sign in sign up
<<
Home , Keratoconus, Refractive error

Downloaded from http://bjo.bmj.com/ on August 21, 2017 - Published by group.bmj.com Clinical science Refractive improvements and safety with topography-guided corneal crosslinking for : 1-year results Maria Nordström, Maria Schiller, Anneli Fredriksson, Anders Behndig

– Department of Clinical Sciences, ABSTRACT the stromal matrix.1 6 CXL has also been shown to , Umeå Purpose To assess the refractive improvements and the cause a gradual post-treatment flattening of the University, Umeå, Sweden corneal endothelial safety of an individualised corneal curvature with regression of the refractive topography-guided regimen for corneal crosslinking in errors over time.2 With the standard ‘Dresden’ Correspondence to Professor Anders Behndig, progressive keratoconus. CXL protocol, the corneal surface is uniformly Department of Clinical Sciences, Methods An open-label prospective randomised irradiated with ultraviolet A (UVA) in a central 9 – Ophthalmology, Umeå clinical trial was performed at the Department of Clinical mm zone at 5.4 J/cm2.179 With such a standard University, Umeå SE-901 87, Sciences, Ophthalmology, Umeå University Hospital, protocol, there are no means to control or regulate Sweden; ​anders.​behndig@​umu.​se Umeå, Sweden. Thirty-seven (50 ) with the post-treatment refractive improvement, and the progressive keratoconus planned for corneal crosslinking degree of improvement often seems clinically Received 20 June 2016 were included. The patients were randomised to unpredictable. Computer simulations of asymmetric Revised 15 October 2016 topography-guided crosslinking (photorefractive UVA irradiation centralised on the ectatic keratoco- Accepted 9 November 2016 intrastromal crosslinking (PiXL); n=25) or uniform 9 mm nus cone suggest that such a treatment pattern may Published Online First fl 8 29 November 2016 crosslinking (corneal crosslinking (CXL); n=25). induce a more pronounced attening. Recent , , keratometry (K1, K2 and Kmax) investigations indicate that higher levels of CXL and corneal endothelial morphometry were assessed energy can be safely delivered,10 which opens up preoperatively and at 1, 3, 6 and 12 months for individualised treatment plans based on the postoperatively. The PiXL treatment involved an with local augmentation of the asymmetrical treatment zone centred on the area of treatment effect in the most ectatic zone. This type maximum corneal steepness with treatment energies of individualised crosslinking has, however, not yet ranging from 7.2 to 15.0 J/cm2; the CXL treatment was been evaluated in a randomised clinical trial. The a uniform 9 mm 5.4 J/cm2 pulsed crosslinking. The main main outcome measures of this study were to outcome measures were changes in refractive errors and evaluate the refractive outcomes of an individua- corneal endothelial cell density. lised topography-guided regimen for corneal cross- Results The spherical refractive errors decreased linking (photorefractive intrastromal crosslinking (p<0.05) and the visual acuity improved (p<0.01) at 3, (PiXL)) in progressive keratoconus compared with 6 and 12 months after PiXL, but not after CXL. The the corresponding outcomes with a uniform, between-groups differences, however, were not accelerated 5.4 J/cm2 pulsed CXL treatment 11 12 significant. K2 and Kmax decreased at 3, 6 and regimen. In addition, we wanted to evaluate 12 months after PiXL (p<0.01), but not after CXL any changes in corneal endothelial morphometry (p<0.01 when comparing the two treatments). No induced by the two treatments. corneal endothelial cell loss was seen after either treatment. MATERIALS AND METHODS Conclusions Individualised topography-based In this open-label, randomised clinical trial, crosslinking treatment centred on the ectatic cone has patients with keratoconus planned for corneal the potential to improve the corneal shape in crosslinking were recruited from the Department keratoconus with decreased spherical refractive errors of Ophthalmology, Umeå University Hospital, and improved visual acuity, without damage to the Sweden,between20March2014and8April . 2015; National Clinical Trials ID: 02514200; Trial registration number NCT02514200, Results. http://ClinicalTrials.gov) by the principal investi- gator (AB). The study comprised 50 eyes of 37 patients with INTRODUCTION progressive keratoconus (33 males, 4 females), aged Keratoconus is a corneal degeneration with pro- 27.7±8.5 years (range: 16–50 years). The power gressive corneal thinning, ectatic corneal protrusion analysis used to determine the group sizes showed and asymmetric corneal irregularity as main fea- that the study design allows for detection of a dif- tures. The refractive consequences of keratoconus ference in spherical equivalent of 2.0 D between are progressive and irregular . treatments and 1.2 D between time points Corneal collagen crosslinking (CXL) with the (α=0.05; power=0.80). After inclusion, patients photosensitizer riboflavin and ultraviolet was were randomised to receive either uniform pulsed fi et al1 2 To cite: Nordström M, rst described by Spoerl and has become a crosslinking with treatment energy of 5.4 J/cm 11 12 Schiller M, Fredriksson A, widely used treatment to halt the disease progres- (CXL; n=25) or individualised topography- et al. Br J Ophthalmol sion by increasing the biomechanical strength of guided corneal crosslinking with asymmetrical 2017;101:920–925. through induction of new covalent bonds in treatment zones and variable treatment energies

920 Nordström M, et al. Br J Ophthalmol 2017;101:920–925. doi:10.1136/bjophthalmol-2016-309210 Nordström M, et al . BrJ Ophthalmol 2017;101:920–925. doi:10.1136/bjophthalmol-2016-309210 Table 1 Changes in refractive errors and keratometry readings after photorefractive intrastromal crosslinking (PiXL) and corneal collagen crosslinking (CXL)

Baseline Difference, 1 month Difference, 3 months Difference, 6 months Difference, 12 months

Mean±SD 95% CI Mean±SD 95% CI p Value Mean±SD 95% CI p Value Mean±SD 95% CI p Value Mean±SD 95% CI p Value

UCVA, logMAR PiXL 0.71±0.66 0.45 to 0.96 −0.05±0.35 −0.19 to +0.10 0.61 −0.17±0.18 −0.24 to −0.09 0.00† −0.35±0.33 −0.48 to −0.22 0.00† −0.31±0.40 −0.49 to −0.13 0.00† CXL 0.69±0.78 0.37 to 1.01 +0.18±0.39 +0.02 to +0.35 0.04* +0.03±0.32 −0.10 to +0.16 0.69 −0.11±0.22 −0.11 to −0.01 0.04* −0.07±0.16 −0.15 to +0.02 0.14 p Value 0.94 0.05* 0.03* 0.01* 0.02* BSCVA, logMAR Downloaded from PiXL 0.28±0.28 0.17 to 0.39 +0.05±0.18 −0.13 to +0.02 0.37 −0.12±0.17 −0.19 to −0.06 0.00† −0.17±0.18 −0.25 to −0.10 0.00† −0.16±0.24 −0.26 to −0.07 0.00† CXL 0.30±0.29 0.19 to 0.42 +0.12±0.24 +0.01 to +0.22 0.03* +0.06±0.25 −0.17 to +0.04 0.23 +0.04±0.23 −0.14 to +0.05 0.39 +0.01±0.29 −0.11 to +0.13 0.86 p Value 0.77 0.02* 0.39 0.05* 0.03* Sphere, D PiXL −0.46±2.50 −1.44 to 0.52 +0.15±1.32 −0.39 to +0.69 0.68 +0.33±0.53 +0.12 to +0.54 0.00* +0.61±0.81 +0.25 to +0.98 0.00† +1.00±1.92 +0.25 to +1.75 0.01*

CXL 0.33±2.43 −0.67 to 1.32 −0.91±4.16 −2.65 to +0.83 0.31 −0.82±3.12 −2.12 to +0.48 0.23 +0.14±2.00 −0.70 to +0.97 0.75 +0.98±2.49 −0.06 to 2.02 0.08 http://bjo.bmj.com/ p Value 0.28 0.27 0.11 0.35 0.97 Cylinder, D PiXL −2.64±1.73 −3.32 to −1.96 −0.25±1.41 −0.83 to +0.33 0.51 +0.23±0.82 −0.10 to +0.55 0.44 +0.14±1.11 −0.31 to +0.58 0.32 −0.32±1.85 −1.05 to +0.41 0.40 CXL −2.54±2.07 −3.39 to −1.70 −0.36±1.73 −1.09 to 0.36 0.33 −0.76±2.15 −1.66 to +0.14 0.10 −0.11±2.55 −1.18 to +0.95 0.83 −0.79±2.54 −1.85 to +0.28 0.16 p Value 0.86 0.72 0.07 0.55 0.48

Spherical equivalent, D onAugust21,2017-Publishedby PiXL −1.95±2.58 −2.96 to −0.94 +0.03±1.61 −0.63 to +0.68 0.96 +0.45±0.73 +0.16 to +0.73 0.03* +0.68±0.85 +0.34 to +1.02 0.00† +0.84±1.84 +0.12 to +1.56 0.03* CXL −0.95±2.28 −1.88 to −0.01 −1.09±3.89 −2.72 to +0.54 0.19 −1.20±3.04 −2.47 to +0.07 0.07 +0.08±1.12 −0.39 to +0.55 0.74 +0.58±2.04 −0.27 to +1.46 0.19 p Value 0.24 0.21 0.03* 0.04* 0.65 K1, D PiXL 46.23±6.14 43.83 to 48.64 +0.04±1.40 −0.51 to +0.59 0.88 −0.17±1.46 −0.75 to +0.40 0.62 +0.09±1.74 −0.61 to +0.78 0.83 −0.01±1.26 −0.50 to +0.49 0.97 CXL 45.80±5.09 43.72 to 47.88 +0.11±0.65 −0.15 to +0.39 0.40 +0.15±0.72 −0.15 to +0.45 0.34 −0.10±0.78 −0.43 to +0.22 0.53 −0.04±0.46 −0.23 to +0.15 0.66 p Value 0.77 0.81 0.38 0.64 0.90 K2, D PiXL 49.60±6.40 47.09 to 52.11 0.00±1.14 −0.44 to +0.45 0.75 −0.91±0.96 −1.29 to −0.53 0.00† −0.87±0.82 −1.20 to −0.54 0.00† −0.42±0.80 −0.74 to −0.11 0.01* CXL 49.40±6.50 46.74 to 52.05 +0.23±4.12 −1.95 to +1.49 0.79 +0.45±0.80 +0.45 to +0.80 0.01* +0.24±0.81 −0.10 to +0.58 0.17 +0.19±0.88 −0.18 to +0.56 0.33 p Value 0.77 0.74 0.00† 0.00† 0.02*

Kmax ,D group.bmj.com PiXL 57.41±10.24 53.39 to 61.42 +0.14±1.78 −0.55 to +0.84 0.51 −0.94±1.28 −1.45 to −0.44 0.00† −1.16±1.37 −1.71 to −0.61 0.00† −1.31±1.52 −1.92 to −0.72 0.00† CXL 57.45±10.69 53.08 to 61.82 +1.29±1.72 +0.57 to +2.01 0.00† +1.41±5.17 −0.75 to +3.57 0.20 +0.39±1.47 −0.22 to +1.01 0.22 +0.30±1.33 −0.26 to +0.86 0.30 p Value 0.90 0.06 0.05* 0.00† 0.00†

True net power, Pmax ,D PiXL 52.34±8.81 48.88 to 55.80 +0.66±1.90 −0.09 to +1.40 0.11 −0.64±0.90 −0.99 to −0.29 0.00† −0.80±1.14 −1.25 to −0.35 0.00† −0.88±1.02 −1.28 to −0.48 0.00† CXL 50.88±8.01 47.74 to 54.02 +0.47±1.03 +0.07 to +0.87 0.04* −0.15±0.65 −0.41 to +0.10 0.27 −0.17±0.59 −0.40 to +0.06 0.17 +0.03±1.16 −0.42 to +0.49 0.89

p Value 0.55 0.60 0.04* 0.03* 0.01* Clinical science

True net power, Pmin ,D PiXL 41.54±5.50 39.38 to 43.69 +0.50±1.59 −0.12 to +1.12 0.15 +0.79±1.45 +0.22 to +1.36 0.01* +0.94±1.90 +0.20 to +1.69 0.02* +1.00±1.22 +0.52 to +1.48 0.00† CXL 41.00±3.64 39.58 to 42.43 +0.05±1.08 +0.38 to +0.47 0.85 −0.10±0.70 −0.38 to +0.17 0.48 −0.03±0.57 −0.26 to +0.19 0.77 −0.07±0.64 −0.32 to +0.19 0.63 p Value 0.69 0.27 0.01* 0.02* 0.00† Bold text indicates significant p values. *p<0.05; †p<0.01.

921 BSCVA, best-corrected visual acuity; logMAR, logarithm of the minimum angle of resolution; UCVA, uncorrected visual acuity. Downloaded from http://bjo.bmj.com/ on August 21, 2017 - Published by group.bmj.com Clinical science

where the Pentacam HR axial curvature dropped off by steps of 2.0 D from the Kmax value, and the treatment energy distribu- tion was based on the maximum keratometry (Kmax) value: 43–47 D, 7.2 J/cm2;48–52 D, 10 J/cm2; and ≥52 D, 15 J/cm2 with treatment times of 8:00, 11:07 and 16:40 min, respect- ively. The machine software used in the present study involved an tracker to ascertain a proper centration of the treatment zones in relation to the . Cyclotorsion was adjusted for manually before starting the treatment. A computer-generated list of the numbers 1–50 in random order was used for randomisation. Patients were included in running numbers according to the list; an even number was treated with CXL and an odd number with PiXL.13 If both eyes of a were included (n=13), the second eye always received the other treatment. Inclusion required a diagnosis of progressive keratoconus with a minimum corneal thickness of 400 mm at the thinnest point after epithelial removal. The kera- toconus diagnosis was based on the Amsler-Krumeich grading14 ‘ ’ fi Figure 1 The change in Kmax at 12 months (Y axis) plotted against and the Total Deviation keratoconus quanti cation value from ‘ ’ the baseline Kmax value (X axis). In photorefractive intrastromal the Belin-Ambrosio enhanced ectasia measurements of the crosslinking-treated eyes (dark dots), there is a significant negative Pentacam HR, plus an altered red reflex and/or distortion of the correlation (R=−0.66; p=0.00), but not in corneal collagen keratometric mires.13 15 Disease progression was verified by − crosslinking-treated eyes (white dots; R= 0.16; p=0.38). increasing corneal steepness (increase in Kmax of ≥1 D in 1 year) and/or thinning of the cornea on repeated Scheimpflug tomog- raphy measurements in 41 eyes; in 9 eyes, the progression was (PiXL, n=25). All treatments were performed with the Avedro verified by increasing keratometric astigmatism and corneal KXL II system (Avedro, Waltham, Massachusetts, USA) and steepness and decreasing best-spectacle-corrected visual acuity. involved removal of the central 9 mm corneal after Exclusion criteria were age under 12, a history of corneal topical anaesthesia with tetracaine, topical application of 0.1% disease such as herpes simplex, significant corneal scarring, dextran-free riboflavin every two minutes during 10 min,11 and pregnancy or lactation, or other conditions preclud- 1 s on/1 s off pulsed 370 nm UVA irradiation of 30 mW/cm2.12 ing a steady gaze, conditions that could interfere with epithelial Postoperatively, all patients were given topical levofloxacin healing, vitamin C supplements within 1 week of the treatment 5 mg/mL three times daily for a week. The treatment time for and a history of previous corneal or cognitive insuffi- the CXL group was 8:00 min. The treatment plan design in the ciency interfering with the informed consent. PiXL group was based on in silico work by Roy and Dupps,8 Data were collected at baseline and at 1, 3, 6 and 12 months. and the sizes and shapes of the treatment zones were calculated The examinations included , determination from Pentacam HR (Oculus, Lynnwood, Washington, USA) of uncorrected visual acuity (UCVA) and best-corrected visual Scheimpflug tomography measurements. The zones were acuity (BSCVA) using the logarithm of the minimum angle of arcuate-shaped, sparing a central 2 mm zone. The size and resolution (logMAR) fast protocol,16 assessment of the intraocu- shape of the treatment zone were determined by the transition lar pressure (IOP) with Goldmann applanation tonometry and

Figure 2 Differential topography of the right and left eyes of the same individual at 12 months post treatment. The right eye was treated with photorefractive intrastromal crosslinking (PiXL), the left with corneal collagen crosslinking (CXL). Note the decreased inferior steepness (negative values) and increased superior steepness (positive values) in the PiXL-treated eye, and a more symmetrical decrease in steepness in the CXL-treated eye. OD, oculus dexter; OS, oculus sinister.

922 Nordström M, et al. Br J Ophthalmol 2017;101:920–925. doi:10.1136/bjophthalmol-2016-309210 Downloaded from http://bjo.bmj.com/ on August 21, 2017 - Published by group.bmj.com Clinical science

Figure 3 Individual photorefractive intrastromal crosslinking treatment plan (left); the larger area was treated with 5.4 J/cm2 and the smaller, inner area with 10 J/cm2. To the right, a map of maximum densitometry 3 months after treatment is shown. Note the increased densitometry in the treated zone. *Pmax; †Pmin. OD, oculus dexter.

keratometry readings (K1, K2 and Kmax). In the PiXL group, one patient in the CXL group missed their visits. No adverse the ‘True Net Power’ feature of the Pentacam HR was used to events occurred. determine the total corneal power in a 1 mm zone centred at Table 1 shows the changes in refractive errors after the two the point of maximum irradiation (Pmax) and at a 1 mm zone treatments. The groups were similar regarding all the refractive 180° from Pmax, at the same distance from the visual axis (Pmin). variables at baseline. UCVA and BSCVA improved, the myopic For comparison, corresponding points were also measured in decreased, the spherical equivalent increased and the CXL-treated . Similarly, the maximum corneal densi- a corneal flattening was seen with decreases in K2 and Kmax at 3– tometry values (corneal light backscatter), expressed as standar- 12 months after PiXL. After CXL, stabilisation but no visual or 15 dised grey scale units (GSUs)) were also assessed at Pmax and refractive improvements were seen. The decreases in UCVA and Pmin in all treated corneas using the Pentacam HR. Central BSCVA seen after CXL at 1 month were not seen after PiXL. corneal endothelial photographs were taken with the Topcon The decrease in Kmax at 12 months was significantly corre- SP-2000P specular microscope (Topcon Europe, Capelle a/d lated to the baseline Kmax values in PiXL (R=−0.66; p=0.00; Ijssel, the Netherlands), and the corneal endothelial morphology figure 1), but not in CXL (R=−0.16; p=0.38; figure 1). In the 2 was calculated from a cluster of 55 cells from each photograph, PiXL eyes treated with 15 J/cm , the reduction in Kmax at as previously detailed.17 The choice of 55 cells per endothelium 12 months was larger than in the PiXL eyes treated with 7.2– was based on a power analysis and was shown to detect a 5% 10 J/cm2 (−1.74±1.66 D vs −0.40±0.53 D; p=0.01). difference in cell count between examinations (α=0.05; After PiXL, an inferior flattening in the Pmax zone and a power=0.80). The endothelial cell count, the hexagon shape superior steepening in the Pmin were noted from 3 months and factor (quantifying the deviation from the ideal hexagonal cell on (table 1). No similar changes were seen after CXL. After 17 shape ) and the degree of cell elongation were calculated. PiXL, the change in true net power in Pmin was inversely corre- Student’s paired or unpaired t-tests were used for statistical lated to the change in true net power in Pmax at 12 months (R= comparisons as appropriate. Correlations were assessed with −0.87; p=0.00), whereas after CXL, a direct correlation Pearson’s correlations. A p value of <0.05 was considered statis- between the changes in Pmin and Pmax was seen (R=0.54; tically significant. p=0.00). Figure 2 exemplifies this difference in refractive outcome between PiXL and CXL. In PiXL-treated eyes, the maximum corneal densitometry RESULTS increase in the Pmax zone was significantly larger than after CXL Two patients randomised to CXL did not accept treatment and from 3 to 6 months, but not in the Pmin zone, meaning that the were excluded. Thus, 25 eyes from 25 patients in the PiXL PiXL treatment rendered an asymmetrical pattern of densitom- group and 23 eyes from 23 patients in the CXL group under- etry increase (figure 3 and table 2). went treatment. Thirteen patients were bilaterally treated, with The measured IOP increased after both treatments (table 2). PiXL in one eye and CXL in the other (see above). At 1 month, The endothelial morphometry showed no detectable endothelial two patients in the PiXL group and one in the CXL group cell loss after either treatment, and the other morphometric missed their return visit; at 3 months one patient in the CXL variables assessed were also unaltered throughout the follow-up group, at 6 months one patient in each group and at 12 months (table 2).

Nordström M, et al. Br J Ophthalmol 2017;101:920–925. doi:10.1136/bjophthalmol-2016-309210 923 924 Clinical science

Table 2 Changes in intraocular pressure values, corneal densitometry and corneal endothelial morphometry after photorefractive intrastromal crosslinking (PiXL) and corneal collagen crosslinking

(CXL) Downloaded from Baseline Difference, 1 month Difference, 3 months Difference, 6 months Difference, 12 months

Mean±SD 95% CI Mean±SD 95% CI p Value Mean±SD 95% CI p Value Mean±SD 95% CI p Value Mean±SD 95% CI p Value

Intraocular pressure, mm Hg PiXL 10.85±1.86 10.12 to 11.58 +1.00±2.50 +0.02 to +2.02 0.03* +2.00±2.66 +0.96 to +3.04 0.00† +3.02±2.95 +0.96 to +3.04 0.00† +3.28±2.21 +2.29 to +4.28 0.00† http://bjo.bmj.com/ CXL 11.96±2.16 11.07 to 12.84 +0.41±2.45 −0.61 to +1.43 0.45 +0.91±2.11 −0.03 to +1.79 0.06 +1.55±2.62 +0.45 to +2.64 0.01* +1.48±2.82 0.00 to +2.95 0.02* p Value 0.07 0.31 0.11 0.07 0.02*

Maximum densitometry, Pmax , GSU PiXL 27.85±4.05 26.27 to 29.44 +9.18±5.46 +7.04 to +11.32 0.00† +5.09±6.11 +2.70 to +7.49 0.00† +3.25±2.86 +2.13 to +4.37 0.00† +0.54±4.83 −1.36 to +2.43 0.58 CXL 27.29±2.78 26.16 to 28.43 +3.74±3.91 +2.14 to +5.33 0.00† +1.62±4.00 −0.01 to +3.26 0.07 −0.72±6.09 −3.21 to +1.77 0.58 +0.45±2.30 −0.49 to +1.39 0.64 Nordström M, et al . BrJ Ophthalmol 2017;101:920–925. doi:10.1136/bjophthalmol-2016-309210 p Value 0.58 0.00† 0.02* 0.01* 0.94 onAugust21,2017-Publishedby Maximum densitometry, Pmin , GSU PiXL 25.73±3.75 24.26 to 27.40 +0.60±4.51 −1.17 to +2.36 0.53 +0.97±3.76 −0.50 to +2.44 0.20 +0.22±2.00 −0.56 to +1.00 0.59 +0.68±3.12 −0.55 to +1.90 0.29 CXL 26.81±3.01 25.58 to 28.04 +2.57±4.24 +0.84 to +4.31 0.01† +1.21±3.19 +0.10 to +2.51 0.09 +0.02±1.81 −0.72 to +0.76 0.95 +0.52±3.56 −0.93 to +1.98 0.47 p Value 0.27 0.14 0.81 0.73 0.88 Endothelial cell count, cells/mm2 PiXL 2362±179 2562 to 2702 +4±121 −44 to +51 0.88 +20±138 −34 to +74 0.49 +17±103 −24 to +57 0.43 −5±92 −41 to +31 0.80 CXL 2768±219 2652 to 2824 +24±197 −53 to +101 0.58 +43±131 −9 to +94 0.13 −12±101 −52 to +28 0.59 +6±66 −19 to +32 0.67 p Value 0.10 0.84 0.46 0.42 0.45 Hexagon shape factor PiXL 0.93±0.33 0.80 to 1.06 +0.03±0.30 −0.09 to +0.15 0.64 −0.01±0.35 −0.15 to +0.12 0.84 0.00±0.25 −0.10 to +0.10 0.96 +0.04±0.53 −0.17 to +0.25 0.73 CXL 0.89±0.42 0.73 to 1.06 +0.24±0.93 −0.12 to +0.61 0.57 −0.01±0.36 −0.15 to +0.13 0.29 +0.07±0.31 −0.05 to +0.19 0.62 +0.11±0.50 −0.08 to +0.31 0.32

p Value 0.65 0.26 0.90 0.21 0.96 group.bmj.com Degree of cell elongation PiXL 0.095±0.009 0.091 to 0.099 +0.003±0.013 −0.002 to +0.008 0.26 +0.004±0.009 0.000 to +0.007 0.05 +0.003±0.011 −0.001 to +0.007 0.19 +0.002±0.012 −0.002 to +0.007 0.33 CXL 0.095±0.009 0.092 to 0.099 +0.003±0.011 −0.001 to +0.007 0.19 0.000±0.009 −0.004 to +0.003 0.91 +0.002±0.005 0.000 to +0.004 0.16 +0.003±0.007 −0.001 to +0.006 0.08 p Value 0.82 0.67 0.19 0.73 0.79 Bold text indicates significant p values. *p<0.05; †p<0.01. GSU, grey scale units. Downloaded from http://bjo.bmj.com/ on August 21, 2017 - Published by group.bmj.com Clinical science

DISCUSSION Contributors MN and MS share first authorship. All authors have made We here show that in progressive keratoconus an individualised, substantial contributions to the conception and design of the work, as well as the acquisition, analysis and interpretation of data for the work. All authors have asymmetric topography-guided corneal crosslinking treatment contributed to drafting the work and revising it critically for important intellectual regimen gives better refractive and visual outcomes at content. Finally, all authors have provided a final approval of the version to be 12 months than a standardised crosslinking regimen. The asym- published and agree to be accountable for all aspects of the work in ensuring that metric treatment effect in PiXL is verified by an inferior flatten- questions related to the accuracy or integrity of any part of the work are ing and a superior steepening of the cornea. On the contrary, a appropriately investigated and resolved. symmetrical treatment pattern (CXL in the present study) also Funding This study was funded by grants from the KMA Fund and Ögonfonden. gives a symmetrical treatment effect. In accordance with our Disclaimer Neither KMA Fund nor Ögonfonden had any role in the design or previous reports,15 18 this study shows that the increase in conduct of the study; the collection, management, analysis or interpretation of the corneal densitometry relates to the CXL treatment effect, which data; the preparation, review or approval of the manuscript; or the decision to submit the manuscript for publication. is in alignment with recent in vivo confocal microscopy findings, where the phenomenon of light scattering is explained as an Competing interests None declared. ‘indirect sign of CXL-induced stromal collagen compaction and Patient consent Obtained. remodeling’.19 Accordingly, with an asymmetrical treatment Ethics approval The study was approved by the Regional Ethical Board in Umeå, pattern as in the present study, an increase in corneal densitom- Sweden, and was performed in accordance with the Declaration of Helsinki. etry is seen only where the cornea is crosslinked, but not Provenance and peer review Not commissioned; externally peer reviewed. outside the treatment zone. Future evaluation of the PiXL regimen could involve, for example, monitoring of the demarca- REFERENCES tion line and epithelial mapping with optical coherence tomog- 1 Spoerl E, Huhle M, Seiler T. Induction of cross-links in corneal tissue. Exp Eye Res raphy, measurements that were not included in the present 1998;66:97–103. study. 2 Meek KM, Hayes S. Corneal cross-linking—a review. Ophthalmic Physiol Opt 2013;33:78–93. In this study, patients with a wide range of keratoconus (KC) 3 Doors M, Tahzib NG, Eggink FA, et al. Use of anterior segment optical coherence severity were included. Differences between treatments may be tomography to study corneal changes after collagen cross-linking. Am J Ophthalmol more readily demonstrated if a study is limited to a certain stage 2009;148:844–51. e842. of the disease, but, on the other hand, the present study design 4 Goldich Y, Marcovich AL, Barkana Y, et al. Clinical and corneal biomechanical fl allowed us to demonstrate the advantages of an individualised changes after collagen cross-linking with ribo avin and UV irradiation in patients with progressive keratoconus: results after 2 years of follow-up. Cornea treatment protocol. 2012;31:609–14. The first weeks after a corneal crosslinking a transient deteri- 5 Raiskup-Wolf F, Hoyer A, Spoerl E, et al. Collagen crosslinking with riboflavin and oration in vision is often seen due to a transient corneal haze,19 ultraviolet-A light in keratoconus: long-term results. J Refract Surg in the present study quantified as increased corneal densitom- 2008;34:796–801. 6 Vinciguerra P, Albè E, Trazza S, et al. Intraoperative and postoperative effects of etry. In this study, the UCVA and the BSCVA worsened at corneal collagen cross-linking on progressive keratoconus. Arch Ophthalmol 1 month after CXL, but not after PiXL. One possible explan- 2009;127:1258–65. ation could be that the central part of the cornea is spared from 7 Ghanem RC, Santhiago MR, Berti T, et al. Topographic, corneal wavefront, and irradiation in PiXL, which may lessen the post-treatment haze in refractive outcomes 2 years after collagen crosslinking for progressive keratoconus. – the central optical zone. Cornea 2014;33:43 8. 8 Roy AS, Dupps WJ Jr. Patient-specific computational modeling of keratoconus In eyes treated with PiXL, comparatively high energy levels progression and differential responses to collagen cross-linking. Invest Ophthalmol are delivered in the thinnest corneal area. In this perspective, Vis Sci 2011;52:9174–87. the treatment effect should be restricted to the anterior stromal 9 Friedman MD, Pertaub R, Usher D, et al. Advanced corneal cross-linking system volume. The short soaking time of 10 min and limited UVA with fluorescence dosimetry. J Ophthalmol 2012;2012:303459. 10 Kanellopoulos AJ, Binder PS. Collagen cross-linking (CCL) with sequential irradiation times ranging from 8 to 16 min may have contribu- topography-guided PRK: a temporizing alternative for keratoconus to penetrating ted to the safety of the treatment protocol. keratoplasty. Cornea 2007;26:891–5. The minor improvements in Kmax and refractive errors after 11 Tomita M, Mita M, Huseynova T. Accelerated versus conventional corneal collagen PiXL may not be noticed by the individual patient but our crosslinking. J Cataract Refract Surg 2014;40:1013–20. impression is that PiXL has a potential of reshaping the KC 12 Mazzotta C, Traversi C, Paradiso AL, et al. Pulsed Light Accelerated Crosslinking versus Continuous Light Accelerated Crosslinking: One-Year Results. J Ophthalmol cornea to a higher degree than standard uniform CXL. In their 2014;2014:604731. 8 finite element model, Roy and Dupps previously found in 13 Beckman Rehnman J, Behndig A, Hallberg P, et al. Increased corneal hysteresis after silico that a customised cone-localised, variable-intensity treat- corneal collagen crosslinking: a study based on applanation resonance technology. ment may render a more pronounced reshaping of the cornea JAMA Ophthalmol 2014;132:1426–32. fl 14 Krumeich JH, Daniel J, Knülle A. Live- for keratoconus. J Cataract with a greater attening effect and reduced aberrations than a Refract Surg 1998;24:456–63. uniform 9 mm treatment. In this respect, our in vivo findings— 15 Rehnman JB, Lindén C, Hallberg P, et al. Treatment effect and corneal light at least partially—confirm the in silico results. scattering with 2 corneal cross-linking protocols: a randomized clinical trial. JAMA A more pronounced flattening in PiXL corneas treated with Ophthalmol 2015;133:1254–60. 15 J/cm2 suggests that the corneal crosslinking effect increases 16 Camparini M, Cassinari P, Ferrigno L, et al. ETDRS-fast: implementing psychophysical adaptive methods to standardized visual acuity measurement with with higher irradiation power. Our impression is that the ETDRS charts. Invest Ophthalmol Vis Sci 2001;42:1226–31. improved refractive effects of PiXL with the present settings, 17 Lundberg B, Behndig A. Intracameral mydriatics in phacoemulsification cataract although significant, are still insufficient in many cases. surgery. J Cataract Refract Surg 2003;29:2366–71. However, the fact that 15 J/cm2 appears to be safe for the 18 Beckman Rehnman J, Janbaz CC, Behndig A, et al. Spatial distribution of corneal light scattering after corneal collagen crosslinking. J Cataract Refract Surg corneal endothelium may open up for further increase of the 2011;37:1939–44. treatment effect. Such an approach could potentially augment 19 Mazzotta C, Hafezi F, Kymionis G, et al. In vivo confocal microscopy after corneal the positive refractive effects from PiXL further. collagen crosslinking. Ocul Surf 2015;13:298–314.

Nordström M, et al. Br J Ophthalmol 2017;101:920–925. doi:10.1136/bjophthalmol-2016-309210 925 Downloaded from http://bjo.bmj.com/ on August 21, 2017 - Published by group.bmj.com

Refractive improvements and safety with topography-guided corneal crosslinking for keratoconus: 1-year results Maria Nordström, Maria Schiller, Anneli Fredriksson and Anders Behndig

Br J Ophthalmol 2017 101: 920-925 originally published online November 29, 2016 doi: 10.1136/bjophthalmol-2016-309210

Updated information and services can be found at: http://bjo.bmj.com/content/101/7/920

These include: References This article cites 19 articles, 2 of which you can access for free at: http://bjo.bmj.com/content/101/7/920#BIBL Email alerting Receive free email alerts when new articles cite this article. Sign up in the service box at the top right corner of the online article.

Topic Articles on similar topics can be found in the following collections Collections (713) Optics and refraction (508) Cornea (524) Ocular surface (618)

Notes

To request permissions go to: http://group.bmj.com/group/rights-licensing/permissions

To order reprints go to: http://journals.bmj.com/cgi/reprintform

To subscribe to BMJ go to: http://group.bmj.com/subscribe/