The Prognostic Value of Persistent Culture
Positivity in Fungal Keratitis in the Mycotic
Antimicrobial Localized Injection Trial
JULIA PICKEL, SHIVANANDA NARAYANA, TIRUVENGADA KRISHNAN, SEEMA RAMAKRISHNAN,
PUJA PRATIVA SAMANTARAY, TRAVIS C. PORCO, TRAVIS REDD, THOMAS M. LIETMAN, AND
JENNIFER ROSE-NUSSBAUMER

PURPOSE: To evaluate the utility of repeat cultures at
days 3 and 7 after starting antifungal medications for
predicting outcomes in fungal keratitis.

DESIGN: Prespecified secondary analysis of the ran-
domized clinical Mycotic Antimicrobial Localized Injec-
tion trial.

METHODS: Patients presenting to Aravind Eye Hospi-
tal, Pondicherry, India, with fungal keratitis and visual
acuity worse than 20/70 received topical natamycin and
were randomized to either receive intrastromal injection
of voriconazole or topical therapy alone. All subjects
received corneal cultures at date of presentation, day
3, and day 7. Outcome measures included 3-week and
3-month visual acuity and scar size, corneal perforation,
and/or the need for therapeutic penetrating keratoplasty
(TPK). Visual acuity and scar size were analyzed with
multiple linear regression controlling for baseline mea-
sures. Survival analysis was used to analyze the risk
of corneal perforation and/or need for TPK.

RESULTS: Of the 70 study subjects with fungal kera-
titis, 25 of 69 (36%) remained culture positive at day 3,
and 20 of 62 (32%) were culture positive at day 7. Cul-
ture positivity at day 3 conferred a hazard ratio of 2.8
for requiring TPK (P [ .03) but was not a statistically
significant predictor of perforation, scar size, or final vi-
sual acuity. Culture positivity at day 7 had a hazard ratio
of 3.5 for requiring TPK (P [ .003). Those with positive
cultures at day 7 had on average 3 logMAR lines worse vi-
sual acuity at 3 months (95% confidence interval 0.9 to
5.2 logMAR lines, P [ .006) and 1.1 mm larger scar
Accepted for publication Feb 24, 2020.
From the Francis I. Proctor Foundation (J.P., T.C.P., T.R., T.M.L., J.R.-
N.), University of California San Francisco, San Francisco, California,
USA; Aravind Eye Hospital, Pondicherry, India (S.N., T.K., S.R.,
P.P.S.); UCSF Epidemiology and Biostatistics (T.C.P., T.M.L.),
University of California San Francisco, San Francisco, California, USA;
UCSF Department of Ophthalmology (T.R., T.M.L., J.R.-N.),
University of California San Francisco, San Francisco, California, USA;
and The Permanente Medical Group, Redwood City, California, USA
(J.R.-N.).
Inquiries to Jennifer Rose-Nussbaumer, UCSF/Proctor Foundation,
Division of Research, The Permanent Medical Group, 513 Parnassus
Ave, S347, San Francisco, CA 94143, USA; e-mail: jennifer.
rose-nussbaumer@ucsf.edu
0002-9394/$36.00 © 2020 ELSEVIER INC. ALL
https://doi.org/10.1016/j.ajo.2020.02.021
size at 3 months after controlling for baseline measures
(95% confidence interval 0.1 to 2.2 mm; P [ .03).

CONCLUSIONS: While not as predictive as day 7 cul-
tures, culture positivity at day 3 after starting treatment
is a significant predictor of the need for TPK in patients
with moderate-to-severe filamentous fungal keratitis.
This has applications for risk stratification, and may facil-
itate earlier consideration of TPK in high-risk
patients. (Am J Ophthalmol 2020;215:1–7. Ó 2020
Elsevier Inc. All rights reserved.)
F
UNGAL KERATITIS REPRESENTS A SIGNIFICANT PUB-
lic health challenge because, compared to bacterial
etiologies, it is particularly difficult to treat and typi-
cally results in worse visual outcomes.1 The Mycotic Ulcer
Treatment Trial (MUTT) established topical natamycin as
the standard of care for managing filamentous fungal kera-
titis. However, many cases remain recalcitrant and continue
to progress despite therapy, resulting in significant visual
morbidity, perforation, and need for surgical intervention.2–
4
It has been theorized that earlier surgical intervention in
recalcitrant cases, before the infection has progressed to the
point of perforation or extension to the limbus or sclera,
may result in better patient outcomes and allow improved
resource allocation in resource-poor settings.5,6 In light of
this, novel diagnostic tools to identify recalcitrant cases as
early as possible are of paramount importance.
Recent work has suggested repeat cultures after initia-
tion of treatment may have excellent prognostic value.
Secondary analyses of the MUTT I and MUTT II trials
identified persistent culture positivity at day 6 after initia-
tion of appropriate medical therapy as a significant predic-
tor of perforation, need for therapeutic penetrating
keratoplasty (TPK), and worse anatomic and visual out-
comes.7,8 However, given the rapid worsening in many
cases as well as the long incubation period of filamentous
fungi in culture, even earlier identification of persistent
culture positivity would be beneficial.
The Mycotic Antimicrobial Localized Injection Trial
(MALIN) found no benefit to adjuvant intrastromal vori-
conazole in addition to topical natamycin in the treatment
of filamentous fungal keratitis in regard to repeat culture
positivity, week 3 or month 3 visual acuity, rate of corneal
RIGHTS RESERVED. 1
perforation, or need for TPK.9 Herein we perform a second-
ary analysis of these patients to evaluate the utility of
persistent culture positivity at day 3 and day 7 in predicting
anatomic and functional outcomes in patients with fungal
keratitis.
METHODS

STUDY PARTICIPANTS: The MALIN was a randomized,
outcome-masked, 2-arm clinical trial comparing the use of
natamycin vs natamycin and intrastromal voriconazole in
the treatment of filamentous fungal corneal ulcers.9 Inclusion
criteria were a presence of smear-positive moderate-to-severe
filamentous fungal corneal ulcer, visual acuity worse than 20/
70 in the affected eye, age between 19 and 80 years, patient
basic understanding of the study and willingness to partici-
pate as determined by the treating physician, and commit-
ment to return for follow-up visits. Patient exclusion
criteria included evidence of a concomitant infection, history
of corneal transplant or recent ophthalmic surgery, lack of
light perception in the affected eye, visual acuity worse
than 20/200 in the unaffected eye, and the presence of cogni-
tive impairment. Written informed consent was obtained
from all participants. All study physicians performing corneal
ulcer scraping and follow-up examinations, microbiologists,
and refractionists were masked to patient study intervention.
Ethical approval for the study was granted from the Univer-
sity of California San Francisco Institutional Review Board
and the Aravind Eye Care System Institutional Review
Board. The trial was registered with clinicaltrials.gov
(NCT02731638). The trial conformed to the ethical guide-
lines outlined in the Declaration of Helsinki.

INTERVENTION: All study visits took place at Aravind
Eye Hospital in Pondicherry, India. Patients reported to
Aravind Eye Hospital for baseline, day 2 (time of intrastro-
mal voriconazole injection in the group randomized to this
intervention), day 3, day 5, day 7, month 1, and month 3
study visits. All patients began 5% natamycin topical ther-
apy hourly while awake beginning at the baseline visit, in
addition to prophylactic topical moxifloxacin 0.5% (Auro-
lab, Madurai, India) every 2 hours while awake and topical
homatropine 2% (Aurolab) 3 times daily. Thirty-five pa-
tients were randomized to receive intrastromal voricona-
zole (Aurolab) and 5% topical natamycin (Aurolab) and
35 were randomized to receive 5% topical natamycin
only, using R Statistical Package, Version 2.12 (R Founda-
tion for Statistical Computing, Vienna, Austria). Patients
randomized to intrastromal voriconazole received an injec-
tion at the baseline, day 3, and day 5 visits. The voricona-
zole injection was prepared by reconstituting white freeze-
dried/lyophilized voriconazole (Aurolab) with lactated
Ringer’s solution (Aurolab) to form a 0.5 mg/mL solution.
A total of 0.1 mL voriconazole solution was injected at the
2 AMERICAN JOURNAL OF OPHTHALMOLOGY
mid-stromal level of the cornea using a 1 mL tuberculin sy-
ringe with a 30 gauge needle (Aurolab) for patients ran-
domized to the intrastromal voriconazole and 5% topical
natamycin study arm.

MICROBIOLOGY METHODS: Corneal ulcers were scraped
and cultured at the baseline visit. Repeat scraping and cul-
ture was performed at the day 3 and day 7 study visits. A
0.5% solution of tetracaine topical anesthetic (Aurolab)
was applied to the affected eye before corneal scrapings
and culture were performed. A sterilized Kimura spatula
was used to scrape from the leading edge to the base of
the ulcer. Two scrapings from the affected eye were directly
smeared on glass slides for Gram staining and potassium hy-
droxide mount. Two further scrapings were directly inocu-
lated on sheep’s blood agar and potato dextrose agar for
fungal culture. Fungal cultures were checked daily for
growth. Cultures were considered positive if there was
any growth on 2 of the 3 media or if there was moderate-
to-heavy growth on at least 1 medium.

OUTCOME ASSESSMENT: The primary outcome in this
secondary analysis was the need for TPK. Other outcomes
included perforation, final visual acuity, scar size, and
epithelial defect size. Best spectacle-corrected visual acuity
was measured at 4 meters at the baseline, month 1, and
month 3 study visits using the protocol from the Age-
Related Eye Disease Study using Early Treatment Diabetic
Retinopathy Study tumbling E charts (Charts 2305 and
2305A; Precision Vision, Woodstock, Illinois, USA).10
The infiltrate/scar size and epithelial defect size of the ul-
cer were defined as the geometric mean of the longest diam-
eter and the maximum distance perpendicular to this
diameter. Baseline infiltrate depth was used as a measure
of ulcer severity. Infiltrate/scar depth was categorized as
anterior third, middle third, or posterior third based on
extension into the corneal stroma, with posterior third
involvement considered the most severe.

STATISTICAL ANALYSIS: The likelihood of persistent
culture positivity among different fungal species was
compared using a 2-sample test of proportions. Survival
analysis evaluating risk of TPK and perforation were
performed using Kaplan-Meier plots and multivariate Cox
hazard ratio (HR) models, controlling for baseline culture
status. Fungal species and baseline ulcer depth were not sig-
nificant predictors in the model. All analyses were conduct-
ed with Stata version 13 (StataCorp, College Station,
Texas, USA) and R (R Foundation, Vienna, Austria). A
2-tailed alpha of 0.05 was considered statistically significant.
RESULTS

DEMOGRAPHICS AND MICROBIOLOGIC RESULTS:
Seventy patients with smear-positive fungal keratitis
JULY 2020
 TABLE 1. Baseline Characteristics According to Repeat Culture Status at Days 3 and 7 After Diagnosis of Filamentous Fungal Keratitis

Day 3 Culture Status Day 7 Culture Status

Baseline Characteristic Positive (n ¼ 25) Negative (n ¼ 44) Positive (n ¼ 20) Negative (n ¼ 42)

Sex (female) 9 (36%) 17 (39%) 8 (40%) 14 (33%)

Age (years) 52.5 6 14.5 51.3 6 10.5 52.4 6 13.0 53.0 6 13.0
Occupation
Agricultural 7 (28%) 18 (41%) 4 (20%) 19 (45%)
Other 18 (72%) 26 (59%) 16 (80%) 23 (55%)
Ulcer location
Central 22 (88%) 39 (89%) 19 (95%) 36 (86%)
Peripheral 3 (12%) 5 (11%) 1 (5%) 6 (14%)
Hypopyon
Yes 19 (76%) 21 (48%) 13 (65%) 25 (60%)
No 6 (24%) 23 (52%) 7 (35%) 17 (40%)
Infiltrate depth
0-33% 9 (36%) 13 (30%) 8 (40%) 13 (31%)
>
_33%-67% 7 (28%) 20 (46%) 7 (35%) 16 (38%)
>
_67%-100% 9 (36%) 11 (25%) 5 (25%) 13 (31%)
Infiltrate diameter (mm) 5.98 6 1.51 5.51 6 1.61 4.87 6 1.53 4.51 6 1.49
Epithelial defect diameter (mm) 5.27 6 1.55 4.80 6 1.25 5.46 6 1.53 4.88 6 1.24

Values for continuous variables are displayed as mean 6 standard deviation.

were included in the study, with a mean 6 standard de-
viation (SD) age of participants of 52 6 13 years and a
range from 19 to 80 years. Forty-four subjects (63%)
were male. Patients with evidence of concomitant bac-
terial infection on smear were ineligible for the study.
Subsequent bacterial cultures were negative for 70 of
70 patients (100%) at baseline, 69 of 69 (100%) at day
3, and 61 of 62 (98%) at day 7. Baseline fungal cultures
were positive for 57 of the 70 participants (81%). Of
those that were culture positive at baseline, 19 grew
Fusarium species (33%), 17 grew Aspergillus species
(30%), and 21 grew other filamentous fungal species
(37%). Twenty-five of 69 subjects remained culture pos-
itive at day 3 (36%). Of these cultures, 3 grew Fusarium
(12%), 13 grew Aspergillus (52%), and 9 grew other fila-
mentous species (36%). Twenty of 65 subjects were
persistently culture-positive at day 7 (30%). Of these,
2 grew Fusarium (10%), 12 grew Aspergillus (60%), and
6 grew other filamentous species (30%). Two of the 19
Fusarium ulcers (11%) remained culture positive at day
7, compared to 12 of the 17 Aspergillus ulcers (71%).
Aspergillus ulcers were significantly more likely to
remain culture positive through day 7 compared to Fusa-
rium ulcers (P < .001). The average time to culture pos-
itivity was 5.8 (3.9) days for baseline cultures, 5.0 (2.6)
days for day 3 cultures, and 4.4 (2.0) days for day 7 cul-
tures. A comparison of baseline characteristics between
subjects based on culture status at days 3 and 7 of exam-
ination is provided in Table 1.

THERAPEUTIC PENETRATING KERATOPLASTY AND
PERFORATION: A total of 28 of the 70 subjects (40%) un-
derwent TPK during the observation period for lack of
response to medical therapy with impending perforation
or limbal involvement. Among these subjects, the median
time to TPK was 30 days, with a range of 3 to 141 days.
Adjusting for baseline culture positivity, subjects with
persistent culture positivity at day 3 were significantly
more likely to require subsequent TPK (HR 2.8, P ¼
.026). This association remained statistically significant
when adjusting for deeper baseline infiltrate depth and
Aspergillus ulcers, each of which were associated with
increased risk of TPK on univariate analysis. The Kaplan-
Meier survival curves for TPK according to day 3 culture
status are depicted in Figure 1. Adjusting for baseline cul-
ture positivity, persistent culture positivity at day 7 was
also strongly associated with subsequent TPK (HR 3.5,
P ¼ .003). The Kaplan-Meier survival curves for TPK ac-
cording to day 7 culture status are depicted in Figure 2.
Six of the 70 subjects (9%) developed corneal perfora-
tion. Among these subjects, the median time to perforation
was 85 days, with a range of 4 to 141 days. Persistent culture
positivity at day 3 was not significantly associated with risk
of perforation (HR 1.1, P ¼ .92), nor was culture positivity
at day 7 (HR 2.6, P ¼ .26). The majority of corneal ulcers
remained smear positive at day 3 (86%; 59/69), and this was
not predictive of need for TPK (HR 0.63, 95% CI 0.2-1.7,
P ¼ .36) or corneal perforation (HR 0.65, 95% CI 0.1-5.9,
P ¼ .7).

VISUAL AND ANATOMIC OUTCOMES: Mean visual acu-
ity at baseline was 1.59 6 0.35 log MAR. At 3 months
follow-up, mean visual acuity was 1.28 6 0.63 log
MAR. Persistent culture positivity at day 3 was not a

VOL. 215 THE PROGNOSTIC VALUE OF REPEAT CULTURE IN FUNGAL KERATITIS 3

FIGURE 1. Kaplan-Meier survival curve comparing the probability of eventuating to therapeutic penetrating keratoplasty based on
3-day culture status.
significant predictor of best-corrected visual acuity at
3 weeks (P ¼ .63) or 3 months (P ¼ .69) after presentation
(Table 2). However, day 7 culture positivity was associ-
ated with significantly worse visual acuity at both week
3 (P ¼ .003) and month 3 (P ¼ .006) (Table 2). After
adjusting for baseline visual acuity, on average subjects
with persistent culture positivity at day 7 had 3 logMAR
lines worse visual acuity compared to those who were cul-
ture negative (95% CI 0.9 to 5.2 logMAR lines, P ¼ .006).
With respect to corneal scar size at 3 months, persistent
culture positivity at day 3 was not a significant predictor of
final scar size (multivariate linear regression adjusting for
baseline infiltrate size, P ¼ .55). After adjusting for base-
line infiltrate size, subjects who remained culture positive
at day 7 had an average final scar size 1.1 mm greater than
those without (95% CI 0.1-2.2 mm; P ¼ .03).
DISCUSSION
THIS PRESPECIFIED SECONDARY ANALYSIS OF A CLINICAL
trial evaluates the prognostic value of persistent culture
positivity at days 3 and 7 of treatment in patients with fila-
mentous fungal keratitis. Key findings include the
following: (1) persistent culture positivity at day 3 of treat-
4 AMERICAN JOURNAL OF OPHTHALMOLOGY
ment confers a significantly increased risk of requiring
TPK; (2) persistent culture positivity at day 7 is strongly
predictive of need for TPK, final visual acuity, and corneal
scar size; and (3) Aspergillus ulcers were more likely to
remain culture positive and were associated with worse out-
comes. These findings confirm the prognostic utility of
repeat cultures at day 7, and are a novel demonstration of
the utility of day 3 cultures for predicting the need for
TPK in fungal keratitis.
Several clinical factors have been evaluated as
possible prognosticators, including visual acuity at pre-
sentation, epithelial defect size, patient age, size and
depth of infiltrate, and the presence of ulcer pigmenta-
tion.11 However, clinical parameters can be problematic
as prognostic markers owing to inter-examiner vari-
ability, even among expert cornea specialists.12,13 Previ-
ous findings indicate that patients who will eventually
need surgery may benefit from earlier intervention
before infection reaches the limbus or results in perfora-
tion.5,6 However, clinicians managing fungal ulcers
have few clinical indicators to determine which pa-
tients are likely to require TPK. Thus, an objective
parameter that predicts need for TPK as early as
possible is highly beneficial. Our results indicate that
persistent culture positivity at day 3 after initiation of
JULY 2020
FIGURE 2. Kaplan-Meier survival curve comparing the probability of eventuating to therapeutic penetrating keratoplasty based on
7-day culture status.

TABLE 2. Visual Acuity 3 Weeks and 3 Months After Fungal Keratitis Infection According to Culture Status at Day 3 and Day 7

Culture Result Week 3 LogMAR BSCVA (Mean 6 SD) Coefficienta 95% CI P

Day 3 culture results 0.05 (-0.16 to 0.27) .63

Positive 1.35 6 0.64
Negative 1.29 6 0.62
Day 7 culture results 0.31 (0.11 to 0.52) .003
Positive 1.67 6 0.30
Negative 1.17 6 0.69

Month 3 LogMAR BSCVA (Mean 6 SD) Coefficient 95% CI P

Day 3 culture results 0.05 (-0.18 to 0.27) .69

Positive 1.32 6 0.61
Negative 1.26 6 0.65
Day 7 culture results 0.31 (0.09 to 0.52) .006
Positive 1.56 6 0.36
Negative 1.09 6 0.70

BSCVA ¼ best spectacle-corrected visual acuity.

a
Multiple linear regression analysis adjusting for baseline BSCVA.

treatment was a statistically significant indicator of It is interesting that smear positivity at 3 days does not
need for subsequent TPK, with a hazard ratio of 2.8. predict the need for TPK. Previous reports have found 7-
This introduces a new and potentially valuable param- day smear results to be predictive of outcomes in fungal
eter to follow in the management of fungal keratitis. keratitis, including the need for TPK.7,8 This finding is

VOL. 215 THE PROGNOSTIC VALUE OF REPEAT CULTURE IN FUNGAL KERATITIS 5

suggestive that while organisms are still present in the ma-
jority of ulcers at 3 days after initiating treatment, many of
them are no longer viable, as demonstrated by the fact that
they do not grow in culture.
Persistent culture positivity at day 7 of treatment also
seems to be predictive of need for surgical intervention,
but additionally predicts final anatomic and visual out-
comes. This is supported by prior studies evaluating the
role of repeat culture at days 6 and 7.7,8 Although day 7
culture positivity has a stronger association with out-
comes, day 3 may be more clinically useful in determining
who might benefit from early surgical intervention.
Fungal cultures can take upwards of 1 week to grow; day
3 cultures are available almost 1 week prior to day 7 cul-
tures during the crucial period of decision-making
regarding response to therapy and early surgical interven-
tion. Many ulcers have already progressed to full-
thickness perforation or threatening limbal involvement
requiring TPK when the results of day 7 cultures are
evident. Outcomes of TPK at this point are poor. These
findings further reinforce the need for additional studies
to determine whether earlier surgical intervention is bet-
ter in fungal keratitis, in which case serial repeat cultures
may play an important role in selecting patients who are
most likely to benefit from early surgery.
This study has several limitations. First, this was a sec-
ondary analysis of a randomized trial, and thus the results
should be interpreted with appropriate caution. All pa-
tients were enrolled in the study in South India, which
may limit the generalizability of these results to other popu-
lations owing to differences in geographic microorganism
diversity, climate, and host genetic factors.14
FUNDING/SUPPORT: NIH, NEI GRANT #EYK23025025RPB DEPARTMENT GRANT TO THE UNIVERSITY OF CALIFORNIA, SAN
Francisco. The authors indicate no funding support. Financial Disclosures: None of the authors has a proprietary/financial interest to disclose. All authors
attest that they meet the current ICMJE criteria for authorship.
REFERENCES
1. Gopinathan U, Garg P, Fernandes M, Sharma S,
Athmanathan S, Rao GN. The epidemiological features
and laboratory results of fungal keratitis. Cornea 2002;21(6):
555–559.
2. Prajna NV, Krishnan T, Mascarenhas J, et al. The Mycotic
Ulcer Treatment Trial. JAMA Ophthalmol 2013;131(4):
422–429.
3. Prajna NV, Krishnan T, Rajaraman R, et al. Effect of oral
voriconazole on fungal keratitis in the Mycotic Ulcer Treat-
ment Trial II (MUTT II): a randomized clinical trial.
JAMA Ophthalmol 2016;134(12):1365–1372.
4. Prajna NV, Krishnan T, Rajaraman R, et al. Predictors of
corneal perforation or need for therapeutic keratoplasty in se-
vere fungal keratitis: a secondary analysis of the Mycotic Ulcer
Treatment Trial II. JAMA Ophthalmol 2017;135(9):987–991.
6 AMERICAN JOURNAL OF OPHTHALMOLOGY
This study demonstrates the value of persistent cul-
ture positivity at day 3 in predicting the need for surgi-
cal intervention in patients with moderate-to-severe
filamentous fungal keratitis, and confirms the value of
day 7 culture positivity in predicting perforation and vi-
sual acuity. These results indicate that obtaining repeat
cultures at days 3 and 7 after initiating therapy for pa-
tients with recalcitrant fungal corneal ulcers is
warranted. These findings also indicate the need for
further studies to determine whether early TPK in
high-risk patients identified by repeat culture may result
in better outcomes.
CRediT AUTHORSHIP CONTRIBUTION
STATEMENT
JULIA PICKEL: INVESTIGATION, DATA CURATION, WRITING
- original draft. Shivananda Narayana: Conceptualization,
Investigation, Writing - review & editing. Tiruvengada
Krishnan: Conceptualization, Supervision, Writing - re-
view & editing. Seema Ramakrishnan: Conceptualization,
Investigation, Writing - review & editing. Puja Prativa
Samantaray: Investigation, Writing - review & editing.
Travis C. Porco: Formal analysis, Writing - review & edit-
ing. Travis Redd: Formal analysis, Data curation, Writing -
review & editing. Thomas M. Lietman: Conceptualiza-
tion, Supervision, Methodology, Writing - review & edit-
ing. Jennifer Rose-Nussbaumer: Conceptualization,
Methodology, Investigation, Funding acquisition, Writing
- review & editing.
5. Sharma N, Jain M, Sehra SV, et al. Outcomes of therapeutic
penetrating keratoplasty from a tertiary eye care centre in
northern India. Cornea 2014;33(2):114–118.
6. Kocluk Y, Sukgen EA. Results of therapeutic penetrating ker-
atoplasty for bacterial and fungal keratitis. Int Ophthalmol
2017;37:1085–1093.
7. Ray KJ, Lalitha P, Prajna NV, et al. The utility of repeat cul-
ture in fungal corneal ulcer management: a secondary analysis
of the MUTT I randomized clinical trial. Am J Ophthalmol
2017;178(415):157–162.
8. Ray KJ, Prajna NV, Lalitha P, et al. The significance of repeat
cultures in the treatment of severe fungal keratitis. Am J
Ophthalmol 2018;189(415):41–46.
9. Narayana S, Krishnan T, Ramakrishnan S, et al. Mycotic
antimicrobial localized injection a randomized clinical trial
evaluating intrastromal injection of voriconazole. Ophthal-
mology 2019;126(8):1084–1089.
JULY 2020
10. Age-Related Eye Disease Study Group. The Age-Related
Eye Disease Study (AREDS): design implications.
AREDS Report No 1. Control Clin Trials 1999;20:
573–600.
11. Prajna NV, Krishnan T, Mascarenhas J, et al. Predictors of
outcome in fungal keratitis. Eye 2012;26(9):1226–1231.
12. Dalmon C, Porco TC, Lietman TM, et al. The clinical
differentiation of bacterial and fungal keratitis: a photo-
graphic survey. Invest Ophthalmol Vis Sci 2012;53(4):
1787–1791.
13. Dahlgren MA, Lingappan A, Wilhelmus KR. The clinical
diagnosis of microbial keratitis. Am J Ophthalmol 2007;
143(6):940–944.
14. Austin A, Lietman T, Rose-Nussbaumer J. Update on the
management of infectious keratitis. Ophthalmology 2017;
124:1678–1689.

VOL. 215 THE PROGNOSTIC VALUE OF REPEAT CULTURE IN FUNGAL KERATITIS 7