Central toxic keratopathy (CTK) is an uncommon complication of refractive surgery characterized by a central corneal opacification with associated stromal loss, striae, and hyperopic shift postoperatively. This phenomenon was first reported in 1998 by Fraenkel et al1 and the term central toxic keratopathy was coined by Sonmez and Maloney2 in 2007. CTK has most often been associated with laser in situ keratomileusis (LASIK) and photorefractive keratectomy (PRK), but has also been reported after small incision lenticule extraction (SMILE),3 contact lens use,4,5 and mechanical debridement.6 CTK can occur sporadically or in outbreaks.7,8 Management options include surgical and non-surgical modalities, but the most appropriate intervention is debated. In this retrospective case series, we report 12 eyes from 8 patients who presented with CTK after FS-LASIK surgery at a single site over 5 years, including an outbreak during this time frame.
Patients and Methods
This 5-year retrospective study reviewed the records of consecutive patients undergoing LASIK by a single surgeon (MM) for myopia, hyperopia, and astigmatism between January 2015 and December 2019 at a single site. The AMO IntraLase iFS 150-Hz excimer laser system (Abbott Medical Optics, Inc) was used for flap creation and the WaveLight EX500 femtosecond laser system (Alcon Laboratories, Inc) was used for stromal ablation in all eyes. In bilateral surgeries, the right eye was always treated first with both femtosecond and excimer lasers. Flap creation was performed at 150 kHz in a raster pattern with a bed energy of 0.75 µJ and a side cut energy of 0.6 µJ. Spot and line separations were 6 µm and 6 µm for the bed cut and 4 µm and 4 µm for the side cut. Hinge position of 90 degrees, hinge angle of 50 degrees, and hinge width of 3.8 mm were used. Targeted flap size of 8.8 mm and thickness of 100 µm was programmed with a 130-degree side cut angle. Deidentified data were analyzed. This study and the consent procedure were approved by the Hoopes Vision Ethics Board and adhered to the tenets of the Declaration of Helsinki.
The standard postoperative treatment protocol included ofloxacin 0.3% (Allergan, Inc) or moxifloxacin 0.5% (Alcon Laboratories, Inc.) four times a day for 1 week. Prednisolone acetate 1% (Allergan, Inc) was also applied topically every 2 hours for the first 24 hours. Patients were seen on postoperative day 1, after which the prednisolone acetate was decreased to four times daily for 1 week and subsequently stopped. If diffuse lamellar keratitis (DLK) or CTK was detected, this regimen was modified as indicated by the clinical course; treatment regimens included a combination of topical and systemic corticosteroids, topical and systemic antibiotics, vitamin C, Coenzyme-Q10, and surgical flap lift and irrigation if indicated. In patients diagnosed as having bilateral CTK, both eyes were treated with the same approach. If DLK or CTK was diagnosed, patients were seen daily until improvement was observed. Patients were then seen every few days until the resolution of disease.
DLK was diagnosed and graded clinically at the time of presentation based on the criteria described by Line-barger et al.9 CTK was diagnosed according to the criteria described by Hazin et al10 and Moshirfar et al.11 These criteria include the sudden presentation of a dense central haze without diffuse inflammation, onset within 2 weeks of surgery, and lack of gradual progression (in contrast to that seen in grades 3 and 4 DLK) (Figure A, available in the online version of this article).
Slit-lamp examination photographs of the (A and C) right and (B and D) left eyes of patient 7. (A) Classic central amorphous corneal opacification of central toxic keratopathy along with surrounding diffuse lamellar keratitis on postoperative day 6. (B) Central corneal opacification with epithelial defect on postoperative day 6. (C and D) Resolving opacification with residual striae in a classic “mud-crack” pattern on postoperative day 14.
Preoperative clinical data collection included the date of LASIK, patient age and sex, uncorrected (UDVA) and corrected (CDVA) distance visual acuity, manifest refraction, slit-lamp examination findings, keratometry, and pachymetry. Postoperative data up to 1 year after LASIK included visual acuity, manifest refraction, slit-lamp examination findings, medical and surgical management, Linebarger et al9 stage of preceding DLK if applicable, the dates of CTK diagnosis and resolution, and postoperative pachymetry and keratometry if available. The maximum hyperopic shift demonstrated throughout the postoperative course was also recorded. All patients had Pentacam Scheimpflug imaging (Oculus Optikgeräte GmbH) performed and some had optical coherence tomography (OCT) (Optovue, Inc). Pachymetry and keratometry measurements were obtained from Scheimpflug imaging and were taken from the thinnest point on the cornea and the corresponding keratometric value at that location, respectively. Expected pachymetry after surgery was calculated by subtracting the ablation depth from the preoperative pachymetry measurement. No intraoperative optical pachymetry was performed. Expected keratometry was calculated by multiplying the attempted spherical equivalent correction by 0.8 and subtracting this number from the pre-operative keratometry.
A total of 12 eyes from 8 patients (of 20,622 LASIK procedures over a 5-year period) were diagnosed as having CTK after FS-LASIK, yielding an occurrence rate of 0.058%. Nine eyes (75%) of 6 patients diagnosed as having CTK underwent LASIK between October 12 and December 20, 2017, likely representing an outbreak (Figure 1). The remaining 3 eyes (25%) of 2 patients developed CTK in 2015 and 2019, respectively, and likely were sporadic. Of the 8 patients, 6 were men and 2 were women. CTK occurred in the right eye eight times and in the left eye four times. CTK occurred bilaterally in 4 patients. Preoperative refractive errors included 4 eyes with myopia and 8 eyes with myopic astigmatism.
Timeline of central toxic keratopathy (CTK) diagnoses after laser in situ keratomileusis (LASIK) from 2015 to 2019 by month and year. The outbreak of CTK from October 10 to December 20, 2017 can be visualized.
Table 1 illustrates the clinical course and management for each eye diagnosed as having CTK. The mean time to diagnosis of CTK was 7 ± 2 days (range: 6 to 14 days). Eight of 12 eyes were diagnosed in 7 days or less. Nine eyes were diagnosed as having DLK of varying stages at the first postoperative visit. Ten eyes were diagnosed as having DLK at some point that preceded CTK. The most severe stage of DLK observed at any point before CTK was stage 2. One patient was diagnosed as having staphylococcal marginal keratitis in both eyes followed by CTK in the right eye. Five eyes developed an epithelial defect between postoperative days 5 and 9. Five eyes were treated surgically with flap lift and irrigation. The remaining 7 eyes were treated non-surgically. The average days after diagnosis to resolution of CTK was 53 days (range: 12 to 88 days) in eyes that underwent flap lift and irrigation compared to 33 days (range: 7 to 59 days) in eyes treated non-surgically. The mean time to resolution of CTK overall was 42 ± 30 days (range: 8 to 78 days).
Clinical Course and Management of CTK
Refraction and visual acuity at the preoperative visit and postoperative visits are listed for each eye in Table 2. The mean total follow-up time was 8 months (range: 3 weeks to 12 months). The maximum hyper-opic shift was +0.63 ± 0.86 diopters (D) (range: −0.75 to 2.38 D) on average. Of the 5 eyes that underwent flap lift and irrigation, the average maximum hyper-opic shift was +0.40 ± 0.57 D (range: −0.38 to +1.13 D), whereas the eyes treated non-surgically had an average maximum hyperopic shift of +0.81 ± 1.07 D (range: −0.75 to +2.38 D). The average final UDVA was 0.11 ± 0.14 logMAR. Six of 12 eyes achieved UDVA of 0.0 logMAR or better. All 5 eyes treated with flap lift and irrigation achieved UDVA of 0.1 logMAR or better; only 3 of 7 (43%) eyes treated non-surgically achieved the same. The average final CDVA was 0.02 ± 0.05 logMAR; 11 eyes had CDVA of 0.0 logMAR or better at the final postoperative visit.
Refraction and Visual Acuity Outcomes After CTK
Figure 2 illustrates the changes in pachymetry and keratometry in the 9 eyes that had these measurements documented postoperatively. Each of these eyes demonstrated thinning of the cornea beyond the expected postoperative thickness. The greatest amount of corneal thinning happened within the first month. At 1 month postoperatively, the eyes that were treated with flap lift and irrigation were on average 43 µm thinner than the expected postoperative pachymetry, whereas eyes treated non-surgically were on average 96 µm thinner. Among the 8 eyes that had follow-up pachymetry performed after 1 month, corneal thickness increased within 3 to 6 months postoperatively. Final pachymetry in patients treated with flap lift and irrigation measured on average 14 µm thinner than the expected postoperative pachymetry, whereas eyes treated non-surgically were on average 28 µm thinner.
Corneal thickness and keratometry after central toxic keratopathy (CTK). This shows the change in corneal thickness at (A) the thinnest point and (B) the corresponding keratometry at that location at the preoperative and postoperative visits. The average documented values of the 9 eyes with recorded measurements (solid line) was compared to the average expected postoperative value (dotted line, marked by *) to demonstrate a trend over time. D = diopters
Most patients also appeared to experience flattening of the anterior corneal curvature beyond what was expected in the first month but gradual steepening subsequently. In the first month postoperatively, keratometry in patients treated with flap lift and irrigation measured on average 3.30 D flatter than the expected postoperative value. In those treated non-surgically, the keratometry was on average 5.50 D flatter than expected. At the final follow-up visit, keratometry in patients treated with flap lift and irrigation measured on average 1.60 D flatter than expected, compared to 1.70 D flatter in those treated non-surgically.
The epidemiology of CTK after FS-LASIK has rarely been described in the literature. The first reported case following the use of a femtosecond laser was described by Hainline et al12 in 2005. A total of 9 eyes were diagnosed as having CTK from a retrospective review of 17,100 microkeratome and FS-LASIK procedures over a span of 10 years. We calculate the rate in this study to be 0.052%. Although the exact occurrence of CTK after FS-LASIK was not reported, 25% of their patients underwent flap creation using this method. In 2008, Stonecipher and Ignacio13 reported an incidence of 0.016% in a series of 6,131 eyes undergoing FS-LASIK, although the period of time was not specified. In the same year, a prospective study of 522 eyes by Hadden et al14 described 4 eyes in 3 patients that developed CTK during a 10-week period. We calculate the rate of CTK in this study to be 0.77%. Of note, we did not calculate a rate of CTK development using data from the literature describing CTK following microkeratome-assisted LASIK, such as the 2020 Sorenson et al15 case series. In our study, we retrospectively reviewed 20,622 FSLASIK procedures over 5 years and calculated an occurrence rate of 0.058%.
Because CTK can happen in clusters, incidence calculations may not accurately represent the probability of CTK arising in the general population and must be interpreted with caution. Evidently, these rates vary widely depending on whether sporadic or outbreak events are considered and could explain why some calculations from the literature are more than 10 times greater than others. Because multiple eyes in our study that were diagnosed as having CTK underwent LASIK on the same day, it may also be useful to consider how often CTK occurs based on surgical days in the calendar year. During our outbreak, CTK events occurred roughly every 10 surgical days in the span of 8 calendar weeks. However, the sporadic cases occurred every 463 days over a total of 5 years.
The pathophysiology of CTK is not well understood; proposed mechanisms include keratocyte apoptosis or enzymatic degradation leading to stromal loss and anterior corneal flattening in the absence of inflammation.2,11,16 Stromal loss may be due to neutrophil degranulation and the release of collagenase and proteolytic enzymes. The remaining keratocytes along with fibroblasts become activated, which can be seen on confocal microscopy and corresponds to the clinical finding of corneal haze.17,18 Other imaging modalities may also help diagnose CTK by revealing these corneal alterations. Keratometry may demonstrate anterior corneal flattening. OCT typically shows corneal thinning in the stroma and possibly the flap (Figure B, available in the online version of this article).11,16 Our patients demonstrated thinning of the cornea in the first month after LASIK, but gradually regenerated corneal thickness; this phenomenon may be due to epithelial regeneration and stromal remodeling.11
Optical coherence tomography (OCT) with corneal thickness measurements of the right eye of patient 7 after central toxic keratopathy (CTK). This demonstrates central haze and significant thinning of the cornea during the first postoperative month.
There is significant evidence that CTK is a distinct disease process from DLK. The conditions share some clinical features, such as central corneal opacification and hyperopic shift, and DLK may often precede CTK.10 Hazin et al10 suggested that timing of presentation (CTK usually yields classic findings within the first week postoperatively, whereas stage 4 DLK rarely presents before postoperative week 3) and the lack of inflammation may help to differentiate these conditions. Although 83% of eyes in our study were diagnosed as having DLK prior to CTK, no patient progressed beyond stage 2 DLK before developing CTK. Additionally, all eyes developed CTK within the first 2 weeks after surgery, whereas it is extremely unlikely that stage 4 DLK would present within 3 weeks of surgery.10 These findings are consistent with the observations made by Hazin et al10 and support the theory that CTK and DLK are distinct entities.
The development of CTK could be initiated by excimer laser photoactivation of certain substances causing a toxic reaction.2 Suspected inciting factors include powder from surgical gloves, surgical marking pens, tonometer disinfectant, postoperative debris from microkeratome blades, and meibomian gland secretions.2,11,14 Although there have been several case series concerning epidemic outbreaks of DLK, there are fewer published on CTK. These clusters of CTK have been attributed to changing brands of surgical gloves7 and altered viscosity of sodium hyaluronate 0.4% drops.8
In our study, a majority of CTK diagnoses were made within a range of 2 months in a single year. When this outbreak was identified, a thorough investigation was performed to detect changes in the surgical environment, femtosecond and excimer laser settings, and materials used. The autoclave at our facility was replaced soon after the outbreak began, because autoclave reservoir biofilms have been implicated in clusters of DLK and CTK.15 However, the outbreak continued despite these changes. Several theories emerged—recent renovation of the air ducts near the surgical suite and an inappropriately concentrated enzymatic cleaner for surgical instruments were implicated. Although it was not possible to confirm an exact cause, no further eyes were diagnosed as having CTK after the appropriate steps were taken to resolve these issues. Of note, each of the cases in our study, whether unilateral or bilateral, involved the development of CTK in the right eye; CTK was never diagnosed in the left eye alone. One possible reason is that our surgeons begin procedures with the right eye. If foreign material contaminated the surgical instruments, whether from a cleaning solution or airborne particles, most of this material would contact the right eye before reaching the left eye. Hainline et al12 noted a similar phenomenon in their series.
Currently, there is no standard of care for CTK. Topical steroids may worsen disease progression in CTK, possibly due to delayed keratocyte proliferation.8 Most of our patients were initially treated with intensive steroid therapy to manage the preceding DLK and subsequently tapered after the diagnosis of CTK was made; this may have affected the healing process. Surgical management in CTK with flap lifting and irrigation is controversial. Some report that surgical management does not improve outcomes and may worsen stromal loss.7,9,19,20 Others report that early surgical management may not worsen stromal loss and may be the preferred approach.13,21,22 Many recommend an approach that involves only close follow-up and monitoring.2,9,11 Our findings indicated that patients who underwent flap lift and irrigation appeared to have a longer disease course compared to those treated non-surgically, but on average experienced less severe hyperopic shifts, decreased corneal thinning, and reduced flattening. Additionally, these patients ultimately achieved better UDVA and CDVA. These benefits could be due to the removal of the inciting agent in the early postoperative period, mitigating its toxic effect at an earlier stage. Our findings indicate that flap lift and irrigation in the management of CTK may not be as detrimental as previously suggested and may even be beneficial in some instances.
Limitations to our study include its retrospective nature, small sample size, and varied duration of follow-up. Patients were not randomized to undergo medical versus surgical treatment. It should also be noted that the average time to diagnosis of CTK was longer in the medical group than in the surgical group (8.1 versus 6.4 days), which may have led to later initiation of treatment in the medical group. Due to the rarity of this condition, despite our relatively large sample size, a comparative analysis could not be performed between patients treated with different modalities. Nevertheless, these results suggest that, in some cases, surgical management of CTK with flap lift and irrigation may improve outcomes in terms of visual acuity, refractive error, and corneal anatomy.
Although our retrospective review suggests that CTK remains an uncommon complication, it is essential to recognize that CTK can occur in clusters. Early identification of these outbreaks can allow appropriate investigation and intervention to prevent further cases. We believe that our outbreak was due to contaminated air or improper dilution of detergent, but our patients did well: most eyes achieved UDVA of 0.1 logMAR or better and all but one eye achieved CDVA of 0.0 logMAR or better. Although a standard of care for CTK has not been established, early recognition and appropriate management, which may include surgical intervention, can lead to acceptable visual outcomes.
- Fraenkel GE, Cohen PR, Sutton GL, Lawless MA, Rogers CM. Central focal interface opacity after laser in situ keratomileusis. J Refract Surg. 1998;14(5):571–576.
- Sonmez B, Maloney RK. Central toxic keratopathy: description of a syndrome in laser refractive surgery. Am J Ophthalmol. 2007;143(3):420–427. doi:10.1016/j.ajo.2006.11.019 [CrossRef]
- Koh K, Jun I, Kim TI, Kim EK, Seo KY. Central toxic keratopathy after small incision lenticule extraction. Korean J Ophthalmol. 2020;34(3):254–255. doi:10.3341/kjo.2019.0122 [CrossRef]
- Moshirfar M, Kurz C, Ghajarnia M. Contact lens-induced keratitis resembling central toxic keratopathy syndrome. Cornea. 2009;28(9):1077–1080. doi:10.1097/ICO.0b013e318197ec3a [CrossRef]
- Hsu M, Tu E, Bouchard C. Confocal microscopy of contact lens keratitis presenting as central toxic keratopathy. Eye Contact Lens. 2011;37(6):377–380. doi:10.1097/ICL.0b013e31821c0401 [CrossRef]
- Ting DSJ, Ghosh S. Central toxic keratopathy after contact lens wear and mechanical debridement: clinical characteristics, and visual and corneal tomographic outcomes. Eye Contact Lens. 2019;45(4):e15–e23. doi:10.1097/ICL.0000000000000575 [CrossRef]
- Marí Cotino JF, Suriano MM, De La Cruz Aguiló RI, Vila-Arteaga J. Central toxic keratopathy: a clinical case series. Br J Ophthalmol. 2013;97(6):701–703. doi:10.1136/bjophthalmol-2012-302732 [CrossRef]
- Jutley G, Aiello F, Robaei D, Maurino V. Central toxic keratopathy after laser in situ keratomileusis. J Cataract Refract Surg. 2014;40(12):1985–1993. doi:10.1016/j.jcrs.2014.03.030 [CrossRef]
- Linebarger EJ, Hardten DR, Lindstrom RL. Diffuse lamellar keratitis: diagnosis and management. J Cataract Refract Surg. 2000;26(7):1072–1077. doi:10.1016/S0886-3350(00)00468-5 [CrossRef]
- Hazin R, Daoud YJ, Khalifa YM. What is central toxic keratopathy syndrome if it is not diffuse lamellar keratitis grade IV?Middle East Afr J Ophthalmol. 2010;17(1):60–62. doi:10.4103/0974-9233.61218 [CrossRef]
- Moshirfar M, Hazin R, Khalifa YM. Central toxic keratopathy. Curr Opin Ophthalmol. 2010;21(4):274–279. doi:10.1097/ICU.0b013e32833a8cb2 [CrossRef]
- Hainline BC, Price MO, Choi DM, Price FW Jr, . Central flap necrosis after LASIK with microkeratome and femtosecond laser created flaps. J Refract Surg. 2007;23(3):233–242. doi:10.3928/1081-597X-20070301-05 [CrossRef]
- Stonecipher K, Ignacio T. Complications and management with the femtosecond laser. In: Alió JL, Azar DT, eds. Management of Complications in Refractive Surgery.Springer-Verlag Publishing; 2008:169–174. doi:10.1007/978-3-540-37584-5_9 [CrossRef]
- Hadden OB, McGhee CNJ, Morris AT, Gray TB, Ring CP, Watson ASJ. Outbreak of diffuse lamellar keratitis caused by marking-pen toxicity. J Cataract Refract Surg. 2008;34(7):1121–1124. doi:10.1016/j.jcrs.2008.03.024 [CrossRef]
- Sorenson AL, Holland S, Tran K, et al. Diffuse lamellar keratitis associated with tabletop autoclave biofilms: case series and review. J Cataract Refract Surg. 2020;46(3):340–349. doi:10.1097/j.jcrs.0000000000000070 [CrossRef]
- Sikder S, Khalifa YM, Neuffer MC, Moshirfar M. Tomographic corneal profile analysis of central toxic keratopathy after LASIK. Cornea. 2012;31(1):48–51. doi:10.1097/ICO.0b013e31821de341 [CrossRef]
- Thornton IL, Foulks GN, Eiferman RA. Confocal microscopy of central toxic keratopathy. Cornea. 2012;31(8):934–936. doi:10.1097/ICO.0b013e3181f7f109 [CrossRef]
- Hau SCH, Allan BD. In vivo confocal microscopy findings in central toxic keratopathy. J Cataract Refract Surg. 2012;38(4):710–712. doi:10.1016/j.jcrs.2012.01.010 [CrossRef]
- Parolini B, Marcon G, Panozzo GA. Central necrotic lamellar inflammation after laser in situ keratomileusis. J Refract Surg. 2001;17(2):110–112.
- Garcia-Gonzalez M, Gil-Cazorla R, Teus MA. Surgical flap amputation for central flap necrosis after laser in situ keratomileusis. J Cataract Refract Surg. 2009;35(11):2018–2021. doi:10.1016/j.jcrs.2009.05.045 [CrossRef]
- Lyle WA, Jin GJC. Results of flap repositioning after laser in situ keratomileusis. J Cataract Refract Surg. 2000;26(10):1451–1457. doi:10.1016/S0886-3350(00)00571-X [CrossRef]
- Tu KL, Aslanides IM. Surgical intervention in central toxic keratopathy. Eur J Ophthalmol. 2012;23(1):0. doi:10.5301/ejo.5000155 [CrossRef]
Clinical Course and Management of CTK
|Patient||Eye||No. of Days to Diagnosis of CTK||Highest Stage of Preceding DLK||Management||No. of Days to Resolution of CTK|
|1||1||4||2||FL + I POD 4, + BCL||88|
|2||2||7||2||FL + I POD 7||12|
|3||7||2||FL + I POD 7||12|
|3||4||7||2||FL + I POD 7, BCL||77|
|4||5||7||2||FL + I POD 7||78|
|8||12||8||2||Medical, BCL ×2||8|
Refraction and Visual Acuity Outcomes After CTK
|Patient||Eye||Laterality||Preop||Postop Day 1||Postop Month 1||Postop Month 3||Final Postop||Max HyperopicShift (SE)|
|UDVA||SE (D) (logMAR)||UDVA (logMAR)||UDVA||SE (D) (logMAR)||UDVA||SE (D) (logMAR)||UDVA (logMAR)||CDVA (logMAR)||SE (D)|
|Average SE (D)||−2.57 ± 1.23 (−4.50 to −0.75)||−0.08 ± 0.78 (−1.25 to +1.13)||+0.23 ± 1.15 (−1.25 to +2.38)||+0.43 ± 0.95 (−1.00 to +2.38)|