• Users Online: 22
  • Home
  • Print this page
  • Email this page
Home About us Editorial board Ahead of print Current issue Search Archives Submit article Instructions Subscribe Contacts Login 

 Table of Contents  
Year : 2019  |  Volume : 20  |  Issue : 3  |  Page : 112-120

Correlation of graft–host interface to refractive outcomes after penetrating keratoplasty using anterior segment optical coherence tomography

Department of Ophthalmology, Menoufia University, Shebeen El-Kom, Egypt

Date of Submission11-Dec-2018
Date of Acceptance20-Mar-2019
Date of Web Publication26-Sep-2019

Correspondence Address:
MD Mohamed S Abd Elaziz
Department of Ophthalmology, Faculty of Medicine, Menoufia University, Shebeen El-Kom 32511
Login to access the Email id

Source of Support: None, Conflict of Interest: None

DOI: 10.4103/DJO.DJO_65_18

Rights and Permissions

Objective The aim of this study was to evaluate the correlation between the graft–host interface and the refractive outcomes after penetrating keratoplasty (PKP) using anterior segment optical coherence tomography.
Patients and methods A total of 40 eyes of 38 patients with different corneal pathologies (bullous keratopathy, seven eyes; corneal scar, 20 eyes; and keratoconus, 13 eyes) who underwent PKP were retrospectively reviewed. The wound profiles of the graft–host junctions were classified into well-apposed junction and malapposed one including gap, step, and protrusion with the aid of spectral domain anterior segment optical coherence tomography. The correlations between clinical outcomes and post-PKP wound characteristics were analyzed.
Results A total of 320 graft–host junctions from 40 eyes were examined, of which, 130 (40.6%) sections showed a well-apposed junction and 190 (59.4%) sections represented a malapposed junction. The most recorded type of malapposition was protrusion (97 sections=30.3%). When referring to the underlying diseases, well-opposed junctions were most frequently met in bullous keratopathy (66.1%). The alignment pattern showed highly significant differences between the preoperative original disease groups (P=0.001). Spherical equivalent and corrected distant visual acuity showed highly significant differences between the alignment groups, with highest myopic shift in the gap malapposition. Graft–host thickness disparities showed no significant correlations with spherical equivalent (r=−0.184, P= 0.256).
Conclusion The alignment pattern of graft–host junctions after PKP varied according to the preoperative original disease and was significantly associated with spherical equivalent and visual acuity.

Keywords: anterior segment optical coherence tomography, graft–host junction, penetrating keratoplasty

How to cite this article:
Elsobky HM, Abd Elaziz MS, El Hamadan IG. Correlation of graft–host interface to refractive outcomes after penetrating keratoplasty using anterior segment optical coherence tomography. Delta J Ophthalmol 2019;20:112-20

How to cite this URL:
Elsobky HM, Abd Elaziz MS, El Hamadan IG. Correlation of graft–host interface to refractive outcomes after penetrating keratoplasty using anterior segment optical coherence tomography. Delta J Ophthalmol [serial online] 2019 [cited 2020 Oct 25];20:112-20. Available from: http://www.djo.eg.net/text.asp?2019/20/3/112/267944

  Introduction Top

The main aim of penetrating keratoplasty (PKP) is to accomplish a precise apposition between the graft and host, which is an important factor affecting the prognosis postoperatively [1]. PKP has a high success rate (80%) in low-risk patients [2], but postoperative astigmatism is an important cause of diminution of vision postoperatively. Sutures (tense, uneven, and strong), misaligned junction between graft and host, and the wound-healing condition (delayed or irregular) cause post-PKP irregular astigmatism [3]. The previous reports noted that the mean post-PKP astigmatism had a range of 3.5–5.6 D following complete suture removal [4],[5],[6],[7]. In addition, retrocorneal membrane formation and graft rejection are caused by delayed or irregular wound healing and improper alignment at the wound interface causing dehiscence of the wound [8]. So, correlation of the graft–host interface is valuable for the postoperative management of patients with PKP.

Hence, the optical coherence tomography (OCT) has a massive recent development, that is, now we can obtain fully detailed anterior segment images with high-resolution using a contactless and noninvasive procedure and thus can assess the wound interface much easier. Previously slit-lamp biomicroscopy was used alone, but it provides limited data about the internal aspect of the wound interface. Now with anterior segment optical coherence tomography (AS-OCT), we can obtain fully detailed cross-sectional images of the cornea including anterior and posterior surfaces [9],[10].

The time-domain technology of AS-OCT is applied for various corneal conditions, such as scar, dystrophy, ectatic disorder, and intrastromal foreign body as a diagnostic and management tool [11].

There are only few studies that have analyzed the wound profile after PKP using AS-OCT [12],[13].

In this study, the morphological characteristics of the graft–host interface were reported by AS-OCT in patients who underwent PKP. These findings were correlated to the clinical outcomes.

  Patients and methods Top

The research was approved by the Ethical Committee of Menoufia University, and a written informed consent was signed by patients before being enrolled in the study.

This is a cross-sectional study that involved patients who underwent PKP at Menoufia University Hospital and who were followed up for at least 1 year with all stitches removed. All patients underwent full ophthalmologic examination including refraction, corrected distance visual acuity, slit-lamp biomicroscopy, intraocular pressure measurement, keratometric astigmatism using automated keratometry, and evaluation of the graft–host interface by a Topcon 3D OCT-2OOO (Topcon Medical Systems Inc. Paramus, New Jersey, USA).

Patients included in this study had keratoconus, corneal opacities, or bullous keratopathy. Patients with unremoved stitches or with other postoperative complications were excluded. The study protocol was approved by the Ophthalmology Department of Menoufia University, Egypt.

Surgical technique

Surgeries were performed by a single surgeon (M.S.A.) under general anesthesia. The donor was excised 0.25 mm larger than the diameter of the recipient using a Moria vacuum punch with a diameter of 7.75 mm (Moria S.A., Antony, France). After marking the center of the recipient, the host tissue was punched using a Moria radial vacuum trephine with the diameter of 7.50 mm. The cut was completed by curved corneal scissors. The wound was sutured by using a 10–0 nylon monofilament in a 16-bite interrupted fashion. All patients included in the study had their sutures removed before one year.

Anterior segment optical coherence tomography acquisition

The Topcon 3D OCT-2OOO was used to acquire the postoperative graft–host junction. It is based on spectral domain technology to obtain high-definition scans. The evaluations were performed after all corneal sutures were removed. One experienced examiner (H.M.E) scanned four high-resolution optical sections in the interval of 45° to obtain eight images of each graft–host junction.

Each graft–host junction was distributed according to the alignment pattern of the corneal endothelial side as follows: well-apposed junction if the corneal internal sides were aligned precisely without disconnection; gap if the Descemet’s membrane and the inner stroma of the donor and recipient were separated, but were aligned; step if the wound interface was not separated, but revealed thickness disparity between the graft and host; and protrusion if the Descemet’s membrane and the stroma protruded toward the anterior chamber ([Figure 1]).
Figure 1 Classification of wound profile using anterior segment optical coherence tomography. (a) Well-apposed junction, (b) gap, (c) step, and (d) protrusion.

Click here to view

The central corneal thickness was measured using the Topcon 3D OCT-2OOO. The thickness disparity at the wound interface was calculated by drawing an imaginary line perpendicular to the outer and inner sides of the cornea. The thickness disparity was calculated by measuring the thicknesses of the donor and recipient corneas, each at the point that was 1 mm away from the merging point on the external side inward. The eight calculated values were averaged ([Figure 2]).
Figure 2 Measurement of thickness disparity between graft and host corneas at the wound interface.

Click here to view

According to the clinical diagnosis before surgery, the patients were allocated into the corneal scar, bullous keratopathy, or keratoconus group. The distribution of alignment patterns in each diagnosis group was evaluated. In addition, the eyes were classified into the well-apposed junction, gap, step, or protrusion group based on the most frequent type among the eight wound alignment patterns in each eye.

Patients’ age, sex, visual acuity, intraocular pressure, spherical equivalent, keratometric astigmatism, central corneal thickness, and thickness disparity at the wound interface were compared between the alignment groups.

Statistical analysis

SPSS version 22 (SPSS Inc., Chicago, Illinois, USA) was used for statistical analysis. For the comparison of the categorical variables between groups, χ2-test was performed. After verifying the normality of distribution using the Shapiro–Wilk normality test, we evaluated the differences in continuous variables between the groups using an independent t-test, Mann–Whitney U-test, or Kruskal–Wallis test. The Pearson correlation coefficient was used to assess the correlating factor associated with thickness disparity at the wound interface. A P value less than 0.05 was considered statistically significant.

  Results Top

This study included 40 eyes of 38 patients. The mean age of the patients was 43.9±15.9 years (range: 18–67 years). The mean follow-up period after PKP was 15.9±2.65 months (range: 12–23 months).

The number of graft–host interface images from the 40 eyes obtained by the Topcon 3D OCT-2OOO was 320 cross-sections, of which, 59.4% (190 cross-sections) had malapposed junction and 40.6% (130 cross-sections) had well-apposed junction. Protrusion (97 cross-sections, 30.3%) was the most frequent misalignment type. Among the alignment patterns, well-apposed junction occurred in 41.9, 66.1, and 25.0% in the corneal scar, bullous keratopathy, and keratoconus groups, respectively ([Table 1]).
Table 1 Prevalence of various alignment patterns found in the internal graft–host junctions according to the preoperative diagnosis

Click here to view

The indications of PKP were corneal scar (20 eyes, 50%), keratoconus (13 eyes, 32.5%), and bullous keratopathy (seven eyes, 17.5%). Comparing the baseline characteristics between the diagnosis groups, statistically highly significant differences (P=0.001) in age, mean time after operation, intraocular pressure, and spherical equivalent were found. There was no significant difference in sex, visual acuity, keratometric astigmatism, and central corneal thicknesses among the diagnosis groups. Thickness disparities between the graft and recipient measured at the wound interface in the corneal scar, bullous keratopathy, and keratoconus groups were 61.60±23.30, 52.00±28.9, and 63.00±26.10 μm, respectively, and showed no differences among the diagnosis groups (P=0.41; [Table 2]).
Table 2 Relation between alignment patterns and clinical data of studied groups

Click here to view

When classifying the eyes according to the most frequent alignment pattern among the eight cross-sections, 18 eyes had the well-apposed junction state (corneal scar, nine eyes; bullous keratopathy, five eyes; and keratoconus, four eyes). Nineteen eyes were classified as the mal-apposed junction state (corneal scar, 10 eyes; bullous keratopathy, two eyes; and keratoconus, seven eyes) and three eyes were classified as the others (corneal scar, one eye, and keratoconus, two eyes). Among the19 eyes with mal-apposed junction, 10 eyes had protrusion, two eyes had gap, and seven eyes had a step. After distributing the eyes into the well-apposed and malapposed junction states, the distribution of the alignment pattern was highly significantly different between the preoperative diagnosis groups (P=0.001). Although the most frequent type of alignment pattern in the corneal scar and bullous keratopathy groups was the well-apposed junction, the keratoconus group showed a higher incidence of malapposed junction ([Figure 3]).
Figure 3 Alignment patterns found in the internal graft–host junctions.

Click here to view

Comparisons of various clinical parameters between the well-apposed junction and the malapposed junction group (combination of gap, step, and protrusion) are summarized in [Table 3]. Spherical equivalent (P=0.005) and thickness disparity at the wound interface (P=0.001) were significantly different between the alignment groups.
Table 3 Correlation between thickness disparity at the wound interface and various parameters

Click here to view

In the analysis of the factors related to the thickness disparity at the wound interface, a highly significant negative correlation between thickness disparity and corrected distance visual acuity among studied groups was found (P=0.001; [Table 3]).

The relation between different alignment patterns to best-corrected visual acuity, spherical equivalent, corneal thickness, and thickness disparity are illustrated in [Figure 4],[Figure 5],[Figure 6],[Figure 7].
Figure 4 Relation between alignment patterns and best-corrected visual acuity (BCVA) of studied groups.

Click here to view
Figure 5 Relation between alignment patterns and spherical equivalent of studied groups.

Click here to view
Figure 6 Relation between alignment patterns and corneal thickness of studied groups.

Click here to view
Figure 7 Relation between alignment patterns and thickness disparity at the wound (μm) among studied groups.

Click here to view

  Discussion Top

Visual outcomes after full-thickness corneal grafting depend largely on the healing pattern of graft–host interface. Postoperative astigmatism (either regular or irregular) counts on proper suturing techniques that should ensure proper tissue alignment; this can now be assessed by using recent advances in AS-OCT [14].

Some reports investigated the alignment pattern of the graft–host interface after PKP using AS-OCT. Kaiserman et al. [6] reported that the degree of malapposition was associated with higher postoperative astigmatism. They handled eight images from one eye as eight different data sets in statistical analysis. However, as these eight data sets were obtained from one eye, it is inconclusive whether the statistical results are really significantly different. Jhanji et al. [5] also conducted a similar study and categorized the alignment pattern into the two groups, step and ledge. As they did not evaluate the characteristics of each alignment group, it was difficult to know whether the alignment pattern after PKP affects the postoperative outcome. Hence, in the present study, we evaluated the wound profile of post-PKP eyes and the factors related to the wound alignment pattern.

As the external alignments are easy to identify during surgery and the epithelial cells have a tendency to fuse together, the anterior surface of the graft–host junction maintains the aligned state without disconnection or protrusion after PKP. In contrast, the endothelial side of the graft–host junction is not well-apposed in many cases [1],[12],[13]. Lang et al. [7] performed a histological analysis of postmortem corneas that underwent PKP and reported misalignment of Descemet’s membrane in 22 of 30 eyes. In the present study, 60.5% of 392 graft–host junctions showed the malapposed state.

Among the malapposed junctions, protrusion was the most common misalignment type. This might be explained by the curling up of large donor on the wound interface. In addition, the tendency for stromal overgrowth induced by incomplete contact between the donor and recipient may affect the wound profile [13]. When a tissue is injured and lost, migration and proliferation of connective tissue occur to fill in the defect as a normal wound-healing process.

Similar to a previous report [12], in the analysis of the alignment pattern according to the preoperative diagnosis, in the present study, 57.2% of the bullous keratopathy group showed well-apposed junction, whereas only 21.2% of the keratoconus group showed well-apposed junction. Furthermore, 43.9% of the graft–host interfaces in the keratoconus group were step, showing a higher incidence than the other groups. As the replaced region in PKP is the only central part of the host cornea, the remaining thin peripheral part might cause a thickness mismatch between the graft and the recipient. In the same context, bullous keratopathy with edematous thick recipient owing to endothelial cell dysfunction was also expected to show a higher incidence of graft–host misalignments like a step or protrusion. However, in contrast to previous reports [12],[13], 57.2% of the graft–host junctions in the bullous keratopathy group, in the present study, showed a well-apposed state. The surgical techniques including suture tension and the severity of bullous keratopathy in this study’s patients might have influenced the alignment patterns.

Additionally, after dividing the post-PKP eyes into two alignment patterns, the well-apposed and mal-apposed junction, we found significant differences in the alignment pattern between the preoperative diagnosis groups. According to Jhanji et al. [5], the degree of graft oversize affects the alignment patterns of a wound interface. The reason was explained by the curling up of the large donor on the posterior surface adjacent to the graft–host junction. To exclude the effect of the degree of graft oversize, we further analyzed the alignment patterns after selecting patients who had the same degree of graft oversize. Although it did not reach statistical significance, we also found a trend of different wound alignment patterns according to the preoperative diagnosis. The result suggests that the underlying diseases might affect the alignment pattern of the graft–host interface.

After dividing the post-PKP eyes into the two groups, the well-apposed junction and malapposed junction group, we found significant differences in spherical equivalent and keratometric astigmatism. Of course, we cannot determine the cause of these differences in spherical equivalent. In addition to the influence of astigmatism, axial length and lens status might have affected the results.

However, regarding keratometric astigmatism, the results of this study suggest that the alignment state has a significant effect on the cylindrical power.

Elskhawy et al. [15] conducted a similar cross-sectional study on 27 eyes for whom PKP was performed. Of these, 22 (18.33%) cross-sections had well-apposed junction and 98 cross-sections had malapposed junction (81.67%). Protrusion (50 cross-sections; 41.67%) was the most frequent misalignment type. This is comparable to the present study regarding the prevalence of protrusion pattern. However, the previous study was limited to keratoconus and corneal scaring for keratoplasty.

Thickness disparity at the wound interface is closely related to the alignment state. As we expected, it was lowest in the well-apposed junction group. One of the interesting findings was that the degree of thickness disparity showed a similar tendency with keratometric astigmatism using a manual keratometer. In the well-apposed junction group with the lowest thickness disparity, the lowest keratometric cylinder power of 2.91±1.41 D was found. In contrast, the protrusion group with the highest disparity showed the highest keratometric cylinder power of 5.25±1.52 D. Calculating the Pearson correlation coefficients showed significant positive correlations between the two variables: degree of thickness disparity and keratometric astigmatism. This finding may explain why the keratoconus group showed the highest keratometric astigmatism of 5.06±1.38 D.

This study has several limitations. First, because the study population consisted of patients with selected underlying diseases and the number of patients in each group was relatively small, the results may not be applicable to all post-PKP cases. A large-scale study on various preoperative clinical diagnoses will be helpful to understand the alignment patterns of the wound interface after PKP. Second, we excluded patients with postoperative complications. Investigating the relationship between the wound profiles and post-PKP complications will be helpful for the management of post-PKP patients. Third, measurement of keratometric astigmatism was performed with a manual keratometer only, whose range of measurement was limited within 3 mm of the central cornea. If an analysis on keratometric astigmatism using other instruments including corneal topography is conducted, a more accurate correlation will be obtained. Finally, as we obtained the eight representative images of the graft–host interface after scanning four optical sections in the interval of 45°, variations in graft–host interfaces may exist that are missed by these images.

  Conclusion Top

The alignment state at the wound interface differed according to the preoperative diagnosis, and the alignment pattern of the graft–host interface significantly correlated with keratometric astigmatism and degree of graft oversize. AS-OCT can be used valuably for the evaluation of the graft–host interface, which is very important for the management of post-PKP patients.

Financial support and sponsorship


Conflicts of interest

There are no conflicts of interest.

  References Top

Binder PS. The effect of suture removal on postkeratoplasty astigmatism. Am J Ophthalmol 1988; 105:637–645.  Back to cited text no. 1
Ing JJ, Ing HH, Nelson LR, Hodge DO, Boume WM. Ten-year postoperative results of penetrating keratoplasty. Ophthalmology 1998; 105:1855–1865.  Back to cited text no. 2
Farid M, Kim M, Steinert RF. Results of penetrating keratoplasty performed with a femtosecond laser zigzag incision initial report. Ophthalmology 2007; 114:2208–2212.  Back to cited text no. 3
Jancevski M, Foster CS. Anterior segment optical coherence tomography. Semin Ophthalmol 2010; 25:317–323.  Back to cited text no. 4
Jhanji V, Constantinou M, Beltz J, Vajpayee RB. Evaluation of posterior wound profile after penetrating keratoplasty using anterior segment optical coherence tomography. Cornea 2011; 30:277–280.  Back to cited text no. 5
Kaiserman I, Bahar I, Rootman DS. Corneal wound malapposition after penetrating keratoplasty: an optical coherence tomography study. Br J Ophthalmol 2008; 92:1103–1107.  Back to cited text no. 6
Lang GK, Green WR, Maumenee AE. Clinicopathologic studies of keratoplasty eyes obtained post mortem. Am J Ophthalmol 1986; 101:28–40.  Back to cited text no. 7
Lim L, Pesudovs K, Coster DJ. Penetrating keratoplasty for keratoconus: visual outcome and success. Ophthalmology 2000; 107:1125–1131.  Back to cited text no. 8
Mohamed S, Lee GK, Rao SK, Wong AL, Cheng AC, Li EY et al. Repeatability and reproducibility of pachymetric mapping with Visante anterior segment-optical coherence tomography. Invest Ophthalmol Vis Sci 2007; 48:5499–5504.  Back to cited text no. 9
Ramos JL, Li Y, Huang D. Clinical and research applications of anterior segment optical coherence tomography − a review. Clin Exp Ophthalmol 2009; 37:81–89.  Back to cited text no. 10
Sayegh RR, Pineda R. Practical applications of anterior segment optical coherence tomography imaging following corneal surgery. Semin Ophthalmol 2012; 27:130–137.  Back to cited text no. 11
Sung MS, Yoon KC. Evaluation of graft-host interface after penetrating keratoplasty using anterior segment optical coherence tomography. Jpn J Ophthalmol 2014; 58:282–289.  Back to cited text no. 12
Troutman RC, Lawless MA. Penetrating keratoplasty for keratoconus. Cornea 1987; 6:298–305.  Back to cited text no. 13
Lim LS, Aung HT, Aung T, Tan DT. Corneal imaging with anterior segment optical coherence tomography for lamellar keratoplasty procedures. Am J Ophthalmol 2008; 145:81–90.  Back to cited text no. 14
Elskhawy ZM, Sabry MM, Alshorbagy MS, Selima AA. Evaluation of patients with keratoplasty in Tanta University Hospital using anterior segment optical coherence tomography. Delta J Ophthalmol 2018; 19:237–242.  Back to cited text no. 15
  [Full text]  


  [Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5], [Figure 6], [Figure 7]

  [Table 1], [Table 2], [Table 3]


Similar in PUBMED
   Search Pubmed for
   Search in Google Scholar for
 Related articles
Access Statistics
Email Alert *
Add to My List *
* Registration required (free)

  In this article
Patients and methods
Article Figures
Article Tables

 Article Access Statistics
    PDF Downloaded82    
    Comments [Add]    

Recommend this journal