Contact Lens F allure in Keratoconus Management Mohamad R. Dana, MD, MPH, l Jamie L. Putz, COT, NCLC, 1 Marlos A. G. Viana, PhD,1,2 Joel Sugar, MD,l Timothy T. McMahon, OD1 Purpose: Surgery is indicated for keratoconus when management with contact lenses fails. The authors sought to determine the relative contribution of various preoperative patient and ocular factors to the ultimate causes of contact lens failure. Methods: The records of unoperated eyes of keratoconus patients whose contact lenses were managed intensively before undergoing penetrating keratoplasty (PK) at the authors' institution between 1981 and 1990 were selected for study. Univariate and multivariate analyses were performed to identify risk factors for ear1y contact lens failure. Results: The records of 99 keratoconic eyes of 75 patients with an average age of 34 years and average keratometry readings of 57.5 diopters at presentation were studied. Cases had been followed for an average of 27 months before PK. The primary reasons for PK were a best-corrected visual acuity of under 20/40 (despite good contact lens fit) causing disability for the patient (43%), contact lens intolerance (32%), frequent lens displacement (13%), and significant peripheral thinning of the cornea (12%). The referral source of the patient, sex, a history of PK in the fellow eye, or of contact lens wear in either eye did not alter the relative contributions of these parameters to surgery. Conclusion: Poor best-corrected visual acuity at presentation, higher keratometry readings (~55 0), age (~40 years), and duration of disease (>5 years) were significantly associated with failure due to poor functional acuity and peripheral thinning, frequently leading to surgery within the first 12 months after presentation. Ophthalmology 1992;99: 1187-1192
Keratoconus is a noninflammatory ectasia of the cornea in which the cornea assumes an irregular conical shape.l-4 Reduction in vision is primarily due to the irregular topography and ensuing astigmatism, and less frequently secondary to corneal scarring. 1 Spectacle correction that Originally received: November 19, 1991. Revision accepted: March 5, 1992. I Department of Ophthalmology and Visual Sciences, The Illinois Eye and Ear Infirmary, University of Illinois at Chicago College of Medicine, Chicago. 2 Epidemiology and Biostatistics Program, University of Illinois at Chicago School of Public Health, Chicago. Presented in part at the Association for Research in Vision and Ophthalmology Annual Meeting, Sarasota, April 1991. Supported in part by core grant EY 1792 from the National Eye Institute/ NIH, Bethesda, Maryland, and by an unrestricted research grant from Research to Prevent Blindness, Inc, New York, New York. Reprint requests to Timothy T. McMahon, OD, Department of Ophthalmology and Visual Sciences, UIC Eye Center, 1855 W Taylor St, Chicago, IL 60612.
may be effective in treating early keratoconic myopic astigmatism 1,5 is inadequate for the majority of patients who present because of high degrees of corneal distortion. Contact lenses playa central role by providing a regular refracting surface in lieu of the distorted keratoconic cornea, thereby delaying or preventing surgery. 1,3,4,8 In a 20year retrospective study, 95% of keratoconus patients were successfully fitted with contact lenses and 81 % achieved an acuity of 20/40 or better. 9 The precise percentage of keratoconus patients who will eventually require surgery is difficult to determine since the total "at risk" population remains unknown. Studies addressing this issue have generally been biased by referral populations that are not representative of the keratoconic population as a whole. 2 ,4,5.7.9 In a 48-year epidemiologic study performed from 1935 to 1982 by Kennedy and co-workers, IO the 20-year cumulative survivorship without transplantation was greater than 80%. Surgical procedures other than penetrating keratoplasty (PK) have been described for the management of keratoconus,II-14 but their applications remain limited. To
1187
Ophthalmology
Volume 99, Number 8, August 1992
this day, PK remains the mainstay of surgical intervention for keratoconus 15- 18 with excellent optical and graft survivorship results. Very few studies have examiQed the management ofa large number of keratoconus patients to specifically analyze the criteria for surgery. To that end, we retrospectively analyzed the records of 99 keratoconic eyes that underwent PK after contact lens management failure. Our goals were to: (I) describe patient and ocular variables characteristic of those with contact lens failure despite initial successful fits, (2) classify the underlying causes for surgical intervention, and (3) relate those causes to patterns of preoperative patient factors that may prove predictive of the ultimate cause for contact lens failure.
Materials and Methods The records of 99 keratoconic eyes of 75 patients who underwent PK in the 10 years between 1981 and 1990 at the University of Illinois Eye and Ear Infirmary were selected for study. Only those patients who had been referred to our Cornea or Contact Lens Services, and who were followed and managed intensively for contact lens (re)fitting before surgery were included. In all patients, the diagnoses of keratoconus made on classic clinical findings of conical stromal thinning, Fleischer ring, subepithelial fibrosis, Vogt's striae, and irregular mires on keratometry were further substantiated by postoperative pathologic confirmation. Exclusion criteria included a history of ocular surgery or significant eye trauma. Furthermore, all those who at the baseline examination were deemed to have corneas that precluded successful contact lens fitting were excluded from study. One investigator (TTM) was responsible for the fitting of all patients for the duration of the 10 years of the study, so that the guidelines followed were internally consistent. Patients were fitted from an extensive inventory oflenses including standard design spherical rigid polymethylmethacrylate and gas-permeable (RGP) lenses (Boston II, Paraperm 02, Polycon II), spherical and toric hydrophilic lenses, and special design (Soper bispheric, bitoric, and "piggyback") lenses. In general, for mild disease, spherical rigid lenses were used, and these were fit with the base curve approximating the flattest keratometry reading, adjusting central and peripheral curvatures to achieve centration and adequate tear exchange by slit-lamp biomicroscopic evaluation of the fluorescein pattern. Except for special circumstances, patients fit with rigid lenses were given RGP, rather than polymethylmethacrylate, lens materials. In select cases when rigid lenses were not well tolerated, hydrophilic lens fittings were attempted. In advanced disease, base curve and peripheral curve designs were altered to allow a "threepoint touch" fit characterized by minimal apical touch and two areas of (mid)peripheral touch with equal distribution of fluorescein to weightbearing areas. 1,3,7 Finally, among those with mild to moderate disease, we generally aimed to have sufficient (superior lens) lid support to increase lens stability and to counter the propensity of in-
1188
ferior lens lift-off. In advanced disease, with corneal curvatures exceeding 52 diopters (D), special design lenses were frequently needed to achieve optimal centration, visual acuity, and comfort. A systematic review of the records by one investigator (MRD) was aimed to catalog the following parameters: (I) demographics, (2) source and reason of referral, (3) pertinent medical history including history of atopic disease, (4) pertinent ophthalmic history including previous PK (in the contralateral eye), previous contact lens wear and optical correction, (5) severity of disease parameters including baseline and preoperative best-corrected visual acuity, average keratometry and cylinder readings at presentation, and duration of disease if known, and (6) measure of effort in lens fitting including the number and types of lens fittings attempted and the duration of followup from baseline examination PK. Attention was subsequently focused to determine the reason(s) why cases had failed contact lenses. First, problem areas such as poor acuity, contact lens-related discomfort, frequent lens decentration, or peripheral thinning were identified. Second, four categories were constructed to establish the leading cause oflens failure: ( I) those with a best-corrected visual acuity of less than 20/40 despite a good contact lens fit, and who subjectively felt disabled by such vision to a degree that they opted for PK; (2) those who were severely contact lens intolerant, and therefore had inadequate functional wearing times, usually of less than 8 hours a day; (3) those with recurrent lens displacement and decentration despite numerous lens trials; and (4) those with advanced disease who showed progressive and marked peripheral corneal thinning, which necessitated grafting. Statistical methods for analysis included the chi-square test (with Yates' correction) for association between discrete variables. Fisher's exact test (two-tailed) was applied when the sample size was not adequate for the chi-square test. Student's t test was used to compare univariate means. Multivariate analysis was performed using a linear regression model relating the timing of surgery to multiple baseline variables.
Results Patient Characteristics The records of 99 keratoconic eyes of 75 patients (52 males, 23 females) who met our inclusion criteria were studied. The age of subjects at the time of presentation to our care ranged from 15 to 75 years (average, 34 years). The patients' average known duration of disease (from time of initial diagnosis until their presentation to us) was 8.4 years, with no significant difference between the sexes. Twenty-one percent of patients had a documented history of atopic disease (allergic rhinitis, asthma, or chronic urticaria) and 18 percent had already undergone PK in the contralateral eye by the time of initial presentation. More than 90% of those studied represented a referral population, with two thirds referred by other ophthal-
Dana et al . Contact Lens Failure in Keratoconus Management mologists (Table I). The reasons for referral were distributed equally between the optometrists and ophthalmologists, with 75% referred primarily for general evaluation and contact lens management.
Ocular Variables Optical Correction Worn. Results of independent evaluation of the 99 keratoconic eyes showed that 85% wore Table 1. Selective Patient and Ocular Characteristics Referring Source Ophthalmologist Optometrist Self Other
,.....
Average Keratometry at Baseline (D) Under 45
45-49 50-54 55-59 60-64 65 or more Indeterminant
Indeterminant/Unrecorded
67% 25% 7% 1% 3% 11% 15% 31%
24% 14% 2% 47% 25% 10%
7% 11%
Optical Correction at Baseline None Spectacles Contact lenses Hard (PMMA) Standard rigid gas-permeable Soft/hydrophilic Unspecified/other (special design)
75% 55% 14% 11% 20%
Ultimate Contact Lens Design Attempted Preoperatively Special design Standard rigid gas-permeable Hard (PMMA) Softjhydrophilic
74% 12% 9% 5%
Number of Lens Trials· per Eye before PK 1 2 3 to 5 6 to 10 More than 10
30% 23% 27% 13% 7%
15% 10%
o = diopters; PMMA = polymethylmethacrylate; PK = penetrating keratoplasty.
• Trials performed after referral to our services.
Baseline
Preoperative
10% 24% 24% 26% 56/0 11%
1% 22% 26% 27% 13% 11%
100%
100%
20/20 to 20/25 20/30 to 20/40 20/50 to 20/60 20/70 to 20/100 20/200 20/400 or less Total
Average Corneal Cylinder at Baseline (D)
1-5 6-10 11-15 16 or more
Table 2. Best-corrected* Baseline and Preoperative Visual Acuity
• Includes contact lens correction.
some type of correction at baseline; 75% were managed with contact lenses (Table I). By the time PK was contemplated, 86% of cases were being managed by standard RGP or special design lenses. Mean Keratometry, Cylinder, and Contact Lens Fittings. The mean keratometry reading at the time of baseline evaluations was 57.5 D, and the mean diopters of cylinder was 7 (Table I). Patients underwent as few as I, and as many as 15 contact lens fittings (mean, 3.8 fits) before keratoplasty, often involving multiple lens designs. Visual Acuity. More than one third (34%) of examined eyes had a best-corrected visual acuity of 20/40 or better at baseline, in contrast to 23% preoperatively (Table 2).
Factors Associated with Causes of Contact Lens Failure For all 99 eyes counted as a whole, the chief cause of lens failure was unacceptably poor vision for 43% (95% confidence interval [CI], 36% to 52%), intractable contact lens intolerance for 32% (95% CI, 25% to 39%), frequent lens displacement for 13% (95% CI, 9% to 20%), and significant peripheral thinning for 12% (95% CI, 7% to 19%). No significant differences existed between the two sexes. Among the 16 patients with a history of atopy, contact lens intolerance was the chief indication for surgical intervention. However, the sample size was too small for the rates to reach significance (P > 0.5). Ten of the 24 (42%) bilaterally operated cases had failed contact lenses because of poor vision in one eye and lens intolerance in the fellow eye. Analysis of data from bilaterally operated eyes during the 10 years of study showed no significant association between causes of lens failure from one eye to the other, validating independent analysis of each eye. Baseline Visual Acuity. The chief cause of failure among those eyes with visual acuity of 20/40 or better at the time of initial presentation was contact lens intolerance (Table 3), in contrast to those who failed lens management due to disablingly poor vision or peripheral thinning (Fisher's exact test, P < 0.05). Average Keratometry. Analysis of mean keratometry readings at baseline (Table 3) showed that more than half of those eyes with mean keratometry readings ofless than
1189
Ophthalmology
Volume 99, Number 8, August 1992
Table 3. Selected Ocular Variables Based on Final Outcomes* by Percent
Baseline acuity 20/20-40 20/50-100 ;:5;20/200 Average keratometry, diopeters 5 Timing of PK in months ;:5;12 >12 PK
=
Poor Vision
Lens Intolerance
Lens Displacement
Thinning
Total
30 50 50
40 30 25
24 6 12.5
6 14 12.5
100 100 100
31 49
52 24
14 13
3 14
100 100
49 20
27 55
13
11
15
10
100 100
52 37
21 40
10 16
17 7
100 100
penetrating keratoplasty .
• See Materials and Methods section for definitions.
55 D became intractably lens intolerant, while less than a quarter of those with higher keratometry readings went on to PK for the same reason (Fisher's exact test, P < 0.01). Conversely, those who underwent PK for poor vision or peripheral thinning tended to have keratometry readings of 55 D or higher. Contact Lens Fitting Trials. There was an inverse relationship between the number of lens fittings and the likelihood of undergoing PK for unacceptably poor vision (Table 3). Conversely, more than five contact lens fittings were associated with lens management failure because of eventual intractable lens intolerance (chi-square test, P < 0.05). Timing of Surgery. The average periods of follow-up from baseline to PK varied widely depending on the cause oflens failure (Table 3), ranging from 16 months for those
Table
4.
Selected Patient Characteristics Based on Final Outcomes* by Percent
Age (yrs) 10 Average
Poor \7ision
Lens Intolerant
43 42 40 50 35 yrs
43 38 37 18 31 yrs
37 53 50 9 yrs
39 7 18 7 yrs
• See Materials and Methods section for definitions.
1190
undergoing PK for thinning to 36 months for those who had surgery because of frequent lens displacement. Significantly, half of those who had surgery in the first year of follow-up represented the low vision group with only a quarter representing those with contact lens intolerance; the relative contributions of the two factors reversing among those who had surgery after 13 months of followup (chi-square test, P < 0.05). Age. The average age for those undergoing PK for thinning (41 years) was the highest compared with that of others having surgery for other reasons (Table 4). There was a significant difference between the odds of undergoing surgery for poor vision or thinning among those older than 40 years of age (71 %) compared with the same odds among patients younger than 40 years (47%) (chisquare, P < 0.05).
Lens Displacement
Thinning
Total
32 yrs
0 10 16 21 41 yrs
100 100 100 100
12 33 14 8 yrs
12 7 18 12 yrs
100 100 100
14 10 17 11
Dana et al . Contact Lens Failure in Keratoconus Management Duration of Disease. The mean known duration of disease before presentation was highest for those undergoing PK for thinning (12 years) compared with those who had surgery for other reasons (Table 4). Expectedly, duration of disease and preoperative age showed a significant positive association, while there was an inverse relationship between disease duration and baseline acuity. Patient sex, source and reason for referral, prior history of contact lens wear or ofPK in the fellow eye, and amount of cylinder present at baseline evaluation showed no significant association with the failure patterns of contact lens management.
Factors Associated witllTiming of Penetrating Keratoplasty The mean follow-up period between the baseline evaluation and PK was 27 months (males, 26 months; females, 29 months). Baseline Visual Acuity. Visual acuity at baseline showed a significant relationship with the length of successful contact lens fit. The average period of follow-up until PK for those with visual acuity of 20/40 or better was 43 (95% CI, 30 to 56) compared with 19 (95% CI, 13 to 24) months for those with poorer vision. Average Keratometry. There was a statistically significant inverse relationship between baseline keratometry readings and the length of time until surgery. Eyes with an average keratometry reading of less than 55 D had a ;mean period to surgery of 42 (95% CI, 29 to 54) months compared with those with keratometry readings of more than 55 D, who had a mean period to surgery of21 (95% CI, 14 to 27) months. Duration of Disease. Known disease duration and the timing of surgery showed an inverse relationship, but this difference did not reach statistical significance (P > 0.5). Application of a linear regression model incorporating mUltiple variables showed the independent retention of statistical significance by average keratometry (P < 0.05) and baseline visual acuity (P < 0.01), with no gain in predictability of surgical need by consideration of age, disease duration, or cylinder. Referral source, as well as a prior history of contact lens wear or of PK in the fellow eye, also failed to demonstrate significant associations with the timing of surgery.
Discussion Advances in lens design have made contact lenses applicable to all but the most advanced, or intractably lens intolerant cases. Various studies2,5,7 have conclusively shown that a majority of those cases with even relatively advanced disease, labeled "contact lens intolerant" according to standard ophthalmic care, can be successfully refitted with contact lenses. Our 10-year series represented a group with advanced disease, 90% representing a referral population. Stringent inclusion criteria assured that only those referral patients followed for intensive contact lens fitting before surgery
by the same practitioners (TTM and JS), and who eventually underwent PK according to the guidelines of one of our investigators (JS), were included in the study. Unlike many retrospective studies suffering from problems of variable outcome measurement and follow-up time, our study's major strength is its internal consistency in outcome definition, along with the fact that surgery itself was the end point in all cases, making the variable followup time not a weakness but simply another parameter to be measured. Our results strongly suggest that even among those with advanced keratoconus, lens intolerance does not represent the chief cause for the cessation of contact lens management. This parallels the findings of Kastl and co-workers9 who, in a 20-year retrospective study, found that poor vision was the primary reason for undergoing PK among the 4.8% minority of cases that failed contact lenses. It differs from the 5-year experience of Smiddy and coworkers7 who found that a majority of their cases that were originally fitted with contact lenses but who went on to PK (n = 51 eyes) did so because oflens intolerance, despite very comparable baseline average keratometries of 56.8 D in their "pre-keratoplasty" group compared with 57.5 D among ours. In the largest series of its kind, a 4-year multicenter retrospective study of the clinical management of keratoconic eyes,2 74% did not require surgery, 21 % underwent PK, and the remainder underwent epikeratoplasty. The investigators concluded that a best-corrected acuity ofless than 20/40, average keratometry of more than 55 D, and a history of contact lens wear were significantly associated with a PK outcome. After an average follow-up period of 16 months, 34 eyes (28 patients) in their series originally fitted with contact lenses became contact lens failures53% because of lens intolerance, 24% because of poor vision, and 12% because of problems with lens centration and handling. Our series of 99 eyes that were originally successfully fitted with contact lenses but which eventually underwent PK is approximately three times the size of the comparable group in the multicenter study cited above,2 with an average follow-up period from presentation to surgery (27 months) almost twice as long. Interestingly, the relative frequencies ofa surprising number of variables including age, sex, atopic disease, referral pattern, duration of disease, baseline visual acuity, and average keratometry show remarkable similarities between the two studies. As expected, and similar to the findings of Lass and co-workers, 2 we found low baseline visual acuity and steep corneal curvature to be predictors of (early) surgical intervention. Our findings also are similar to those of Lass and colleagues in failing to demonstrate a significant predictive role for sex, atopy, or corneal cylinder. Corneal cylinder seems to be an unreliable predictor of outcome, not because irregular corneal curvature is unimportant in the disease pathophysiology, but because of the inherent limitations of conventional (central) keratometry in delineating corneal topographic changes. Our experience indicates a significant association between a baseline best-corrected visual acuity of less than
1191
Ophthalmology
Volume 99, Number 8, August 1992
20/40, an average keratometry of more than 55 D, and contact lens failure due to a poor functional acuity despite a good fit. It is noteworthy that keratometry remains a poor indicator of the eventual cause of lens failure. This is so because whereas the probability of lens intolerance increases with increasing corneal curvature, its relative contribution to lens failure decreases, thereby making it the outcome most weakly associated with higher keratometry readings-in contrast to peripheral thinning or low acuity. Our results indicate that those patients with a longer disease duration (more than 5 years) are more likely to undergo PK for poor vision or thinning compared with lens intolerance. Second, a longer disease duration is predictive of an earlier surgical outcome. This is not surprising, as those with a longer duration of keratoconus represent a subset of patients with more advanced disease who present with poorer vision and steeper corneas. Interestingly, a prior history of PK (in the fellow eye) at baseline fails to demonstrate any significant association with disease duration, baseline acuity, ultimate cause of contact lens failure, or timing of surgery. Although a history of PK in one eye could signify a more longstanding and advanced disease state, and possibly decrease the threshold for referral (bringing the case to attention earlier), surgery in one eye also could significantly ameliorate functional vision to a point where the seeking of surgical attention for the unoperated eye is delayed. We conclude that, based on our experience with the contact lens management of advanced keratoconic eyes, they may be classified into two groups. The first group includes older patients with a longer duration of disease and high baseline keratometry readings who tend to have a lower visual acuity at presentation. These patients, for the most part, undergo fewer lens fittings and have surgery (PK) early, usually for low functional vision or for peripheral thinning. The second group includes younger patients with a more recently diagnosed condition who have lower baseline keratometry readings and who tend to have a better visual acuity at presentation. The atopic cases tend to cluster in this group. These patients undergo multiple lens fittings and tend to eventually have surgery many months after their presentation, usually failing their contact lens management secondary to intractable lens intolerance. These constructs are compatible with the known natural history of keratoconus, in which increasing duration of disease (associated with a rise in age) leads to higher corneal curvatures with a concomitant decrease in acuity, all combining to increase the odds of contact lens management failure. The encouraging experience with contact lenses in the management of even advanced keratoconic disease bodes. well for patients, who can expect to maintain good vision (and thereby delay or avoid surgery) with an intensive effort at contact lens fitting. The indications for surgical intervention are few, but a small albeit significant proportion of cases with advanced disease may reach that end point. To our knowledge, our study represents the first exclusive and systematic analysis of that group of
1192
keratoconic cases who eventually fail contact lens management and undergo PK. By elaborating on the patterns of contact lens failure in keratoconus management and outlining the risk factors for the various failure outcomes, we hope to have provided significant natural history data on which realistic clinical goals and expectations may be constructed.
References 1. Belin MW. Optical and surgical correction of keratoconus. Focal Points 1988: Clinical Modules for Ophthalmologists. San Francisco: American Academy of Ophthalmology, 1988; vol. VI, module 11. 2. Lass JH, Lembach RG, Park SB, et aI. Clinical management of keratoconus. A multicenter analysis. Ophthalmology 1990;97:433-45. 3. Krachmer JH, Feder RS, Belin MW. Keratoconus and related noninflammatory corneal thinning disorders. Surv Ophthalmol 1984;28:293-322. 4. Lawless M, Coster DJ, Phillips AJ, Loane M. Keratoconus: diagnosis and management. Aust NZ J Ophthalmol 1989; 17: 33-60. 5. Belin MW, Fowler WC, Chambers WA. Keratoconus: evaluation of recent trends in the surgical and nonsurgical correction of keratoconus. Ophthalmology 1988;95:335-9. 6. Buxton IN, Keates RH, Hoefle FB. The contact lens correction of keratoconus. In: Dabezies OH Jr, ed. Contact Lenses: The CLAO Guide to Basic Science and Clinical Practice. Vol. 2. Orlando: Grone & Stratton, 1984; chap. 55. 7. Smiddy WE, Hamburg TR, Kracher GP, Stark WJ. Keratoconus: contact lens or keratoplasty? Ophthalmology 1988;95:487-92. 8. Carney LG. Contact lens correction of visual loss in keratoconus. Acta Ophthalmol 1982;60:795-802. 9. Kastl PR, Donzis PB, Cole HP III, et al. A 20-year retrospective study of the use of contact lenses in keratoconus. CLAO J 1987;13:102-4. 10. Kennedy RH, Bourne WM, Dyer JA. A 48-year clinical and epidemiologic study of keratoconus. Am J Ophthalmol 1986;101:267-73. 11. Arentsen JJ, Laibson PRo Thermokeratoplasty for keratoconus. Am J Ophthalmol 1976;82:447-9. 12. Wood TO. Lamellar transplants in keratoconus. Am J Ophthalmol 1977;83:543-5. 13. McDonald MB, Kaufman HE, Durrie DS, et al. Epikeratophakia for keratoconus. The nationwide study. Arch Ophthalmol 1986; 104: 1294-1300. 14. McDonald MB, Safir A, Waring GO III, et aI. A preliminary comparative study of epikeratophakia or penetrating keratoplasty for keratoconus [letter]. Am J Ophthalmol 1987;103:467. 15.· Payne JW. Primary penetrating keratoplasty for keratoconus: a long-term followup. Cornea 1982;1:21-7. 16. Paglen PG, Fine M, Abbott RL, Webster RG Jr. The prognosis for keratoplasty for keratoconus. Ophthalmology 1982;89:651-4. 17. Troutman RC, Lawless MA. Penetrating keratoplasty for keratoconus. Cornea 1987;6:298-305. 18. Steinert RF, Wagoner MD. Long-term comparison of epikeratoplasty and penetrating keratoplasty for keratoconus. Arch Ophthalmol 1988;106:493-6.
Keratoconus is a noninflammatory ectasia of the cornea in which the cornea assumes an irregular conical shape.l-4 Reduction in vision is primarily due to the irregular topography and ensuing astigmatism, and less frequently secondary to corneal scarring. 1 Spectacle correction that Originally received: November 19, 1991. Revision accepted: March 5, 1992. I Department of Ophthalmology and Visual Sciences, The Illinois Eye and Ear Infirmary, University of Illinois at Chicago College of Medicine, Chicago. 2 Epidemiology and Biostatistics Program, University of Illinois at Chicago School of Public Health, Chicago. Presented in part at the Association for Research in Vision and Ophthalmology Annual Meeting, Sarasota, April 1991. Supported in part by core grant EY 1792 from the National Eye Institute/ NIH, Bethesda, Maryland, and by an unrestricted research grant from Research to Prevent Blindness, Inc, New York, New York. Reprint requests to Timothy T. McMahon, OD, Department of Ophthalmology and Visual Sciences, UIC Eye Center, 1855 W Taylor St, Chicago, IL 60612.
may be effective in treating early keratoconic myopic astigmatism 1,5 is inadequate for the majority of patients who present because of high degrees of corneal distortion. Contact lenses playa central role by providing a regular refracting surface in lieu of the distorted keratoconic cornea, thereby delaying or preventing surgery. 1,3,4,8 In a 20year retrospective study, 95% of keratoconus patients were successfully fitted with contact lenses and 81 % achieved an acuity of 20/40 or better. 9 The precise percentage of keratoconus patients who will eventually require surgery is difficult to determine since the total "at risk" population remains unknown. Studies addressing this issue have generally been biased by referral populations that are not representative of the keratoconic population as a whole. 2 ,4,5.7.9 In a 48-year epidemiologic study performed from 1935 to 1982 by Kennedy and co-workers, IO the 20-year cumulative survivorship without transplantation was greater than 80%. Surgical procedures other than penetrating keratoplasty (PK) have been described for the management of keratoconus,II-14 but their applications remain limited. To
1187
Ophthalmology
Volume 99, Number 8, August 1992
this day, PK remains the mainstay of surgical intervention for keratoconus 15- 18 with excellent optical and graft survivorship results. Very few studies have examiQed the management ofa large number of keratoconus patients to specifically analyze the criteria for surgery. To that end, we retrospectively analyzed the records of 99 keratoconic eyes that underwent PK after contact lens management failure. Our goals were to: (I) describe patient and ocular variables characteristic of those with contact lens failure despite initial successful fits, (2) classify the underlying causes for surgical intervention, and (3) relate those causes to patterns of preoperative patient factors that may prove predictive of the ultimate cause for contact lens failure.
Materials and Methods The records of 99 keratoconic eyes of 75 patients who underwent PK in the 10 years between 1981 and 1990 at the University of Illinois Eye and Ear Infirmary were selected for study. Only those patients who had been referred to our Cornea or Contact Lens Services, and who were followed and managed intensively for contact lens (re)fitting before surgery were included. In all patients, the diagnoses of keratoconus made on classic clinical findings of conical stromal thinning, Fleischer ring, subepithelial fibrosis, Vogt's striae, and irregular mires on keratometry were further substantiated by postoperative pathologic confirmation. Exclusion criteria included a history of ocular surgery or significant eye trauma. Furthermore, all those who at the baseline examination were deemed to have corneas that precluded successful contact lens fitting were excluded from study. One investigator (TTM) was responsible for the fitting of all patients for the duration of the 10 years of the study, so that the guidelines followed were internally consistent. Patients were fitted from an extensive inventory oflenses including standard design spherical rigid polymethylmethacrylate and gas-permeable (RGP) lenses (Boston II, Paraperm 02, Polycon II), spherical and toric hydrophilic lenses, and special design (Soper bispheric, bitoric, and "piggyback") lenses. In general, for mild disease, spherical rigid lenses were used, and these were fit with the base curve approximating the flattest keratometry reading, adjusting central and peripheral curvatures to achieve centration and adequate tear exchange by slit-lamp biomicroscopic evaluation of the fluorescein pattern. Except for special circumstances, patients fit with rigid lenses were given RGP, rather than polymethylmethacrylate, lens materials. In select cases when rigid lenses were not well tolerated, hydrophilic lens fittings were attempted. In advanced disease, base curve and peripheral curve designs were altered to allow a "threepoint touch" fit characterized by minimal apical touch and two areas of (mid)peripheral touch with equal distribution of fluorescein to weightbearing areas. 1,3,7 Finally, among those with mild to moderate disease, we generally aimed to have sufficient (superior lens) lid support to increase lens stability and to counter the propensity of in-
1188
ferior lens lift-off. In advanced disease, with corneal curvatures exceeding 52 diopters (D), special design lenses were frequently needed to achieve optimal centration, visual acuity, and comfort. A systematic review of the records by one investigator (MRD) was aimed to catalog the following parameters: (I) demographics, (2) source and reason of referral, (3) pertinent medical history including history of atopic disease, (4) pertinent ophthalmic history including previous PK (in the contralateral eye), previous contact lens wear and optical correction, (5) severity of disease parameters including baseline and preoperative best-corrected visual acuity, average keratometry and cylinder readings at presentation, and duration of disease if known, and (6) measure of effort in lens fitting including the number and types of lens fittings attempted and the duration of followup from baseline examination PK. Attention was subsequently focused to determine the reason(s) why cases had failed contact lenses. First, problem areas such as poor acuity, contact lens-related discomfort, frequent lens decentration, or peripheral thinning were identified. Second, four categories were constructed to establish the leading cause oflens failure: ( I) those with a best-corrected visual acuity of less than 20/40 despite a good contact lens fit, and who subjectively felt disabled by such vision to a degree that they opted for PK; (2) those who were severely contact lens intolerant, and therefore had inadequate functional wearing times, usually of less than 8 hours a day; (3) those with recurrent lens displacement and decentration despite numerous lens trials; and (4) those with advanced disease who showed progressive and marked peripheral corneal thinning, which necessitated grafting. Statistical methods for analysis included the chi-square test (with Yates' correction) for association between discrete variables. Fisher's exact test (two-tailed) was applied when the sample size was not adequate for the chi-square test. Student's t test was used to compare univariate means. Multivariate analysis was performed using a linear regression model relating the timing of surgery to multiple baseline variables.
Results Patient Characteristics The records of 99 keratoconic eyes of 75 patients (52 males, 23 females) who met our inclusion criteria were studied. The age of subjects at the time of presentation to our care ranged from 15 to 75 years (average, 34 years). The patients' average known duration of disease (from time of initial diagnosis until their presentation to us) was 8.4 years, with no significant difference between the sexes. Twenty-one percent of patients had a documented history of atopic disease (allergic rhinitis, asthma, or chronic urticaria) and 18 percent had already undergone PK in the contralateral eye by the time of initial presentation. More than 90% of those studied represented a referral population, with two thirds referred by other ophthal-
Dana et al . Contact Lens Failure in Keratoconus Management mologists (Table I). The reasons for referral were distributed equally between the optometrists and ophthalmologists, with 75% referred primarily for general evaluation and contact lens management.
Ocular Variables Optical Correction Worn. Results of independent evaluation of the 99 keratoconic eyes showed that 85% wore Table 1. Selective Patient and Ocular Characteristics Referring Source Ophthalmologist Optometrist Self Other
,.....
Average Keratometry at Baseline (D) Under 45
45-49 50-54 55-59 60-64 65 or more Indeterminant
Indeterminant/Unrecorded
67% 25% 7% 1% 3% 11% 15% 31%
24% 14% 2% 47% 25% 10%
7% 11%
Optical Correction at Baseline None Spectacles Contact lenses Hard (PMMA) Standard rigid gas-permeable Soft/hydrophilic Unspecified/other (special design)
75% 55% 14% 11% 20%
Ultimate Contact Lens Design Attempted Preoperatively Special design Standard rigid gas-permeable Hard (PMMA) Softjhydrophilic
74% 12% 9% 5%
Number of Lens Trials· per Eye before PK 1 2 3 to 5 6 to 10 More than 10
30% 23% 27% 13% 7%
15% 10%
o = diopters; PMMA = polymethylmethacrylate; PK = penetrating keratoplasty.
• Trials performed after referral to our services.
Baseline
Preoperative
10% 24% 24% 26% 56/0 11%
1% 22% 26% 27% 13% 11%
100%
100%
20/20 to 20/25 20/30 to 20/40 20/50 to 20/60 20/70 to 20/100 20/200 20/400 or less Total
Average Corneal Cylinder at Baseline (D)
1-5 6-10 11-15 16 or more
Table 2. Best-corrected* Baseline and Preoperative Visual Acuity
• Includes contact lens correction.
some type of correction at baseline; 75% were managed with contact lenses (Table I). By the time PK was contemplated, 86% of cases were being managed by standard RGP or special design lenses. Mean Keratometry, Cylinder, and Contact Lens Fittings. The mean keratometry reading at the time of baseline evaluations was 57.5 D, and the mean diopters of cylinder was 7 (Table I). Patients underwent as few as I, and as many as 15 contact lens fittings (mean, 3.8 fits) before keratoplasty, often involving multiple lens designs. Visual Acuity. More than one third (34%) of examined eyes had a best-corrected visual acuity of 20/40 or better at baseline, in contrast to 23% preoperatively (Table 2).
Factors Associated with Causes of Contact Lens Failure For all 99 eyes counted as a whole, the chief cause of lens failure was unacceptably poor vision for 43% (95% confidence interval [CI], 36% to 52%), intractable contact lens intolerance for 32% (95% CI, 25% to 39%), frequent lens displacement for 13% (95% CI, 9% to 20%), and significant peripheral thinning for 12% (95% CI, 7% to 19%). No significant differences existed between the two sexes. Among the 16 patients with a history of atopy, contact lens intolerance was the chief indication for surgical intervention. However, the sample size was too small for the rates to reach significance (P > 0.5). Ten of the 24 (42%) bilaterally operated cases had failed contact lenses because of poor vision in one eye and lens intolerance in the fellow eye. Analysis of data from bilaterally operated eyes during the 10 years of study showed no significant association between causes of lens failure from one eye to the other, validating independent analysis of each eye. Baseline Visual Acuity. The chief cause of failure among those eyes with visual acuity of 20/40 or better at the time of initial presentation was contact lens intolerance (Table 3), in contrast to those who failed lens management due to disablingly poor vision or peripheral thinning (Fisher's exact test, P < 0.05). Average Keratometry. Analysis of mean keratometry readings at baseline (Table 3) showed that more than half of those eyes with mean keratometry readings ofless than
1189
Ophthalmology
Volume 99, Number 8, August 1992
Table 3. Selected Ocular Variables Based on Final Outcomes* by Percent
Baseline acuity 20/20-40 20/50-100 ;:5;20/200 Average keratometry, diopeters 5 Timing of PK in months ;:5;12 >12 PK
=
Poor Vision
Lens Intolerance
Lens Displacement
Thinning
Total
30 50 50
40 30 25
24 6 12.5
6 14 12.5
100 100 100
31 49
52 24
14 13
3 14
100 100
49 20
27 55
13
11
15
10
100 100
52 37
21 40
10 16
17 7
100 100
penetrating keratoplasty .
• See Materials and Methods section for definitions.
55 D became intractably lens intolerant, while less than a quarter of those with higher keratometry readings went on to PK for the same reason (Fisher's exact test, P < 0.01). Conversely, those who underwent PK for poor vision or peripheral thinning tended to have keratometry readings of 55 D or higher. Contact Lens Fitting Trials. There was an inverse relationship between the number of lens fittings and the likelihood of undergoing PK for unacceptably poor vision (Table 3). Conversely, more than five contact lens fittings were associated with lens management failure because of eventual intractable lens intolerance (chi-square test, P < 0.05). Timing of Surgery. The average periods of follow-up from baseline to PK varied widely depending on the cause oflens failure (Table 3), ranging from 16 months for those
Table
4.
Selected Patient Characteristics Based on Final Outcomes* by Percent
Age (yrs) 10 Average
Poor \7ision
Lens Intolerant
43 42 40 50 35 yrs
43 38 37 18 31 yrs
37 53 50 9 yrs
39 7 18 7 yrs
• See Materials and Methods section for definitions.
1190
undergoing PK for thinning to 36 months for those who had surgery because of frequent lens displacement. Significantly, half of those who had surgery in the first year of follow-up represented the low vision group with only a quarter representing those with contact lens intolerance; the relative contributions of the two factors reversing among those who had surgery after 13 months of followup (chi-square test, P < 0.05). Age. The average age for those undergoing PK for thinning (41 years) was the highest compared with that of others having surgery for other reasons (Table 4). There was a significant difference between the odds of undergoing surgery for poor vision or thinning among those older than 40 years of age (71 %) compared with the same odds among patients younger than 40 years (47%) (chisquare, P < 0.05).
Lens Displacement
Thinning
Total
32 yrs
0 10 16 21 41 yrs
100 100 100 100
12 33 14 8 yrs
12 7 18 12 yrs
100 100 100
14 10 17 11
Dana et al . Contact Lens Failure in Keratoconus Management Duration of Disease. The mean known duration of disease before presentation was highest for those undergoing PK for thinning (12 years) compared with those who had surgery for other reasons (Table 4). Expectedly, duration of disease and preoperative age showed a significant positive association, while there was an inverse relationship between disease duration and baseline acuity. Patient sex, source and reason for referral, prior history of contact lens wear or ofPK in the fellow eye, and amount of cylinder present at baseline evaluation showed no significant association with the failure patterns of contact lens management.
Factors Associated witllTiming of Penetrating Keratoplasty The mean follow-up period between the baseline evaluation and PK was 27 months (males, 26 months; females, 29 months). Baseline Visual Acuity. Visual acuity at baseline showed a significant relationship with the length of successful contact lens fit. The average period of follow-up until PK for those with visual acuity of 20/40 or better was 43 (95% CI, 30 to 56) compared with 19 (95% CI, 13 to 24) months for those with poorer vision. Average Keratometry. There was a statistically significant inverse relationship between baseline keratometry readings and the length of time until surgery. Eyes with an average keratometry reading of less than 55 D had a ;mean period to surgery of 42 (95% CI, 29 to 54) months compared with those with keratometry readings of more than 55 D, who had a mean period to surgery of21 (95% CI, 14 to 27) months. Duration of Disease. Known disease duration and the timing of surgery showed an inverse relationship, but this difference did not reach statistical significance (P > 0.5). Application of a linear regression model incorporating mUltiple variables showed the independent retention of statistical significance by average keratometry (P < 0.05) and baseline visual acuity (P < 0.01), with no gain in predictability of surgical need by consideration of age, disease duration, or cylinder. Referral source, as well as a prior history of contact lens wear or of PK in the fellow eye, also failed to demonstrate significant associations with the timing of surgery.
Discussion Advances in lens design have made contact lenses applicable to all but the most advanced, or intractably lens intolerant cases. Various studies2,5,7 have conclusively shown that a majority of those cases with even relatively advanced disease, labeled "contact lens intolerant" according to standard ophthalmic care, can be successfully refitted with contact lenses. Our 10-year series represented a group with advanced disease, 90% representing a referral population. Stringent inclusion criteria assured that only those referral patients followed for intensive contact lens fitting before surgery
by the same practitioners (TTM and JS), and who eventually underwent PK according to the guidelines of one of our investigators (JS), were included in the study. Unlike many retrospective studies suffering from problems of variable outcome measurement and follow-up time, our study's major strength is its internal consistency in outcome definition, along with the fact that surgery itself was the end point in all cases, making the variable followup time not a weakness but simply another parameter to be measured. Our results strongly suggest that even among those with advanced keratoconus, lens intolerance does not represent the chief cause for the cessation of contact lens management. This parallels the findings of Kastl and co-workers9 who, in a 20-year retrospective study, found that poor vision was the primary reason for undergoing PK among the 4.8% minority of cases that failed contact lenses. It differs from the 5-year experience of Smiddy and coworkers7 who found that a majority of their cases that were originally fitted with contact lenses but who went on to PK (n = 51 eyes) did so because oflens intolerance, despite very comparable baseline average keratometries of 56.8 D in their "pre-keratoplasty" group compared with 57.5 D among ours. In the largest series of its kind, a 4-year multicenter retrospective study of the clinical management of keratoconic eyes,2 74% did not require surgery, 21 % underwent PK, and the remainder underwent epikeratoplasty. The investigators concluded that a best-corrected acuity ofless than 20/40, average keratometry of more than 55 D, and a history of contact lens wear were significantly associated with a PK outcome. After an average follow-up period of 16 months, 34 eyes (28 patients) in their series originally fitted with contact lenses became contact lens failures53% because of lens intolerance, 24% because of poor vision, and 12% because of problems with lens centration and handling. Our series of 99 eyes that were originally successfully fitted with contact lenses but which eventually underwent PK is approximately three times the size of the comparable group in the multicenter study cited above,2 with an average follow-up period from presentation to surgery (27 months) almost twice as long. Interestingly, the relative frequencies ofa surprising number of variables including age, sex, atopic disease, referral pattern, duration of disease, baseline visual acuity, and average keratometry show remarkable similarities between the two studies. As expected, and similar to the findings of Lass and co-workers, 2 we found low baseline visual acuity and steep corneal curvature to be predictors of (early) surgical intervention. Our findings also are similar to those of Lass and colleagues in failing to demonstrate a significant predictive role for sex, atopy, or corneal cylinder. Corneal cylinder seems to be an unreliable predictor of outcome, not because irregular corneal curvature is unimportant in the disease pathophysiology, but because of the inherent limitations of conventional (central) keratometry in delineating corneal topographic changes. Our experience indicates a significant association between a baseline best-corrected visual acuity of less than
1191
Ophthalmology
Volume 99, Number 8, August 1992
20/40, an average keratometry of more than 55 D, and contact lens failure due to a poor functional acuity despite a good fit. It is noteworthy that keratometry remains a poor indicator of the eventual cause of lens failure. This is so because whereas the probability of lens intolerance increases with increasing corneal curvature, its relative contribution to lens failure decreases, thereby making it the outcome most weakly associated with higher keratometry readings-in contrast to peripheral thinning or low acuity. Our results indicate that those patients with a longer disease duration (more than 5 years) are more likely to undergo PK for poor vision or thinning compared with lens intolerance. Second, a longer disease duration is predictive of an earlier surgical outcome. This is not surprising, as those with a longer duration of keratoconus represent a subset of patients with more advanced disease who present with poorer vision and steeper corneas. Interestingly, a prior history of PK (in the fellow eye) at baseline fails to demonstrate any significant association with disease duration, baseline acuity, ultimate cause of contact lens failure, or timing of surgery. Although a history of PK in one eye could signify a more longstanding and advanced disease state, and possibly decrease the threshold for referral (bringing the case to attention earlier), surgery in one eye also could significantly ameliorate functional vision to a point where the seeking of surgical attention for the unoperated eye is delayed. We conclude that, based on our experience with the contact lens management of advanced keratoconic eyes, they may be classified into two groups. The first group includes older patients with a longer duration of disease and high baseline keratometry readings who tend to have a lower visual acuity at presentation. These patients, for the most part, undergo fewer lens fittings and have surgery (PK) early, usually for low functional vision or for peripheral thinning. The second group includes younger patients with a more recently diagnosed condition who have lower baseline keratometry readings and who tend to have a better visual acuity at presentation. The atopic cases tend to cluster in this group. These patients undergo multiple lens fittings and tend to eventually have surgery many months after their presentation, usually failing their contact lens management secondary to intractable lens intolerance. These constructs are compatible with the known natural history of keratoconus, in which increasing duration of disease (associated with a rise in age) leads to higher corneal curvatures with a concomitant decrease in acuity, all combining to increase the odds of contact lens management failure. The encouraging experience with contact lenses in the management of even advanced keratoconic disease bodes. well for patients, who can expect to maintain good vision (and thereby delay or avoid surgery) with an intensive effort at contact lens fitting. The indications for surgical intervention are few, but a small albeit significant proportion of cases with advanced disease may reach that end point. To our knowledge, our study represents the first exclusive and systematic analysis of that group of
1192
keratoconic cases who eventually fail contact lens management and undergo PK. By elaborating on the patterns of contact lens failure in keratoconus management and outlining the risk factors for the various failure outcomes, we hope to have provided significant natural history data on which realistic clinical goals and expectations may be constructed.
References 1. Belin MW. Optical and surgical correction of keratoconus. Focal Points 1988: Clinical Modules for Ophthalmologists. San Francisco: American Academy of Ophthalmology, 1988; vol. VI, module 11. 2. Lass JH, Lembach RG, Park SB, et aI. Clinical management of keratoconus. A multicenter analysis. Ophthalmology 1990;97:433-45. 3. Krachmer JH, Feder RS, Belin MW. Keratoconus and related noninflammatory corneal thinning disorders. Surv Ophthalmol 1984;28:293-322. 4. Lawless M, Coster DJ, Phillips AJ, Loane M. Keratoconus: diagnosis and management. Aust NZ J Ophthalmol 1989; 17: 33-60. 5. Belin MW, Fowler WC, Chambers WA. Keratoconus: evaluation of recent trends in the surgical and nonsurgical correction of keratoconus. Ophthalmology 1988;95:335-9. 6. Buxton IN, Keates RH, Hoefle FB. The contact lens correction of keratoconus. In: Dabezies OH Jr, ed. Contact Lenses: The CLAO Guide to Basic Science and Clinical Practice. Vol. 2. Orlando: Grone & Stratton, 1984; chap. 55. 7. Smiddy WE, Hamburg TR, Kracher GP, Stark WJ. Keratoconus: contact lens or keratoplasty? Ophthalmology 1988;95:487-92. 8. Carney LG. Contact lens correction of visual loss in keratoconus. Acta Ophthalmol 1982;60:795-802. 9. Kastl PR, Donzis PB, Cole HP III, et al. A 20-year retrospective study of the use of contact lenses in keratoconus. CLAO J 1987;13:102-4. 10. Kennedy RH, Bourne WM, Dyer JA. A 48-year clinical and epidemiologic study of keratoconus. Am J Ophthalmol 1986;101:267-73. 11. Arentsen JJ, Laibson PRo Thermokeratoplasty for keratoconus. Am J Ophthalmol 1976;82:447-9. 12. Wood TO. Lamellar transplants in keratoconus. Am J Ophthalmol 1977;83:543-5. 13. McDonald MB, Kaufman HE, Durrie DS, et al. Epikeratophakia for keratoconus. The nationwide study. Arch Ophthalmol 1986; 104: 1294-1300. 14. McDonald MB, Safir A, Waring GO III, et aI. A preliminary comparative study of epikeratophakia or penetrating keratoplasty for keratoconus [letter]. Am J Ophthalmol 1987;103:467. 15.· Payne JW. Primary penetrating keratoplasty for keratoconus: a long-term followup. Cornea 1982;1:21-7. 16. Paglen PG, Fine M, Abbott RL, Webster RG Jr. The prognosis for keratoplasty for keratoconus. Ophthalmology 1982;89:651-4. 17. Troutman RC, Lawless MA. Penetrating keratoplasty for keratoconus. Cornea 1987;6:298-305. 18. Steinert RF, Wagoner MD. Long-term comparison of epikeratoplasty and penetrating keratoplasty for keratoconus. Arch Ophthalmol 1988;106:493-6.
