Dr Raul Martin, IOBA Eye Institute, C/ Paseo Belen 17, E-47011 Valladolid, SPAIN. E-mail: firstname.lastname@example.org
Purpose: The aim was to describe the forward shift of the posterior corneal surface one year after myopic laser in situ keratomileusis (LASIK) relative to the estimated residual stromal bed thickness (RBT) and to correlate RBT with the ablation percentage per total corneal thickness (CT) in eyes without significant changes in anterior best-fit sphere (BFS) and corneal thickness one year after LASIK.
Methods: The anterior and posterior BFS and central and peripheral corneal thickness were measured in 86 eyes from 43 patients (mean age 32.07 ± 7.01 years), who underwent uneventful LASIK for myopia using Orbscan before and one, six and 12 months after LASIK. The patients were divided into three groups based on estimated RBT (Group 1 with 16 eyes [19%], RBT less than 250 µm, Group 2 with 52 eyes [58%], RBT from 250 to 300 µm and Group 3 with 20 eyes [23%], RBT greater than 300 µm) and into two groups based on ablation percentage (Group A, 52 eyes [60%], less than 20% and Group B, 34 eyes [40%], more than 20%).
Results: The mean posterior forward shifting was 9.4 ± 14.3 µm (range -27 to 80 µm) (p < 0.05), 4.0 ± 9.8 µm (range -34 to 24 µm) (p > 0.05) and 3.1 ± 8.8 µm (range -37 to 21 µm) (p > 0.05) at one, six and 12 months following LASIK, respectively. The posterior corneal forward shift was higher and showed an increasing trend in Group 1 (r2= 0.111), Group 2 (r2= 0.185) and Group B (r2= 0.156). Non-significant (p > 0.05) posterior elevation was found in Group 3. The anterior BFS and central and peripheral corneal thickness did not show statistically significant differences (p > 0.05) for all post-LASIK follow-up visits.
Conclusions: The posterior corneal displacement measured with Orbscan after LASIK was time dependent, with a different trend between estimated residual stromal bed thickness that protruded at an early stage but then returned to original levels 12 months after LASIK in eyes without post-LASIK anterior topographic or pachymetric changes.
Laser in situ keratomileusis (LASIK) is a successful surgical method for the correction of myopia; however, this procedure can weaken the stroma beneath and there might be increased risk for keratectasia post-operatively.1–3
Corneal ectasia after LASIK results in an increase in myopia and astigmatism and loss of visual acuity due to progressive corneal steepening that takes place centrally or inferiorly1–5 and can occur early2 or be delayed up to several months4 after the surgery (mean of 13 months after surgery).5 The incidence of keratectasia after myopic LASIK has been estimated to be 0.04 to 0.6 per cent.2,3,6
Unfortunately, the exact causes of iatrogenic keratectasia after myopic LASIK are unknown.2,6–9 There have been several studies2–4,8,10,11 insisting that high residual stromal bed thickness (RBT) is necessary to prevent the keratectasia. Additionally, some authors8 recommend an ablation ratio per total central cornea thickness (CCT) of less than 10 per cent of pre-operative CCT. There are known risk factors for keratectasia, such as pre-operative keratoconus or forme fruste keratoconus, thin corneas, high intraocular pressure, a high degree of myopia and age,2,4,12 but none of these is an absolute predictor that an ectasia will occur after LASIK.7,13
Forward shifting of the posterior cornea after LASIK and ectasia have been correlated with RBT10–17 because the flap does not contribute to the biomechanical stability of the cornea after LASIK.10–12 Few studies have described the posterior elevation of the cornea more than six months after LASIK, which is important, as keratectasia can occur six to 20 months after LASIK.1,4,5
The purpose of this study was to describe the forward shift of the posterior corneal surface one year after LASIK and to correlate the corneal shift with RBT and the ablation percentage per total corneal thickness in eyes without significant changes in anterior best-fit sphere (BFS) and CCT after uneventful myopic LASIK.
This was a retrospective, non-randomised and open-label study that included four visits: a baseline visit (before LASIK) and visits at one, six and 12 months after uneventful myopic LASIK to study the posterior corneal surface measured with Orbscan topography.
Eighty-six eyes of 43 non-consecutive patients, who underwent uneventful LASIK for myopia, were examined retrospectively. The mean age was 32.07 ± 7.01 years (range 19 to 52 years). The mean spherical equivalent was -4.92 ± 2.24 D (range -1.25 to -10.75 D), with a mean RBT of 274.77 ± 36.41 µm (range 182 to 378 µm), and the ablation depth was 98.78 ± 26.12 µm (range 38 to 143 µm).
The spherical equivalent changed from -4.92 ± 2.24 D pre-operatively to -0.18 ± 0.56 D at one month post-operation; these values remained stable at six (–0.19 ± 0.80 D) and 12 months (–0.18 ± 0.62 D) after surgery.
The criteria for inclusion were no family history of keratoconus or evidence of external corneal disease, no chronic trauma (eye rubbing) or ocular pathology, including glaucomatous optic nerve damage. Before surgery, no eyes showed the topographic appearance of keratoconus or suspect keratoconus as evaluated with Orbscan topography (Orbscan II, Bausch & Lomb, Rochester, NY, USA, version 3.12). Additionally, patients with surgical complications such as an incomplete flap, epithelial ingrowths and diffuse lamellar keratitis were excluded.
Informed consent was obtained from subjects to use their data and the Human Sciences Ethics Committee of the University of Valladolid approved the study. All subjects were treated in accordance with the Declaration of Helsinki.
The patients were divided into three groups based on the estimated RBT (Table 1):
Table 1. Pre-operative parameter values of the patients based on the residual corneal thickness
Group 1 (<250 µm)
Group 2 (250–300 µm)
Group 3 (>300 µm)
Mean ± SD
Mean ± SD
Mean ± SD
* p > 0.05 ANOVA. BFS: best-fit sphere, EE: spherical equivalent, CCT: central corneal thickness, IOP: intraocular pressure, RBT: residual stromal bed thickness, SD: standard deviation.
1Group 1, 16 eyes (19 per cent) with RBT less than 250 µm
2Group 2, 50 eyes (58 per cent) with RBT from 250 to 300 µm
3Group 3, 20 eyes (23 per cent) with RBT greater than 300 µm.
Estimated RBT was not calculated based on an intra-operative measurement, but rather it was calculated pre-operatively using the difference between the pre-operative Orbscan CCT with the predicted flap thickness and the predicted ablation depth as assessed by the laser.
Subjects were divided into two further groups based on the ablation percentage per total corneal thickness:
1Group A, 52 eyes (60 per cent) with less than 20 per cent ablation
2Group B, 34 eyes (40 per cent) with more than 20 per cent ablation.
Ablation percentage was calculated by the relationship between the laser ablation depths with pre-operative Orbscan CCT (Table 2).
Table 2. Pre-operative parameter values of the patients based on the ablated percentage per total cornea
Group A (<20%)
Group B (>20%)
Mean ± SD
Mean ± SD
* p > 0.05 analysis of variance (ANOVA). BFS: best-fit sphere, EE: spherical equivalent, CCT: central corneal thickness, IOP: intraocular pressure, RBT: residual stromal bed thickness, SD: standard deviation.
The same experienced surgeon performed all procedures (from 2006 to 2009). Standard LASIK surgery using a narrow beam, flying-spot excimer laser with eye tracker assistance (Chiron Technolas 217-C with Plano Scan program; Bausch & Lomb Surgical, Rochester, NY, USA) was conducted. This laser has an emission wavelength of 193 nm, a fixed-pulse repetition rate of 50 Hz and a radiance exposure of 120 mJ/cm2. Flaps were performed with suction rings of 8.5 mm (22 per cent of eyes) or 9.5 mm (78 per cent of eyes) in diameter. A microkeratome (Hansatome Microkeratome; Bausch & Lomb) with a 160 microns (63 per cent of eyes) or 180 microns (37 per cent of eyes) depth was used. Photo-ablation was applied to an optical zone with a diameter ranging from 4.5 to 7.0 mm (mean 6.3 ± 0.8 mm).
Instrumentation and main outcome measures
An Orbscan topographic system was used for anterior and posterior topographic analysis and optical pachymetry with a 95 per cent acoustic factor to determine central and peripheral corneal thickness (3.5 mm from the apex in superior, inferior, nasal and temporal locations). The procedure involving the Orbscan II has been described previously.18
The forward shift of the posterior corneal surface was determined with Orbscan posterior BFS. The posterior corneal elevation was defined as the value relative to the BFS of a single map and was used to compare the pre-operative and post-operative posterior corneal surface. An experienced operator conducted the Orbscan exploration in all of the study visits.
Statistical analysis was performed using the SPSS 14.0 (SPSS Chicago, IL, USA) statistical package for Windows.
The normality of the data was tested by the Kolmogorov–Smirnov test. Multivariate analysis of variance (ANOVA with a Bonferroni correction for multiple comparisons) was used to detect differences in the forward shift of the posterior corneal surface, in the anterior BFS and in the Orbscan central and peripheral corneal thickness in each visit and in each study group. A p-value of less than 0.05 was considered statistically significant.
Simple regression analysis was performed to investigate the amount of forward shift of the posterior corneal surface after surgery in each study group. Finally, an r2 correlation coefficient was determined. A p-value of less than 0.05 was considered statistically significant.
We found statistically significant differences (p < 0.01) between visits in the forward shift of the posterior cornea after LASIK with 9.4 ± 14.3 µm (range -27 to 80 µm) at one month, 4.0 ± 9.8 µm (range -34 to 24 µm), at six months and 3.1 ± 8.8 µm (range -37 to 21 µm) at 12 months of mean forward shift relative to pre-operative posterior elevation BFS. A multiple analysis comparison showed that the differences were statistically significant only between the pre-operative visit and the one-month follow up (p < 0.01). After six months (p = 0.05) and 12 months (p = 0.29), the differences were not statistically significant. Figure 1 shows the evolution of the forward shifting of the posterior corneal surface during post-operative visits for all patients.
Residual bed thickness differences
The forward shift of the posterior cornea relative to estimated RBT groups was as follows (Figure 2):
Group 1 (estimated RBT less than 250 µm): 12.4 ± 16.1 µm (range -12 to 69 µm) at one month, 3.9 ± 11.2 µm (range -32 to 19 µm) at six months and 0.5 ± 11.4 µm (range -37 to 11 µm) at 12 months (p = 0.01). A multiple comparison showed statistically significant differences at the one month (p = 0.02) visit; however, there were no statistical differences at six (p = 1.00) and 12 months (p = 1.00).
Group 2 (estimated RBT between 250 and 300 µm): 10.3 ± 15.8 µm (range -27 to 80 µm) at 1 month, 6.0 ± 8.7 µm (range -20 to 24 µm) at 6 months and 5.3 ± 7.3 µm (range -20 to 21 µm) at 12 months (p < 0.01). A multiple comparison showed statistical differences at the one-month (p < 0.01) and six-months (p = 0.02) visits. There were no statistical differences at 12 months (p = 0.06).
Group 3 (estimated RBT greater than 300 µm): 4.4 ± 5.2 µm (range -5 to 13 µm) at one month, -1.1 ± 9.6 µm (range -34 to 12 µm) at six months and 0.6 ± 8.6 µm (range -16 to 14 µm) at 12 months without statistical differences (p = 0.05) between visits.
Using simple regression analysis (Figure 3), the change in the posterior corneal forward shift one year after LASIK showed a possible trend toward increasing posterior corneal elevation when the estimated RBT was lower than 250 µm (r2= 0.111, p = 0.26) and between 250 to 300 µm (r2= 0.185, p < 0.01). Nevertheless, in the groups with estimated RBT greater than 300 µm there was no change in the posterior corneal forward shift (r2= 0.002, p < 0.86).
Ablation percentage differences
The forward shift of the posterior cornea relative to the ablation percentage groups was as follows (Figure 4).
Group A (ablation depth less than 20 per cent): 8.3 ± 12.4 µm (range -5 to 69 µm) at one month, 2.6 ± 8.5 µm (range -34 to 24 µm) at six months and 2.0 ± 6.8 µm (range -16 to 18 µm) at 12 months (p < 0.01). A multiple comparison showed statistical differences at one month (p < 0.01). There were no statistical differences at six months (p = 0.66) or 12 months (p = 1.00).
Group B (ablation depth greater than 20 per cent): 10.9 ± 16.6 µm (range -27 to 80 µm) at one month, 6.0 ± 11.3 µm (range -32 to 24 µm) at six months and 4.5 ± 10.9 µm (range -37 to 21 µm) at 12 months (p < 0.01). A multiple comparison showed statistical differences at one month (p < 0.02). There were no statistical differences at six months (p = 0.20) and 12 months (p = 0.70).
Using simple regression analysis (Figure 5), the change in the posterior corneal forward shift one year after LASIK showed a trend toward higher posterior corneal change in patients with an ablation percentage greater than 20 per cent (r2= 0.156, p = 0.01) and when the ablation percentage was less than 20 per cent (r2= 0.09, p = 0.10).
Central and peripheral corneal thickness changes
There were no statistically significant differences among the changes in the central and peripheral corneal thicknesses (Table 3) for the post-LASIK visits (p > 0.05). The differences were statistically significant only between the pre-operative visit and the post-LASIK visits (p < 0.05) for the central, superior and inferior corneal thicknesses. The nasal and temporal thicknesses did not show statistically significant differences with pre- and post-LASIK pachymetry (p > 0.05). CCT was not different between the RBT groups (Figure 6 and Table 3) and peripheral corneal thickness showed a similar trend (Table 3).
Table 3. Change in corneal thickness measured central and peripherally with Orbscan, one year after LASIK
Differences between the baseline (pre-LASIK) and post-LASIK values are presented. * p-value analysis of variance (ANOVA) with Bonferroni correction between post-LASIK follow-up visits. LASIK: laser in situ keratomileusis, SD: standard deviation.
Central thickness (µm)
-57.1 ± 39.9
-49.9 ± 33.4
-61.6 ± 37.8
p = 1.00
Superior tthickness (µm)
-11.3 ± 44.4
-2.3 ± 38.7
-21.9 ± 42.4
p > 0.05
Inferior thickness (µm)
-10.6 ± 31.1
-17.3 ± 37.2
-22.3 ± 44.9
p = 1.00
Nasal thickness (µm)
-11.3 ± 66.0
-12.8 ± 58.5
-14.5 ± 63.6
p = 1.00
Temporal thickness (µm)
-7.6 ± 48.7
-4.4 ± 44.4
-8.2 ± 56.2
p = 1.00
Anterior BFS evolution
There were no statistically significant differences between the anterior BFS in post-LASIK visits (p = 1.00) (Table 4 and Figure 7). The differences were statistically significant only between the pre-operative visit and the post-LASIK visits (p < 0.01).
Table 4. Change in anterior and posterior BFS measured with Orbscan one year after LASIK
Differences between the baseline (pre-LASIK) and post-LASIK values are presented. * p-value analysis of variance (ANOVA) with Bonferroni correction between post-LASIK follow-up visits. ** Differences were statistically significant only at the one-month visit. BFS: best-fit sphere, LASIK: laser in situ keratomileusis, RBT: residual stromal bed thickness, SD: standard deviation.
The change in anterior BFS after LASIK for each estimated RBT group is summarised in Table 4; however, there were no statistically significant differences (p = 1.00).
LASIK is one of the most useful surgical procedures used to correct myopia; however, the potential for complications remains. One of these complications is the risk of iatrogenic keratectasia,1 which has a low incidence of less than one per cent2,4,6 but has a significant impact on refractive surgeons and patients.8 Unfortunately, the exact aetiology of iatrogenic keratectasia is unknown.2,3,5–9
The RBT is especially important after LASIK because the posterior stroma is structurally and functionally different from the anterior stroma.9 Corneal biomechanical strength is significantly greater in the anterior 40 per cent of the corneal stroma than in the posterior 60 per cent, with a high density of keratocytes.4,19,20 A further deficiency of keratocytes21 after LASIK surgery could affect the corneal biomechanical strength. Meghpara and colleagues7 reported that the development of keratectasia after LASIK might be due to the stretching of the collagen fibrils in addition to compression, secondary to the loss of structural resistance of the cornea.
Because the ectasia appears on the posterior surface of the cornea1,2,6,17 and not on the anterior topographic measurements, the forward shift of the posterior corneal surface was measured with an Orbscan system to detect the posterior evolution one year after uneventful LASIK with different RBT. Similar to previous reports, we found a posterior BFS increase lower than 80 µm with different follow-up times.5,10,12–15 Our results present a trend towards posterior elevation that increased in the first month after LASIK and reduced with time, without significant differences one year after LASIK in all RBT groups.
Although safe limits for RBT have not been precisely determined,9,13 current opinion estimates the minimal thickness of the bed for long-term stability to be 200 to 300 µm to prevent development of ectasia;9 however, the calculation of residual stromal thickness can be controversial.8 If the residual thickness is measured immediately after laser ablation, the results could be less (thinner) because the laser beam dehydrates the stroma.22 A study of untreated and re-treated corneas implicating stromal hydration23 reported that LASIK significantly increases the refractive index of the treated stromal bed.
Therefore, the real value of RBT could be different from the pre-operative expected values as defined for our study groups. This would depend on the real amount of excimer laser ablation and variable flap thickness made by the same microkeratome.3 Different studies have reported that microkeratomes do not always produce a corneal flap of the intended thickness.8,24,25 A previous report has shown a standard deviation of 19 to 23 microns in the flap thickness measurements,26 so the real value of RBT could be different to the estimated value. For these reasons, the exact thickness of the flap and the residual stromal thickness of all patients undergoing LASIK would be practically impossible to determine.8
Thus, it is necessary to provide simple tools for surgeons to use pre-operatively. We used Orbscan pachymetry (with a 0.95 acoustic factor) to determine CCT and to estimate the RBT because it is a non-invasive, reproducible system to measure CCT,27 posterior curvature of the cornea28 and central and peripheral contact lens-induced corneal swelling.27,29 The use of the 0.95 acoustic factor could be controversial because this factor had no effect on the data interpretation obtained from repetitive measurements in the same patient.30 The presumed RBT was calculated as the difference between pre-operative CCT minus the intended flap thickness and predicted central laser ablation depth. This presumed RBT value could be calculated easily before surgery and helps a surgeon's decision and patient's management to avoid the forward shift of the posterior cornea. Also, Saad and Gatinel31 have recently proposed the use of a neural network to corneal analysis that could be used to improve the detection of corneas at risk for refractive surgery.
The use of the Orbscan device for assessing post-LASIK posterior corneal elevation might be controversial because the accuracy of the Orbscan in assessing the posterior corneal surface is a subject of debate.16,32 Different hypotheses have been proposed to explain the problems with the Orbscan in the assessment of posterior curvature after LASIK.33 The effect of eyelids, eyelashes and reflections could provide an incomplete map, especially in the periphery of the cornea. If the peripheral data are missing, the elevation maps could be affected.34 Also, Orbscan pachymetry after LASIK tends to underestimate the corneal thickness compared with ultrasonic pachymetry.35,36 Ciolino and colleagues32 proposed that the forward protrusion of the posterior corneal surface could be due to an erroneous measurement of the posterior corneal elevation because the anterior cornea is easier to assess and Orbscan treats the anterior and posterior surface fits independently.16 We found significant differences in posterior forward protrusion without differences in CCT one, six and 12 months after LASIK. Thus, the accuracy of the pachymetry cannot explain the posterior bulging differences found in the present study.
Also, post-LASIK posterior corneal elevation measured using the Orbscan could be influenced by the altered anterior surface refractive power or corneal swelling induced after ablation, especially in the periphery of the cornea.16,33 Different reports34,37 have concluded that there is no evidence of surgically induced changes in the posterior surface one week34,37 and one month34 after LASIK surgery with Pentacam measurements. Pentacam showed higher accuracy in posterior corneal elevation measurement than Orbscan in eyes after photorefractive keratectomy,38 suggesting that the light scatter could affect the Orbscan image acquisition. We did not find changes in the anterior surface (anterior BFS) measured with the Orbscan for the follow-up visits included in this study (p = 1.00), according to the results from Nishimura and colleagues;34 however, a change in corneal magnification induced by anterior refractive correction33 would not explain the posterior changes detected in this study.
Grzybowski and colleagues16 proposed that the post-operative increase in posterior elevation after LASIK is dominated by backward peripheral movement due to peripheral corneal swelling. Reinstein and colleagues39 described an increase in peripheral swelling of 10 to 20 µm after LASIK, which was in line with the model by Grzybowski and colleagues. For these reasons, we measured central and peripheral corneal thickness after LASIK to corroborate whether swelling of the peripheral stroma changed after LASIK. The Orbscan showed a high repeatability to detect contact lens-induced peripheral corneal swelling of less than five per cent (less than 20 µm of corneal thickness change).27 We found less peripheral corneal thickness without significant modifications after LASIK (Table 3). These results suggest that peripheral corneal swelling cannot explain the apparent steepening of the posterior curvature,33 although biomechanical remodeling implies that mechanical aspects could be responsible for the transient posterior forward shift found in the present study. However, peripheral corneal thickness could be affected by the transitions ablation optic zone (from 4.5 to 7.0 mm), thus corneal swelling in the periphery could be difficult to measure with the change in corneal thickness.
Our results confirm that an estimated RBT greater than 300 µm showed no significant posterior forward shift (p = 0.05 and r2= 0.002, p < 0.86). Patients with estimated RBT less than 300 µm showed a significant posterior forward shift in the first month after LASIK (p < 0.05), but this difference was not significant one year after surgery (p > 0.05) and occurred without significant changes in the anterior BFS or central and peripheral corneal thicknesses after LASIK (Figure 6 and Tables 3 and 4). We found an inverse, statistically significant relationship, which suggested a transitory increase in the posterior forward shift with low estimated RBT after LASIK (Figure 3). Similar relationships between one week and three months after LASIK without differences between RBT groups have been described.8,11,15 This could indicate the necessity of longer follow-up times. Our results show the evolution of the posterior forward shift without significant changes in anterior BFS and CCT after LASIK and confirm the recovery of pre-operative posterior curvature values one year after LASIK, in accordance with previous reports,32,40,41 and it is similar to the evolution after photorefractive keratectomy laser procedures.42
The main difference between the present study and previous reports is a different trend in changes to the posterior curvature between an estimated RBT that returned to original levels after surgery in eyes without changes in anterior BFS and corneal thickness after LASIK.
This significant change in the early onset of posterior forward shift could be related to corneal swelling induced after ablation16,23 or Orbscan repeatability.43 Thus, early onset of posterior changes should be interpreted with caution and studies with different techniques with high accuracy and repeatability for posterior corneal analysis and longer follow up could be necessary. Early anterior bulging of the posterior corneal surface after LASIK does not imply evidence of corneal ectasia.
Until the risk factors that affect the development of ectasia after LASIK can be clearly determined, the best recommendation is to conduct an exhaustive assessment of the corneal topography to avoid LASIK in eyes with abnormal corneal topography44 and not leave the cornea with an excessively thin residual thickness.2,9,45
The results of the present study could help surgeons to perform LASIK with an estimated RBT greater than 300 µm using Orbscan CCT measurements, so as not to induce significant changes in the posterior curvature of the cornea, even in the early period after LASIK.
In conclusion, the present study describes the trend in the forward shift of the posterior corneal surface one year after uneventful myopic LASIK relative to an estimated RBT and the ablation percentage estimated with an Orbscan scanning slit system without significant changes in the anterior BFS and central and peripheral corneal thickness after LASIK. The corneal bulge measured with the Orbscan device after LASIK was time dependent, with a different trend between the estimated RBT that protruded at an early stage but then returned to original levels 12 months after LASIK.