|
|
ORIGINAL ARTICLE |
|
Year : 1997 | Volume
: 45
| Issue : 1 | Page : 19-24 |
|
Cornea stress test--evaluation of corneal endothelial function in vivo by contact lens induced stress
JS Saini, S Mittal, M Anand
Dept. of Ophthalmology, Postgraduate Institute of Medical Education and Research, Chandigarh, India
Correspondence Address: J S Saini Dept. of Ophthalmology, Postgraduate Institute of Medical Education and Research, Chandigarh India
Source of Support: None, Conflict of Interest: None | Check |
PMID: 9475007
Reliable and valid assessment of corneal endothelial function is a critical input for diagnosing, prognosticating and monitoring progression of disorders affecting corneal endothelium. In 123 eyes, corneal endothelial function was assessed employing data from the corneal hydration recovery dynamics. Serial pachometric readings were recorded on Haag-Striet pachometer with Mishima-Hedbys modification before and after two hours of thick soft contact lens wear. Percentage Recovery Per Hour (PRPH) was derived from raw data as an index of endothelial function. Assessed PRPH in pseudophakic corneal oedema and Fuchs' endothelial dystrophy eyes (35.9 +/- 9.8%) was significantly lower than normal controls (61.9 +/- 10.5%). On employing receiver operation characteristics curve analysis the tested results demonstrated high sensitivity (87%) and specificity (92%) for detection of low endothelial function at PRPH cut off of 47.5%. Using this PRPH cut off, 80% of Fuchs' endothelial dystrophy and 93.3% of pseudophakic corneal oedema eyes could be demonstrated to have low endothelial function. A total of 66.7% of diabetic eyes also demonstrated PRPH of lower than 47.5%. Clear corneal grafts demonstrated PRPH values of 24.6% to 73.0%. Of 6 corneal grafts that demonstrated initial PRPH of lower than 47.5%, 4 failed within 4 to 6 months. Our data demonstrated high sensitivity and specificity of this corneal stress test. PRPH index was useful in quantifying endothelial function in clinical disorders including diabetes mellitus. The index PRPH was demonstrated to be useful in monitoring and prognosticating outcome of corneal grafts. Keywords: Corneal endothelium, Corneal hydration, Pseudophakic corneal oedema, Fuchs′ dystrophy, Diabetes Mellitus, Diabetic retinopathy, Corneal grafts
How to cite this article: Saini J S, Mittal S, Anand M. Cornea stress test--evaluation of corneal endothelial function in vivo by contact lens induced stress. Indian J Ophthalmol 1997;45:19-24 |
In-vivo reliable and valid assessment of functional status of corneal endothelium will assist the clinician in monitoring the progression of diseases affecting corneal endothelium, choosing the appropriate intervention and predicting the likelihood of corneal decompensation. The quantification of corneal endothelial function can also help in identifying grafts at higher risk of failure[1]. Optical pachometry provides a measure of corneal hydration which indicates the state of endothelial function but it is often not affected in the early stages of disease[2]. Fluorimetric analysis of corneal endothelium is hampered by the large sources of inherent error and potential toxicity of fluorescein[3]. Morphological evaluation of cornea by clinical biomicroscopic findings and specular microscopy has not been confirmed to adequately reflect the corneal endothelial function and paradoxical results have been obtained in some studies[4]. In vivo corneal endothelial function assessment by regulating corneal hydration has been shown to be affected in old age[5], Fuchs' dystrophy[6], contact lens wear[7] and diabetes mellitus[8]. Cornea stress test assumes that observed differences in the dynamics of recovery from induced corneal swelling represents primarily the difference in the contribution by endothelium to hydration control. We assessed the corneal endothelial function in eyes of subjects with pseudophakic corneal oedema and Fuchs' dystrophy confirmed biomicroscopically, and normals to validate the test. We also evaluated endothelial function in subjects with diabetes mellitus and in clear corneal grafts where there were no biomicroscopic abnormalities. The above data was compared with normal volunteers with an objective to assess clinical usefulness of this corneal stress test procedure.
Materials and methods | | |
One hundred and twenty three eyes of 123 subjects were evaluated in this study. Among those included were 15 eyes with pseudophakic corneal oedema, 15 with Fuchs' endothelial dystrophy, 23 with clear transplants, 45 eyes of diabetic patients and 25 eyes of normal subjects. The disease state was clinically diagnosed and was independently confirmed by two different trained observers with 100% agreement. The criteria for the clinical diagnosis of Fuchs' dystrophy included the presence of central corneal guttata, epithelial microcysts, Descemet's wrinkles and stromal haze. Diabetes mellitus was diagnosed on the basis of fasting blood sugar level of more than 140 mg% and post prandial blood sugar level of more than 200 mg%. The normal controls were recruited from staff at the hospital and their relatives. We recruited a sample stratified by age. The affected eye or at random either eye (when both eyes were equal) was recruited for the study. Before participation, informed consent was obtained from each subject and tenets of declaration of Helnsinki were followed. The study protocol was approved by Institutional Ethics and Review Committee. Subjects with glaucoma, prior contact lens wear, ocular trauma and intraocular surgery (except for the specified surgery in groups) were excluded.
Assessment of corneal endothelial function was done independently by the same observer for all the eyes. The observer was masked for the clinical status of the subject.
Corneal Pachometry | | |
Optical pachometer (Haag Striet, Bern, Switzerland) equipped with small light emitting diodes to ensure reproducibility, eye fixation and correct alignment was employed[10]. Each corneal thickness measurement at a certain time point was obtained as two sets of ten readings, carried out within four minutes. Each set of ten readings was arranged and at every time point two means of ten readings were obtained. Between each reading the instrument was refocussed and realigned and between each set of ten replicate measurements the subject's head was retracted and then repositioned in the instrument. For each enrolled eye, a measure of baseline open eye steady state corneal thickness was obtained by doing optical pachometry after the subjects were awake for 6 to 8 hours. This was done to avoid possible thickness changes after eye closure during the night. In each eye, at least ten replicate thickness measurements were performed during two hours to establish baseline open eye steady state corneal thickness.
Cornea Stress Test | | |
Corneal swelling was induced by applying a soft stress contact lens and then patching the eye for two hours. A contact lens of low water content(38%) having a central thickness of 0.3 mm and a diameter of 13.8 mm with oxygen transmissibility (DK/L) of 14x10-9 (cm2/sec) (ml O2/ml x mm Hg) was used. The lenses were available in the posterior curve radii of 8.5, 8.7, 8.9 and 9.1 mm. Before applying the patch each lens was checked for its centration.
The lenses were removed after two hours and the deswelling phase was recorded by obtaining corneal thickness measurements immediately and at every 30 minute interval for three hours.
Analysis of Data | | |
An exponential curve was obtained for each set of recovery data during the deswelling phase using the least squares analysis method. The representative pattern of deswelling response in three groups is depicted in [Figure - 1]. The deswelling response can be described by
The values of x, y and z were obtained using non linear regression procedures and least squares analysis. The values of x and y are directly interpretable but z can be interpreted easily when it is converted to alternative forms. One of the forms is Percentage Recovery Per Hour (PRPH) which is the deswelling that occurs in any one hour interval when expressed as percentage of the swelling that exists at the start of interval and ranges from 0 to 100%.
Time taken for 95% recovery (T95%) is the time required by the cornea to recover back to 95% of its baseline open eye steady state corneal thickness. This can be computed from 'z' by
Baseline open eye steady state thickness is defined as the thickness of cornea after being awake for atleast six hours and is obtained from multiple pachometry readings before applying stress test lens. Mean of ten replicate measurements was taken to arrive at the final baseline open eye steady state thickness.
Percentage swelling induced is the swelling caused by stress test when expressed in terms of baseline open eye steady state thickness. This is given as
Comparisons among the eyes with different clinical corneal endothelial status and normal eyes were made by performing an analysis of variance (F-test). For simultaneous multiple comparisons, Bonferroni modification of student 't' test was applied.
Results | | |
The parameters of corneal endothelial function are summarised in [Table - 1] and depicted on [Figure - 1]. The values of baseline open eye steady state corneal thickness (x), PRPH time taken for 95% recovery (T95%) are normally distributed. The baseline open eye steady state corneal thickness was significantly more in eyes with pseudosphakic corneal oedema (0.77 ± 0.12 mm) than normal eyes (0.52 ± 0.02 mm) ( p< 0.001). Assessed PRPH was significantly lower in eyes with pseudophakic corneal oedema (31.6 ± 9.6%) and Fuchs' endothelial dystrophy (40.2 ± 8.1%) as compared with normal controls (61.9 ± 10.5%) (p<0.001). Similarly T95% was significantly higher in eyes with pseudophakic corneal oedema and Fuchs' dystrophy as compared with normal controls (p<.0001). Eyes with diabetic retinopathy demonstrated significantly lower PRPH values when compared to normal eyes and those of diabetic subjects without retinopathy. (P<.001). Percentage swelling induced was not significantly different among the eyes evaluated. Eyes having lowest PRPH belonged to the pseudophakic corneal oedema and Fuchs' dystrophy groups, those with morderately low PRPH were clear cornea transplants and eyes of diabetes mellitus subjects.
Validity of test procedure indicates whether we are actually measuring what we are attempting to measure.[11] To test the validity of this corneal stress test procedure, we mixed the data of pseudophakic corneal oedema and Fuchs' dystrophy eyes with normal eyes. To determine the usefulness of PRPH in identifying eyes with normal and abnormal endothelium function we set various cut off values. By varying the cut off value of PRPH and determining the sensitivity and specificity values at each cut off PRPH, we generated a receiver operation characteristics curve[12] for PRPH [Figure - 2]. With a PRPH cut off measure 47.5%, both the specificity (87%) and sensitivity (92%) for detecting clinical dysfunction was high [Table - 2] which suggests reasonable validity of this test procedure.
On employing this PRPH cut off value of 47.5 % to eyes with clear corneal transplants and those of diabetic subjects, 53% of the eyes demonstrated low endothelial function whereas baseline open eye steady state corneal thickness (≥ 0.55) was abnormal in only 16% of the eyes [Figure - 3]. Even clinically normal appearance of endothelium could thus be demonstrated to have low endothelium function.
Assessed PRPH was useful in identifying endothelial dysfunction in eyes demonstrating clinically early changes and minimal stromal oedema. In eyes demonstrating baseline corneal thickness of more than 0.75 mm, all eyes demonstrate critically low PRPH (≤ 47.5%). However, when baseline corneal thickness was less than 0.55 mm, assessed PRPH ranged from 26.7% to 85.6% [Table - 3].
Discussion | | |
Many investigators have, in past, pursued the development of a non-invasive and reliable test procedure to evaluate the corneal endothelial function.[1][2][3][4][5][6][7][8],[15],[16] Baseline pachometry, fluorometry and quantitative specular microscopy have been employed but have been found inadequate.[2][3][4] A reliable assessment of corneal endothelial function in vivo will be helpful in monitoring and choosing appropriate management strategies in diseases of corneal endothelium. Recently, Polse and colleagues[5] have described a test protocol to measure the endothelial function which involves essentially inducing controlled corneal oedema by contact lens wear and subsequently measuring deswelling to arrive at PRPH as an index of endothelial function. The procedure has demonstrated low endothelial function in old age [5,] Fuchs' dystrophy,[6] and long term contact lens wearers.[7] Our observations demonstrate that PRPH is significantly low in pseudophakic corneal oedema and Fuchs' dystrophy. At PRPH cut off value of 47.5%, 93% of clinical pseudophakic corneal oedema and 80% of Fuchs' dystrophy eyes demonstrated quantifiable low endothelial function.
In clear corneal grafts and subjects of diabetes mellitus, there is substantially more variable distribution of PRPH suggesting that normal biomicroscopic assessment of endothelium is compatible with wide range of quantified endothelial function. Higher baseline corneal thickness is generally associated with lower PRPH. In eyes with baseline corneal thickness of more than 0.55mm, clinical biomicroscopy can identify abnormality almost always but when baseline pachometry was less than 0.55 mm, a wide range of PRPH was detected. Thus baseline corneal thickness is not as good an indicator of endothelial function as cornea stress test in early stages of endothelial dysfunction.
The test protocol may marginally affect the calculated PRPH. We generally followed the modified procedure described by Nieuwendaal and colleagues[7] which employs a pre test baseline open eye steady state corneal thickness instead of theoretical open eye steady state corneal thickness employed by Polse et al[5]. Pretest baseline thickness includes residual corneal thickness and tends to give higher recovery rates. The test protocol followed by us involved minimum six hours waking period before thickness measurements are done. The longer waking period theoretically helps to reduce overnight residual swelling more than the two hours waking period described by Nieuwendaal and colleagues[7]. Our PRPH values in normal control eyes are comparable to those described in similar age groups of Nieuwendaal et al[7] and higher than of Polse et al.[5],[6] The test protocol in our study can be completed in a shorter time period and is more convenient for patients than the longer time period with other protocols. As long as test procedures are administered in similar manner to age matched subjects in normal and diseased states, conclusions are valid. We suggest that each test protocol should obtain age specific values in normal eyes before attempting interpretation of test results in diseased states. The T95% is an arithmetical transformation of PRPH and gives no additional information. Our results demonstrate that T95% reflects the assessed PRPH values.
The quantum of swelling induced is limited by the quality of endothelium and proteoglycans in stroma. Decreased stromal proteoglycan and the consequent diminished capacity to retain fluid[15] in stroma can limit the swelling induced. In our study the swelling induced is not significantly different in the three groups. It is not a good indicator of endothelial function.
Corneal thickness, in past, has been used as an indicator of corneal endothelial function. Corneal thickness measurements in eyes at risk for graft rejection had been identified as a test of prognostic value[1]. In vivo assessment of corneal endothelial function by estimating PRPH is a better alternative and may prove to be of better prognostic value.
References | | |
1. | McDonnel PJ, Enger C, Stark WJ: Corneal thickness changes after high risk penetrating keratoplasty. Arch Ophthalmol 111:1374-1381, 1993. |
2. | Mishima S: Corneal thickness. Surv Ophthalmol 13:57-96, 1968. |
3. | Brubaker RF, Maurice DM, McLaren JW: Fluorometry of anterior segment. In: Non invasive diagnostic techniques in Ophthalmology. Barry R. Masters (ed), New York, Springer - Verlag Inc, 1990; pp 248-280. |
4. | Sato T: Studies on the endothelium of the corneal graft. Jpn J Ophthalmol 22:114-117, 1978. |
5. | Polse KA, Brand RJ, Mandell RB, et al. Age differences in corneal hydration control. Invest Ophthalmol Vis Sci 30:392-399, 1989. |
6. | Mandell RB, Polse KA, Brand RJ, et al: Corneal hydration control in Fuchs' dystrophy. Invest Ophthalmol Vis Sci 30:845-852, 1989. |
7. | Nieuwendaal CP, Odenthal MTP, Kok JHC, et al. Morphology and function of corneal endothelium after long term contact lens wear. Invest Ophthalmol Vis Sci 35:3071-3077, 1994. |
8. | Herse P, Hooker B: Corneal edema recovery dynamics in diabetes. Is the Alloxan induced diabetic rabbit a useful model ? Invest Ophthalmol Vis Sci 35:310-313, 1994. |
9. | Adamis AP, Filatov V, Tripathi BJ, Tripathi RC: Fuch's endothelial dystrophy of cornea. Surv Ophthalmol 38:149-168, 1993. |
10. | Mandell RB, Poise KA, Bonanno J: Reassessment of optical pachometry. In: H. Dwight Cavanagh (ed), The cornea: Transactions of the World Congress on the cornea - III, New York, Raven Press Ltd., 1988, pp 201-205. |
11. | Bourke GJ, Daly LE, McGilvray J: Interpretation and use of medical statistics (3rd ed), Oxford, Blackwell Scientific Publications, 1985, p. 238. |
12. | Saini JS and Khandalvala B: Corneal epithelial fragility in diabetes mellitus. Can J Ophthalmol 30:142-146, 1995. |
13. | Burns RR, Bourne WM, Brubaker RF: Endothelial function in patients with cornea guttata. Invest Ophthalmol Vis Sci 20:77-85, 1981. |
14. | Shaw EL, Rao GN, Arthur EJ, Aquavella JV: The functional reserve of corneal endothelium. Ophthalmology 85:640-649, 1978. |
15. | Kangas TA, Edelhauser HF, Twining SS, O'Brien WJ: Loss of stromal glycosaminoglycans during corneal edema. Invest Ophthalmol Vis Sci 31:1994-2002, 1990. |
16. | Schultz RO, Matsuda M, Yee RW, et al. Corneal endothelial changes in Type I and Type II diabetes mellitus. Am J Ophthalmol 98:401-410, 1984. |
[Figure - 1], [Figure - 2], [Figure - 3], [Figure - 4], [Figure - 5], [Figure - 6]
[Table - 1], [Table - 2], [Table - 3]
|