Hyperopic shifts and anterior curvature flattening after Descemet-stripping endothelial keratoplasty and cataract surgery in 2 patients with a history of myopic LASIK

Doherty, Terrence J. MD

Author Information
doi: 10.1097/j.jcro.0000000000000038
  • Free


Endothelial keratoplasty has proven to be a successful treatment for patients with corneal endothelial disease such as Fuchs endothelial corneal dystrophy. Compared with penetrating keratoplasty, endothelial keratoplasty provides comparable final corrected distance visual acuity (CDVA) but with faster visual recovery and less-induced irregular astigmatism and other suture-related complications (although some suturing techniques might provide similarly rapid recovery from penetrating keratoplasty).1 Descemet-stripping endothelial keratoplasty (DSEK) involves removing the central portion of Descemet membrane and transplanting a partial-thickness donor graft consisting of endothelial cells carried on a thin layer of posterior stroma. The graft can vary in thickness and contour.

The addition of the graft tissue decreases the posterior radius of curvature, resulting in an overall decrease of the total corneal power and a hyperopic refractive shift. The amount of postoperative hyperopia varies by graft thickness and can be as much as several diopters.2–4 Multiple studies have consistently demonstrated that, unlike the posterior curvature, the anterior curvature of post-DSEK corneas remains relatively stable and unchanged from preoperative measurements, contributing very little to the hyperopic changes. This might not be the case if the patient has had previous myopic LASIK, however.

In this study, I present a small case series in which 4 eyes of 2 patients with Fuchs endothelial corneal dystrophy and a history of myopic LASIK experienced significant flattening of the anterior corneal curvature after undergoing cataract surgery with combined or sequential DSEK, with a much greater-than-anticipated hyperopic result.


Case 1

A 53-year-old man presented to the clinic with complaints of difficulty driving, increased glare, and trouble watching television. He reported having had bilateral LASIK surgery at another site about 14 years earlier. His vision had been satisfactory until these new symptoms started increasing over the past year. He denied any other ocular history or surgeries and reported no family history of eye disease.

Examination revealed a manifest refraction of −0.25 −1.75 × 90 with 20/25 vision in the right eye and −0.50 −1.75 × 91 with 20/20 vision in the left eye. Nuclear sclerotic cataracts and moderate 2+ central corneal guttata were noted in both eyes. No significant corneal edema was noted. Glare testing revealed a reduction of vision to 20/100 in the right eye and 20/50 in the left eye. Endothelial cell counts were 2163/mm2 in the right eye and 2025/mm2 in the left eye. The posterior examination was normal in both eyes.

The decision was made to perform cataract surgery, starting with the right eye. Corneal topographies demonstrated bilateral myopic ablation patterns that appeared well centered (Figure 1). Axial length, as measured by partial coherence interferometry (IOLMaster, Carl Zeiss Meditec AG) was 26.95 mm in the right eye and 26.79 mm in the left eye. Anterior chamber depth was 3.37 mm in the right eye and 2.89 mm in the left eye. The horizontal white-to-white diameter in the right eye was 11.6 mm. The Haigis-L formula and IOL calculations from the ASCRS for previous myopic LASIK were used to calculate IOL power. All calculations were found to be relatively similar. The surgeon chose an initial refractive target of −2.50 D to counterbalance any hyperopia that could result from the history of myopic LASIK and possible need for DSEK in the future. The patient had also expressed desire for spectacle independence, if possible, and the surgeon was aiming to achieve a myopic outcome to improve the accuracy of a potential photorefractive keratectomy procedure.

Figure 1.
Figure 1.:
Case 1 topography maps. Top: Preoperative topography demonstrating a well-centered myopic ablation pattern in both eyes. Amount of correction was not known at the time of the patient’s presentation. Bottom: After Descemet-stripping endothelial keratoplasty, a significant decrease in the SimK readings and corneal power is noted. A myopic ablation pattern is still apparent.

Cataract surgery in the right eye was uneventful. A Tecnis Z9002 3-piece silicone IOL (Johnson & Johnson Vision Care, Inc.) with a power of 20.5 D was inserted into the capsular bag. Postoperatively, the IOL was observed to be well centered, with the optic edge overlapped by the anterior capsule for a full 360 degree. The postoperative refractive error was +0.50 − 1.25 × 95, which was much less myopic than planned. Despite having CDVA of 20/25 with this refraction, the patient still reported continued visual symptoms and overall dissatisfaction with visual quality.

Six months postoperatively, CDVA reduced to 20/50 with worsening fluctuations due to his Fuchs endothelial corneal dystrophy. An uneventful DSEK was performed in the right eye using an 8.5-mm graft. The graft was 132 μm thick and was noted to be perfectly centered postoperatively. The refraction changed drastically 5 months after DSEK to +3.00 − 1.75 × 150 with CDVA of 20/30. The posterior segment examination was normal, and there was no significant irregular astigmatism. Further diagnostic tests were not performed. The left eye had a relatively stable refraction of −0.25 − 2.00 × 86, creating an intolerable amount of anisometropia. Over a 3-month span, the patient tried multiple pairs of spectacles, all of which were unsatisfactory. He underwent a successful piggyback IOL implant 5 months after DSEK, which resulted in uncorrected visual acuity (UCVA) of 20/30 and a refraction of +0.50–1.00 × 70.

About 1 year later, the decision was made to perform combined cataract and DSEK surgery on the patient’s second eye to alleviate the anisometropia and visual quality complaints. CDVA in this eye was 20/25-2. An IOL was chosen in the same manner as for the first eye, with a calculated postoperative refractive target of −3.00 D in anticipation that the left eye would potentially heal in a manner similar to the right As had occurred with the right eye, the patient experienced a much more hyperopic result in the left eye than would have been predicted. Six months postoperatively, the refraction was +2.25 − 0.50 × 94 with UCVA of 20/30. Ocular coherence tomography imaging of the maculae performed at that time showed normal foveal depressions with a central thickness of 300 μm in the right eye and 284 μm in the left eye. An Objective Scatter Index of 7.0 was measured with the Optical Quality Analysis HD double-pass system (Visiometrics), indicating significant scatter in the left eye. This was thought to be due to a small amount of interface haze and mild posterior capsular opacification in that eye. There was no significant ocular surface dryness. Six months after DSEK/cataract surgery, a piggyback IOL was implanted in the left eye, improving the UCVA and resulting in a refraction of +0.25 − 0.50 × 90.

The patient was lost to follow-up 3 months after piggyback IOL surgery when he moved to a different state. His last refraction in our practice 8 months after DSEK and 3 months after the piggyback IOL in the left eye was +0.50 − 1.00 × 70, with 20/20-2 CDVA in the right eye and +1.25 − 2.00 × 66 with 20/25-2 CDVA in the left eye.

Comparison of preoperative and postoperative corneal topography maps showed a large decrease in the anterior curvature was noted in both eyes (Figure 1). The right eye had a steep simulated keratometry (SimK) reading that went from 41.20 to 36.72 D and corneal power reading that changed from 40.65 D preoperatively to 33.86 D 4 months after DSEK. Similarly, the steep K in the left eye changed from 41.2 to 38.7 D and power decreased from 40.98 D preoperatively to 35.14 D 3 months after DSEK. Despite some fluctuation, the right eye topography seemed to have stabilized about 11 months after the final surgery (piggyback IOL) in that eye. The final corneal measurements performed for this patient were high-resolution Scheimpflug tomography maps (Pentacam HR, Oculus Optikgeräte GmbH) taken 30 months after DSEK in the right eye and 6 months after DSEK/cataract in the left eye. They showed SimK readings of 37.1/36.8 D in the right eye and 36.9/36.5 D in the left eye (Figure 2). This flattening and reduction of the corneal power was seen as a likely contributing factor for the higher-than-expected hyperopic result.

Figure 2.
Figure 2.:
Case 1 final tomography maps taken 30 months after DSEK in the right eye and 6 months after DSEK in the left eye. Significant anterior flattening compared with preoperative measurements is still present (DSEK = Descemet-stripping endothelial keratoplasty).

Case 2

A 61-year-old woman presented with progressively worse vision in both eyes. She had a history of myopic LASIK many years before, but these records were not available. On examination, she was found to have significant cataracts and 3 to 4+ central corneal guttata with mild cornea swelling characteristic of Fuchs endothelial corneal dystrophy. Endothelial cell counts were not measured. The decision was made to perform combined cataract extraction and DSEK in both eyes. IOL power calculations were performed using the calculators on the ASCRS website for previous myopic LASIK and Holladay 2 formula with equivalent K reading (EKR65) SimK values from the Pentacam tomography.5 A higher-than-average myopic target of −1.50 D was chosen, in anticipation that the history of myopic LASIK in the presence of advanced Fuchs endothelial corneal dystrophy would cause a significant hyperopic error.

Combined surgery on the left eye was performed without difficulty. The graft thickness for the DSEK was 69 μm. Three months postoperatively, the patient’s refraction in that eye was +2.50 − 0.50 × 45. This reduced 4 months later to +1.50 − 0.75 × 75 with a CDVA of 20/25-. Ocular coherence tomography of the macula revealed a normal foveal contour and central thickness of 267 μm in the right eye and 281 μm in the left eye. Preoperative and postoperative tomography maps are shown in Figure 3. There was a measured 1.60 D flattening in the anterior corneal curvature after the procedures. The postoperative map also demonstrated a more characteristic central flattening of a myopic LASIK ablation compared with that of the preoperative map.

Figure 3.
Figure 3.:
Case 2 topography maps. Top: Preoperative topography shows a mild myopic ablation pattern in the right eye. The left eye does not display a typical myopic ablation pattern and actually measures slightly steeper centrally. Bottom: Three months after DSEK, the left eye now demonstrates a more characteristic myopic ablation pattern with a central flattening of about 1.60 D. The right eye displays a prominent central island that was completely masked before the DSEK procedure and subsequent resolution of corneal swelling (DSEK = Descemet-stripping endothelial keratoplasty).

This patient had a similar combined procedure in the right eye 7 months later with an IOL power targeted for −2.00 D. The graft thickness in this case was 105 μm. Four months after the procedure, the patient’s refraction was plano −0.50 × 081 with CDVA of 20/30-2. Again, there was a marked change in the anterior curvature maps, this time unveiling a noticeable central island from her LASIK procedure that seemed to be completely masked in the preoperative maps.


This case report demonstrates a significant shift in the anterior curvature and topographical characteristics of 2 patients with Fuchs endothelial corneal dystrophy and a history of myopic LASIK who underwent posterior lamellar keratoplasty (DSEK) and cataract surgery. Both cases had postoperative refractive results that were significantly more hyperopic than anticipated. This was likely due in part to significant flattening of the anterior cornea.

Multiple studies have shown insignificant anterior corneal changes after posterior lamellar procedures.3,4,6,7 We propose that eyes with central corneal ablations might be more susceptible to the edematous changes that occur with Fuchs endothelial corneal dystrophy. The edema might create a central steepening anteriorly, creating myopic shift that is subsequently alleviated once the edema is resolved with keratoplasty. This is similar to the posterior flattening that occurs with Fuchs endothelial corneal dystrophy progression.8

It is well established that the most significant mechanism for hyperopic shifts after DSEK is a steepening of the posterior corneal curvature from the added tissue thickness and graft contours that are thicker peripherally. This results in a smaller radius of curvature and higher negative power of the posterior surface, thereby decreasing the total corneal power.3

Development and implementation of ultrathin grafts for DSEK and Descemet membrane endothelial keratoplasty (DMEK) minimize the changes to the posterior curvature. However, hyperopic surprises have still not been completely eliminated with these procedures.6,7 Improving the predictability of the corneal refractive power is especially important when posterior keratoplasty is combined with cataract surgery. Despite improved results compared with DSEK, refractive errors in both the myopic and hyperopic direction can still occur.6–8

Fritz et al. very eloquently demonstrated that refractive errors after combined cases with DMEK can be better predicted by looking at the preoperative posterior Q values.7 They found that corneas with a positive posterior Q value, indicating a flatter, more oblate curvature, were up to 6× more likely to experience a hyperopic shift after combined DMEK/cataract surgery. The central posterior flattening from edematous swelling is a known manifestation of Fuchs endothelial corneal dystrophy progression.3,6 Once the edema has been resolved, the posterior curvature steepens, resulting in a reduction in total corneal power. The study noted, however, that anterior curvatures and Q values were not significantly different postoperatively and did not help predict refractive errors. Their series did not involve patients who had undergone previous refractive surgery.

Pre-LASIK data and tomography were not available to us for the cases in this report. However, the long axial length and large anterior chamber of the first case suggest a highly myopic eye and significant ablation depth. We would hypothesize that, in addition to the posterior edematous changes described by Fritz et al., these patients also experienced anterior bulging in the weakened central ablated cornea, creating a more prolate (−Q value) and steepened anterior surface. Both of these corneal changes will create a myopic shift that would potentially be alleviated with lamellar keratoplasty. Interestingly, the postoperative topography from case 2 showed that the endothelial keratoplasty unveiled a central island that was completely masked prior to the DSEK (Figure 3). This perhaps suggests that the more ablated portion of the cornea is more prone to swelling and increased steepening.

A study by Zeidenweber et al. looked at the outcomes of DMEK in 21 eyes with previous laser refractive surgery.8 They showed that the average anterior and posterior corneal curvatures in these patients did not significantly change 6 months postoperatively and that combined cataract surgeries with DMEK are safe and predictable. However, of the 21 eyes, 4 experienced a flattening of more than 1.00 D, with the most significant case showing a 2.40 D change. The study also did not mention whether any of these cases underwent a myopic or hyperopic correction, nor the preoperative spherical equivalent or ablation depth. It is possible that all these factors could play a part in the predicting which patients will experience a more significant shift in their corneal power after lamellar keratoplasty.

It is not unlikely that many more patients with subclinical Fuchs endothelial corneal dystrophy who have undergone laser refractive surgery will present in later years for cataract surgery and lamellar keratoplasty. Although this report is limited by the available clinical history, variable graft sizes, and lack of standardized advanced imaging of the corneas, it is presented as a counterpoint to the recent studies that demonstrate that the average refractive outcome of a combined cataract/DMEK procedure in eyes with previous laser refractive surgery is predictable with little change to the corneal contour.6,8 Despite these findings, there are still a significant number of outlying cases that can have significant changes to the anterior cornea and high refractive errors. This could be associated with ablation depth, age at the time of refractive surgery, and/or physiological characteristics of the cornea. Because imaging and diagnostic technology advances and electronic medical records allow a more streamlined sharing of information, it will be important to better delineate refractive outcomes in these patients.


  • Descemet-stripping endothelial keratoplasty (DSEK) typically produces a final corrected distance visual acuity that is comparable with that achieved with penetrating keratoplasty, but with faster visual recovery and less-induced irregular astigmatism.
  • The addition of graft tissue in DSEK results in a hyperopic refractive shift due to changes in the posterior curvature.
  • The anterior curvature of post-DSEK corneas remains relatively stable and unchanged from preoperative measurements, contributing very little to the hyperopic changes in corneas that have not undergone previous refractive surgery.


  • Myopic LASIK might make the cornea more prone to swelling and anterior steepening as Fuchs endothelial corneal dystrophy progresses.
  • These cases suggest that, with a subsequent DSEK, the reduction in swelling in the postrefractive cornea might result in significant and unexpected flattening of the anterior corneal curvature.
  • Post-myopic LASIK patients might be more prone to hyperopic errors than patients with no history of corneal refractive surgery after combined cataract–endothelial keratoplasty surgery.


1. Nanavaty MA, Wang X, Shortt A. Endothelial keratoplasty versus penetrating keratoplasty for Fuchs endothelial dystrophy. Cochrane Database Syst Rev. 2014;2:CD008420
2. Dupps WJ Jr, Qian Y, Meisler DM. Multivariate model of refractive shift in Descemet-stripping automated endothelial keratoplasty. J Cataract Refract Surg. 2008;34:578–584
3. Rao SK, Leung CKS, Cheung CYL, Li EYM, Cheng ACK, Lam PTH, Lam DSC. Descemet stripping endothelial keratoplasty: effect of the surgical procedure on corneal optics. Am J Ophthalmol. 2008;145:991–996
4. Clemmensen K, Ivarsen A, Hjortdal J. Changes in corneal power after Descemet stripping automated endothelial keratoplasty. J Refract Surg. 2015; 31:807–812
5. Holladay JT. Holladay IOL Consultant User’s and Reference Manual. Houston, TX, Holladay Lasik Institute, 1999
6. Ham L, Dapena I. Refractive change and stability after Descemet membrane endothelial keratoplasty. Effect of corneal dehydration-induced hyperopic shift on intraocular lens power calculation. J Cataract Refract Surg. 2011;37:1455–1464
7. Fritz M, Grewing V, Böhringer D, Lapp T, Maier P, Reinhard T, Wacker K. Avoiding hyperopic surprises after Descemet membrane endothelial keratoplasty in Fuchs dystrophy eyes by assessing corneal shape. Am J Ophthalmol. 2019;197:1–6
8. Zeidenweber DA, Mayko ZM, Straiko MD, Terry MA. Descemet membrane endothelial keratoplasty in eyes with previous laser refractive surgery: outcomes and complications. Cornea 2017;36:1302–1307
Copyright © 2020 Published by Wolters Kluwer on behalf of ASCRS and ESCRS
Data is temporarily unavailable. Please try again soon.