- Email: [email protected]
Oral versus Topical Diclofenac for Pain Prevention during Panretinal Photocoagulation
Oral versus Topical Diclofenac for Pain Prevention during Panretinal Photocoagulation Peter A. Zakrzewski, MD,1 Heather L. O’Donnell, BSc,1 Wai-Ching Lam, MD, FRCSC2 Purpose: To investigate the effect of pretreatment oral and topical diclofenac on pain reduction during panretinal laser photocoagulation (PRP) for proliferative diabetic retinopathy (PDR). Design: Prospective, randomized, double-masked, placebo-controlled clinical trial. Participants and Controls: A total of 90 patients with PDR requiring PRP for the first time were assigned randomly to 1 of 3 study groups: oral diclofenac (n ⫽ 30), topical diclofenac (n ⫽ 31), or placebo (n ⫽ 29). Methods: Study medications were administrated before the first PRP treatment, and pain levels experienced during and 15 minutes after PRP were recorded on a visual analog scale (VAS). Pain levels during a second PRP session, performed on a later date with no pretreatment medications, also were recorded on a VAS. Main Outcome Measures: The primary outcome measures were the mean VAS pain scores during the first PRP treatment. Secondary outcome measures were the mean VAS pain scores 15 minutes after the first PRP and during the second PRP, and reported side effects after the first PRP. Results: Mean VAS pain scores during the first PRP were: oral diclofenac, 25.7⫾19.9; topical diclofenac, 33.8⫾27.9; and placebo, 41.3⫾31.0. The pain score difference between oral diclofenac and placebo was both clinically significant (ⱖ13) and statistically significant (P ⫽ 0.02), whereas differences between oral and topical diclofenac (P ⫽ 0.20) and topical diclofenac and placebo (P ⫽ 0.33) were not. Multivariate regression analysis for age, gender, and total laser energy demonstrated lower pain levels for both oral diclofenac (P ⫽ 0.015) and topical diclofenac (P⬍0.0001) versus placebo, but no difference between oral and topical diclofenac (P ⫽ 0.67). For the first PRP, all 3 groups had lower mean pain scores at 15 minutes after treatment compared with during treatment (Pⱕ0.0003). Mean pain scores were higher during the second compared with the first PRP for the oral diclofenac (P ⫽ 0.02) and placebo (P ⫽ 0.05) groups. No significant rate difference for any side effect was found between groups. Conclusions: When given in a single dose, oral diclofenac is an effective pretreatment analgesic agent for reducing the pain experienced during PRP for PDR. Financial Disclosure(s): The author(s) have no proprietary or commercial interest in any materials discussed in this article. Ophthalmology 2009;116:1168 –1174 © 2009 by the American Academy of Ophthalmology.
Panretinal photocoagulation (PRP) is an effective treatment for reducing the risk of vision loss in proliferative diabetic retinopathy (PDR).1–3 Patients frequently experience pain during the procedure, which typically is in the moderate range.4 – 8 Preprocedure or rescue retrobulbar or peribulbar injection anesthesia can be used to reduce pain during the procedure, but it is invasive and imparts the risk of serious, although rare, complications.9 –11 Other pre-PRP analgesic techniques investigated include sub-Tenon anesthesia delivery,12 transcutaneous electrical nerve stimulation,13 and inhalation of Entonox (50:50 nitrous oxide:oxygen gas)5; however, these methods are limited by their need for special equipment, invasiveness, or both. Various medications that are easily administered and impart only minimal risk have been studied as pretreatment agents, including nonsteroidal anti-inflammatory drugs (NSAIDs; topical diclofenac,4 topical ketorolac,6 oral mefenamic acid,8 and intramuscular ketorolac8), oral acetaminophen,7,8 and oral diazepam.8 Among these medications, only topical diclofenac has demonstrated statistically significant pain reduction during PRP.4
1168
© 2009 by the American Academy of Ophthalmology Published by Elsevier Inc.
Nonsteroidal anti-inflammatory drugs comprise a family of compounds with proven analgesic, anti-inflammatory, and antipyretic effects that can be used systemically or locally. They produce their clinical effects mainly by inhibiting the cyclooxygenase pathway, which leads to reduction in prostaglandin formation, a potent mediator of the pain pathway.14 –16 Diclofenac is a commonly used NSAID available in both oral and topical forms. The oral form is a potent analgesic agent effective in many conditions,14,17 including postoperative pain when given as a single dose.18,19 The topical form is indicated for postoperative inflammation after cataract surgery and for relief of pain and photophobia after corneal refractive surgery.15,16 Oral NSAIDs reach targeted tissues via the systemic circulation, whereas topical NSAIDs directly penetrate the cornea to reach intraocular tissues.20 The use of topical NSAIDs as analgesics in posterior segment procedures has not been well studied. In rabbit models, after topical administration, the NSAIDs ketorolac and flurbiprofen were found throughout the eye, including the choroid and retina; howISSN 0161-6420/09/$–see front matter doi:10.1016/j.ophtha.2009.01.022
Zakrzewski et al 䡠 Oral vs. Topical Diclofenac for Pain Prevention during PRP ever, lower levels were found in the vitreous than in the aqueous humor.21,22 After preoperative administration, the topical form of the NSAID indomethacin was shown to have higher aqueous levels than its oral form.23 It is unknown whether oral or topical NSAID administration would result in higher concentrations at the level of the long posterior ciliary nerves in the suprachoroidal space, the purported site of pain generation during PRP.24 No studies to date have directly compared the oral and topical forms of an NSAID in reducing pain during PRP. The purpose of this study was to investigate the analgesic effects of pretreatment oral and topical diclofenac during and after PRP for PDR.
Patients and Methods This prospective, randomized, double-masked, 3-armed placebocontrolled trial was carried out using consecutive patients from the practice of 1 vitreoretinal specialist (WCL) in the Department of Ophthalmology and Visual Sciences at the University of Toronto between November 2003 and March 2007. Ethics approval was granted by the University of Toronto University Health Network Research Ethics Board. Approval for the clinical trial was granted by the Health Canada (federal government) Therapeutic Products Directorate, which is where the trial also was formally registered and is equivalent to an Investigational New Drug application process in the United States. Informed consent was obtained from all patients who participated in the study. The research described herein adhered to the tenets of the Declaration of Helsinki. Literature searches were carried out in the PubMed and MEDLINE databases in June 2002 and March 2008 to uncover all previously published articles describing pain modification techniques for PRP using the search terms panretinal photocoagulation, laser photocoagulation, and pain. Inclusion criteria were the diagnosis of PDR requiring PRP treatment for the first time and the ability to understand the nature of the study and to complete all measurement instruments. Exclusion criteria were previous PRP in either eye and contraindications to oral NSAIDS, including pregnancy, history of peptic ulcer disease, bleeding disorders, asthma, or renal failure; known allergy or hypersensitivity to NSAIDs or acetylsalicylic acid; and current use of any narcotic, NSAID (except acetylsalicylic acid), digitalis, lithium, probenecid, methotrexate, phenytoin, or cyclosporine. Age and gender were collected as demographic data. Only 1 eye per patient was enrolled in the study. Patients were randomized into 1 of 3 study groups: (1) oral diclofenac potassium (immediate release) 50 mg (Voltaren Rapide; Novartis International AG, Basel, Switzerland) plus topical placebo; (2) oral placebo plus nonpreserved topical diclofenac sodium 0.1% (Voltaren Ophtha, Novartis International AG); or (3) oral placebo plus topical placebo. The oral diclofenac and oral placebo were prepared in unmarked capsules and were indistinguishable in appearance and odor from each other. The topical placebo used was Bion Tears (Alcon, Inc., Fort Worth, TX), which is a nonpreserved artificial tear available in single-dose packages. With stickers covering the product labels, these single-dose packages were indistinguishable in appearance from the topical diclofenac single-dose packages used in the study. Patients receiving topical diclofenac may have experienced more stinging on administration than placebo. Recruited patients were randomized (by electronic number generation) in blocks of 6 to 1 of the 3 study groups. Study medication packages containing the appropriate oral and topical products were premade by a hospital pharmacist, and none of the treating physicians, study nurses, or patients knew what study
group they were part of. The randomization code linking patients to study groups was stored locked in the hospital pharmacist’s office and was not broken until all patients had completed testing. A total of 14 physicians (1 staff physician, 13 ophthalmology fellows and residents) took part in the enrollment, PRP treatment, and pain measurement testing of study patients. All study patients were treated in the same manner and all medications were administered by the study nurse. Pupillary dilation was achieved with 1 drop each successively of proparacaine 0.5% (Alcaine; Alcon, Inc.), tropicamide 1% (Mydriacyl; Alcon, Inc.), and phenylephrine 2.5% (Mydfrin; Alcon, Inc.) into the study eye a minimum of 30 minutes before PRP. Repeat administration of dilating drops was allowed in cases of inadequate pupillary dilation. Both the oral study medication (given with water) and 1 drop of the topical study medication were given 60 minutes before the onset of PRP, followed by a second drop of the topical study medication 5 to 10 minutes before the PRP was started. Another drop of proparacaine was given just before PRP to facilitate the placement of the fundus contact lens used to guide laser administration. The PRP was performed using an argon laser with settings based on individual user preferences, with a maximum of 600 laser burns allowed during the treatment session. One of yellow, yellowgreen, or green wavelengths was used; spot sizes ranged from 200 to 500 m; and laser durations ranged from 0.07 to 0.15 seconds. The energy per burn (mW) and total number of burns were recorded for each patient, allowing for the calculation of total laser energy delivered (total number of burns ⫻ energy per burn) during each PRP session. The treating physician was asked to record if rescue injection anesthesia was required to complete the PRP. Immediately after the PRP session, study patients were asked to rate the level of pain they had felt during the PRP treatment session on the visual analog scale (VAS), a pain measurement tool previously validated as a reliable, sensitive measure of pain and change in pain.25 It consists of a 100-mm line anchored at one end by the label no pain at all and at the other end by the label the worst pain imaginable. Patients were asked to mark the line with an X to indicate pain intensity. The position of the X was measured by a masked third party and was converted to a score from 0 to 100. Fifteen minutes after the completion of the PRP treatment, study patients were asked to rate their current level of pain on a second VAS and to report any ocular or systemic side effects that were not present before the start of the study. A checklist of expected possible side effects was provided, and space also was given for subjects to write in any side effects being experienced that were not on the checklist. When study patients returned to the clinic for their second PRP treatment session in the study eye, they were not given any pretreatment medications except for the topical proparacaine and dilating drop regimen as described previously. The range of laser settings used was the same as above, and the number of burns again was limited to 600. The energy per burn and the number of burns were recorded, again allowing for the calculation of total laser energy delivered. After this second PRP session, study patients were asked to rate the level of pain experienced during the PRP on a third VAS. The primary outcome measure was the VAS pain score for each group during the first PRP treatment session. Secondary outcome measures were the VAS pain scores 15 minutes after the first PRP session, VAS pain scores during the second PRP treatment, the need for injection anesthesia during the first PRP session, and ocular and systemic side effects experienced during the first session. All study data were masked and were analyzed only at the completion of the study by a noninvolved third party. During the design of the study, a 3-way analysis of variance power calculation (80% power, 5% significance level) determined
1169
Ophthalmology
Volume 116, Number 6, June 2009
Table 1. Demographics and Mean Total Laser Energy in Each Panretinal Photocoagulation Session for Each Study Group Group 1
Group 2
Group 3
Mean age (yrs)⫾SD 59.8⫾12.7 59.0⫾14.6 56.6⫾16.9 Male:female (% male) 20:10 (66.7) 21:10 (67.7) 17:12 (58.6) First PRP mean total 177.1⫾63.3 193.8⫾119.4 188.3⫾82.8 energy (mW/1000)⫾SD Second PRP mean 179.8⫾61.5 178.0⫾65.4 157.3⫾59.1 total energy (mW/1000)⫾SD
P Value 0.71 0.73 0.82 0.35
PRP ⫽ panretinal photocoagulation; SD ⫽ standard deviation.
that 30 patients per study group would be needed to detect a 13-point statistically significant difference in mean VAS pain scores between groups. A 13-point difference was chosen because it represented the most conservative value within the range of published minimum VAS pain score differences (9 –13) required to achieve clinical significance.26 –29 The 1-way analysis of variance test was used to compare differences in mean patient ages and total laser energies between groups, and a chi-square test was used to compare gender proportions. Mean VAS pain scores were compared between groups with the t test. Multivariate regression analysis of pain scores for age, gender, and total laser energy was performed assuming exponential distribution with the parameters estimated by the maximum likelihood method (PROC GENMOD), where P values were generated by the chi-square test. To allow for proper statistical regression analysis, it was necessary to avoid pain scores with a value of 0; thus, a transformation was used with all new pain scores equaling the old pain scores plus 1. The Fisher exact test was used to analyze differences in side-effect proportions between groups experienced during the first PRP session. All statistical analyses were performed using the SAS software system, version 9.1 (SAS Inc., Cary, NC).
Results A total of 90 patients were recruited into the study, with 30 in the oral diclofenac group, 31 in the topical diclofenac group, and 29 in
the placebo group. No patients changed groups at any time during the study. All 90 patients took their appropriate pretreatment medications (and thus were analyzed as intention to treat), underwent their first PRP treatment, and completed the 2 pain questionnaires corresponding to the first PRP session. Three patients (1 from each study group) did not have pain questionnaires completed for a second PRP session. Two of these patients underwent a second PRP treatment, but erroneously were not asked to complete a pain questionnaire, whereas the third patient had a vitreous hemorrhage and did not undergo follow-up PRP. The other 87 patients underwent a second PRP session and completed the corresponding pain questionnaire. The range of elapsed time between the first and second PRP treatments was 2 days to 15 weeks; however, for 80 of the 87 patients, the elapsed time was between 1 and 5 weeks. Six patients underwent the second PRP less than 1 week after the first, and 1 underwent the second PRP more than 5 weeks after the first (15 weeks). The mean age, gender proportion, and total laser energies for each PRP session for each study group are presented in Table 1. None of the differences between groups were statistically significant. Mean VAS pain scores for each group during the first PRP treatment session were: oral diclofenac, 25.7⫾19.9; topical diclofenac, 33.8⫾27.9; and placebo, 41.3⫾31.0 (Fig 1A). The lower pain rating for oral diclofenac compared with placebo was statistically significant (P ⫽ 0.02), whereas differences between oral diclofenac and topical diclofenac (P ⫽ 0.20) and between topical diclofenac and placebo (P ⫽ 0.33) were not significant. To account for possible effects on pain levels by differences in age, gender proportion, and total laser energy between groups, a multivariate regression analysis of the VAS pain scores for these variables was performed. It demonstrated statistically significant pain reductions for oral diclofenac (P ⫽ 0.01) and topical diclofenac (P⬍0.0001) versus placebo, but no difference between oral and topical diclofenac (P ⫽ 0.67). The mean VAS pain scores 15 minutes after the completion of the first PRP session were: oral diclofenac, 8.8⫾9.7; topical diclofenac, 10.1⫾13.9; and placebo, 15.2⫾18.5 (Fig 1B). None of the pain score differences between groups were statistically significant (oral diclofenac vs. placebo, P ⫽ 0.10; oral vs. topical diclofenac, P ⫽ 0.67; topical diclofenac vs. placebo, P ⫽ 0.24). Comparison between groups with multivariate regression for age, gender, and total laser energy also failed to demonstrate differences between groups (oral diclofenac vs. placebo, P ⫽ 0.18; topical diclofenac vs. placebo, P ⫽ 0.26; oral vs. topical diclofenac, P ⫽ 0.85).
Figure 1. Bar graphs showing mean visual analog scale (VAS) pain scores (bracketed by standard errors) for each group (A) during the first panretinal photocoagulation (PRP), (B) 15 minutes after the first PRP, and (C) during the second PRP.
1170
Zakrzewski et al 䡠 Oral vs. Topical Diclofenac for Pain Prevention during PRP Mean VAS pain levels reported during the second PRP laser session (with no pretreatment medications) were: oral diclofenac, 38.5⫾27.0; topical diclofenac, 39.5⫾29.1; and placebo, 54.5⫾27.9 (Fig 1C). Compared with placebo, the lower pain ratings for both oral diclofenac (P ⫽ 0.03) and topical diclofenac (P ⫽ 0.04) were significant, whereas there was no significant difference between oral and topical diclofenac (P ⫽ 0.88). Comparison between groups with the multivariate regression analysis for age, gender, and total laser energy demonstrated statistically less pain for oral diclofenac versus placebo (P ⫽ 0.007), but no differences between topical diclofenac and placebo (P ⫽ 0.07) and oral and topical diclofenac (P ⫽ 0.78). For each study group, comparisons were made between VAS pain scores during the first PRP and those recorded 15 minutes after treatment and between pain scores for the first and second PRP sessions. All 3 study groups demonstrated statistically significant reductions in pain at 15 minutes after treatment compared with during the first treatment (oral diclofenac, P ⫽ 0.0001; topical diclofenac, P⬍0.0001; placebo, P ⫽ 0.0003). Both the oral diclofenac (P ⫽ 0.02) and placebo (P ⫽ 0.05) groups demonstrated an increase in pain from the first to the second PRP. The difference between the 2 PRP sessions for the topical diclofenac group was not significant (P ⫽ 0.33). Regression analysis accounting for total energy differences between the 2 PRP sessions found the pain increase to be statistically significant for oral diclofenac (P ⫽ 0.05), but not for topical diclofenac (P ⫽ 0.54) or placebo (P ⫽ 0.12). One patient in the study (topical diclofenac group) required a rescue retrobulbar anesthesia injection to complete the first PRP treatment session. This patient’s reported VAS pain scores during and 15 minutes after the first PRP session were 99 and 3.5, respectively. For the second PRP session, the patient reported a pain score of 77, but did not require injection anesthesia to complete treatment. Side effects reported by each study group for the first PRP treatment session are shown in Table 2. Although there were more headaches, less eye stinging, and less photophobia in the oral diclofenac group compared with the topical diclofenac and placebo groups, none of the differences reached statistical significance.
Discussion This study sought to investigate the level of analgesia provided by the oral and topical forms of the NSAID diclofenac when given before PRP for PDR. The idea for the study was contrived because patients typically experience pain during PRP, but there is no consensus favoring any particular pretreatment analgesic agent. In this study, plus the 4 published randomized trials that investigated pretreatment analgesic medications for PRP using the VAS scale, mean placebo group pain scores range from 37.3 to 53.1,4,6 – 8 which falls in the moderate range of pain (31– 69).27 The most common descriptions of pain experienced during PRP included sharp, flashing, pricking, tiring, and piercing.7 Finding a safe, inexpensive, and easy-to-administer method to reduce the pain intensity felt during PRP would make the treatment more comfortable for patients, likely would decrease the need for rescue injection anesthesia, and may improve compliance with follow-up PRP treatments. Oral and topical diclofenac were chosen for this study because oral diclofenac is effective in reducing postoperative pain from various procedures when given as a single preoperative dose18 and topical diclofenac was the only
Table 2. Number of Subjects in Each Group Reporting Side Effects for the First Panretinal Photocoagulation Treatment Session Side Effect Headache Eye stinging Dizziness/lightheadedness Fatigue Photophobia Nausea Agitated/restless Eye itchiness Heartburn Other† Eye pain Eye stickiness Seeing colors Tearing Dry eye Facial pain
Group 1 (n ⴝ 30)
Group 2 (n ⴝ 31)
Group 3 (n ⴝ 29)
8 5 3 3 1 1 0 0 0
4 (P ⫽ 0.21)* 11 (P ⫽ 0.15)* 2 2 6 (P ⫽ 0.10)* 0 2 1 0
4 (P ⫽ 0.33)* 8 (P ⫽ 0.36)* 3 1 5 (P ⫽ 0.10)* 2 3 0 1
2 1 1 0 0 0
0 0 0 1 1 0
3 0 0 0 0 1
*P values represent statistical analysis with the Fisher exact test of the differences in proportions between group 1 and groups 2 or 3. † Indicates a symptom written in by subject(s) that was not offered on the questionnaire checklist.
medication found to decrease pain during PRP when given before treatment.4 The immediate-release form of oral diclofenac was used because it has the fastest time of onset (detectable in plasma within 10 minutes of ingestion with a peak plasma concentration between 20 and 60 minutes after dose),30 allowing for the least required time lag between drug administration and onset of PRP. Two doses of topical diclofenac were given (at 60 minutes and 5–10 minutes before treatment) to ensure that both the short-term and long-term potential analgesic effects afforded by the medication were accounted for in the study. The results from the study demonstrated that pretreatment oral diclofenac provides statistically significant pain reduction during PRP treatment for PDR compared with placebo. This finding was evident when the raw study data were analyzed and also when a multivariate regression for age, gender, and total laser energy was performed. Topical diclofenac, however, produced statistically significant pain reduction versus placebo with the multivariate regression, but not when the raw study data was analyzed without the regression analysis. Differences between oral and topical diclofenac were not statistically significant with either analysis. The multivariate regression analysis was carried out in an attempt to remove possible confounding of pain perception resulting from differences in age, gender, and total laser energy between study groups. To what extent each of these variables had on pain levels recorded in this study is unknown. Both of the statistical analysis techniques (multivariate regression or raw data alone without the regression) are important in interpreting the study results; thus, both techniques are reported throughout the study. The 2 techniques should be evaluated together when making conclusions; however, the presence or absence of clinical signifi-
1171
Ophthalmology
Volume 116, Number 6, June 2009
cance also is important to consider. For example, topical diclofenac was statistically superior to placebo in pain reduction during the first PRP with multivariate regression, but not with raw data alone. This makes interpretation difficult; however, the pain score difference between the groups was not clinically significant, so the overall conclusion was that topical diclofenac was not superior to placebo. Determining the required difference in VAS pain scores between groups to constitute actual clinical significance is based on several published studies on the issue. The published minimum VAS pain score differences found to reflect clinical significance ranges between 9 and 13 on the 0 to 100 scale.26 –29 Although these studies took place in an emergency room setting measuring changes in acute pain over time for individual patients, these values are the best available predictor of the pain score differences necessary in this study to detect clinical significance. Examining the pain levels for the first PRP, the pain score difference of 15.6 between oral diclofenac and placebo achieved clinical significance, whereas differences between topical diclofenac and placebo (7.5) and between oral and topical diclofenac (8.1) did not. Interestingly, the VAS pain score difference between topical diclofenac and placebo in this study was similar to that in the study by Weinberger et al4 (8.9). In that study, the difference was found to be statistically significant; however, using the published range of minimum required pain score differences of 9 to 13, it does not quite reach clinical significance. At 15 minutes after the first PRP, pain scores for all 3 groups were in the mild range (⬍31),27 with the highest being 15.2 (placebo). None of the differences in pain scores between groups at this time point were significant. Meanwhile, pain reductions for the 3 groups from during the first PRP to 15 minutes after were all clinically and statistically significant. This implies that pain experienced during PRP diminishes fairly rapidly to a mild level, although the findings cannot be extrapolated past 15 minutes. Vaideanu et al7 found the mean pain score for the placebo group at 24 hours after PRP to be approximately half of that immediately after the laser treatment (VAS scores of 27 and 52, respectively). This amount of pain at 24 hours is almost double the placebo group pain score 15 minutes after laser treatment (15.2), a difference that may be explained by heterogeneity of study populations, or possibly a later increase in laserinduced pain. Further study is required to determine what happens to pain levels between 15 minutes and 24 hours after treatment. Comparing pain scores between the 2 PRP sessions, the statistically higher pain score for oral diclofenac during the second PRP likely is explained by the lack of pretreatment medication given before the second PRP. For the placebo group, there are 3 possible explanations for why the second PRP pain score was higher than the first. First, subjects received placebo oral and topical medications before the first PRP, but not the second; thus, the placebo effect would have been present only for the first PRP. The placebo effect is a well-studied phenomenon and is believed to alter pain perception in 3 ways: (1) changing the prestimulus expectation of pain relief; (2) modifying the actual level of pain experienced; and (3) distorting the poststimulus pain rat-
1172
ing.31 Second, subjects might have experienced sufficient pain during the first PRP session to develop a central sensitization of pain associated with the procedure, which might have led to a heightened perception of pain experienced during the second PRP treatment.32 Third, the pain score differences between the 2 PRP sessions were not significant for the placebo group when a regression analysis accounting for differences in total laser energy between the 2 sessions was performed. This implies that the pain differences seen with the raw data may be the result of the higher mean laser energy delivered during the second PRP compared with the first. With respect to safety, no significant differences were found between the groups for any particular side effect. No serious safety concerns were identified, although patients were monitored only to 15 minutes after PRP. Using a single dose of oral diclofenac or 2 doses of topical diclofenac would be very unlikely to cause any serious adverse events. A large review study analyzed 528 patients receiving a single dose of oral diclofenac for postoperative pain and found no more side effects with diclofenac than placebo, with neither experiencing any serious side effects.18 Side effects resulting from topical diclofenac typically are local and benign and include stinging, conjunctival hyperemia, and reversible keratitis.15,33 Of the 177 total PRP sessions in this study, only 1 required rescue injection anesthesia to complete laser treatment. Although effective in reducing pain during PRP,8 injection anesthesia carries the risk of rare, although serious, complications of brainstem anesthesia,11 respiratory arrest,10 and globe perforation.9 Thus, the safest course of action likely would be to use injection anesthesia only as a rescue procedure when pain is severe enough to compromise completion of PRP treatment. Besides pretreatment analgesia for lower pain experiences during PRP, researchers also have begun investigating different laser parameters in an attempt to reduce pain. Pain during PRP is believed to arise from the photocoagulation of ciliary nerves running in the suprachoroidal space, and more pain typically is experienced with more anterior application and along the horizontal meridians.4,24 Avoiding these areas, if possible, may lead to decreased pain. A small study (20 patients) comparing short laser exposure PRP (0.02 seconds) versus standard exposure (0.1 seconds) with a 532-nm, frequency-doubled, neodymium:yttrium–aluminum– garnet laser found statistically less pain in the shortexposure group.34 Another study compared different diode laser pulse shapes for PRP and demonstrated more pain in the standard square-wave mode than in the altered shapedwave and micropulse modes.35 Combining optimal laser parameters with effective pretreatment analgesia may result in even lower pain levels during PRP than either method alone. It is important to consider the waiting period between the administration of oral diclofenac and the PRP treatment when taking into account pain reduction for patients in a real clinic setting. Based on half-life data for the medication, a waiting period of 60 minutes was chosen. It is unknown what effect a shorter or longer waiting time would has on pain reduction; thus, the best chance to duplicate the results from this study would be to use a 60-minute waiting
Zakrzewski et al 䡠 Oral vs. Topical Diclofenac for Pain Prevention during PRP period between medication administration and PRP application. In some clinic settings, PRP treatments are given immediately after the diagnosis of PDR. In these settings, adding a waiting period may not be practical and any benefit from pain reduction may be outweighed by an increased patient frustration from having a longer appointment time. These issues must be considered when deciding whether to incorporate such a pain reduction strategy in each individual clinic. A limitation of the study is that patients receiving topical diclofenac may have experienced minor stinging, leading them to believe they were receiving actual medication. What effect this may have had on recorded pain levels is unknown. Another potential limitation was the large number of treating physicians, which was an unavoidable consequence in a busy vitreoretinal practice in an academic setting. Each treating physician likely dealt with patient reports of pain differently, with the content and nature of their responses possibly leading to an alteration of patients’ perceived and recorded levels of pain. Each treating physician also likely had his or her own end point for what constitutes a sufficient PRP laser burn, which might have led to differences in the total amount of laser energy delivered per session and might have affected the level of pain experienced. The authors attempted to correct for this with regression analyses. Another limitation of the study is the variable laser settings used by each treating physician. At the time the study was designed, there were no published studies examining the effect of laser exposure times on pain levels. As mentioned before, a shorter laser exposure time was found to correlate with lower pain levels during PRP with the neodymium:yttrium–aluminum– garnet laser.35 It is unclear whether this finding can be extrapolated to the argon laser used in this study, and even if so, whether these study results would have been altered by any differences in laser exposure times between groups. Whether the analgesic benefit of oral diclofenac found in this study for PRP can be extrapolated to other oral NSAIDs is not known. In the study by Wu et al,8 pre-PRP administration of oral mefenamic acid, an NSAID, did not provide pain reduction compared with the control group. However, a direct comparison of single doses of oral diclofenac versus oral ibuprofen for postoperative pain found no significant analgesic differences between the 2, with both demonstrating significant pain reductions compared with placebo.18 No studies to date have compared ibuprofen with either placebo or diclofenac for pain reduction during PRP. In conclusion, this study demonstrated that a single pretreatment dose of oral diclofenac provides clinically and statistically significant pain reduction during PRP for PDR. Although pretreatment topical diclofenac provided pain reduction that reached statistical significance using a multivariate regression analysis, the actual amount of pain reduction was not clinically significant. Based on the results of this study, the principal author (WCL) now offers all patients requiring PRP the option of a pretreatment dose of oral diclofenac potassium 50 mg provided that there are no contraindications to oral NSAIDs (listed in “Methods” as study exclusion criteria).
References 1. Diabetic Retinopathy Study Research Group. Photocoagulation treatment of 7 proliferative diabetic retinopathy: clinical application of Diabetic Retinopathy Study 8 (DRS) findings. DRS report number 8. Ophthalmology 1981;88:583– 600. 2. Early photocoagulation for diabetic retinopathy. ETDRS report number 9. Early Treatment Diabetic Retinopathy Study Research Group. Ophthalmology 1991;98:766 – 85. 3. Mohamed Q, Gillies MC, Wong TY. Management of diabetic retinopathy: a systematic review. JAMA 2007;298:902–16. 4. Weinberger D, Ron Y, Lichter H, et al. Analgesic effect of topical sodium diclofenac 0.1% drops during retinal laser photocoagulation. Br J Ophthalmol 2000;84:135–7. 5. Cook HL, Newsom RS, Mensah E, et al. Entonox as an analgesic agent during panretinal photocoagulation. Br J Ophthalmol 2002;86:1107– 8. 6. Esgin H, Samut HS. Topical ketorolac 0.5% for ocular pain relief during scatter laser photocoagulation with 532 nm green laser. J Ocul Pharmacol Ther 2006;22:460 – 4. 7. Vaideanu D, Taylor P, McAndrew P, et al. Double masked randomised controlled trial to assess the effectiveness of paracetamol in reducing pain in panretinal photocoagulation. Br J Ophthalmol 2006;90:713–7. 8. Wu WC, Hsu KH, Chen TL, et al. Interventions for relieving pain associated with panretinal photocoagulation: a prospective randomized trial. Eye 2006;20:712–9. 9. Hay A, Flynn HW Jr, Hoffman JI, Rivera AH. Needle penetration of the globe during retrobulbar and peribulbar injections. Ophthalmology 1991;98:1017–24. 10. Smith JL. Retrobulbar bupivacaine can cause respiratory arrest. Ann Ophthalmol 1982;14:1005– 6. 11. Javitt JC, Addiego R, Friedberg HL, et al. Brain stem anesthesia after retrobulbar block. Ophthalmology 1987;94:718 –24. 12. Stevens JD, Foss AJ, Hamilton AM. No-needle one-quadrant sub-tenon anaesthesia for panretinal photocoagulation. Eye 1993;7:768 –71. 13. Whitacre MM. The effect of transcutaneous electrical nerve stimulation on ocular pain. Ophthalmic Surg 1991;22:462– 6. 14. Skoutakis VA, Carter CA, Mickle TR, et al. Review of diclofenac and evaluation of its place in therapy as a nonsteroidal anti-inflammatory agent. Drug Intell Clin Pharm 1988;22: 850 –9. 15. Gaynes BI, Fiscella R. Topical nonsteroidal anti-inflammatory drugs for ophthalmic use: a safety review. Drug Saf 2002;25: 233–50. 16. Colin J. The role of NSAIDs in the management of postoperative ophthalmic inflammation. Drugs 2007;67:1291–308. 17. Todd PA, Sorkin EM. Diclofenac sodium: a reappraisal of its pharmacodynamic and pharmacokinetic properties, and therapeutic efficacy. Drugs 1988;35:244 – 85. 18. Collins SL, Moore RA, McQuay HJ, et al. Single dose oral ibuprofen and diclofenac for postoperative pain. Cochrane Database Syst Rev 2000(2):CD001548. 19. Barden J, Edwards J, Moore RA, McQuay HJ. Single dose oral diclofenac for postoperative pain. Cochrane Database Syst Rev 2004(2):CD004768. 20. Goa KL, Chrisp P. Ocular diclofenac: a review of its pharmacology and clinical use in cataract surgery, and potential in other inflammatory ocular conditions. Drugs Aging 1992;2:473– 86. 21. Ling TL, Combs DL. Ocular bioavailability and tissue distribution of [14C]ketorolac tromethamine in rabbits. J Pharm Sci 1987;76:289 –94.
1173
Ophthalmology
Volume 116, Number 6, June 2009
22. Anderson JA, Chen CC, Vita JB, Shackleton M. Disposition of topical flurbiprofen in normal and aphakic rabbit eyes. Arch Ophthalmol 1982;100:642–5. 23. Sanders DR, Goldstick B, Kraff C, et al. Aqueous penetration of oral and topical indomethacin in humans. Arch Ophthalmol 1983;101:1614 – 6. 24. Bloom SM, Brucker AJ. Laser Surgery of the Posterior Segment. 2nd ed. Philadelphia: Lippincott; 1997:382. 25. Ohnhaus EE, Adler R. Methodological problems in the measurement of pain: a comparison between the verbal rating scale and the visual analogue scale. Pain 1975;1:379 – 84. 26. Todd KH, Funk KG, Funk JP, Bonacci R. Clinical significance of reported changes in pain severity. Ann Emerg Med 1996; 27:485–9. 27. Kelly AM. Does the clinically significant difference in visual analog scale pain scores vary with gender, age, or cause of pain? Acad Emerg Med 1998;5:1086 –90. 28. Kelly AM. The minimum clinically significant difference in visual analogue scale pain score does not differ with severity of pain. Emerg Med J 2001;18:205–7.
29. Gallagher EJ, Liebman M, Bijur PE. Prospective validation of clinically important changes in pain severity measured on a visual analog scale. Ann Emerg Med 2001;38:633– 8. 30. Diclofenac. Compendium of Pharmaceuticals and Specialties. Toronto: Canadian Pharmacists Association; 2002:1847–51. 31. Kong J, Kaptchuk TJ, Polich G, et al. Placebo analgesia: findings from brain imaging studies and emerging hypotheses. Rev Neurosci 2007;18:173–90. 32. Yunus MB. Role of central sensitization in symptoms beyond muscle pain, and the evaluation of a patient with widespread pain. Best Pract Res Clin Rheumatol 2007;21:481–97. 33. Koay P. The emerging roles of topical non-steroidal antiinflammatory agents in ophthalmology. Br J Ophthalmol 1996; 80:480 –5. 34. Al-Hussainy S, Dodson PM, Gibson JM. Pain response and follow-up of patients undergoing panretinal laser photocoagulation with reduced exposure times. Eye 2008;22:96 –9. 35. Friberg TR, Venkatesh S. Alteration of pulse configuration affects the pain response during diode laser photocoagulation. Lasers Surg Med 1995;16:380 –3.
Footnotes and Financial Disclosures Originally received: July 25, 2008. Final revision: January 14, 2009. Accepted: January 20, 2009. Available online: April 19, 2009.
Manuscript no. 2008-897.
1
Department of Ophthalmology, University of British Columbia, Vancouver, Canada. 2
Department of Ophthalmology, Toronto Western Hospital, University of Toronto, Toronto, Canada. Presented as a poster at: American Academy of Ophthalmology Annual Meeting, November 2008, Atlanta, Georgia.
1174
Financial Disclosure(s): The author(s) have no proprietary or commercial interest in any materials discussed in this article. Supported by the Academic Enrichment Fund, Department of Ophthalmology, Toronto Western Hospital/University Health Network, University of Toronto, Toronto, Canada. The funding organization had no role in the design or conduct of this research. Correspondence: Wai-Ching Lam, MD, FRCSC, Toronto Western Hospital, 399 Bathurst Street, East Wing 6-432, Toronto, ON, Canada, M5T 2S8. E-mail: waiching. [email protected].
Panretinal photocoagulation (PRP) is an effective treatment for reducing the risk of vision loss in proliferative diabetic retinopathy (PDR).1–3 Patients frequently experience pain during the procedure, which typically is in the moderate range.4 – 8 Preprocedure or rescue retrobulbar or peribulbar injection anesthesia can be used to reduce pain during the procedure, but it is invasive and imparts the risk of serious, although rare, complications.9 –11 Other pre-PRP analgesic techniques investigated include sub-Tenon anesthesia delivery,12 transcutaneous electrical nerve stimulation,13 and inhalation of Entonox (50:50 nitrous oxide:oxygen gas)5; however, these methods are limited by their need for special equipment, invasiveness, or both. Various medications that are easily administered and impart only minimal risk have been studied as pretreatment agents, including nonsteroidal anti-inflammatory drugs (NSAIDs; topical diclofenac,4 topical ketorolac,6 oral mefenamic acid,8 and intramuscular ketorolac8), oral acetaminophen,7,8 and oral diazepam.8 Among these medications, only topical diclofenac has demonstrated statistically significant pain reduction during PRP.4
1168
© 2009 by the American Academy of Ophthalmology Published by Elsevier Inc.
Nonsteroidal anti-inflammatory drugs comprise a family of compounds with proven analgesic, anti-inflammatory, and antipyretic effects that can be used systemically or locally. They produce their clinical effects mainly by inhibiting the cyclooxygenase pathway, which leads to reduction in prostaglandin formation, a potent mediator of the pain pathway.14 –16 Diclofenac is a commonly used NSAID available in both oral and topical forms. The oral form is a potent analgesic agent effective in many conditions,14,17 including postoperative pain when given as a single dose.18,19 The topical form is indicated for postoperative inflammation after cataract surgery and for relief of pain and photophobia after corneal refractive surgery.15,16 Oral NSAIDs reach targeted tissues via the systemic circulation, whereas topical NSAIDs directly penetrate the cornea to reach intraocular tissues.20 The use of topical NSAIDs as analgesics in posterior segment procedures has not been well studied. In rabbit models, after topical administration, the NSAIDs ketorolac and flurbiprofen were found throughout the eye, including the choroid and retina; howISSN 0161-6420/09/$–see front matter doi:10.1016/j.ophtha.2009.01.022
Zakrzewski et al 䡠 Oral vs. Topical Diclofenac for Pain Prevention during PRP ever, lower levels were found in the vitreous than in the aqueous humor.21,22 After preoperative administration, the topical form of the NSAID indomethacin was shown to have higher aqueous levels than its oral form.23 It is unknown whether oral or topical NSAID administration would result in higher concentrations at the level of the long posterior ciliary nerves in the suprachoroidal space, the purported site of pain generation during PRP.24 No studies to date have directly compared the oral and topical forms of an NSAID in reducing pain during PRP. The purpose of this study was to investigate the analgesic effects of pretreatment oral and topical diclofenac during and after PRP for PDR.
Patients and Methods This prospective, randomized, double-masked, 3-armed placebocontrolled trial was carried out using consecutive patients from the practice of 1 vitreoretinal specialist (WCL) in the Department of Ophthalmology and Visual Sciences at the University of Toronto between November 2003 and March 2007. Ethics approval was granted by the University of Toronto University Health Network Research Ethics Board. Approval for the clinical trial was granted by the Health Canada (federal government) Therapeutic Products Directorate, which is where the trial also was formally registered and is equivalent to an Investigational New Drug application process in the United States. Informed consent was obtained from all patients who participated in the study. The research described herein adhered to the tenets of the Declaration of Helsinki. Literature searches were carried out in the PubMed and MEDLINE databases in June 2002 and March 2008 to uncover all previously published articles describing pain modification techniques for PRP using the search terms panretinal photocoagulation, laser photocoagulation, and pain. Inclusion criteria were the diagnosis of PDR requiring PRP treatment for the first time and the ability to understand the nature of the study and to complete all measurement instruments. Exclusion criteria were previous PRP in either eye and contraindications to oral NSAIDS, including pregnancy, history of peptic ulcer disease, bleeding disorders, asthma, or renal failure; known allergy or hypersensitivity to NSAIDs or acetylsalicylic acid; and current use of any narcotic, NSAID (except acetylsalicylic acid), digitalis, lithium, probenecid, methotrexate, phenytoin, or cyclosporine. Age and gender were collected as demographic data. Only 1 eye per patient was enrolled in the study. Patients were randomized into 1 of 3 study groups: (1) oral diclofenac potassium (immediate release) 50 mg (Voltaren Rapide; Novartis International AG, Basel, Switzerland) plus topical placebo; (2) oral placebo plus nonpreserved topical diclofenac sodium 0.1% (Voltaren Ophtha, Novartis International AG); or (3) oral placebo plus topical placebo. The oral diclofenac and oral placebo were prepared in unmarked capsules and were indistinguishable in appearance and odor from each other. The topical placebo used was Bion Tears (Alcon, Inc., Fort Worth, TX), which is a nonpreserved artificial tear available in single-dose packages. With stickers covering the product labels, these single-dose packages were indistinguishable in appearance from the topical diclofenac single-dose packages used in the study. Patients receiving topical diclofenac may have experienced more stinging on administration than placebo. Recruited patients were randomized (by electronic number generation) in blocks of 6 to 1 of the 3 study groups. Study medication packages containing the appropriate oral and topical products were premade by a hospital pharmacist, and none of the treating physicians, study nurses, or patients knew what study
group they were part of. The randomization code linking patients to study groups was stored locked in the hospital pharmacist’s office and was not broken until all patients had completed testing. A total of 14 physicians (1 staff physician, 13 ophthalmology fellows and residents) took part in the enrollment, PRP treatment, and pain measurement testing of study patients. All study patients were treated in the same manner and all medications were administered by the study nurse. Pupillary dilation was achieved with 1 drop each successively of proparacaine 0.5% (Alcaine; Alcon, Inc.), tropicamide 1% (Mydriacyl; Alcon, Inc.), and phenylephrine 2.5% (Mydfrin; Alcon, Inc.) into the study eye a minimum of 30 minutes before PRP. Repeat administration of dilating drops was allowed in cases of inadequate pupillary dilation. Both the oral study medication (given with water) and 1 drop of the topical study medication were given 60 minutes before the onset of PRP, followed by a second drop of the topical study medication 5 to 10 minutes before the PRP was started. Another drop of proparacaine was given just before PRP to facilitate the placement of the fundus contact lens used to guide laser administration. The PRP was performed using an argon laser with settings based on individual user preferences, with a maximum of 600 laser burns allowed during the treatment session. One of yellow, yellowgreen, or green wavelengths was used; spot sizes ranged from 200 to 500 m; and laser durations ranged from 0.07 to 0.15 seconds. The energy per burn (mW) and total number of burns were recorded for each patient, allowing for the calculation of total laser energy delivered (total number of burns ⫻ energy per burn) during each PRP session. The treating physician was asked to record if rescue injection anesthesia was required to complete the PRP. Immediately after the PRP session, study patients were asked to rate the level of pain they had felt during the PRP treatment session on the visual analog scale (VAS), a pain measurement tool previously validated as a reliable, sensitive measure of pain and change in pain.25 It consists of a 100-mm line anchored at one end by the label no pain at all and at the other end by the label the worst pain imaginable. Patients were asked to mark the line with an X to indicate pain intensity. The position of the X was measured by a masked third party and was converted to a score from 0 to 100. Fifteen minutes after the completion of the PRP treatment, study patients were asked to rate their current level of pain on a second VAS and to report any ocular or systemic side effects that were not present before the start of the study. A checklist of expected possible side effects was provided, and space also was given for subjects to write in any side effects being experienced that were not on the checklist. When study patients returned to the clinic for their second PRP treatment session in the study eye, they were not given any pretreatment medications except for the topical proparacaine and dilating drop regimen as described previously. The range of laser settings used was the same as above, and the number of burns again was limited to 600. The energy per burn and the number of burns were recorded, again allowing for the calculation of total laser energy delivered. After this second PRP session, study patients were asked to rate the level of pain experienced during the PRP on a third VAS. The primary outcome measure was the VAS pain score for each group during the first PRP treatment session. Secondary outcome measures were the VAS pain scores 15 minutes after the first PRP session, VAS pain scores during the second PRP treatment, the need for injection anesthesia during the first PRP session, and ocular and systemic side effects experienced during the first session. All study data were masked and were analyzed only at the completion of the study by a noninvolved third party. During the design of the study, a 3-way analysis of variance power calculation (80% power, 5% significance level) determined
1169
Ophthalmology
Volume 116, Number 6, June 2009
Table 1. Demographics and Mean Total Laser Energy in Each Panretinal Photocoagulation Session for Each Study Group Group 1
Group 2
Group 3
Mean age (yrs)⫾SD 59.8⫾12.7 59.0⫾14.6 56.6⫾16.9 Male:female (% male) 20:10 (66.7) 21:10 (67.7) 17:12 (58.6) First PRP mean total 177.1⫾63.3 193.8⫾119.4 188.3⫾82.8 energy (mW/1000)⫾SD Second PRP mean 179.8⫾61.5 178.0⫾65.4 157.3⫾59.1 total energy (mW/1000)⫾SD
P Value 0.71 0.73 0.82 0.35
PRP ⫽ panretinal photocoagulation; SD ⫽ standard deviation.
that 30 patients per study group would be needed to detect a 13-point statistically significant difference in mean VAS pain scores between groups. A 13-point difference was chosen because it represented the most conservative value within the range of published minimum VAS pain score differences (9 –13) required to achieve clinical significance.26 –29 The 1-way analysis of variance test was used to compare differences in mean patient ages and total laser energies between groups, and a chi-square test was used to compare gender proportions. Mean VAS pain scores were compared between groups with the t test. Multivariate regression analysis of pain scores for age, gender, and total laser energy was performed assuming exponential distribution with the parameters estimated by the maximum likelihood method (PROC GENMOD), where P values were generated by the chi-square test. To allow for proper statistical regression analysis, it was necessary to avoid pain scores with a value of 0; thus, a transformation was used with all new pain scores equaling the old pain scores plus 1. The Fisher exact test was used to analyze differences in side-effect proportions between groups experienced during the first PRP session. All statistical analyses were performed using the SAS software system, version 9.1 (SAS Inc., Cary, NC).
Results A total of 90 patients were recruited into the study, with 30 in the oral diclofenac group, 31 in the topical diclofenac group, and 29 in
the placebo group. No patients changed groups at any time during the study. All 90 patients took their appropriate pretreatment medications (and thus were analyzed as intention to treat), underwent their first PRP treatment, and completed the 2 pain questionnaires corresponding to the first PRP session. Three patients (1 from each study group) did not have pain questionnaires completed for a second PRP session. Two of these patients underwent a second PRP treatment, but erroneously were not asked to complete a pain questionnaire, whereas the third patient had a vitreous hemorrhage and did not undergo follow-up PRP. The other 87 patients underwent a second PRP session and completed the corresponding pain questionnaire. The range of elapsed time between the first and second PRP treatments was 2 days to 15 weeks; however, for 80 of the 87 patients, the elapsed time was between 1 and 5 weeks. Six patients underwent the second PRP less than 1 week after the first, and 1 underwent the second PRP more than 5 weeks after the first (15 weeks). The mean age, gender proportion, and total laser energies for each PRP session for each study group are presented in Table 1. None of the differences between groups were statistically significant. Mean VAS pain scores for each group during the first PRP treatment session were: oral diclofenac, 25.7⫾19.9; topical diclofenac, 33.8⫾27.9; and placebo, 41.3⫾31.0 (Fig 1A). The lower pain rating for oral diclofenac compared with placebo was statistically significant (P ⫽ 0.02), whereas differences between oral diclofenac and topical diclofenac (P ⫽ 0.20) and between topical diclofenac and placebo (P ⫽ 0.33) were not significant. To account for possible effects on pain levels by differences in age, gender proportion, and total laser energy between groups, a multivariate regression analysis of the VAS pain scores for these variables was performed. It demonstrated statistically significant pain reductions for oral diclofenac (P ⫽ 0.01) and topical diclofenac (P⬍0.0001) versus placebo, but no difference between oral and topical diclofenac (P ⫽ 0.67). The mean VAS pain scores 15 minutes after the completion of the first PRP session were: oral diclofenac, 8.8⫾9.7; topical diclofenac, 10.1⫾13.9; and placebo, 15.2⫾18.5 (Fig 1B). None of the pain score differences between groups were statistically significant (oral diclofenac vs. placebo, P ⫽ 0.10; oral vs. topical diclofenac, P ⫽ 0.67; topical diclofenac vs. placebo, P ⫽ 0.24). Comparison between groups with multivariate regression for age, gender, and total laser energy also failed to demonstrate differences between groups (oral diclofenac vs. placebo, P ⫽ 0.18; topical diclofenac vs. placebo, P ⫽ 0.26; oral vs. topical diclofenac, P ⫽ 0.85).
Figure 1. Bar graphs showing mean visual analog scale (VAS) pain scores (bracketed by standard errors) for each group (A) during the first panretinal photocoagulation (PRP), (B) 15 minutes after the first PRP, and (C) during the second PRP.
1170
Zakrzewski et al 䡠 Oral vs. Topical Diclofenac for Pain Prevention during PRP Mean VAS pain levels reported during the second PRP laser session (with no pretreatment medications) were: oral diclofenac, 38.5⫾27.0; topical diclofenac, 39.5⫾29.1; and placebo, 54.5⫾27.9 (Fig 1C). Compared with placebo, the lower pain ratings for both oral diclofenac (P ⫽ 0.03) and topical diclofenac (P ⫽ 0.04) were significant, whereas there was no significant difference between oral and topical diclofenac (P ⫽ 0.88). Comparison between groups with the multivariate regression analysis for age, gender, and total laser energy demonstrated statistically less pain for oral diclofenac versus placebo (P ⫽ 0.007), but no differences between topical diclofenac and placebo (P ⫽ 0.07) and oral and topical diclofenac (P ⫽ 0.78). For each study group, comparisons were made between VAS pain scores during the first PRP and those recorded 15 minutes after treatment and between pain scores for the first and second PRP sessions. All 3 study groups demonstrated statistically significant reductions in pain at 15 minutes after treatment compared with during the first treatment (oral diclofenac, P ⫽ 0.0001; topical diclofenac, P⬍0.0001; placebo, P ⫽ 0.0003). Both the oral diclofenac (P ⫽ 0.02) and placebo (P ⫽ 0.05) groups demonstrated an increase in pain from the first to the second PRP. The difference between the 2 PRP sessions for the topical diclofenac group was not significant (P ⫽ 0.33). Regression analysis accounting for total energy differences between the 2 PRP sessions found the pain increase to be statistically significant for oral diclofenac (P ⫽ 0.05), but not for topical diclofenac (P ⫽ 0.54) or placebo (P ⫽ 0.12). One patient in the study (topical diclofenac group) required a rescue retrobulbar anesthesia injection to complete the first PRP treatment session. This patient’s reported VAS pain scores during and 15 minutes after the first PRP session were 99 and 3.5, respectively. For the second PRP session, the patient reported a pain score of 77, but did not require injection anesthesia to complete treatment. Side effects reported by each study group for the first PRP treatment session are shown in Table 2. Although there were more headaches, less eye stinging, and less photophobia in the oral diclofenac group compared with the topical diclofenac and placebo groups, none of the differences reached statistical significance.
Discussion This study sought to investigate the level of analgesia provided by the oral and topical forms of the NSAID diclofenac when given before PRP for PDR. The idea for the study was contrived because patients typically experience pain during PRP, but there is no consensus favoring any particular pretreatment analgesic agent. In this study, plus the 4 published randomized trials that investigated pretreatment analgesic medications for PRP using the VAS scale, mean placebo group pain scores range from 37.3 to 53.1,4,6 – 8 which falls in the moderate range of pain (31– 69).27 The most common descriptions of pain experienced during PRP included sharp, flashing, pricking, tiring, and piercing.7 Finding a safe, inexpensive, and easy-to-administer method to reduce the pain intensity felt during PRP would make the treatment more comfortable for patients, likely would decrease the need for rescue injection anesthesia, and may improve compliance with follow-up PRP treatments. Oral and topical diclofenac were chosen for this study because oral diclofenac is effective in reducing postoperative pain from various procedures when given as a single preoperative dose18 and topical diclofenac was the only
Table 2. Number of Subjects in Each Group Reporting Side Effects for the First Panretinal Photocoagulation Treatment Session Side Effect Headache Eye stinging Dizziness/lightheadedness Fatigue Photophobia Nausea Agitated/restless Eye itchiness Heartburn Other† Eye pain Eye stickiness Seeing colors Tearing Dry eye Facial pain
Group 1 (n ⴝ 30)
Group 2 (n ⴝ 31)
Group 3 (n ⴝ 29)
8 5 3 3 1 1 0 0 0
4 (P ⫽ 0.21)* 11 (P ⫽ 0.15)* 2 2 6 (P ⫽ 0.10)* 0 2 1 0
4 (P ⫽ 0.33)* 8 (P ⫽ 0.36)* 3 1 5 (P ⫽ 0.10)* 2 3 0 1
2 1 1 0 0 0
0 0 0 1 1 0
3 0 0 0 0 1
*P values represent statistical analysis with the Fisher exact test of the differences in proportions between group 1 and groups 2 or 3. † Indicates a symptom written in by subject(s) that was not offered on the questionnaire checklist.
medication found to decrease pain during PRP when given before treatment.4 The immediate-release form of oral diclofenac was used because it has the fastest time of onset (detectable in plasma within 10 minutes of ingestion with a peak plasma concentration between 20 and 60 minutes after dose),30 allowing for the least required time lag between drug administration and onset of PRP. Two doses of topical diclofenac were given (at 60 minutes and 5–10 minutes before treatment) to ensure that both the short-term and long-term potential analgesic effects afforded by the medication were accounted for in the study. The results from the study demonstrated that pretreatment oral diclofenac provides statistically significant pain reduction during PRP treatment for PDR compared with placebo. This finding was evident when the raw study data were analyzed and also when a multivariate regression for age, gender, and total laser energy was performed. Topical diclofenac, however, produced statistically significant pain reduction versus placebo with the multivariate regression, but not when the raw study data was analyzed without the regression analysis. Differences between oral and topical diclofenac were not statistically significant with either analysis. The multivariate regression analysis was carried out in an attempt to remove possible confounding of pain perception resulting from differences in age, gender, and total laser energy between study groups. To what extent each of these variables had on pain levels recorded in this study is unknown. Both of the statistical analysis techniques (multivariate regression or raw data alone without the regression) are important in interpreting the study results; thus, both techniques are reported throughout the study. The 2 techniques should be evaluated together when making conclusions; however, the presence or absence of clinical signifi-
1171
Ophthalmology
Volume 116, Number 6, June 2009
cance also is important to consider. For example, topical diclofenac was statistically superior to placebo in pain reduction during the first PRP with multivariate regression, but not with raw data alone. This makes interpretation difficult; however, the pain score difference between the groups was not clinically significant, so the overall conclusion was that topical diclofenac was not superior to placebo. Determining the required difference in VAS pain scores between groups to constitute actual clinical significance is based on several published studies on the issue. The published minimum VAS pain score differences found to reflect clinical significance ranges between 9 and 13 on the 0 to 100 scale.26 –29 Although these studies took place in an emergency room setting measuring changes in acute pain over time for individual patients, these values are the best available predictor of the pain score differences necessary in this study to detect clinical significance. Examining the pain levels for the first PRP, the pain score difference of 15.6 between oral diclofenac and placebo achieved clinical significance, whereas differences between topical diclofenac and placebo (7.5) and between oral and topical diclofenac (8.1) did not. Interestingly, the VAS pain score difference between topical diclofenac and placebo in this study was similar to that in the study by Weinberger et al4 (8.9). In that study, the difference was found to be statistically significant; however, using the published range of minimum required pain score differences of 9 to 13, it does not quite reach clinical significance. At 15 minutes after the first PRP, pain scores for all 3 groups were in the mild range (⬍31),27 with the highest being 15.2 (placebo). None of the differences in pain scores between groups at this time point were significant. Meanwhile, pain reductions for the 3 groups from during the first PRP to 15 minutes after were all clinically and statistically significant. This implies that pain experienced during PRP diminishes fairly rapidly to a mild level, although the findings cannot be extrapolated past 15 minutes. Vaideanu et al7 found the mean pain score for the placebo group at 24 hours after PRP to be approximately half of that immediately after the laser treatment (VAS scores of 27 and 52, respectively). This amount of pain at 24 hours is almost double the placebo group pain score 15 minutes after laser treatment (15.2), a difference that may be explained by heterogeneity of study populations, or possibly a later increase in laserinduced pain. Further study is required to determine what happens to pain levels between 15 minutes and 24 hours after treatment. Comparing pain scores between the 2 PRP sessions, the statistically higher pain score for oral diclofenac during the second PRP likely is explained by the lack of pretreatment medication given before the second PRP. For the placebo group, there are 3 possible explanations for why the second PRP pain score was higher than the first. First, subjects received placebo oral and topical medications before the first PRP, but not the second; thus, the placebo effect would have been present only for the first PRP. The placebo effect is a well-studied phenomenon and is believed to alter pain perception in 3 ways: (1) changing the prestimulus expectation of pain relief; (2) modifying the actual level of pain experienced; and (3) distorting the poststimulus pain rat-
1172
ing.31 Second, subjects might have experienced sufficient pain during the first PRP session to develop a central sensitization of pain associated with the procedure, which might have led to a heightened perception of pain experienced during the second PRP treatment.32 Third, the pain score differences between the 2 PRP sessions were not significant for the placebo group when a regression analysis accounting for differences in total laser energy between the 2 sessions was performed. This implies that the pain differences seen with the raw data may be the result of the higher mean laser energy delivered during the second PRP compared with the first. With respect to safety, no significant differences were found between the groups for any particular side effect. No serious safety concerns were identified, although patients were monitored only to 15 minutes after PRP. Using a single dose of oral diclofenac or 2 doses of topical diclofenac would be very unlikely to cause any serious adverse events. A large review study analyzed 528 patients receiving a single dose of oral diclofenac for postoperative pain and found no more side effects with diclofenac than placebo, with neither experiencing any serious side effects.18 Side effects resulting from topical diclofenac typically are local and benign and include stinging, conjunctival hyperemia, and reversible keratitis.15,33 Of the 177 total PRP sessions in this study, only 1 required rescue injection anesthesia to complete laser treatment. Although effective in reducing pain during PRP,8 injection anesthesia carries the risk of rare, although serious, complications of brainstem anesthesia,11 respiratory arrest,10 and globe perforation.9 Thus, the safest course of action likely would be to use injection anesthesia only as a rescue procedure when pain is severe enough to compromise completion of PRP treatment. Besides pretreatment analgesia for lower pain experiences during PRP, researchers also have begun investigating different laser parameters in an attempt to reduce pain. Pain during PRP is believed to arise from the photocoagulation of ciliary nerves running in the suprachoroidal space, and more pain typically is experienced with more anterior application and along the horizontal meridians.4,24 Avoiding these areas, if possible, may lead to decreased pain. A small study (20 patients) comparing short laser exposure PRP (0.02 seconds) versus standard exposure (0.1 seconds) with a 532-nm, frequency-doubled, neodymium:yttrium–aluminum– garnet laser found statistically less pain in the shortexposure group.34 Another study compared different diode laser pulse shapes for PRP and demonstrated more pain in the standard square-wave mode than in the altered shapedwave and micropulse modes.35 Combining optimal laser parameters with effective pretreatment analgesia may result in even lower pain levels during PRP than either method alone. It is important to consider the waiting period between the administration of oral diclofenac and the PRP treatment when taking into account pain reduction for patients in a real clinic setting. Based on half-life data for the medication, a waiting period of 60 minutes was chosen. It is unknown what effect a shorter or longer waiting time would has on pain reduction; thus, the best chance to duplicate the results from this study would be to use a 60-minute waiting
Zakrzewski et al 䡠 Oral vs. Topical Diclofenac for Pain Prevention during PRP period between medication administration and PRP application. In some clinic settings, PRP treatments are given immediately after the diagnosis of PDR. In these settings, adding a waiting period may not be practical and any benefit from pain reduction may be outweighed by an increased patient frustration from having a longer appointment time. These issues must be considered when deciding whether to incorporate such a pain reduction strategy in each individual clinic. A limitation of the study is that patients receiving topical diclofenac may have experienced minor stinging, leading them to believe they were receiving actual medication. What effect this may have had on recorded pain levels is unknown. Another potential limitation was the large number of treating physicians, which was an unavoidable consequence in a busy vitreoretinal practice in an academic setting. Each treating physician likely dealt with patient reports of pain differently, with the content and nature of their responses possibly leading to an alteration of patients’ perceived and recorded levels of pain. Each treating physician also likely had his or her own end point for what constitutes a sufficient PRP laser burn, which might have led to differences in the total amount of laser energy delivered per session and might have affected the level of pain experienced. The authors attempted to correct for this with regression analyses. Another limitation of the study is the variable laser settings used by each treating physician. At the time the study was designed, there were no published studies examining the effect of laser exposure times on pain levels. As mentioned before, a shorter laser exposure time was found to correlate with lower pain levels during PRP with the neodymium:yttrium–aluminum– garnet laser.35 It is unclear whether this finding can be extrapolated to the argon laser used in this study, and even if so, whether these study results would have been altered by any differences in laser exposure times between groups. Whether the analgesic benefit of oral diclofenac found in this study for PRP can be extrapolated to other oral NSAIDs is not known. In the study by Wu et al,8 pre-PRP administration of oral mefenamic acid, an NSAID, did not provide pain reduction compared with the control group. However, a direct comparison of single doses of oral diclofenac versus oral ibuprofen for postoperative pain found no significant analgesic differences between the 2, with both demonstrating significant pain reductions compared with placebo.18 No studies to date have compared ibuprofen with either placebo or diclofenac for pain reduction during PRP. In conclusion, this study demonstrated that a single pretreatment dose of oral diclofenac provides clinically and statistically significant pain reduction during PRP for PDR. Although pretreatment topical diclofenac provided pain reduction that reached statistical significance using a multivariate regression analysis, the actual amount of pain reduction was not clinically significant. Based on the results of this study, the principal author (WCL) now offers all patients requiring PRP the option of a pretreatment dose of oral diclofenac potassium 50 mg provided that there are no contraindications to oral NSAIDs (listed in “Methods” as study exclusion criteria).
References 1. Diabetic Retinopathy Study Research Group. Photocoagulation treatment of 7 proliferative diabetic retinopathy: clinical application of Diabetic Retinopathy Study 8 (DRS) findings. DRS report number 8. Ophthalmology 1981;88:583– 600. 2. Early photocoagulation for diabetic retinopathy. ETDRS report number 9. Early Treatment Diabetic Retinopathy Study Research Group. Ophthalmology 1991;98:766 – 85. 3. Mohamed Q, Gillies MC, Wong TY. Management of diabetic retinopathy: a systematic review. JAMA 2007;298:902–16. 4. Weinberger D, Ron Y, Lichter H, et al. Analgesic effect of topical sodium diclofenac 0.1% drops during retinal laser photocoagulation. Br J Ophthalmol 2000;84:135–7. 5. Cook HL, Newsom RS, Mensah E, et al. Entonox as an analgesic agent during panretinal photocoagulation. Br J Ophthalmol 2002;86:1107– 8. 6. Esgin H, Samut HS. Topical ketorolac 0.5% for ocular pain relief during scatter laser photocoagulation with 532 nm green laser. J Ocul Pharmacol Ther 2006;22:460 – 4. 7. Vaideanu D, Taylor P, McAndrew P, et al. Double masked randomised controlled trial to assess the effectiveness of paracetamol in reducing pain in panretinal photocoagulation. Br J Ophthalmol 2006;90:713–7. 8. Wu WC, Hsu KH, Chen TL, et al. Interventions for relieving pain associated with panretinal photocoagulation: a prospective randomized trial. Eye 2006;20:712–9. 9. Hay A, Flynn HW Jr, Hoffman JI, Rivera AH. Needle penetration of the globe during retrobulbar and peribulbar injections. Ophthalmology 1991;98:1017–24. 10. Smith JL. Retrobulbar bupivacaine can cause respiratory arrest. Ann Ophthalmol 1982;14:1005– 6. 11. Javitt JC, Addiego R, Friedberg HL, et al. Brain stem anesthesia after retrobulbar block. Ophthalmology 1987;94:718 –24. 12. Stevens JD, Foss AJ, Hamilton AM. No-needle one-quadrant sub-tenon anaesthesia for panretinal photocoagulation. Eye 1993;7:768 –71. 13. Whitacre MM. The effect of transcutaneous electrical nerve stimulation on ocular pain. Ophthalmic Surg 1991;22:462– 6. 14. Skoutakis VA, Carter CA, Mickle TR, et al. Review of diclofenac and evaluation of its place in therapy as a nonsteroidal anti-inflammatory agent. Drug Intell Clin Pharm 1988;22: 850 –9. 15. Gaynes BI, Fiscella R. Topical nonsteroidal anti-inflammatory drugs for ophthalmic use: a safety review. Drug Saf 2002;25: 233–50. 16. Colin J. The role of NSAIDs in the management of postoperative ophthalmic inflammation. Drugs 2007;67:1291–308. 17. Todd PA, Sorkin EM. Diclofenac sodium: a reappraisal of its pharmacodynamic and pharmacokinetic properties, and therapeutic efficacy. Drugs 1988;35:244 – 85. 18. Collins SL, Moore RA, McQuay HJ, et al. Single dose oral ibuprofen and diclofenac for postoperative pain. Cochrane Database Syst Rev 2000(2):CD001548. 19. Barden J, Edwards J, Moore RA, McQuay HJ. Single dose oral diclofenac for postoperative pain. Cochrane Database Syst Rev 2004(2):CD004768. 20. Goa KL, Chrisp P. Ocular diclofenac: a review of its pharmacology and clinical use in cataract surgery, and potential in other inflammatory ocular conditions. Drugs Aging 1992;2:473– 86. 21. Ling TL, Combs DL. Ocular bioavailability and tissue distribution of [14C]ketorolac tromethamine in rabbits. J Pharm Sci 1987;76:289 –94.
1173
Ophthalmology
Volume 116, Number 6, June 2009
22. Anderson JA, Chen CC, Vita JB, Shackleton M. Disposition of topical flurbiprofen in normal and aphakic rabbit eyes. Arch Ophthalmol 1982;100:642–5. 23. Sanders DR, Goldstick B, Kraff C, et al. Aqueous penetration of oral and topical indomethacin in humans. Arch Ophthalmol 1983;101:1614 – 6. 24. Bloom SM, Brucker AJ. Laser Surgery of the Posterior Segment. 2nd ed. Philadelphia: Lippincott; 1997:382. 25. Ohnhaus EE, Adler R. Methodological problems in the measurement of pain: a comparison between the verbal rating scale and the visual analogue scale. Pain 1975;1:379 – 84. 26. Todd KH, Funk KG, Funk JP, Bonacci R. Clinical significance of reported changes in pain severity. Ann Emerg Med 1996; 27:485–9. 27. Kelly AM. Does the clinically significant difference in visual analog scale pain scores vary with gender, age, or cause of pain? Acad Emerg Med 1998;5:1086 –90. 28. Kelly AM. The minimum clinically significant difference in visual analogue scale pain score does not differ with severity of pain. Emerg Med J 2001;18:205–7.
29. Gallagher EJ, Liebman M, Bijur PE. Prospective validation of clinically important changes in pain severity measured on a visual analog scale. Ann Emerg Med 2001;38:633– 8. 30. Diclofenac. Compendium of Pharmaceuticals and Specialties. Toronto: Canadian Pharmacists Association; 2002:1847–51. 31. Kong J, Kaptchuk TJ, Polich G, et al. Placebo analgesia: findings from brain imaging studies and emerging hypotheses. Rev Neurosci 2007;18:173–90. 32. Yunus MB. Role of central sensitization in symptoms beyond muscle pain, and the evaluation of a patient with widespread pain. Best Pract Res Clin Rheumatol 2007;21:481–97. 33. Koay P. The emerging roles of topical non-steroidal antiinflammatory agents in ophthalmology. Br J Ophthalmol 1996; 80:480 –5. 34. Al-Hussainy S, Dodson PM, Gibson JM. Pain response and follow-up of patients undergoing panretinal laser photocoagulation with reduced exposure times. Eye 2008;22:96 –9. 35. Friberg TR, Venkatesh S. Alteration of pulse configuration affects the pain response during diode laser photocoagulation. Lasers Surg Med 1995;16:380 –3.
Footnotes and Financial Disclosures Originally received: July 25, 2008. Final revision: January 14, 2009. Accepted: January 20, 2009. Available online: April 19, 2009.
Manuscript no. 2008-897.
1
Department of Ophthalmology, University of British Columbia, Vancouver, Canada. 2
Department of Ophthalmology, Toronto Western Hospital, University of Toronto, Toronto, Canada. Presented as a poster at: American Academy of Ophthalmology Annual Meeting, November 2008, Atlanta, Georgia.
1174
Financial Disclosure(s): The author(s) have no proprietary or commercial interest in any materials discussed in this article. Supported by the Academic Enrichment Fund, Department of Ophthalmology, Toronto Western Hospital/University Health Network, University of Toronto, Toronto, Canada. The funding organization had no role in the design or conduct of this research. Correspondence: Wai-Ching Lam, MD, FRCSC, Toronto Western Hospital, 399 Bathurst Street, East Wing 6-432, Toronto, ON, Canada, M5T 2S8. E-mail: waiching. [email protected].