POINT: THE CASE FOR CYCLOPENTOLATE
Paula McDowell, OD, FAAO
In a standard annual eye examination with dilation, providers have a wide variety of dilating and cycloplegic agents from which to choose. Tropicamide or cyclopentolate of varying percentages (0.5% and 1%) are the most common in pediatric optometry practices, but there is poor agreement among providers as to which drop or combination should be used for baseline and subsequent examinations. Both the American Optometric Association and the American Academy of Ophthalmology recommend 1% cyclopentolate for pediatric examinations in their evidence-based clinical practice pediatric examination guidelines and pediatric Preferred Practice Patterns, respectively.1,2 Specifically, the American Academy of Ophthalmology recommends 1% cyclopentolate for children aged 1 to 12 years,1 and the American Optometric Association recommends tropicamide only as a substitute in nonstrabismic children when cyclopentolate is not available or contraindicated.2 The benefit of any cycloplegic agent is the short-term paralysis of the accommodative system, allowing the clinician to better analyze the refractive error without the complicating factor of input from fluctuating accommodation. Each agent has a varying time of onset and strength, with cyclopentolate having a longer onset and recovery but less residual accommodation, and tropicamide having a faster onset and recovery but more residual accommodation.3
Use of Cyclopentolate
Although much literature uses cyclopentolate as the standard for pediatric research, anecdotal evidence suggests that practitioners are hesitant to embrace its use in daily practice. Cyclopentolate takes longer to obtain its full cycloplegic effect compared with other drops and has a higher frequency of side effects compared with other agents such as tropicamide. Despite these limitations, there are several reasons why cyclopentolate should be used as the standard drop for baseline pediatric eye examinations. The greatest benefit to cyclopentolate compared with tropicamide is the more effective reduction of accommodation, which is important for a variety of reasons when evaluating pediatric patients.
The main purpose of the initial pediatric examination is to rule out amblyogenic risk factors such as high refractive error, anisometropia, or strabismus. To fully understand a patient’s risk for amblyopia, a clinician must be aware of the comprehensive refractive, accommodative, and binocularity data that are available for that patient. There are multiple recommendations and pediatric guidelines that outline the risks of developing anisometropic or isoametropic refractive amblyopia.2,4 Using 1% cyclopentolate allows the clinician to determine the total refractive error without the influence of accommodation. The clinician can then have confidence in the diagnosis of refractive amblyopia, particularly in hyperopes who make up the highest prevalence of this condition.5 Comparing cycloplegic refractive error to the dry refractive findings allows the clinician to evaluate the full clinical picture of how the accommodative and binocular vision systems may be influenced by uncorrected refractive error. For instance, a patient presenting with intermittent esotropia may benefit from a hyperopic prescription and determining the full magnitude of hyperopia will influence discussions with patients about management options and expected prognosis. There is sometimes debate about how well cyclopentolate and tropicamide inhibit ocular accommodation. Some studies suggest that tropicamide may allow up to 6.5 D of residual accommodation following instillation, with cyclopentolate only up to 1.75 D.3,6 When considering the entire clinical picture and interaction between ocular alignment and refractive error, this is a clinically significant difference. The concern is that pediatric patients with high hyperopia and strong accommodative ability may be more likely to overaccommodate in the presence of a milder cycloplegic agent such as tropicamide. They may also be less reliable historians than adults, making it difficult to appreciate how subtle symptoms may indicate an overactive accommodative system secondary to latent hyperopia.
The Pediatric Eye Disease Investigator Group Amblyopia Treatment Studies have defined a standard treatment approach for patients with amblyopia.7,8 Protocols are agreed upon by the consensus of expert pediatric optometrists and ophthalmologists who work in academic, hospital-based, and other community settings. For the Amblyopia Treatment Studies, study participants must have a baseline cycloplegic examination (specified by 1% cyclopentolate) on which the glasses prescription is finalized. Hyperopic refractive error can be reduced bilaterally up to a diopter and a half, based on the cycloplegic refraction. A full cycloplegic examination allows clinicians to more accurately capture the full amount of anisometropia, so that the final prescription determined from that examination equalizes the accommodative stimulus. If a weaker agent is used and the reduced hyperopic prescription is based on an undercorrected evaluation, the final prescription may be well below the 1.5 D standard and may also result in inaccuracies in anisometropia. The Amblyopia Treatment Studies have shown that amblyopia resolves with glasses correction alone (up to 33% of anisometropic amblyopes, and up to 74% of isoametropic amblyopes),9 and having an accurate cycloplegic refraction as the baseline is critical to determine that prescription.
The Amblyopia Treatment Studies are not the only clinical trials that use cyclopentolate when evaluating refractive error in children. The Multi-Ethnic Pediatric Eye Disease and Baltimore Pediatric Eye Disease Studies10 and the Berkeley Infant Biometry Study11 used cyclopentolate 1% to evaluate refractive error. In patients with hyperopia, the Multi-Ethnic Pediatric Eye Disease and Baltimore Pediatric Eye Disease Studies found that children with uncorrected hyperopia of approximately 3 D or higher have an increased risk of developing esotropia and that risk increases exponentially with increasing hyperopia.12 In order to have more informed conversations with parents about the risk of esotropia and the role of accommodation in esotropia, the clinician should be aware of the full refractive error that is present and how accommodation is impacting ocular alignment. The Berkeley Infant Biometry Study studied refractive error in infants and children aged 4 months to 6.5 years, and much of our understanding of infant emmetropization and the decrease in hyperopia in the first 2 years of life comes from these data.11 More recent studies evaluating high hyperopia in infants show that retinoscopy with 1% cyclopentolate found up to a diopter more hyperopic spherical equivalent as compared with the 1% tropicamide (0.44 ±0.54 D, P < .001).13 A diopter difference in hyperopia, whether unilateral or bilateral, can mean the difference between concern for amblyopia or not. When prescribing for our youngest patients and assessing risk without subjective input, cyclopentolate allows full and more accurate assessment of risk factors.
In order to mimic these high-level study results across several refractive errors and ages in a clinical practice, following the same clinical protocol with cyclopentolate will allow results that more closely match those of the clinical trials.
Efficacy of Other Drop Regimens
There are 2 systematic reviews and meta-analyses that conclude that tropicamide is as effective as cyclopentolate for determining postcycloplegic spherical equivalent distance refractive error.14,15 However, when looking at these reviews more closely, the reader may have hesitancy in adopting this conclusion for pediatric care. A systematic review by Bist et al14 included only 4 studies following the study selection process for systematic reviews , and the Yazdani et al review15 only included 6 studies, with 3 studies overlapping in the Bist and Yazdani analyses. One of the studies included in both systematic reviews enrolled 21- to 50-year-old myopes, a population in which meaningful differences in refractive outcomes between cycloplegic agents are unlikely. Thus, including this study does not help determine which agent is most appropriate to use for children.16 Of the remaining 6 total studies included in the systematic reviews, 3 were case controls studies (increased risk for bias) instead of randomized control trials, and only 1 of the studies included infants.14,15 It is important to note that none of the studies in either review included patients with high hyperopia (>5 D), strabismus, or amblyogenic risks, conditions in which it is the most important to obtain a baseline cycloplegic refraction. The other randomized controlled trial from Twelker and Mutti,17 included in both reviews, found a mean noncycloplegic refractive error of +0.94 ± 1.19 D in their 29 infants between the ages of 4 and 7 months, which is also a lower-risk group in terms of expected refractive error. The Rajan et al study18 included in the Yazdani review included 25 children aged 6 to 15 years with hyperopia and had the most significant favorability toward cyclopentolate compared with tropicamide, even though only 0.5% cyclopentolate concentration was used, with a mean of 0.54 ± 0.18 D difference between agents. Although the reviews conclude that tropicamide is as effective as cyclopentolate for children and adults, Yazdani et al15 specifically notes, “…these results should be used cautiously in infants and in patients with high hyperopia or strabismus when using tropicamide as the sole cycloplegic agent especially in situations that the findings are variable.”
Time and Convenience
It would be easy to conclude that a drop with a shorter time of onset means a more efficient examination for both the provider and the patient. However, if a clinician uses a milder agent and discovers that there is a larger hyperopic shift than expected, or if the agent wears off prior to the time that the clinician is ready to check cycloplegic findings, the clinician may be less confident in the final prescription. This may prompt the provider to bring the patient back for a follow-up examination to re-check the prescription with full cycloplegia. If the clinician does finalize a prescription in which they are not confident, this may also warrant an extra follow-up that may not have been necessary otherwise. In either case, there is time added for both the clinician and patient/parent, which could have been avoided by using 1% cyclopentolate initially. The added follow-up with another drop procedure may also increase patient anxiety and will add to time away from school or work for the patient and parent, respectively. This said, most sources state a maximum cycloplegic effect of 30 minutes for tropicamide and 25 to 75 minutes for cyclopentolate, the lower end of which is not significantly different than tropicamide in the first place.3,19,20
Historically, literature suggests that, for full effect, cyclopentolate 1% be administered as follows: one drop in each eye, followed by a second drop 5 minutes later.1,10,11 If clinicians are also using other pharmacological agents for tonometry or mydriasis, this ends up being a large quantity of drops. If the clinician or their technician is waiting a full 5 minutes in between each drop, the factor of time also quickly adds up for both parties. Fortunately, there are multiple studies that indicate that the standard drop protocol of 2 drops of cyclopentolate may not be needed, which may reduce the amount of time for the patient in-office and side effect profile of the cyclopentolate. First, Hug and Olitsky21 found that 1 drop of 1% cyclopentolate is as effective as multiple drops in pediatric populations and that cycloplegia has an adequate mydriatic effect to reduce the need for a separate mydriatic agent (slightly less effective mydriasis, but not statistically significant in dark irides), reducing the number of drops needed for evaluation. Second, Zurevinsky et al22 found that a drop on the lids is as effective as a drop directly on the cornea, even in cases with amblyopia. Third, 2 studies have shown that spray is as effective in pediatric patients as drops in reaching full cycloplegia.23,24 The latter 2 strategies of closed eyes or use of a spray can both increase patient comfort and decrease anxiety about the drop instillation. However, the spray may not produce the same cycloplegic effectivity as a drop in patients with dark irides.24 The reduced number of drops will also improve the experience of the provider or technician, which may allow practitioners increased comfort in using 1% cyclopentolate more readily.
Caution in Certain Populations
Although this author supports the use of 1% cyclopentolate in most standard examinations, clinicians should be aware of the increased risk of adverse events of cyclopentolate in certain populations, in which cyclopentolate should be avoided. Young (aged younger than 3 months)1 and premature3 infants and children with Down syndrome and other significant developmental delays are at higher risk for the adverse events effects of language problems, excessive fatigue, decreased heart rate, or ataxia with the use of 1% cyclopentolate.3 Adverse reactions can also be noted in neurotypical children; however, these are usually transient and mild,3,25 and a closer look at the literature may minimize concern for clinicians. For instance, Imai et al surveyed parents about side effects of their children following a 2-drop dose of 1% cyclopentolate. There was an 18.3% frequency of side effects in the 646 patients surveyed, with the most common symptoms being conjunctival injection, drowsiness, and facial flush (noted mainly in children aged younger than 4 years).25 Heart rate and body temperature were measured for the children with facial flush but were all found to be normal. Additionally, despite including 55 subjects with central nervous system complications, this was not found to be a significant risk factor for symptoms.25 Contreras-Salinas et al3 provides a mini-review of the benefits and risks of cyclopentolate, with particular attention to systemic side effects. Although they include 106 references in their review, the most common side effect of language problems is only mentioned 12 times, with many of those being isolated case reports. The side effect profile increases with multiple drops, or higher concentrations of cyclopentolate, but is still minimal. For instance, Kim, Kim, and Lee26 reported a 10.83% rate of side effects when children were dosed with 5 instillations of 1% cyclopentolate, separated by 5 minutes each. Even with this exceedingly high dosage, events were all singular (isolated to 1 side effect per child) and resolved without further intervention needed.27 Clinicians should note that cycloplegic agents by nature will all cause the expected but transient side effects of blur at near and light sensitivity, and further adverse events are rare but possible in all agents (tropicamide, cyclopentolate, and atropine).28 The systemic absorption of agents can be further reduced through the use of punctal occlusion following drop instillation.
Summary
Cyclopentolate should be used as standard of care for all pediatric baseline examinations. The clinician can then confidently use the cyclopentolate results, with review of the other clinical data, to finalize an appropriate diagnosis, prescription, and management plan for the pediatric patient. Once a baseline is established, cyclopentolate does not need to be used at every examination thereafter but should be considered when large fluctuations in refractive error conflict with previous refractive findings or current binocular data. Side effects of cyclopentolate are exceedingly rare and transient when they occur. Use of a single drop with instillation modifications can increase patient and practitioner comfort in use of this agent.
COUNTERPOINT: THE CASE FOR TROPICAMIDE
Marc Taub, OD, MS, FAAO, FCOVD, FNAP
Dilation with a cycloplegic agent is a standard of care in optometry, as part of the examination of a pediatric patient.2 We use the dilation to obtain a view of the posterior segment and aid in prescribing. Children can be challenging to examine, and the use of a cycloplegic agent can reduce the fluctuations in the accommodative system, allowing the practitioner to obtain cleaner data more confidently. Once the clinician decides to use that data in prescribing, the question arises as to which agent is the most appropriate. Although cyclopentolate is considered the gold standard,2 the evidence does not support that it provides any greater clinical benefit, in many situations, than a drop we use every day—tropicamide.
Systemic Side Effects
Any medication put into the eye has potential systemic side effects. Facial flush, dizziness, tachycardia, palpitations, trembling, falling, and impact on the central nervous system leading to delirium, drowsiness, and sleep are documented side effects of cyclopentolate.19 Imai et al25 instilled 2 drops of 1% cyclopentolate in 646 patients aged 0 to 15 years (mean age: 7.0 years). The frequency of reaction was 18.3% overall, with conjunctival injection (10.5%), drowsiness (6.8%), and facial flush (2.2%) being the most common reactions. Facial flush was mostly observed in children aged younger than 4 years. Central nervous system complications were not a significant element for any of the symptoms. Tropicamide poses a lower risk of systemic complications, but these include allergic reactions, drowsiness, and irritation.19 There is a lower likelihood of central nervous system impact with tropicamide. Yolton et al29 and Garston30 reported no serious adverse experiences in 15,000 and 10,000 applications of tropicamide, respectively. Van Minderhout et al27 compared 2 drops of 1% cyclopentolate (C + C, n = 408) vs 1 drop each of 1% cyclopentolate and 1% tropicamide (C + T, n = 504). The study population age range was 3 to 14 years. The adverse reactions were reported in 10.3% and 4.8% of the C + C and C + T groups, respectively. The most common adverse finding was severe to moderate drowsiness in the C + C group (5.4%), which was found most often in children aged 3 to 6 years and in children with low body mass index. Severe to moderate drowsiness, hyperactivity, and/or behavioral problems were found significantly less in the C + T group.
Duration and Onset of Cycloplegia
It is well documented that cyclopentolate is slower in onset and has a longer duration of action than tropicamide.19 Cyclopentolate provides cycloplegia for 12 to 24 hours compared with tropicamide, which lasts for 4 to 10 hours. This means that the associated blurry vision and photosensitivity caused by cyclopentolate could be spread over 2 school days depending on the time of the examination. Since there is a difference in the amount of time for the agents to reach maximum effectiveness—up to 50 minutes for cyclopentolate vs 20 to 30 minutes for tropicamide20—this can lead to decreased attention and compliance, especially in young patients.
Differences in Pupil Size
One of the primary purposes of these pharmacological agents is pupil dilation, which makes the question of pupil size worthy of investigation. Anderson et al31 studied 45 subjects, aged 4 to 32 years, with dark irides. They compared 2 drop combinations: 1% tropicamide and 2.5% phenylephrine vs 1% tropicamide and 1% cyclopentolate. Each patient had both combinations, 1 in each eye. Photographs were taken at various intervals up to 60 minutes. Ninety-eight percent achieved a 6-mm dilation with either combination of drops. Only 80% reached 7 mm with 1% tropicamide and 2.5% phenylephrine and 58% with 1% tropicamide and 1% cyclopentolate (P = .0062). The time to reach 6 mm was not statistically significant, but the time to reach 7 mm was both statistically and clinically significant (P = .0325, 1% tropicamide and 2.5% phenylephrine 32 min vs 1% tropicamide and 1% cyclopentolate 52 min). Age was not a factor. Cyclopentolate did not aid in achieving a larger pupil size and had a slower time to attain that size.
In contrast, Apt and Hendrick32 compared pupil size with cyclopentolate hydrochloride 0.5%, tropicamide 1%, and tropicamide 0.5%, each with phenylephrine 2.5% in 70 individuals, aged 16 to 84 years (mean: 54 years). Unlike the study by Anderson et al,31 the study included different iris colors (37-blue, 12-hazel, 31-brown). Each of the combinations were tested with and without proparacaine. Each individual received 2 drop combinations, 1 in each eye, depending on the study group. No difference in pupil size was found between cyclopentolate and tropicamide 0.5%, either with or without proparacaine (P < .05) at 30, 45, and 60 minutes postinstallation. At 30 minutes, the dilation was 0.3 mm less with cyclopentolate when proparacaine was not used and 0.1 mm less when it was used. Age was a factor (P < .05) but iris color was not (P < .05).
Differences in Refraction
Two recently published meta-analyses came to the same conclusion regarding cycloplegia for refractive purposes: Tropicamide is a viable substitute for cyclopentolate but should be used cautiously with specific populations, including patients with strabismus or high hyperopia, especially infants.14,15 Bist et al14 analyzed 4 randomized-controlled studies from 1993 to 2024, whereas Yazdani et al15 included 6 studies (3 randomized-controlled and 3 case-control studies) from 1993 to 2011. Three studies overlapped between the 2 meta-analyses. The ages of the patients studied ranged from several months old to adults and included all types of refractive conditions. It is important to keep in mind that there is a difference between clinical and statistical significance. For the purpose of this discussion, as proposed by Guha et al,33 a dioptric difference of less than 0.50 is considered clinically insignificant.
A study by Egashira et al,34 which was included in both meta-analyses, compared tropicamide and cyclopentolate in a population of 20 nonstrabismic, nonamblyopic, hyperopic 6- to 12-year-olds. The mean and range of the refractive error was +1.48 ± 1.10 D and +0.25 D to +4.50 D, respectively. Several measures were taken, including retinoscopy, distance subjective refraction, distance autorefraction, and accommodative amplitude (objectively and subjectively). Cycloplegic retinoscopy and distance subjective refraction showed no difference (P = .13, P = .42). However, distance autorefraction showed a statistical (P = .045) but not clinical significance with cyclopentolate producing more hyperopia (0.14 ± 0.30 D). Accommodative amplitude showed significance only in the objective measurement (subjective P = .08, objective P = .013). Time was a considered factor, as all measurements were made at 30 and 60 minutes.
Al-Thawabieh et al20 studied 185 eyes from 94 children aged 3 to 16 years. Those with amblyopia and strabismus were included except if there was a history of surgery. Subjects were randomized to tropicamide or cyclopentolate, and autorefraction was completed using the Nidek-AR 1000 (NIDEK Co., Gamagori, Aichi, Japan) 30 minutes after tropicamide and 60 minutes after cyclopentolate administration. The same group, on a second visit, was given the opposite drop and the same testing performed. The primary outcome was the change in spherical equivalent refractive error. The refractive conditions were classified as high myopia (SE < −5.0 D), myopia (−5.0 D ≤ SE < −0.50 D), emmetropia −0.50 D ≤ SE ≤ +0.50 D), hyperopia (+0.50 D < SE ≤ +5.0 D), and high hyperopia (SE > +5.0 D). The average hyperopic change was 1.15 ± 1.2 D for cyclopentolate and 1.04 ± 1.2 D for tropicamide. The 0.11D difference was statistically significant (P < .001) but not clinically significant. On closer examination of the 5 refractive categories, the largest difference was observed in the high myopia group (0.15 D), with a group range of 0.05 to 0.15 D. All 5 categories showed statistical significance, but clinical significance was lacking again.
Is Preauction Necessary With Certain Populations?
Regarding the use of tropicamide for refraction, practitioners preach caution with high hyperopes, infants, and strabismics, but whether this rings true is the question. As seen in the Al-Thawabieh et al20 study, in the group with high hyperopia, participants demonstrated 0.08 D more hyperopia with cyclopentolate when compared with tropicamide. Morrison and Muti13 studied 473 infants undergoing their 2-month well visit. Refractive error was measured by retinoscopy and autorefraction after participants received 1% tropicamide at the screening. Thirty-five infants with +5.0 D or more in the most hyperopic meridian were reexamined at a later visit using 1% cyclopentolate. In comparing the 2 agents on the 2 types of measurement in the 35 highly farsighted infants, the cyclopentolate refractive error reading ranged from 0.40 D to 0.52 D more hyperopic than the tropicamide readings. The mean refractive error measured with cyclopentolate was 0.40 D more hyperopic compared with tropicamide (P < .001), and whereas statistical significance cannot be disputed, clinical significance can be. Only finding 0.40 to 0.52 D more hyperopia in highly farsighted infants may not be enough to justify the use of cyclopentolate.
Twelker and Mutti17 examined 29 nonstrabismic infants aged 4 to 7 months using a near noncycloplegic technique (a modified Mohindra technique in which the examiner neutralized the reflex with loose lenses in a dark room and 0.75 D was subtracted from the finding)35 and cycloplegic retinoscopy, with both tropicamide and cyclopentolate, on separate visits. They concluded that there was no significant difference (P = .65) between the 2 agents (+1.88 D with cyclopentolate and +1.81 D with tropicamide). As might be expected, there was a difference between both agents and the noncycloplegic technique, which the cycloplegic refractive error showing more hyperopia than the noncycloplegic assessment. The study found that tropicamide was just as effective as cyclopentolate for measuring refractive conditions in healthy, nonstrabismic infants.
Data are lacking for the last group of concern—those with strabismus. Lyu et al36 reviewed the records of 34 children (3.75 to 5 years) with both esotropia and hyperopia. A prism cover test was completed both before and after cycloplegia with tropicamide and phenylephrine as part of a comprehensive examination. An increased angle of deviation was defined as 10 prism diopters or greater after cycloplegia. Thirteen patients, all of whom had accommodative esotropia, increased in esodeviation under cycloplegia. The deviations changed from 15 prism diopters at distance and near to 30 prism diopters at distance and 35 prism diopters at near. This is obviously a large change and one that is not a shock to optometrists who treat children. What is unclear is whether the use of cyclopentolate would have led to a further increase in esodeviation, as this comparison was not made in this study.
Conclusion
For many clinicians, the use of a cycloplegic agent, such as tropicamide or cyclopentolate, is necessary and desirable to assist in determining the refractive correction. Although cyclopentolate is considered the gold standard, there is growing support for the use of tropicamide as an alternative. It has a better safety profile and a shorter duration of action, with minimal to no clinical significance between the 2 pharmacological agents. It is time to rethink our reliance on cyclopentolate as the only agent to assess refractive error in our youngest patients, especially those without amblyopia or strabismus.
Conflicts of Interest
The authors declare no conflicts of interest.
Disclosure of Funding
The authors declare no funding sources.