Graefe's Arch Clin Exp Ophthalmol (1998)236:822±828 � Springer-Verlag 1998 C L I N I C A L I N V E S T I G A T I O N
Christopher WirbelauerHans ho*raufJohann RoiderHorst Laqua
Corneal shape changesafter pars plana vitrectomy
Received: 2 February 1998Accepted: 26 February 1998
The authors have no commercial,proprietary or financial interest in anyresearch or devices described in thepresented study.Presented in part at the 95th Annual Meetingof the German Ophthalmological Society,Berlin, September 1997
C. Wirbelauer ()) ´ H. ho*raufJ. Roider ´ H. LaquaAugenklinik,Medizinische Universität zu Lübeck,Ratzeburger Allee 160,D-23538 Lübeck, GermanyTel. +49-451-5002211fax +49-451-5003085e-mail [emailprotected]
Abstract l Background: The pur-pose of this prospective, controlled,clinical study was to investigate cor-neal shape changes due to pars planavitrectomy (PPV) in patients withpotential visual improvement post-operatively. l Methods: A total of36 consecutive patients undergoingconventional PPV combined with orwithout macular surgery were stud-ied. Sequential determinations of thecorneal curvature by manual kera-tometry and corneal topography wereperformed preoperatively, during thefirst postoperative week, at 4 weeksand after 2±8 months (mean 4months). l Results: The mean surg-ically induced keratometric astigma-tism was 2.92�1.98 diopters (D)(P<0.0001) during the first postop-erative week. After 4 weeks and4 months the values decreased to1.01�0.97 D and 0.67�0.43 D, re-spectively. Videokeratographic anal-ysis confirmed significant curvaturechanges, with corneal steepening(P<0.008) which corresponded to the
superonasal and temporal semime-ridian, and flattening (P<0.008)along the inferior and inferonasalsemimeridians. Corneal changes per-sisted in some cases for severalweeks, partly in an asymmetric andirregular configuration. Shifts in axisto against-the-rule and oblique me-ridians were noted postoperativelywith redistribution within preopera-tive values in 53% of the cases at 4weeks after surgery. Subgroup anal-ysis revealed that suture diameter, aswell as the use of gas endotampon-ade, influenced the induced astigma-tism.l Conclusion: A substantial increaseof the corneal astigmatism and distinctshape changes can occur after PPV inthe immediate postoperative period.Consecutive stabilization at preopera-tive values was observed after severalweeks. The time course of the cornealcurvature alterations should be con-sidered in the postoperative manage-ment to detect refractive causes ofinadequate visual acuity.
Introduction
The scleral incisions in pars plana vitrectomy (PPV) caninfluence the corneal curvature and can be an importantfactor in the visual outcome of surgery [7, 12]. Centralcorneal changes are optically the most important for theformation of the macular image and in certain condi-tions might produce transitory severe refractive impair-ment of postoperative vision. Although some research-
ers [12] regard the changes after PPV as slight and tran-sitory, having little long-term clinical consequences, thepatients' visual acuity would be certainly impaired se-verely by induced changes in corneal shape that are bothasymmetric and irregular. Rapid visual rehabilitation isimportant in patients with potentially good macularfunction undergoing PPV, and knowledge of the postvit-rectomy response would be useful in guiding postoper-ative management. This study sought to investigatethe effect of incision and wound closure on corneal
823
shape changes associated with routine PPV postopera-tively.
Patients and methods
Patients and clinical examination
Thirty-six consecutive phakic patients scheduled for elective vitrec-tomy were studied prospectively in a controlled clinical setting.Each patient underwent a complete preoperative ophthalmologic ex-amination, including anterior segment biomicroscopy, applanationtonometry (Goldmann), and a binocular fundus examination. Thecriterion for eligibility was the presence of a vitreoretinal diseasewith potential for improvement of the visual function postoperative-ly. The indications for surgery are listed in Table 1. None of the pa-tients had a history of previous ocular surgery or diseases affectingthe cornea. Postoperative examinations were performed within thefirst week, at 4 weeks, and on average 4 months (range 2±8 months)after PPV. Twenty-three patients (64%) included in the study hadcomplete follow-up visits.
Surgical procedure
After informed consent had been obtained, all patients underwent anidentical standardized surgical procedure including conventionalthree-port PPV without additional episcleral buckling. After perito-my the episcleral blood vessels were coagulated with light wet-fieldbipolar cautery to achieve sufficient hemostasis. In normal-sizedeyes, the sclerotomies were made 3.5 mm from the limbus. A micro-vitreal knife with a blade 0.9 mm in width was directed perpendic-ular to the sclera to achieve a stab-type incision parallel to the lim-bus for use of surgical instruments. The sclerotomy for the infusioncannula was made in the inferotemporal quadrant and the infusionline was sutured during the operation with 6-0 polyglactin sutures.In deep-set eyes or narrow eyelids a more temporal location was pre-ferred. Vitrectomy instruments were introduced through the sclerot-omy sites in the superonasal and superotemporal quadrant. Intraoc-ular manipulations with removal of epiretinal membranes, cryocoag-ulation, photocoagulation, fluid replacement or additional intraocu-lar gas exchange to tamponade the retina (C2F6) were combined asrequired. The intraocular solutions used were at room temperature.On completion of the procedure the sclerotomies were closed in afigure-of-eight fashion with braided, absorbable polyglactin sutures(Vicryl, Ethicon, Norderstedt, Germany). The scleral bites extendedto at least one-half of the depth of the sclera with suture diametersfrom 6-0 to 8-0. The suture ends were then trimmed to the knot.An effort was made to use identical surgical techniques in all cases,performed only by experienced surgeons.
All patients were given prednisolone acetate 1% drops five timesdaily and hydrocortisone acetate 2.5% ointment at night, taperedover 4±6 weeks.
Corneal shape analysis
Total absolute corneal astigmatism was determined in all eyes bymanual keratometry (Ophthalmometer, Zeiss, Oberkochen, Germa-ny). In 28 eyes (78% of the cases), simulated keratometry readingsof the keratometric equivalent at the central 3-mm optical zone wereobtained from videokeratography (Eyemap EH-290, Version 5.02,Alcon, Fort Worth, Tex.). Both instruments were calibrated accord-ing to the manufacturer's instructions prior to the study. Data usedfor analysis were the average of three consecutive manual keratome-try readings and two corneal topographic maps. Tear film abnormal-ities, improper focusing, or eccentric fixation, which are important
sources of error using corneal topography, were noted. The surgical-ly induced astigmatism was calculated using the vector analysismethod described by Jaffe [11].
The stored data from the highest quality videokeratograph werethen further processed for qualitative assessment using the absolutescale map and the differential map. Qualitative analysis was carriedout according to the previously described objective pattern classifi-cation system, with the round and oval categories combined [4]. Toconfirm the topographic findings each color-coded map was care-fully interpreted using both the 0.25-diopters (D) and 1.0-D inter-vals.
Videokeratographic corneal dioptric analysis was performed atpoints at a radial distance of 1.5 mm and 2.5 mm from the cornealapex (i.e. 3.0- and 5.0-mm optical zones) at 45� intervals. This cor-responds to 16 points on 8 semimeridians. For each of these points,the preoperative curvature was subtracted from the postoperativecurvature to calculate the surgically induced dioptric change [14].To improve analysis all left eyes were mathematically treated asright eyes.
Statistical analysis
Changes in the individual groups were analyzed with the Wilcoxonsigned-rank test (intragroup differences). The Bonferroni methodwas used to adjust the probability value for multiple comparisons.Accordingly, a P value of less than 0.008 was considered statistical-ly significant.
Where appropriate statistical analyses were performed to com-pare various subgroups using the chi-square test for categoric vari-ables and the Mann-Whitney-U test for the measurement data (inter-group differences). Statistical differences in which the P value wasbelow 0.05 were considered significant.
Results
A total of 36 eyes of 36 patients participated in this study.Of these, 22 (61%) were female and 14 (39%) were male.The average (�SD) patient age was 62.1�13.4 years(range 32±78 years). The patients' demographic data arelisted in Table 2. No statistically significant differenceswere observed between the subgroups regarding age, gen-der, ratio of right and left eyes, or preoperative keratomet-ric cylinder.
Table 3 details the total keratometric and videokerato-graphic astigmatic changes at each time point. The meankeratometric absolute astigmatism rose to 2.95�1.88 D(P<0.0001) at 1 week after surgery, then decreased grad-ually, recovering to the preoperative level within 4 months.In 34 (94%) of the patients the absolute astigmatism rised,
Table 1 Indications for surgery
Diagnosis No.of patients
% of totalpatients
Idiopathic macular hole 12 34Idiopathic epiretinal membrane 8 22Vitreous opacities 3 8Vitreous hemorrhage (due to diabetesmellitus or other causes)
13 36
824
and the proportion of patients with absolute keratometriccylinder of more than 2.0 D during the first week was 23eyes (64%). The mean surgically induced astigmatismwas 2.92�1.98 D, with a maximum of 8.82 D in the imme-diate postoperative period. Only 6 patients (17%) had morethan 2.00 D of astigmatism 4 weeks after surgery. Cornealtopography confirmed these values (Table 3). The totalmean keratometric flat meridian decreased significantlypostoperatively from 42.88�1.37 D to 42.01�1.77 D(P<0.001), regaining preoperative values after 4 weeks(42.85�1.44). The corneal refractive power changed from43.24�1.35 D to 43.48�1.52 D (P=0.005) at 1 week withcontinued steepening after 4 weeks (43.45�1.41 D,P=0.003). The corneal power returned to preoperative val-ues at 4 months (43.30�1.53 D).
The mean changes in the keratometric and surgicallyinduced cylinder, stratified into several groups based onthe injection of gas or the diameter of the used sutures,are listed in Tables 4 and 5. Among the treated patients,15 (42%) required the use of intraocular gas. TheMann-Whitney-U test found a statistically significant dif-ference (P=0.023) in the change of the amount of absoluteastigmatism after 4 weeks between the different sub-groups, suggesting that the postoperative astigmatic decaypattern was longer with the use of gas.
Patients with 6-0 sutures (n=8) showed a tendency toinduce curvature changes of approximately 1 D higherthan patients with 7-0 (n=5) or 8-0 (n=23) suture diameter(Table 5). These differences were surgeon independent.While the relatively small number of patients in each sub-
Table 2 Patients' demographicdata (mean�standard deviation) n Age (years)a Gender (F:M)b OD/OSb Preoperative keratometric cylinder
(D)a
Total 36 62.1�13.4 22:14 19/17 0.71�0.51
SubgroupsControl 21 59.5�15.0 12:9 10/11 0.74�0.52Gas 15 65.7�10.1 10:5 9/6 0.66�0.52
Sutures6-0 8 60.9�16.4 4:4 6/2 0.72�0.787-0 5 61.2�5.9 4:1 4/1 0.83�0.578-0 23 62.7�13.9 14:9 9/14 0.68�0.40
a Mann-Whitney-U test: no sig-nificant differencesb Chi-square test: no significantdifferences
Table 3 Keratometric and video-keratographic cylinder over timeexpressed as mean and standarddeviation in patients undergoingpars plana vitrectomy
Keratometric cylinder (D) Videokeratographic cylinder (D)
Absolute Induceda Absolute Induceda
Preoperative 0.71�0.51 0.79�0.47Postoperative1 week 2.95�1.88* 2.92�1.98 3.30�2.01* 3.40�2.104 weeks 1.21�1.03 1.01�0.97 1.36�0.98* 1.19�1.074 months 0.79�0.59 0.67�0.43 0.88�0.57 0.70�0.63
a Calculated by the vector analysis method* P<0.008, postoperative compared with preoperative values (Wilcoxon signed-rank test)
Table 4 Keratometric cylinderover time expressed as meanand standard deviation in pa-tients undergoing pars planavitrectomy with and without gasendotamponade
Keratometric cylinder (D)
Control group (n=21) Gas group (n=15)
Absolute Induceda Absolute Induceda
Preoperative 0.74�0.52 0.66�0.52Postoperative1 week 2.52�1.40* 2.54�1.59 3.63�2.36* 3.52�2.424 weeks 0.86�0.66 0.68�0.45 1.77�1.28+ 1.53�1.324 months 0.64�0.43 0.58�0.42 0.99�0.73 0.78�0.43
a Calculated by the vector analysis method* P<0.008, postoperative compared with preoperative values (Wilcoxon signed-rank test)+ P<0.05, changes from preoperative values: control group vs gas group (Mann-Whitney-U test)
825
group precluded analysis of statistical significance, thetrend seems to indicate the presence of meaningful clini-cal differences between the subgroups at 1 week after sur-gery.
Shifts in axis to against-the-rule and oblique meridianscould be noted 1 week after surgery (Table 6). Most ofthe orientations were in meridians corresponding to oneof the sclerotomy sites. The axis distributions later re-turned to preoperative values in 53% of the cases. How-ever, further changes of the main meridian in several pa-tients suggest prolonged axis instability (Table 6).
The videokeratographic corneal dioptric powerchange at the 16 corneal locations was greatest in the su-peronasal (45�) and temporal (180�) semimeridians (Ta-ble 7). In these regions significant mean corneal steepen-ing (P<0.008) occurred with remarkable changes that ex-
ceeded 6.0 D in some cases. Furthermore, mean steepen-ing was present along the nasal (0�) and inferotemporal(225�) semimeridians. In the 225� semimeridian, largerstandard deviations indicated a larger degree of variabil-ity in the induced changes. These alterations were cou-pled on the opposite side along the inferior (270�) and in-feronasal (315�) semimeridians with marked mean flat-tening (P<0.008). A variable gradual decrease or in-crease in astigmatic changes was noted as the measure-ments moved from the corneal center to paracentral re-gions. At 4 weeks postoperatively, the maximal dioptricchange was 2.84 D and the mean amount of changewas low in most cases. However, there was still signifi-cant steepening compared with baseline values at the su-peronasal (45�) and temporal (180�) points. Regionalchanges of the dioptric power were still detectable after4 months, with significant flattening (P<0.008) oppositethe sclerotomy positions (0� and 315� points).
The distribution of topographic patterns is shown inTable 8. Corneal topography changed from a hom*ogene-ous contour preoperatively to predominantly asymmetri-cal bow tie (57%) and irregular patterns (25%) at 1 week.Eyes with an irregular pattern were consistently irregularon both of the maps that were taken. Of the 28 eyes, 25(89%) presented such sclerotomy-induced corneal chang-es. Wound closure induced considerable segmental cen-tral and peripheral steepening in the temporal quadrantsof the globe. However, there were marked differences
Table 5 Keratometric cylinderover time expressed as meanand standard deviation in pa-tients undergoing pars planavitrectomy with different suturediameters
Keratometric cylinder (D)
6-0 (n=8) 7-0 (n=5) 8-0 (n=23)
Absolute Induceda Absolute Induceda Absolute Induceda
Preoperative 0.72�0.78 0.83�0.57 0.68�0.40Postoperative1 week 3.68�2.23 3.51�2.54 2.38�1.60 2.30�1.27 2.80�1.80 2.84�1.904 weeks 1.13�1.25 1.03�0.86 0.88�0.57 0.52�0.36 1.32�1.05 1.11�1.094 months 0.37�0.51 0.73�0.50 0.69�0.57 0.39�0.25 0.93�0.59 0.72�0.44
a Calculated by the vector analy-sis method
Table 6 Axis distributions over time [number of eyes (%)]. WTRwith the rule (0�±30�, 150�±179�), ATR against the rule (60�±120�), OBL oblique axis (31�±59�, 121�±149�)
WTR ATR OBL
Preoperative 19 (53) 10 (28) 7 (19)Postoperative1 week 4 (11) 20 (56) 12 (33)4 weeks 17 (47) 16 (44) 3 (9)4 months 10 (43) 5 (22) 8 (35)
Table 7 Videokeratographicsurgically induced changes incorneal power (D) at a radialdistance of 1.5 mm and 2.5 mmfrom the apex corneae overtime, expressed as mean andstandard deviation, in patientsundergoing pars plana vitrecto-my. Values derived by sub-tracting preoperative from post-operative values. Positive valuesindicate steepening. Negativevalues indicate flattening. Alleyes converted mathematicallyas right eyes
Mean change (D)
1 week 4 weeks 4 months
1.5 mm 2.5 mm 1.5 mm 2.5 mm 1.5 mm 2.5 mm
Semimeridian (�)0 0.66�1.24 0.04�1.55 0.35�0.93 0.03�1.04 �0.71�0.72* �0.81�0.99*
45 1.48�1.88* 0.59�1.45 0.73�1.05* 0.17�1.49 �0.44�0.64 �0.73�0.8590 0.01�1.65 �0.50�1.65 0.21�1.14 �0.09�1.10 �0.29�0.84 �0.51�0.88
135 �0.09�1.13 0.08�1.40 0.23�0.94 0.39�1.08 �0.21�0.72 �0.21�0.63180 1.25�1.58* 1.65�1.46* 0.65�0.98 0.79�1.16* �0.10�0.72 �0.14�0.80225 0.63�2.23 0.93�2.14 0.41�0.92 0.23�1.31 �0.23�0.65 �0.40�0.84270 �1.05�1.55* �0.87�1.70 0.13�1.14 0.07�1.30 �0.25�0.96 �0.51�0.97315 �0.96�1.35* �0.92�1.52 �0.01�0.97 �0.11�1.04 �0.38�0.66 �0.63�0.58*
* P<0.008, postoperative compared with preoperative values (Wilcoxon signed-rank test)
826
in response among the eyes. Higher rates of wound-relat-ed steepening occurred in eyes with larger suture diame-ters. In most patients the sclerotomy-induced cornealshape changes altered over time and decreased gradually,but in some cases asymmetric corneal astigmatism (21%)or irregular focal steepening and flattening (14%) persist-ed for up to 4 weeks after surgery. Resolution of thesedistortions was observed in all patients and the corneashad recovered their normal shape by the last follow up.
Discussion
The results of this study describing the time course of cor-neal shape changes induced by PPV demonstrated that,despite the small sclerotomies at a location relatively pos-terior to the central cornea, significant alterations are pro-duced in the immediate postoperative period by this com-monly used surgical procedure. The corneal surface dy-namics and wound healing response after PPV are affect-ed by several factors related primarily to the wound con-struction. The effect of incision size, position, and config-uration on local globe deformations is well known fromcataract surgery [2, 11]. Other responsible factors includewound edema, cautery, different tensile strengths and tis-sue gripping characteristics of suture materials, sutureplacement, knot-tying technique, and tension as well asnumber of sutures. The use of corticosteroids postopera-tively also has an effect on wound healing.
The mean induced astigmatism during the first post-opertive week was approximately 3.0 D with a maximumchange of almost 9.0 D, comparable with the findings inprevious studies using braided absorbable sutures [7].The fact that more than 90% of eyes displayed a rise in-dicates that astigmatism will almost always increase afteruncomplicated PPV with traditional wound closure. Asfor the specific response of the cornea to PPV, we iden-tified several typical patterns of changes in the course ofthe healing process. In the immediate postoperative peri-od frequent segmental steepening in the superonasal, in-ferotemporal, and temporal semimeridians was noted.Thus, the sclerotomies seemed to induce an early tran-sient change in astigmatism. The tissue compressioncaused by the radial force of the suture has been provento be the most important factor influencing the cornealcurvature in the meridian of surgery [1, 19, 22]. Local-ized compression of the tissue depressed and flattened
the sclera, which caused central corneal steepening inthe meridian adjacent to the incision. It has been calculat-ed that a scleral difference of 0.1 mm can induce an av-erage change of 1 D [1]. However, changes seem unpre-dictable in scleral tissue with great interindividual differ-ences. Opposite the sclerotomies, significant flatteningwas observed in the inferonasal and inferior locations.The smallest changes occurred in the superotemporal re-gion, probably due to the simultaneous effects of adjacentsuture-induced steepening and flattening from the or-thogonal sclerotomies.
These effects were increased by scleral cautery nearthe incision, which leads to thermal contracture of thetreated tissue with immediate central steepening [3].The effects of cautery on scleral tissue are transient andare difficult to control, but the use of a larger cauteriza-tion zone seems to partially counteract the thermal con-tracture effect [3]. In cataract surgery a 50% reductionin the amount of induced cylinder change was noticedwhen scleral cautery was limited to the sclera posteriorto the insertion of Tenon's fascia and was not performedall the way anteriorly to the limbus [10]. This suggeststhat a more posterior cauterization might reduce the astig-matic effects on the cornea.
In several patients the sclerotomies produced a certaindegree of corneal distortion, and thus, led to asymmetricor irregular patterns. Many of the changes observed incorneal topography images confirmed the stress distribu-tion with complex corneal alterations. These relate to theinteractions of three scleral incisions and their closurewith sutures. Additionally, in the case of 0.9-mm sclero-tomies, the forces are applied more focally with possiblesuture-related wound misalignment and greater distortionof the cornea [7, 13]. Regional differences in the organi-zation and the biomechanical properties of the scleral fi-brils might also influence corneal structural interactionsand stability [8, 15, 20].
The modifications of corneal shape decreased gradual-ly and evolved over an extended period of time duringthe process of wound remodelling. As the scleral incisionhealed, the wound tissue edema cleared, the polyglactinsutures were absorbed, releasing the tension, and the col-lagen contracture from cautery dissipated. The relaxationof the wound produced flattening during the further ob-servation period. In several cases an irregular patternemerged after 4 weeks, with areas of focal steepeningand flattening causing distortion of the optical axis.
Table 8 Topographic patterndistribution over time [numberof eyes (%)]
hom*ogeneousa Symmetric bow tie Asymmetric bow tie Irregular
Preoperative 24 (86) 3 (11) 1 (3) 0 (0)Postoperative1 week 2 (7) 3 (11) 15 (57) 7 (25)4 weeks 15 (54) 3 (11) 6 (21) 4 (14)4 months 18 (95) 1 (5) 0 (0) 0 (0)
a hom*ogeneous comprisesround and oval patterns
827
These areas probably corresponded to the locations of in-tact sutures still exerting tension on the cornea, due totemporal differences in suture hydrolysis. The cornealchanges also resulted in prolonged instability of the mainmeridian, with frequent shifts to against-the-rule or obli-que orientations, as described previously [7]. Further-more, regional dioptric instability was reflected by sig-nificant corneal flattening in the nasal and inferonasal ar-eas after 4 months. An interesting finding is the relation-ship between additional gas injection and significant ag-gravation of the corneal surface changes. Greater earlysurgically induced cylinder corresponded to greater andprolonged astigmatic decay after surgery. This confirmsa stronger compression of the sutures intraoperativelyto avoid loss of gas when the infusion cannula was re-moved [7]. In most patients we demonstrated a substan-tial ability of the cornea to resume a normal contour afterthe sutures were dissolved.
Progressive disintegration of the suture material dueto chemical hydrolysis and physical changes seems ofparticular importance. Using polyglactin sutures, tensilestrength is reduced to 50% after approximately 2 weeksand completely resolves after 4 weeks, with final mate-rial absorption after 70 days (according to Ethicon). Thedramatic increase in astigmatism due to overtight suturesshould, therefore, be no reason to selectively cut or re-move the sutures postoperatively. Our results indicatethat 8-0 polyglactin was responsible for less pronouncedtransient suture-related astigmatism than 6-0 sutures.The smaller diameter sutures not only result in weakerwound compression initially, but because of the lessermass of foreign material the sutures are probably hydro-lyzed more quickly. This means that the visual advanta-ges of 8-0 sutures could be clinically relevant in the firstseveral weeks after surgery. The suture tension decay ofvarious materials and differential use of suture combina-tions possibly affecting the wound healing processshould be considered in future refinements. The use ofnonabsorbable monofilament sutures is not recommend-ed, as a high and long-standing astigmatism can be in-duced [6, 9]. Our study may be criticised for not requir-ing a uniform suture diameter. The protocol design rec-ognized the relatively small sample size and the limita-tions this places on subgroup analysis. However, this in-consistency relates to intraoperative decisions on thepart of surgeons and reflects the diversity of clinical sit-uations that can be encountered during posterior eye seg-ment surgery.
Modifications of suturing technique to construct aphysically stable incision system in accord with basickeratorefractive concepts could markedly influence cor-neal curvature [19]. Circumferential wound appositionparallel to the limbus to decrease the radial tension vectorof the suture has been recently developed for scleral pock-et incisions [17, 21]. Suture bites of variable depth or theplacement of interrupted sutures in the negative cylinder
axis intraoperatively to selectively steepen the corneamight also positively affect the astigmatic change [18,19]. However, the unreliability of the latter, with frequentdifficulties in titrating the sutures, is not practicable forroutine surgery.
Intraoperative control of corneal astigmatism duringPPV would be of benefit to adjust the suture tension undertonometric and keratometric guidance before tying theknot. To prevent excessive compression of the wound, asimple way of estimating intraoperative changes in astig-matism is to use a ring of light projected onto the corneaor to perform measurements with specific surgical kerato-meters [16]. However, in a traditional stab-type incisionconfiguration, the difficulty lies in tensioning the finalwound closure after repressurizing the eye to produce wa-tertight integrity without leakage and to provide accept-able corneal changes. Recent encouraging reports indicatethat it is clinically possible to avoid sutures after PPVthrough the use of self-sealing sclerotomies [5]. The ad-vantages of reduced intraoperative wound leakage couldalso exert a favorable influence on the corneal shape post-operatively.
In the effort to minimize the sclerotomy-induced cor-neal shape changes with traditional wound managementand improve visual recovery, several recommendationscan be made. The first is to select appropriate absorbablesutures of the lowest cross-sectional diameter possible.The second is to avoid any excessive tightening of theprecisely placed suture. Thirdly, cauterization should beperformed with the lowest power setting capable ofachieving hemostasis, and preferably not up to the limbus.
In conclusion, our study showed that in patients withgood visual potential, PPV produced significant cornealshape changes with a high degree of regular and in somecases irregular components distributed around the visualaxis. Changes evolved as a function of time, and the cor-nea needed several weeks to recover and reach a stableform. Because possible significant curvature changes de-lay optical rehabilitation after PPV, keratometry readingsand in selected cases also corneal topography might de-tect refractive causes of inadequate postoperative visualacuity. The observation of continuous corneal shapechanges should, furthermore, prevent unnecessary pre-scriptions for glasses with possible poor toleration ofthe resultant spectacle correction before the dioptric sys-tem has stabilized. A thorough understanding and aware-ness of the possible corneal shape changes induced byPPV can minimize these alterations and help to achievea rapid and stable visual result.
828
References
1. Arciniegas A, Amaya LE (1984) Ex-perimental modification of the cornealcurvature by means of scleral surgery.Ann Ophthalmol 16:1155±1166
2. Armeniades CD, Boriek A, Knolle GE(1990) Effect of incision length, loca-tion, and shape on local corneoscleraldeformation during cataract surgery. JCataract Refract Surg 16:83±87
3. Bergmann MT, Koch DD, Zeiter JH(1988) The effect of scleral cautery oncorneal astigmatism in cadaver eyes.Ophthalmic Surg 19:259±262
4. Bogan SJ, Waring GO III, Ibrahim O,Drews C, Curtis L (1990) Classificationof normal corneal topography based oncomputer-assisted videokeratography.Arch Ophthalmol 108:945±949
5. Chen JC (1992) Sutureless pars planavitrectomy through self-sealing sclero-tomies. Arch Ophthalmol 114:1273±1275
6. Cravy TV (1989) Long-term cornealastigmatism related to selected elastic,monofilament, nonabsorbable sutures. JCataract Refract Surg 15:61±69
7. Eckert T, Eckardt C (1996) Verhaltendes Hornhautastigmatismus nach Pars-plana-Vitrektomie mit oder ohnegleichzeitiger Kataraktoperation. Oph-thalmologe 93:38±44
8. Friberg TR, Lace JW (1988) A com-parison of the elastic properties of hu-man choroid and sclera. Exp Eye Res47:429±436
9. Gimbel HV, Raanan MG, DeLuca M(1992) Effect of suture material onpostoperative astigmatism. J CataractRefract Surg 18:42±50
10. Grabow HB (1991) Early results of 500cases of no-stitch cataract surgery. JCataract Refract Surg 17:726±730
11. Jaffe NS (1981) Postoperative cornealastigmatism. In: Jaffe NS (ed) Cataractsurgery and its complications. Mosby,St. Louis, pp 92±110
12. Jampel HD, Thompson JT, Nunez M,Michels RG (1987) Corneal astigmaticchanges after pars plana vitrectomy.Retina 7:223±226
13. Koch DD, Del Pero RA, Wong TC,McCulloch RR, Weaver TA (1987)Scleral flap surgery for modification ofcorneal astigmatism. Am J Ophthalmol104:259±264
14. Koch DD, Haft EA, Gay C (1993)Computerized videokeratographic anal-ysis of corneal topographic changes in-duced by sutured and unsutured scleralpocket incisions. J Cataract Refract Surg19:166±169
15. Komai Y, Ushiki T (1991) The three-dimensional organization of collagenfibrils in the human cornea and sclera.Invest Ophthalmol Vis Sci 32:2244±2258
16. Masket S (1988) Correlation betweenintraoperative and early postoperativekeratometry. J Cataract Refract Surg14:277±280
17. Masket S (1991) Horizontal anchor su-ture closure method for small incisioncataract surgery. J Cataract Refract Surg17:689±695
18. Roper-Hall MJ, Atkins AD (1985)Control of astigmatism after surgery andtrauma: a new technique. Br J Ophthal-mology 69:352±359
19. Rowsey JJ (1983) Ten caveats inkeratorefractive surgery. Ophthalmolo-gy 90:148±155
20. Smolek MK (1993) Interlamellar cohe-sive strength in the vertical meridian ofhuman eye bank corneas. InvestOphthalmol Vis Sci 34:2962±2969
21. Steinert RF, Brint SF, White SM, FineIH (1991) Astigmatism after small inci-sion cataract surgery ± a prospective,randomized, multicenter comparison of4- and 6.5-mm incisions. Ophthalmolo-gy 98:417±424
22. Van Rij G, Waring GO III (1984)Changes in corneal curvature by suturesand incisions. Am J Ophthalmol98:773±783
