Recurrent Keratoconus: Corneal Transplants For Keratoconus Develop Tomographic Ectatic Changes

Abstract

Purpose:

To evaluate postoperative Scheimpflug imaging changes during the first five years following penetrating keratoplasty (PK) in patients with keratoconus (KC).

Methods:

This retrospective, interventional case series includes 31 eyes of 31 patients that underwent their first PK with a history of KC. Post-operative Scheimpflug imaging was performed 3 months after removal of the last suture (baseline) and then repeated 3 and 5 years following the PK. Demographic data, donor and host trephination diameter, and Scheimpflug imaging (Pentacam HR, Oculus, Germany) parameters indicative of ectasia were analyzed to evaluate postoperative graft changes that occur following PK.

Results:

The maximal keratometry (Kmax) progressed significantly between baseline (53.5±6.1D) and postoperative year 3 and year 5 (56.5±6.1D and 58.8±7.9D, p<0.001). Significant changes were also observed for the anterior best-fit-sphere and posterior best-fit-sphere (p<0.001 for 3 and 5 years compared to baseline). Kmax increased by at least 2 diopters for 74.2% of patients and up to 7 diopters or more 25.8% of the patients. A significant inverse correlation was observed for host trephine size and progression of Kmax (r=−0.52, p=0.01), which indicated that larger host trephination size was associated with a smaller increase in postoperative Kmax.

Conclusions:

Tomographic graft changes indicative of ectasia were observed within 3 to 5 years following PK in KC patients. These changes were observed more frequently and sooner following corneal transplants than previously reported.

Precis:

Scheimpflug corneal imaging demonstrated ectatic progression within 3 to 5 years after penetrating keratoplasty in keratoconus patients; thus providing evidence that ectasic changes occur more frequently and earlier than previously reported.

INTRODUCTION

The high survival rate of penetrating keratoplasties (PKs) in patients with keratoconus has been well studied ^1–10, however, in comparison, there are relatively few reports on the incidence of corneal ectasia in this patient population. Over the last 30 years, there have been a few studies following PK in KC eyes that yielded conflicting results on the incidence of ectactic changes ^11–15. A retrospective study published in 1995 analyzed mean keratometric values over a 7-year period using a placido disc based keratometer and found that the post-PK eyes did not show an increase in the mean keratometric values, but that the eyes became progressively more myopic following the corneal transplant ¹⁴. In contrast, a study published in 2003 analyzed refractive astigmatism after PK in KC eyes throughout a 20-year follow-up period and found that keratometric astigmatism remained stable for the first seven years, but then began to progressively increase 10 years after suture removal ¹⁵.

Due to technological developments, we now have corneal imaging that has a more sensitive detection of ectactic changes and can potentially identify ectasia within the PK graft. These corneal changes include elevation of the anterior and posterior corneal surface, corneal thinning, apex de-centration, increase in vertical corneal asymmetry and an enhanced identification of ectasia utilizing modified reference spheres etc.¹⁶ This retrospective study evaluates tomographic changes measured by Scheimpflug-based corneal imaging (Pentacam HR, Oculus, Germany) to examine whether tomographic changes indicative of corneal ectasia are present in grafts during the first five years after PK in patients with KC.

MATERIALS AND METHODS

The study adhered to the tenets of the Declaration of Helsinki and was approved by the Institutional Review Board of Massachusetts Eye and Ear Infirmary (IRB number: 2019P000259).

Subjects

The medical records of consecutive eyes of patients with KC who underwent PK between January 2010 and June 2015 at Massachusetts Eye and Ear Infirmary were reviewed. Eyes were excluded with previous PK, eyes that underwent DALK, and eyes that had combined glaucoma or cataract surgery at the time of PK. The Baseline examination was defined as the first examination that occurred 3 months after the last suture was removed from the PK. Tomography examinations were reviewed at baseline, 3 years (32–40 months) and 4.5 years (54–62 months) postoperatively. Eyes were excluded that did not have corneal imaging performed during the three time periods.

Outcome parameters

Data collected included sex, age at time of surgery, phakic or pseudophakic status, donor and host trephination diameter, best spectacles corrected visual acuity (BSCVA), subjective refraction, tomographic and topographic (Pentacam HR, Oculus, Germany) parameters indicative of ectasia such as anterior keratometry (Ant Km), maximal keratometry (Kmax), anterior best-fit-sphere (Ant BFS) using 8 mm optical zone, posterior keratometry (Post Km), posterior best-fit-sphere (Post BFS) using 8 mm optical zone, thinnest local pachymeter (ThinPachy), anterior smallest radius of curvature (RminAnt) using 8 mm optical zone, posterior average radii of curvature (PRC) using 8 mm optical zone., index of surface variance (ISV), index of height asymmetry (IHA), index of vertical asymmetry (IVA), and index of height decentration (IHD).

Surgical technique

The PKs were performed by three experienced surgeons according to previously described techniques ¹⁷. All surgeries were performed under retrobulbar anesthesia of 2% of lidocaine and 0.5 to 0.75% of bupivacaine. A Hessburg-Baron suction trephine was used to cut a full-thickness depth circular incision that was centered on the cornea. Excision of the recipient corneal button was completed with curved corneal scissors. The donor button was trephined with a Barron punch blade with a diameter that was 0.25 or 0.50 mm greater than that used to trephine the host cornea. The oversized graft was sutured in place with 16 interrupted 10–0 nylon sutures in all of the eyes. All sutures were completely removed 12–16 months postoperatively.

Statistical Analysis

Descriptive statistics were used to compare baseline and follow-up characteristics. The Mann-Whitney U-test and paired Student’s t-test were applied for continuous variables and the chi-square or Fisher's exact test when appropriate for nominal variables. All statistical analyses performed were two-sided and statistical significance was set at p-value of 0.05. Data were analyzed with the Minitab, version 16 (Minitab Inc, State College, Pennsylvania, USA).

RESULTS

120 eyes were identified as having KC and a PK procedure between January 2010 and June 2015. According to the criteria, eyes with previous PK, eyes that underwent DALK, eyes that had combined glaucoma or cataract surgery at the time of PK, and eyes which did not have tomography examinations performed during the three postoperative time periods of interest were excluded . This study included the remaining 31 consecutive eyes from 31 patients whose demographic and preoperative clinical characteristics are described in Table 1.

Table 1.

Demographic, clinical, and procedural characteristics from the eyes with a history of keratoconus that underwent penetrating keratoplasty

Parameter	N (%) / Mean ± SD
Male	19 (61%)
Age at time of surgery (years)	44.3±17.0
Mean Host size trephine (mm)	8.0±0.3
Mean Donor size trephine (mm)	8.4±0.3
Phakic eyes	31 (100%)
Preoperative BSCVA (LogMAR)	0.86±0.37

Legend: Best spherical corrected visual acuity (BCVA), Standard deviation (SD), Logarithm of the Minimum Angle of Resolution (LogMAR).

Keratometry Parameters

The baseline Scheimpflug-based Kmax, which is an indicator of ectatic progression ¹⁸, was obtained on the first visit following removal of the last suture. Compared to the baseline Kmax of 53.5±6.1 D (mean ±standard deviation), mean Kmax (56.5±6.1) significantly increased (p<0.001 using paired t-test) after 3 years (Figure 1). The mean Kmax also significantly increased between 3 years and 4.5 years (58.9±7.9, p<0.001) after PK. Twenty-four eyes (77.4%) demonstrated a Kmax increase of 1 D or more, 23 eyes (74.2%) of 2 D or more, 19 eyes (61.3%) of 3 D or more, 11 eyes (35.5%) of 5 D or more and 8 eyes (25.8%) of 7 D or more at 5 years postoperatively. Of the 26 eyes that had pre-operative imaging performed, there was no significant difference in preoperative Kmax between patients who had a greater than 2D progression in Kmax (71.4+/−13.2) and those that did not (74.3+/−13.9) (p=0.68; data not shown).

Figure 1. Topographic and tomographic indices following penetrating keratoplasty in eyes with a history of keratoconus.

Topographic and tomographic parameters were used to investigate trends in postoperative keratometry. A statically significant change was observed for Kmax (maximal keratometry), Ant BFS (anterior best-fit-sphere), and Post BFS (posterior best-fit-sphere) over the first five postoperative years after penetrating keratoplasty. P-values were calculated by the Mann-Whitney U-test and paired Student’s t-test were applied for continuous variables and the chi-square or Fisher's exact test when appropriate for nominal variables. Top, Kmax. Middle, Ant BFS. Bottom, Post BFS.

Additional tomographic parameters were analyzed to investigate trends in postoperative keratometry (See Table 2 and Figure 1). Ant BFS and Post BFS became significantly steeper from baseline to the 3-year post-operative examination (p<0.001 for Ant BFS and p<0.004 Post BFS); and these values continued to become steeper between the 3-year and 5-year (p=0.001 for Ant BFS and p=0.001 for Post BFS) post-operative examinations. RminAnt demonstrated a statistically significant steepening when the 5-year was compared to 3-year post-operative examinations (p=0.003). The thinnest corneal pachymetry value significantly increased after 3 years compared to baseline (p=0.02), but this difference was not observed at 5 years.

Table 2. Vision, refraction, and tomographic data following penetrating keratoplasty in eyes with a history of keratoconus.

Kmax (maximal keratometry) and other tomographic and topographic parameters were used to investigate trends in postoperative keratometry and the correlations between other visual and keratometry outcomes. Kmax, Ant BFS (anterior best fit sphere), Post BFS (posterior best fit sphere), and showed a statistically significant change over the first five postoperative years. P-values were calculated by the Mann-Whitney U-test and paired Student’s t-test were applied for continuous variables and the chi-square or Fisher's exact test when appropriate for nominal variables.

Parameter	Baseline examination	3 years	P-value comparing baseline to 3 years	4.5 years	P-value comparing baseline to 4.5 years	P-value comparing 3 years to 4.5 years
Visual acuity and subjective refraction
BCVA (LogMAR)	0.3±0.2	0.2±0.2	0.008	0.2±0.1	0.84	0.04
SE (D)	−2.4±3.7	−2.1±20.6	0.30	−1.3±2.7	0.35	0.34
Sphere (D)	−0.5±3.5	3.5±20.4	0.35	0.7±2.6	0.40	0.22
Cylinder (D)	−3.8±2.2	−3.1±2.5	0.17	−3.9±1.8	0.15	0.73
Tomographic indices
Ant Km (D)	44.2±3.6	44.8±3.8	0.12	45.1±4.2	0.06	0.047
Kmax	53.5±6.1	56.5±6.1	<0.001	58.9±7.9	0.007	<0.001
AntBFS (mm)	7.6±1.0	7.3±0.9	<0.001	7.1±0.6	0.09	<0.001
AntAst (D)	5.5±5.0	5.9±4.4	0.57	5.9±4.6	0.68	0.80
PostKm (D)	7.7±7.2	6.5±0.5	0.37	6.3±1.1	0.46	0.30
PostBFS (mm)	6.2±0.6	5.9±0.5	0.004	5.9±0.5	0.21	0.001
PostAst	5.7±18.8	0.8±0.5	0.21	1.0±0.8	0.23	0.84
ThinPachy (µm)	504.7±49.9	523.2±52.7	0.02	518.1±97.2	0.63	0.57
RminAnt	6.4±0.7	7.0±5.2	0.52	5.4±1.5	0.03	0.003
ARC	7.9±1.2	7.6±1.2	0.10	7.6±1.1	0.45	0.049
PRC	5.7±0.6	5.6±0.6	0.66	6.0±1.7	0.39	0.49
ISV	106.4±95.6	101.1±90.0	0.52	107.3±73.4	0.89	0.71
IVA	1.0±1.3	0.9±1.2	0.16	1.0±1.0	0.63	0.24
IHA	35.1±37.6	33.9±31.6	0.84	45.3±43.5	0.15	0.48
IHD	0.1±0.1	0.1±0.1	0.62	0.1±0.2	0.15	0.16

Legend: Best-corrected visual acuity (BCVA), Logarithm of the Minimum Angle of Resolution (LogMAR), Anterior keratometry (Ant Km), Maximal keratometry (Kmax), Anterior best-fit-sphere (Ant BFS), Posterior keratometry (Post Km), Posterior best-fit-sphere (Post BFS), Thinnest Local Pachymeter (ThinPachy), Anterior smallest radius (RminAnt), Anterior average radii of curvature (ARC), Posterior average radii of curvature (PRC), Index of surface variance (ISV), Index of height asymmetry (IHA), Index of vertical asymmetry (IVA), Index of height decentration (IHD).

Host trephine size

We analyzed the correlation between host trephination size and the postoperative Kmax to determine whether excision of a larger diameter of host tissue reduced corneal steepening. The host trephination size ranged from 7.5mm to 9.00mm in diameter. There was a significant inverse correlation between host trephine size and progression of Kmax (r=−0.52, p=0.01) indicating that larger host trephination size was associated with less progression of postoperative Kmax (Figure 2).

Figure 2. The correlation between host trephination size and progression of postoperative maximal keratometry (Kmax) following penetrating keratoplasty in eyes with a history of keratoconus.

A range of 7.5mm-9.00mm diameter was used for host trephination. There was a significant inverse correlation between the diameter of the host trephine and progression of maximal keratometry (Kmax) (r=−0.52, p=0.01) indicating that the larger host trephination size was associated with less postoperative Kmax progression. P-values were calculated by the Mann-Whitney U-test and paired Student’s t-test were applied for continuous variables and the chi-square or Fisher's exact test when appropriate for nominal variables.

Graft Oversize

The difference between the diameter of the trephine used to excise the donor tissue (transplanted corneal donor button) and the diameter of the trephine used to remove tissue from the host cornea (oversize) was 0.25mm for 10 eyes and 0.50mm for 21 eyes. The changes of the Kmax value during the study period was 5.9 ± 6.7 for the 0.25 group and 5.9 ± 5.0 for the 0.50 group (p=1.00, calculated by Student t test comparing 0.25 to 0.50 mm trephine size).

Age

The association between age and evidence of ectatic changes was analyzed to determine whether there was a positive correlation between age and progression of keratoconus, as previously reported. In this study, the correlation between age and progression of postoperative Kmax was not statistically significant (r=−0.20, p=0.38, Table 2).

DISCUSSION

This study sought to determine whether rotating Scheimpflug-imaging detected tomographic changes consistent with corneal ectasia during the first 5 years following PK in patients with KC. In general, rotating Scheimpflug-imaging enables the diagnosis of corneal ectasia prior to the presentation of clinical findings and symptoms ¹⁶. As a result, Scheimpflug imaging is commonly used as a screening tool to identify corneal ectasia prior to corneal refractive surgery ¹⁶. In addition, corneal ectasia (e.g., KN) progression can be detected using Scheimpflug-imaging ¹⁹.

This study found that the anterior curvature and posterior curvature of PK grafts in KC eyes increased during the first 4.5 postoperative years. Compared to baseline, the Kmax increased by an average of 5 diopters 4.5 years after PK, with a corresponding significant decrease in AntBFS and PostBFS. Over 70% (16 out of 22 eyes) demonstrated more than 2 D increased Kmax after 4.5 years and 36% (8 of 22 eyes) of the eyes exhibited at least 7 D of increased Kmax. The average Kmax at the baseline measurement was 53.5 ± 6.1, which is steeper than the average cornea. Potential reasons for this include that the anterior corneal steepening had already started at 1 year after surgery and that there could be focal areas of corneal steepening that would increase the Kmax.

We found that the RminAnt and ARC also decreased significantly, indicating an increase in steepness, whereas PRC, ISV, IHA, IHD and IVA did not demonstrate significant changes. These conflicting outcomes with regard to topometric indices are not surprising since the indices and their ranges are based on untreated eyes, and thus were not designed to detect ectasia changes after PK in keratoconus patients. Currently no indices after PK exist to objectively evaluate ectatic changes after PK in KC patients.

A significant inverse correlation was observed between the size of the host trephine and progression of Kmax (r=−0.52, p=0.01), which indicated that larger host trephination diameter was associated with less postoperative ectasia. This result is consistent with a previous study that found that larger graft size resulted in superior visual outcomes^20,21. In contrast, our results are inconsistent with a different study that used an excimer laser for trephination of the donor and host tissue and found that a smaller graft diameter resulted in flatter curvature and greater topographic irregularity²². The fact that larger trephinations have lower recurrence rates could suggest that progression of the host rim could play a role in the ectasia. While our results and other have shown that larger graft size may be advantageous for keratoconus eyes, there may be a limit to the trephination size out of concern that a larger graft could result in a higher rate of immunologic rejection^5,23,24. Therefore, it has been suggested that for eyes with keratoconus, grafts should be as large as possible for less postoperative ectasia, and as small as needed to avoid immunologic rejection ^23,25. With regard to corneal crosslinking, it is unclear if there would be an advantage to treating a larger area of the cornea to help reduce progression of keratoconus. But no studies on that specific topic are available yet and that further research is needed.

The absolute change in Kmax during the follow-up was not significantly different between patients in which 0.25 mm or 0.5 mm donor oversizing was used during keratoplasty. Previous reports show that reducing recipient-donor trephine disparity can decrease myopia in KC patients after PK ^26,27, or undersized grafts can lead to hypermetropic shifts. ²⁸　In this study we compared the keratometry values to a baseline examination taken at least three months after suture removal, i.e., at least 15 months postoperatively. There is a possibility that the effect of graft-host oversizing or localized graft-host disparity on the graft’s geometry may have already been present by the observation period of the study and may not have progressed during the subsequent follow up period. Therefore, it can be said that our result is not necessarily inconsistent with previous reports.

The study provides evidence of corneal steepening over time by observing that there is an increase in Kmax, a decrease in anterior BFS and posterior BFS, indicating a recurrence of ectasia after PK. Limitations to this study include its retrospective nature, the rather small sample size analyzed and the lack of a control group (non-KC patients). There are no prior studies available reporting the prevalence of post-keratoplasty ectasia in non-KC patients except for one case report where the donor eye was obtained from a patient with KC ²⁹. While there are at least two case reports describing recurrent ectasia after DALK in KC eyes ^30,31, this study excluded DALK procedures because these cases could have introduced confounding effects. For instance, multivalent analyses found that DALK cases have spherical equivalents that were less negative than PKPs ³² and reduced refractive astigmatism³³ compared to PKs. Moreover, oblate asymmetric bow tie was more commonly seen in DALK than in PKs.³⁴ Due to the retrospective nature of the study, vision and refraction were not included in the analysis of this study because they were not performed or recorded in a standardized manner.

It is important to note that patients included in this cohort did not undergo collagen cross-linking, which was FDA approved in 2016. Although collagen cross-linking does not commonly include the corneal periphery, in some patients it may include the area that will become the graft-host junction. Future research may examine whether collagen cross-linking has any effect in preventing post-keratoplasty ectasia in patients with KC.

In conclusion, the majority of KC eyes demonstrate ectatic changes within the first 5 years after PK, thus earlier than previously reported. Future long-term studies are needed to determine if the ectatic changes observed by tomography precede clinical findings consistent with recurrence of ectasia in the PK grafts of eyes with KC.