Oxymetazoline: reduction of subconjunctival hemorrhage

incidence after intravitreal injections

Gerardo Gonzalez-Saldivar, MD,*,1 Ingrid Yazmin Pita-Ortiz, MD,y Erick Omar Flores-Villalobos, MD,y
Jes
us Noel Jaurrieta-Hinojos, MD,y Itzel Espinosa-Soto, MD,y Geovanni Rios-Nequis, MD,y
Abel Ramirez-Estudillo, MD,y Martin Jimenez-Rodriguez, MDy


ABSTRACT  RESUM

E
Objective: Subconjunctival hemorrhage (SCH) is an important minor side effect that might affect patient compliance to antivascular
endothelial growth factor (anti-VEGF) intravitreal injection treatment (IVI). We sought to compare SCH incidence and pain score
responses after topical oxymetazoline in na€ıve patients undergoing a single IVI of ranibizumab for diabetic macular edema.
Methods: Prospective, randomized, double-blinded, single centre study. One hundred two patients na€ıve to anti-VEGF were assigned to
receive either topical oxymetazoline or placebo 30 minutes before IVI. SCH incidence and area were measured by slit lamp 24 hours
after, and pain was evaluated 5 minutes and 24 hours after.
Results: SCH incidence was reported on 72% in control group versus 51% in oxymetazoline group (p = 0.037). Mean size of SCH was 16.82
mm2 in control group versus 12.55 mm2 in oxymetazoline group (p = 0.394). Prevalence of local pain in the overall study population was
60%. No significant statistical difference was achieved between groups 5 minutes or 24 hours after IVI in either pain scale evaluation.
Conclusion: Administration of topical oxymetazoline 30 minutes before IVI is a single, harmless, cost-effective intervention that
decreases the incidence of subconjunctival hemorrhage. This may considerably improve patient treatment satisfaction and promote
compliance to IVI therapy.

Objectif: L'hémorragie sous-conjonctivale (HSC) constitue un effet secondaire mineur mais non négligeable qui risque d'influer sur
l'observance des patients qui reçoivent un inhibiteur du facteur de croissance endothélial vasculaire (VEGF, pour vascular endothelial
growth factor) par injection intravitréenne (IIV). Nous avons comparé l'incidence d'HSC et mesuré le score de douleur après l'adminis-
tration topique d'oxymétazoline chez des patients qui en étaient à leur premier traitement et qui devaient recevoir une IIV unique de
ranibizumab dans le traitement de l’œdème maculaire diabétique.
Méthodes: Cette étude prospective, randomisée, à double insu réalisée dans un seul centre réunissait 102 patients qui n'avaient jamais
reçu d'anti-VEGF et qui ont été répartis en 2 groupes : oxymétazoline topique ou placebo 30 minutes avant l'IIV. L'incidence et l’éten-
due de l'HSC ont été mesurées par lampe à fente 24 heures plus tard, et la douleur a été évaluée 5 minutes et 24 heures plus tard.
Résultats: L'incidence d'HSC s'est élevée à 72 % dans le groupe témoin, comparativement à 51 % dans le groupe oxymétazoline
(p = 0,037). L’étendue moyenne de l'HSC était de 16,82 mm2 dans le groupe témoin, comparativement à 12,55 mm2 dans le groupe oxy-
métazoline (p = 0,394). La prévalence de douleur locale chez l'ensemble des sujets était de 60 %. Aucune différence statistiquement signif-
icative n'a été enregistrée entre les groupes 5 minutes ou 24 heures après l'IIV selon l'une ou l'autre échelle d’évaluation de la douleur.
Conclusion: L'administration d'oxymétazoline topique 30 minutes avant l'IIV est une solution simple, sûre et peu coûteuse qui réduit
l'incidence d'hémorragie sous-conjonctivale. Voilà qui peut améliorer considérablement la satisfaction du patient envers son traite-
ment et favoriser son observance envers les IIV.

In recent years, intravitreal injections of antivascular endothelial
growth factor (anti-VEGF) have gained importance in the
treatment of various retinal pathologies. Intravitreal injections
are an effective and safe method for delivering intraocular ther-
apeutic agents.1 Major side effects that have been reported are
endophthalmitis (0.01%), lens injury (0.01%), retinal detach-
ment (0.04%), and inflammation (0.14%).2 However, minor
side effects that are not always mentioned and that have a great
impact on the patient’s quality of life for days or even several
weeks are subconjunctival hemorrhage (SCH) and local pain.2,3
Although considered minor side effects, SCH and local
pain have an important effect during intravitreal injec-
tion. Pain might affect comfort and treatment compliance
and has been reported in more than 85% of procedures
with a visual analog scale of 17.4 § 17.1 (range, 084). 2
On the other hand, SCH with a wide incidence range of
0.03% to 52% might be considered only a cosmetic prob-
lem as it resolves within days with no treatment or conse-
quence at all; however, patient treatment satisfaction may
be altered as some patients avoid social and occupational
interaction until it resolves. 24
Oxymetazoline (OXY) is a selective a1-adrenergic recep-
tor agonist and an a2-adrenergic receptor partial agonist
that causes potent vasoconstriction in ophthalmic as well
as in other tissues.5 a1-agonists may decrease the incidence
of SCH because of their potent vasoconstrictive effect.
This vasoconstriction has been exploited to reduce bleeding
in laser in situ keratomileusis, cataract, and strabismus sur-
gery, but never in the prevention of SCH in intravitreal
injections.6

© 2018 Canadian Ophthalmological Society.

Published by Elsevier Inc. All rights reserved.
https://doi.org/10.1016/j.jcjo.2018.09.006
ISSN 0008-4182

CAN J OPHTHALMOL—VOL. 54, NO. 4, AUGUST 2019 513

OXY for reduction of SCH after intravitreal injections—Gonzalez-Saldivar et al.
As a consequence of the aforementioned, we sought, as a
primary endpoint, to compare the incidence of SCH and
pain score in na€ıve patients undergoing a single intravitreal
injection of ranibizumab for the treatment of diabetic mac-
ular edema. As a secondary outcome, the area of SCH was
evaluated.
TAGEDH1PATIENTS AND METHODSTAGEDEN
Study population
This study was a prospective, randomized, double blind,
single centre study. A total of 102 eyes of 102 patients na€ıve
to treatment with visual complaints due to diabetic macular
edema (DME) between October and December 2017 were
enrolled in the study. Ethical committee approval and
informed consent from all patients were obtained. The study
adhered to the Helsinki Declaration.
Patients over the age of 18 years with diabetic macular
edema na€ıve to intravitreal anti-VEGF injection treatment
were included. DME was defined as a central macular thick-
ness of 315 nm on optical coherence tomography (Spectralis;
Heidelberg Engineering, Heidelberg, Germany). Patients with
a mean arterial pressure >20 mm Hg, neovascular glaucoma,
history of any ocular surgery 6 months before inclusion, or
using antiplatelet or anticoagulant therapy were excluded.
Study design
Patients were randomly assigned into 2 experimental
groups: study group 1 received 0.05 mL of OXY (Borca-
thOX oxymetazoline clorhydrate 0.25 mg/1 mL; Laborator-
ios Sophia, Guadalajara, M
exico) and group 2 (control)
received 0.05 mL of topical saline solution (sodium chloride
[NaCl] 0.9%). Patients, investigators, and pollsters were
masked to treatment. Both groups received their correspond-
ing treatment 30 minutes before intravitreal injection. All
intravitreal injection procedures were standardized and iden-
tical for all the participants in either group. Each patient
received conventional anaesthesia (PONTI Ofteno tetra-
caine hydrochloride 5 mg/1 mL; Laboratorios Sophia)
5 minutes before the procedure. Regular disinfection regi-
men with topical 10% povidone-iodine for the eyelids and
eyelashes several minutes before the procedure was per-
formed. A lid speculum was employed and 4% povidone-
iodine in the cul-de-sac before the procedure was placed.
Ranibizumab 0.5 mg/0.05 mL was injected in the superior-
temporal quadrant by either of 2 retina specialists (I.Y.P.O.,
E.O.F.V.), 4 mm or 3.5 mm from the corneal limbus in
phakic or pseudophakic patients, respectively, with a
30-gauge needle in both groups.
Study measurements
Subconjunctival hemorrhage
Investigators who were masked to the experimental design
evaluated SCH incidence and severity/size 24 hours after the
514 CAN J OPHTHALMOL—VOL. 54, NO. 4, AUGUST 2019
procedure. The area of SCH was determined by multiplying
the maximal lengths of the horizontal and vertical axis of the
SCH 90 degrees one from the other.
Pain
Five minutes after intravitreal injection application for each
eye, the patients were asked for a verbal description of their
pain as “none,” “mild,” “moderate,” or “severe” through a verbal
rating scale (VRS). Participants were also asked to specify the
severity of pain through a numerical pain rating scale (NRS-
11) indicating their score from “0” representing “no pain at all”
to “10” representing “the most severe pain they have experi-
enced.” Outcomes were obtained and analyzed.
Nonparticipating personnel, 5 minutes and 24 hours after
intravitreal injection, applied questionnaires.
Statistics
The statistical analysis was done using SPSS version 20
(IBM Corp, Armonk, N.Y.). We conducted a descriptive sta-
tistical analysis for quantitative variables with measures of cen-
tral tendency and dispersion. In the case of qualitative
variables, frequencies and percentages were obtained and data
were compared using Pearson’s x2 test or Fisher’s exact test
for 2 £ 2 tables. For quantitative comparative variables, the
1-sample Kolmog
orov-Smirnov test was used to test the distri-
bution of data, and depending on the result, we used unpaired
Student’s t or the Mann-Whitney test. A p value <0.05 was
considered statistically significant.
TAGEDH1RESULTSTAGEDEN
A total of 102 eyes from 102 participants (52 females and
50 males) diagnosed with diabetic macular edema and na€ıve
to any therapy were treated with intravitreal injections of
ranibizumab between October and December 2017. The
mean age of the whole study group was 66.3 years (range, 38.7
to 90.6 years). Mean arterial pressure was 104 mm Hg.
The whole studied population is shown in Table 1. There
was no significant statistical difference between the 2 study
groups in any of our variables such as sex, age, treated eye, or
mean blood pressure. Incidence of SCH in the overall study
population that underwent intravitreal injections of ranibizu-
mab was 61%. SCH was reported in 36 out of 50 (72%)
patients in the control group (NaCl) in contrast to 27 out of
52 (51%) in the experimental group (OXY), reaching a signif-
icant difference of p = 0.037 (Table 2). Mean size of the sub-
conjunctival hemorrhage in the study population was 14.99
mm2; no significant difference was achieved when comparing
both groups (p = 0.394), the NaCl group with an area of 16.8
mm2 (SD § 22.9) and the OXY group with an area of 12.55
mm2 (SD § 16.51).
Prevalence of local pain in the overall study population was
60%. When pain perception was assessed by VRS, most patients
reported “mild pain” 5 minutes after intravitreal injection, and
no significant difference was found between groups (p = 0.208).
 OXY for reduction of SCH after intravitreal injections—Gonzalez-Saldivar et al.

Table 1—Baseline characteristics

Variable OXY (n = 52) Control (n = 50) Total (n = 102) p Value
Age, years, mean § SD 64.23 § 10.8 68.47 § 10.39 66.31 § 10.76 0.052
Right eye, n (%) 31 (59.6) 28 (56) 59 (57.8) 0.712
Left eye, n (%) 21 (40.4) 22 (44) 43 (42.2)
SP, mm Hg, mean § SD 146 § 23 149 § 20 147 (§22) 0.478
DP, mm Hg, mean § SD 83 § 11 82 § 12 83 § 11 0.759
MAP, mm Hg, mean § SD 103.89 § 9.23 104.45 § 10.08 104.17 § 9.61 0.77
OXY, oxymetazoline; SP, systolic pressure; DP, diastolic pressure; MAP, mean arterial pressure.

Pain was reported as “no pain” in most of our patients when with saline solution to prevent and decrease the SCH area.8
interrogated 24 hours after application (p = 0.52). “Severe pain” Both incidence and area of SCH were statistically different
was not reported by any of the patients. When pain perception with a p = 0.003 and p = 0.001, respectively.6
was evaluated by NRS-11, the NaCl group reported a mean pain In this study, we prospectively examined the therapeutic
score of 3 (§3) and 1 (§2) for 5 minutes and 24 hours, respec- effect of OXY on the incidence and area of SCH after intravi-
tively. On the other hand, the OXY group reported 2 (§2) and treal injections. Total incidence of SCH was 61%, which is
1 (§2) for 5 minutes and 24 hours, respectively. No significant significantly higher than that of Lagstein et al. and Kim et al.,
difference was found (Table 3). who reported 41% and 21%, respectively. A reason might be
that the indication for IVI therapy in all our patients was dia-
betic macular edema, which is a microvasculature consequence
TAGEDH1DISCUSSIONTAGEDEN of a systemic disease that might be accompanied by platelet
dysfunction or capillary fragility, contrary to Lagstein et al.
SCH is considered a harmless side effect of intravitreal injec-
and Kim et al., who had diverse indications for IVI therapy,
tions; however, it may cause aesthetic problems and anxiety in
including local pathologies such as exudative age-related mac-
patients, which might lead to anti-VEGF cyclic treatment inter-
ular degeneration. In our study, the OXY treatment group’s
ruption.2 SCH is a result of direct damage to conjunctival ves-
incidence of SCH was significantly lower when compared
sels by the injection needle. Activation of a-1 adrenergic
with NaCl (p = 0.037), having an effect similar to brimoni-
receptors causes vasoconstriction and blanching of conjunctiva
dine. On the other hand, when comparing area, there was no
making SCH less likely to occur.7 Various methods have been
significant difference between both groups in this study, a
examined to prevent SCH after intravitreous injections. Lag-
finding that differs from the brimonidine effect, where besides
stein et al. in 2017 compared the use of apraclonidine and
incidence, the area was also significantly lower when com-
saline solution in order to prevent SCH, but no significant sta-
pared with NaCl. Apraclonidine, in contrast, did not reduce
tistical difference was observed both in incidence (p = 0.503)
neither incidence nor area when compared with NaCl.
and the SCH area (p = 0.253).2 On the other hand, Kim et al.
As a secondary endpoint, pain was analyzed, having an inci-
reported in 2011 the prophylactic use of brimonidine compared
dence in the overall study population of 60.8% (61) and 39.2%
(40) at 5 minutes and 24 hours, respectively; no statistical dif-
Table 2—Incidence of subconjunctival hemorrhage and area
ference was observed comparing both groups by either scale
Variable OXY (n = 52) Control (n = 50) p Value
after intravitreal injection. Lagstein et al. reported a mean
SCH; n (%) 27 (51) 36 (72) 0.037*
NO SCH; n (%) 25 (48.1) 14 (28) NRS-11 score 30 minutes after IVT of 1.69 (SD § 1.44) with
SCH area mm2; mean § SD 12.55 § 16.51 16.82 § 22.99 0.394 topical apraclonidine versus 3.28 (SD § 2.27) in the control
OXY, oxymetazoline; SCH, subconjunctival hemorrhage.
*
p value <0.05 was considered statistically significant.
group, achieving a statistical difference (p < 0.001).2 We found
a greater mean NRS-11 score than that reported by Lagstein
et al. Although OXY is a selective a1 and partial a2 adrenergic
Table 3—Pain scores receptor agonist, which causes potent vasoconstriction, it does
Incidence of Pain VRS not have an analgesic effect on the eye.
5 minutes 24 hours A bottle of OXY drops (each with a volume of 10 mL) has
a cost around $11.40 USD in Mexico. If 1 drop is placed on
OXY n, Control n, p OXY n, Control n, p
(%) n = 52 (%) n = 50 (%) n = 52 (%) n=50 each patient, it produces a potential cost of $0.05 USD per
None 24 (46.2) 16 (32) 0.208 28 (53.8) 34 (68) 0.52 patient. When compared with apraclonidine $0.29 or brimo-
Mild 20 (38.5) 29 (58) 16 (30.8) 10 (20) nidine $0.34, we can assume that this is a secure and inexpen-
Moderate 7 (13.5) 5 (10) 7 (13.5) 5 (10)
Severe 1 (1.9) 0 (0) 1 (1.9) 1 (2) sive method of prophylaxis for SCH that could lead to better
patient satisfaction and quality of life.9
NRS-11

OXY (n = 52) Control (n = 50) p

5 minutes, mean § SD 2 (§2) 3 (§3) 0.253
TAGEDH1CONCLUSIONTAGEDEN
24 hours, mean § SD 1 (§2) 1 (§2) 0.373 Administering topical OXY 30 minutes before intravitreal
OXY, oxymetazoline; VRS, Verbal Rating Scale Score; NRS-11, numerical pain rating scale. injections decreases the incidence of SCH. This is an inexpensive

CAN J OPHTHALMOL—VOL. 54, NO. 4, AUGUST 2019 515

OXY for reduction of SCH after intravitreal injections—Gonzalez-Saldivar et al.

and harmless procedure that might considerably improve
patient’s comfort and promote compliance to treatment.
TAGEDH1REFERENCESTAGEDEN
8. Kim CS, Nam KY, Kim JY. Effect of prophylactic topical brimonidine
(0.15%) administration on the development of subconjunctival hemor-
rhage after intravitreal injection. Retina. 2011;31:389–92.
9. Stavert B, McGuinness MB, Harper CA, Guymer RH, Finger RP. Cardio-
vascular adverse effects of phenylephrine eyedrops: a systematic review and
meta-analysis. JAMA Ophthalmol. 2015;133:647–52.

1. Fasih U, Shaikh N, Rahman A, Sultan S, Fehmi MS, Shaikh A. A one-year

follow-up study of ocular and systemic complications of intravitreal injec-
tion of bevacizumab (Avastin). J Pak Med Assoc. 2013;63:707–10.
2. Lagstein O, Artzi BA, Achiron A, et al. Topical apraclonidine reduces pain
after intravitreal injections: a double-blind randomized controlled trial. Footnotes and Disclosure:
Retina. 2017;37:1575–80.
3. Yun C, Oh J, Hwang SY, Kim SW, Huh K. Subconjunctival hemorrhage The authors have no proprietary or commercial interest in any mate-
after intravitreal injection of anti-vascular endothelial growth factor. rials discussed in this article.
Graefes Arch Clin Exp Ophthalmol. 2015;253:1465–70.
4. Loukopoulos V, Meier C, Gerding H. Hemorrhagic complications after
From the *Ophthalmology Unit, Fundaci
on Hospital Nuestra
intravitreal injections of ranibizumab in patients under coumarin-type nora de la Luz, IAP, Ciudad de M
exico, Mexico; yDivision of Ret-
Se~
anticoagulation. Klin Monbl Augenheilkd. 2010;227:289–91. ina, University of Toronto, St. Michael’s Hospital, Toronto, Ont.
5. Xuan B, Chiou GC. Efficacy of oxymetazoline eye drop in non-infectious Originally received Jul. 19, 2018. Final revision Sep. 17, 2018.
conjunctivitis, the most common cause of acute red eyes. J Ocul Pharmacol
Ther. 1997;13:363–7.
Accepted Sep. 18, 2018.
6. Gu B, Dou X, Sima J, Peng N. The application of naphcon eye drops dur- Correspondence to Gerardo Gonz
alez-Saldivar, MD, Fundaci
on
ing lasik surgery. Yan Ke Xue Bao. 2004;20:206. Hospital Nuestra Se~ nora de La Luz, IAP, Departamento de Retina,
7. Xuan B, Chiou GC. Efficacy of oxymetazoline eye drop in non-infectious Calle Ezequiel Montes # 135, Deleg. Cuauht
emoc Colonia Tabaca-
conjunctivitis, the most common cause of acute red eyes. J Ocul Pharmacol
Ther. 1997;13:363–7.
lera, Cd. de M
exico, M
exico C.P. 06030; gzzgerardo@hotmail.com

516 CAN J OPHTHALMOL—VOL. 54, NO. 4, AUGUST 2019