19

Cataract surgery on diabeticpatients. A prospective evaluationof risk factors and complicationsPer Flesner1, Birgit Sander1, Vibeke Henning1,Hans-Henrik Parving2, Morten Dornonville de la Cour3 andHenrik Lund-Andersen1

1Department of Ophthalmology, Herlev University Hospital, Copenhagen, Denmark2Steno Diabetes Centre, Gentofte, Denmark, 3Department of Ophthalmology,Rigshospitalet HS, Copenhagen, Denmark

ABSTRACT.Purpose: This study presents an evaluation of cataract surgery on diabetic pa-tients. One experienced surgeon carried out phaco emulsification on all subjectsand the same surface-coated one-piece PMMA-lens-type was implanted. Thelens fluorescence and the blood-aqueous barrier (BAB) were then evaluated asexperimental preoperative risk indicators.Results: During follow-up, 10 out of 39 diabetic patients progressed unilaterallyin diabetic retinopathy or developed macular oedema, a significant relative risk.Neither lens fluorescence, BAB, HbA1c, level of retinopathy, type/duration ofdiabetes, diabetes treatment or antihypertensive treatment differed significantlybetween the group of patients with postoperative progression of retinopathy/macular oedema and those without. Results indicated NIDDM (non-insulin-dependent diabetes mellitus/type 2 diabetes) patients might have increased riskof a postoperative macular oedema.Conclusion: When diabetic retinopathy (DR) is not in a proliferative phase itshould not be regarded as a contraindication to modern cataract surgery. Nei-ther lens fluorescence nor BAB is valuable as a risk indicator for postoperativeprogression of DR.

Key words: diabetes – cataract – lens fluorescence – blood-aqueous barrier (BAB) – retinopathy.

Acta Ophthalmol. Scand. 2002: 80: 19–24Copyright c Acta Ophthalmol Scand 2002. ISSN 1395-3907

Diabetic patients have an increasedrisk of developing cataract (Oimo-

mi et al. 1988; Sensi et al. 1995; Ramalhoet al. 1996). Because of the risk of post-operative progression of retinopathy anddevelopment of macular oedema re-ported by some studies (Dowler et al.1995; Pollack et al. 1992; Menchini et al.1993; Krupsky et al. 1997), cataractsurgery in diabetic patients used to beconsidered high-risk surgery. Neovascu-lar glaucoma and continuing retinal pro-

liferation have also been reported aftercataract surgery in diabetic patients (Sa-diq et al. 1995; Pollack et al. 1991; Schatzet al. 1994). These unintended postopera-tive events are associated with a poor vis-ual prognosis. Recent studies, often usingthe technique of phaco emulsification,have reported fewer postoperative com-plications and generally more encour-aging results (Henricsson et al. 1996;Wagner et al. 1996; Antcliff et al. 1997;Cunliffe et al. 1997). The progression of

diabetic retinopathy (DR) is associatedwith the degree of glycaemic control, dur-ation of diabetes and untreated hyperten-sion (Klein et al. (III) (1984); Klein et al.(IV) 1984; Parving et al. 1989; Sjølieet al. 1997; Stephenson et al. 1995).Rarely has the association between pro-gression of DR after cataract surgery andthe degree of glycaemic control beenmade (Henricsson et al. 1996).

This study evaluates the relative risk ofprogression of DR after cataract surgerywhen using small-incision phaco emulsi-fication technique and a surface coatedIOL. Patients are operated in one eyeonly and the fellow eye serves as a controlduring a follow-up period of 6months. Apreoperative, non-invasive measurementof the auto fluorescence of the lens isevaluated as a possible risk indicator forcataract surgery in diabetic patients. Theauto fluorescence of the human lens indiabetics has been demonstrated to de-pend upon age, duration of diabetes andthe degree of glycaemic control (HbA1c)(Bleeker et al. 1986; Kjer et al. 1987;Larsen et al. 1989; Koefoed Theil et al.1996; Sparrow et al. 1997).

Diabetic retinopathy and macularoedema are partly explained by a break-down of the blood-ocular barriers(blood-retina barrier, BRB). In the an-terior segment of the eye, the correspond-ing barrier is the blood-aqueous barrier,(BAB). BAB is also evaluated as a pre-operative risk indicator for progressionof DR after cataract surgery in diabeticpatients (Boot et al. 1989; Moriarty et al.1994; Schalnus & Ohrloff 1995).

20

Patients and methodsA total of 41 patients were enrolled intothis prospective study. Two patients werenot able to attend the full postoperativefollow-up period of 6months. One ofthese died 3months after cataract surgeryand the other had a retinal detachmenton the eye operated on 1month after cat-aract surgery. Thirty-nine patients com-pleted 6months follow-up. Twenty-ninepatients were NIDDM (non-insulin-de-pendent diabetes mellitus/type2 diabetes)patients, 12 IDDM (insulin-dependentdiabetes mellitus/type1 diabetes). Theclinical characteristics of the patients en-rolled are shown in Table1.

The type of diabetes was based uponthe patient’s age at the onset of diabetesand the type of antihyperglycaemic treat-ment. Patients whose diabetes began be-fore the age of 40 and who received insu-lin treatment were classified as belongingto the IDDM group. Patients whose dia-betes onset after the age of 40 and whowere treated only with diet or diet andoral anti hyperglycaemic treatment orwith insulin were classified within theNIDDM group.

Retinopathy was graded using theEURODIAB IDDM complications studygrading system (Aldington et al. 1995).Macular oedema was defined in accord-ance with standard early treatment dia-betes diabetic retinopathy study group(ETDRS) criteria as retinal thickening in-volving the centre of the macula; hardexudates within 500mm of the centre ofthe macula; an area of macular oedemagreater than one disc area but within onedisc diameter of the centre of the macula.

Table 1. Clinical characteristics of patients (N Ω 39) who fulfilled the complete follow-up period (6months). (mean (SD))

Mean (SD)

Patients (male/female) 39 (20/19)Age (year.) 70 (8.5)Duration of diabetes (year.) 14.8 (13.8)HbA1c (%) 8.3 (1.8)IDDM (%) 28NIDDM (%) 72Insulin treatment (%) 41Oral treatment (%) 44Diet alone (%) 15Antihypertensive treatment (%) 72S-creatinin (mmol/L) 91 (21.4)U-albumin (mmol/L) 5 (16.4)BAB (x 10ª6 minª1) 671 (629)Lens fluorescense (ng/mL) 552 (305)

Examination procedure

A full medical and ocular history wastaken before cataract surgery togetherwith blood and urine samples to deter-mine HbA1c, S-creatinin and U-albumin.A complete ocular examination was per-formed initially and at each follow-up ex-amination (at 1week, 1month, 3monthsand 6months). This included determi-nation of best-corrected visual acuity(VA) (ETDRS), tonometry, slit-lamp ex-amination and retinal biomicroscopy.

Before cataract surgery, lens autofluorescence was determined using a com-mercially available fluorophotometer,Fluorotron Master (OcuMetrics Inc., SanJose, CA, USA) The majority of eyes (33)were measured with a Fluorotron Masterusing laser light at 488nm as an excitationsource. The remaining patients were meas-ured with a broadband excitation from430nm. The same equipment was used todetermine BAB permeability. BAB wascalculated using software produced by vanBest et al. (van Best et al. 1996).

Fluorescein angiographies were per-formed at baseline and 6months post-operative.

Colour fundus photographs at anangle of 60æ were taken at all clinical ex-aminations with a Topcon fundus cam-era. The photographs were taken in ac-cordance with the EURODIAB gradingsystem, which includes a picture of themacular region and a disc-nasal pictureleading to placement on a 6-level retino-pathy scale ranging from no retinopathyto proliferative retinopathy.

The investigator and a trained nursegraded the photographs without access

to clinical patient data. When gradinglevels were divergent, a trained ophthal-mologist carried out the final grading (sixof the 41 patients referred to in the pres-ent material were graded this way).

Macular oedema was clinically evalu-ated by biomicroscopy by the investigatorand by fluorescein angiographies by atrained ophthalmologist without anyknowledge of clinical patient data.

Two patient groups are defined in thefollowing text.

The first of these includes patientswho postoperatively progress unilaterallyin retinopathy in the eye operated on, orwho progress at least two grading stepsmore in the eye operated on than in thenonoperated eye. It also includes patientswho postoperatively develop a unilateralmacular oedema. This group is referredto as the ‘complicated group’.

The second group includes patientswith no postoperative progression of reti-nopathy and no macular oedema. It alsoincludes patients with bilateral pro-gression and/or bilateral macular oede-ma. This group is called the ‘non-compli-cated’ group.

Surgical procedure

Surgical procedure consisted of capsu-lorhexis, small-incision phaco emulsifi-cation and in-the-bag implantation of aone-piece, surface modified (heparincoated) IOL (Pharmacia A/S Corpor-ation, NJ, USA).

A three-step self-sealing 3.5mm scleraltunnel incision and a side port incisionfor the left hand instrument were made.Central circular capsulorhexis (CCC) wascarried out; the nucleus was removedusing phaco emulsification and a ’divideand conquer’ technique. The capsularbag was cleaned up with irrigation/aspir-ation. The tunnel incision was enlargedto 5.2mm and an intraocular lens withan optical diameter of 5mm (Pharmacia/UpjohnA 809C) was implanted. BSS wasinjected through the side port incisionuntil normal intraocular tension. Twentymg Hexamycin and 2mg Decadron wereinjected intraconjunctivally. A horizontalsuture was used.

Statistical methods

The progression of DR in the eye oper-ated on is compared against its pro-gression in the control eye by calculationof relative risk (RR). The significance ofthis parameter is evaluated by Fisher’s ex-act test (two-tailed). The Mann–WhitneyU-test is used to evaluate the different

21

Table 2. Retinopathy before cataract surgery; operated eye and control eye. (median (range))

Preoperative retinopathy Planned surgery Control

Median (range) 1(0–5) 0 (0–5)0 (no retinopathy)% 49 511–2 (simplex retinopathy)% 24 223 (preproliferative retinopathy)% 0 2.54 (photocoagulated retina)% 17 155 (proliferative retinopathy)% 10 10

stages of DR. Visual acuity, age, durationof diabetes, HbA1c, S-creatinine, U-albu-min, BAB and lens fluorescence werecompared with Student’s t-test. A com-plete analysis of all data was carried outby estimating the relative risk of eachparameter.

ResultsAt baseline, 50% of the patients had noretinopathy, 23% had background retino-pathy, 2.5% had preproliferative retinopa-thy, 16% were previously photocoagu-lated and 10% had quiescent, proliferat-ive retinopathy (Table2).

At the final follow-up examination, fivepatients had unilaterally progressed in re-tinopathy and one patient had progressedtwo grading steps more in the eye operatedon than in the control eye. Four patients(all NIDDM) had a unilateral macularoedema in the eye operated on. These 10patients were assigned to the complicatedgroup. Twenty-nine patients were assignedto the noncomplicated group. In the non-complicated group, one patient post-operatively developed both bilateral pro-gression of DR and macular oedema andanother patient developed bilateral macu-lar oedema. The data of the two patientgroups are shown in Table3.

No clinical characteristics differ sig-

Table 3. Clinical characteristics of patients assigned to either the ‘complicated’ or the ‘noncomplicated’ group (mean (SD)). P . 0.2 for all parametersexcept lens fluorescence where P Ω 0.11 (a Ω 0.05).

Mean (SD) Complicated Non-complicated Difference of mean

Patients (male/female) 10 (4/6) 29 (15/14)Age (year.) 70.5 (10.8) 70 (7.6) 0.5Duration of diabetes (year.) 13.4 (7.8) 14.8 (15.5) 1.4HbA1c (%) 8.6 (1.6) 8.3 (1.9) 0.3S-creatinin (mmol/L) 95 (30.7) 90 (29.6) 5U-albumin (mmol/L) 0.78 (1.3) 4.0 (14.1) 3.22Insulin treatment (%) 40 40 0IDDM (%) 30 28 2Antihypertensive treatment (%) 80 72 8Mean blood pressure (mmHg) 138 (21.8) 135 (19.4) 3BAB (x 10ª6 minª1) 707 (375) 602 (586) 105Lens fluorescence (ng/mL) 674 (337) 526 (287) 148

nificantly between the two patientgroups, although lens fluorescence tendstowards a higher level in the complicatedgroup. An estimate of the relative risk(RR) of each parameter showed a nonsig-nificant tendency towards HbA1c as themost potent risk factor for progression ofretinopathy after cataract surgery.

Relative risk

It appeared during the follow-up periodthat cataract surgery seems to induce anincreased risk of progression of DR ordevelopment of macular oedema (Table4).

Data analysis confirms that the calcu-lated relative risk (RR) of progression ofDR or development of oedema is signifi-cantly higher in the eye operated on thanin the control eye: RRΩ3.25 (95% CI:1.14–9.3) Fisher’s exact test, two-tailed,PΩ0026 (Table 5).

Visual acuity

Regarding the whole patient group (NΩ39), the majority of patients (36 out of39) gained enhanced VA after cataractsurgery (see Fig.1)

Postoperative VA was significantlybetter at the six month follow-up than ithad been preoperatively (t-test, P,0001)and also significantly better than on thecontrol eye (t-test, P,0001). VA meas-ured one week after surgery is identical

to that observed at the six-month follow-up.

Three patients experienced postopera-tive reduction of VA, all caused by post-operative macular oedema. Only one pa-tient developed a unilateral oedema (onthe eye operated on). The other two de-veloped bilateral oedemas and one ofthem simultaneously developed a bilat-eral progression of retinopathy. Thegroup is mainly characterised by a rela-tively high level of HbA1c (8.4–9.9).

Macular oedema

Four out of 39 patients developed unilat-eral macular oedema on the eye operatedon during the follow-up period.

All of these are NIDDM patients, withan average duration of diabetes of 15yearsand HbA1c of 7–10.2%. The oedema wassmall but significant in all patients. Noneof the patients developed a simultaneousprogression of retinopathy. All patientsdeveloped the macular oedema after 3months follow-up or more, which was par-alleled in their VA (Fig.2).

Lens fluorescence and BAB

Levels of lens fluorescence showed con-siderable interindividual variation withgood correlation between each patient’stwo eyes (Pearson rΩ0.87) (Table6).

BAB examinations also showed con-siderable interindividual variation (Table7), and, except in the case of one patient,showed a high level of correlation betweeneach patient’s eyes (Pearson rΩ0.85). Theexceptional patient developed a retinal de-tachment 1month after cataract surgeryon the eye that preoperatively had demon-strated very high leakage of BAB. All pa-tients preoperatively had normal IOP andno patients had the type of cataract thatcould potentially compromise the flow ofaqueous matter in the anterior eye.

22

Table 4. The relative risk (RR) of each parameter comparing patients from the ‘complicated’ andthe ‘noncomplicated’ groups.

Complication RR CI (95%) Chi-square P Ω

HbA1c . 8 2.66 0.6–12.5 0.17Duration .10 years 0.75 0.2–3.1 0.88IDDM 1.00 0.2–4.8 0.69Hypertension 1.66 0.3–9.3 0.73Retinopathy $2 1.00 0.2–4.9 0.73Kd . 700 1.66 0.4–7.8 0.63Lensflu .700 1.66 0.3–8.7 0.70

Table 5. Incidence of progression of retinopathy and macular oedemas during 6 months follow-upafter cataract surgery. Incidences are given for the eye operated on and for the control eye.

Cataract surgery Control eye Total

Progression of DR 7 2 9Oedema 6 2 8Total 13 4 17

DiscussionMost patients (36 out of 39) in our studyachieved improved and good VA aftercataract surgery. Postoperative reductionof VA was in two out of three casescaused by macular oedema that was bi-lateral and therefore probably not causedby cataract surgery.

Our material also represents the risk ofprogression of retinopathy and/or devel-opment of macular oedema after cataract

Fig. 1. Visual acuity pre (x-axis) and postoperatively (y-axis) of the 39 patients who completedthe follow-up period. Visual acuity is given as the number of letters read on an ETDRS board(85 letters , 6/6 on a Snellen board).

surgery. The risk for the eye undergoingsurgery is significant compared to thatfor the control eye, but, in comparison torisk identified in previous reports, anychanges are moderate and controllable.

Using the described surgical routine(one experienced surgeon, phaco tech-nique, surface modified IOL) neither thelevel of retinopathy, the duration of dia-betes history, nor the degree of glycaemiccontrol influence the risk of progressionof retinopathy during follow-up. This re-

Table 6. Lens fluorescence of the eye operatedon and the control eye.

Lens fluorescence Operated Control(ng/mL) eye eye

Mean 562 542Max 1473 1408Min 133 196

Table 7. Blood-aqueous barrier of the eye oper-ated on and the control eye.

BAB Operated Controleye eye

Mean ( ¿ 10ª6minª1) 723 619Max 3483 2088Min 57 57

sult is similar to that identified in otherrecent reports (Kato et al. 1999). Patientswith preoperative proliferative diabeticretinopathy were all treated with panreti-nal photocoagulation and had quiescentproliferative retinopathy at the time ofsurgery. Four patients developed unilat-eral macular oedema after more than 3months during follow-up.

Lens fluorescence is thought to parallelthe lifelong level of glycation and corre-sponds to a lifetime HbA1c: in accord-ance with the results of Henricsson et al.(1996) it is considered to represent a riskfactor for progression of retinopathyafter cataract surgery. The complicatedgroup in the present study revealed a ten-dency towards a higher level of lensfluorescence. Different types of cataractin otherwise comparable patients differsignificantly in fluorescence level (Vanden Brom et al. 1990; Siik et al. 1993);this may be one reason for the consider-able interindividual variation. The rel-evance of lens fluorescence as a preopera-tive risk indicator before cataract surgerymust be modified by several parameterssuch as age, duration and type of diabetesif a preoperative measurement of lensfluorescence is to be considered a reliablerisk indicator for postoperative pro-gression of retinopathy and/or develop-ment of macular oedema.

The complicated group also tended to-wards a higher level of permeability of theBAB. This difference is, however, far frombeing statistically significant. We find thisa little surprising, since BAB evaluates theactual status of the ocular vessels in theanterior eye, and we would have expectedany great leakage to imply an increasedrisk of postoperative progression of retin-

23

Fig. 2. Visual acuity, measured at baseline and during follow-up (1 week, 1, 3 and 6 months), ofpatients who develop unilateral macular oedema after cataract surgery. Visual acuity is given asthe number of letters read on an ETDRS board (85 letters , 6/6 on a Snellen board).

opathy and/or development of macularoedema. This hypothesis cannot be con-firmed in our material. Moreover, BABhas a very large interindividual variation,both in our material and in healthy per-sons (van Best et al. 1993).

None of our patients developed neo-vascular glaucoma, continuing retinalneovascularisation or prolonged post-operative ocular inflammation followingcataract surgery, as previously reportedin some studies (Pollack et al. 1991;Schatz et al. 1994; Sadiq et al. 1995). Webelieve the surgery may be so lenient thatthe evaluation of lens fluorescence andBAB is compromised. A less experiencedsurgeon might perhaps have provokedmore per- and postoperative compli-cations which might have made evalu-ation of lens fluorescence and BAB sizemore efficient as possible risk indicators.It would also have been interesting tocompare these per- and postoperativecomplications arising in surgery with avery experienced surgeon and one less ex-perienced. For ethical reasons we aban-doned this study model.

ConclusionThis study reports cataract surgery indiabetic patients as producing good vis-ual prognosis. Only a few patients pro-gress in retinopathy or develop macularoedema after cataract surgery performedby an experienced surgeon using the tech-nique of phaco emulsification and a sur-face modified IOL. The changes that doappear are generally moderate and ac-cessible for retinal photocoagulation.

However, cataract surgery still presents asignificantly higher risk of progression ofretinopathy and/or development of macu-lar oedema than that experienced by thecontrol eye. NIDDM patients in particu-lar appear to be at increased risk of devel-oping macular oedema after cataractsurgery, implying that postoperative fol-low-up should last at least three to 6months as the few macular oedemas re-ported in the present material all ap-peared after more than 3months obser-vation.

Neither lens fluorescence nor BAB ap-pear to have any potential as risk indi-cators for diabetic patients scheduled forcataract surgery. Ocular diabetic sequel(retinopathy level), general diabeticcharacteristics (duration of diabetes, gly-caemic control, antihypertensive treat-ment) are not significant risk indicatorswhen cataract surgery is performed indiabetics who have received relevant pre-operative retinal photocoagulation.

Diabetes should no longer be con-sidered a contraindication to cataractsurgery when attention is paid to the preand postoperative control of retinopathy.

AcknowledgementWe would like to thank photographer HansHenrik Petersen for skilful technical assistance.

ReferencesAldington SJ, Kohner EM, Meuer S, Klein

R & Sjclie AK (1995): Methodology for reti-nal photography and assessment of diabetic

retinopathy: the EURODIAB IDDM com-plications study. Diabetologia 38: 437–444.

Antcliff RJ, Poulson A & Flanagan DW(1997): Phacoemulsification in diabetics Eye10: 737–741.

van Best J, del Castillo JB, Diestelhorst M etal. (1996): Diffusion coefficient through theblood-aqueous barrier using a standard pro-tocol. Br J Ophthalmol 80: 356–362.

van Best J, Mota MC & Larsen M (1993):Manual of Ocular Fluorometry. EuropeanCommunity Biomedical Programme OnOcular Fluorometry.

Bleeker JC, van Best JA, Vrij L, van der VeldeEA & Oosterhuis JA (1986): Autofluoresc-ence of the lens in diabetic and healthy sub-jects by fluorophotometry. Invest Ophthal-mol Vis Sci 27: 791–794.

Boot JP, van Gerven JM, van Best JA, Vrij L,Lemkes HH & Oosterhuis JA (1989): Bloodretinal and blood aqueous barriers in dia-betics by fluorophotometry. Doc Ophthal-mol 71: 19–27.

Cunliffe IA, Flanagan DW, George ND, Ag-garwaal RJ & Moore AT (1997): Extracap-sular cataract surgery with lens implantationin diabetics with and without proliferativeretinopathy. Br J Ophthalmol 75: 9–12.

Dowler JG, Hykin PG, Lightman SL & Hamil-ton AM (1995): Visual acuity followingextracapsular cataract extraction in dia-betes: a meta-analysis. Eye 9: 313–317.

Henricsson M, Heijl A & Janzon L (1996):Diabetic retinopathy before and after catar-act surgery. Br J Ophthalmol 80: 789–793.

Kato S, Fukada Y, Hori S, Tanaka Y & OshikaT (2000): Influence of phacoemulsificationand intraocular lens implantation on thecourse of diabetic retinopathy. J CataractRefract Surg 25: 788–793.

Kjer B, Larsen M, Bendtson I, Binder C, Dal-gaard P & Lund-Andersen H (1987): Lensautofluorescence in diabetes compared withthe level of glycosylated hemoglobin A1c.Acta Ophthalmol Suppl 182: 100–102.

Klein R, Klein BE, Moss SE, Davis MD &DeMets DL (1984): The Wisconsin epide-miologic study of diabetic retinopathy III.Prevalence and risk of diabetic retinopathywhen age at diagnosis is 30 or more years.Arch Ophthalmol 102: 527–532.

Klein R, Klein BE, Moss SE, Davis MD &DeMets DL (1984): The Wisconsin epide-miologic study of diabetic retinopathy IV.Diabetic macular oedema. Ophthalmol 91:1464–1474.

Koefoed Theil P, Hansen T, Larsen M, Peders-en O & Lund-Andersen H (1996): Lens au-tofluorescence is increased in newly diag-nosed patients with NIDDM. Diabetologia39: 1524–1527.

Krupsky S, Zalish M, Oliver M & Pollack A(1997): Anterior segment complications indiabetic patients following extracapsularcataract extraction and posterior chamberintraocular lens implantation. OphthalmicSurg 22: 526–530.

Larsen M, Kjer B, Bendtson I, Dalgaard P &

24

Lund-Andersen H (1989): Lens fluorescencein relation to metabolic control of insulin-dependent diabetes mellitus. Arch Ophthal-mol 107: 59–62.

Menchini U, Bandello F, Brancato R, Camesa-sca FI & Galdini M (1993): Cystoid macularoedema after extracapsular cataract extrac-tion and intraocular lens implantation indiabetic patients without retinopathy. Br JOphthalmol 77: 208–211.

Moriarty AP, Spalton DJ, Moriarty BJ, Shil-ling JS, Ffytche TJ & Bulsara M (1994):Studies of the blood-aqueous barrier in dia-betes mellitus. Am J Ophthalmol 117: 768–771.

Oimomi M, Maeda Y, Hata F et al. (1988):Glycation of cataractous lens in non-dia-betic senile subjects and in diabetic patients.Exp Eye Res 46: 415–420.

Parving HH, Larsen M, Hommel E & Lund-Andersen H (1989): Effect of antihyperten-sive treatment on blood-retinal barrier per-meability to fluorescein in hypertensive type 1(insulin-dependent) diabetic patients withbackground retinopathy. Diabetologia 32:440–444.

Pollack A, Dotan S & Oliver M (1991): Pro-gression of diabetic retinopathy after cataractextraction. Br J Ophthalmol 75: 547–551.

Pollack A, Leiba H, Bukelman A & Oliver M(1992): Cystoid macular oedema following

cataract extraction in patients with diabetes.Br J Ophthalmol 76: 221–224.

Ramalho JS, Marques C, Pereira PC & MotaMC (1996): Role of glycation in human lensprotein structure change. Eur J Ophthalmol6: 155–161.

Sadiq SA, Chatterjee A & Vernon SA (1995):Progression of diabetic retinopathy and rub-eotic glaucoma following cataract surgery.Eye 9: 728–738.

Schalnus R & Ohrloff C (1995): The blood-ocular barrier in type I diabetes without dia-betic retinopathy: permeability measure-ments using fluorophotometry. OphthalmicRes 27: 116–123.

Schatz H, Atienza D, McDonald HR & John-son RN (1994): Severe diabetic retinopathyafter cataract surgery. Am J Ophthalmol117: 314–321.

Sensi M, Pricci F, Pugliese G et al. (1995): Roleof advanced glycation end-products (AGE)in late diabetic complications. Diabetes ResClin Pract 28: 9–17.

Siik S, Airaksinen PJ, Tuulonen A & NieminenH (1993): Autofluorescence in cataractoushuman lens and its relationship to lightscatter. Acta Ophthalmol (Copenh) 71: 388–392.

Sjølie AK, Stephenson J, Aldington S et al.(1997): Retinopathy and vision loss in insu-lin-dependent diabetes in Europe. The

EURODIAB IDDM Complications StudyOphthalmol 104: 252–260.

Sparrow JM, Bron AJ, Brown NA & Neil HA(1997): Autofluorescence of the crystallinelens in early and late onset diabetes. Br JOphthalmol 76: 25–31.

Stephenson JM, Fuller JH, Viberti GC, Sj√lieAK & Navalesi R (1995): Blood pressure, re-tinopathy and urinary albumin excretion inIDDM: the EURODIAB IDDM Compli-cations Study. Diabetologia 38: 599–603.

Van den Brom HJ, Kooijman AC & BlanksmaLJ (1990): Clinical and physical measure-ments of the cataractous lens. Doc Ophthal-mol 75: 247–258.

Wagner T, Knaflic D, Rauber M & Mester U(1996): Influence of cataract surgery on thediabetic eye: a prospective study. Ger JOphthalmol 5: 79–83.

Received on May 15th, 2001.Accepted on September 17th, 2001.

Correspondence:Per Flesner, MDTagensvej 38 3 thDK-2200 Copenhagen NDenmarkTel: π45 35 82 29 13e-mail: Flesner/dadlnet.dk