Human cataract formation

Ciba Foundation symposium 106

1984

Pitman London

Human cataract formation

The Ciba Foundation is an international scientific and educational charity. It was established in 1947 by the Swiss chemical and pharmaceutical company of CIBA Limited-now CIBA-GEIGY Limited. The Foundation operates independently in London under English trust law.

The Ciba Foundation exists to promote international cooperation in biological, medical and chemical research. It organizes about eight international multidisciplinary symposia each year on topics that seem ready for discussion by a small group of research workers. The papers and discussions are published in the Ciba Foundation symposium series. The Foundation also holds many shorter meetings (not published), organized by the Foundation itself or by outside scientific organizations. The staff always welcome suggestions for future meetings.

The Foundation’s house at 41 Portland Place, London, WIN 4BN, provides facilities for all the meetings. Its library, open seven days a week to any graduate in science or medicine, also provides information on scientific meetings throughout the world and answers general enquiries on biomedical and chemical subjects. Scientists from any part of the world may stay in the house during working visits to London.

Human cataract formation

Ciba Foundation symposium 106

1984

Pitman London

0 Ciba Foundation 1984

ISBN 0 272 79175 8

Published in July 1984 by Pitman Publishing Ltd., 128 Long Acre, London WC2E 9AN, UK. Distributed in North America by CIBA Pharmaceutical Company (Medical Education Division), Post Office Box 12832, Newark, NJ 07101, USA

Suggested series entry for library catalogues: Ciba Foundation symposia

Ciba Foundation symposium 106 x + 284 pages, 90 figures, 15 tables

British Library cataloguing in publication data: Symposium on Human Cataract Formation.

(1983: Ciba Foundation. London) Human cataract formation.-(Ciba Foundation symposium; 106) 1. Cataract I. Title 11. Nugent, Jonathan 111. Whelan, Julie IV. Series 617.1'42 RE45 1

Printed in Great Britain at The Pitman Press, Bath

Contents

Symposium on Human Cataract Formation, held at the Ciba Foundation, London, 25-27 October 1983

Editors: Jonathan Nugent (Organizer) and Julie Whelan

A. Spector Introduction 1

L. T. Chylack, Jr Classification of human cataractous change by the American Cooperative Cataract Research Group method 3 Discussion 17

R. M. Clayton, J. Cuthbert, J. Seth, C. I. Phillips, R. S. Bartholomew and J. McK. Reid Epidemiological and other studies in the assessment of factors contributing to cataractogenesis 25 Discussion 40

A. Spector Oxidation and cataract 48 Discussion 55

V. N. Reddy and F. J . Giblin Metabolism and function of glutathione in the lens 65 Discussion 83

R. F. Borkman Cataracts and photochemical damage in the lens 88 Discussion 99

P. F. Kador and J. H. Kinoshita Diabetic and galactosaemic cataracts 110 Discussion 123

G. Duncan and T. J. C. Jacob Calcium and the physiology of cataract 132 Discussion 148

General discussion Causes and mechanisms of human cataract formation 153

vi CONTENTS

H. Maisel Cytoskeletal proteins of the ageing human lens 163 Discussion 174

H. Bloemendal, G. A. M. Berbers, W. W. de Jong, F. C. S. Ramaekers, A. J. Interaction of crystallins with M. Vermorken, I. Dunia and E. L. Benedetti

the cytoskeletal-plasma membrane complex of the bovine lens Discussion 186

177

J. Piatigorsky, J. M. Nickerson,C. R. King, G. Inana, J. F. Hejtmancik, J. W. Hawkins, T. Borras, T. Shinohara, G. Wistow and B. Norman Crystallin genes: templates for lens transparency Discussion 204

191

J. G. G. Schoenmakers, J. T. den Dunnen, R. J. M. Moormann, R. Jongbloed, R. W. van Leen and N. H. Lubsen The crystallin gene families 208

L. Summers, C. Slingsby, H. White, M. Narebor, D. Moss, L. Miller, D. Mahadevan, P. F. Lindley, H. Driessen, T. L. Blundell, J. T. den Dunnen, R. J. M. Moormann, R. W. van Leen and J. G. G. Schoenmakers The molecular structures and interactions of bovine and human 7- crystallins 219 Discussion 230

G. B. Benedek The molecular basis of cataract formation 237 Discussion 240

W. H. Garner The application of non-invasive techniques to the study of cataract development on the metabolic and the protein molecular level 248 Discussion 259

Final general discussion Sequence of events in cataract formation 266 Strategies for future research on the human lens 271

Index of contributors 275

Subject index 277

Participants

G. B. Benedek Department of Physics, Room 13-2005, Massachusetts Insti- tute of Technology, Cambridge, Massachusetts 02139, USA

H. Bloemendal Department of Biochemistry, University of Nijmegen, Postbus 9101, Geert Grooteplein noord 21. 6500 HB Nijmegen. The Netherlands

T. L. Blundell Laboratory of Molecular Biology, Department of Crystal- lography, Birkbeck College, University of London, Malet Street, London WClE 7HX, UK

R. F. Borkman School of Chemistry, Georgia Institute of Technology, Atlanta, Georgia 30332, USA

L. T. Chylack, Jr Howe Laboratory of Ophthalmology, Harvard Medical School, Massachusetts Eye and Ear Infirmary, 243 Charles Street, Boston, Massachusetts 02114. USA

R. M. Clayton Department of Genetics, University of Edinburgh, Institute of Animal Genetics, King’s Buildings, West Mains Road, Edinburgh EH9 3JN, UK

M. J. C. Crabbe Nuffield Laboratory of Ophthalmology, University of Oxford, Walton Street, Oxford OX2 6AW, UK

G. Duncan School of Biological Sciences, University of East Anglia, Norwich NR4 7TJ, UK

W. H. Garner Department of Ophthalmology, College of Physicians & Surgeons of Columbia University, 630 West 168th Street, New York, NY 10032, USA

F. J. Giblin Institute of Biological Sciences, Oakland University, Rochester, Michigan 48063, USA

vii

... PARTICIPANTS Vl l l

K. Gupta (Ciba Foundation Bursar) Ocular Biochemistry Section, Dr R . P. Centre for Ophthalmic Sciences, All-India Institute of Medical Sciences, Ansari Nagar, New Delhi-110029, India

J. J. Harding Nuffield Laboratory of Ophthalmology. University of Oxford, Walton Street, Oxford OX2 6AW, UK

0. Hockwin Institute of Experimental Ophthalmology, University of Bonn, Sigmund Freud Str. 25,5300 Bonn 1, Federal Republic of Germany

J. Horwitz Department of Ophthalmology/Biophysics, Jules Stein Eye In- stitute, UCLA School of Medicine, Center for the Health Sciences, Los Angeles, California 90024, USA

P. F. Kador Laboratory of Vision Research, Bldg 6, Room 237, National Eye Institute, National Institutes of Health, Bethesda, Maryland 20205, USA

S. Lerman Department of Ophthalmology, Emory University Medical School, Laboratory for Ophthalmic Research, 1708 Haygood Drive NE, Atlanta, Georgia 30322, USA

H. Maisel Department of Anatomy. Wayne State University, School of Medicine, 540 East Canfield Avenue, Detroit, Michigan 48201, USA

G. Maraini Institute of Ophthalmology, University of Parma. Via A. Gram- sci 14,43100 Parma, Italy

D. C. Minassian International Centre for Eye Health, Institute of Ophthalmology, 27 Cayton Street, London EClV 9EJ, UK

C. A. Paterson Departments of Ophthalmology and Physiology, University of Colorado Health Sciences Center, Box B205, 4200 East Ninth Avenue, Denver, Colorado 80262, USA

B. T. Philipson Department of Ophthalmology, Karolinska Institute, Karolinska Hospital, 104 01 Stockholm, Sweden

J. Piatigorsky Laboratory of Molecular & Developmental Biology, Bldg 6, Room 201, National Eye Institute, National Institutes of Health, Bethesda, Maryland 20205, USA

V. N. Reddy Institute of Biological Sciences, Oakland University, Rochester, Michigan 48063, USA

PARTICIPANTS IX

J. G. G. Schoenmakers Laboratory of Molecular Biology, Catholic Univer- sity of Nijmegen, Toernooiveld. 6525 ED Nijmegen. The Netherlands

C. Slingsby Department of Crystallography, Birkbeck College, University of London, Malet Street, London WClE 7HX, UK

A. Spector (Chairman) Biochemistry and Molecular Biology Laboratory, Department of Ophthalmology, College of Physicians & Surgeons of Columbia University, 630 West 168th Street, New York, NY 10032, USA

R. A. Weale Department of Visual Science. Institute of Ophthalmology, Judd Street, London WClH 9QS, UK

This Page Intentionally Left Blank

Introduction

A. SPECTOR

Biochemistry and Molecular Biology Laboratory, Department of Ophthalmology, College of Physicians & Surgeons of Columbia University, 630 West 1681h Slreet, New York, NY 10032, U S A

I984 Human cataract formalion. Pitman, London (Ciba Foundation symposium 106) p 1-2

It is ten years since the first Ciba Foundation symposium on the human lens and cataract (Ciba Foundation 1973). At that time, little work had been done on the human lens. Most studies had utilized bovine, rat and rabbit material. General aspects of metabolism, protein chemistry and morphology had been deline- ated. An understanding of diabetic cataract was emerging, and some concepts had been cautiously advanced to explain aspects of cataract development.

The Ciba Foundation symposium in 1973 changed the emphasis, changed the perception of the field. A number of laboratories turned to the direct study of the human lens and human cataract. Some scientists who had studied the animal lens for a considerable period now began to investigate the human lens for the first time. Stimulated by the Ciba Foundation meeting and the discus- sions on the classification of cataract, American cataract researchers joined together to collaborate on the study of human lenses. A system of classifying human lenses was developed by Dr Leo Chylack under this programme, and an effective computer network for analysing, storing and sharing data was de- veloped. Cooperative programmes also emerged in Japan and in Europe, to some extent stimulated and encouraged by the experience of the American Cooperative Cataract Research Group.

The success of the CCRG classification system has led to the tendency to relate the type of opacity to the initiating cause. It should be remembered, however, that opacification is the final step in cataract formation. For example, in animal models of sugar-induced cataract, the sudden opacification of the inner region of the lens is far removed from the initiating site. Caution must be exercised in presuming that an opaque region of the tissue can reveal the initiating events that have led to its present status, or even suggest where in the lens, or elsewhere in the organism, the process began.

Considerable progress has been made in the past ten years. It has also become evident that a simple extrapolation of the animal data to the human lens is not possible. There are a number of reasons for this. Long-term ageing changes have been found which had not been anticipated from studies with

1

2 SPECTOR

relatively short-lived mammals. Aspects of human lens biology, such as growth, hydration, relative enzyme concentrations and changes in protein structure, are found to be different. Furthermore, we have been confronted with a diversity of cataracts, at least in terms of their morphological features.

We are now at another cross-roads. New surgical procedures for extracting cataracts, which result in the destruction of the tissue, are becoming popular. This development lends an urgency to the need to discover and improve non-invasive techniques for studying the lens in the human as well as in animal models. Recently, light-scattering techniques have been refined and applied at both the molecular and cellular levels. New techniques, such as nuclear magne- tic resonance and fluorescence spectroscopy, are gradually emerging as power- ful non-invasive analytical tools. We shall become increasingly dependent on such techniques, and also on areas of the world where intracapsular cataract extraction remains popular. The non-invasive methods will also have a central role in confirming the observations obtained in the test-tube and in culture systems with intact lenses or with epithelial cell preparations. They should provide important insights into the effectiveness of new drugs in retarding or preventing cataracts.

There is another aspect to the present cross-roads confronting lens re- searchers. A sufficient body of information is becoming available for a critical evaluation to be made of mechanisms in the development of cataract. Indeed, while solutions elude us, hypotheses abound. There now are a host of concepts to explain cataract, including osmotic imbalance related to the accumulation of sugar alcohols via aldose reductase; oxidative insult; the development of high molecular weight protein aggregates and the phase separation of protein com- ponents in the cytoplasm; calcium imbalance; genetic defects; alterations in the cytoskeleton; insufficiency of glucose causing metabolic imbalance and insta- bility of certain enzymes; subliminal factors associated with ageing, diet, and drug intake; and many others. It is time to begin an evaluation of these concepts with the information now available, and to question their validity. It is probable that some of them are related and are part of a general route of opacification. It is important to design experiments and model systems for the purpose of testing our concepts of cataract formation, to allow us to reject or strengthen our present ideas.

We shall consider some of these problems at this meeting. If we are success- ful, perhaps this symposium will give a new direction to cataract research as did its predecessor, a decade ago.

REFERENCE

Ciba Foundation 1973 The human lens-in relation to cataract. Associated Scientific Publishers, Amsterdam (Ciba Found Symp 19)

Classification of human cataractous change by the American Cooperative Cataract Research Group method

LEO T. CHYLACK. JR.

Howe Laboratory of Ophthalmology, 243 Charles Slreet, Boston, Massachusetts 02114, U S A

Abstract. The American Cooperative Cataract Research Group (CCRG) has adopted a system of classifying cataractous changes in excised human cataracts that is based on separ- ate and independent photographic documentation of opacification and nuclear colour. The classification data can be simplified according to the needs of the scientist in his or her effort to measure clinically or scientifically significant associations between laboratory measures and cataractous change. The association between nuclear colour and the extent of cortical opacification and the intensity of nuclear opacification has been studied and found to be insignificant. These results justify the recommendation that nuclear colour be abandoned as the single index of severity of any type of senile cataractous change.

The adaptability of this system to in vivo use in epidemiological and other studies of the natural history of cataractous change is discussed. Its limitations are outlined. The system may offer a basis for international cooperation in cataract classification.

1984 Human cataract formailon Pitman. London (Ciba Foundation symposium 106) p 3-24

In 1976, at the annual meeting of the Association for Research in Vision and Ophthalmology in the US, an informal meeting of lens scientists discussed the feasibility of increasing the emphasis on human, as opposed to animal, cataract research. There was considerable enthusiasm for this idea, and shortly there- after the Cooperative Cararact Research Group (CCRG) was formed. In 1980, funding as a consortium of 23 laboratories was obtained from the National Eye Institute. Initially the major objectives of the consortium were:(l) to increase the supply of human cataracts available to the lens scientist, so that studies of biochemistry, physiology, anatomy and so on could proceed in paral- lel with animal experimentation under way in the laboratory; (2) to develop a sound method of classifying cataractous changes in the human lens; and (3) to develop a method of exchanging information among CCRG scientists. In the

3

4 CHYLACK

initial grant application, a section on in vivo studies was included but was later removed, because it was judged to be premature. The system of cataract classification is presented in this symposium in its present state of development, but it is undergoing continuous improvement. It is appropriate that the system be discussed here, since the symposium aims to evaluate critically various hypotheses of cataract formation; the CCRG classification system may provide an infrastructure for just such an evaluation. I shall discuss the suitability of this system for use in vivo and mention alternative or supplementary non-invasive techniques for gaining information about the mechanisms of cataract forma- tion. I hope to convince you that the CCRG classification system is scientifically sound, and to persuade you to adopt it as the system of classification of cata- racts in laboratories in Europe as well as in the US.

Background

For decades clinicians have described cataractous changes with a series of internationally understood and accepted terms (subcapsular, anterior, equa- torial and posterior cortical, supranuclear and nuclear opacification). Other terms implying aetiological significance (such as radiation cataract, steroid cataract, or traumatic cataract) have been used, but I do not regard this clinical effort to be legitimate ‘classification’. I have chosen to define classification as the description and organized grouping of the discrete features of lens opacities. The clinician derives no significant benefit from the rigorous applica- tion of an ‘organized’ system of terminology. In response to the unsuitability of the clinical terminology, Pirie (1968) suggested organizing insoluble lens proteins according to the intensity of colour (from yellow to black) of the nuclei from which the protein was derived. Others (Duncan & Bushell 1979, Mar- cantonio et a1 1980, Duncan 1981) have proposed similar or somewhat amplified classification systems. Each of these systems uses either observation or photography of the cataract against a white background and assumes that the intensity of nuclear colour, on a spectrum of pale yellow through to yellow- brown and black, is a measure of the severity of cataract formation. None of these scientists has experimentally tested the validity of this assumption. Other weaknesses in some of the European proposals for cataract classification are the use of cataracts which have been frozen or stored in salt solution. We suspected that such treatment would obscure or at least drastically alter the appearance of the cataract.

In planning the design of the CCRG classification system, I stressed the importance of:

(1) Photographing cataracts against a black background with white fluor-

CLASSIFICATION OF HUMAN CATARACT 5

escent light. We found these cataractous changes to be invisible when the lens was viewed against a white background.

(2) Obtaining a separate photograph of the lens against a white background, so as to estimate nuclear colour.

( 3 ) Photographing the lens immediately after extraction from the globe; to facilitate this, the technicians and the photographic apparatus were present in the operating rooms of the Massachusetts Eye and Ear Infirmary.

(4) Describing semi-quantitatively the extent of subcapsular, cortical and supranuclear opacification with measurable variables, and the relative in- tensity of nuclear opacification and colour with ranked or ordinal variables.

( 5 ) Creating a permanent set of stereoscopic 35 mm colour transparencies of cataracts that can be used in classifying cataracts and can be recalled and reviewed if revision of the system of classification is indicated.

(6) Facilitating the simplification of the raw classification data. The system has been described at each stage of its evaluation (Chylack 1978,

Chylack et al 1983, 1984b). The chosen terminology puts all cataracts into one of three groups, as follows:

Hypermature (H): a totally opaque lens that has undergone marked swelling in the antero-posterior dimension.

Mature (M): a totally opaque lens in which no recognizable normal anatomical zone remains, but in which there is no appreciable antero-posterior swelling.

Immature (I): an opacity which does not totally obscure all normal anatomical regions of the lens.

All immature cataracts (I) possess some amount of normal lens anatomy, and the cataractous changes are grouped into the following zones (see Fig. 1): CXA (anterior cortex), CXE (equatorial cortex), CXP (posterior cortex), SCA (subcapsular anterior), SCP (subcapsular posterior). SN (supranuclear)

FIG. 1. Diagram of the principal anatomical zones of the human lens as used in the CCRG classification scheme. SCA. SCP: subcapsular, anterior and posterior respectively: CXA. CXE, CXP: anterior, equatorial and posterior cortical zones respectively; SN: supranuclear; N: nuclear. Left view: sagittal; right view: frontal with anterior surface up.

6 CHYLACK

and N (nuclear). Nuclear colour is designated by NS (nuclear sclerosis). Subcapsular opacification is graded by relating the opaque zone to a series of concentric circles (F’ig. 2). The outer circle represents the equatorial circle of the lens. CXA, CXE, CXP and SN opacifications are graded by the cumula- tive extent of quadrantic involvement (subscripts 1-4) occupied by the opacity. Nuclear cataract (N) is graded by the density, as judged by the degree (1-4) to which a slit-beam image passing through the lens is obscured. NS is graded by the colour on an eight-step spectrum from clear to very pale yellow, pale

% of total area of circle

FIG. 2. An aid to the estimation of the area involved in a subcapsular cataract. A set of concentric circles, the largest of which represents the equatorial circle of the lens. The smaller circles are designated with a number representing the percentage of the area of the equatorial circle occupied by the designated circle. The classifier uses these circles to estimate the approximate size of an SCA or SCP opacity in a human lens.

yellow, yellow, dark yellow, very dark yellow, brown and black. The classifier uses six stereoscopic views of the lens and records the classification on cards containing identification information and clinical data as well as the classifi- cation codes. A raw classification might appear, simply, as ‘M’, or be as complex as ‘I SCA, CXA2 N, NS4’ or ‘I SN CXE2 CXPz N1 NS8’.

‘M’ is a totally opaque lens in which the colour of the nucleus cannot be graded accurately because of the light scattering properties of the perinuclear cortex.

‘I SCA3 CXA2 N, NS,’ is an immature cataract in which normal and opacified zones are seen. The ‘SCA,’ term indicates that the approximate area of the anterior subcapsular opacity is 3% of the area of the equatorial circle of the lens (Fig. 2). The ‘CXA2’ term indicates that approximately two quadrants, or 50%

CLASSIFICATION OF HUMAN CATARACT 7

of the anterior cortex (Fig. l), is occupied by opacity. The opacity need not be exactly quadrantic in shape but should occupy 50% of the CXA zone. The ‘N4‘ term describes a nuclear opacity through which a slit beam cannot pass. This is evident in a slit-view of the cataract against a black background in which the beam does not reach and, therefore, cannot be reflected from the black surface underlying the lens. The ‘NS4’ term indicates that the colour of the nucleus against a white background is yellow.

The ‘I SN, CXEz CXPzN, NSx’ classification denotes an immature cataract in which three quadrants of the supranuclear zone ‘SN?’ are opaque. If the SN zones contain clear and opaque regions, the cumulative extent of the opacity is used. Likewise ‘CXE?’ and ‘CXP:’ indicate two quadrants-worth of equatorial and posterior cortical opacity respectively. The ‘N,’ term describes a faint nuclear opacity, one in which a slit beam passes through the nucleus to the black surface beneath the lens and appears slightly blurred. If the blur was moderate, the term would be ‘N2’; if the blur was marked, but not total, the term would be ‘N?’. The ‘NS,‘ term indicates the colour of the nucleus to be black or blackish brown against a white background.

When the CCRG Consortium began functioning, I was the only classifier; all lens photos were sent to me, classified and returned to the scientist of origin. I made an attempt to test the consistency with which I was classifying lenses. A subset of 82 lenses, representing the same proportion of simple and compound cataracts as in the total population, was classified by me on two occasions, 48 hours apart. The sequence of cataracts was changed at the second effort, and the results of the first effort were not available the second time. The two sets of classification data were compared and errors analysed. The results (Chylack et al 1983) demonstrated a high degree of consistency in my classification, with the following exceptions:

(1) It was difficult to distinguish consistently between anterior subcapsular opacities (SCA) and anterior cortical opacities (CXA), and between the pos- terior subcapsular and posterior cortical opacities (SCP and CXP). Since this distinction is likely not to be important. i t was not regarded as a major weakness.

(2) When nuclear colour (NS) was estimated, there were 20/73 cases in which the NS subscript differed by one grade; 9/82 lenses were H or M, and in these, NS cannot be classified accurately. In no case did the estimate of NS differ by more than one grade. In using the data on nuclear colour it therefore seemed best to condense the eight steps in the NS index to four.

With increased acceptance of the CCRG classification system it became possible to test the ability of other scientists to classify cataractous change consistently. With the help of Dr Bernard Rosner, a protocol was developed in which three examiners from outside Boston came to Boston for two days of intensive instruction in the technique of classification. Each examiner classified

8 CHYLACK

225 cataracts overt hree days and the same 25 cataracts each day for three days. The 25 were interspersed among the other 75 so as not to be identifiable by the examiner from1 one day to the next. Three examiners from within the Howe Laboratory, including myself, will repeat the exercise with the same lenses and the results will be analysed for intra-observer and inter-observer consistency.

One of the first studies of cataracts classified by the CCRG protocol has examined the association of nuclear colour (sclerosis) with the extent of cata- ract formation, age and visual acuity. In this study, 1976 immature human cataracts extracted mtracapsularly were classified and data on cataract location and extent, nuclear colour, preoperative visual acuity, age and sex were organized and stored in the PROPHET system. The data were examined for relationships between nuclear colour and the age of the cataractous lens, and the extent of opacification in seven anatomical regions, including the degree of nuclear opacification and the preoperative visual acuity.

We found that nuclear colour correlates with age in a curvilinear manner. Nuclear yellowing increases gradually with increasing nuclear opacification, but the colour change is so slight as to be useless for the purposes of deducing the intensity of nuclear opacification from the colour of the nucleus. There was no association between the extent of anterior cortical, equatorial cortical, posterior cortical, subcapsular or supranuclear opacification and nuclear colour. Nuclear colour impairs vision only for the range from dark yellow through to black.

These results jusl ify the recommendation that nuclear colour be abandoned as the single index of the severity of any type of senile cataractous change (Chylack et a1 198413).

This study examined the significance of associations among data gathered during the classification exercise; no laboratory analyses were needed to complete the study. Unsimplified classification data were used. However, as one attempts to study associations between cataract class and laboratory measures, it becomes apparent that the needs of one scientist may differ radically from those of another.

For the anatomist dealing with specific regional abnormalities, the raw classification, in spite of its complexity, is adequate for a study of the associa- tion between anatomical change and cataract classification (Harding et a1 1980, 1982). For the biochemist, simplification is usually necessary for a study of the association between a biochemical parameter and cataract class; there are too few lenses per class at the most complex stage of the classification for mean- ingful statistical analyses to be done. While the need for a simplification of the CCRG classification system was apparent at the outset, the means of simplify- ing it were slow to appear. A simplification scheme has been devised and published (Chylack. et a1 1984a). It condenses the subdivisions of anatomical

CLASSIFICATION OF HUMAN CATARACT 9

zones in a complex classification, thereby reducing the number of terms used to describe each cataract. For example:

SCA and SCP sc CXA, CXE and CXP CXAEP SC and CXAEP CXS and SN > CX (cortical) CX and N 9 CXN (corticonuclear) H and I CAT (cataract)

CXS (superficial cortical)

The staging of the simplification is shown in Fig. 3. It is clear that at Stage I , the large number of unordered combinations (> 1000) precludes study of associa- tions among cataract classes and laboratory measures; there are simply too few lenses in any one class to achieve statistical significance, much less any scienti- fically or clinically important differences. As one uses the simpler classifications (Stages 11-IX) and the number of possible unordered combinations of classifi- cation terms decreases, the number of lenses per class increases and mean- ingful associations can be measured. Unavoidably, one trades away the chance to correlate a biochemical measure with a specific zone of opacification when one gains the chance to make any correlation at all. This is to some degree a disadvantage to the biochemist.

A new and potentially superior method of simplifying cataract classification data is now being tested in our laboratory. In essence, it adds 1-3 retro-illumi- nation photographs of the cataract to the standard series of photos.

With these new views one has photographs of the opacity in reflected and transmitted light, each of which may have a different appearance and be more or less useful to the biophysicist or biochemist studying cataract formation. The cataract is placed within a 9.25 mm hole in an opaque black disc immersed in normal saline solution in the standard CCRG classification apparatus; a black and white 35 mm transparency is taken with the camera focused on the prin- cipal opacity (or opacities, if anterior, equatorial and posterior planes are involved). In the same opaque disc is mounted a grey-scale step-wedge which is also incorporated into each photograph. The photographs are then digitized, subjected to a gamma correction and analysed by computerized image analysis in a manner similar to that described for in vivo cataract classification later in this chapter.

The image of the cataract captures all the features of both nuclear and cortical opacification. Because of the limited depth of focus of the photo- graphic system, not all zones of the lens will be in focus in a single photo. Thus, three photos may be necessary to capture focused views of all parts of a complex lens opacity.

The great advantage of this technique is its ability to express photographic data (the appearance of a cataract) in quantitative numerical terms (depth of

Scheme for Simplifvina Raw CCRG Classification Data c 0

Classification Terms Nuclear Mature- No of Sclerosis Hypermature

Staqe Definition or Maneuver Combinations Immature Cataracts Index Cataracts

I

ll

m

m

Y

m

m

mn

Ix

Terms with subscriots

Drop subscripts

Condense fSCA+SCP)to S(

Condense (CXA+CXE-tCXF to CXAEP

Condense all superficial cortical opacities ( S C X X A E P ) = CXS

Condense all cortical opacities (CXS+SN)=CA

Condense all rmmoture cataracts (CX*N)=I

Condense mature and hypermoture cataracts (M+H) = HM

Condense all HM ond I cataracts (I+HM) =CAT

.I000

127

65

I5

7

3

3

2

I -

cY I

NSI4

NS

ti5

NS

NS

NS

NS

M H

M H

M ii

M H

M H

M H

ChT

FIG. 3. Scheme for simplifying the raw (Stage I) CCRG classification data. Definitions of H , M , I , SCA, SCP, CXA, CXE. CXP, SN, N and NS can he found in the text. HM is acomhined class of all hypermature (H) and mature (M) lenses. CAT is a combined class in which thc morphological characteristics of the cataract are not considered.

CLASSIFICATION OF HUMAN CATARACT 11

grey colour for each pixel, and the total number of pixels per each grey zone) for each opacity. A single, albeit complex cataract, can thereby be charac- terized by a histogram of the number of pixels for each of 256 grey zones. Such a histogram can be characterized by a mean, median, standard deviation about a mean, etc., and these data can easily be compared to data derived from labora- tory analyses of whole or discrete parts of a lens.

In studies of how a cataract degrades the image passing through the lens, it may prove to be essential to have a view, and measure, of the cataract from the retina‘s ‘point of view’. A classification photo with light transmitted through the lens might be such a view.

Recently, significant progress has been made by Horwitz and Takemoto (Takemoto et al 1982) in microdissecting opaque and clear zones from human cataracts. This approach offers an alternative to the previously described systems of whole-lens classification; it allows the clear zones of the cortex or the supranuclear zone to serve as the control for the opaque counterpart zones. While technically difficult, this is such a direct solution to the problem as to be far more appealing than opting for a harvest of several hundred lenses needed for studying the same association. Some studies require whole lenses and for these the only alternative to a decreasing supply of intracapsularly extracted cataracts is an augmented supply of eye bank, tissue bank or autopsy lenses. In the US, few eye banks are able to supply scientists with large numbers of lenses; however, the National Diabetes Research Interchange (NDRI) is a new tissue bank with a remarkably successful record for the distribution of diabetic tissues to researchers in the US.

So far, I have dealt with classification of extracted human cataracts. The CCRG system can also be applied in vivo. A comparison of the preoperative and postoperative features of the cataract has been made. Except for an identifiable ‘cryo’ mark, the appearance preoperatively was the same as post- operatively. The system has been used in my practice to record preoperative features of cataracts and nuclear colour. The principal limitation to the use of this system in vivo is pupil size which, if less than 9.0mm on maximal dilata- tion, will conceal the equatorial and perhaps the supranuclear features of the opacity. This is the only major limitation to the CCRG system’s use in vivo.

Although feasible, in vivo application of this classification system may not always be desirable. For example, if a study demands cataract classification no more precise than is possible with the CCRG system (i.e. epidemiological studies of the incidence and prevalence of certain types of cataract) then, in vivo, the CCRG system suffices. However, if greater precision is needed, such as might be required in measuring the subtle impact of an anticataract drug on a cataract’s growth or the toxic potential of an environmental or dietary substance, a more precise system will be needed. In fact, many toxicological studies and therapeutic trials require precision and sensitivity great enough

12 CHYLACK

FIG. 4. A Topcon SL-45 (Schiempflug) photograph of a human cataract. Top left: axis of slit beam relative to the eye (90" = vertical slit) and area for patient identification information. Centre left: grey-scale step-wedge for gamma correction or standardization of scanning microdissection. Centre right: in focus photograph of cornea, anterior chamber and entire slit view of lens.

to measure very small changes in the extent of cataracts or changes in the precataractous stages of lens growth.

For high precision in dealing with the extent of cataracts, computerized image analysis of standardized lens photographs may suffice. We have been experimenting with such a system. High quality, artifact-free, slit and retro- illumination photographs can be obtained with the Topcon SL-45 (Fig. 4) (Hockwin et a1 1978) and the Zeiss photo-slit-lamp fitted with Kawara optics (crossed polarizing and orange filters in the slit-lamp path) (Kawara & Obaz- awa 1980) (Fig. 5). Nuclear cataracts have polar symmetry and, therefore, a single slit-lamp photograph suffices as a representation of the whole opacity.

CLASSIFICATION OF HUMAN CATARACT

FIG. 5. A retro-illumination photograph of a human cortical cataract taken with a Zeiss photo-slit- lamp fitted with Kawara-Obazaua optics.

Cortical, supranuclear and subcapsular cataracts have no polar symmetry. and neither one nor several slit images would suffice as a true representation of the cataract. Retro-illumination photographs are capable of capturing the features of all cortical cataracts with 1-3 photographs, depending on the level (anterior, equatorial or posterior) of the opacity. I t is possible to digitize these photo- graphic images (Fig. 6), mask out the areas to be ignored (the cortex in the SL-45 photos and the iris in the retro-illumination photos), and analyse the residual image in terms of the number of pixels per each of 256 levels of grey (on a 256-step spectrum from white to black) (Fig. 7). The data can be expressed as the mean grey level, with the standard deviation about the mean, or the top or bottom 5% of pixels, or in many other ways. It is possible to measure cataractous changes over a period of time with this system and. per- haps, to detect a favourable effect of a potential anticataract drug or an adverse effect of a lens toxin. At present we are measuring the variance (sources of error and variability) of such a system used by different photographers, different film developers, and different image analysers.

14 CHYLACK

FIG. 6 . A digitized image of Fig. 5 with a circular area outlined within the pupil within which an image analysis for the grey-scale density of each pixel and the number of pixels per grey step can be done.

Considerable progress has been made in developing non-invasive techniques of measuring pre-cataractous lens changes. Some of the most promising of these are:

(1) Quasielastic light scattering (Tanaka & Benedek 1975, Tanaka & Ishimoto 1977, Weiss et a1 1982).

(2) Tyndall photometry (Blumental et a1 1977). (3) Specular microscopy (Bron & Matsuda 1981). Later in this symposium, Dr Garner will present some of the exciting

developments in the field of nuclear magnetic resonance spectroscopy, ultra- violet fluorescence and Raman spectroscopy of the lens in situ (see p 248-259). Such techniques provide biochemical data as a function of time without des- troying the tissue, a13 is done at present with techniques yielding the same type of data.

In conclusion, I blzlieve that in the past ten years the lens research commun- ity has experienced major evolutionary change. No longer are we confining our

CLASSIFICATION OF HUMAN CATARACT

FIG. 7. The same digitized image as in Fig. 6, except that the area outside the circular mask has been eliminated, and upon which has been superimposed a histogram of the number of pixels at each of 256 grades of grey.

experiments to animal lenses maintained in tissue culture; we have begun to examine human cataractogenesis in vitro and in vivo and we have even developed non-invasive technologies for the study of pre-cataractous changes in vivo. The CCRG cataract classification system is merely a device designed to facilitate this transition. I believe it is scientifically sound, and will be useful to

16 CHYLACK

scientists faced withi the task of correlating laboratory and/or clinical measure- ments with the type and/or the extent of cataractous change.

Ackno wledgernents

I should like to acknowledge the support of the Brigham Surgical Group Foundation, the National Eye Institute (Research Grants EY01276 and EY03247), the surgeons and nurses of the operating rooms at the Massachusfetts Eye and Ear Infirmary and the Brigham and Women's Hospital, and the National Diabetes Research Interchange (NDRI).

REFERENCES

Bron AJ, Matsuda K 1981 Specular microscopy of the human lens. Trans Ophthalmol SOC UK

Blumental M, Bodenheimer J , Krushelensky A , Rothkoff L 1977 New Tyndall photometer. Arch

Chylack LT Jr 1978 Classification of human cataracts. Arch Ophthalmol96:888-892 Chylack LT Jr, Lee MR., Tung WH, Cheng H M 1983 Classification of human senile cataractous

change by the American Cooperative Cataract Research Group (CCRG) method. I. Instru- mentation and technique. Invest Ophthalmol Vis Sci 24:424-431

Chylack LT J r , White 0, Neff RK, Tung WH 1984a Classification of human senile cataractous change by the American Cooperative Cataract Research Group (CCRG) method. 11. Staged simplification. Invest Ophthalmol Vis Sci 25:166-173

Chylack LT Jr, Ransil B.J, White 0 1984b Classification of human senile cataractous change by the American Cooperative Cataract Research Group (CCRG) method. 111. The association of nuclear color (sclerosis) with extent of cataract formation, age and visual acuity. Invest Oph- thalmol Vis Sci 25:171-180

Duncan G 1981 O n classifying human cataractous lenses. In: Duncan G (ed) Mechanisms of cataract formation in the human lens. Academic Press, London, p 1-5

Duncan G , Bushell A R '1979 Relationships between colour, sodium and protein content in individ- ual senile human cataractous lenses. Ophthalmic Res 11:397-404

Harding CV, Chylack LT Jr, Susan SR, Decker JG, Lo WK 1980 Morphological changes in cataract: the ultrastructure of human lens opacities, localized by CCRG procedures. In: Srivas- tava SK (ed) Red blood cell and lens metabolism. Elsevier North-Holland Biomedical Press, Amsterdam, p 27

Harding CV, Chylack LT Jr, Susan SR, Lo WK, Bobrowski WF 1982 Elemental and ultrastruc- tural analysis of specific human lens opacities. Invest Ophthalmol Vis Sci 23:l-13

Hockwin 0, Dragomirescu V, Koch H R , Sasaki K 1978 Followup method for the documentation of lens opacities with, a new photographic equipment (TOPCON lens densitograph). Abstr XXIII Int Congr Ophthalmol, Kyoto, Japan. Excerpta Medica, Amsterdam. ICS no. 442, p 182 (Exhibit)

Kawara T, Obazawa H 11980 A new method for retroillumination photography of cataractous lens opacities. Am J Ophthalmol90:186-189

Marcantonio JM, Duncan G, Davies PD, Bushell A R 1980 Classification of human senile cataracts by nuclear colour and sodium content. Exp Eye Res 31:227-237

101:163-169

Ophthalmol95 : 323-3;!4

CLASSIFICATION OF HUMAN CATARACT 17

Pirie A 1968 Colour and solubility of the proteins of human cataracts. Invest Ophthalmol7:634-650 Takemoto LJ. Hansen JS, Horwitz J 1982 Biochemical analysis of microdissected sections from the

Tanaka T, Benedek G B 1975 Observation of protein diffusivity in intact human, bovine lenses with

Tanaka T, Ishimoto C 1977 In-vivo observation of protein diffusivity in rabbit lenses. Invest

Weiss JN, Nishio I , Clark JI , Tanaka T. Benedek G B , Giblin FJ , Reddy VN 1982 Early detection

normal and cataractous human lens. Curr Eye Res 2:443-450

application to cataract. Invest Ophthalmol 14:449-456

Ophthalmol 16: 135-140

of cataractogenesis by laser light scattering. Invest Ophthalmol Vis Sci (ARVO suppl) p 41

DISCUSSION

Spector: How satisfactory is this classification scheme? Another scheme was previously developed by Dr Antoinette Pirie and modified by Dr Duncan and others (Pirie 1968, Marcantonio et a1 1980). I would like to hear Dr Duncan’s views and to establish whether we can arrive at a uniform classification system.

Duncan: My first comment on Leo Chylack’s paper would be that I know of nobody, in Europe or elsewhere, who now uses colour as the single means of classifying lenses. In the classification system that we proposed and tested (Marcantonio et a1 1980) we used both in vivo and in vitro photography to assess not only lens colour, but also nuclear (group 11) and cortical (group I) scatter. We also suggested that lens sodium concentration should be used as a measure of the extent of osmotic disruption of the lens.

Spector: But should we be using colour at all in classifying lenses? From Dr Chylack’s results, nuclear colour is apparently not of much value in classifying most types of cataract.

Chyfack: Nuclear colouration is a separate change from lens opacification; the development of colour and of opacity probably entails two different proces- ses. Therefore the use of colour as a measure of opacification is incorrect. Nuclear colour certainly should be recorded, because several features of lens ageing are related to colour development.

I agree that sodium estimation is useful in a retrospective classification of lenses but, since it involves destroying the lens, it is not a useful measure for classifying lenses in vitro.

Duncan: The cortical changes in cataract are indeed complex, and we agree that they are adequately described by the CCRG system. Examples of the application of this system to European lenses can be seen in Fig. 6 of Duncan & Jacob (this volume). I do not agree, however, that measuring lens sodium necessarily involves destroying the lens. Since 23Na has a strong NMR signal (Reisen et a1 1970), it should be possible to measure changes in human lens

18 DISCUSSION

sodium content in vivo. We shall also describe an in vitro, minimal perturbation method for measuring sodium activity using very finely tipped glass microelec- trodes (Duncan & Jacob, this volume).

Weule: What worries me about the CCRG system is that it leaves the patient hanging in mid-air. Patients are not complaining about coloured lenses; if they were, we would have a cataract extraction every time we wear sun-glasses! The patient becomes c’oncerned when his ability to see detail is impaired. No classification system takes cognizance of this. Admittedly, we are concerned with visual acuity, but this can be impaired for more than one reason. The retina may be defective when there is a perfect lens. If one relies just on visual acuity, a wrong coriclusion may be reached. There are, of course, a variety of objective methods (of obtaining the data that Dr Chylack has presented. If one relies on a person’s judgement and he is colour blind, a wrong result will be obtained. Colour and transmission curves can be measured photo-electrically. One may ask: how much light can be transmitted through a cataract? If the cataract is very dense the answer will be, not much; but the lens can be sliced and the transmission that gets through half a millimetre, for example , can be measured.

But I am concerned even more about the patient: he complains only when he cannot see detail, and the loss of detail can be quantified very accurately by modern image-forming techniques (Weale 1983). We have used a modification of these gratings (Fig. 1). This target can be photographed through a lens and, depending on the quality of the lens, one can resolve either the top line or the next line, or the next line, and so on. If one wants to analyse a cataract one can slice the lens until some image is formed by the apparatus. In this manner, even ‘opacities’ can be described quantitatively.

Chy1uck:The CCRG system is semi-quantitative. I recognize the limitations of this system, which is why I emphasized a shift from using a subjective description of an image viewed by reflected light to the computerized analysis of an image of the cataract taken with transmitted light. This will give us a

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 l l l 1 l l l l l l l l l l l l l l l l l l l l l l l l l l

9 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 9 7 7- I I I I I I I I I I I I I

-5 5- - - - - - - m

3 1 1 1 1 - I I I I I I 3 I - - - - - 1

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

IIIIIIIIII - - - - - - - - - - - - - FIG.1 (Wecille). Grating used to assess light transmission through lenses.