesion formation. In a vascular injury model, global MRP-14 deficiency reduced leukocyte accumulation and atherosclerotic area. Because the study was performed using a global deletion approach, the specific role of platelet-derived MRP-14 could not be elucidated (34).
Taken together, our data underscore the important role of MRP-14 in synchronizing platelet activity and directing monocytic inflammation in PAD. Our results demonstrate a novel effector role of platelets in PAD and indicate that targeting MRP-14 in platelets may dampen inflammation and have therapeutic value in PAD.
COMPETENCY IN MEDICAL KNOWLEDGE: Patients with symptomatic PAD have increased platelet activity and high platelet MRP-14 levels that promote platelet–monocyte aggregates and are associated with severe chronic limb ischemia and major adverse clinical limb events. Circulating levels of MRP-14 may have clinical utility as a biomarker that correlates with the severity of PAD.
TRANSLATIONAL OUTLOOK: Clinical studies are needed to assess the therapeutic value of MRP-14 to reduce the complications associated with PAD.
The authors thank Drs. Adriana Heguy, Aristotelis Tsirigos, and Tenzin Lhakhang from the NYU Genome Technology Center for RNA sequencing data.
This work was supported in part by the National Heart, Lung, and Blood Institute of the National Institute of Health (R00 HL125667 to Dr. Ramkhelawon) and (R01HL114978 to Dr. Berger) and the American Heart Association Clinical Research Program (13CRP14410042 to Dr. Berger). The authors have reported that they have no relationships relevant to the contents of this paper to disclose. Drs. Ramkhelawon and Berger contributed equally to this work and are joint senior authors.
- Abbreviations and Acronyms
- carotid artery stenosis
- critical limb ischemia
- leukocyte–platelet aggregate(s)
- major adverse cardiovascular and limb events
- monocyte–platelet aggregate(s)
- myeloid-related protein
- peripheral artery disease
- platelet releasate
- Received July 27, 2017.
- Revision received September 28, 2017.
- Accepted October 23, 2017.
- 2018 American College of Cardiology Foundation
- Wagner D.D.,
- Burger P.C.
- Furman M.I.,
- Benoit S.E.,
- Barnard M.R.,
- et al.
- McDonald B.,
- Davis R.P.,
- Kim S.J.,
- et al.
- Shah B.,
- Allen N.,
- Harchandani B.,
- et al.
- Montenont E.,
- Echagarruga C.,
- Allen N.,
- Araldi E.,
- Suarez Y.,
- Berger J.S.
- Freedman J.E.,
- Larson M.G.,
- Tanriverdi K.,
- et al.
- Gnatenko D.V.,
- Dunn J.J.,
- McCorkle S.R.,
- Weissmann D.,
- Perrotta P.L.,
- Bahou W.F.
- McDermott M.M.
- Brevetti G.,
- Giugliano G.,
- Brevetti L.,
- Hiatt W.R.
- Healy A.M.,
- Pickard M.D.,
- Pradhan A.D.,
- et al.
- Hiatt W.R.,
- Fowkes F.G.,
- Heizer G.,
- et al.
- Jones W.S.,
- Baumgartner I.,
- Hiatt W.R.,
- et al.
- Hessian P.A.,
- Edgeworth J.,
- Hogg N.