Content uploaded by Jed A Lusthaus
Author content
All content in this area was uploaded by Jed A Lusthaus on Sep 05, 2020
Content may be subject to copyright.
Downloaded from https://journals.lww.com/glaucomajournal by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD3Ypodx1mzGi35KRSdy/ke8wX301pMtsuRKzjiqbJleSSl2aq2yP4GcQ== on 09/05/2020
Downloadedfromhttps://journals.lww.com/glaucomajournal by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD3Ypodx1mzGi35KRSdy/ke8wX301pMtsuRKzjiqbJleSSl2aq2yP4GcQ== on 09/05/2020
The Effects of Trabecular Bypass Surgery on Conventional
Aqueous Outflow, Visualized by Hemoglobin Video Imaging
Jed A. Lusthaus, MBBS, MPH, FRANZCO,*†
Paul A.R. Meyer, MD, FRCP,‡§ Tasneem Z. Khatib, MD, MBBCh,∥¶
and Keith R. Martin, MA, DM, FRCOphth, FRANZCO∥¶#**††
Precis: Hemoglobin Video Imaging (HVI) provides a noninvasive
method to quantify aqueous outflow (AO) perioperatively. Tra-
becular bypass surgery (TBS) is able to improve, and in some cases
re-establish, conventional AO.
Purpose: The purpose of this study was to use HVI to illustrate and
quantify effects of TBS on AO through the episcleral venous system.
Design: This is a prospective observational cohort study.
Participants: Patients were recruited from Sydney Eye Hospital,
Australia. The study included 29 eyes from 25 patients, 15 with
glaucoma and 14 normal controls. TBS (iStent Inject) was per-
formed on 14 glaucomatous eyes (9 combined phacoemulsification/
TBS and 5 standalone TBS). Cataract surgery alone was performed
on the remaining eye from the glaucoma group and 2 eyes from the
control group.
Methods: We used HVI, a novel clinic-based tool, to visualize and
quantify AO perioperatively during routine follow-up to 6 months.
Angiographic blood flow patterns were observed within prominent
aqueous veins on the nasal and temporal ocular surface. Aqueous
column cross-section area (AqCA) was compared before and after
surgery.
Main Outcome Measures: AqCA, number of aqueous veins, intra-
ocular pressure (IOP) before and after surgery, and number of IOP-
lowering medications.
Results: Patients with glaucoma had reduced AqCA compared with
normal controls (P=0.00001). TBS increased AqCA in 13 eyes at
1 month (n =14; P<0.002), suggesting improved AO. This effect
was maintained at 6 months in 7 eyes (n =9, P≤0.05). All patients
with unrecordable AO before surgery (n =3; 2 standalone TBS, 1
combined cataract/TBS) established measurable flow after TBS.
IOP and/or medication burden became reduced in every patient
undergoing TBS. Cataract surgery alone (n =3) increased AqCA in
nasal and temporal vessels at 4 weeks after surgery.
Conclusions: HVI provides a safe method for detecting and mon-
itoring AO perioperatively in an outpatient setting. Improvement of
AO into the episcleral venous system is expected after TBS and can
be visualized with HVI. TBS is able to improve, and in some cases
re-establish, conventional AO. Cataract surgery may augment this.
Some aqueous veins were first seen after TBS and these patients had
unstable postoperative IOP control, which possibly suggests reor-
ganization of aqueous homeostatic mechanisms. HVI may confirm
adequacy of surgery during short-term follow-up, but further work
is required to assess the potential of HVI to predict surgical out-
comes and assist with personalized treatment decisions.
Key Words: glaucoma, aqueous outflow, trabecular bypass surgery
(J Glaucoma 2020;29:656–665)
The incorporation of minimally invasive glaucoma sur-
gery (MIGS) devices into the glaucoma management
algorithm has been challenging, in part due to an incomplete
understanding of aqueous flow dynamics within the epis-
cleral venous system (EVS).
Whereas in previous years the most widely used surgical
options, trabeculectomy or tube shunt surgery, bypassed the
conventional drainage system, clinicians must now decide
whether to enhance or bypass the eye’s natural drainage system
and which tools to use. These decisions are presently made in
the absence of a quantitative assessment of the episcleral venous
system’s capacity to accept additional aqueous, and this may
contribute to variable intraocular pressure (IOP) results between
patients.1–3It is our hypothesis that visualization of aqueous
flow within the EVS during the perioperative period may pro-
vide information to assist with surgical decision-making.
The anatomic pathway of aqueous drainage from the
eye is well documented,4–9but our physiological under-
standing remains limited. Noninvasive assessment in vivo
has been challenging and further work in this area is
required to identify the significance of changes within the
outflow system in relation to open-angle glaucoma. In this
study we used Hemoglobin Video Imaging (HVI) to study
aqueous outflow (AO) patterns within the EVS before and
after trabecular bypass surgery (TBS) with iStent Inject
(Glaukos Corporation, USA) and cataract surgery. Success
rates (unmedicated IOP reduction ≥20%) with iStent Inject
are ∼76% to 78%.10,11 Our aim is to characterize the peri-
operative changes in aqueous drainage into the episcleral
veins, following interventions that facilitate outflow into
Schlemm canal. We expect this information will enhance
our existing knowledge of the pathophysiology of glaucoma
and the consequences of surgical interventions, leading to
improved surgical outcomes.
DOI: 10.1097/IJG.0000000000001561
Received for publication September 1, 2019; accepted May 9, 2020.
From the *Glaucoma Unit, Sydney Eye Hospital; †Discipline of Oph-
thalmology, The University of Sydney, Sydney, NSW; #Department
of Ophthalmology and Surgery, University of Melbourne; **Centre
for Eye Research Australia, The Royal Victorian Eye and Ear
Hospital, Melbourne, Vic., Australia; Departments of ‡Engineering;
§Medicine; ∥John van Geest Centre for Brain Repair; ††Wellcome
Trust—MRC Cambridge Stem Cell Institute, University of Cam-
bridge; and ¶Eye Department, Cambridge University Hospitals
NHS Foundation Trust, Cambridge, UK.
Supported by: (1) Hemoglobin Video Imaging facilities funded by Sydney
Eye Hospital Foundation, Carl Zeiss Meditec, and Glaukos Corpo-
ration. (2) iStent Inject devices for standalone cases were donated by
Glaukos Corporation. (3) A core support grant from the Wellcome
Trust and MRC to the Wellcome Trust—Medical Research Council
Cambridge Stem Cell Institute.
Disclosure: The authors declare no conflict of interest.
Reprints: Jed A. Lusthaus, MBBS, MPH, FRANZCO, Eyehaus 73-109
Belmore Road, Randwick 2031, NSW, Australia (e-mail: jed.
lusthaus@gmail.com).
Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.
ORIGINAL STUDY
656
|
www.glaucomajournal.com J Glaucoma Volume 29, Number 8, August 2020
Copyright r2020 Wolters Kluwer Health, Inc. All rights reserved.
METHODS
This study was undertaken in accordance with the
Declaration of Helsinki and approved by the Human Research
Ethics Committee, Prince of Wales Hospital, South Eastern
Sydney Local Health District, Sydney, Australia (LNR 16/224).
Written informed consent was obtained from all participants.
PARTICIPANTS
HVI was performed on 29 eyes, including 15 with
glaucoma and 14 normal controls (Table 1). Fourteen
consecutive glaucoma patients went on to have TBS. One
glaucoma patient and 2 normal controls underwent cataract
surgery with intraocular lens insertion only. All controls
were phakic and had IOP <21 mm Hg. Patients undergoing
TBS had open-angle glaucoma of varying aetiologies. All
patients continued IOP-lowering treatment until the morn-
ing of surgery. In addition to the 29 study eyes, HVI was
performed on one patient with ocular hypertension (OHT)
who was subsequently commenced on brimonidine tartrate
0.15% in both eyes.
Imaging Protocol
The HVI setup has been described previously and images
were obtained using the same methods.12,13 In summary, a
monochromatic Prosilica GC1380H camera mounted on a
Zeiss SL130 slit lamp with a bandpass filter transmitting
wavelengths from 540 to 580 nm was used to capture the real-
time image series. No drops or dyes were required before HVI.
The patient sat upright at the slit lamp in a darkened room and
the episcleral venous system was mapped by imaging the entire
limbus of each eye. One-minute videos (1800 frames) of the
nasal and temporal conjunctival and episcleral microcirculations
were recorded. Screening to identify aqueous veins was per-
formed at a nominal magnification of ×12. Areas of AO were
identified and examined more closely using ×20 magnification,
at which each pixel covers 4 μm
2
.
Participants underwent preoperative HVI within 3 months
of their surgery date. HVI was repeated at each regular follow-
up visit during the postoperative period (1 wk, 4 wk, 3 mo, and
6 mo). HVI was also performed 1 day after surgery where
possible (n =4). The control group underwent HVI on a single
occasion. The values obtained from these patients formed
comparative data for preoperative glaucomatous patients. The
OHT patient was imaged before, and 6 weeks after, com-
mencement of brimonidine.
In some cases, identification of an aqueous-carrying
vein was only possible retrospectively, once flow had been
established. This may represent recruitment of a normal
vein by aqueous, or reperfusion of a pre-existing aqueous
vein after TBS. In such cases there were occasions when
the vein had not been the main object of attention in the
preoperative angiogram and preoperative images were
poorly focused; however, all were adequate for study.
Surgical Protocol
Peribulbar or sub-Tenon block was used to achieve anes-
thesia. A temporal corneal approach was used in all surgical
cases. Nine cases underwent phacoemulsification and insertion
of posterior chamber intraocular lens before TBS. In the
remaining 5 cases, TBS was an isolated procedure (standalone).
1.4% sodium hyaluronate maintained the anterior chamber and
a Volk Transcend Vold Gonio intraoperative gonioscopy lens
was used for angle visualization. All patients had TBS with 2
iStents injected into Schlemm canal within the nasal quadrant,
between 1 and 2 clock hours apart. Stent positioning was not
based on preoperative aqueous vein patterns, because this study
was undertaken to demonstrate real-world aqueous flow char-
acteristics associated with TBS. Further studies are planned to
address targeted stent insertion.
In all cases, blood was seen to reflux into the anterior
chamber during stent insertion. Further 1.4% sodium hyaluro-
nate was injected to tamponade blood reflux if visualization for
second stent insertion required improvement. Viscoelastic was
removed and the anterior chamber was pressurized with bal-
anced salt solution to reduce the risk of hyphaema during the
early postoperative period. Prophylactic intracameral cefazolin
0.5 mg/0.1 mL was injected at the completion of surgery.
All IOP-lowering treatment was stopped in every study
eye on the day of surgery. The decision whether to recom-
mence IOP-lowering treatment during the postoperative
period was made on an individual basis, depending on the
severity of glaucoma and the IOP level. All patients were
treated with guttae chloramphenicol 4 times a day for 1
week and a weaning course of guttae dexamethasone 0.1%
(Novartis) for 4 weeks. Gonioscopy was performed at every
postoperative visit to confirm positioning of each stent.
Image Analysis
As blood-filled episcleral veins join those containing
aqueous, the stream of aqueous migrates towards the center of
the vessel, while blood remains at its periphery: the paraxial
zone of low pixel density defines the aqueous stream (Fig. 1C).
In HVI, outside the aqueous stream, pixel density is
proportionate to the depth of red cells in the blood column.
Transepts of aqueous veins show increasing density from the
periphery until the low-density central aqueous column and
we have shown that the aqueous column diameter is accu-
rately defined by the separation between the 2 points of
maximum pixel density.13 Images of aqueous veins were
processed in accordance with this previously described
technique, the aqueous column cross-section area (AqCA)
TABLE 1. Preoperative Characteristics
Glaucoma
(N =15)
Normal Controls
(N =14)
Age (y) 68.2 ± 18.0 56.0 ± 19.5
Diagnosis
POAG 7 No glaucoma
XFG 5
Angle recession 2
Uveitis 1
Sex
Male 9 4
Female 6 10
Preoperative mean IOP
(mm Hg)
23.2 ± 10.3 13.6 ± 3.3
No. medications mean 3.3 ± 1.3 0
AqCA mean (μm
2
) 107 ± 181 910 ± 475
Visual field mean (dB) −5.79 ± 5.65 Not performed
Intervention
Combined 9
Standalone TBS 5
Cataract surgery 1 2
Combined medications count as 2 medications.
Oral medications count as 1 medication.
AqCA indicates aqueous column cross-sectional area; Combined, pha-
coemulsification, intraocular lens insertion and iStent Inject; IOP, intraocular
pressure; POAG, primary open-angle glaucoma; Standalone TBS, iStent
Inject only; TBS, trabecular bypass surgery; XFG, exfoliative glaucoma.
J Glaucoma Volume 29, Number 8, August 2020 Aqueous Outflow Visualization After TBS
Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved. www.glaucomajournal.com
|
657
Copyright r2020 Wolters Kluwer Health, Inc. All rights reserved.
being measured to reflect aqueous flow.13 We have already
shown this to correlate inversely with IOP immediately after
selective laser trabeculoplasty.13
Sometimes aqueous stratifies to one side of the episcleral
vessel, giving the appearance of separated layers of aqueous and
blood, a phenomenon previously described by Ascher.14 Despite
this, HVI can detect the thin stream of erythrocytes that sepa-
rates aqueous from the vessel wall, and this permits measure-
ment of AqCA (Fig. 2). Aqueous veins may be devoid of blood
as they emerge from Schlemm canal; however, all our meas-
urements were made after erythrocytes had entered the vessel.
The number and location of aqueous veins were tallied
for each eye and compared between groups. When multiple
smaller aqueous veins drained into one larger episcleral vein,
this was recorded as a single aqueous vein.
Image J software was used to generate the transept of
pixel density (ie, depth of red blood cells) across aqueous
veins. The average of 3 measurements was used to represent
aqueous column diameter, which was converted to AqCA.
Measurements were taken before surgery and at each post-
operative follow-up visit. If AqCA was unmeasurable then flow
was characterized as visible or unrecordable. This distinction
was made to denote a difference between cases where aqueous
was visualized (Fig. 1B), but flow appeared slow (AqCA =3,
derived from nominal aqueous column diameter of 1), com-
pared with vessels where there was neither blanching nor a
defined aqueous column (Fig. 1A) (AqCA =0).
AqCA measurements were recorded from the most
prominent aqueous vein (single vein with the greatest
AqCA). AqCA was correlated with IOP and medication
FIGURE 1. Stages of aqueous vein revitalization. A, Unrecordable
preoperative aqueous outflow (black arrows). B, Flow re-
establishes with blanching seen at week 1. C, Laminar flow at
week 4 with linear transept representing site of aqueous column
cross-section area measurement. Figure 1 can be viewed in color
online at www.glaucomajournal.com.
FIGURE 2. Stratification of aqueous and blood flow. A thin stream
of erythrocytes (black arrow) enables measurement of aqueous
column diameter. Reprinted image still taken from supplementary
video, Khatib et al13 under the terms of the CC BY license (https://
creativecommons.org/licenses/by/4.0/).
FIGURE 3. Comparison of aqueous column cross-section area
(AqCA) between eyes with and without glaucoma. AqCA was
significantly reduced in glaucomatous eyes (P<0.0001). Black
line represents median (control =676 μm
2
, glaucoma =3μm
2
).
Lusthaus et al J Glaucoma Volume 29, Number 8, August 2020
658
|
www.glaucomajournal.com Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.
Copyright r2020 Wolters Kluwer Health, Inc. All rights reserved.
reduction for intervention eyes. Statistical analysis between
preoperative and postoperative AqCA and IOP was
undertaken using a 2-tailed, Mann-Whitney test, Wilcoxon
signed-rank test and Spearman rank order correlation
coefficient (Microsoft Excel version 14.6.7).
RESULTS
Aqueous column diameter measurements were repeat-
able for each eye. Fluctuations in flow occurred due to the
pulsatile and dynamic nature of AO. However, variations at
a single timepoint were limited to <40% of the measured
column diameter, which is consistent with previous work.13
Distribution of Aqueous Veins
There was no significant difference in the number of
aqueous veins visualized in the 2 groups (P=0.30). Glau-
comatous eyes were seen to have between 0 and 6; eyes
without glaucoma had between 2 and 5.
Nasal fields contained approximately twice as many
aqueous veins as did the temporal fields; however, at least 1
temporal aqueous vein was identified in 8 glaucomatous
patients before and after surgery, and in 11 controls.
The vein with the largest aqueous column diameter was
located within 2 clock hours of the nasal meridian in every
glaucomatous and control eye.
Reduction of AqCA in Glaucomatous Eyes
Before surgery, AqCA was significantly reduced in
glaucomatous eyes compared with normal controls
(P<0.0001) (Fig. 3). AqCA was inversely correlated with
preoperative IOP (N =29; r
s
=−0.6; P<0.001) and number
of medications (N =29; r
s
=−0.8; P<00001).
Correlations between AqCA and postoperative IOP
did not reach significance and are difficult to interpret due to
recommencement of IOP-lowering treatment in some
patients.
Effectiveness of TBS
In the glaucoma group, of the 14 patients that under-
went TBS, at the 6 months postoperative follow-up period
(n =9), 6 patients achieved ≥20% reduction in IOP without
medication. Median preoperative IOP was 20.5 mm Hg on 3
medications, including 3 patients requiring oral acetazola-
mide. Postoperative IOP (Fig. 4) measured at 1 day was
13 mm Hg (N =14), 1 week 19 mm Hg (N =14), 4 weeks
19.5 mm Hg (N =14), 3 months 13 mm Hg (N =10), and
6 months 14 mm Hg (N =9). Significant reductions in IOP
and number of medications were seen at 3 (N =10; P<0.05)
and 6 months (N =9; P<0.05).
It was necessary to recommence IOP-lowering
medication during the postoperative period in 4 cases: 3
following standalone TBS and 1 after combined phacoe-
mulsification/TBS. Despite measurable AqCA before
surgery, the latter was the only patient who did not dem-
onstrate an improvement in AqCA at any point during the
study. Micropulse cyclodiode laser was required after
6 months of follow-up, which enabled cessation of oral
acetazolamide. This was the only patient who required an
additional IOP-lowering procedure postoperatively.
The aqueous column was measurable at the end of each
patient’s observation period, with the exception of one case.
This patient had refused traditional glaucoma drainage
surgery and before TBS had IOP of 30 mm Hg on 4 IOP-
lowering agents (including oral acetazolamide) with no
visualized AO. Despite establishment of AO with isolated
TBS, he required recommencement of brimonidine and
fixed-combination brinzolamide/timolol after 1 month. He
stopped all treatment after 5 months and, at his 6-month
review, his IOP measured 45 mm Hg and AO could not be
visualized. Recommencement of topical treatment con-
trolled his IOP (14 mm Hg 1 mo later) and AO was re-
established.
No intraoperative complications occurred in any
patient. All stents appeared well-positioned at each follow-
up, and with no visible obstruction. Comparison between
combined phacoemulsification/TBS and standalone TBS
was limited in this study due to small sample size. Detailed
comparison is planned for a future study.
Re-establishment of AO
Preoperative AO was unrecordable in 3 patients. All epis-
cleral veins appeared full of blood and IOP was uncontrolled on
FIGURE 4. Median postoperative intraocular pressure (IOP) reduction at each timepoint, where the number of patients taking IOP-
lowering medications is in brackets, and error bars represent SE. Figure 4 can be viewed in color online at www.glaucomajournal.com.
J Glaucoma Volume 29, Number 8, August 2020 Aqueous Outflow Visualization After TBS
Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved. www.glaucomajournal.com
|
659
Copyright r2020 Wolters Kluwer Health, Inc. All rights reserved.
maximal tolerated medical therapy, including oral acetazolamide
in 2 patients. The decision to avoid bleb-forming surgeries in
these cases was based on patient preference, despite patients being
given advice to undergo traditional drainage surgery. All 3
patients developed measurable laminar flow at 1 month following
TBS, indicating a return of aqueous to blood-filled aqueous veins
(Figs. 1, 5, 6).
Patterns of Aqueous Flow After TBS
After TBS, visible AO always remained most prom-
inent in the nasal quadrant; however, AqCA of the largest
temporal vein increased significantly in 5 patients (all after
combined phacoemulsification/TBS), 4 weeks after surgery
(P<0.03). The remaining patients did not demonstrate
temporal AqCA improvement. Standalone TBS was not
associated with improved temporal AqCA in any case.
Mean nasal AqCA was similar between combined
(N =9) and standalone cases (N =5) preoperatively (126 and
94 μm
2
;P=0.42) and at 1 month (349 and 404 μm
2
;P=0.69),
respectively. Direct comparison between the groups beyond
1 month was not possible due to small sample size in the
standalone TBS group.
HVIwasperformedin3casesonthefirst postoperative
day. It was not possible in the remainder of the cases due to
subconjunctival hemorrhage or patient preference. In all 3 cases
there was dilation and tortuosity of episcleral vasculature, which
is a characteristic feature of inflammation.15 AO was not
detectable, but all 3 cases had IOP ≤14 mm Hg and laminar
flow recovered within 1 week (Fig. 7).
Laminar flow was also confirmed 1 week after TBS in
all patients who had visible preoperative AO anywhere in
the nasal quadrant. In some patients, a second episcleral
vessel, devoid of aqueous before surgery, demonstrated
laminar flow at 1 week (Fig. 8).
The recovery of AO in the 3 patients with no preoperative
evidence of aqueous veins was particularly interesting. One case
recovered laminar aqueous flow by 1 week postoperatively, but
2 passed through a phase in which episcleral vessels became
blanched (week 1 postoperative), and had developed laminar
aqueous flow by week 4 (Figs. 1, 6).
Flow within the conjunctival and episcleral vasculature
altered in every patient following TBS. In 1 case, a pre-
viously invisible or closed aqueous vein demonstrated lam-
inar flow 1 week after surgery (Fig. 9). There was additional
improvement in flow through 2 other aqueous veins that had
been evident preoperatively. The patient had IOP of
16 mm Hg on 3 IOP-lowering agents before TBS, and this
fell to 11 mm Hg, where it remained without any treatment
throughout the postoperative period.
Time Course of Change
AqCA increased significantly during the study period
(Fig. 10) with improvement at 4 weeks (N =14; P=0.002),
FIGURE 5. Re-establishment of aqueous outflow in an eye with-
out visible preoperative flow (stars to assist with orientation). Flow
absence seen with Hemoglobin Video Imaging throughout the
circumference of the eye. A, No flow seen before trabecular
bypass surgery. Image extracted from ×12 magnification survey of
the circumference of the eye. B, Flow re-established by week 1
(black arrows) recorded with ×20 magnification.
FIGURE 6. Blanching of a large episcleral vein due to resumption
of aqueous flow following trabecular bypass surgery. A, Episcleral
vein engorged with blood preoperatively. B, Episcleral vessel
almost completely disappears due to aqueous fill and dilution or
displacement of red blood cells (black arrows identify vessel).
Lusthaus et al J Glaucoma Volume 29, Number 8, August 2020
660
|
www.glaucomajournal.com Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.
Copyright r2020 Wolters Kluwer Health, Inc. All rights reserved.
3 months (N =10; P<0.05), and 6 months (N =9; P<0.05).
AqCA did not increase suddenly after TBS, but improved
gradually over months (Fig. 11). The time-course of this
varied between cases.
Patterns of Aqueous Flow After Cataract Surgery
and Brimonidine
An increase of AqCA in nasal and temporal vessels was
seen in all 3 cases (1 with glaucoma, 2 without) who had
isolated cataract surgery. In the glaucomatous patient,
AqCA improved from visible only (AqCA =3μm
2
) in each
vessel to 547 μm
2
in the nasal aqueous vein and 1168 μm
2
in
the temporal aqueous vein after 4 weeks. IOP was
16 mm Hg on 3 IOP-lowering agents before surgery, and
22 mm Hg unmedicated at 4 weeks after surgery. Similarly,
both patients without glaucoma who had cataract surgery
demonstrated an improvement in AqCA after 4 weeks;
patient 1 improved from 1427 μm
2
in both vessels to 1712
μm
2
in the nasal vessel and 2188 μm
2
in the temporal vessel,
although in patient 2 the nasal vessel improved from 1010 to
1202 μm
2
, and the temporal vessel from 324 to 392 μm
2
.
Longer-term follow-up and a larger cohort are planned to
compare AO in patients having cataract surgery or TBS.
In a single patient with OHT, the effect of brimonidine
commencement on AqCA of the largest aqueous vein was
examined in both eyes. IOP reduced from 23 mm Hg right
eye and 22 mm Hg left eye before treatment to 17 mm Hg in
both eyes after 6 weeks of brimonidine. AqCA increased
>2.5 times in both vessels studied; from 232 to 634 μm
2
in
the right eye and 324 to 835 μm
2
in the left eye.
DISCUSSION
We used HVI, a noninvasive outpatient technique, to
visualize aqueous drainage within the EVS before and after
TBS. Angiographic AO patterns have been shown to improve
following TBS,6but real-time analysis of AO in physiological
conditions has only recently been introduced.12,13 Our study
AB
CD
F
E
FIGURE 7. Acute changes in episcleral blood and aqueous flow following trabecular bypass surgery (TBS) in 2 separate patients. Sites of
aqueous column cross-section area measurement marked with linear transept. A and B, Preoperative scant aqueous outflow. C and D,
Dilation and filling of episcleral vein seen the day after TBS. E and F, Laminar flow established at 1 week review. Figure 7 can be viewed in
color online at www.glaucomajournal.com.
J Glaucoma Volume 29, Number 8, August 2020 Aqueous Outflow Visualization After TBS
Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved. www.glaucomajournal.com
|
661
Copyright r2020 Wolters Kluwer Health, Inc. All rights reserved.
suggests AO improves in eyes with established preoperative flow,
andisrestoredineyeswithoutrecordableflow preoperatively.
This study compared aqueous flow patterns in normal
and glaucomatous eyes, and then observed a cohort of
glaucomatous patients for up to 6 months after TBS. A very
broad range of glaucoma severity was included. The sample
size was modest, but our results were consistent. For the
purpose of this study, we used the cross-section area of the
aqueous column to quantify AO. This approach is here
validated by the significant difference in AqCA between
control and glaucomatous eyes.
We were unable to include a medication washout period
for any patient, and so the possibility remains that some changes
depicted in the EVS may have been influenced by cessation or
recommencement of topical IOP-lowering therapies. Cessation
of brimonidine at the time of surgery could relax episcleral
vasoconstriction and assist AO, and it is conceivable that this
explains reperfusion of the previously invisible aqueous vein in
Figure 9. Nevertheless, both eyes from our OHT patient showed
an increase in AqCA, 6 weeks after brimonidine commence-
ment, suggesting an improvement in AO. Johnstone et al16 also
demonstrated an acute increase in aqueous discharge into the
episcleral venous system 2 hours after instillation of brimonidine
in 8 normal subjects. The longer-term effect of its use has not
been reported to our knowledge. Angiographic responses to
different topical treatments have not yet been studied method-
ically. No patient was using local AO-promoting therapies such
as latanoprostene bunod or Rho-kinase inhibitors (these are not
yet available in Australia).
Cataract surgery appeared to improve aqueous drainage
into nasal and temporal episcleral veins, whereas standalone TBS
was associated with only nasal AqCA improvement. This
FIGURE 8. Recovery of aqueous outflow (AO) within 1 week of
trabecular bypass surgery. Linear marker indicates point of
aqueous column cross-section area measurement. A, Preoperative
angiogram showing an episcleral vessel with unrecordable AO,
however the patient had visible AO within another vessel in the
nasal quadrant. B, AO re-established 1 week after trabecular
bypass surgery. C, Further improvement of AO 1 month after TBS.
Figure 8 can be viewed in color online at www.glaucomajournal.
com.
FIGURE 9. Reperfusion of aqueous vein (black arrows) seen 1
week after trabecular bypass surgery (stars to assist with ori-
entation). Brimonidine had been taken for 3 years before tra-
becular bypass surgery and was stopped on the day of surgery.
A, Preoperative angiogram. B, Postoperative angiogram.
Lusthaus et al J Glaucoma Volume 29, Number 8, August 2020
662
|
www.glaucomajournal.com Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.
Copyright r2020 Wolters Kluwer Health, Inc. All rights reserved.
suggests that cataract surgery may augment AO when combined
with TBS; AqCA of 5 combined phacoemulsification/TBS cases
improved in nasal and temporal aqueous veins. We acknowledge
that further studies are required to separate this effect from iStent
insertionincombinedcases.
A number of studies on cadaveric eyes have clearly dem-
onstrated the anatomy of the conventional AO pathway, dating
back to Ascher and Ashton early work.4,5,14 More recently,
intraoperative aqueous angiography, using fluorescein and
indocyanine green at the time of TBS,6has shown detailed
outflow pathways predominantly in the nasal quadrant, con-
sistent with previous in vivo7–9,17,18 and ex vivo19,20 distal out-
flow work. Our study confirms segmental AO with a higher
incidence of aqueous veins in the nasal episcleral vasculature in
patients with and without glaucoma.
Our studies of aqueous veins in 15 glaucomatous partic-
ipants showed widely variable preoperative AO patterns, ranging
from no visible flow to healthy laminar flow. This was sig-
nificantly different from the 14 control subjects, who all
demonstrated good AO. AqCA was significantly reduced in
glaucomatous eyes compared with controls, an observation that
will be studied in more detail in future work. AO is dynamic and
likely to undergo diurnal variation, which also needs to be
considered.
Most of the TBS cohort required large numbers of
IOP-lowering agents before surgery, despite which some
patients still had very high preoperative IOPs. TBS is
commonly advocated in mild to moderate glaucoma.21,22
We have demonstrated that TBS can re-establish AO in a
range of glaucoma patients, but this study cohort is not a
true representation of real-world case selection. Three
patients from the TBS cohort were initially advised to have
glaucoma drainage surgery. None of these patients would
agree to bleb-forming surgery, but were willing to undergo
TBS. This provided a unique opportunity to assess the IOP
and HVI responses in eyes with poor IOP control.
Proof of aqueous vein reperfusion has been elegantly
demonstrated by Huang et al,6but in nonphysiological
states. Our study showed re-establishment of aqueous flow
after TBS in all 3 patients who had lacked aqueous veins
preoperatively. TBS also increased AqCA regardless of the
state of preoperative aqueous drainage. However, aqueous
flow did not suddenly increase as intraoperative angiog-
raphy might predict.6Aqueous vein congestion, seen in all 4
patients imaged on the day after surgery, was presumed to
be related to postoperative inflammation. IOP was low in
these cases, but aqueous was not visible. By 1 week, there
was recovery of AO and normalization of the episcleral
vasculature in most patients. Gradual improvement in AO
occurred over weeks to months, and in some cases fluctua-
tion in flow occurred.
The speed and quality of aqueous vein revitalization
following TBS differed between patients and likely
depended on the functionality of Schlemm canal and the
EVS. In some patients, laminar aqueous flow only became
apparent after 4 weeks, possibly indicating gradual reper-
fusion of collector channels and/or Schlemm canal, or may
reflect attempts by the eye to establish a new homeostatic
balance of aqueous flow. Generalized blanching of vessels,
seen in some patients, may reflect low aqueous flow velocity,
with failure of discrete aqueous and erythrocyte columns to
form; alternatively, aqueous channels that have not yet
joined the episcleral venous circulation. Aqueous flow
accelerates during the postoperative period and flow lami-
nae develop within episcleral vessels. This variability in flow
may explain IOP variations in the early postoperative
period.
The postoperative variation in IOP results and flow
dynamics suggests that multiple mechanisms may contribute
to IOP dysregulation, other than TM dysfunction. These
may include disorganized control of episcleral venous pres-
sure and aqueous production, as well as effects from the use
FIGURE 10. Gradual improvement in aqueous column cross-sectional area (AqCA) following trabecular bypass surgery after 4 weeks
(N =14; P=0.002), 3 months (N =10; P<0.05), and 6 months (N =9; P<0.05). Black lines represent median AqCA.
J Glaucoma Volume 29, Number 8, August 2020 Aqueous Outflow Visualization After TBS
Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved. www.glaucomajournal.com
|
663
Copyright r2020 Wolters Kluwer Health, Inc. All rights reserved.
of postoperative topical steroid and the cessation of IOP-
lowering drops. It is clear that the eye may take time to
recover from the shock of a sudden change in fluid dynamics
caused by TBS. We hypothesize that the EVS may partic-
ipate in the detection and control of IOP.
The patency and resilience of Schlemm canal almost
certainly plays a significant role in aqueous drainage.23–25
Evolving techniques to image the trabecular meshwork and
Schlemm canal, such as phase-sensitive OCT,21 may
complement HVI.
Many interventions to control IOP affect AO and,
using HVI, we have been able to characterize some of these
changes. In this study of patients undergoing TBS, we have
observed an increase in aqueous column diameter, and the
restoration of laminar aqueous flow where it had not been
visible preoperatively. Furthermore, this is the first report of
a noninvasive examination of aqueous veins in which
evolution of surgically induced changes in AO can be
monitored and compared with characteristics recorded
preoperatively. In this way, HVI can be applied to other
techniques that manipulate AO; and subsequent findings
may contribute to a greater physiological understanding of
IOP homeostasis. Further work is required to identify spe-
cific HVI features that are predictive of surgical success or
failure. HVI is also a promising technique for comparing
different MIGS devices and for investigating the targeting of
TBS according to preoperative AO.
Using HVI, we have demonstrated impaired episcleral
aqueous flow in glaucoma, and the manner of its restoration by
TBS. This clinical technique enables the physiology of AO to be
studied and quantified to establish normal drainage, define
pathology and monitor therapeutic interventions.
REFERENCES
1. Shah M, Campos-Möller X, Werner L, et al. Midterm failure of
combined phacoemulsification with trabecular microbypass
stenting: clinicopathological analysis. J Cataract Refract Surg.
2018;44:654–657.
2. Samuelson TW, Katz LJ, Wells JM, et al. Randomized
evaluation of the trabecular micro-bypass stent with phacoe-
mulsification in patients with glaucoma and cataract. Oph-
thalmology. 2011;118:459–467.
3. Craven ER, Katz LJ, Wells JM, et al. Cataract surgery with
trabecular micro-bypass stent implantation in patients with
FIGURE 11. Improvement of aqueous outflow following trabecular bypass surgery as evidenced by gradual aqueous column cross-
section area (AqCA) increase during the first 3 months after surgery. Linear transept is the site where AqCA measurement was taken.
0=Preoperative laminar flow. AqCA increases 1 week after trabecular bypass surgery and this is maintained after 4 and 12 weeks. Figure
11 can be viewed in color online at www.glaucomajournal.com.
Lusthaus et al J Glaucoma Volume 29, Number 8, August 2020
664
|
www.glaucomajournal.com Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.
Copyright r2020 Wolters Kluwer Health, Inc. All rights reserved.
mild-to-moderate open-angle glaucoma and cataract: two-year
follow-up. J Cataract Refract Surg. 2012;38:1339–1345.
4. Ascher KW. Aqueous veins and their significance for patho-
genesis of glaucoma. Arch Ophthalmol. 1949;42:66–76.
5. Ashton N. Anatomical study of Schlemm’s canal and aqueous veins
by means of neoprene casts. Br J Ophthalmol. 1951;35:291–303.
6. Huang AS, Penteado RC, Papoyan V, et al. Aqueous angio-
graphic outflow improvement after trabecular micro-bypass in
glaucoma patients. Ophthalmol Glaucoma. 2019;2:11–21.
7. Carreon T, van der Merwe E, Fellman RL, et al. Aqueous
outflow—a continuum from trabecular meshwork to episcleral
veins. Prog Retin Eye Res. 2017;57:108–133.
8. Huang AS, Camp A, Xu BY, et al. Aqueous angiography:
aqueous humor outflow imaging in live human subjects.
Ophthalmology. 2017;124:1249–1251.
9. Kagemann L, Wollstein G, Ishikawa H, et al. Identification and
assessment of Schlemm’s canal by spectral-domain optical coherence
tomography. Invest Ophthalmol Vis Sci. 2010;51:4054–4059.
10. Hengerer FH, Auffarth GU, Riffel C, et al. Prospective, non-
randomized, 36-month study of second-generation trabecular
micro-bypass stents with phacoemulsification in eyes with
various types of glaucoma. Ophthalmol Therp. 2018;7:405–415.
11. Samuelson TW, Sarkisian SR, Lubeck DM, et al. Prospective,
randomized, controlled pivotal trial of an ab interno implanted
trabecular micro-bypass in primary open-angle glaucoma and
cataract. Ophthalmology. 2019;126:811–821.
12. Meyer PAR. Re-orchestration of blood flow by micro-
circulations. Eye. 2018;32:222–229.
13. Khatib TZ, Meyer PAR, Lusthaus JA, et al. Haemoglobin
Video Imaging provides novel in vivo high-resolution imaging
and quantification of human aqueous outflow in glaucoma
patients. Ophthalmol Glaucoma. 2019;2:327–335.
14. Ascher KW. The Aqueous Veins: I. Physiologic importance of
the visible elimination of intraocular fluid. Am J Ophthalmol.
2018;192:29–54.
15. Meyer PA. Anterior segment vascular imaging. In: Easty DL,
Sparrow JM, eds. Oxford Textbook of Ophthalmology. Oxford,
UK: Oxford University Press; 1999:160–174.
16. Johnstone MA, Jamil A, Martin E, et al. Brimonidine-
dependent pulsatile aqueous discharge to the episcleral veins.
Invest Ophthalmol Vis Sci. 2006;47S:253.
17. Loewen RT, Brown EN, Roy P, et al. Regionally discrete
aqueous humor outflow quantification using fluorescein canalo-
grams. PLoS One. 2016;11:e0151754.
18. Fellman RL, Grover DS. Episcleral venous fluid wave in the
living human eye adjacent to microinvasive glaucoma surgery
(MIGS) supports laboratory research: outflow is limited
circumferentially, conserved distally, and favored inferonasally.
J Glaucoma. 2019;28:139–145.
19. Akagi T, Uji A, Huang AS, et al. Conjunctival and intrascleral
vasculatures assessed using anterior segment optic coherence
tomography angiography in normal eyes. Am J Ophthalmol.
2018;196:1–9.
20. Waxman S, Loewen RT, Dang Y, et al. High-resolution, three-
dimensional reconstruction of the outflow tract demonstrates
segmental differences in cleared eyes. Invest Ophthalmol Vis Sci.
2018;59:2371–2380.
21. Katz LJ, Erb C, Guillamet AC, et al. Long-term titrated IOP
control with one, two, or three trabecular micro-bypass stents in
open-angle glaucoma subjects on topical hypotensive medica-
tion: 42-month outcomes. Clin Ophthalmol. 2018;12:255–262.
22. Malvankar-Mehta MS, Chen YN, Iordanous Y, et al. iStent as
a solo procedure for glaucoma patients: a systematic review and
meta-analysis. PLoS One. 2015;10:e0128146.
23. Johnstone MA. Intraocular pressure regulation: findings of
pulse-dependent trabecular meshwork motion lead to unifying
concepts of intraocular pressure homeostasis. J Ocul Pharmacol
Ther. 2014;30:88–93.
24. Overby DR, Stamer WD, Johnson M. The changing paradigm
of outflow resistance generation: towards synergistic models of
the JCT and inner wall endothelium. Exp Eye Res. 2009;88:
656–670.
25. Li P, Reif R, Zhi Z, et al. Phase-sensitive optical coherence
tomography characterization of pulse-induce trabecular mesh-
work displacement in ex vivo nonhuman primate eyes. J
Biomed Opt. 2012;17:076026.
J Glaucoma Volume 29, Number 8, August 2020 Aqueous Outflow Visualization After TBS
Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved. www.glaucomajournal.com
|
665
Copyright r2020 Wolters Kluwer Health, Inc. All rights reserved.