ASSOCIATION OF PREVASCULAR VITREOUS FISSURES AND CISTERNS WITH VITREOUS DEGENERATION AS ASSESSED BY SWEPT SOURCE OPTICAL COHERENCE TOMOGRAPHY CLAUDINE E. PANG, MD,*† KAREN B. SCHAAL, MD,*† MICHAEL ENGELBERT, MD, PHD*†‡ Purpose: To demonstrate the presence of prevascular vitreous fissures (PVF) and posterior vitreous cisterns in vivo and correlate with the degree of vitreous degeneration (VD). Methods: This was a cross-sectional study using Topcon Deep Range Imaging OCT-1 Atlantis 3D swept source optical coherence tomography for acquiring scans of posterior vitreous covering an 18 · 18-mm area in 104 eyes of 52 healthy volunteers without posterior vitreous detachment. Results: We observed that increasing age was associated with higher VD grades (P , 0.05). Prevascular vitreous fissures, characterized by areas of lower optical density overlying the retinal blood vessels, were identified in 93 (89%) eyes, and the presence of PVF correlated with lower VD grades (P , 0.05). Presence of cisterns correlated with higher VD grades (P , 0.05). All eyes with absence of PVF were found to have established cisterns. Prevascular vitreous fissures were connected with cisterns in 44 of the 71 (62%) eyes with cisterns, while the base of the cistern was directly above retinal blood vessels in 38 (54%) eyes, which suggests that the cisterns could be derived from PVF. Conclusion: Swept source optical coherence tomography imaging can identify PVF and cisterns occurring in the context of age-related VD, and PVF appeared to be possible precursors of cisterns. RETINA 35:1875–1882, 2015
T
and depth and hence appreciate the characteristics of the vitreous in an unprecedented way.2 Over the course of time, several investigators1,3 have made considerable contributions to the knowledge of the vitreous body. Eisner4,5 first suggested the presence of areas of optical transparency overlying the superficial retinal blood vessels, which extended into the vitreous body. Eisner termed them “prevascular vitreous fissures” (PVF) or prevascular “Lücken,” the English translation of which would be “gaps.” He suggested that those were the “result of decreased framework formation within those areas where the retina is covered by a blood vessel,”4 presumably because of a lower concentration of hyalocytes producing vitreous matrix. Worst6 used an ink injection technique in cadaver eyes to demonstrate a complex arrangement of liquefied spaces within the vitreous cavity and termed them cisterns. Fine7 confirmed the existence of these cisternal spaces in vivo with
he vitreous body occupies the largest volume in the eye, yet it arguably is the least understood compartment within it. This is mainly because it consists primarily of water and as such, many ex vivo histological studies have been unsuccessful in studying its details.1 With the advent of swept source optical coherence tomography (SS-OCT), it is possible to image the vitreous in vivo with greater resolution From the *Vitreous Retina Macula Consultants of New York, New York, New York; †LuEsther T. Mertz Retinal Research Center, Manhattan Eye, Ear and Throat Hospital, New York, New York; and ‡Department of Ophthalmology, New York University School of Medicine, New York, New York. Supported by LuEsther T. Mertz Retinal Research Center and The Macula Foundation, Inc, New York, NY. The funding organization had no role in the design or conduct of this research. None of the authors have any conflicting interests to disclose. Reprint requests: Michael Engelbert, MD, PhD, Vitreous Retina Macula Consultants of New York, 460 Park Avenue, Fifth Floor, New York, NY 10022; e-mail: [email protected]
1875
1876 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES
intraoperative triamcinolone staining. The origin of those cisterns, and their relationship to PVF, if any, is unknown. At the present time, there is a revival of interest in the vitreous anatomy, as greater understanding of vitreous structures and their age-related changes may be relevant for intravitreal drug delivery including the newest vitreolytic agents.8 Itakura et al9 first reported the finding of a posterior precortical vitreous pocket and the connecting channel to Cloquet’s canal using SS-OCT. Since then, several publications have used SS-OCT to examine the vitreous in health and disease.9–14 However, although the premacular bursa and its relationship to Cloquet’s canal have been the subject of relatively intense study,2,9,10 PVF and vitreous cisterns have not been described using this noninvasive imaging method. The aim of this study was to characterize PVF and vitreous cisternal spaces in vivo with SS-OCT imaging, which may be relevant to vitreoretinal interface diseases and vitreous pharmacodynamics.
Methods Institutional Review Board approval was obtained through Western IRB. Informed consent was obtained from all volunteers. This study complied with Health Insurance Portability and Accountability Act and adhered to the tenets of the Declaration of Helsinki. The posterior vitreous of healthy subjects aged from 18 to 65 years without posterior vitreous separation was imaged using the swept source Deep Range Imaging OCT-1 Atlantis 3D Optical Coherence Tomography (Topcon Medical Systems, Oakland, NJ). The study population consisted of volunteers recruited at Vitreous Retina Macula Consultants of New York from March through December 2013. Eyes with any degree of posterior vitreous separation on OCT, history of vitreoretinal pathology, or ocular/ systemic comorbidities known to affect the vitreous, such as diabetes or high myopia (defined as an axial length [AL] greater than 28 mm) were excluded from the study. Axial length was determined using partial coherence laser interferometry (Zeiss IOLMaster; Carl Zeiss AG, Oberkochen, Germany). To acquire the vitreous images, 5-line-cross scan patterns with a spacing of 1.5 mm between each line of the cross scan pattern were acquired as described previously.2 Briefly, a scan width of 12 mm was chosen, and 32 A-scans were averaged for each line of the 5-line-cross. The OCT focus was selected to be on the vitreous and contrast adjusted with the proprietary viewing software to visualize vitreous structures. Six 5-line-cross scans were acquired for each eye to
2015 VOLUME 35 NUMBER 9
create composite maps of the vitreous anatomy at the posterior pole over an approximately 18 · 18-mm area. All scan lines were examined to minimize the possibility of missing connections between the liquefied spaces of interest. All eyes were graded regarding the degree of vitreous degeneration (VD) using a grading scheme developed by the authors.2 Although the premacular bursa described by Worst1 and the posterior precortical vitreous pocket described by Itakura et al9 are likely the same structure defined differently, this study uses the term premacular bursa to be consistent with the authors’ previous work.2 Two graders (C.E.P. and K.B.S.) independently analyzed all SS-OCT scans, examining the characteristics and connections of visible liquefied vitreous spaces and categorizing them according to the following scheme: vitreous degeneration Grade 0 is characterized by a premacular bursa with no connection to neighboring spaces on any of the scans acquired. Grade 1 was defined as visible neighboring spaces in close proximity to the premacular bursa, often with speckled hyperreflectivity at the edges but no connections to the premacular bursa. Grade 2 was defined as more advanced VD with connections of these shallow spaces with the premacular bursa. Grade 3 was defined as a connection between the premacular bursa and a more central much larger lacuna, possibly from the coalescence of several cisternal spaces, indicating more advanced liquefaction (Figure 1). Any discrepancies between the two graders were adjudicated by the senior author. In all eyes, PVF were identified as wedge-shaped areas of qualitatively lower reflectivity on SS-OCT overlying the superficial retinal blood vessels within the vitreous cortex and extending into the vitreous body. The relationships of the various vitreous spaces to each other are illustrated in Figure 2. The premacular bursa and preoptic area of Martegiani were identified as optically empty spaces above the macula and optic disc, respectively. Ovoid or shallow optically empty spaces, which were often multiple and with or without irregular and hyperreflective edges, overlying the premacular bursa were taken to be fissure planes in the context of VD. Optically empty spaces with or without connections to the premacular bursa or lacunae that were in proximity to the retinal surface and peripheral to the premacular bursa were identified as cisterns. The characteristics of the cisternal spaces, such as location, size, shape, and their connections to the premacular bursa were recorded. The size of the cistern was taken to be the approximate width, which was calculated by multiplying the scan spacing of 1.5 mm by the number of scans where the cistern was present, and depth, which was measured using the
PREVASCULAR VITREOUS FISSURES AND CISTERNS PANG ET AL
1877
Fig. 2. The illustration in A and corresponding schematic vertical SSOCT scan in B through the area indicated by the dotted line in A define the various liquefied spaces of interest. The premacular bursa and preoptic area of Martegiani were identified as optically empty spaces above the macula and optic disc, respectively (green). Ovoid or shallow optically empty spaces, which were often multiple and with or without irregular and hyperreflective edges, overlying the premacular bursa (asterisk) were taken to be the central lacuna in the context of vitreous degeneration (yellow). Optically empty spaces with or without connections to the premacular bursa or lacuna that appeared adjacent to the aforementioned spaces and in the peripheral retina were identified as cisterns (blue). Prevascular vitreous fissures were identified as wedgeshaped areas of qualitatively lower reflectivity on SS-OCT overlying the superficial retinal blood vessels (red). Optically empty spaces with or without connections to the premacular bursa or lacunae that were in proximity to the retinal surface and peripheral to the premacular bursa were identified as cisterns (magenta).
Fig. 1. Swept source OCT images showing the grading of vitreous degeneration. A. Demonstrates Grade 0 characterized by the presence of a premacular bursa (asterisk) with no neighboring spaces in its vicinity. Formed vitreous can be seen between the premacular bursa and a larger lacuna (triangle). B. Demonstrates Grade 1 characterized by visible shallow neighboring spaces with speckled hyperreflectivity (arrows) at the edges in close proximity, but without connections to the premacular bursa (asterisk). C. Demonstrates Grade 2 characterized by more advanced VD with connections of these shallow spaces with the premacular bursa (asterisk). D. Demonstrates Grade 3 characterized by a connection between the premacular bursa (asterisk) and a more central larger lacuna.
caliper function provided by the SS-OCT software, after adjusting the B-scan scale to 1:1, to avoid false measurements due to an anteroposterior magnification of the B-scan. To exclude the possibility that connections between spaces identified with the above protocol were missed with 1.5-mm spaced horizontal and vertical line sampling, 0.25-mm spaced scans over the area of interest were obtained in cases where no connection could be identified. In eyes where PVF were identified, the presence or absence of connections with an adjacent cisternal space was determined. Statistical analysis was performed with GraphPad Prism software version 6.0 (La Jolla, CA) using Pearson chi-square test for nominal variables and one-way
analysis of variance for calculating the difference between means. Calculations were performed on variables including VD grade, age, AL, presence of PVF, presence of cisterns, presence of connections, and size of cisterns to elucidate the presence of correlation, accepting P , 0.05 as statistically significant. Results Vitreous Degeneration Correlated With Age A total of 52 subjects (18 males, 34 females) aged 21 to 60 (34 ± 9) years with ALs ranging from 20.6 mm to 27.6 mm (24.2 ± 1.4 mm; 4 patients had anisometropia .0.5 but less than 1 mm) were included in this study. All eyes were categorized according to the degree of VD, and the mean ages and mean ALs were tabulated. There were 21 of 104 (20%) eyes with VD Grade 0, 28 (27%) eyes with VD Grade 1, 31 (30%) eyes with VD Grade 2, and 24 (23%) eyes with VD Grade 3 (Table 1). Using one-way analysis of variance, there was a statistically significant difference between the mean ages of each grade of VD showing that a higher grade correlated with more advanced age (P , 0.05). The mean ALs were observed to increase with increasing grades of VD, although this was not statistically significant (P = 0.24).
1878 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES
2015 VOLUME 35 NUMBER 9
Table 1. Correlation of Vitreous Degeneration Grades With Age, Axial Length, PVF, and Cisterns Vitreous Degeneration Grade
Grade 0
Grade 1
Grade 2
Grade 3
Stat (P)
No. of eyes Mean age ± SD, years Mean AL ±SD, mm No. of eyes with any PVF (%) No. of eyes with superior and inferior PVF (%) No. of eyes with only superior PVF (%) No. of eyes with only inferior PVF (%) No. of eyes with no cisterns (%) No. of eyes with cisterns with no connections (%) No. of eyes with cisterns with connections (%) Mean no. of cisterns Mean width of cisterns ±SD, mm Mean depth of cisterns ±SD, mm
21 28 ± 4.4 23.9 ± 1.57 21 (100) 11 (52) 5 (24) 5 (24) 21 (100) 0 (0) 0 (0) 0 0 0
28 32 ± 6.8 23.9 ± 1.24 28 (100) 14 (50) 7 (25) 7 (25) 12 (43) 9 (32) 7 (25) 1 4.6 ± 2.3 704 ± 218
31 34 ± 7.7 24.3 ± 1.60 27 (87) 14 (45) 4 (13) 9 (19) 0 (0) 3 (10) 28 (90) 1 7.6 ± 3.6 778 ± 299
24 41 ± 10.1 24.6 ± 1.17 17 (71) 3 (13) 7 (29) 7 (29) 0 (0) 0 (0) 24 (100) 2 7.6 ± 3.6 848 ± 296
— ,0.05 ,0.24 ,0.05 — — — ,0.05 ,0.05 ,0.05 ,0.13 ,0.05 ,0.41
Prevascular Vitreous Fissures Correlated With Lower Vitreous Degeneration Grades Prevascular vitreous fissures, characterized by areas of lower optical density overlying superficial blood vessels (Figures 3 and 4), were identified in 93 of 104 (89%) eyes and were located in both or either superior or inferior quadrants (Table 1). Prevascular vitreous fissures were not seen above every single blood vessel, but when present, they were always located above a blood vessel. Prevascular vitreous fissures were present in all 21 eyes with Grade 0, all 28 eyes with Grade 1, 27 of 31 (87%) eyes with Grade 2, and 17 of 31 (71%) eyes with Grade 3. All 11 eyes without PVF had VD Grade 2 or 3 with established posterior vitreous cisterns. Of note, PVF were present in all of the 33 eyes without cisterns, which were either Grade 0 or 1. The presence of PVF correlated with lower grades of VD (P = 0.02). Cisterns Correlated With Higher Vitreous Degeneration Grades Of the 104 eyes, 71 (68%) eyes had one or several cisterns with or without connections to the premacular bursa or a more central lacuna (Figures 3 and 5). None
Fig. 3. Bar graph plotting the prevalence of PVF and cisterns against the grades of vitreous degeneration. The presence of PVF correlated inversely with higher grades of VD (P = 0.02), whereas the presence of cisterns correlated positively with increasing VD grades (P , 0.05).
of the 21 eyes with VD Grade 0 had visible cisterns. In the eyes with VD Grade 1, 12 eyes (43%) had no visible cisterns, 9 eyes (32%) had cisterns without visible connections between the cisternal space to the premacular bursa or central lacuna, and 7 eyes (25%) had cisterns with connections between the cisternal space and the premacular space or central lacuna. All 31 eyes with VD Grade 2 had cisterns, and all but 3 of these eyes had visible connections between the cisternal space and the premacular bursa. All 24 eyes with VD Grade 3 were observed to have cisterns with connections between the cisternal space and the premacular bursa (Table 1). The presence of cisterns and their connections to the premacular bursa correlated positively with increasing VD grades (P , 0.05). Prevascular Vitreous Fissures Were Related to Cisterns by Their Connections and Location In 20 eyes (17 eyes with VD Grade 0 and 3 eyes with VD Grade 1), PVF were observed in the absence of any cisterns. However, PVF and cisterns could be observed to occur concurrently. In 44 (62%) of the 71 eyes with cisterns, PVF were seen to be connected with the cisternal space. In 38 (54%) of the 71 eyes with cisterns, the base of the cistern was located directly above the blood vessels, the location of presumed antecedent PVF. In only 10 eyes, the cistern was neither seen directly overlying a blood vessel nor connected to a PVF. In these cases, the cisterns were seen deeper in the central vitreous and away from the vitreous cortex, hence could not be localized adjacent to any retinal blood vessel (Figures 6 and 7). A Greater Number and Larger Size of Cisterns Were Seen With Higher Vitreous Degeneration Grades Cisterns were observed to occur superiorly in all of the 71 eyes, of which 66 eyes had a single superior
PREVASCULAR VITREOUS FISSURES AND CISTERNS PANG ET AL
1879
Fig. 4. Swept source OCT images of 3 different subjects showing prevascular fissures (PVF) and their location above retinal blood vessels (black arrowheads). Infrared images in A, C, and E with green arrows indicating scan position corresponding to vertical SS-OCT scans in B, D, and F. A and B demonstrate PVF overlying 2 retinal blood vessels in the left eye of a 32-year-old woman with an axial length (AL) of 24.5 mm. C and D demonstrate PVF overlying 3 retinal blood vessels in the right eye of a 36year-old man with an AL of 24.1 mm. E and F demonstrate PVF overlying 2 retinal blood vessels in the left eye of a 36-year-old man with an AL of 25.5 mm. The black arrow points to a blood vessel with no PVF above it.
cistern and 5 eyes had 2 superior cisterns. A single cistern was found superotemporal to the fovea in 24 eyes, superior to the fovea in 38 eyes, and superonasal to the fovea in 4 eyes. In the 5 cases where 2 superior cisterns were present, they were located superotemporal and superonasal to the fovea. Nineteen of the 71 (26.8%) eyes had an inferior cistern in addition to the superior cistern(s). Only 1 eye had a total of 3 cisterns present, 2 superior and 1 inferior (Figure 8). All eyes with inferior cisterns were VD Grade 2 or 3. The shape of the cisterns that had connections to the premacular bursa in 59 eyes was found to assume a “bowl” configuration, while that of cisterns that had no connections in 12 eyes assumed a “teardrop” configuration (Figure 5). A greater number of cisterns was seen with higher VD grades, although this was not statistically significant (P = 0.13). The mean width of cisterns was seen to increase with increasing grades of VD (P , 0.05). The mean depth was also observed to increase with increasing grades of VD, although there was no statistical significance (P = 0.41) (Table 1). Discussion Eisner4 first proposed that PVF are areas of optical translucency in the vitreous overlying the normal inner
retinal surface. They had not been characterized before or confirmed since. We have found in our study that SS-OCT can identify prevascular areas of decreased reflectivity probably corresponding to what Eisner observed. We also found that these PVF are less prevalent with increasing vitreous syneresis and synchysis. It is possible that these areas of lower optical density in Eisner’s experiments and lower reflectivity on SSOCT in this study are gaps between the fibrils produced by the hyalocytes in the cortical vitreous, which extend into the vitreous cavity. Eisner4 demonstrates this by showing a lack of hyalocytes and fibrils overlying the retinal vessels and experimentally produced similar areas of lower optical density in the vitreous following photocoagulation of the retina and destruction of the overlying posterior hyaloid. Worst’s6 pioneering work characterized the vitreous cisternal anatomy using ink injection technique in postmortem eyes. He reported that the vitreous center consists of a core surrounded by a number of pouchlike extensions, which he termed cisterns and divided into 3 separate circles. These may or may not correspond to the liquefied spaces, most commonly called “lacunae,” which can be seen with the slit lamp or on ultrasound. In addition to these 2 anterior cisternal circles, a third posterior one was found to surround
1880 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES
2015 VOLUME 35 NUMBER 9
Fig. 5. Two subjects showing the relationship between premacular bursa (asterisks) and cisterns (white triangles). Infrared images in A and C with green arrows indicating scan position corresponding to vertical SS-OCT scans in B and D. A and B show a teardrop configured cistern with no visible connection to the premacular bursa in the right eye of a 36year-old man with an axial length (AL) of 24.1 mm. C and D show a bowl-configured cistern with visible connection to the premacular bursa in the left eye of a 51-year-old man with an AL of 24.2 mm.
the premacular bursa. As the SS-OCT has only an imaging depth of 2.6 mm, this study could only examine Worst’s third-order cisterns at the posterior pole. Worst6 proposed that 12 petal-like cisterns surrounded Fig. 6. Swept source OCT images of 4 different subjects showing various relationships between PVF overlying retinal blood vessels (black arrowheads), cisterns (white triangles), and premacular bursa (asterisks). Infrared images in A, C, E, and G with green arrows indicating scan position corresponding to vertical SS-OCT scans in B, D, F, and H. A and B demonstrate PVF occurring concurrently with a cistern in the left eye of a 27-year-old man with axial length (AL) of 23.1 mm. Note how the base of the cistern is directly above a retinal blood vessel. C and D demonstrate a cistern with its base directly overlying a retinal blood vessel in the left eye of a 36-year-old woman with an AL of 24.6 mm. No PVF was present in this case. E and F demonstrate PVF overlying 3 retinal blood vessels and an adjacent cistern with its base directly overlying a retinal blood vessel in the right eye of a 38year-old woman with an AL of 24.7 mm. Note how the PVF is connected to the cistern (doubleheaded arrow). G and H demonstrate PVF overlying 3 retinal blood vessels and a cistern that is located deeper into the vitreous, away from the vitreous cortex, in the left eye of a 36-year-old woman with an AL of 25.5 mm. In this case, no correlation of its base with the location of the retinal blood vessel could be made.
the premacular bursa and preoptic area of Martegiani. However, in our study, we did not observe all 12 cisterns in any of the eyes we studied. There were only 1 to 3 cisterns in any one eye. The majority of the
PREVASCULAR VITREOUS FISSURES AND CISTERNS PANG ET AL
1881
Fig. 7. Swept source OCT images showing the correlation of PVF overlying retinal blood vessels (black arrowheads) and cisterns (white triangles) with progression of vitreous degeneration (VD). A, B, C, and D are 4 different subjects representative of each grade and are shown with 3 SS-OCT vertical scans. A. Demonstrates the presence of PVF and no cisterns in the left eye of a 24-year-old man with an axial length (AL) of 23.6 mm and Grade 0 VD. The third scan to the right was taken with the patient’s eye in elevation to demonstrate a superior PVF. B. Demonstrates the presence of PVF and no cisterns in the left eye of a 27-year-old man with an AL of 25.1 mm and Grade 1 VD. C. Demonstrates the presence of PVF and a cistern in the right eye of a 34year-old woman with an AL of 23.6 mm and Grade 2 VD. Note how the adjacent PVF lying above the retinal blood vessels have connections to the cistern (double-headed arrow). D. Demonstrates the presence of PVF and a cistern in the right eye of a 38-year-old woman with an AL of 24.7 mm and Grade 3 VD. Note how the base of the cistern is directly overlying a retinal blood vessel, and the adjacent PVF have connections to the premacular bursa.
cisterns appeared in the superior quadrants, and all eyes with inferior cisterns had superior cisterns concurrently. We speculate that this could be due to the denser vitreous gravitating inferiorly; hence, the areas of liquefaction occur superiorly first. This study suggests that the cisterns observed by Worst may be agerelated liquefied lacunae anterior to the area that could be scanned with SS-OCT. Worst6 hypothesized that the cisterns were all interconnected through the center of the vitreous body. Fine7 reported one case in which triamcinolone acetonide was injected into the posterior vitreous and visualized the premacular bursa and surrounding cisterns, which appeared to be connected. However, this study demonstrates that the interconnecting channels
between cisternal spaces and premacular bursa are not always present. Because the presence of cisternal spaces and their connections to the premacular bursa correlate well with the grade of VD, we hypothesize that they occur in the context of VD. They are not present in young formed vitreous gel. Based on this hypothesis, we expected the prevalence and size of the cisterns to correlate with the degree of VD, and we did indeed find a correlation using our grading scheme. Worst’s observations of many more posterior cisterns5 could be explained by ex vivo postmortem decay artifact, or that the positive pressure of ink injection ballooned these cisterns, as they are areas of more liquefied vitreous. Fine’s observations7 may have a similar explanation given the turbulent flow
1882 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES
2015 VOLUME 35 NUMBER 9
In conclusion, we observed that the vitreous cisternal spaces are generally not present in young eyes and are therefore probably a result of more advanced VD in the context of age-related liquefied lacuna. Cisterns can occur singly and are not always connected to the premacular bursa. The interconnections are observed in the context of VD. In addition, PVF can be observed by SS-OCT imaging and appear to be the precursor of cisterns. SS-OCT analysis of the posterior vitreous anatomy may be helpful in the grading of VD and understanding of vitreoretinal interface diseases. Key words: vitreous, cistern, lacuna, prevascular vitreous fissure, premacular bursa, posterior precortical vitreous pocket, swept source optical coherence tomography. References
Fig. 8. Schematic diagram showing the distribution of cisterns. Sixtyfive percent of eyes had a single superior cistern. This cistern could be located superotemporal, superior, or superonasal to the fovea. Seven percent had 2 superior cisterns, 26% had 2 cisterns, one located superiorly and one inferiorly, and only 2% had 3 cisterns.
encountered during vitrectomy and the unroofing of real and potential spaces by the vitrectomy cutter. Because 1) PVF were found to be connected with cisterns in more than half of the eyes with cisterns, 2) PVF were less prevalent in eyes with established cisterns, and 3) the base of some cisterns was seen to be overlying blood vessels that may correspond to the location of presumed antecedent PVF, we hypothesize that PVF are the precursors to cisterns. In further support of this theory, we observed that all 33 eyes without cisterns were found to have PVF instead. A limitation of this study is its cross-sectional nature, which does not permit determination of cause and effect. Longitudinal data are not available given the recent advent of SS-OCT. A prospective study following the evolution of PVF and cisterns in individuals would be necessary to definitively demonstrate that PVF are precursors of cisterns. Although we use the most comprehensive scanning protocol used by any group examining the vitreous with SS-OCT to date, we cannot exclude that tenuous connections between PVF, cisterns, and other liquefied vitreous spaces were missed. The presence of PVF and cisterns and their relationship may be relevant for efficacy of intravitreal drugs because intravitreal fluorescent particles have been shown to have distributions relating to cisternal anatomy.15 It can be extrapolated that the PVF and cisterns may serve as vitreous pockets for accumulation of inflammatory mediators.
1. Jongebloed WL, Worst JG. The cisternal anatomy of the vitreous body. Doc Ophthalmol 1987;67:183–196. 2. Schaal KB, Pang CE, Pozzoni MC, Engelbert M. The premacular bursa’s shape revealed in vivo by swept source optical coherence tomography. Ophthalmology 2014;121:1020–1028. 3. Sebag J, Balazs EA. Human vitreous fibres and vitreoretinal disease. Trans Ophthalmol Soc U K 1985;104:123–128. 4. Eisner G. Clinical anatomy of the vitreous. In: Jacobiec FA, ed. Ocular Anatomy, Embryology, and Teratology. Philadelphia, PA: Harper & Row Publishers; 1982:391–424. 5. Eisner G. Biomicroscopy of the Peripheral Fundus. New York, NY: Springer-Verlag; 1979:20–21; 106–107. 6. Worst JG. Cisternal systems of the fully developed vitreous body in the young adult. Trans Ophthalmol Soc U K 1977;97:550–554. 7. Fine HF, Spaide RF. Visualization of the posterior precortical vitreous pocket in vivo with triamcinolone. Arch Ophthalmol 2006;124:1663. 8. Duker JS, Moshfegi AA. Ocriplasmin: a medical or surgical therapy. Retina 2013;33:2001–2002. 9. Itakura H, Kishi S, Li D, Akiyama H. Observation of posterior precortical vitreous pocket using swept-source optical coherence tomography. Invest Ophthalmol Vis Sci 2013;54:3102–3107. 10. Li D, Kishi S, Itakura H, et al. Posterior precortical vitreous pockets and connecting channels in children on swept-source optical coherence tomography. Invest Ophthalmol Vis Sci 2014;55:2412–2416. 11. Itakura H, Kishi S, Li D, et al. Vitreous changes in high myopia observed by swept-source optical coherence tomography. Invest Ophthalmol Vis Sci 2014;55:1447–1452. 12. Stanga PE, Sala-Puigdollers A, Caputo S, et al. In vivo imaging of cortical vitreous using 1050-nm swept-source deep range imaging optical coherence tomography. Am J Ophthalmol 2014;157:397–404. 13. Liu JJ, Witkin AJ, Adhi M, et al. Enhanced vitreous imaging in healthy eyes using swept-source optical coherence tomography. PLoS One 2014;9:e102950. 14. Alasil T, Adhi M, Liu JJ, et al. Spectral domain and swept-source OCT imaging of asteroid hyalosis: a case report. Ophthalmic Surg Lasers Imaging Retina 2014;45:459–461. 15. Laude A, Tan LE, Wilson CG, et al. Intravitreal therapy for neovascular age-related macular degeneration and interindividual variations in vitreous pharmacokinetics. Prog Retin Eye Res 2010;29:466–475.
T
and depth and hence appreciate the characteristics of the vitreous in an unprecedented way.2 Over the course of time, several investigators1,3 have made considerable contributions to the knowledge of the vitreous body. Eisner4,5 first suggested the presence of areas of optical transparency overlying the superficial retinal blood vessels, which extended into the vitreous body. Eisner termed them “prevascular vitreous fissures” (PVF) or prevascular “Lücken,” the English translation of which would be “gaps.” He suggested that those were the “result of decreased framework formation within those areas where the retina is covered by a blood vessel,”4 presumably because of a lower concentration of hyalocytes producing vitreous matrix. Worst6 used an ink injection technique in cadaver eyes to demonstrate a complex arrangement of liquefied spaces within the vitreous cavity and termed them cisterns. Fine7 confirmed the existence of these cisternal spaces in vivo with
he vitreous body occupies the largest volume in the eye, yet it arguably is the least understood compartment within it. This is mainly because it consists primarily of water and as such, many ex vivo histological studies have been unsuccessful in studying its details.1 With the advent of swept source optical coherence tomography (SS-OCT), it is possible to image the vitreous in vivo with greater resolution From the *Vitreous Retina Macula Consultants of New York, New York, New York; †LuEsther T. Mertz Retinal Research Center, Manhattan Eye, Ear and Throat Hospital, New York, New York; and ‡Department of Ophthalmology, New York University School of Medicine, New York, New York. Supported by LuEsther T. Mertz Retinal Research Center and The Macula Foundation, Inc, New York, NY. The funding organization had no role in the design or conduct of this research. None of the authors have any conflicting interests to disclose. Reprint requests: Michael Engelbert, MD, PhD, Vitreous Retina Macula Consultants of New York, 460 Park Avenue, Fifth Floor, New York, NY 10022; e-mail: [email protected]
1875
1876 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES
intraoperative triamcinolone staining. The origin of those cisterns, and their relationship to PVF, if any, is unknown. At the present time, there is a revival of interest in the vitreous anatomy, as greater understanding of vitreous structures and their age-related changes may be relevant for intravitreal drug delivery including the newest vitreolytic agents.8 Itakura et al9 first reported the finding of a posterior precortical vitreous pocket and the connecting channel to Cloquet’s canal using SS-OCT. Since then, several publications have used SS-OCT to examine the vitreous in health and disease.9–14 However, although the premacular bursa and its relationship to Cloquet’s canal have been the subject of relatively intense study,2,9,10 PVF and vitreous cisterns have not been described using this noninvasive imaging method. The aim of this study was to characterize PVF and vitreous cisternal spaces in vivo with SS-OCT imaging, which may be relevant to vitreoretinal interface diseases and vitreous pharmacodynamics.
Methods Institutional Review Board approval was obtained through Western IRB. Informed consent was obtained from all volunteers. This study complied with Health Insurance Portability and Accountability Act and adhered to the tenets of the Declaration of Helsinki. The posterior vitreous of healthy subjects aged from 18 to 65 years without posterior vitreous separation was imaged using the swept source Deep Range Imaging OCT-1 Atlantis 3D Optical Coherence Tomography (Topcon Medical Systems, Oakland, NJ). The study population consisted of volunteers recruited at Vitreous Retina Macula Consultants of New York from March through December 2013. Eyes with any degree of posterior vitreous separation on OCT, history of vitreoretinal pathology, or ocular/ systemic comorbidities known to affect the vitreous, such as diabetes or high myopia (defined as an axial length [AL] greater than 28 mm) were excluded from the study. Axial length was determined using partial coherence laser interferometry (Zeiss IOLMaster; Carl Zeiss AG, Oberkochen, Germany). To acquire the vitreous images, 5-line-cross scan patterns with a spacing of 1.5 mm between each line of the cross scan pattern were acquired as described previously.2 Briefly, a scan width of 12 mm was chosen, and 32 A-scans were averaged for each line of the 5-line-cross. The OCT focus was selected to be on the vitreous and contrast adjusted with the proprietary viewing software to visualize vitreous structures. Six 5-line-cross scans were acquired for each eye to
2015 VOLUME 35 NUMBER 9
create composite maps of the vitreous anatomy at the posterior pole over an approximately 18 · 18-mm area. All scan lines were examined to minimize the possibility of missing connections between the liquefied spaces of interest. All eyes were graded regarding the degree of vitreous degeneration (VD) using a grading scheme developed by the authors.2 Although the premacular bursa described by Worst1 and the posterior precortical vitreous pocket described by Itakura et al9 are likely the same structure defined differently, this study uses the term premacular bursa to be consistent with the authors’ previous work.2 Two graders (C.E.P. and K.B.S.) independently analyzed all SS-OCT scans, examining the characteristics and connections of visible liquefied vitreous spaces and categorizing them according to the following scheme: vitreous degeneration Grade 0 is characterized by a premacular bursa with no connection to neighboring spaces on any of the scans acquired. Grade 1 was defined as visible neighboring spaces in close proximity to the premacular bursa, often with speckled hyperreflectivity at the edges but no connections to the premacular bursa. Grade 2 was defined as more advanced VD with connections of these shallow spaces with the premacular bursa. Grade 3 was defined as a connection between the premacular bursa and a more central much larger lacuna, possibly from the coalescence of several cisternal spaces, indicating more advanced liquefaction (Figure 1). Any discrepancies between the two graders were adjudicated by the senior author. In all eyes, PVF were identified as wedge-shaped areas of qualitatively lower reflectivity on SS-OCT overlying the superficial retinal blood vessels within the vitreous cortex and extending into the vitreous body. The relationships of the various vitreous spaces to each other are illustrated in Figure 2. The premacular bursa and preoptic area of Martegiani were identified as optically empty spaces above the macula and optic disc, respectively. Ovoid or shallow optically empty spaces, which were often multiple and with or without irregular and hyperreflective edges, overlying the premacular bursa were taken to be fissure planes in the context of VD. Optically empty spaces with or without connections to the premacular bursa or lacunae that were in proximity to the retinal surface and peripheral to the premacular bursa were identified as cisterns. The characteristics of the cisternal spaces, such as location, size, shape, and their connections to the premacular bursa were recorded. The size of the cistern was taken to be the approximate width, which was calculated by multiplying the scan spacing of 1.5 mm by the number of scans where the cistern was present, and depth, which was measured using the
PREVASCULAR VITREOUS FISSURES AND CISTERNS PANG ET AL
1877
Fig. 2. The illustration in A and corresponding schematic vertical SSOCT scan in B through the area indicated by the dotted line in A define the various liquefied spaces of interest. The premacular bursa and preoptic area of Martegiani were identified as optically empty spaces above the macula and optic disc, respectively (green). Ovoid or shallow optically empty spaces, which were often multiple and with or without irregular and hyperreflective edges, overlying the premacular bursa (asterisk) were taken to be the central lacuna in the context of vitreous degeneration (yellow). Optically empty spaces with or without connections to the premacular bursa or lacuna that appeared adjacent to the aforementioned spaces and in the peripheral retina were identified as cisterns (blue). Prevascular vitreous fissures were identified as wedgeshaped areas of qualitatively lower reflectivity on SS-OCT overlying the superficial retinal blood vessels (red). Optically empty spaces with or without connections to the premacular bursa or lacunae that were in proximity to the retinal surface and peripheral to the premacular bursa were identified as cisterns (magenta).
Fig. 1. Swept source OCT images showing the grading of vitreous degeneration. A. Demonstrates Grade 0 characterized by the presence of a premacular bursa (asterisk) with no neighboring spaces in its vicinity. Formed vitreous can be seen between the premacular bursa and a larger lacuna (triangle). B. Demonstrates Grade 1 characterized by visible shallow neighboring spaces with speckled hyperreflectivity (arrows) at the edges in close proximity, but without connections to the premacular bursa (asterisk). C. Demonstrates Grade 2 characterized by more advanced VD with connections of these shallow spaces with the premacular bursa (asterisk). D. Demonstrates Grade 3 characterized by a connection between the premacular bursa (asterisk) and a more central larger lacuna.
caliper function provided by the SS-OCT software, after adjusting the B-scan scale to 1:1, to avoid false measurements due to an anteroposterior magnification of the B-scan. To exclude the possibility that connections between spaces identified with the above protocol were missed with 1.5-mm spaced horizontal and vertical line sampling, 0.25-mm spaced scans over the area of interest were obtained in cases where no connection could be identified. In eyes where PVF were identified, the presence or absence of connections with an adjacent cisternal space was determined. Statistical analysis was performed with GraphPad Prism software version 6.0 (La Jolla, CA) using Pearson chi-square test for nominal variables and one-way
analysis of variance for calculating the difference between means. Calculations were performed on variables including VD grade, age, AL, presence of PVF, presence of cisterns, presence of connections, and size of cisterns to elucidate the presence of correlation, accepting P , 0.05 as statistically significant. Results Vitreous Degeneration Correlated With Age A total of 52 subjects (18 males, 34 females) aged 21 to 60 (34 ± 9) years with ALs ranging from 20.6 mm to 27.6 mm (24.2 ± 1.4 mm; 4 patients had anisometropia .0.5 but less than 1 mm) were included in this study. All eyes were categorized according to the degree of VD, and the mean ages and mean ALs were tabulated. There were 21 of 104 (20%) eyes with VD Grade 0, 28 (27%) eyes with VD Grade 1, 31 (30%) eyes with VD Grade 2, and 24 (23%) eyes with VD Grade 3 (Table 1). Using one-way analysis of variance, there was a statistically significant difference between the mean ages of each grade of VD showing that a higher grade correlated with more advanced age (P , 0.05). The mean ALs were observed to increase with increasing grades of VD, although this was not statistically significant (P = 0.24).
1878 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES
2015 VOLUME 35 NUMBER 9
Table 1. Correlation of Vitreous Degeneration Grades With Age, Axial Length, PVF, and Cisterns Vitreous Degeneration Grade
Grade 0
Grade 1
Grade 2
Grade 3
Stat (P)
No. of eyes Mean age ± SD, years Mean AL ±SD, mm No. of eyes with any PVF (%) No. of eyes with superior and inferior PVF (%) No. of eyes with only superior PVF (%) No. of eyes with only inferior PVF (%) No. of eyes with no cisterns (%) No. of eyes with cisterns with no connections (%) No. of eyes with cisterns with connections (%) Mean no. of cisterns Mean width of cisterns ±SD, mm Mean depth of cisterns ±SD, mm
21 28 ± 4.4 23.9 ± 1.57 21 (100) 11 (52) 5 (24) 5 (24) 21 (100) 0 (0) 0 (0) 0 0 0
28 32 ± 6.8 23.9 ± 1.24 28 (100) 14 (50) 7 (25) 7 (25) 12 (43) 9 (32) 7 (25) 1 4.6 ± 2.3 704 ± 218
31 34 ± 7.7 24.3 ± 1.60 27 (87) 14 (45) 4 (13) 9 (19) 0 (0) 3 (10) 28 (90) 1 7.6 ± 3.6 778 ± 299
24 41 ± 10.1 24.6 ± 1.17 17 (71) 3 (13) 7 (29) 7 (29) 0 (0) 0 (0) 24 (100) 2 7.6 ± 3.6 848 ± 296
— ,0.05 ,0.24 ,0.05 — — — ,0.05 ,0.05 ,0.05 ,0.13 ,0.05 ,0.41
Prevascular Vitreous Fissures Correlated With Lower Vitreous Degeneration Grades Prevascular vitreous fissures, characterized by areas of lower optical density overlying superficial blood vessels (Figures 3 and 4), were identified in 93 of 104 (89%) eyes and were located in both or either superior or inferior quadrants (Table 1). Prevascular vitreous fissures were not seen above every single blood vessel, but when present, they were always located above a blood vessel. Prevascular vitreous fissures were present in all 21 eyes with Grade 0, all 28 eyes with Grade 1, 27 of 31 (87%) eyes with Grade 2, and 17 of 31 (71%) eyes with Grade 3. All 11 eyes without PVF had VD Grade 2 or 3 with established posterior vitreous cisterns. Of note, PVF were present in all of the 33 eyes without cisterns, which were either Grade 0 or 1. The presence of PVF correlated with lower grades of VD (P = 0.02). Cisterns Correlated With Higher Vitreous Degeneration Grades Of the 104 eyes, 71 (68%) eyes had one or several cisterns with or without connections to the premacular bursa or a more central lacuna (Figures 3 and 5). None
Fig. 3. Bar graph plotting the prevalence of PVF and cisterns against the grades of vitreous degeneration. The presence of PVF correlated inversely with higher grades of VD (P = 0.02), whereas the presence of cisterns correlated positively with increasing VD grades (P , 0.05).
of the 21 eyes with VD Grade 0 had visible cisterns. In the eyes with VD Grade 1, 12 eyes (43%) had no visible cisterns, 9 eyes (32%) had cisterns without visible connections between the cisternal space to the premacular bursa or central lacuna, and 7 eyes (25%) had cisterns with connections between the cisternal space and the premacular space or central lacuna. All 31 eyes with VD Grade 2 had cisterns, and all but 3 of these eyes had visible connections between the cisternal space and the premacular bursa. All 24 eyes with VD Grade 3 were observed to have cisterns with connections between the cisternal space and the premacular bursa (Table 1). The presence of cisterns and their connections to the premacular bursa correlated positively with increasing VD grades (P , 0.05). Prevascular Vitreous Fissures Were Related to Cisterns by Their Connections and Location In 20 eyes (17 eyes with VD Grade 0 and 3 eyes with VD Grade 1), PVF were observed in the absence of any cisterns. However, PVF and cisterns could be observed to occur concurrently. In 44 (62%) of the 71 eyes with cisterns, PVF were seen to be connected with the cisternal space. In 38 (54%) of the 71 eyes with cisterns, the base of the cistern was located directly above the blood vessels, the location of presumed antecedent PVF. In only 10 eyes, the cistern was neither seen directly overlying a blood vessel nor connected to a PVF. In these cases, the cisterns were seen deeper in the central vitreous and away from the vitreous cortex, hence could not be localized adjacent to any retinal blood vessel (Figures 6 and 7). A Greater Number and Larger Size of Cisterns Were Seen With Higher Vitreous Degeneration Grades Cisterns were observed to occur superiorly in all of the 71 eyes, of which 66 eyes had a single superior
PREVASCULAR VITREOUS FISSURES AND CISTERNS PANG ET AL
1879
Fig. 4. Swept source OCT images of 3 different subjects showing prevascular fissures (PVF) and their location above retinal blood vessels (black arrowheads). Infrared images in A, C, and E with green arrows indicating scan position corresponding to vertical SS-OCT scans in B, D, and F. A and B demonstrate PVF overlying 2 retinal blood vessels in the left eye of a 32-year-old woman with an axial length (AL) of 24.5 mm. C and D demonstrate PVF overlying 3 retinal blood vessels in the right eye of a 36year-old man with an AL of 24.1 mm. E and F demonstrate PVF overlying 2 retinal blood vessels in the left eye of a 36-year-old man with an AL of 25.5 mm. The black arrow points to a blood vessel with no PVF above it.
cistern and 5 eyes had 2 superior cisterns. A single cistern was found superotemporal to the fovea in 24 eyes, superior to the fovea in 38 eyes, and superonasal to the fovea in 4 eyes. In the 5 cases where 2 superior cisterns were present, they were located superotemporal and superonasal to the fovea. Nineteen of the 71 (26.8%) eyes had an inferior cistern in addition to the superior cistern(s). Only 1 eye had a total of 3 cisterns present, 2 superior and 1 inferior (Figure 8). All eyes with inferior cisterns were VD Grade 2 or 3. The shape of the cisterns that had connections to the premacular bursa in 59 eyes was found to assume a “bowl” configuration, while that of cisterns that had no connections in 12 eyes assumed a “teardrop” configuration (Figure 5). A greater number of cisterns was seen with higher VD grades, although this was not statistically significant (P = 0.13). The mean width of cisterns was seen to increase with increasing grades of VD (P , 0.05). The mean depth was also observed to increase with increasing grades of VD, although there was no statistical significance (P = 0.41) (Table 1). Discussion Eisner4 first proposed that PVF are areas of optical translucency in the vitreous overlying the normal inner
retinal surface. They had not been characterized before or confirmed since. We have found in our study that SS-OCT can identify prevascular areas of decreased reflectivity probably corresponding to what Eisner observed. We also found that these PVF are less prevalent with increasing vitreous syneresis and synchysis. It is possible that these areas of lower optical density in Eisner’s experiments and lower reflectivity on SSOCT in this study are gaps between the fibrils produced by the hyalocytes in the cortical vitreous, which extend into the vitreous cavity. Eisner4 demonstrates this by showing a lack of hyalocytes and fibrils overlying the retinal vessels and experimentally produced similar areas of lower optical density in the vitreous following photocoagulation of the retina and destruction of the overlying posterior hyaloid. Worst’s6 pioneering work characterized the vitreous cisternal anatomy using ink injection technique in postmortem eyes. He reported that the vitreous center consists of a core surrounded by a number of pouchlike extensions, which he termed cisterns and divided into 3 separate circles. These may or may not correspond to the liquefied spaces, most commonly called “lacunae,” which can be seen with the slit lamp or on ultrasound. In addition to these 2 anterior cisternal circles, a third posterior one was found to surround
1880 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES
2015 VOLUME 35 NUMBER 9
Fig. 5. Two subjects showing the relationship between premacular bursa (asterisks) and cisterns (white triangles). Infrared images in A and C with green arrows indicating scan position corresponding to vertical SS-OCT scans in B and D. A and B show a teardrop configured cistern with no visible connection to the premacular bursa in the right eye of a 36year-old man with an axial length (AL) of 24.1 mm. C and D show a bowl-configured cistern with visible connection to the premacular bursa in the left eye of a 51-year-old man with an AL of 24.2 mm.
the premacular bursa. As the SS-OCT has only an imaging depth of 2.6 mm, this study could only examine Worst’s third-order cisterns at the posterior pole. Worst6 proposed that 12 petal-like cisterns surrounded Fig. 6. Swept source OCT images of 4 different subjects showing various relationships between PVF overlying retinal blood vessels (black arrowheads), cisterns (white triangles), and premacular bursa (asterisks). Infrared images in A, C, E, and G with green arrows indicating scan position corresponding to vertical SS-OCT scans in B, D, F, and H. A and B demonstrate PVF occurring concurrently with a cistern in the left eye of a 27-year-old man with axial length (AL) of 23.1 mm. Note how the base of the cistern is directly above a retinal blood vessel. C and D demonstrate a cistern with its base directly overlying a retinal blood vessel in the left eye of a 36-year-old woman with an AL of 24.6 mm. No PVF was present in this case. E and F demonstrate PVF overlying 3 retinal blood vessels and an adjacent cistern with its base directly overlying a retinal blood vessel in the right eye of a 38year-old woman with an AL of 24.7 mm. Note how the PVF is connected to the cistern (doubleheaded arrow). G and H demonstrate PVF overlying 3 retinal blood vessels and a cistern that is located deeper into the vitreous, away from the vitreous cortex, in the left eye of a 36-year-old woman with an AL of 25.5 mm. In this case, no correlation of its base with the location of the retinal blood vessel could be made.
the premacular bursa and preoptic area of Martegiani. However, in our study, we did not observe all 12 cisterns in any of the eyes we studied. There were only 1 to 3 cisterns in any one eye. The majority of the
PREVASCULAR VITREOUS FISSURES AND CISTERNS PANG ET AL
1881
Fig. 7. Swept source OCT images showing the correlation of PVF overlying retinal blood vessels (black arrowheads) and cisterns (white triangles) with progression of vitreous degeneration (VD). A, B, C, and D are 4 different subjects representative of each grade and are shown with 3 SS-OCT vertical scans. A. Demonstrates the presence of PVF and no cisterns in the left eye of a 24-year-old man with an axial length (AL) of 23.6 mm and Grade 0 VD. The third scan to the right was taken with the patient’s eye in elevation to demonstrate a superior PVF. B. Demonstrates the presence of PVF and no cisterns in the left eye of a 27-year-old man with an AL of 25.1 mm and Grade 1 VD. C. Demonstrates the presence of PVF and a cistern in the right eye of a 34year-old woman with an AL of 23.6 mm and Grade 2 VD. Note how the adjacent PVF lying above the retinal blood vessels have connections to the cistern (double-headed arrow). D. Demonstrates the presence of PVF and a cistern in the right eye of a 38-year-old woman with an AL of 24.7 mm and Grade 3 VD. Note how the base of the cistern is directly overlying a retinal blood vessel, and the adjacent PVF have connections to the premacular bursa.
cisterns appeared in the superior quadrants, and all eyes with inferior cisterns had superior cisterns concurrently. We speculate that this could be due to the denser vitreous gravitating inferiorly; hence, the areas of liquefaction occur superiorly first. This study suggests that the cisterns observed by Worst may be agerelated liquefied lacunae anterior to the area that could be scanned with SS-OCT. Worst6 hypothesized that the cisterns were all interconnected through the center of the vitreous body. Fine7 reported one case in which triamcinolone acetonide was injected into the posterior vitreous and visualized the premacular bursa and surrounding cisterns, which appeared to be connected. However, this study demonstrates that the interconnecting channels
between cisternal spaces and premacular bursa are not always present. Because the presence of cisternal spaces and their connections to the premacular bursa correlate well with the grade of VD, we hypothesize that they occur in the context of VD. They are not present in young formed vitreous gel. Based on this hypothesis, we expected the prevalence and size of the cisterns to correlate with the degree of VD, and we did indeed find a correlation using our grading scheme. Worst’s observations of many more posterior cisterns5 could be explained by ex vivo postmortem decay artifact, or that the positive pressure of ink injection ballooned these cisterns, as they are areas of more liquefied vitreous. Fine’s observations7 may have a similar explanation given the turbulent flow
1882 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES
2015 VOLUME 35 NUMBER 9
In conclusion, we observed that the vitreous cisternal spaces are generally not present in young eyes and are therefore probably a result of more advanced VD in the context of age-related liquefied lacuna. Cisterns can occur singly and are not always connected to the premacular bursa. The interconnections are observed in the context of VD. In addition, PVF can be observed by SS-OCT imaging and appear to be the precursor of cisterns. SS-OCT analysis of the posterior vitreous anatomy may be helpful in the grading of VD and understanding of vitreoretinal interface diseases. Key words: vitreous, cistern, lacuna, prevascular vitreous fissure, premacular bursa, posterior precortical vitreous pocket, swept source optical coherence tomography. References
Fig. 8. Schematic diagram showing the distribution of cisterns. Sixtyfive percent of eyes had a single superior cistern. This cistern could be located superotemporal, superior, or superonasal to the fovea. Seven percent had 2 superior cisterns, 26% had 2 cisterns, one located superiorly and one inferiorly, and only 2% had 3 cisterns.
encountered during vitrectomy and the unroofing of real and potential spaces by the vitrectomy cutter. Because 1) PVF were found to be connected with cisterns in more than half of the eyes with cisterns, 2) PVF were less prevalent in eyes with established cisterns, and 3) the base of some cisterns was seen to be overlying blood vessels that may correspond to the location of presumed antecedent PVF, we hypothesize that PVF are the precursors to cisterns. In further support of this theory, we observed that all 33 eyes without cisterns were found to have PVF instead. A limitation of this study is its cross-sectional nature, which does not permit determination of cause and effect. Longitudinal data are not available given the recent advent of SS-OCT. A prospective study following the evolution of PVF and cisterns in individuals would be necessary to definitively demonstrate that PVF are precursors of cisterns. Although we use the most comprehensive scanning protocol used by any group examining the vitreous with SS-OCT to date, we cannot exclude that tenuous connections between PVF, cisterns, and other liquefied vitreous spaces were missed. The presence of PVF and cisterns and their relationship may be relevant for efficacy of intravitreal drugs because intravitreal fluorescent particles have been shown to have distributions relating to cisternal anatomy.15 It can be extrapolated that the PVF and cisterns may serve as vitreous pockets for accumulation of inflammatory mediators.
1. Jongebloed WL, Worst JG. The cisternal anatomy of the vitreous body. Doc Ophthalmol 1987;67:183–196. 2. Schaal KB, Pang CE, Pozzoni MC, Engelbert M. The premacular bursa’s shape revealed in vivo by swept source optical coherence tomography. Ophthalmology 2014;121:1020–1028. 3. Sebag J, Balazs EA. Human vitreous fibres and vitreoretinal disease. Trans Ophthalmol Soc U K 1985;104:123–128. 4. Eisner G. Clinical anatomy of the vitreous. In: Jacobiec FA, ed. Ocular Anatomy, Embryology, and Teratology. Philadelphia, PA: Harper & Row Publishers; 1982:391–424. 5. Eisner G. Biomicroscopy of the Peripheral Fundus. New York, NY: Springer-Verlag; 1979:20–21; 106–107. 6. Worst JG. Cisternal systems of the fully developed vitreous body in the young adult. Trans Ophthalmol Soc U K 1977;97:550–554. 7. Fine HF, Spaide RF. Visualization of the posterior precortical vitreous pocket in vivo with triamcinolone. Arch Ophthalmol 2006;124:1663. 8. Duker JS, Moshfegi AA. Ocriplasmin: a medical or surgical therapy. Retina 2013;33:2001–2002. 9. Itakura H, Kishi S, Li D, Akiyama H. Observation of posterior precortical vitreous pocket using swept-source optical coherence tomography. Invest Ophthalmol Vis Sci 2013;54:3102–3107. 10. Li D, Kishi S, Itakura H, et al. Posterior precortical vitreous pockets and connecting channels in children on swept-source optical coherence tomography. Invest Ophthalmol Vis Sci 2014;55:2412–2416. 11. Itakura H, Kishi S, Li D, et al. Vitreous changes in high myopia observed by swept-source optical coherence tomography. Invest Ophthalmol Vis Sci 2014;55:1447–1452. 12. Stanga PE, Sala-Puigdollers A, Caputo S, et al. In vivo imaging of cortical vitreous using 1050-nm swept-source deep range imaging optical coherence tomography. Am J Ophthalmol 2014;157:397–404. 13. Liu JJ, Witkin AJ, Adhi M, et al. Enhanced vitreous imaging in healthy eyes using swept-source optical coherence tomography. PLoS One 2014;9:e102950. 14. Alasil T, Adhi M, Liu JJ, et al. Spectral domain and swept-source OCT imaging of asteroid hyalosis: a case report. Ophthalmic Surg Lasers Imaging Retina 2014;45:459–461. 15. Laude A, Tan LE, Wilson CG, et al. Intravitreal therapy for neovascular age-related macular degeneration and interindividual variations in vitreous pharmacokinetics. Prog Retin Eye Res 2010;29:466–475.
