Growth patterns of choroidal neovascularization. Each specimen is shown on a sponge (S). A, Subretinal pigment epithelium (type 1) growth pattern (arrow) beneath retinal pigment epithelium (RPE) (asterisk). B, Subretinal (type 2) growth pattern (arrow) overlying RPE (asterisk). C, Combined growth pattern with subretinal (arrow) and sub-RPE (double arrow) components. The RPE (asterisk) is in the center of the specimen (all, hematoxylin-eosin, original magnification ×25).
Submacular Surgery Trials Research Group. Histopathologic and Ultrastructural Features of Surgically Excised Subfoveal Choroidal Neovascular LesionsSubmacular Surgery Trials Report No. 7. Arch Ophthalmol. 2005;123(7):914-921. doi:10.1001/archopht.123.7.914
To identify the histologic and ultrastructural features of surgically excised subfoveal choroidal neovascular lesions from patients enrolled in the Submacular Surgery Trials and to compare them with clinical data.
Surgically excised subfoveal choroidal neovascular lesions from patients enrolled in the Submacular Surgery Trials group N trial (lesion predominantly choroidal neovascularization [CNV] with evidence of classic CNV from age-related macular degeneration), group B trial (lesion predominantly hemorrhagic from age-related macular degeneration), and group H trial (idiopathic subfoveal CNV or subfoveal CNV from ocular histoplasmosis syndrome) between October 1, 1999, and September 1, 2001, were submitted to the pathology center. The lesion growth pattern (subretinal pigment epithelial [sub-RPE], subretinal, combined, or indeterminate) and the cellular and extracellular constituents were classified independently. Demographic, clinical, and fluorescein angiographic characteristics of patients, eyes, and lesions, respectively, were compared with the pathologic features.
Of 269 patients assigned to surgery during the 24 months that pathologic specimens were collected, surgical specimens from study eyes of 199 were submitted to the pathology center. Of the 199 routine histologic specimens processed, 144 (72%) were classified as CNV, 51 (26%) as fibrocellular tissue, and 4 (2%) as hemorrhage. The median specimen size was smaller in group H (932 × 208 μm) than in groups N (1980 × 325 μm) and B (1800 × 395 μm). The CNV growth pattern was determined in 91 (46%) of 199 specimens. Of 159 group N and group B lesions, 76 (48%) had an indeterminate growth pattern, 28 (18%) had a sub-RPE growth pattern, and 33 (21%) had sub-RPE and subretinal growth patterns. Of 40 group H lesions, 32 (80%) had an indeterminate growth pattern, 7 (18%) had a subretinal growth pattern, and 1 (2%) had a combined sub-RPE and subretinal pattern. Based on electron microscopy, the most common cellular lesion components were RPE, macrophages, erythrocytes, fibrocytes, and vascular endothelium; the most common extracellular components were 24-nm collagen and fibrin. Basal laminar and linear deposits were found in 80% (40/50) and 16% (8/49) of group N specimens, 66% (43/65) and 5% (3/65) of group B specimens, and 8% (2/26) and 0% (0/26) of group H specimens, respectively.
Most surgically excised subfoveal specimens had evidence of CNV or tissue associated with CNV. The constituents in CNV were consistent with granulation tissue proliferation. The presence of basal deposits in surgically excised specimens suggested a clinical diagnosis of age-related macular degeneration, even when blood was the predominant component of the lesion. Correlation of growth patterns above or below the RPE with fluorescein angiographic patterns of classic or occult CNV was limited because most specimens had insufficient material to determine these patterns.
Choroidal neovascular lesions have been surgically excised for more than a decade.1- 3 These lesions develop in patients with age-related macular degeneration (AMD),3 ocular histoplasmosis syndrome (OHS),2 and other conditions, and they may be associated with hemorrhage.1 Although the pathologic features of AMD4 and OHS5 have been well-documented, many of these studies examined choroidal neovascularization (CNV) in eyes obtained post mortem. Recent evidence6,7 has shown that CNV is a dynamic process with several stages. The initiation stage occurs when angiogenic cytokines first induce CNV development. The active stage occurs when the CNV grows to a point where there is equilibrium between angiogenic cytokines and antiangiogenic cytokines. The inactive stage occurs when antiangiogenic cytokines cause CNV regression.
Histopathologic and ultrastructural studies of surgically excised CNV have revealed that most lesions contain inflammatory cells, retinal pigment epithelium (RPE), fibrocytes, collagen, and fibrin, regardless of the underlying disease.6- 12 Studies of postmortem eyes also have revealed inflammatory cells in CNV.4 In addition, since the advent of submacular surgery, sub-RPE (type 1), subretinal (type 2), and combined sub-RPE and subretinal growth patterns of CNV have been recognized.4,7,13,14 These growth patterns are described in detail elsewhere.7 Some investigators have suggested that the growth pattern correlates with the fundus appearance14 and the angiographic pattern of fluorescein leakage.15
Since 1997, the Submacular Surgery Trials (SST) Research Group has conducted multicenter randomized clinical trials to evaluate whether submacular surgery to remove subfoveal choroidal neovascular lesions increases the chance of stable or improved visual acuity compared with observation in patients with AMD and a predominantly neovascular lesion with evidence of classic CNV (group N), OHS or idiopathic CNV (group H), or a predominantly hemorrhagic lesion attributed to AMD (group B). The histologic and ultrastructural findings of surgically excised CNV obtained from some patients enrolled in the SST pilot study, initiated in 1993, have been reported.16 Herein we report the histologic and ultrastructural findings in CNV specimens collected prospectively from patients who enrolled in the SST at participating clinical centers between October 1, 1999 (27 months after accrual in the SST began on April 1, 1997), through September 1, 2001 (when patient accrual ended).
The SST clinical centers began enrolling patients in the group H trial on April 1, 1997, and in the group N and group B trials on July 1, 1998. Before enrollment, written consent was obtained from each patient participating in the SST. The SST pathology center (L.F. Montgomery Ophthalmic Pathology Laboratory, Atlanta, Ga) was ready to receive surgical specimens as of October 1, 1999, as additional approval to submit specimens for central processing and evaluation was obtained from the institutional review board of each participating institution. Consent to the submission of surgical specimens continued until the end of accrual on September 1, 2001. The last surgical specimen was received at the pathology center on October 9, 2001.
The SST surgeons at 21 participating institutions submitted specimens in glutaraldehyde/formalin (10% neutral buffered formalin–2.5% glutaraldehyde). The surgeons placed the specimens on a microsurgical sponge (Weck cell sponge; Edward Weck, Inc, Research Triangle Park, NC), choroidal side down when possible, with an accompanying diagram to assist the pathologists with orientation of the specimen. The patient’s SST identification code and group (N, H, or B) along with the clinical center and surgeon’s name were provided to the pathology center. All CNV specimens were measured, photographed, and processed. Specimens that were determined by gross examination to be fragmented, very small, or not well oriented on the sponge were processed for routine histologic examination and for transmission electron microscopy (TEM). Specimens that were judged by a pathologist (H.E.G.) by gross examination to be intact and well oriented on the Weck cell sponge were also processed for routine histologic examination but were then serial sectioned for 2-dimensional reconstruction.
For routine histologic examination, the specimens were dehydrated in increasing concentrations of alcohol, cleared in xylene, and embedded in paraffin. A microtome (Thermo Shandon Inc, Pittsburgh, Pa) was used to prepare two 5-μm-thick sections through the widest portion of the specimen, and the slides were stained with hematoxylin-eosin. The presence, size, and growth pattern of CNV (sub-RPE, subretinal, combined sub-RPE and subretinal, and indeterminate) were assessed as follows. Specimens were interpreted as representing CNV when vascular channels were present. Specimens were interpreted as being consistent with CNV when no vascular channels were present but fibrocellular tissue, hemorrhage, or both were present. Sections were evaluated for greatest linear diameter and maximum thickness using a standard reticule and microscope (Olympus BHTU; Olympus, Tokyo, Japan). Anatomic landmarks that enabled the characterization of growth patterns included RPE, basal laminar deposits, Bruch membrane, photoreceptor outer segments, and interphotoreceptor matrix. The orientation of the specimen with its outer surface placed on a Weck cell sponge also assisted in determining the growth patterns. Vascular channels observed only external to the RPE were interpreted as type 1 growth pattern (Figure,A); vascular channels observed only internal to the RPE, which usually appeared as a reflected layer of RPE, as type 2 growth pattern (Figure 1B); and vessels observed external and internal to the RPE as combined growth pattern (Figure 1C).
Specimens selected for TEM were postfixed with cacodylate buffer (0.1 M) and 1% osmium tetroxide solution. Standard dehydration of the specimen was performed before it was embedded in epoxy resin, sectioned, and stained with toluidine blue. These toluidine blue–stained sections were evaluated in the same manner as the routine histology specimens, thus permitting evaluation of the growth pattern. Semithin (0.1-μm) sections were cut and stained with uranyl acetate–lead citrate. A minimum of 20 micrographs per specimen were examined. Previously reported criteria16 were used to identify specific cell types and extracellular material (eg, RPE, vascular endothelium, fibrocytes, macrophages, myofibroblasts, glial cells, photoreceptors, lymphocytes, and extracellular components, including collagen, fibrin, basal laminar deposits, basal linear deposits, fragments of the Bruch membrane, and the choroid).
After the routine histology sections were obtained for initial diagnosis, 2-dimensional reconstructions were created in selected cases. Serial 5-μm-thick sections were made using a microtome (Thermo Shandon Inc), numbered, and sequentially stained with hematoxylin-eosin, periodic acid–Schiff, Masson trichrome, and Prussian blue for iron; 1 section was left unstained. The slides were sequentially examined using a microscope (Olympus BHTU) with a standard reticule, and landmarks (RPE, vascular channels, and the Bruch membrane) were mapped as previously described17,18 for future comparison with postsurgery angiograms.
Two observers (H.E.G. and W.R.G.) examined all the routine histology slides, toluidine-stained sections, and TEM findings. Based on all the slides available, the largest diameter and thickness of each specimen were determined, and an average of the 2 observers’ measurements was recorded. The presence and growth pattern of CNV was compiled in composite form after discrepancies between the observers were resolved by mutual agreement. A final histologic diagnosis was reported for each specimen. In some cases, serial sections made for 2-dimensional reconstruction provided additional evidence of CNV growth patterns. Representative serial sections of those specimens and a random sample of specimens that did not provide additional evidence of CNV growth patterns were reviewed by the 2 observers. After adjudication, a final determination of the growth pattern was recorded.
Baseline demographic, clinical, and fluorescein angiographic characteristics were recorded before randomization for the SST. The design and procedures used in the SST have been described elsewhere.19 Briefly, best-corrected visual acuity was measured using modified Bailey-Lovie charts (Early Treatment Diabetic Retinopathy Study charts)20 by SST-certified vision examiners. Baseline fluorescein angiograms were submitted to the SST photograph reading center (The Wilmer Eye Institute, Baltimore, Md), where they were reviewed by trained readers.
Data analysis included calculation of counts and percentages for categorical variables and calculation of distributional characteristics, such as median and interquartile range, for continuous variables. Because of the descriptive nature of this article, no formal statistical test results are reported. Data analysis was performed using a software program (SAS; SAS Institute Inc, Cary, NC) on a UNIX platform.
A total of 199 specimens, representing 74% of all lesions excised from the eyes of patients enrolled in the SST between October 1, 1999, and September 1, 2001, were processed. Three additional specimens were submitted to the pathology center but could not be processed because the specimen was either too small (2 cases) or was lost during processing (1 case). The characteristics of the 199 patients from whom specimens were processed are summarized in Table 1. As expected, group H patients were younger than those in groups N and B. The sex distribution was similar in all 3 trials, with slightly more specimens from women than from men in all groups. Visual acuities were considerably better in study eyes of patients in group H (median, 20/64) than in groups N (median, 20/200) and B (median, 20/250) and likely reflect, in part, the differences among the trials in visual acuity requirements for eligibility (20/50 to 20/800 for group H, 20/100 to 20/800 for group N, and 20/100 to light perception for group B).
The histologic findings in group N, B, and H specimens are given in Table 2 by whether the histologic growth pattern was determined. Lesion configuration could not be determined in 30 (34%) of 87 group N specimens, 46 (64%) of 72 group B specimens, and 32 (80%) of 40 group H specimens. Overall, the median specimen size was smaller in group H (932 × 208 μm) than in group N (1980 × 325 μm) or group B (1800 × 395 μm). Group B and group H cases with gradable lesion configuration had larger median specimen diameters (group B: 2000 vs 1725 μm and group H: 1295 vs 880 μm) and smaller median thicknesses (group B: 364 vs 440 μm and group H: 190 vs 208 μm) than cases in which lesion configuration could not be determined. Group N cases with lesion configuration were also thinner (median thickness: 325 vs 348 μm) but had smaller median diameters (1980 vs 2065 μm) than group N cases with indeterminate lesion configuration. The observable lesion configurations in each group are given in Table 2. Of 199 cases, the final histopathologic diagnosis was CNV for 144 (72%), fibrocellular tissue for 51 (26%), and hemorrhage with no CNV for 4 (2%). Although CNV was diagnosed in most specimens, vessels were sparse in almost all of these specimens. When identified, the vessels usually were located in the center of the specimen evaluated. There were no unexpected histopathologic diagnoses, except for 1 lymphocytic infiltrate consistent with uveitis in a group H patient. Review of this patient’s medical record by the submitting ophthalmologist confirmed that the patient had uveitis. There also was 1 specimen that consisted solely of surgically excised Bruch membrane and choroid.
The cellular and extracellular components of 143 specimens selected for electron microscopy are given in Table 3. For all groups, RPE, macrophages, vascular endothelium, fibrocytes, erythrocytes, collagen, and fibrin were found in most (>50%) of the specimens. Basal laminar deposits were present in 80% (40/50) of group N, 66% (43/65) of group B, and 8% (2/26) of group H. Basal linear deposits were present in 16% (8/49) and 5% (3/65) of specimens in groups N and B, respectively; none were found in group H specimens.
The angiographic lesion characteristics compared with the histologic CNV growth patterns for 199 specimens are given in Table 4. Although numerous lesions had histologically indeterminate growth patterns, of the 91 (46%) of 199 lesions in which a growth pattern could be determined, an angiographic diagnosis of AMD was made in eyes with lesions that had histologic sub-RPE, subretinal, or combined growth patterns. An angiographic diagnosis of OHS was made only in eyes with subretinal (n = 5) and combined (n = 1) growth patterns of CNV, and idiopathic CNV was diagnosed only when the CNV was subretinal (n = 2). The histologic CNV growth pattern in most CNV diagnosed angiographically to be OHS was indeterminate (30 of 36), and the growth pattern of half of the CNV diagnosed angiographically as idiopathic was indeterminate (2 of 4). Angiographic patterns of classic CNV, occult CNV, or both were observed in most of the eyes for which specimens were submitted. Twenty-two percent (6/27) of the specimens with a sub-RPE growth pattern had occult CNV only without a classic component, and 37% (10/27) of the specimens with a subretinal growth pattern had classic CNV only without an occult component. Of 189 cases that had gradable angiograms for comparison with CNV growth pattern, 53% (n=101) had an indeterminate growth pattern. Blood composing most of the lesion and larger lesion size (>12 mm) were more common in the sub-RPE, combined, and indeterminate growth patterns than in the subretinal growth pattern.
This case series provides the largest histopathologic study of surgically excised neovascular lesions reported to date, to our knowledge, with 199 specimens examined. In 2 other studies,11,16 123 and 78 specimens were examined. In this prospective study, 2 experienced ophthalmic pathologists examined surgically excised CNV specimens from a broad spectrum of clinical centers throughout the United States accompanied by fluorescein angiographic findings recorded independently by experienced retinal photograph readers. The ratio of CNV specimens examined by TEM to those studied by histologic examination alone in the present study was 2.5:1 and is similar to the ratio in another large series.11 The proportion of CNV specimens from group N (AMD) vs group H (OHS and idiopathic CNV) was similar to that of the SST pilot study.16 In the present study, 72% of the specimens were classified as CNV, 26% as fibrocellular tissue, and 2% as hemorrhage. This finding confirms a high accuracy in the clinical diagnosis of CNV based on specific fluorescein angiographic definitions,21 as the fibrocellular tissue with hemorrhage is likely to be associated with CNV.16 There was a lower percentage of fibrovascular tissue in this study compared with the SST pilot study,16 probably because specimens from group B (submacular hemorrhage) eyes consisted mostly of hemorrhage and fibrocellular tissue. Previous histologic studies examined only surgically excised specimens from eyes in which CNV was visible clinically or angiographically. The most common constituents of surgically excised neovascular lesions in this study were RPE, macrophages, vascular endothelium, fibrocytes, erythrocytes, collagen, and fibrin. These components have previously been shown to be common in surgically excised CNV8- 12,16 and are found in granulation tissue (ie, a wound repair response). Blood vessels identified in the specimens were more infrequent than some SST Research Group members might have expected from typical fluorescein angiograms of group H or group N lesions. Basal laminar deposits were found more often in lesions excised from group N and group B patients than from group H patients; basal linear deposits were found only in specimens from groups N and B. It has been shown that these basal deposits are found in patients with AMD,16 although it has been suggested that basal linear deposits may be more specific for AMD22 and that basal laminar deposits may be associated with aging but not necessarily specific for AMD.22 Of note in this study was the high percentage of inflammatory cells in specimens regardless of the underlying disease. Macrophages were present in the center in 84% (41/49) of group N specimens, 76% (50/66) of group B specimens, and 85% (23/27) of group H specimens. Choroidal neovascularization is hypothesized to be a dynamic process that includes several stages (macrophage recruitment into the CNV, inflammatory active, and inflammatory inactive [involutional]).6 Fibrin, which was present in 79% of all CNV specimens on TEM, likely serves as a scaffold for CNV growth.6 Future studies are planned to address the biological features of CNV that correspond to fundus appearance, angiographic pattern, and clinical outcomes after submacular surgery.
An important weakness of the present study is that the growth pattern in most specimens (54%) could not be determined, thereby limiting conclusions and comparisons with the angiographic appearance of CNV. Strengths of this study are that the specimens were obtained in controlled, prospective, multicenter trials with defined protocols and that histopathologic examinations and photograph gradings were performed independently. This study supports a high clinical diagnostic accuracy by SST ophthalmologists of CNV based on specific fluorescein angiographic patterns of leakage, that CNV composition is consistent with granulation tissue proliferation, and that the growth patterns of surgically excised CNV obtained from patients enrolled in the SST may correspond to the biological features of CNV in specific diseases.7
The median size of the CNV tissue from patients in groups N and B was larger than that from group H patients. This finding is consistent with that from the SST pilot study, in which CNV from patients with AMD was larger than that from patients without AMD,16 and with the distributions of lesion sizes observed on fluorescein angiograms.
Choroidal neovascularization may grow predominantly between the RPE and the Bruch membrane (sub-RPE or type 1 pattern), between the retina and the RPE (subretinal or type 2 pattern), or with approximately equal components (combined pattern).7,13,14 Theoretical constructs of CNV growth patterns represent a spectrum, and there are many variations.7 There is some evidence that the “classic” fluorescein angiographic pattern is associated with a subretinal CNV growth pattern,15 although this association was not confirmed by our study (Table 4). There were cases in this series with purely classic CNV on angiography that were determined to have a sub-RPE growth pattern. In this study, we classified the growth pattern of the surgically excised CNV in 66% (57/87) of group N patients, a higher rate than the SST pilot study.16 However, most specimens from group H and group B eyes could not be classified. This discrepancy was likely due to the large size of group N lesions and the presence of basal deposits that helped with orientation.
The biological features of CNV growth in the context of underlying disease7 seem to correspond to the size and configuration of the surgically excised specimen. The CNV growth between the RPE and the Bruch membrane (type 1 pattern), as may occur in AMD, starts with multiple ingrowth sites.23 A cleavage plane between basal laminar deposits and the Bruch membrane facilitates lateral extension of the CNV into the sub-RPE space.4,7,14 The combined growth pattern in AMD occurs when there is extension of the sub-RPE pattern into the subretinal space. Therefore, the appearance of the sub-RPE (type 1), subretinal (type 2), or combined CNV growth pattern may be expected in surgically excised lesions from group N. It seems that subretinal hemorrhage may be associated with the type 1 growth pattern in patients with AMD (group B). The subretinal (type 2) growth pattern likely occurs with single or few ingrowth sites, as occurs in a focal disease process, such as OHS.7 There may be a small sub-RPE component of the CNV in these cases, which does not coalesce in the sub-RPE space as in eyes with AMD.7 The subretinal (type 2) and combined growth patterns may be expected in surgically excised lesions from group H patients. These findings are reflected in the SST photograph reading center diagnoses (Table 4).
Correlations of histopathologic CNV growth pattern and angiographic appearance of CNV are limited because most eyes contained classic and occult CNV, regardless of the growth pattern (Table 4). However, classic and occult CNV were observed in more eyes that had lesions with a combined growth pattern (91%, 31/34) than in eyes with other growth patterns. Classic, occult, and combinations of classic and occult angiographic patterns may be associated with any of the histologic CNV growth patterns. Comparisons of 2-dimensional reconstruction histologic maps constructed for eyes that have had submacular surgery24,25 with preoperative and postoperative angiograms may provide some insight regarding growth pattern vs angiographic appearance. There have been 2 studies24,25 correlating 2-dimensional reconstructions of the histologic appearance and angiograms of eyes obtained post mortem after submacular surgery. In 1 study,24 a preoperative angiogram showed a classic pattern of CNV, and the surgically excised lesions showed a combined growth pattern. A postoperative angiogram showed pigment mottling of the RPE. The eye, obtained 5 months after surgery, showed repopulation of two thirds of the area of surgical excision with attenuated RPE and an area of sub-RPE CNV. The other study25 showed angiographically classic CNV and RPE pigment mottling before surgery. It was impossible to determine the growth pattern from the surgically excised lesion. A postsurgery angiogram showed an RPE defect. Examination of the postmortem eye 3 years after surgery showed an RPE defect and clinically undetected sub-RPE CNV with an associated retinal anastamosis.25 Studies like these,24,25 in which the appearance of the preoperative and postoperative angiographic features, surgically excised specimen, and postmortem eye can be compared, will provide insight into pathologic correlates of angiographic features.
Correspondence: Hans E. Grossniklaus, MD, L. F. Montgomery Ophthalmic Pathology Laboratory, BT428 Emory Eye Center, 1365 Clifton Rd NE, Atlanta, GA 30322 (firstname.lastname@example.org).
Submitted for Publication: April 5, 2004; final revision received September 1, 2004; accepted September 13, 2004.
Funding/Support: The SST is sponsored by the National Eye Institute, National Institutes of Health, US Department of Health and Human Services, Bethesda, Md, through cooperative agreements U10 EY12662 with Emory University, Atlanta, and U10 EY11547, EY11557, and EY11558 with The Johns Hopkins University, Baltimore.
Clinical centers with personnel who contributed data for this study are listed in alphabetical order by city. The number of specimens submitted to the SST pathology center is given in parentheses. Detailed information about investigators at each clinical center and at other resource centers has been published elsewhere (Submacular Surgery Trials Research Group. Clinical trial performance of community- vs university-based practices in the Submacular Surgery Trials [SST]: SST report No. 2. Arch Ophthalmol. 2004;122:857-863).
Emory University Eye Center, Atlanta (15); The Wilmer Eye Institute, Baltimore (14); Illinois Retina Associates, Chicago and Harvey (5); Cole Eye Institute, Cleveland, Ohio (4); Retina Associates of Cleveland (4); Ohio State University, Columbus (3); Texas Retina Associates, Dallas (5); Duke University Eye Center, Durham, NC (8); The University of Iowa Department of Ophthalmology, Iowa City (1); Southeastern Retina Associates, Knoxville, Tenn (22); Retina and Vitreous Associates of Kentucky, Lexington (12); Jules Stein Eye Institute, Los Angeles, Calif (7); Vitreoretinal Surgery PA, Minneapolis, Minn (9); McGee Eye Institute, Oklahoma City, Okla (8); Retinal Consultants of Arizona, Phoenix (25); Retina Vitreous Consultants, Pittsburgh, Pa (11); Oregon Health Sciences University, Portland (16); Associated Retinal Consultants, Royal Oak, Mich (7); Barnes Retina Institute, St Louis, Mo (10); West Coast Retina Medical Group Inc, San Francisco, Calif (6); and St Vincent Mercy Medical Center, Retina Vitreous Associates, Toledo, Ohio (10).
Emory University L.F. Montgomery Ophthalmic Pathology Laboratory
Principal Investigator: Hans E. Grossniklaus, MD. SST Coordinator: Pingbo Liu. Pathologist: W. Richard Green, MD. Epidemiologist: Päivi H. Miskala, PhD. Consultants: Neil M. Bressler, MD; Susan B. Bressler, MD; and Barbara S. Hawkins, PhD.
SST Executive Committee That Reviewed and Approved The Manuscript
Ex Officio Members: Neil M. Bressler, MD (chairperson); Eric B. Bass, MD, MPH; Susan B. Bressler, MD; Hans E. Grossniklaus, MD; Julia A. Haller, MD; Barbara S. Hawkins, PhD; Carol M. Mangione, MD, MSPH; Peggy R. Orr, MPH, COMT; Maryann Redford, DDS, MPH; Paul Sternberg, Jr, MD; and Matthew A. Thomas, MD. Other Members: Jayne M. Brown; Nancy M. Holekamp, MD; H. Richard McDonald, MD; and David J. Wilson, MD.
Financial Disclosure: None.