Daniel Kook, MD, PhD, FEBO
The problem in this case is an overcorrection after myopic femtosecond-assisted laser in situ keratomileusis procedure in the left eye, which occurs less often than undercorrection. In general, reasons for overcorrection are due to preoperative errors or conditions, intraoperative mistakes or complications, or postoperative complications. Postoperative corneal tomography in this patient shows a centered ablation and the difference between the K values preoperatively (43.80 diopters [D] and 43.50 D) and postoperatively (36.60 D and 36.10 D) that is in line with the intended correction of −7.75 D spherical equivalent. Postoperative slitlamp examination showed a clear cornea. Therefore, it seems that the underlying problem was not the laser treatment or postoperative complications (eg, diffuse lamellar keratitis) but the preoperative refraction, which was presumably incorrectly high.
Before planning retreatment, the underlying problem, or possibly a combination of different problems, must be identified. Possible reasons to be ruled out in this specific situation should include preexisting central corneal warpage due to insufficient contact lens break (preoperative corneal tomography); underlying subclinical corneal dystrophy (eg, Cogan's dystrophy); higher corneal vertex distance during visual acuity testing than a vertex distance of 12.0 mm in the laser protocol, which would however only explain a minor role in the overcorrection; accommodation (preoperative cycloplegic refraction being 1.00 D less myopic, compared with refraction of the other eye); pseudomyopia due to underlying large exophoria or intermittent exotropia of the left eye, leading to accommodative convergence; or systemic conditions like hyperglycemia, unknown pregnancy, or breastfeeding.
Refractive error must be stable before retreatment. As the UDVA was 20/25 1 week postoperatively, refraction at that timepoint was probably different than refraction at 1 month postoperatively. At 7 months postoperatively refraction again changed to the now hyperopic astigmatism. Tomography images and epithelial mapping would help to address this change in refraction to corneal remodeling. In any case, I would strongly advise cycloplegic refraction of both eyes before retreatment of the left eye.
The procedure of choice for retreatment also depends on the identified cause. In general, as in this case with hyperopic astigmatism being the underlying type of refractive error to be corrected, my first preferred procedure would be a relift of the flap if estimated residual stromal bed (RSB) thickness before retreatment was around 300 μm. Of note, hyperopic ablation would not markedly decrease central stromal thickness. I would choose the same optical zone as in the primary treatment because a larger optical zone will probably not fit to the flap diameter and ensure very good centration of the laser. My second preferred procedure, if RSB thickness was critical, such as in the case of a thick flap or a case of a large mesopic pupil that may require enlargement of the optical zone, would be surface ablation with use of mitomycin-C. Target refraction in both procedures depends on the refraction of the other eye treated in order to avoid postoperative anisometropia.
Disclosures: D. Kook has received fees for lectures for Alcon Laboratories, Inc. and Carl Zeiss Meditec AG.
Rohit Shetty, FRCS(G), PhD Vaitheeswaran G. Lalgudi, MBBS
The hyperopic shift in refraction in this post-LASIK eye is because of two issues. First, the cycloplegic refraction is −6.50/−0.50 @ 68, which is different by 1.00 D from the refraction treated (−7.50 −0.50 × 75). This contributes 1.00 D of overcorrection. Second, myopes who use overcorrected spectacles frequently have accommodative and convergence insufficiency. Before surgery, these issues are masked by the myopia as such and the base-in-prism effect of concave spectacle lenses. Postoperatively, accommodative and convergence insufficiency worsen and can lead to the patient manifesting a hyperopic refractive error for distance along with poor near vision. The refractive error of +2.50 −1.25 × 140 at 1-year follow-up is a combination of the 1.00 D of overcorrection and the refractive error induced by accommodative/convergence insufficiency. Our refractive and orthoptics departments have previously encountered such induction of hyperopia and astigmatism due to accommodation insufficiency in post-refractive surgery patients. Even without overcorrection, myopes with accommodative/convergence issues can present with hyperopic shifts post-refractive surgery.
Correction of this patient's refractive error should not involve a laser recorrection as the first step. A thorough orthoptic evaluation should be done first. Most patients tend toward emmetropia and with appropriate exercises can address the underlying orthoptic issues. While initiating orthoptic exercises, spectacle correction of the full power may be required. The patient can be reassured that during therapy spectacle power will be gradually tapered. In our experience, most patients achieve significant reduction in refractive error and no longer require a corrective procedure. If a certain amount of refractive error persists (eg, 1.00 to 1.50 D) despite the full course of orthoptic exercises, correction can be planned for the residual error.
Residual error of hyperopia, confirmed by a repeat cycloplegic refraction, is best addressed by a flap lift and excimer laser ablation of the bed. Surface ablations are not ideal for this, as epithelial remodeling will lead to regression. Another available option is a hyperopic phakic intraocular lens, an implantable collamer lens (ICL), if the anterior chamber depth is adequate.
To summarize, hyperopic shift post myopic refractive surgery due to overcorrection or excessive corneal flattening may be incorrect. A meticulous stepwise approach along with orthoptic evaluation helps to solve the problem.
Disclosures: Neither author has a proprietary or financial interest in any material or method mentioned.
Samuel Arba Mosquera, MSc, PhD
This is an interesting case of an apparently unexpected overcorrection in 1 eye in a young woman, leading to patient dissatisfaction from a relatively early follow-up time; I have a few comments about the case.
The reported temperature and relative humidity do not correspond one-to-one to the values presented in Figure 1, although both sets were within normal ranges. This may be because the reported values are from an external measurement in the operating room, while those in Figure 1 were recorded inside the system at a position close to the laser exit.
It is interesting that there was a refraction disparity of −2.75 −1.00 × 8 in the left eye (higher myopia), although the CDVA was 20/16 and it was reported that there were no abnormalities in her medical history (ie, no signs of amblyopia). The manifest refraction was also −0.75 −0. 25 × 119, more myopic than the cycloplegic refraction, while the laser was set for correcting the manifest refraction. From the cycloplegic difference, up to about +1.00 D overcorrection could have been expected. Also, the strong disparity between refractions in both eyes without signs of comorbidities or pathologies were a warning flag. The left eye refraction at least should have been confirmed by other means. Otherwise, the patient history and treatment outcomes were not significant, within normal and/or expected parameters.
In comparing the change in K values after applying corrections, although the preoperative K readings were from the Sirius (Schwind eye-tech-solutions GmbH & Co. KG) while the postoperative K readings were from the Pentacam, a refractive change of −9.00 −.50 × 72 can be measured in the cornea, an overcorrection of only +1.25 D above the actual plan.1 To correct for the cycloplegic refraction, one could estimate a postoperative refraction of +2.25 D −0.25 D × 8, not far from that at the 1-month and 1-year postoperative follow-ups.
At 1-year postoperatively the patient manifests +2.25 D −1.25 D × 140 in the left eye, reaching a CDVA of 20/16. Calculating the cycloplegic difference from preoperative to postoperative refraction, there is an estimated change of −8.0 D −2.0 D × 54, with the same spherical equivalent as seen in the change in K values but with a much stronger, and some 20 degrees rotated, cylinder change.
Figure 2 may be interpreted as a slight decentration superiorly, an estimate of up to 0.5 mm superior decentration, but this is difficult to confirm from sagittal curvature without preoperative topography. Preoperative topographies and local curvature maps would be helpful and perhaps particularly relevant for the 1-year measured refraction.
In my opinion, the overcorrection might be due to overplanning. My recommendation for correction is anterior segment optical coherence tomography and detailed diagnostic evaluations. But one could check whether a simple +1.50 D reaches good VA, as well as comparing the CDVA and quality of vision for +2.25 D −0. 25D × 8 with +2.50 D −1.25 D × 140. Placing a rigid gas-permeable contact lens in the left eye to see if quality of vision improves can help determine if corneal wavefront-guided (CWFG) ablation is required.
If CWFG ablation is required, the residual hyperopia can be treated either with CWFG or aberration-free ablation after determining a refraction closer to that predicted by topography. An optical zone of 6.7 mm should be used with CWFG ablation, and a 6.3 mm with aberration-free ablation. Likely relift would be the most reasonable technique. If CWFG ablation is required, then transepithelial photorefractive keratectomy can be also considered, despite the hyperopic refraction.1
Disclosures: The author has no financial or proprietary interest in any material or method mentioned.
1. Adib-Moghaddam S, Arba-Mosquera S, Walter-Fincke R, Soleyman-Jahi S, Adili-Aghdam F. Transepithelial photorefractive keratectomy for hyperopia: a 12-month bicentral study. J Refract Surg 2016;32:172–180
José F. Alfonso MD, PhD
This case presents a common situation that results when an excimer laser is used to correct a myopic anisometropia over 3.00 D. Analysing the provided information, the postoperative hyperopia could be caused by different issues.
The manifest refraction value was −7.50 D, whereas the cycloplegic refraction was −6.50 D. Based on my experience, in patients with anisometropia of ≥3.00 D using the cycloplegic refraction is recommended; however, in this case the manifest refraction was the target refraction. In my opinion, the target refraction should have been at most −7.00 D.
There was not an axial length (AL) measurement to verify the myopic difference between both eyes. Preoperative AL measurement should be mandatory before any surgical refractive procedure. In this case, both eyes had similar K values and there should be correlation between the difference in refraction and the difference in AL of both eyes. Unfortunately, if there is no AL measurement, an error in the refractive estimation could go unnoticed.
Corneal hydration levels affect the efficiency of laser ablation and consequently refractive outcomes. This becomes more important when, as in this case, the amount of ablation is large (>100 μm). Corneal stromal dehydration during surgery could result in more corneal tissue removal than was programmed.1 Another possibility is that the photoablation profile designed to correct against-the-rule myopic astigmatism tends to induce slight hyperopia (+0.25 to +0.50 D).
The postoperative flat K of 36.10 D indicates the cornea flattened more than required, which means the hyperopia is real. According to my estimates, given the preoperative characteristics of this case, the postoperative flat K to achieve emmetropia should be around 37.50 D. Thus, this residual hyperopia could come from a double source: (1) there was an overcorrection because the target of correction was the manifest refraction, and (2) there was an excessive absorption of energy in the corneal stroma.
For corneal recovery after an excimer laser procedure to treat myopic refractions over 6.00 D, it can take up to 6 months for the corneal epithelium to compensate for photoablation-induced flattening. This case may be within this postoperative period. Also, the higher the preoperative myopia, the slower the recovery process for the cornea, as the amount of corneal tissue removed impacts the time it takes for the cornea crystalline lens to adapt to the new corneal shape.
I have different approaches to solve this postoperative complication. The conservative approach is to wait 3 to 6 months and see if the epithelium compensates for the corneal flattening; the lens accommodation should be fully recovered during this same time period. Keratometry and near vision should be closely monitored.
As the residual hyperopia is low, a relift of the flap would be easy with a good prognosis. There are usually no intraoperative complications, although an interface epithelization is possible. It is important to consider that the ablation profile for retreatment could induce some visual disorders. Also, this retreatment could be inconvenient if the patient requires refractive lensectomy in the future.
Another option is to implant an ICL. This is a safe and effective procedure to correct residual refractive errors after LASIK. I would choose the ICL for this patient, because it will correct the refractive error and perhaps other changes that could occur in the future. However, it is important to note that the anterior chamber depth must be greater than 3.00 mm and an iridectomy should be performed.
Disclosures: The author has no financial or proprietary interest in any material or method mentioned.
1. Kim WS, Jo JM. Corneal hydration affects ablation during laser in situ keratomileusis surgery. Cornea 2001;20:394–397
Joaquim Neto Murta, MD, PhD
To have a clear idea of what happened with this patient more information from preoperative Scheimpflug topography for K measurements and corneal thickness measurements is necessary. In the patient’s clinical history, were the following possibilities investigated: previous ocular trauma, inflammatory episodes, or even pregnancy, which could have been interrupted?
I think that this patient was not a good candidate for refractive surgery because of the observed difference of cycloplegic and subjective refraction of 1.00 D. It would be advisable to check the reasons for such a difference and wait. On the other hand, even though I do not know the preoperative corneal thickness and K measurements, I would have to consider an ICL for this patient. The RSB is very low. What was the thickness of the lenticule? Was surgery performed by microkeratome or femtosecond laser?
Possible primary causes for this overcorrection are corneal microstriae (a theoretical cause but unlikely), a technical problem with the excimer laser (I have had this problem twice), or medication. Topiramate, for example, can induce myopia. With the suspension of the drug, hyperopia would be induced.
My advice for correction of the refractive error for this patient is implantation of an ICL.
Disclosures: The author has no financial or proprietary interest in any material or method mentioned.
Scott MacRae, MD
Rochester, New York, USA
There are several clues to what may have caused the overcorrection in the left eye. Perhaps the most obvious is the cycloplegic refraction in the left eye, which is 1.00 D less myopic, suggesting that the patient is accommodating. This should have prompted the surgeon's team to automatically bring in the patient for a repeat post-cycloplegic manifest in both eyes, fogging the eye by adding enough plus power to minimize the patient's tendency to accommodate. I typically start out showing the patient the 20/40 line with1.25 D more hyperopic than the manifest, telling the patient that the image will be intentionally blurry and proceeding slowly; the patient may jump 2 to 3 lines down the chart as their accommodation relaxes.
The second interesting discrepancy is the 1.25 D of cylinder in the right eye but only 0.50 D of cylinder in the left eye. One common problem with manifest refraction, even with experienced surgeons, is the tendency to masquerade unrecognized myopic cylinder by adding myopic sphere. Did the left eye have 1.25 D cylinder that was unrecognized and then was masqueraded by overcorrecting the patient with myopic sphere? This leads to a hyperopic overcorrection and overcorrection of the cylinder in the opposite meridian. I have found that high-quality wavefront sensors are very sensitive to detecting larger amounts of cylinder with a more correct axis than the refractionist detects, and when a repeat refraction is performed and the undetected cylinder correction is incorporated, a better correction is obtained. The third interesting preoperative finding is the marked anisometropia, which should raise concerns about the treatment.
In terms of retreatment, the patient should have several repeat manifest refractions by an experienced refractionist, fogging the eye as noted above and pushing plus power since the patient probably tends to accommodate. Once a consistent refraction is achieved, a hyperopic treatment could be performed. I reduce the hyperopic treatment by 3% to 40% when retreating an overcorrected myope because I am “putting downhill” and there may be a tendency to overcorrect the patient in the opposite direction.
Disclosure: The author has no financial or proprietary interest in any material or method mentioned.
Jorge Alió, MD, PhD
There is no information about the preoperative topography even though adequate cycloplegic refraction is provided. With no information about the right eye, it is assumed that this is a purely unilateral overcorrection related to factors that affect only the left eye. The reasons for overcorrection such as this are environmental conditions, which were borderline in this case but cannot justify the overcorrection and should have eventually affected the other eye; the surgeon performing excessive drying during surgery; postoperative keratitis; or post-LASIK necrotizing keratitis, a severe form of diffuse lamellar keratitis that leaves a scar at the LASIK interface with central corneal flattening, which is not observed in this patient as biomicroscopy seems to be normal. With all this in mind, the corneal topography map looks too flattened for a standard cornea, in which a −7.50 D refraction should have been performed. Rather, it seems that overcorrection has occurred.
Because of missing data and limited information, it can only be presumed that this is an idiopathic overcorrection related to factors that affected the left eye during surgery, most likely desiccation. Desiccation can be caused by excessive time taken during ablation, stopping the ablation due to factors related to the surgery (eg, patient cooperation), or drying the surface with a sponge. This case shows an adequate residual corneal thickness assuming it was operated with a regular flap of 100 to 110 μm. An overcorrection of +2.5 −1.25 will ablate basically one meridian with less effect at the center and no problem of RSB.
My approach to retreatment would be to reevaluate the flap, applying the refraction that is mentioned in the postoperative refractive condition of the patient, with a large optical zone in order to ablate the peripheral part of the cornea, which should be affected by postoperative spherical aberration. This patient lacks aberrometry readings, but total eye aberrometry would likely show excessive amounts of posterior surgical aberration. I would plan for a customized total wavefront-guided excimer laser treatment.1
Disclosures: The author has no propietary or financial interest in any material or method mentioned.
Chayet AS, Torres LF, Lopez J. Refractive miscalculation with refractive surprise: sphere. In: Alio JL, Azar D, eds. Management of complications in refractive surgery. 2nd ed. Cham, Switzerland: Springer; 2018:141–145
Erik L. Mertens, MD, FEBO
This 31-year-old woman presented with refractive high myopia and anisometropia in the left eye. Cycloplegic refraction clearly shows 1.00 D less myopia and should have been taken into account. The surgery itself seemed to be uneventful.
Now why did the patient end up with hyperopia ? One diopter of overcorrection can be explained by not taking into account the cycloplegic refraction. The biomicroscopic examination tells us there is no epithelial ingrowth or flap irregularities. What would have been very helpful is an epithelial thickness map and the refractive power of this post-LASIK epithelium. This is invaluable information and every refractive surgeon should take a look at the epithelial thickness map preoperatively and postoperatively.
Other information which is missing is the nibut, tearfilm osmolarity, meibography pictures,…. How is the ocular surface behaving? When examining eligible candidates for surgery, great care should be taken to examine the quality of the tear film. It is not uncommon to find a 1.00 D or more change in refraction and keratometry if the ocular surface is optimized before performing preoperative measurements. At 7 months postoperatively the refraction shows a diagonal astigmatism and astigmatism is also shown by the posterior elevation map (Pentacam).
This is a case where I would not touch the cornea again although there is not enough detailed information for a final conclusion. If the anterior chamber depth is sufficiently deep and the refraction is at least stable in between consultation visits at a 3-month interval, I would opt to implant a poster chamber hyperopic toric phakic IOL.
Disclosures: Dr. Mertens is a consultant to STAAR Surgical.
Rudy M.M.A. Nuijts, MD, PhD
Maastricht, the Netherlends
This case points at an important practical problem, that is, how to handle and interpret variations in refraction error preoperatively. Repeating measurements when there is a large disparity between manifest and cycloplegic measurements is an absolute necessity before one decides on the amount of refractive error to correct. As was pointed out by one of the respondents, fogging the eye by adding enough plus power to minimize the patient's tendency to accommodate is absolutely mandatory, an art that might be overlooked in this era of focus on auto-refractokeratometers. Another interesting topic that was raised by the respondents is the preoperative anisometropia and the potential relationship with accommodation insufficiency in post-refractive surgery patients when frinstens preoperative exotropia would not have been recognized. This points at the role of orthoptists in setting the treatment strategy for this particular group of anisometropic patients.
Depending on the agreement between various repetitive refractions after 6 months by an experienced optometrist, we will set the refractive error for a hyperopic treatment, preferably with a relift of the flap.