Treatment Outcomes in the Primary Tube

Versus Trabeculectomy (PTVT) Study after

3 Years of Follow-up
Steven J. Gedde, MD,1 William J. Feuer, MS,1 Kin Sheng Lim, MD,2 Keith Barton, MD,3 Saurabh Goyal, MD,4
Iqbal I. Ahmed, MD,5 James D. Brandt, MD,6 for the Primary Tube Versus Trabeculectomy Study Group

Purpose: To report 3-year results of the Primary Tube Versus Trabeculectomy (PTVT) Study.
Design: Unmasked multicenter randomized clinical trial.
Participants: Two hundred forty-two eyes of 242 patients with medically uncontrolled glaucoma and no
previous incisional ocular surgery, including 125 in the tube group and 117 in the trabeculectomy group.
Methods: Patients were enrolled at 16 clinical centers and were assigned randomly to treatment with a tube
shunt (350-mm2 Baerveldt glaucoma implant) or trabeculectomy with mitomycin C (MMC; 0.4 mg/ml for 2
minutes).
Main Outcome Measures: The primary outcome measure was the rate of surgical failure, defined as
intraocular pressure (IOP) of more than 21 mmHg or reduced less than 20% from baseline, IOP of 5 mmHg or less,
reoperation for glaucoma, or loss of light perception vision. Secondary outcome measures included IOP, glau-
coma medical therapy, visual acuity, and surgical complications.
Results: The cumulative probability of failure after 3 years of follow-up was 33% in the tube group and 28%
in the trabeculectomy group (P ¼ 0.17; hazard ratio, 1.39; 95% confidence interval, 0.9e2.2). Mean  standard
deviation IOP was 14.04.2 mmHg in the tube group and 12.14.8 mmHg in the trabeculectomy group at 3 years
(P ¼ 0.008), and the number of glaucoma medications was 2.11.4 in the tube group and 1.21.5 in the tra-
beculectomy group (P < 0.001). Serious complications requiring reoperation or producing loss of 2 or more
Snellen lines developed in 3 patients (2%) in the tube group and 9 patients (8%) in the trabeculectomy group
(P ¼ 0.11).
Conclusions: There was no significant difference in the rate of surgical failure between the 2 surgical pro-
cedures at 3 years. Trabeculectomy with MMC achieved lower IOP with use of fewer glaucoma medications
compared with tube shunt surgery after 3 years of follow-up in the PTVT Study. Serious complications producing
vision loss or requiring reoperation occurred with similar frequency after both surgical
procedures. Ophthalmology 2019;-:1e13 ª 2019 by the American Academy of Ophthalmology

Supplemental material available at www.aaojournal.org.

The surgical options for managing glaucoma have expanded Johnson Vision, Santa Ana, CA) or trabeculectomy with
markedly in recent years. However, trabeculectomy and MMC. The goal of this investigator-initiated study was to
tube shunt implantation remain the most effective proced- offer information that will assist in surgical decision making
ures to reduce intraocular pressure (IOP). Medicare claims in similar patient groups. The methodology and outcomes
data1 and surveys of the American Glaucoma Society during the first postoperative year are described in previous
membership2e5 demonstrate that tube shunts are being publications.6,7 The primary outcome measure in the PTVT
increasingly used as an alternative to trabeculectomy, Study was the cumulative rate of surgical failure at 1 year.
especially in eyes with refractory glaucoma. Glaucoma The present report provides 3-year follow-up data on
surgeons have differing opinions about the preferred initial enrolled patients.
operation for medically uncontrolled glaucoma in eyes
without prior ocular surgery.4,5
The Primary Tube Versus Trabeculectomy (PTVT) Methods
Study is a multicenter randomized clinical trial comparing
The study was approved by the institutional review board at each
the safety and efficacy of tube shunt surgery and trabecu- clinical center before recruitment was initiated (see Appendix,
lectomy with mitomycin C (MMC) as a primary incisional available at www.aaojournal.org). Written informed consent was
glaucoma procedure. Patients with medically uncontrolled obtained from all participants for both treatment and participation
IOP were enrolled and assigned randomly to treatment with in the research. The study adhered to the tenets of the
a 350-mm2 Baerveldt glaucoma implant (Johnson & Declaration of Helsinki and the Health Insurance Portability and

ª 2019 by the American Academy of Ophthalmology https://doi.org/10.1016/j.ophtha.2019.10.002 1

Published by Elsevier Inc. ISSN 0161-6420/19
 Ophthalmology Volume -, Number -, Month 2019

Accountability Act. This study is registered at clinicaltrials.gov qualified successes. Study outcomes were reviewed regularly by an
(identifier, NCT00666237). The design and methods of the independent safety and data monitoring committee.
PTVT Study were described previously in detail,6 and they are Reoperation for glaucoma or a complication was defined as
summarized as follows. additional surgery performed in the operating room. Cyclo-
destruction also was considered a reoperation for glaucoma, and a
Eligibility Criteria vitreous tap with injection of intravitreal antibiotics was a reoper-
ation for a complication, regardless of whether these procedures
Patients 18 to 85 years of age who not had undergone previous were performed in the clinic or operating room. Interventions
incisional ocular surgery and demonstrated inadequately controlled performed at the slit lamp, such as needling procedures or refor-
glaucoma with IOP of 18 mmHg or more and 40 mmHg or less on mation of the anterior chamber, were not considered reoperations.
tolerated medical therapy were eligible for the study. Exclusion Early postoperative complications were defined as surgical com-
criteria included no light perception vision, pregnant or nursing plications developing within the first month after randomized
women, narrow anterior chamber angle, iris neovascularization or surgical treatment, and late postoperative complications were
proliferative retinopathy, iridocorneal endothelial syndrome, complications that occurred more than 1 month after glaucoma
epithelial or fibrous downgrowth, chronic or recurrent uveitis, surgery. Surgical complications that developed during the first
steroid-induced glaucoma, severe posterior blepharitis, unwilling- postoperative month and persisted with longer follow-up were
ness to discontinue contact lens use after surgery, previous counted only as early postoperative complications. Persistent
cyclodestructive procedure, conjunctival scarring resulting from diplopia, persistent corneal edema, and dysesthesia were defined as
prior ocular trauma or cicatrizing disease precluding a superior the postoperative development of these complications and their
trabeculectomy, functionally significant cataract, need for glau- presence at the 6-month follow-up visit or thereafter. Eyes that
coma surgery combined with other ocular procedures or anticipated showed positive Seidel results within the first month of follow-up
need for additional ocular surgery, unwillingness or inability to were classified as having wound leaks, and those with a positive
give consent, unwillingness to accept randomization, or inability to Seidel test after 1 month were categorized as having bleb leaks.
return for scheduled protocol visits. Only 1 eye of eligible patients Serious complications were defined as surgical complications that
was included in the study. produced loss of 2 or more lines of Snellen VA, required reoper-
ation to manage the complication, or both. Patients who underwent
additional surgery were censored from analysis of complications
Randomization and Treatment after the reoperation. Cataracts were considered to have progressed
The PTVT Study was conducted at 16 clinical centers. Eligibility if there was loss of 2 or more Snellen lines that was attributed to
was confirmed independently at the statistical coordinating center. cataract at the 6-month follow-up visit or thereafter, or if cataract
Enrolled patients were assigned randomly to treatment with a surgery was performed.
Baerveldt glaucoma implant (350 mm2) or trabeculectomy with
MMC (0.4 mg/ml for 2 minutes). Randomization was performed
within each clinical center and stratum using a permuted variable Sample Size Calculations
block size (n ¼ 2, 4, or 6) design. The 3 strata were designed to
Sample size calculations were performed based on projected dif-
include age, race, and presence of failed filtering surgery in the
ferences in failure rates between treatment groups. Enrollment of
nonstudy eye. Neither the patient nor the clinician was masked to
88 patients in each treatment group was expected to detect a
the randomization assignment.
relative risk of failure of 2.0 at 5 years assuming a 20% failure rate
in the lower-risk group with a 2-sided significance level of 0.05, a
Patient Visits power of 0.80, and analysis with the Yates-corrected chi-square
test. A total of 242 patients were recruited for the study to allow for
Baseline demographic and clinical information was obtained for a drop-out rate of 6% per year.
enrolled patients. Follow-up visits were scheduled at 1 day, 1
week, 1 month, 3 months, 6 months, 1 year, 18 months, 2 years, 3
years, 4 years, and 5 years after surgery. Data were collected with Statistical Analysis
standardized forms at each follow-up visit. Additional information
was acquired for patients undergoing a reoperation, including the Univariate comparisons between treatment groups were per-
date of surgery, type of procedure, and IOP level and number of formed with the 2-sided Student t test for continuous variables
glaucoma medications immediately before reoperation. and the Fisher exact or chi-square test (asymptotic, Yates cor-
rected, or exact permutation as appropriate) for categorical
Outcome Measures variables. Snellen VA measurements were converted to loga-
rithm of the minimum angle of resolution (logMAR) equivalents
The primary outcome measure in the PTVT Study was the cu- for the purpose of data analysis. The time to failure was defined
mulative rate of surgical failure at 1 year, but additional reporting as the time from surgical treatment to reoperation for glaucoma,
of data was planned for 3 and 5 years during the design of the trial. loss of light perception vision, or the first of 2 consecutive study
Secondary outcome measures included IOP, use of glaucoma visits after 3 months in which the patient showed persistent
medical therapy, visual acuity (VA), and surgical complications. hypotony (i.e., IOP 5 mmHg) or inadequately reduced IOP
Failure was defined a priori as IOP of more than 21 mmHg or less (i.e., IOP >21 mmHg or reduced <20% below baseline).
than 20% reduction below baseline on 2 consecutive follow-up Treatment comparisons of cumulative rate of failure, reoperation
visits after 3 months, IOP of 5 mmHg or less on 2 consecutive for glaucoma or complications, and cataract extraction were
follow-up visits after 3 months, reoperation for glaucoma, or loss assessed with the stratified Kaplan-Meier survival analysis log-
of light perception vision. Eyes that had not failed by the previous rank test. There was no prespecified statistical analysis plan;
criteria and were not receiving supplemental medical therapy were however, the design and analysis of data from the PTVT Study
considered complete successes. Eyes that had not failed but were modeled after the previously published Tube Versus Tra-
received supplemental medical therapy were categorized as beculectomy (TVT) Study.

2
 Gedde et al 
PTVT 3-Year Outcomes

Results upper IOP limit distinguishing success from failure was

changed. When inadequate IOP reduction was defined as IOP
Recruitment and Retention of more than 17 mmHg or reduced less than 20% from
baseline on 2 consecutive follow-up visits after 3 months, the
A total of 242 eyes of 242 patients were enrolled and underwent cumulative probability of failure at 3 years was 37% in the tube
surgical treatment between May 2008 and March 2015, including group and 28% in the trabeculectomy group (P ¼ 0.047, log-rank
125 patients in the tube group and 117 patients in the trabeculec- test adjusted for stratum; HR, 1.59; 95% CI, 1.0e2.5). When
tomy group. The progress of patients in the study is shown in inadequate IOP reduction was defined as IOP of more than 14
Figure S1 (available at www.aaojournal.org). During the first 3 mmHg on 2 consecutive visits after 3 months, the cumulative
years, 7% of follow-up visits were missed because of deaths and probability of failure was 49% in the tube group and 49% in the
losses to follow-up in both the tube and trabeculectomy groups trabeculectomy group at 3 years (P ¼ 0.65, log-rank test adjusted
(P ¼ 0.95, chi-square test). for stratum; HR, 1.09; 95% CI, 0.8e1.6).
The reasons for classification as a treatment failure are listed
Protocol Violations in Table 3. Inadequate IOP reduction (i.e., IOP >21 mmHg or
reduced <20% below baseline on 2 consecutive follow-up
One patient who was randomized to the trabeculectomy group
visits after 3 months) was the most common cause for failure
underwent LASIK in the study eye and should have been excluded
during the first 3 years of follow-up in both treatment groups,
because of previous incisional ocular surgery. Two patients were
occurring in 26 patients in the tube group and 18 patients in the
randomized to the trabeculectomy group but underwent placement
trabeculectomy group. Fifteen patients in the tube group and 8
of a tube shunt because of surgeon error. All patients were
patients in the trabeculectomy group were classified as having
analyzed according to the treatment group to which they were
failed treatment because they underwent a reoperation for glau-
assigned by randomization in an intention-to-treat analysis. None
coma. Persistent hypotony (i.e., IOP 5 mmHg on 2 consecutive
of the 3 patients who violated the study protocol experienced
visits after 3 months) was the cause for treatment failure in 4
treatment failure or underwent additional ocular surgery.
patients in the trabeculectomy group. Loss of VA from baseline
was seen in all of the patients with hypotony failure. No failures
Baseline Characteristics occurred in either treatment group because of loss of light
The baseline characteristics of the study population are presented perception vision. A difference in the distribution of reasons for
in Table 1. The mean  standard deviation (SD) age of the study failure was seen between treatment groups, with more failures
population at enrollment was 61.411.8 years, and 160 (66%) occurring because of glaucoma reoperations in the tube group
patients were men. The mean  SD IOP of the overall study and more hypotony failures in the trabeculectomy group (P ¼
group was 23.65.3 mmHg, and the mean  SD number of 0.050, exact permutation chi-square test).
glaucoma medications was 3.21.1. The most common
diagnosis was primary open-angle glaucoma in 218 eyes (90%).
The mean  SD mean deviation with Humphrey visual field testing Intraocular Pressure
was e14.69.9 dB. There were no significant differences in any of
Baseline and follow-up IOP measurements for the tube and tra-
the demographic or ocular features between treatment groups at
beculectomy groups are provided in Table 4, and they are also
enrollment.
graphically presented in Figure 4 and Figure S5 (available at
www.aaojournal.org). Patients who underwent additional
Treatment Outcomes glaucoma surgery were censored from analysis after reoperation.
The outcomes of randomized patients unadjusted for follow-up At 3 years, mean  SD IOP was 14.04.2 mmHg in the tube
time are presented in Table S2 (available at group and 12.14.8 mmHg in the trabeculectomy group
www.aaojournal.org). All patients who completed 3-year follow- (P ¼ 0.008, Student t test). Among patients who completed
up visits, who experienced a prior failure, or both were included 3-year follow-up visits, mean  SD IOP reduction from baseline
in this analysis. At 3 years, treatment failure had occurred in 41 was 9.26.7 mmHg (39%) in the tube group and 11.17.7 mmHg
patients (39%) in the tube group and 30 patients (30%) in the (46%) in the trabeculectomy group (P ¼ 0.08, Student t test). The
trabeculectomy group (P ¼ 0.17, logistic regression analysis trabeculectomy group showed lower mean IOPs than the tube
adjusted for stratum). In the tube group, 14 patients (13%) were group at all follow-up visits during the first 3 years of the study,
classified as complete successes, and 51 patients (48%) were and these differences were statistically significant, with the
qualified successes. In the trabeculectomy group, 44 patients (44%) exception of the 18-month and 2-year time points. At 3 years, 54
were complete successes, and 27 patients (27%) were qualified patients (61%) in the tube group and 66 patients (74%) in the
successes. The rate of complete success was significantly higher in trabeculectomy group showed an IOP of 14 mmHg or less (P ¼
the trabeculectomy group relative to the tube group (P < 0.001, 0.1, chi-square test).
logistic regression analysis adjusted for stratum). An analysis was performed carrying the last observation
Kaplan-Meier survival analysis also was used to compare fail- forward, which included the last IOP measurement before
ure rates between the 2 treatment groups, and the results are pre- glaucoma reoperation for patients who underwent additional
sented in Figure 2. The cumulative probability of failure was 33% glaucoma surgery and the last study visit for patients with
in the tube group and 28% in the trabeculectomy group at 3 years missing follow-up. At 3 years, mean  SD IOP was 15.15.7
(P ¼ 0.17, log-rank test adjusted for stratum; hazard ratio [HR], mmHg in the tube group and 13.15.4 mmHg in the trabecu-
1.39; 95% confidence interval [CI], 0.9e2.2). No significant dif- lectomy group (P ¼ 0.005, Student t test) with the last obser-
ferences in treatment efficacy were found between strata (P ¼ 0.17, vation carried forward. An assessment of IOP also was made for
Cox regression analysis). all patients, including those who underwent further surgery for
Figure 3 shows the failure rates for the 2 treatment groups glaucoma. At 3 years, mean  SD IOP was 14.64.6 mmHg in
with alternative outcome criteria. Patients with persistent the tube group and 12.24.7 mmHg in the trabeculectomy
hypotony, reoperation for glaucoma, or loss of light perception group (P  0.001, Student t test) considering all medical and
vision were still classified as treatment failures. However, the surgical management.

3
 Ophthalmology Volume -, Number -, Month 2019

Table 1. Baseline Characteristics of Primary Tube Versus Trabeculectomy Study Patients

Tube Group (n [ 125) Trabeculectomy Group (n [ 117)

Age (yrs)
Mean  SD 62.011.4 60.812.3
Median (range) 62 (28e85) 61 (21e85)
Gender, no. (%)
Male 84 (67) 76 (65)
Female 41 (33) 41 (35)
Race, no. (%)
Black 59 (47) 57 (49)
White 50 (40) 45 (39)
Hispanic 9 (7) 6 (5)
Asian 6 (5) 7 (6)
Other 1 (1) 2 (2)
Hypertension, no. (%) 63 (50) 55 (47)
Diabetes mellitus, no. (%) 18 (14) 27 (23)
Study eye, no. (%)
Right 68 (54) 60 (51)
Left 57 (46) 57 (49)
IOP (mmHg)
Mean  SD 23.34.9 23.95.7
Median (range) 22 (18e40) 22 (18e40)
Central corneal thickness (mm)
Mean  SD 52537 52433
Median (range) 530 (419e602) 522 (431e600)
Glaucoma medications
Mean  SD 3.11.1 3.21.1
Median (range) 3 (0e6) 3 (0e5)
Diagnosis, no. (%)
POAG 109 (87) 109 (93)
CACG 5 (4) 3 (3)
PG 4 (3) 2 (2)
PXFG 4 (3) 1 (1)
Other 3 (2) 2 (2)
Previous ocular laser treatment, no. (%)
LTP 34 (27) 29 (25)
LPI 11 (9) 2 (2)
Other 9 (7) 5 (4)
ETDRS VA, mean  SD 7320 7320
Snellen VA
logMAR, mean  SD 0.200.42 0.250.51
Median 20/25 20/25
Range 20/13eHM 20/13eLP
Cataract, no. (%) 76 (61) 65 (56)
Mild 62 (82) 50 (77)
Moderate 13 (17) 14 (22)
Severe 1 (1) 1 (1)
Diplopia, no. (%) 8 (6) 4 (3)
Humphrey visual fields, mean  SD
MD e14.510.2 e14.79.7
PSD 7.713.86 8.193.57
Stratum*
1 55 (44) 53 (45)
2 8 (6) 5 (4)
3 62 (50) 59 (50)

CACG ¼ chronic angle-closure glaucoma; CF ¼ counting fingers; ETDRS ¼ Early Treatment Diabetic Retinopathy Study; IOP ¼ intraocular pressure;
logMAR ¼ logarithm of the minimum angle of resolution; LP ¼ light perception; LPI ¼ laser iridotomy; LTP ¼ laser trabeculoplasty; MD ¼ mean de-
viation; PG ¼ pigmentary glaucoma; POAG ¼ primary open-angle glaucoma; PSD ¼ pattern standard deviation; PXFG ¼ pseudoexfoliation glaucoma;
SD ¼ standard deviation; VA ¼ visual acuity.
*Stratum 1 ¼ no failed glaucoma surgery in fellow eye, age 50 years or older, and nonblack race; stratum 2 ¼ failed glaucoma surgery in fellow eye; stratum
3 ¼ no failed glaucoma surgery in fellow eye, age younger than 50 years, black race, or a combination thereof.

4
 Gedde et al 
PTVT 3-Year Outcomes

Figure 2. Kaplan-Meier plots showing the cumulative probability of failure in the Primary Tube Versus Trabeculectomy Study. The number of patients at
risk at each follow-up visit is shown at the bottom.

Medical Therapy shunt in 6 patients, endoscopic cyclophotocoagulation in

conjunction with cataract extraction in 4 patients, transscleral
Table 4 shows the number of glaucoma medications in the tube and cyclophotocoagulation in 4 patients, trabeculectomy with MMC in
trabeculectomy groups at baseline and follow-up. Patients who 3 patients, insertion of an InnFocus microshunt with MMC in 1
underwent additional glaucoma surgery were censored from anal- patient, and implantation of a Xen gel stent with MMC in 1 patient.
ysis after reoperation. At 3 years, the mean  SD number of One of the patients who underwent a second tube shunt procedure
glaucoma medications was 2.11.4 in the tube group and 1.21.5 subsequently underwent endoscopic cyclophotocoagulation as a
in the trabeculectomy group (P < 0.001, Student t test). Among second glaucoma reoperation. Ten patients in the trabeculectomy
patients with qualified success at 3 years, the mean  SD number group underwent glaucoma reoperations, including tube shunt
of glaucoma medications was 2.51.1 in the tube group and placement in 8 patients, a repeat trabeculectomy with MMC in 1
2.50.9 in the trabeculectomy group (P ¼ 0.97). The mean  SD patient, and trabeculectomy revision in 1 patient.
number of glaucoma medications decreased from baseline by Because the surgeon was not masked to the treatment assign-
1.01.6 in the tube group and 2.01.7 in the trabeculectomy ment, a potential bias existed with respect to the decision to reo-
group in patients who completed 3-year follow-up visits (P < perate for glaucoma. To evaluate for selection bias, the IOP levels
0.001, Student t test). Significantly greater use of glaucoma med- were compared between treatment groups in patients who under-
ical therapy was observed in the tube group compared with the went glaucoma reoperation, as well as those who failed because of
trabeculectomy group at all follow-up visits during the first 3 years inadequate IOP reduction but did not have additional glaucoma
of the study. surgery. The mean  SD IOP was 20.84.1 mmHg for the 18
The mean  SD number of glaucoma medications was 2.11.4 patients in the tube group and 22.86.3 mmHg for the 10 patients
in the tube group and 1.21.5 in the trabeculectomy group at 3 in the trabeculectomy group at the time of reoperation for glau-
years (P < 0.001, Student t test) with the last observation carried coma (P ¼ 0.36, Student t test). The IOP levels also were
forward. When patients who underwent additional glaucoma sur- compared between the 23 patients in the tube group and 16 patients
gery were included in the analysis, the mean  SD number of in the trabeculectomy group who failed because of inadequate IOP
glaucoma medications was 2.11.5 in the tube group and 1.21.5 reduction but who did not undergo additional glaucoma surgery
in the trabeculectomy group (P < 0.001, Student t test). during the first 3 years of follow-up. In this patient subgroup, the
mean  SD IOP was 18.33.6 mmHg in the tube group and
18.22.2 in the trabeculectomy group (P ¼ 0.92, Student t test).
Reoperation for Glaucoma
The 3-year cumulative reoperation rate for glaucoma with Kaplan- Visual Acuity
Meier survival analysis was 16% in the tube group and 9% in the
trabeculectomy group (P ¼ 0.15, log-rank test adjusted for stra- Table 5 shows VA results. Among patients who completed 3-year
tum). A total of 18 patients in the tube group underwent additional follow-up visits, mean  SD logMAR Snellen VA decreased
glaucoma surgery, which involved placement of a second tube 0.160.40 units from baseline (P  0.001, paired t test), and mean

5
 Ophthalmology Volume -, Number -, Month 2019

Figure 3. Kaplan-Meier plots showing the cumulative probability of failure in the Primary Tube Versus Trabeculectomy Study defining inadequate
intraocular pressure (IOP) reduction as (A) IOP of more than 17 mmHg or reduced less than 20% below baseline or (B) IOP of more than 14 mmHg.
Inadequate IOP reduction criteria must have been present on 2 consecutive visits after 3 months to qualify as failure. Patients with persistent hypotony,
reoperation for glaucoma, and loss of light perception vision were classified as experiencing treatment failure. The number of patients at risk at each follow-
up visit is shown at the bottom.

6
 Gedde et al 
PTVT 3-Year Outcomes

Table 3. Reasons for Treatment Failure in the Primary Tube Table 4. Intraocular Pressure and Medical Therapy at Baseline
Versus Trabeculectomy Study and Follow-up in the Primary Tube Versus Trabeculectomy Study

Tube Group Trabeculectomy Group Tube Trabeculectomy

(n [ 41) (n [ 30) Group Group P Value*
Inadequate IOP reduction* 26 (63) 18 (60) Baseline
Reoperation for glaucoma 15 (37) 8 (27) IOP (mmHg) 23.34.9 23.95.7 0.35
Persistent hypotonyy 0 4 (13) Glaucoma medications 3.11.1 3.21.1 0.56
Loss of light perception 0 0 No. 125 117
1 day
IOP (mm Hg) 19.09.7 16.39.2 0.03
IOP ¼ intraocular pressure. No. 125 116
Data are presented as number (percentage). P ¼ 0.050 for the difference in
1 wk
distribution of reasons for failure between treatment groups (exact 42
permutation chi-square test). IOP (mmHg) 18.28.5 15.19.2 0.007
*Intraocular pressure >21 mmHg or reduced <20% below baseline on 2 Glaucoma medications 1.11.4 0.10.5 <0.001
consecutive follow-up visits after 3 months. No. 120 116
y 1 mo
Intraocular pressure 5 mmHg on 2 consecutive follow-up visits after 3
months. IOP (mmHg) 19.77.3 13.16.3 <0.001
Glaucoma medications 1.41.5 0.20.8 <0.001
No. 124 115
3 mos
IOP (mmHg) 18.05.9 12.54.9 <0.001
 SD Early Treatment Diabetic Retinopathy Study (ETDRS) VA Glaucoma medications 1.91.4 0.61.2 <0.001
was reduced by 6.017 letters from baseline (P ¼ 0.002, paired t No. 121 113
test) in the tube group. In the trabeculectomy group, mean  SD 6 mos
logMAR Snellen VA decreased 0.120.37 units (P ¼ 0.002, IOP (mmHg) 14.74.4 12.84.8 0.003
paired t test) and mean  SD ETDRS VA declined 6.916 letters Glaucoma medications 2.11.4 0.61.2 <0.001
(P < 0.001, paired t test) from baseline to the 3-year follow-up No. 112 109
visit. No significant differences in Snellen VA (P ¼ 0.74, Stu- 1 yr
dent t test) or ETDRS VA (P ¼ 0.95, Student t test) were seen IOP (mmHg) 13.84.1 12.44.4 0.01
between the tube and trabeculectomy groups at 3 years. The Glaucoma medications 2.11.4 0.91.4 <0.001
changes in Snellen VA (P ¼ 0.44, Student t test) and ETDRS VA No. 108 105
(P ¼ 0.74, Student t test) from baseline also were similar between 18 mos
treatment groups among patients who completed 3 years of follow- IOP (mmHg) 13.54.1 12.84.7 0.32
up. Glaucoma medications 2.01.3 0.81.3 <0.001
Loss of 2 or more Snellen lines from baseline had occurred in No. 100 97
2 yrs
17 patients (14%) in the tube group and 18 patients (15%) in the
IOP (mmHg) 13.74.0 12.95.2 0.24
trabeculectomy group at 3 years (P ¼ 0.83, chi-square test). The
Glaucoma medications 2.21.4 1.01.5 <0.001
examining clinician was asked to provide an explanation for this No. 100 98
reduction in VA, which may have included complications that 3 yrs
occurred after randomized surgical treatment or after additional IOP (mmHg) 14.04.2 12.14.8 0.008
ocular surgery. The most frequent cause of vision loss during the Glaucoma medications 2.11.4 1.21.5 <0.001
first 3 years of the study was cataract in 8 patients in the tube No. 88 89
group and 7 patients in the trabeculectomy group. Vision loss
resulting from glaucoma developed in 8 patients in the tube
group and 6 patients in the trabeculectomy group. Macular dis- IOP ¼ intraocular pressure.
Data are mean  standard deviation. Data censored after a reoperation for
ease produced vision loss in 5 patients in the trabeculectomy
glaucoma.
group, including 2 patients with hypotony maculopathy, 1 patient *Student t test.
with cystoid macular edema, 1 patient with central serous cho-
rioretinopathy, and 1 patient with an epiretinal membrane. Other
causes of vision loss in the trabeculectomy group were branch
retinal vein occlusion in 1 patient, central retinal artery occlusion
in 1 patient, and corneal edema in 1 patient. The cause of vision Postoperative Complications
loss was unknown in 1 patient in the trabeculectomy group. In
the tube group, additional causes of vision loss included epi- Postoperative complications encountered during the first 3 years of
retinal membrane in 1 patient and posterior capsular opacification the PTVT Study are shown in Table 7. A total of 36 early
in 1 patient. postoperative complications were reported in 25 patients (20%)
in the tube group, and 55 early complications were noted in 39
patients (33%) in the trabeculectomy group (P ¼ 0.028, chi-
Postoperative Interventions square test). Wound leak (P < 0.001, chi-square test) and encap-
sulated bleb (P ¼ 0.009, chi-square test) were early postoperative
Postoperative interventions are listed in Table 6. A total of 108 complications that were significantly more common in the trabe-
interventions were performed in 77 patients (62%) in the tube culectomy group compared with the tube group. No early post-
group, and 126 interventions were undertaken in 77 patients operative complications occurred with significantly greater
(66%) in the trabeculectomy group (P ¼ 0.58, chi-square test). frequency in the tube group than the trabeculectomy group. A total
Rip-cord removal and laser suture lysis were the most common of 30 late postoperative complications were seen in 27 patients
interventions in the tube and trabeculectomy groups, respectively. (22%) in the tube group, and 31 complications were observed in 29

7
 Ophthalmology Volume -, Number -, Month 2019

Figure 4. Graph showing intraocular pressure (IOP) at baseline and follow-up in the Primary Tube Versus Trabeculectomy Study. Data are presented as
mean  standard error of the mean and are censored after a reoperation for glaucoma.

patients (25%) in the trabeculectomy group (P ¼ 0.65, chi-square trabeculectomy group experienced 1 or more surgical complica-
test). No late postoperative complication occurred with signifi- tions after surgery (P ¼ 0.046, chi-square test).
cantly higher frequency in either treatment group. Several patients Table 8 shows serious complications resulting in reoperation,
in each treatment group developed both early and late post- vision loss, or both. Serious complications were observed in 3
operative complications. During the first 3 years of follow-up, 43 patients (2%) in the tube group and 9 patients (8%) in the
patients (34%) in the tube group and 56 patients (48%) in the trabeculectomy group (P ¼ 0.11, Fisher exact test). Reoperation
for complications was performed in 3 patients in the tube group
and 9 patients in the trabeculectomy group. In the
trabeculectomy group, loss of 2 or more Snellen lines developed
Table 5. Visual Acuity Results in the Primary Tube Versus because of hypotony maculopathy in 1 patient and a combination
Trabeculectomy Study

Tube Group Trabeculectomy

(n [ 125) Group (n [ 117) P Value Table 6. Postoperative Interventions in the Primary Tube Versus
Trabeculectomy Study
ETDRS VA, mean  SD
Baseline 7320 7320 0.96* Tube Group Trabeculectomy
3 yrs 7321 7123 0.95* (n [ 125) Group (n [ 117)
Changey e6.017 e6.916 0.74*
Snellen VA, logMAR Removal of rip-cord 65 (52) NA
mean  SD Laser suture lysis 23 (18) 34 (29)
Baseline 0.200.42 0.250.51 0.42* Removal of releasable suture NA 21 (18)
3 yrs 0.300.57 0.330.61 0.74* 5-fluorouracil injection 0 29 (25)
Changey 0.160.40 0.120.37 0.44* Mitomycin C injection 0 2 (2)
Loss of 2 Snellen lines, 17 (14) 18 (15) 0.83x Needling 0 23 (20)
no. (%)z Anterior chamber reformation 10 (8) 6 (5)
Cataract 8 7 Paracentesis 7 (6) 2 (2)
Glaucoma 8 6 Suture wound 2 (2) 3 (3)
Macular disease 1 5 Avastin injection 0 3 (3)
Other 1 3 Subconjunctival steroid injection 0 1 (1)
Unknown 0 1 Laser iridotomy 0 1 (1)
Laser iridoplasty 1 (1) 1 (1)
Total no. of patients with 77 (62) 77 (66)
ETDRS ¼ Early Treatment Diabetic Retinopathy Study; logMAR ¼ log- postoperative interventions*,y
arithm of the minimum angle of resolution; SD ¼ standard deviation;
VA ¼ visual acuity.
*Student t test. NA ¼ not applicable.
y
Not all patients seen at baseline returned for 3-year visits, so the baseline Data are number (percentage).
mean minus the 3-year mean does not equal the change. *Some patients underwent more than 1 intervention.
z y
Some patients had more than 1 reason for decreased vision. P ¼ 0.58 for the difference in total number of patients with postoperative
x
Chi-square test. interventions between treatment groups (chi-square test).

8
 Gedde et al 
PTVT 3-Year Outcomes

Table 7. Surgical Complications in the Primary Tube Versus Trabeculectomy Study

Tube Group (n [ 125) Trabeculectomy Group (n [ 117)

Early postoperative complications*
Shallow or flat anterior chamber 13 (10) 11 (9)
Choroidal effusion 9 (7) 12 (10)
Wound leak 1 (1) 14 (12)
Hyphema 8 (6) 5 (4)
Encapsulated bleb 0 8 (7)
Hypotony maculopathy 1 (1) 3 (3)
Wound dehiscence 2 (2) 0
Aqueous misdirection 0 1 (1)
Corneal dellen 1 (1) 0
Cystoid macular edema 1 (1) 0
Suture-related infection 0 1 (1)
Total number of patients with early postoperative complicationsy,z 25 (20) 39 (33)
Late postoperative complicationsx
Encapsulated bleb 14 (11) 14 (12)
Persistent diplopia 3 (2) 4 (3)
Shallow or flat anterior chamber 3 (2) 3 (3)
Dysesthesia 1 (1) 3 (3)
Hypotony maculopathy 0 3 (3)
Iritis 2 (2) 1 (1)
Choroidal effusion 2 (2) 0
Cystoid macular edema 2 (2) 0
Persistent corneal edema 0 2 (2)
Conjunctival cyst 1 (1) 0
Anterior bleb migration/irregular astigmatism 0 1 (1)
Plate erosion 1 (1) NA
Tube retraction 1 (1) NA
Total number of patients with late postoperative complicationsy,jj 27 (22) 29 (25)
Total no. of patients with postoperative complications{,# 43 (34) 56 (48)

NA ¼ not applicable.
Data are number of patients (percentage). Data censored after a reoperation.
*Onset 1 month.
y
Some patients had more than 1 complication.
z
P ¼ 0.028 for the difference in rates of early postoperative complications between treatment groups (chi-square test).
x
Onset >1 month.
k
P ¼ 0.65 for the difference in rates of late postoperative complications between treatment groups (chi-square test).
{
Some patients had early and late postoperative complications.
#
P ¼ 0.046 for the difference in overall rates of postoperative complications between treatment groups (chi-square test).

of hypotony maculopathy and corneal edema in 1 patient. This exposure of the end plate. A phacoemulsification cataract extrac-
vision loss persisted despite trabeculectomy revision for tion and endoscopic cyclophotocoagulation was performed at the
hypotony maculopathy in both patients. time of shunt removal, and the patient was classified as a failure
because of additional glaucoma surgery.

Reoperation for Complications Cataract Progression

The 3-year cumulative reoperation rate for complications from The 3-year cumulative rate of cataract extraction using Kaplan-
Kaplan-Meier survival analysis was 3% in the tube group and 8% Meier survival analysis was 32% in the tube group and 33% in
in the trabeculectomy group (P ¼ 0.059, log-rank test). A total of 3 the trabeculectomy group (P ¼ 0.42, log-rank test adjusted for
patients (2%) in the tube group and 9 patients (8%) in the trabe- stratum). Cataract surgery was performed in 43 patients in the tube
culectomy group underwent additional surgery to manage post- group and 36 patients in the trabeculectomy group during the first 3
operative complications. Five patients in the trabeculectomy group years of follow-up. An additional 8 patients in the tube group and 7
underwent bleb revision, including 4 patients with wound leaks in patients in the trabeculectomy group experienced loss of 2 or more
whom management with a bandage contact lens or suturing at the lines of Snellen VA attributed to cataract. Cataract progression
slit lamp failed and 1 patient with anterior migration of the bleb occurred in 51 patients (41%) in the tube group and 42 patients
with irregular astigmatism. Trabeculectomy revision was per- (36%) in the trabeculectomy group after 3 years (P ¼ 0.61, chi-
formed in 3 patients for hypotony maculopathy, and 1 patient square test).
underwent a trabeculectomy revision and anterior chamber
washout for an 8-ball hyphema. A conjunctival cyst was excised in Sensitivity Analysis
1 patient in the tube group, and 1 patient underwent tube revision
for retraction of the tube retraction from the anterior chamber. One Although the principal analysis was performed on observed data,
patient in the tube group underwent removal of the tube shunt for we also performed a sensitivity analysis in which multiple

9
 Ophthalmology Volume -, Number -, Month 2019

Table 8. Serious Complications Associated with Reoperation, trabeculectomy group after 1 year of follow-up.7 However,
Vision Loss, or Both in the Primary Tube Versus Trabeculectomy the cumulative probability of failure increased to 33% in
Study the tube group and 28% in the trabeculectomy group at 3
years, a difference that was no longer statistically
Tube Group Trabeculectomy
(n [ 125) Group (n [ 117)
significant. The outcome measures for the PTVT Study
were developed a priori, and our definitions for failure and
Reoperation for complications 3 (2) 9 (8) success are consistent with recommendations from the
8-ball hyphema 0 1 World Glaucoma Association for the reporting of
Anterior migration of bleb 0 1
or irregular astigmatism outcomes of glaucoma surgical trials.8
Conjunctival cyst 1 0 Tube shunt surgery and trabeculectomy with MMC were
Hypotony maculopathy 0 3 both effective in lowering IOP as a primary glaucoma pro-
Plate exposure 1 NA cedure. Among patients who completed 3-year follow-up
Tube retraction 1 NA visits, IOP was reduced by 39% in the tube group and
Wound leak 0 4 46% in the trabeculectomy group. Most patients in both
Vision loss of 2 Snellen lines 2 (2)*
Corneal edema 0 1
treatment groups showed an IOP of 14 mmHg or less at 3
Hypotony maculopathy 0 2 years. Mean IOPs were lower in the trabeculectomy group
Total no. of patients with 3 (2) 9 (8)z compared with the tube group at all time points, and the
serious complicationsy differences were statistically significant with the exception
of 18 months and 2 years. The lower IOPs in the trabecu-
NA ¼ not applicable. lectomy group were achieved with significantly fewer
Data are number of patients (percentage). Data censored after a glaucoma medications relative to the tube group.
reoperation. Treatment success was subdivided into complete and
*One patient in the trabeculectomy group experienced vision loss that was
attributed to 2 complications.
qualified successes, based on the use of supplemental
y
P ¼ 0.11 for the difference in serious complication rates between treat- medical therapy. Although the overall rates of success were
ment groups (chi-square test). similar between the 2 treatment groups, the rate of complete
z
Two patients in the trabeculectomy group had both a reoperation for a success was higher in the trabeculectomy group than the
complication and vision loss. tube group. This is consistent with the observed greater use
of glaucoma medications by the tube group at 3 years. The
trabeculectomy group showed a progressive increase in the
imputation was used to estimate missing IOP and number of mean number of glaucoma medications during 3 years of
glaucoma medications. In the imputed dataset, the averages by follow-up, whereas use of glaucoma medical therapy was
study visit and treatment group differed from the observed ones fairly constant in the tube group.
listed in Table 4 by 0.5 mmHg or more and 0.2 medication or We recognize the difficulty in defining success by an
fewer. All postoperative mean IOPs and number of medications
arbitrary IOP level, because individuals vary in the sus-
were significantly different between treatment groups with the
exception of mean IOP at 18 months and 2 years, as was the ceptibility of their optic nerves to the damaging effect of
case with the observed data. However, some of the differences IOP. Results from several multicenter randomized clinical
were slightly more and others slightly less significant with the trials have suggested that IOP of 21 mmHg or less may not
imputed data. The cumulative probability of failure at 3 years be adequate to prevent glaucomatous progression in many
was 34% in the tube group and 28% in the trabeculectomy group patients.9e11 Planned secondary analyses were performed to
(P ¼ 0.20; HR, 1.35; 95% CI, 0.9e2.1), leaving the HR and determine if the PTVT Study results changed if more
associated statistical significance almost unchanged with the stringent IOP criteria were applied to define success. Similar
imputed data. failure rates were seen in both treatment groups when the
upper IOP level defining success was reduced to 14 mmHg.
However, the tube group showed a significantly higher
Discussion failure rate than the trabeculectomy group when the upper
IOP level defining failure was decreased to 17 mmHg.
The PTVT Study is an unmasked multicenter clinical trial The distribution of reasons for failure were different
that prospectively enrolled patients with medically uncon- between the 2 treatment groups. Failure because of persis-
trolled glaucoma who had not previously undergone inci- tent hypotony occurred more frequently in the trabeculec-
sional ocular surgery and that randomized them to treatment tomy group, and failure because of a glaucoma reoperation
with a 350-mm2 Baerveldt glaucoma implant or a trabecu- was more common in the tube group. It has been argued that
lectomy with MMC. Some patients in the study who initially hypotony may be an acceptable outcome of glaucoma sur-
experienced successful surgical treatment subsequently gery if it is not associated with vision loss.12 It is noteworthy
experienced treatment failure. Unfortunately, loss of IOP that all of the patients whose treatment failed because of
control may occur over time with all currently available hypotony also experienced vision loss. Inadequate IOP
glaucoma procedures. The primary outcome measure in the reduction was the most common reason for failure in both
PTVT Study was the cumulative rate of surgical failure at 1 treatment groups.
year, but additional reporting of data was planned at 3 and 5 The rate of reoperation for glaucoma was similar between
years during the design of the trial.6 The failure rate was the 2 treatment groups. Patients who experience trabecu-
significantly higher in the tube group compared with the lectomy failure and require additional glaucoma surgery

10
 Gedde et al 
PTVT 3-Year Outcomes

generally undergo repeat trabeculectomy or tube shunt The benefit of a glaucoma procedure in reducing IOP
placement. However, additional glaucoma surgery in eyes in must be balanced against the risk of adverse events. The
which tube shunt surgery has failed frequently is more rates of early complications and overall complications dur-
complex and traditionally has involved placement of a ing the first 3 years of follow-up were higher in the trabe-
second tube shunt or cyclodestruction.13e15 Subconjunctival culectomy group than the tube group. All surgical
filtering surgery may be a feasible option in some patients complications are not equal in severity. We defined serious
who have undergone primary tube shunt implantation. complications as postoperative events that produced loss of
Reoperations for glaucoma included trabeculectomy with 2 or more lines of Snellen VA, required reoperation to
MMC was performed in 3 patients, an InnFocus microshunt manage the complication, or both, as was done in the TVT
(Santen Pharmaeutical Co., Osaka, Japan) was inserted in 1 Study20 and Ahmed Baerveldt Comparison Study.21 No
patient with MMC, and a Xen gel stent (Allergan PLC, significant difference in the incidence of serious
Irvine, CA, USA) was implanted with MMC in 1 patient in complications was seen between treatment groups. It is
the tube group. Because investigators in the PTVT Study noteworthy that no patients in the tube group
were not masked to the treatment assignment and the de- demonstrated persistent corneal edema after 3 years of
cision to reoperate was left to the surgeon’s discretion, a follow-up, because late corneal decompensation is a
potential bias existed in the decision to reoperate for glau- known complication of tube shunt implantation.
coma. We explored the possibility that surgeons may have A large number of postoperative complications were
had a higher threshold for performing additional glaucoma observed in the PTVT Study, but most were transient and
surgery in the tube group than the trabeculectomy group. self-limited. High rates of complications also were seen in
The mean IOP before reoperation was similar in the tube the TVT Study,20 the Ahmed Baerveldt Comparison
and trabeculectomy groups, and no significant difference Study,21 the Ahmed Versus Baerveldt Study,22 the
was seen between treatment groups in mean IOP among Collaborative Initial Glaucoma Treatment Study,23 and the
patients whose treatment failed because of inadequate IOP Advanced Glaucoma Intervention Study.24 Prospective
reduction but did not undergo additional glaucoma surgery. studies are expected to report higher complication rates
We found no evidence of selection bias for additional than retrospective case series. Surgical complications, such
glaucoma surgery, although the small number of glaucoma as choroidal effusions and bleb leaks, may be overlooked
reoperations limits the ability to detect differences between unless attention is directed specifically toward their
treatment groups. detection. Furthermore, complications may not be
Reduction of VA occurred in the tube group and trabe- documented in the medical record even when observed, if
culectomy group during the first 3 years of follow-up. they are believed to be insignificant.
Snellen and ETDRS VA were similar in both treatment Retrospective and prospective studies have compared
groups at 3 years, and no significant difference in the rate of trabeculectomy with tube shunt surgery. Wilson et al25
vision loss was observed between treatment groups. All conducted a prospective clinical trial in Sri Lanka
patients in the PTVT Study were phakic at enrollment, and randomizing unoperated eyes with primary angle-closure
vision loss of 2 or more Snellen lines was attributed most glaucoma and primary open-angle glaucoma to initial tra-
frequently to cataract by the examining clinicians. Cataracts beculectomy or placement of the Ahmed glaucoma valve
were considered to have progressed if loss of 2 or more lines (New World Medical, Inc, Rancho Cucamonga, CA). Sur-
of Snellen VA was attributed to lens opacification or if gical success, mean IOP, glaucoma medications, VA, and
cataract surgery was performed. Cataract progression was postoperative complications were comparable with both
common during the first 3 years of follow-up but occurred at surgical procedures at the final follow-up period (41e52
a similar rate in the both treatment groups. Multiple studies months). The TVT Study was a multicenter randomized
also have reported that glaucoma surgery is associated with clinical trial comparing Baerveldt implantation and trabe-
the development of cataract.9,16e21 culectomy with MMC in eyes with previous cataract sur-
Postoperative interventions were undertaken with similar gery, glaucoma surgery, or both.26 The rate of surgical
frequency in both treatment groups. Most interventions were success with survival analysis was higher with tube shunt
performed in the early postoperative period and have been surgery than trabeculectomy with MMC throughout 5
reported previously.7 Rip-cord removal and laser suture years of follow-up, although similar mean IOP and use of
lysis were the most common interventions in the tube and glaucoma medications were observed with both surgical
trabeculectomy groups, respectively. The Baerveldt glau- procedures at 5 years. Early postoperative complications
coma implant is a nonvalved tube shunt that requires tem- occurred more frequently after trabeculectomy with MMC
porary restriction of flow through the device until than tube shunt placement, but both procedures showed
encapsulation of the end plate occurs. Use of an intraluminal similar rates of late and serious complications after 5
rip-cord was the most common method for temporary tube years.20 Molteno et al27 reported the results of a prospective,
occlusion in the study. Trabeculectomy is the only glaucoma nonrandomized study comparing primary trabeculectomy
operation that allows titration of the IOP-lowering effect and primary insertion of the Molteno implant (Molteno
after surgery. Patients respond differently to glaucoma sur- Ophthalmic Limited, Dunedin, New Zealand). The
gery, so the ability to increase filtration selectively after cumulative rate of failure was higher after trabeculectomy
surgery is a valuable feature of trabeculectomy. relative to Molteno implant placement during 20 years of

11
 Ophthalmology Volume -, Number -, Month 2019

follow-up, but no significant differences were seen between References

the 2 procedures in mean IOP, glaucoma medical therapy,
complications, or vision loss. Panarelli et al28
1. Arora KS, Robin AL, Corcoran KJ, et al. Use of various
retrospectively reviewed the outcomes of Baerveldt
glaucoma surgeries and procedures in Medicare beneficiaries
implant placement and trabeculectomy with MMC in eyes from 1994 to 2012. Ophthalmology. 2015;122:1615e1624.
without previous ocular surgery. Similar rates of surgical 2. Chen PP, Yamamoto T, Sawada A, et al. Use of antifibrosis
success and complications were observed with both agents and glaucoma drainage devices in the American and
procedures, but trabeculectomy produced greater IOP Japanese Glaucoma Societies. J Glaucoma. 2007;16:
reduction with use of fewer glaucoma mediations than 14e19.
Baerveldt implantation after 3 years of follow-up. The dif- 3. Joshi AB, Parrish RK, Feuer WF. 2002 survey of the Amer-
ferences in results between the PTVT Study and other in- ican Glaucoma Society: practice preferences for glaucoma
vestigations comparing tube shunts and trabeculectomy with surgery and antifibrotic use. J Glaucoma. 2005;14:172e174.
MMC may relate to variations in study design, patient 4. Desai MA, Gedde SJ, Feuer WJ, et al. Practice preferences for
glaucoma surgery: a survey of the American Glaucoma Soci-
populations, disease stage, surgical technique, length of
ety in 2008. Ophthalmic Surg Lasers Imaging. 2011;42:
follow-up, and definitions of success and failure. 202e208.
The PTVT Study has several limitations. The study 5. Vinod K, Gedde SJ, Feuer WJ, et al. Practice preferences for
enrolled patients who met specific inclusion and exclusion glaucoma surgery: a survey of the American Glaucoma Soci-
criteria, and they underwent a standardized trabeculectomy ety. J Glaucoma. 2017;26:687e693.
with MMC or 350-mm2 Baerveldt glaucoma implantation. 6. Gedde SJ, Chen PP, Heuer DK, et al. The Primary Tube
Study results cannot be generalized to different patient groups Versus Trabeculectomy (PTVT) Study: methodology of a
or implant types. Although aspects of both surgical proced- multicenter randomized clinical trial comparing tube shunt
ures were standardized, some variation in technique occurred surgery and trabeculectomy with mitomycin C. Ophthal-
because surgeons were allowed sufficient latitude to perform mology. 2018;125:774e781.
7. Gedde SJ, Feuer WJ, Shi W, et al. Treatment outcomes in the
operations in a manner in which they were comfortable. There
Primary Tube Versus Trabeculectomy Study after 1 year of
were no standardized definitions or quantification of surgical follow-up. Ophthalmology. 2018;125:650e663.
complications. The low incidence of certain complications 8. Heuer DK, Barton K, Grehn F, et al. Consensus on definitions
and the small size of many patient subgroups limit the power of success. In: Shaarawy TM, Sherwood MB, Grehn F, eds.
of the study to detect significant differences. The patients and Guidelines on Design and Reporting of Glaucoma Surgical
investigators were not masked to the randomized treatment Trials. Amsterdam, The Netherlands: Kugler; 2008:15e24.
assignment, and this is a potential source of bias. The analysis 9. Lichter PR, Musch DC, Gillespie BW, et al. Interim clinical
of the 3-year PTVT outcomes data included a variety of hy- outcomes in the Collaborative Initial Glaucoma Treatment
pothesis tests and are susceptible to multiple test artifacts. The Study comparing initial treatment randomized to medications
principal outcome assessment was the comparison of cumu- or surgery. Ophthalmology. 2001;108:1943e1953.
10. The AGIS Investigators. The Advanced Glaucoma Interven-
lative failure rates between treatment groups. Because we
tion Study (AGIS): 7. The relationship between control of
previously published a 1-year comparison and plan to publish intraocular pressure and visual field deterioration. Am J Oph-
a 5-year one as well, concerned readers may wish to use 0.017 thalmol. 2000;130:429e440.
as a critical value. Interpretation of most of the comparisons of 11. Heijl A, Leske MC, Bengtsson B, et al. Reduction of intra-
secondary outcomes (e.g., IOP and medication use) do not ocular pressure and glaucoma progression: results from the
change with any reasonable P value adjustment. However, Early Manifest Glaucoma Trial. Arch Ophthalmol. 2002;120:
some of the tests resulted in P values close to the usual 0.05 1268e1279.
level of statistical significance. These may best be considered 12. Tseng VL, Kim CH, Romero PT, et al. Risk factors and long-
hypothesis generating and should be buttressed by replication term outcomes in patients with low intraocular pressure after
in other studies. trabeculectomy. Ophthalmology. 2017;124:1457e1465.
13. Saheb H, Gedde SJ, Schiffman JC, et al. Outcomes of glau-
The PTVT Study did not demonstrate clear superiority of
coma reoperations in the Tube Versus Trabeculectomy (TVT)
one glaucoma operation over the other. The rate of surgical Study. Am J Ophthalmol. 2014;157:1179e1189.
success was not significantly different between tube shunt 14. Schaefer JL, Levine MA, Martorana G, et al. Failed glaucoma
surgery and trabeculectomy with MMC at 3 years. Both drainage implant: long-term outcomes of a second glaucoma
surgical procedures were effective in lowering IOP in patients drainage device versus cyclophotocoagulation. Br J Oph-
who had not previously undergone incisional glaucoma sur- thalmol. 2015;99:1718e1724.
gery. Greater IOP reduction with use of fewer glaucoma 15. Levinson JD, Giangiacomo AL, Beck AD, et al. A comparison
medications was achieved with trabeculectomy with MMC of sequential glaucoma drainage device implantation versus
compared with tube shunt implantation. However, trabecu- cyclophotocoagulation following failure of a primary drainage
lectomy with MMC also was associated with higher rates of device. J Glaucoma. 2017;26:311e314.
16. Collaborative Normal-Tension Glaucoma Study Group.
early postoperative complications and overall complications
Comparison of glaucomatous progression between untreated
during 3 years of follow-up. Surgeon skill and experience patients with normal-tension glaucoma and patients with
with each operation is an important additional consideration therapeutically reduced intraocular pressures. Am J Oph-
when selecting a glaucoma surgical procedure but was not thalmol. 1998;126:487e497.
evaluated in this study. We plan to report 5-year results of the 17. The AGIS Investigators. The Advanced Glaucoma Interven-
trial in the future, because long-term data are needed to tion Study, 6: effect of cataract on visual field and visual
evaluate fully these traditional glaucoma operations. acuity. Arch Ophthalmol. 2000;118:1639e1652.

12
 Gedde et al 
PTVT 3-Year Outcomes

18. The AGIS Investigators. The Advanced Glaucoma Interven- 24. Investigators AGIS. The Advanced Glaucoma Intervention
tion Study: 8. Risk of cataract formation after trabeculectomy. Study (AGIS): 11. Risk factors for failure of trabeculectomy
Arch Ophthalmol. 2001;119:1771e1779. and argon laser trabeculoplasty. Am J Ophthalmol. 2002;134:
19. Hylton C, Congdon N, Friedman D, et al. Cataract after glau- 481e498.
coma filtration surgery. Am J Ophthalmol. 2003;135:231e232. 25. Wilson MR, Mendis U, Paliwal A, Haynatzka V. Long-term
20. Gedde SJ, Herndon LW, Brandt JD, et al. Postoperative follow-up of primary glaucoma surgery with Ahmed glaucoma
complications in the Tube Versus Trabeculectomy (TVT) valve implant versus trabeculectomy. Am J Ophthalmol.
Study during five years of follow-up. Am J Ophthalmol. 2003;136:464e470.
2012;153:804e814. 26. Gedde SJ, Schiffman JC, Feuer WJ, et al. Treatment outcomes
21. Budenz DL, Feuer WJ, Barton K, et al. Postoperative com- in the Tube Versus Trabeculectomy (TVT) Study after five
plications in the Ahmed Baerveldt Comparison Study during years of follow-up. Am J Ophthalmol. 2012;153:789e803.
five years of follow-up. Am J Ophthalmol. 2016;163:75e82. 27. Molteno AC, Bevin TH, Herbison P, Husni MA. Long-term
22. Christakis PG, Kalenak JW, Tsai JC, et al. The Ahmed Versus results of primary trabeculectomies and Molteno implants for
Baerveldt Study: five-year treatment outcomes. Ophthal- primary open-angle glaucoma. Arch Ophthalmol. 2011;129:
mology. 2016;123:2093e2102. 1444e1450.
23. Jampel HD, Musch DC, Gillespie BW, et al. Perioperative 28. Panarelli JF, Banitt MR, Gedde SJ, et al. A retrospective
complications of trabeculectomy in the Collaborative Initial comparison of primary Baerveldt implantation versus trabe-
Glaucoma Treatment Study (CIGTS). Am J Ophthalmol. culectomy with mitomycin C. Ophthalmology. 2016;123:
2005;140:16e22. 789e795.

Footnotes and Financial Disclosures

Originally received: July 11, 2019. HUMAN SUBJECTS: Human subjects were included in this study. Insti-
Final revision: September 4, 2019. tutional Review Board approval was obtained at each Clinical Center before
Accepted: October 1, 2019. recruitment was initiated. All research complied with the Health Insurance
Available online: ---. Manuscript no. 2019-1514. Portability and Accountability (HIPAA) Act of 1996 and adhered to the
1
Bascom Palmer Eye Institute, University of Miami Miller School of tenets of the Declaration of Helsinki. All participants provided informed
Medicine, Miami, Florida. consent.
2 No animal subjects were included in this study.
St. Thomas Hospital, London, United Kingdom.
Author Contributions:
3
Moorfields Eye Hospital, London, United Kingdom. Conception and design: Gedde, Feuer, Barton, Brandt
4
Queen Mary’s Hospital, Sidcup, United Kingdom. Analysis and interpretation: Gedde, Feuer, Lim, Barton, Goyal, Ahmed,
5
University of Toronto, Toronto, Canada. Brandt
6
Department of Ophthalmology, University of California, Davis, Sacra- Data collection: Gedde, Feuer, Lim, Barton, Goyal, Ahmed, Brandt
mento, California. Obtained funding: Feuer, Ahmed
Overall responsibility: Gedde, Feuer, Lim, Barton, Goyal, Ahmed, Brandt
Presented at: American Glaucoma Society Annual Meeting, March 2019,
San Francisco, California. Abbreviations and Acronyms:
Financial Disclosure(s): CI ¼ confidence interval; ETDRS ¼ Early Treatment Diabetic Retinopathy
The author(s) have made the following disclosure(s): W.J.F.: Financial Study; HR ¼ hazard ratio; IOP ¼ intraocular pressure;
support e Johnson & Johnson Vision. logMAR ¼ logarithm of the minimum angle of resolution;
MMC ¼ mitomycin C; PTVT ¼ Primary Tube Versus Trabeculectomy;
I.I.K.A.: Consultant e Johnson & Johnson Vision. SD ¼ standard deviation; TVT ¼ Tube versus Trabeculectomy;
Supported by Johnson & Johnson Vision, Santa Ana, California; the Na- VA ¼ visual acuity.
tional Eye Institute, National Institutes of Health, Bethesda, Maryland
(grant no.: EY014801); and Research to Prevent Blindness, Inc., New York, Correspondence:
New York. The sponsors had no role in the design or conduct of this Steven J. Gedde, MD, Bascom Palmer Eye Institute, University of Miami
Miller School of Medicine, 900 NW 17th Street, Miami, FL 33136. E-mail:
research.
[email protected].

13