DIAGNOSTIC PARS PLANA VITRECTOMY

AND AQUEOUS ANALYSES IN PATIENTS

WITH UVEITIS OF UNKNOWN CAUSE
ABDELKARIM OAHALOU, MD,* PETER A. W. J. F. SCHELLEKENS, MD,*
JOLANDA D. DE GROOT-MIJNES, PHD,*† ANIKI ROTHOVA, MD, PHD*‡

Purpose: To compare the yield of diagnostic pars plana vitrectomy (PPV) with the yield
of aqueous analyses in patients with uveitis of unknown cause.
Methods: Seventy-five consecutive patients (84 eyes) with uveitis involving posterior eye
segment who undergo a diagnostic PPV from 2005 through 2009 were retrospectively
reviewed. Vitreous specimens were simultaneously analyzed by microbiological culture,
flow cytometry, and cytology as well as by polymerase chain reaction and for intraocular
antibody production by Goldmann–Witmer coefficient. In 53 eyes, both aqueous and vit-
reous samples were assessed. The primary outcome measure was the comparison
between vitreous and aqueous analyses.
Results: Vitreous analysis was positive in 18 of 84 eyes (21%). Positive results indicated
infectious uveitis in 12 of 18 cases (67%) and lymphoma in 6 of 18 (33%) cases. Of the 53
eyes with both aqueous and vitreous samples available, aqueous analysis revealed the
diagnosis in 6 of 53 eyes and vitreous in 9 of 53 eyes. Unilateral uveitis (P = 0.022),
panuveitis and uveitis posterior (P # 0.001), preoperative immunosuppressive therapy
(P = 0.004), and increasing age (P = 0.018) were associated with an increased diagnostic
yield of PPV. Overall, 1 year after PPV, median visual acuity improved from 20/200 to 20/80
(Snellen, P # 0.001). Of 18 patients who were on immunosuppressive treatment before
PPV, 8 (44%) were able to stop immunosuppressive therapy during 1-year follow-up. The
complications of PPV consisted predominantly of cataract development (33/65, 51%).
Conclusion: Diagnostic PPV with the analysis of vitreous fluid by multiple laboratories
for infectious and malignant disorders was useful in diagnosing uveitis of unknown cause.
Previous aqueous analysis was especially valuable for the diagnosis of intraocular
infections and may therefore decrease the number of patients who would otherwise
undergo an invasive diagnostic PPV. Furthermore, PPV was associated with improved
visual acuity and decreased use of immunosuppressive therapy.
RETINA 34:108–114, 2014

D etermining the cause of uveitis has important im-

plications for its treatment and prognosis and for
an early diagnosis of associated systemic disease.
When patients with unknown uveitis do not respond
to conventional therapy, an indolent infection or malig-
nancy is often suspected. In such cases, a diagnostic
pars plana vitrectomy (PPV) can be helpful. The results
From the *Department of Ophthalmology and †Department of
Virology, University Medical Centre Utrecht, Utrecht, The Nether-
lands; and ‡Department of Ophthalmology, Erasmus Medical Cen-
tre, Rotterdam, The Netherlands.
None of the authors have any financial/conflicting interests to
disclose.
Reprint requests: Abdelkarim Oahalou, MD, University Medical
Center Utrecht, PO Box 85500, 3508 GA Utrecht, The Netherlands;
e-mail: a.oahalou@umcutrecht.nl
of a diagnostic PPV depend on the included population
and analyses performed; previously reported diagnos-
tic yields varied widely from 14.3% to 61.5%.1–8 The
less invasive aqueous analysis obtained by anterior
chamber tap is also used for the diagnosis of intraoc-
ular infections. Positive results of aqueous analysis
have been reported by several studies.9–15
Only a few studies reported on the comparison
between the polymerase chain reaction (PCR) yield of
aqueous and vitreous diagnostic assessments. Matos
et al reported on the correlation between PCR anal-
ysis of aqueous and vitreous samples in 16 patients
with intraocular inflammation and AIDS. Aqueous
PCR analysis was useful in ocular toxoplasmosis, with
a sensitivity of 75%. Polymerase chain reaction analysis

108
 DIAGNOSTIC PPV IN PATIENTS WITH UVEITIS  OAHALOU ET AL
in the vitreous was more helpful as an auxiliary diag-
nostic test in the cases of cytomegalovirus (CMV) and
acute retinal necrosis than PCR analysis of aqueous.16
In contrast, Smith et al found that CMV PCR of aqueous
and vitreous is an excellent marker for active retinitis in
patients with AIDS.17 In a study of Pathanapitoon et al,18
PCR analysis in the detection of CMV in vitreous and
aqueous samples had an equal specificity of 93% and
a sensitivity of 67% and 37%, respectively. However,
Santos et al19 reported a higher yield of vitreous PCR
compared with aqueous PCR analysis in patients with
uveitis. The purpose of this study was to compare the
yield of diagnostic PPV with the yield of aqueous anal-
yses in patients with uveitis of unknown cause in which
careful clinical and laboratory evaluation failed to iden-
tify a diagnosis.
Patients and Methods
The medical records and laboratory data of all
75 patients (84 eyes) who underwent diagnostic PPV
from 2005 through 2009 at the University Medical
Center Utrecht, Utrecht, the Netherlands, were
reviewed retrospectively. The patients for this study
were identified from laboratory records, which listed
all patients who had undergone diagnostic PPV over
this period. Excluded from the study were patients
with posttraumatic or acute postsurgical endophthal-
mitis and patients who underwent only therapeutic
PPV.
The clinical records of the patients were reviewed
for age, gender, previous vitreoretinal surgery, relevant
medical history, type of uveitis, duration of uveitis,
laterality, suspected diagnosis, previous aqueous anal-
ysis, previous uveitis screening, pre- and postoperative
immunosuppressive therapy, pre- and postoperative
visual acuity, results of vitreous analyses, final diag-
nosis, complications, and additional vitreoretinal sur-
gery. Our series included two immunosuppressed
patients (both organ transplant recipients).
This study was performed in concordance with the
declaration of Helsinki and the regulations of the local
medical ethics committee. A 20-gauge PPV was
performed according to a standard 3-port approach
with direct visualization of the instruments. At the start
of the PPV, undiluted vitreous was manually aspirated
into a 1-mL syringe connected to the vitrectomy
handpiece. To prevent hypotonia, external pressure
was given by the surgeon. The infusion was started
immediately after aspiration was completed. Vitreous
samples were delivered to the clinical laboratories
immediately after collection without waiting for the
end of surgery. After collecting the undiluted vitreous
109
samples, a complete PPV was performed. Sixteen eyes
underwent simultaneous PPV and cataract extraction.
Ten eyes received silicone oil because of severe
exudative retinal detachment or extensive acute retinal
necrosis. Aqueous humor for the analysis was obtained
by an anterior chamber puncture.
Almost all vitreous samples underwent the following
5 examinations: (1) real-time PCR analysis for Herpes
simplex virus (HSV) types 1 and 2, Varicella zoster virus
(VZV), CMV, and for the parasite Toxoplasma gondii
(T. gondii). (2) Goldmann–Witmer coefficient (GWC)
analysis was performed by determining specific antibody
titers against HSV, VZV, CMV, and T. gondii and total
immunoglobulin G concentrations in serum and vitreous
samples by enzyme-linked immunosorbent assay. The
GWC was calculated as follows [specific IgG vitreous/
total IgG vitreous]/[specific IgG serum/total IgG serum].
A GWC value of .3 was considered indicative of intra-
ocular antibody production.11 (3) Bacterial and fungal
cultures of vitreous contents were performed at the
microbiology laboratory according to the standard tech-
niques. (4) Cytological evaluation was used to identify
tumor cells suspected for malignancy. Slides were
stained with Papanicolaou stain and reviewed by an
experienced cytopathologist. (5) Part of the vitreous
(±100 mL) was stored in pure fetal calf serum for imme-
diate analysis by flow cytometry. Cells were counted and
stained for common leukocyte markers. T cells were
analyzed for CD4/CD8 and CD2/CD7 ratios, and B cells
for kappa/lambda light chain ratios according to the gen-
eral diagnostic leukemia/lymphoma panel of cell diag-
nostics at the University Medical Centre Utrecht. The
mean number of tests performed on each vitreous sample
was 4.8. In sporadic cases, specific tests were not being
performed either because of insufficient sample volume
or because of a very low pretest diagnostic probability.
Polymerase chain reaction and GWC analysis were per-
formed in 80 vitreous specimens (95%). Eighty-two
samples (98%) were analyzed by cultures. Cytological
analysis and flow cytometry were performed in 82 (98%)
and 76 samples (90%), respectively. Sixty five of the 84
vitreous samples (77%) underwent all 5 diagnostic tests,
whereas #4 tests were performed in 19 vitreous samples
(23%). Both PCR and GWC analyses were performed in
49 of 53 aqueous samples (92%).
Final diagnoses were assigned to each case on the
basis of all available information, which included the
patient’s history, previous diagnostic tests, clinical
course, response to treatment, and the vitreous analysis.
The primary outcome measure was the comparison
between vitreous and aqueous analyses. Sensitivity and
specificity of vitreous and aqueous analyses for diagnos-
ing infectious uveitis were calculated. Sensitivity was
defined as true-positive tests divided by all true-positive
110 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES  2014  VOLUME 34  NUMBER 1

and false-negative tests. Specificity was defined as true- Vitreous Analysis in 84 Eyes
negative tests divided by all true-negative and false-
Vitreous analysis was positive in 21% (18/84) of all
positive tests. The final diagnosis as mentioned before
eyes and included infectious uveitis in 12 of 18
was considered as the golden standard.
samples (67%) and lymphoma in 6 of 18 samples
Secondary outcome measures were the diagnostic
(33%). Goldmann–Witmer coefficient was the most
yield of PPV, the change in visual acuity, and
informative assay (9/80, 11% of all; 9/18, 50% of
immunosuppressive therapy at 1 year after PPV.
positive vitreous samples). Polymerase chain reaction
Surgical complications and additional vitreoretinal
analysis (5/80, 6% of all; 5/18, 28% of positive vitre-
surgery during the first year of follow-up were also
ous samples) and flow cytometry (4/76, 6% of all;
reported. Subgroup analysis was performed to identify
4/18, 22% of positive vitreous samples) were equally
patient characteristics, which predict a high diagnostic
informative. Two percent of the positive results were
yield. These characteristics included gender, age,
revealed by cytological analysis and 2% by cultures
uveitis location (anterior, intermediate, posterior, or
(both 2/82, 2% of all; 2/18, 11% of positive vitreous
panuveitis), laterality, preoperative immunosuppres-
samples).
sive therapy, and duration of uveitis till surgery.
Snellen visual acuity was converted to logarithm of
the minimum angle of resolution for the statistical Aqueous Analysis in 53 Eyes
analysis. Changes in visual acuity were analyzed with Aqueous analysis was positive in 11% (6/53).
the paired t-test. Multivariate regression analysis Goldmann–Witmer coefficient was the most informa-
(MRA) was used for the detection of predictive fac- tive assay (5/53, 9% of all; 5/6, 83% of positive aque-
tors. A value of P , 0.05 was considered statistically ous samples) compared with PCR analysis (3/53, 6%
significant. of all, 50% of positive aqueous samples).

Results Comparison of Aqueous and Vitreous Analysis in

53 Eyes With Both Samples Available
Demographic data for all 75 patients (84 eyes) are
Ten of 53 eyes exhibited positive results by aqueous
listed in Table 1. Thirty-nine patients were men and
and/or vitreous examination (Table 2). Vitreous anal-
the average age was 57.6 (range, 19–81). Previous
ysis was positive in 9 of 53 eyes (17% of all; 9/10 eyes
aqueous analysis was performed in 37 eyes, and in
with positive results by aqueous and/or vitreous sam-
16 eyes, aqueous analysis was performed together with
ples) and aqueous analysis in 6 of 53 eyes (11% of all;
vitreous analyses, resulting in 53 eyes in which both
6/10 eyes with positive results by aqueous and/or vit-
samples were assessed. In 31 eyes, the vitreous biop-
reous samples, P = 0.289, Fisher exact test). Overall, 5
sies were not accompanied by a previously or simul-
of the 10 eyes with positive results had the same re-
taneously taken aqueous sample. A total of 400
sults by both assessments (VZV, n = 2; T. gondii, n = 2
diagnostic tests were performed on the 84 vitreous
and HSV, n = 1), 4 samples were positive in vitreous
specimens.
(bacterial endophthalmitis, Candida endophthalmitis,
T. gondii, and lymphoma) and 1 in aqueous solely
(T. gondii).
Table 1. Patient Characteristics Goldmann–Witmer coefficient and PCR analysis
Patients/eyes (n) 75/84 were performed in both vitreous and aqueous samples
Mean age (year) 57.6 (range, (Table 3). Four eyes revealed positive GWC results by
19–81) both assessments (VZV, n = 2; T. gondii, n = 2), 1 eye
Male, n (%) 39 (52)
Uveitis type, n (%)
had a positive GWC result in vitreous (T. gondii) and 1
Intermediate 23 (27) eye in aqueous solely (T. gondii).
Posterior 14 (17) Three eyes revealed positive PCR results by both
Panuveitis 47 (56) assessments (VZV, HSV, and T. gondii) and 2 eyes
Unilateral involvement, n (%) 33 (35) had positive PCR results in vitreous (T. gondii, n = 2)
Previous aqueous analysis, n (%) 37 (44)
alone.
Aqueous analysis during PPV, n (%) 16 (19)
Previous standard uveitis screening, n (%) 60 (80) Sensitivity and specificity of aqueous analysis for
Preoperative IS therapy, n (%) 18 (24) diagnosing infectious uveitis were 86% and 100%,
Preoperative mean VA (Snellen) 20/200 respectively. If vitreous analysis instead of final diag-
Follow-up (median months) 11.6 nosis was considered as the golden standard sensitivity
IS, immunosuppressive; VA, visual acuity. and specificity of aqueous analysis decreased to 71%
 DIAGNOSTIC PPV IN PATIENTS WITH UVEITIS  OAHALOU ET AL
Table 2. Yield of Vitreous and Aqueous Analysis in 53 Eyes in Which Both Samples Were Analyzed
Eyes (n)
Vitreous 53
Aqueous 53
Both vitreous and aqueous 53
*These 15 positive results were found in 10 eyes: 5 eyes had the same results by both assessments, 4 eyes had only positive vitreous
results, and 1 eye showed only a positive aqueous result.
and 98%, respectively. Sensitivity and specificity of
vitreous analysis for diagnosing infectious uveitis were
100%.
Forty four of the 53 vitreous samples (83%)
underwent all 5 diagnostic tests. Missing tests were
cytometry (n = 8), cytology (n = 1), GWC (n = 2), and
PCR (n = 2). Only 4 of the 53 (8%) aqueous samples
had missing PCR and GWC tests.
Yield Predictors
Overall, patients with unilateral uveitis (P = 0.022,
MRA), panuveitis and uveitis posterior (P # 0.001,
MRA), preoperative immunosuppressive therapy (P =
0.004, MRA), and increasing age (P = 0.018, MRA)
were associated with an increased diagnostic yield of
PPV. Of the 18 positive results, no predictors could be
identified for the eyes with positive results for infec-
tion (n = 12). Vitreous analysis was positive in 29% of
the eyes (9/31) without previous aqueous examination.
The diagnostic yield of PPV was lower in eyes with
previous negative aqueous analysis (3/32, 9%; P =
0.046, Fisher exact test). These vitreous samples were
positive for lymphoma, endogenous bacterial endoph-
thalmitis, and T. gondii infection.
Final Diagnosis
Based on the combination of all diagnostic proce-
dures available, which included patient’s history, pre-
vious diagnostic tests, clinical course, response to
treatment, and the vitreous analysis, a final diagnosis
was obtained in 25 of 84 eyes (30%).
Table 3. Yield of PCR and GWC Analysis in 53 Eyes in
Which Both Samples Were Analyzed
Eyes with PCR and/or
Positive Test PCR+ GWC+ GWC+
Eyes with vitreous+ 5 5 6 tinal membrane was seen in six eyes; three of these
Eyes with aqueous + 3 5 6 underwent additional peeling of the limiting internal
Three eyes revealed positive PCR results by both assessments
(VZV, HSV, and T. gondii) and 2 eyes had positive PCR results in
vitreous (T. gondii, n = 2) alone. Four eyes revealed positive GWC
results by both assessments (VZV, n = 2; T. gondii, n = 2), 1 eye
had a positive GWC result in vitreous (T. gondii), and 1 eye in
aqueous solely (T. gondii).
111
Positive Results Infections Malignancy
9/53 8/9 1/9
6/53 6/6 0/6
15/53* 14/15 1/15
In the presence of other negative tests, vitreous
analysis established a final diagnosis in 18% of the
eyes (15/84 of all, 15/25, 60% of those with estab-
lished cause). In 10 additional cases, negative vitreous
examination in combination with additional tests
revealed the final diagnosis. These cases included lym-
phoma (n = 2), sarcoid-associated uveitis (n = 4), bird-
shot chorioretinopathy (n = 1), Behçet disease (n = 1),
and phacogenic uveitis (n = 1).
Visual Outcomes
Preoperative and postoperative visual acuities are
given in Table 4. The overall mean preoperative vision
was 20/200 (Snellen). Visual acuity improved from
20/200 to 20/80 after 1 year (Snellen, P # 0.001,
paired t-test).
Immunosuppressive Therapy
Eighteen of the 75 patients (24%) used immuno-
suppressive therapy before surgery. During the follow-
up of 1 year, 8 of these 18 patients (44%) were able to
stop immunosuppressive therapy. Therapy could be
reduced in 3 patients (17%). In 7 patients (39%),
therapy could not be stopped or reduced during
follow-up.
Complications
Complications because of the diagnostic PPV after 1
year of follow-up are listed in Table 5. Overall, com-
plications occurred in 45 of 84 eyes (54%). Thirty
three of the 65 preoperative phakic eyes (51%) devel-
oped cataract of which 23 underwent subsequent cat-
aract extraction with intraocular lens implantation.
Retinal detachment occurred in two eyes of which
one was repaired without incident (the other patient
refused any surgical therapy). Formation of an epire-
membrane. Decreased intraocular pressure (,4 mmHg)
was seen in 2 eyes. One eye recovered after an addi-
tional PPV with peeling of the limiting internal mem-
brane because of a macular epiretinal membrane
formation. The second eye had leakage from one of
112 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES  2014  VOLUME 34  NUMBER 1

Table 4. Visual Outcomes (Snellen) of Patients Who Underwent Diagnostic Vitrectomy

Number of Eyes Mean Preoperative Visual Acuity (n) Mean 1 Year Postoperative VA (n)
Total 84 20/200 (84) 20/80 (52); P # 0.001

the sclerotomies and recovered spontaneously. Vitre-
ous hemorrhage in 2 eyes dissolved spontaneously.
Besides the aforementioned procedures, one eye had
an additional diagnostic PPV because of an insuffi-
cient amount of the first vitreous specimen to perform
all diagnostic procedures. In six eyes, silicone oil was
removed and one eye underwent vitreoretinal surgery
with limiting internal membrane peeling for persistent
cystoid macular edema. In total, additional vitreoreti-
nal surgery was needed in 13 of 84 eyes (15%).
Discussion
Our study shows that the diagnostic PPV in patients
with severe uveitis with the vitreous specimens
examined in multiple laboratories (including cytolog-
ical, microbiological, and uveitis assessments) is
a procedure that identifies a cause of uveitis in 21%
of the eyes. In patients with previous negative analysis
of aqueous, the diagnostic value of PPV decreased to
9% that can be explained by ruling out of infections by
aqueous analyses.
The primary aim of diagnostic PPV is to obtain
qualitatively and quantitatively adequate amounts of
material for identifying uveitis of unknown origin.
Previous studies reporting on the diagnostic yield of
PPV showed that analysis of vitreous fluid is useful in
identifying intraocular infection or malignancy.1–8 The
vitreous yield was similar to that reported by Margolis
et al7 (20%) but much lower than that reported by
Davis et al6 (61.5%). The large range of positive
results is certainly influenced by the selection of pa-
tients and analyses performed. In studies that selected
particularly cases suspected for infectious uveitis and/
or lymphoma, a higher diagnostic yield was to be
expected. Our study contains mainly eyes of patients
with severe chronic uveitis in whom extensive screen-
ing (and in some cases aqueous analyses) was already
performed, which may explain the relatively low diag-
nostic yield compared with previously published
studies. The differences in previous diagnostic exami-
nations make direct comparison of various studies dif-
ficult. The aqueous yield was lower than that reported
by several other studies.9−15
Only a few studies reported on the comparison
between the yield of vitreous and aqueous ana-
lyses.16–19 Our aim was not to compare diagnostic
values of PPV with aqueous analyses (this would be
very difficult because not all cases with suspected
infectious uveitis undergo diagnostic PPV) but to eval-
uate the spectrum of vitreous and aqueous results in
patients with both assessments available. Mostly, the
analysis of aqueous fluid addresses the eventual infec-
tious causes by PCR and GWC examinations. In the
present series, of the 53 eyes with both aqueous and
vitreous samples available, aqueous analysis revealed
the diagnosis in 6 of 53 eyes and vitreous in 9 of 53
eyes. Five of the 10 eyes with positive results had the
same results by both assessments (VZV, n = 2;
T. gondii, n = 2; and HSV, n = 1). These results show
that aqueous analyses might establish a diagnosis of
uveitis (infectious) in a considerable percentage of
patients (in the present series 6/10, 60%) without per-
forming an invasive diagnostic PPV. Goldmann–
Witmer coefficient and PCR analysis was performed
in both vitreous and aqueous samples (Table 5).
In 1 eye with positive GWC and PCR results in
vitreous solely, the associated aqueous sample analyzed
was obtained in our laboratory after a delay of 1 week
after sampling, which might have caused the negative
aqueous result. In another eye with a positive PCR result
in vitreous alone, aqueous analysis revealed a positive
GWC. These findings suggest that both PCR and GWC
determination might be performed for comprehensive
diagnosis of intraocular infections and support the
usefulness of aqueous analyses.
In the previous studies including that of Davis et al,
the choice of diagnostic tests was mainly based on
clinical suspicion.4,6,7 In our series, however, we did
not choose one single test to be performed based on
the pretest likelihood but combined five types of tests
performed on vitreous samples to obtain maximum
diagnostic value of PPV. Therefore, the limited vol-
ume of specimens used for specific tests may have had

Table 5. Complications in Patients Who Underwent Diagnostic Vitrectomy

Number of Eyes Cataract, N (%) RD, N (%) ERM, N (%) VH, N (%) IOP (,5 mmHg), N (%) Total, N (%)
Total 84 33/65 (51)* 2/84 (2) 6/84 (7) 2/84 (2) 2/84 (2) 45/84 (54)
*Thirty-three cataracts among 65 phakic patients (51%).
RD, retinal detachment; ERM, epiretinal membrane; VH, vitreous hemorrhage; IOP, intraocular pressure.
 DIAGNOSTIC PPV IN PATIENTS WITH UVEITIS  OAHALOU ET AL
a negative effect on our results (e.g., in lymphoma
where the maximum volume might have been associ-
ated with the higher percentage of positive results).
Preoperative immunosuppressive therapy (P =
0.004, MRA) was associated with an increased diag-
nostic yield. It is feasible that the group of patients
with insufficient therapeutical effect of immunosup-
pressive treatments harbor the cases of not diagnosed
infections or malignancies. Unilateral uveitis (P =
0.022, MRA), panuveitis and uveitis posterior (P #
0.001, MRA), and increasing age (P = 0.018, MRA)
were also associated with an increased diagnostic yield
of PPV.
A secondary aim was to ascertain the visual out-
comes of patients and the change in immunosuppres-
sive therapy after diagnostic PPV. A beneficial effect
of PPV on visual acuity was previously reported by
others.3,7,20,21 Visual acuity and inflammatory activity
showed a slight improvement after PPV. Visual acuity
improved from 20/200 to 20/80 after 1 year (Snellen,
P # 0.001, paired t-test). The change in visual acuity
is not only influenced by the diagnostic intervention
but may represent a result of many other factors such
as the cause and location of disease, lens status, and
postvitrectomy treatment. During the follow-up, 8 of
18 patients (44%) were able to stop immunosuppres-
sive therapy and therapy could be reduced in 3 of 18
patients (16.7%). This beneficial effect is similar to the
results of therapeutic vitrectomy.20,22,23
Next to its beneficial effects, diagnostic PPV has
some potential vision compromising risks.22,24 Com-
plication rates in this study were similar to that
reported by other studies.6,22,24 These complications
might have been influenced in part by intraocular
inflammation and additional surgical procedures (sili-
cone oil) in cases with severe exudative retinal detach-
ment or extensive acute retinal necrosis. Almost all
complications were manageable.
The limitations of our study include its retrospective
design, which is less accurate than the prospective
studies. One concern with the 5-test battery is the
viable yield of a part of the sample used for specific
examinations. Using small amounts of vitreous might
have had induced false-negative results. As described
before, there are some missing tests that could
introduce bias in our results. Since missing tests
consisted mainly of flow cytometry analysis (essential
for lymphoma), and most of the missing tests had
a very low pretest diagnostic probability based on
clinical suspicion, we believe that any bias would be
of a minimal level.
Since smaller gauge instruments were not yet
available, we used a 20-gauge PPV for vitreoretinal
surgery. Smaller gauge instruments may shorten
113
operating time, improve patient comfort, and speed
visual recovery. However, no significant differences in
complication rates were reported.25,26 A smaller gauge
PPV does not protect against cataract, which in our
study was the main complication.27
As in other studies, the criteria for diagnostic PPV in
uveitis were not clearly established. In our institution,
we perform the aqueous analysis as a second step of
diagnostic process in uveitis, preceded by general
screening examinations. In the case of negative
aqueous analysis and if uveitis is severe and chronic,
we proceed to diagnostic PPV as a third step.
However, we perform primary diagnostic PPV in very
severe cases when retinal evaluation is not possible
and in those with a strong suspicion of malignancy. A
PPV provides a large volume of sample and may
improve visual acuity. However, the drawback is the
invasiveness of the procedure with its potential
adverse effects. Aqueous analysis is less invasive
and may yield results faster. However, it provides
a small volume of the sample and therefore a limited
number of examinations are possible. Even with the
limited number of patients included, we have clearly
shown that performing aqueous analysis before PPV is
especially valuable for the diagnosis of intraocular
infections and may therefore decrease the number of
patients who would otherwise undergo diagnostic
PPV.
In conclusion, we showed that diagnostic PPV with
analysis of vitreous fluid by multiple tests is useful in
diagnosing uveitis of unknown cause. During the
diagnostic process for uveitis, aqueous analysis might
establish a diagnosis of (infectious) uveitis in a con-
siderable percentage of patients and makes the inva-
sive diagnostic PPV less frequently necessary.
Diagnostic PPV is mainly valuable for diagnosing
lymphoma and other disorders not assessed by aque-
ous examinations. Moreover, diagnostic PPV was
associated with improved visual acuity and decreased
use of immunosuppressive therapy.
Key words: diagnostic vitrectomy, aqueous analy-
sis, uveitis, visual acuity.
References
1. Priem H, Verbraeken H, De Laey JJ. Diagnostic problems in
chronic vitreous inflammation. Graefes Arch Clin Exp Oph-
thalmol 1993;231:453–456.
2. Palexas GN, Green WR, Goldberg MF, Ding Y. Diagnostic
pars plana vitrectomy report of a 21-year retrospective study.
Trans Am Ophtalmol Soc 1995;93:281–308.
3. Verbraeken H. Diagnostic vitrectomy and chronic uveitis.
Graefes Arch Clin Exp Ophtalmol 1996;234:S2–S7.
4. Mruthyunjaya P, Jumper JM, McCallum R, et al. Diagnostic
yield of vitrectomy in eyes with suspected posterior segment
114 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES  2014  VOLUME 34  NUMBER 1
infection or malignancy. Ophthalmology 2002;109:1123–
1129.
5. Coupland SE, Bechrakis NE, Anastassiou G, et al. Evaluation
of vitrectomy specimens and chorioretinal biopsies in the diag-
nosis of primary intraocular lymphoma in patients with mas-
querade syndrome. Graefes Arch Clin Exp Ophtalmol 2003;
241:860–870.
6. Davis JL, Miller DM, Ruiz P. Diagnostic testing of vitrectomy
specimens. Am J Ophtalmol 2005;140:822–829.
7. Margolis R, Brasil OF, Lowder CY, et al. Vitrectomy for the
diagnosis and management of uveitis of unknown cause. Oph-
thalmology 2007;114:1893–1897.
8. Margolis R. Diagnostic vitrectomy for the diagnosis and man-
agement of posterior uveitis of unknown etiology. Curr Opin
Ophtalmol 2008;19:218–224.
9. Haut J, Roman S, Morin Y, et al. Search for etiology in 110
cases of uveitis: value of punctures of the aqueous humor and
vitreous body [French]. J Fr Ophtalmol 1995;18:292–304.
10. De Boer JH, Verhagen C, Bruinenberg M, et al. Serologic and
polymerase chain reaction analysis of intraocular fluids in the
diagnosis of infectious uveïtis. Am J Ophthalmol 1996;121:
650–658.
11. Liekfeld A, Schweig F, Jaeckel C, et al. Intraocular antibody
production in intraocular inflammation. Graefes Arch Clin Exp
Ophthalmol 2000;238:222–227.
12. Vasseneix C, Bodaghi B, Muraine M, et al. Intraocular fluids
analysis for etiologic diagnosis of presumed infectious uveitis.
J Fr Ophtalmol 2006;29:398–403.
13. De Groot-Mijnes JDF, Rothova A, van Loon AM, et al. Poly-
merase chain reaction and Goldmann-Witmer coefficient anal-
ysis are complimentary for the diagnosis of infectious uveitis.
Am J Ophthalmol 2006;141:313–318.
14. Rothova A, de Boer JH, ten Dam-van Loon NH, et al. Useful-
ness of aqueous humor analysis for the diagnosis of posterior
uveitis. Ophthalmology 2008;115:306–311.
15. Harper TW, Miller D, Schiffman JC, Davis JL. Polymerase
chain reaction analysis of aqueous and vitreous specimens in
the diagnosis of posterior segment infectious uveitis. Am J
Ophthalmol 2009;147:140–147.
16. Matos K, Muccioli C, Belfort Junior R, et al. Correlation
between clinical diagnosis and PCR analysis of serum, aque-
ous, and vitreous samples in patients with inflammatory eye
disease. Arq Bras Oftalmol 2007;70:109–114.
17. Smith IL, Macdonald JC, Freeman WR, et al. Cytomegalovirus
(CMV) retinitis activity is accurately reflected by the presence
and level of CMV DNA in aqueous humor and vitreous.
J Infect Dis 1999;179:1249–1253.
18. Pathanapitoon K, Ausayakhun S, Kunavisarut P, et al. Detec-
tion of cytomegalovirus in vitreous, aqueous and conjunctiva
by polymerase chain reaction (PCR). J Med Assoc Thai 2005;
88:228–232.
19. Santos FF, Commodaro AG, Souza AV, et al. Real-time PCR
in infectious uveitis as an alternative diagnosis. Arq Bras Of-
talmol Arq Bras Oftalmol 2011;74:258–261.
20. Stavrou P, Baltatzis S, Letko E, et al. Pars plana vitrectomy in
patients with intermediate uveitis. Ocul Immunol Inflamm
2001;9:141–151.
21. Tranos P, Scott R, Zambarajki H, et al. The effect of pars plana
vitrectomy on cystoid macular oedema associated with chronic
uveitis: a randomised, controlled pilot study. Br J Ophthalmol
2006;90:1107–1110.
22. Verbraeken H. Therapeutic pars plana vitrectomy for chronic
uveitis: a retrospective study of the long-term results. Graefes
Arch Clin Exp Ophtalmol 1996;234:288–293.
23. Wiechens B, Nölle B, Reichelt JA. Pars-plana vitrectomy in
cystoid macular edema associated with intermediate uveitis.
Graefes Arch Clin Exp Ophtalmol 2001;239:474–481.
24. Heiligenhaus A, Bornfeld N, Foerster MH, Wessing A. Long-
term results of pars plana vitrectomy in the management of
complicated uveitis. Br J Ophthalmol 1994;78:549–554.
25. Williams GA. 25-, 23-, or 20-gauge instrumentation for vitre-
ous surgery? Eye 2008;22:1263–1266.
26. Recchia FM, Scott IU, Brown GC, et al. Small-gauge pars
plana vitrectomy: a report by the American Academy of Oph-
thalmology. Ohthalmology 2010;117:1851–1857.
27. Almony A, Holekamp NM, Bai F, et al. Small-gauge vitrec-
tomy does not protect against nuclear sclerotic cataract. Retina
2012;32:499–505.