- Email: [email protected]
The repair response following Nd:YAG laser sclerostomy Ab interno in rabbits
Exp. Eye Res. (1995). 61. 311-321
The Repair Response Following N d ' Y A G Laser Sclerostomy Ab Interno in Rabbits MILKO
E. I L I E V ~*, E U G E N V A N
DER Z Y P E N a, F R A N Z ENGLAND a
FANKHAUSER
b
AND CERI
aInstitute of Anatomy, University of Bern," bLindenhof Hospital," Bern, Switzerland (Received Lund 2 February 1995 and accepted in revised form 31 March 1995) This study was undertaken to examine ultrastructurally the course of tissue regeneration after thermallyinduced laser sclerostomy and compare it with the post-operative clinical events in a rabbit model. Applying a continuous wave neodymium:YAG (Nd:YAG) laser, two sclerostomies were created ab interno on one eye in each of ten pigmented rabbits using a 200-/~m-diameter quartz optical fiber; the unoperated fellow-eyes served as controls. Intraocular pressure (IOP) was measured daily before and after surgery. Animals were observed post-operatively for up to 16 days. Well-defined filtering blebs and a low complication rate demonstrated the success of the procedure, lOP was significantly lowered during the entire course of the observation period but after the fifth day, the conjunctival blebs had disappeared. The morphology of the changes in the sclerostomy fistulas were analysed at 2-day intervals by light and electron microscopy. Immediately after surgery, the canal was ensheathed by an approximately lO0-/zmthick layer of coagulated collagenous tissue in which two zones could be distinguished according to the intensity of damage. Within 5 days. the inner and outer canal openings were invaded by macrophages and fibroblasts originating from the iris root and episclera, respectively. The former cells were engaged in the phagocytosis of disintegrated collagen adjacent to the lumen. More distally located fibrils which had incurred less severe damage, had retained their fibrillar structure but had lost banding periodicity. After thermal stress had faded, they appeared to undergo a process of repolymerisation. By day 10, the lumen had become occluded by a loose meshwork of phagocytes, fibroblasts and proliferating capillaries. These new vessels and the loose nature of the canal-occluding framework and of the surrounding regenerating collagenous tissue could have further permitted percolation and transport of aqueous humor, since IOP remained low, despite the disappearance of filtering blebs. Although the time course of repair is more rapid in rabbits than in humans, the data gleaned nonetheless yield valid information respecting the sequence of events following thermally-induced scleral fistula. © 1995 Academic Press Limited Key words: sclerostomy; ab interno: laser: Nd:YAG laser: rabbits: healing process: thermal damage; ultrastructure: glaucoma: intraocular pressure.
1. Introduction Apart from animal experiments, full-thickness sclerostomy has been investigated in some clinical trials for the surgical treatment of refractory g l a u c o m a with promising results (Wilson and Javitt, 1 9 9 0 : I w a c h and Hoskins. 1 9 9 2 ; Kampmeier et al., 1 9 9 3 ; S c h u m a n et al., 1993). This minor and low-traumatic surgical intervention m a y be achieved by m e a n s of different lasers (Jaffe et al., 1 9 8 8 : Mtiller-Stolzenburg et al. 1991 ; Hill et al., 1 9 9 2 : Karp et al., 1993) or by sophisticated surgical equipment (Brown et al., 1987). However, the most c o m m o n long-term post-operative complication of filtering s u r g e r y - - s c a r r i n g with consequent bleb f a i l u r e - - c a n n o t be avoided by a n y of the sclerostomy techniques currently employed. Nonetheless, some authors have found that thermallaser sclerostomies (cw-Nd:YAG, and diode laser) in rabbits and m o n k e y s maintain patency longer and exhibit less inflammatory reaction t h a n conventional * For correspondence at: Institute of Anatomy, University of Bern. Department of Topographical Anatomy, B6hlstrasse 26, CH3000 Bern 9, Switzerland. 0 0 1 4 - 4 8 3 5 / 9 5 / 0 9 0 3 1 1 + 11 $12.00/0
procedure filters (Higginbotham et al., 1988, 1 9 9 0 : Karp et al., 1993). It has been discussed that collateral thermal d a m a g e produced in the walls of scleral fistulas by laser radiation m a y delay the repair response and thus prolong bleb survival, a l t h o u g h this question remains controversial (Hill et al., 1 9 9 2 ; Wetzel and Scheu, 1 9 9 3 ; D~irr and Fankhauser, 1994). Thermal-mode lasers (leading to evaporation and coagulation of tissue), rather than Q-switchedmode lasers (very short pulses of high energy causing optical breakdown and mechanical tearing of tissue), have the advantage of coagulating blood vessels and reducing the risk of bleeding, but thermally-induced sclerostomies also close with time. In an attempt to reduce post-operative scarring, steroids and antimetabolites have been applied with moderate success (Tawakol et al., 1 9 8 9 ; Shirato et al., 1 9 9 0 ; Karp et al., 1994). However, before interference with the undesirable process of canal obliteration can hope to be really successful, a detailed analysis of the course of events provoked by the intervention is essential. The purpose of this study was to delineate ultrastructurally the course of tissue repair following © 1995 Academic Press Limited
312
M E ILIEV
ET A L
lOP [mm Hg]
20 15
t ~
8-~_8,_--=-8_-~_8~,
1
~. O ~O O~O..-.~O/O
R4
20 15
o~58:-=8
:_:~..-:~
•
]
~. _ ~ o -
- o - ~o~
~u~
~" o ~ ~
o - -o~
-o-
_ o
•
No...__o/°""~--o____e/ e ~ o/ R8
2O
,
"/o"
o'__o'__:_
o__o.-O--o
o
o o l . L : :-
o
RIO
o
,,.
°11R
o
.
",
20 15
o
J°o
i\ "°"
o o
j.
•
o_ o
-o
oC
°
.-.-.-.--
I
o
o.._.-o~...o/O~o
1
,..,.
o /
o
O. _',~8 _ -o-
~-~e
l
R16
_o_
0 /
~- ~ o ~ e ~
e~o/e
jo~o--o~
o _ -o e_.._e-..__e ~ o
^
e ~ e ~ ,
"~-~e
i
i
i
i
1
i
i
i
i
i
i
I
I
I
I
I
I
I
I
i
-4
-3
-2
-1
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
i
16 D A Y S
t
oP
FIG. l. Graphs representing individually for five rabbits (R4, RS, RIO, RI2 and R16) the daily IOP measurements of both eyes (O, sclerostomy and O, control eye) 4 days before operation (OP) and over the entire subsequent period prior to death, on days 4, 8, 10, 12 and 16, respectively. Two sclerostomies were created per test eye, the total energy required for the realization of each being 2-4-4.8], and the combined dose received per eye being R4 : 7.2J: R8 : 6"OJ;R10: 7.2J: RI 2 : 7.2J; R16 : 7.2J. The IOP difference between sclerostomy eye and control eye, and the difference between pre- and post-operative IOP level of the operated eye were significant in all rabbits (P < 0.01, Student's t-test). Nd:YAG laser thermal sclerostomy ab interno and to compare it with the post-operative clinical events. 2. Materials and Methods
Ten adult pigmented rabbits (weighing between 3"9 and 5-5 kg) underwent laser surgery in the left eye: the unoperated fellow-eyes served as controls. In each experimental eye. two sclerostomies (one at 10 o'clock and the other at 2 o'clock of the limbus circumference) were created ab interno with a commercial continuous wave 1064 nm Nd:YAG laser (Meridian AG, BernLiebefeld) using the following parameters (optimized in preliminary investigations on cadaver rabbit eyes): pulse duration: 2 0 0 m s : pulse power: 6 W; pulse energy: l'2J; pulse frequency: 2 H z ; spot size: 0.1 m m : pulses lbr perforation: 2 - 4 ; energy range required for perforation: 2.4-4.8J. The 200-#m-
diameter naked quartz optical fiber (Fiberguide, numerical aperture: 0-22) was protracted from a guiding 0 - 6 m m (23-gauge) Atkinson retrobulbar needle mounted on a hand-held probe (see below for details of procedure). In accordance with Swiss Federal Regulations, animals were maintained on a 12-h light-dark cycle and under the care of a qualified person. They were killed randomly at 2-day intervals (including the day of surgery) up to 16 days post-operatively. In addition to this series, one animal was killed on the fifth postoperative day. Intraocular pressure (IOP) was measured on the 4 days prior to, and on the day of surgery, and subsequently on a daily basis until the time of death, using a Digilab pneumotonometer type Micro One. Tonometry was performed on both eyes under topical anesthesia with 0.4% oxybuprocain (Novesin) by the
REPAIR RESPONSE AFTER LASER SCLEROSTOMY
313
FIG. 2. (A) Light micrograph of obliquely-sectioned sclerostomy fistula immediately after surgery. Arrows (- ÷) indicate the depth of the area of collateral thermal damage ensheathing the canal. Z1 (zone 1), the zone of most severe thermal damage exhibiting very intense staining with Toluidine Blue; Z2 (zone 2), the next zone of less severe thermal damage and less intense staining; S, normal sclera; CB, ciliary body. Semithin section (1/¢m). Negative magnification: x 20. Bar, 100/~m. (B) Transmission electron micrograph of the bordering to the lumen (L) zone 1 of the same fistula as in (A): V, vaporization vacuoles; CC, finely granulated mass of coagulated collagen. Negative magnification: x 15 700. Bar. 1 /sm. same person and at the same time on each day (0900 hr). The average of four consecutive measurements was taken as the final value. After the fourth post-operative day, digital pressure was applied to both eyes for 30 sec following the IOP measurement.
Sclerostomy Procedure On the day before surgery, antibiotic eye drops (Spersapolymyxin) were instilled three times at regular intervals in the experimental eye. Preoperatively, miosis was induced with topical 2% pilocarpine to minimize the risk of lens damage. Rabbits were anaesthetized by intramuscular injection of a Xylacin [Rompun: 6 m g ( k g body weight) -1] Ketamin (Narketan; 40 mg kg -1) mixture, and operations performed under sterile conditions. In order to maintain its depth, the anterior chamber was infused with a balanced salt solution (BSS) using a 0.8 mm VenflonPTFE catheter introduced through a paracentesis at the corneal periphery. The conjunctiva was ballooned with cooled (4°C) BSS at the anticipated sclerostomy sites (10 and 2 o'clock) in order to avoid injury by the laser beam; particular care was taken to ensure that the perilimbal region was adequately elevated. The guiding needle enclosing the retracted laser fiber was
inserted 180 ° from the sclerostomy site through a 0"6mm-diameter self-healing paracentesis: it was then advanced parallel to the iris plane and brought into contact with corneo-scleral tissue. The fiber alone was forwarded by gentle manual pressure and pulses of radiation were delivered simultaneously until the fiber tip appeared under the ballooned conjunctiva. The fiber was then retracted into the needle, and the latter withdrawn from the eye. A filtering bleb was formed immediately, thus affirming fistula patency. The lO o'clock sclerostomy was produced first: after stabilization of anterior chamber pressure and depth, the second fistula was created. After gradually decreasing the pressure of the infusion, the latter was discontinued. A drop of gentamicine (Opthagram) and a smear of antibiotic ointment (Neotracin] were placed on the eye and a monoculous made. No steroids were used. Gentamicine and 0"4% oxybuprocain drops were instilled three times daily for 3 days in the operated eye; thereafter no medication was used. Animals were killed by an overdose of pentobarbital administered intravenously. Immediately after death, 0.3 ml of 2.5% glutaraldehyde solution (in 0.1 M sodium cacodylate buffer, pH 7"4) was injected into the anterior chamber of each eye to initiate fixation. Eyes were enucleated and maintained in fixative for
314
M.E. I L I E V ET AL.
FI(;. 3. Transmission electron micrographs of the same fistula as in Fig. 2(A). (A) Longitudinally and obliquely (the bottom lefthand corner) sectioned collagen fibrils in zone 2 of the coagulation sheath. Loss of periodic banding, marked swelling, and minimal or absent interfibrillar space characterize the fibrils in this zone: nevertheless, their orientation within the bundle can still be recognized (arrows). Arrowheads indicate cell remnants. Negative magnification: xlO3OO. Bar. 1/~m. (B) Longitudinally- and cross-sectioned collagen fibrils in the transition zone between zone 2 and the normal sclera: damaged fibrils that have losl their periodic banding alternate with normal, cross-striated ones. Negative magnification: × 20 500. Bar, O' 5 / ~m.
REPAIR RESPON'SE AFTER LASER SCLEROSTOMY
1 hr to harden. They were then dissected (care being taken to avoid tissue dehydration), and zones containing sclerostomy canals as well as control regions were further fixed in 2'5% glutaraldehyde solution overnight. Tissue blocks were washed three times, each for a period of 5 min, in 0' 1 M sodium cacodylate buffer (pH 7"4,340 mosmol), post-fixed in 1% osmium tetroxide solution (340 mosmol) for 2 hr, again rinsed. dehydrated in a graded series of increasing ethanol concentration, and embedded in epoxy resin. Semithin (1/~m) sections were cut on an ultramicrotome and stained with 1% Toluidine Blue for light microscopy. For transmission electron microscopy, ultrathin ( ~ 70 nm) sections were cut, stained with 6% uranyl acetate and lead citrate (after Reynolds), and examined in a Philips EM300 electron microscope,
3. Results
Clinical Course With the exception of one eye, which developed a large intraoperative hyphaema, surgery was performed without complications. Post-operatively, the hyphaemic eye did not produce a filtering bleb, and IOP rose to 35 mmHg; by the sixth day the level had normalized, but this eye was excluded from the final evaluation. On the first post-operative day, all other eyes had restored anterior chambers, well-developed
315
filtering blebs and significantly lowered IOPs. In two cases, no irritation of the anterior eye segment was noted : in all other eyes, a slight or moderate fibrinous exudation was observed within the anterior chamber. A localized anterior synechia developed at the site of paracentesis (for infusion) in four eyes. [nspection of the sclerostomy region revealed a localized nonperforating thermal damage of the iris periphery associated with 14 of the 20 fistulas. No corneal oedema or lens opacification was noted. No symptoms of a deteriorating general status were observed in any of the rabbits during the perioperative period. In all eyes. filtering blebs disappeared within 5 days of surgery although the external fistula was clearly visible under the conjunctiva and IOP remained low. Manual pressure applied for 30 sec daily did not help to prolong the duration of bleb maintenance. On the first post-operative day. IOP was 6-8 mmHg in all eyes (Fig. 1), and throughout the entire observation period for each rabbit, the value for the operated eye was significantly lower than that in the control eye (Student's t-test, P < 0'01). Post-operative IOP levels of the sclerostomy eyes were also significantly lower than their preoperative values (P < 0.01 in all cases).
Morphological Findings For descriptive purposes, the continuous course of
FIG. 4. Transmission electron micrograph of the lumen of sclerostomy fistula 4 days after surgery. Fragments of strongly coagulated collagen (arrowheads) loosened from the canal wall are being phagocytosed. Negative magnification: x 10000. Bar, 1/~m.
316
M.E. I L I E V ET AL.
FIG. 5. (A) Light micrograph {cross section) of the outer opening of a sclerostomy fistula 8 days after surgery. Cells from the episclera (ES) penetrate into the lumen: the integrity of the layer of strongly coagulated collagen (Z1) is undermined. The zone of less severe thermal damage (Z2) exhibits no more high affinity for the stain: it is characterized by a looser tissue structure. and is invaded by cells originating from the surrounding normal sclera (S). Co. conjunctiva. Toluidine Blue: negative
REPAIR RESPON'SE AFTER LASER S C L E R O S T O M Y
317
FIO. 6. Transmission electron micrographs of collagen fibrils [(A) longitudinally-sectioned: (B) cross-sectioned] in zone 2 of the coagulation sheath 6 days after surgery. Periodic banding is observed again: however, numerous electron-loose gaps (arrowheads) between the microfibrils within a fibril suggest that microfibril aggregation is imperfect. In cross-section (B), these gaps render a characteristic (' Swiss cheese') appearance: El, elastic fibrils. Negative magnification: (A) x 32 OOO; (B) × 28 000. Bars, 0"5/zm.
the repair process is subdivided into three (partially overlapping) phases. Phase 1 (days 0-4). Immediately after creation, the transscleral fistula is lined by a n a r r o w layer ( ~ 1 5 2 0 # m in depth) of intensely heat-damaged collagenous tissue w h i c h has a high affinity for the stain toluidine blue [zone 1 in Fig. 2(A)]. Fibrillar structure is completely destroyed and collagenous material appears ultrastructurally as a h o m o g e n e o u s (or finely granulated) electron-dense mass embedded within which are a few cell r e m n a n t s [Fig. 2(B)]. Adjacent to zone 1 is a wider area of less severe (moderate) thermal d a m a g e (zone 2: 8 0 - 9 0 / ~ m in depth) characterized by less intense staining but also by h o m o g e n e o u s appearance in the light microscope [Z2 in Fig. 2(A)]. In electron micrographs of zone 2, fibrillar structures and their orientation within a bundle can still be discerned, but they have u n d e r g o n e
marked swelling and have lost their periodic banding : the interfibrillar space is minimal or absent [Fig. 3(A)], Sclerocytes are destroyed across the whole depth of this zone. In its outer layers [Fig. 3(B)], cross-striated fibrils m a y again be distinguished and mark the transition to normal scleral tissue. During the first 3 - 4 post-operative days, the canal lumen remains cell free. R e m n a n t s of cells and coagulated collagenous tissue begin to loosen from the canal walls. Phase 2 (days 5-7), The first cells are found to penetrate into the canal via its inner and outer openings. These are phagocytes that line the canal walls and begin to digest coagulated material (Fig. 4). Increasing in number, they spread toward the center of the lumen, thus reaching also the loosened particles. Phagocytosing cells near the inner canal opening originate from the iris root and m a y be identified as
magnification: × 33. Bar, 1O0 tlm. (B) The cell network filling the lumen of the fistula in (C) ill higher magnification. Toluidine Blue; negative magnification: x 100. Bar, 20/ml. (C) Light micrograph (cross-section) of the inner opening o[ a sclerostomy fistula 5 days after surgery. Clump cells and fibroblasts stream from the ciliary body (CB) into the canal, line its wall (arrowheads) and occupy the lumen. The sheath of collateral thermal damage (Z1, zone 1, Z2, zone 2) appears cell free. Semithin section (1 inn) stained with Toluidine Blue. Negative magnification: x 33. Bar, l O0/ml.
318
M,E. I L I E V ET AL.
FIG. 7. Light micrograph of cross-sectioned sclerostomy fistula 16 days after surgery. Arrowheads mark the border between normal sclera {S} and the area of original damage (approximately 400/~m in diameter), the latter differing by its loose structure. A rest of coagulated collagen tissue (arrow) is encompassed in a network of cells and blood vessels (asterisks). The border between original lumen and its wall cannot be recognized. Toluidine Blue; negative magnification: x 50. Bar. 100/~m.
clump cells ()wing to their characteristic appearance (high content of pigment granules). As such. they are clearly distinguishable from macrophages originating from the episciera (outer opening of canal). Populations of fibroblasts are observed at both ends of the fistula from where they stream into the lumen (Fig. 5). At the same time. swelling of collagen fibrils in zone 2 of the coagulated tissue sheath decreases, and interfibrillar spaces consequently enlarge: the characteristic periodic banding pattern reappears [Fig. 6(A)]. However. these fibrils differ from normal collagen fibrils by numerous electron loose gaps which, in cross section, render a characteristic ('Swiss cheese') appearance [Fig. 6(B)]. Compared to normal periodicity {640 A), the observed period of banding is decreased {480 A). At this stage, no or very few cells are observed in zone 2 in the light and electron microscope. Phase 3 (days 8-16). Fibroblasts from the surrounding normal sclera invade the outer zone of the coagulation sheath: fibroblasts within the lumen begin to penetrate into the layer of intensely damaged collagen [Fig. 5(A)]: after day 12, its integrity has been completely undermined, remnants of coagulated tissue being surrounded by a network of cells (Fig. 7). The whole fistula is filled with a loose scaffolding of cells. In cross-sections, the original border between the lumen and its wall can no longer be discerned with
clarity. However, the entire area of primary damage (canal and collateral zones together) may be distinguished from the normal tissue by its loose structure (Fig, 7). After day 8. small blood vessels are observed within the lumenal region [Fig. 8(A) and (B)]: the density of this vascular network being m u c h higher than that of the normal sclera. Also found here are polymorphic fibroblasts with a highly developed rough endoplasmatic reticulum and large so-called fibroclasts (Fankhauser et al., 1986: van der Zypen, 1987) containing phagocytosed fragments of mature collagen fibrils [Fig. 8(C) and (D)].
4. Discussion
Although repair is more rapid in rabbits than in h u m a n s (Seetner and Morin, 1979; Skuta and Parish, 1987 : Jaffe et al.. 1988), the data gleaned nonetheless yield valid information on the sequence of events involved in the scarification process. A pigmented species was chosen for this study in order to establish a closer basis for comparison between the experimental model and h u m a n s with respect to uveal pigmentation and radiation-absorbing properties of tissues. The total energy required for perforation of the sclera (2.4 to 4"8]) is comparable with that reported by other
R E P A I R RESPON'SE AFTER LASER S C L E R O S T O M Y
319
Fit;. 8. Micrographs illustrating the repair process in the lmnen of sclerostomy fstulas during phase 3 {8-16 days after surgery). (AI Light micrograph of newly built vessels (arrowheads) in the lumen 10 days alter surgery. \7. an episcleral vein giving sprouts into the lumen. Toluidine Blue: negative magnification: ×132. Bar. 20/ml. (B) Transmission electron micrograph of a venule in the lumen of the same fistula as in A. Arrow, endothelium cell. Negative magnification : × 4600. Bar, 1 /¢m. (C) Transmission electron micrograph of a fibroblast in the lumen 12 days after surgery. Highly developed rough endoplasmatic reticulum (ergastoplasma) implies an active synthesis of collagen precursors. Negative magnification: × 10000. Bar, 1/¢m. (D) Transmission electron micrograph of a fibroclast in the lumen 16 days after surgery. Arrowheads indicate crosssectioned fragments of mature collagen fibrils that have been phagocytosed. Negative magnification: × 20 500. Bar, 0"5/¢m.
320
investigators using a contact Nd:YAG laser in pigmented rabbits (Federman et al., 1987) and is lower than that used in albino rabbits (Higginbotham et al., 1988; Hill et al., 1992). The technique of bringing the fiber tip into contact with the tissue and exerting a gentle pressure while delivering laser pulses (pressure-facilitated laser sclerostomy) is important with the cw-Nd: YAG source (Dfirr et al., 1992). The pressure-facilitated technique compensates for the low water-absorbing properties of Nd:YAG laser light and makes possible the creation of a patent fistula by a few pulses. Dfirr et al. (1992) and England et al. (1994) have shown that the extent of the collateral thermal damage realized by Nd:YAG laser is comparable to the one incurred to the sclera by Ho:YAG and diode laser light despite the difference in the energy required for perforation. Jaffe et al. (1988, 1989) and Wet~el et al. (1990), using argon laser, have reported a similar depth of the sheath of collateral damage around sclerostomy fistulas. Therefore, bearing in mind possible phase-duration variances related to different optical fiber diameters and surgical techniques, the observed course of tissue regeneration may be generally referred to the thermal-mode laser sclerostomies. In one animal, an intraoperative hyphaema accounted for sclerostomy failure. In the other nine eyes, operations could be performed without serious complications. The most frequent side effect was the non-perforating burn of the iris periphery which was most probably due to reflected radiation and seems to be common with this type of laser source (Hill et al., 1992). In terms of clinics, a shortcoming of this side effect of the Nd:YAG irradiation may be seen in provoking of an inflammatory cell reaction in the vicinity of the inner canal opening; a possible advantage might be a decreased incidence of iris prolapse in the inner opening because of a certain flattening of the iris surface (we have observed no iris prolapses in our study). Steroids and antimetabolites were intentionally not used in this study. Filtering blebs disappeared within 5 days of surgery. The time course of bleb failure in our study is comparable to that reported by other authors under similar postoperative medication in rabbits (Jaffe et al., 1988; Karp et al., 1994) and in cynomologous monkeys (Latina et al., 1988). In all our animals, IOP reduction lasted longer than the conjunctival blebs. The same observation has been made also by Higginbotham et al. (1988), Mfiller-Stolzenburg et al. (1991) and Karp et al. (1994). On basis of the morphological findings, we suggest a possible explanation of this phenomenon below. Using the described parameters, pressure-facilitated (contact) Nd:YAG laser sclerostomy leads to the creation of a fistula 2 0 0 - 2 2 0 / ~ m in diameter surrounded by an approximately 100-/tm-thick layer of thermally damaged tissue. The coagulated sheath appears to act as a barrier to cell penetration from the
M. E. I L I E V ET AL.
undamaged sclera during the first 6 - 7 post-operative days. The area of collateral damage may be subdivided into two zones according to the intensity of thermal insult. In the zone abutting on the canal, the collagen fibrils had undergone disintegration into a finely granulated mass of tissue. On the basis of previous studies (van der Zypen, 1985) this morphological manifestation suggests that temperatures as high as 180°C were incurred here. Collagenous tissue which has undergone such thermal insult probably loses the capacity for regeneration since no fibrillar structures were observed in this zone during the post-operative follow-up period. The granular mass was phagocytosed by cells migrating from the iris root and episclera, and after about 2 weeks was completely removed. The tissue next to zone 1 had undergone moderate thermal damage (zone 2, local temperatures in the range of 80-150°C: van der Zypen, 1985). Here, the fibrillar nature of collagen was still recognizable but periodic banding was lost. The course of events observed within this region suggests that in about a week collagen fibrils undergo a process of repolymerization. It might be supposed that, immediately after laser impact, the microfibrils within gross fibrils lose their cross bindings and are displaced laterally to one another, thus causing loss of typical periodicity, but retain their axial orientation. As thermal stress fades in a few days, banding periodicity begins to reappear implying that micro fibrils had reassembled. That the phenomenon observed represents polymerization rather than new synthesis is corroborated by the finding that no cells were observed within zone 2 at this stage. Imperfections in aggregation were evidenced by the presence of numerous gaps between microfibrils (' Swiss cheese' appearance). By the second or third post-operative week, the canal lumen was filled with a framework of proliferating cells and capillaries. The presence of highly developed rough endoplasmatic reticulum (ergastoplasma) in the fibroblasts occupying the lumen and of so called fibroclasts implies that an active turnover of collagen is taking place. The follow-up period of 16 days was too short to provide evidence about the long-term fate of the canal occluding tissue. The major objective of this morphological study was to concentrate on the first postoperative weeks when essential tissue repair reactions are activated and changes are the most intense. The repair response after pressure-facilitated thermal Nd:YAG laser sclerostomy may be summarized as follows: (a) removal of intensely heat-damaged collagenous tissue lining the canal walls by phagocytosis and passive washing out of loosened fragments with the aqueous humor; (b) supposed repolymerization of moderately heat-damaged collagen fibrils; (c) neosynthesis of collagen by fibroblasts migrating into the canal initially from the iris root and episclera and later from the surrounding undamaged
R E P A I R R E S P O N S E AFTER LASER S C L E R O S T O M Y
sclera a n d (d) n e o v a s c u l a r i z a t i o n to form a m u c h denser n e t w o r k of vessels t h a n occurs in the n o r m a l sclera. The clinical discrepancy of persisting low intraocular pressure despite the disappearance of filtering blebs m a y be explained o n the basis of the morphological findings: a l t h o u g h the blebs appeared to be blocked by cells m i g r a t i n g into the o p e n i n g s of the canal, a n d later into its l u m e n , the loose n a t u r e of the occluding cell framework a n d of the s u r r o u n d i n g r e g e n e r a t i n g collagenous tissue, together with the dense n e t w o r k of proliferating capillaries, m i g h t nevertheless allow slow percolation a n d t r a n s p o r t of a q u e o u s h u m o r . F u r t h e r investigations with tracer substances will, however, be necessary to prove this hypothesis.
Acknowledgements The authors would like to thank Ulrich Dtirr and PierreDavid Henchoz (from Meridian AG) for their technical support, and Patricia Huggler for her laboratory expertise. This investigation was supported by grants from the Swiss National Science Foundation (No. 32-36062.92) and the Swiss Commission for the Advancement of Scientific Research (Nos. 2339.1 and 2743.2).
References Brown, R. H., Denham, D. B., Brunner, W. E., Lynch, W. E., Ouigley, H.A. and Parel, ].-M. (1987). Internal sclereetomy for glaucoma filtering surgery with an automated trephine. Arch. Ophthalmol. 105, 133-6. Dfirr, U., Fankhauser, F., England, C., Kwasniewska, S. and van der Zypen, E. (1992). Experimental Nd:YAG diode and Ho:YAG laser sclerostomy performed ab interno and ab externo. Lasers Light Ophthahnol. 5, 1-11. D/irr, U. and Fankhauser, F. (1994). Optical and thermal mechanisms related to the design of laser sclerostomy. German 1. Ophthalmol. 3, 202-11. England, C., van der Zypen, E., Fankhauser, F. and Kwasniewska. S. (1994). Ho:YAG, Nd:YAG and diode laser sclerostomy ab internol: histological and ultrastructural findings in cadavar porcine eyes. Lasers Light Ophthalmol. 6, 183-94. Fankhauser, F., van der Zypen, E. and Kwasniewska, S. (1986). Thermal effects on the trabecular meshwork induced by laser irradiation: clinical implications deduced from ultrastructural studies on the Macaca speciosa monkey. Trans. Ophthalmol. Soc. U.K. 105, 555-61. Federman, ]. L., Wilson, R. P., Fumitaka, A. and Peyman, G.A. (1987). Contact laser: thermal sclerostomy ab interno. Ophthalmic Surg. 18, 726-7. Higginbotham, E.]., Kao, G. and Peyman, G. (1988). Internal sclerostomy with the Nd:YAG contact laser versus thermal sclerostomy in rabbits. Ophthalmology 95, 385-90. Higginbotham, E. ]., Zou, X. and Edward, D. (1990). Internal sclerostomy with the Nd:YAG contact laser versus thermal sclerostomy in cynomolgus monkeys. Lasers Light Ophthalmol. 3, 281-6.
321
Hill, R.A., Ozler, A.S., Baerveldt, G., Viscardi, J.J.. Keates, R.H., Lee, M., Harrington, ]. A. and Berns, M.W. (1992). Ab-interno neodymium:YAG versus erbium:YAG laser sclerostomies in a rabbit model. Ophthalmic Surg. 23, 192L7. Iwach, A. G. and Hoskins, Jr., H. D. (1992). Holmium laser sclerostomy in glaucoma management: a new tool for an old procedure. Bull. Soc. belge Ophtahnol. 244, 187-92. ]affe, G. L., Williams, G. A., Mieler, W. F. and Radius, R. L. (1988). Ab interno sclerostomy with a high-powered argon endolaser. Am J. OphthahnoL 106, 391-6. Jaffe, G.L., Mieler, W.F., Radius, R.L., Kineaid, M.C., Massaro, B. M. and Abrams, G. W. (1989). Ab interno sclerostomy with a high-powered argon endolaser: Clinicopathologic correlation. Arch. Ophthahnol. 107, 1183-5. Kampmeier, ]., Burgass, W., Schfitte. E.. SchrOder, D., Hibst, R. and Steiner, R. (1993). Vergleichende Studie zur Abinterno Sklerostomie mit dem Excimer- und Er:YAGLaser. Ophthahnologe 90, 594-8. Karp, C. L., Higginbotham, E. l., Edward, D.P. and Musch, D. C. (1993). Diode laser surgery: ab interno and ab externo versus conventional surgery in rabbits. Ophthahnology 100, 1567-73. Latina, M., Goode, S., de Kater, A. W., Long. F. H., Deutsch, T. F. and Epstein, D. L. (1988). Experimental ab interno sclerostomies using a pulsed-dye laser. Lasers Surg. Med. 8, 2 3 3 4 0 . M/iller-Stolzenburg, N., yon Haebler, M., Buchwald, H. ]., Bergmann, A. ]., Kruse, H, A. and M/iller, G. ]. (1991). Ab-interno Sklerostomie mit dem Excimerlaser fiber Quarzfaser am Albinokaninchen. Fortschr. Ophthalmol. 88, 762-9. Schuman, ]. S., Stinson, W. G., Hutchinson, B. T., Bellows, A. R., Puliafito, C.A. and Lytle. R. (1993). Holmium laser sclerostomy : success and complications. Ophthalmology 100, 1060-5. Seetner, A. and Morin, ]. D. (1979). Healing of trabeculectomies in rabbits. Can. ]. Ophthahnol. 14, 121-5. Shirato, S., Adachi, M. and Yamashita. H. (1990). Internal sclerostomy with argon contact laser--animal experiment using 5-fiuorouracil. ]pn. ]. Ophthahnol. 34, 381-7. Skuta, G. L. and Parrish II, R. K. (1987). Wound healing in glaucoma filtering surgery. Surv. Ophthahnol. 32, 149-70. Tawakol, M. E., Peyman, G. A., Oncel, M. and Khoobehli, B. (1989). External limbal sclerostomy with contact Nd:YAG laser versus surgical knife. Intern. Ophthahnol. 13, 205-8. van der Zypen, E. (1985). The use of laser in eye surgery: morphological principles. Int. Ophthahnol. Clin. 25, 21-52. van der Zypen, E. (1987). The effects of lasers on outflow structures. In: Glaucoma update Ill, (Ed. Krieglstein, G. K.) Pp. 169-76. Springer-Verlag : Berlin/Heidelberg. Wetzel, W., Schumacher, C., Duncker, G. arid Dreyer, H. P. (1990). Rasterelektronenmikroskopische Befunde nach experimenteller Laser-Sklerostomie ab interno. Fortschr. Ophthalmol. 87, 417-20. Wetzel, W. and Scheu, M. (1993 ). A new application system for laser sclerostomy ab externo. Lasers Light Ophthahnol. 5, 193-8. Wilson, P.R. and ]avitt, ]. C. (1990). Ab interno laser sclerostomy in aphakic patients with glaucoma and chronic inflammation. Am. 1. OphthahnoI. 110, 178-84.
The Repair Response Following N d ' Y A G Laser Sclerostomy Ab Interno in Rabbits MILKO
E. I L I E V ~*, E U G E N V A N
DER Z Y P E N a, F R A N Z ENGLAND a
FANKHAUSER
b
AND CERI
aInstitute of Anatomy, University of Bern," bLindenhof Hospital," Bern, Switzerland (Received Lund 2 February 1995 and accepted in revised form 31 March 1995) This study was undertaken to examine ultrastructurally the course of tissue regeneration after thermallyinduced laser sclerostomy and compare it with the post-operative clinical events in a rabbit model. Applying a continuous wave neodymium:YAG (Nd:YAG) laser, two sclerostomies were created ab interno on one eye in each of ten pigmented rabbits using a 200-/~m-diameter quartz optical fiber; the unoperated fellow-eyes served as controls. Intraocular pressure (IOP) was measured daily before and after surgery. Animals were observed post-operatively for up to 16 days. Well-defined filtering blebs and a low complication rate demonstrated the success of the procedure, lOP was significantly lowered during the entire course of the observation period but after the fifth day, the conjunctival blebs had disappeared. The morphology of the changes in the sclerostomy fistulas were analysed at 2-day intervals by light and electron microscopy. Immediately after surgery, the canal was ensheathed by an approximately lO0-/zmthick layer of coagulated collagenous tissue in which two zones could be distinguished according to the intensity of damage. Within 5 days. the inner and outer canal openings were invaded by macrophages and fibroblasts originating from the iris root and episclera, respectively. The former cells were engaged in the phagocytosis of disintegrated collagen adjacent to the lumen. More distally located fibrils which had incurred less severe damage, had retained their fibrillar structure but had lost banding periodicity. After thermal stress had faded, they appeared to undergo a process of repolymerisation. By day 10, the lumen had become occluded by a loose meshwork of phagocytes, fibroblasts and proliferating capillaries. These new vessels and the loose nature of the canal-occluding framework and of the surrounding regenerating collagenous tissue could have further permitted percolation and transport of aqueous humor, since IOP remained low, despite the disappearance of filtering blebs. Although the time course of repair is more rapid in rabbits than in humans, the data gleaned nonetheless yield valid information respecting the sequence of events following thermally-induced scleral fistula. © 1995 Academic Press Limited Key words: sclerostomy; ab interno: laser: Nd:YAG laser: rabbits: healing process: thermal damage; ultrastructure: glaucoma: intraocular pressure.
1. Introduction Apart from animal experiments, full-thickness sclerostomy has been investigated in some clinical trials for the surgical treatment of refractory g l a u c o m a with promising results (Wilson and Javitt, 1 9 9 0 : I w a c h and Hoskins. 1 9 9 2 ; Kampmeier et al., 1 9 9 3 ; S c h u m a n et al., 1993). This minor and low-traumatic surgical intervention m a y be achieved by m e a n s of different lasers (Jaffe et al., 1 9 8 8 : Mtiller-Stolzenburg et al. 1991 ; Hill et al., 1 9 9 2 : Karp et al., 1993) or by sophisticated surgical equipment (Brown et al., 1987). However, the most c o m m o n long-term post-operative complication of filtering s u r g e r y - - s c a r r i n g with consequent bleb f a i l u r e - - c a n n o t be avoided by a n y of the sclerostomy techniques currently employed. Nonetheless, some authors have found that thermallaser sclerostomies (cw-Nd:YAG, and diode laser) in rabbits and m o n k e y s maintain patency longer and exhibit less inflammatory reaction t h a n conventional * For correspondence at: Institute of Anatomy, University of Bern. Department of Topographical Anatomy, B6hlstrasse 26, CH3000 Bern 9, Switzerland. 0 0 1 4 - 4 8 3 5 / 9 5 / 0 9 0 3 1 1 + 11 $12.00/0
procedure filters (Higginbotham et al., 1988, 1 9 9 0 : Karp et al., 1993). It has been discussed that collateral thermal d a m a g e produced in the walls of scleral fistulas by laser radiation m a y delay the repair response and thus prolong bleb survival, a l t h o u g h this question remains controversial (Hill et al., 1 9 9 2 ; Wetzel and Scheu, 1 9 9 3 ; D~irr and Fankhauser, 1994). Thermal-mode lasers (leading to evaporation and coagulation of tissue), rather than Q-switchedmode lasers (very short pulses of high energy causing optical breakdown and mechanical tearing of tissue), have the advantage of coagulating blood vessels and reducing the risk of bleeding, but thermally-induced sclerostomies also close with time. In an attempt to reduce post-operative scarring, steroids and antimetabolites have been applied with moderate success (Tawakol et al., 1 9 8 9 ; Shirato et al., 1 9 9 0 ; Karp et al., 1994). However, before interference with the undesirable process of canal obliteration can hope to be really successful, a detailed analysis of the course of events provoked by the intervention is essential. The purpose of this study was to delineate ultrastructurally the course of tissue repair following © 1995 Academic Press Limited
312
M E ILIEV
ET A L
lOP [mm Hg]
20 15
t ~
8-~_8,_--=-8_-~_8~,
1
~. O ~O O~O..-.~O/O
R4
20 15
o~58:-=8
:_:~..-:~
•
]
~. _ ~ o -
- o - ~o~
~u~
~" o ~ ~
o - -o~
-o-
_ o
•
No...__o/°""~--o____e/ e ~ o/ R8
2O
,
"/o"
o'__o'__:_
o__o.-O--o
o
o o l . L : :-
o
RIO
o
,,.
°11R
o
.
",
20 15
o
J°o
i\ "°"
o o
j.
•
o_ o
-o
oC
°
.-.-.-.--
I
o
o.._.-o~...o/O~o
1
,..,.
o /
o
O. _',~8 _ -o-
~-~e
l
R16
_o_
0 /
~- ~ o ~ e ~
e~o/e
jo~o--o~
o _ -o e_.._e-..__e ~ o
^
e ~ e ~ ,
"~-~e
i
i
i
i
1
i
i
i
i
i
i
I
I
I
I
I
I
I
I
i
-4
-3
-2
-1
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
i
16 D A Y S
t
oP
FIG. l. Graphs representing individually for five rabbits (R4, RS, RIO, RI2 and R16) the daily IOP measurements of both eyes (O, sclerostomy and O, control eye) 4 days before operation (OP) and over the entire subsequent period prior to death, on days 4, 8, 10, 12 and 16, respectively. Two sclerostomies were created per test eye, the total energy required for the realization of each being 2-4-4.8], and the combined dose received per eye being R4 : 7.2J: R8 : 6"OJ;R10: 7.2J: RI 2 : 7.2J; R16 : 7.2J. The IOP difference between sclerostomy eye and control eye, and the difference between pre- and post-operative IOP level of the operated eye were significant in all rabbits (P < 0.01, Student's t-test). Nd:YAG laser thermal sclerostomy ab interno and to compare it with the post-operative clinical events. 2. Materials and Methods
Ten adult pigmented rabbits (weighing between 3"9 and 5-5 kg) underwent laser surgery in the left eye: the unoperated fellow-eyes served as controls. In each experimental eye. two sclerostomies (one at 10 o'clock and the other at 2 o'clock of the limbus circumference) were created ab interno with a commercial continuous wave 1064 nm Nd:YAG laser (Meridian AG, BernLiebefeld) using the following parameters (optimized in preliminary investigations on cadaver rabbit eyes): pulse duration: 2 0 0 m s : pulse power: 6 W; pulse energy: l'2J; pulse frequency: 2 H z ; spot size: 0.1 m m : pulses lbr perforation: 2 - 4 ; energy range required for perforation: 2.4-4.8J. The 200-#m-
diameter naked quartz optical fiber (Fiberguide, numerical aperture: 0-22) was protracted from a guiding 0 - 6 m m (23-gauge) Atkinson retrobulbar needle mounted on a hand-held probe (see below for details of procedure). In accordance with Swiss Federal Regulations, animals were maintained on a 12-h light-dark cycle and under the care of a qualified person. They were killed randomly at 2-day intervals (including the day of surgery) up to 16 days post-operatively. In addition to this series, one animal was killed on the fifth postoperative day. Intraocular pressure (IOP) was measured on the 4 days prior to, and on the day of surgery, and subsequently on a daily basis until the time of death, using a Digilab pneumotonometer type Micro One. Tonometry was performed on both eyes under topical anesthesia with 0.4% oxybuprocain (Novesin) by the
REPAIR RESPONSE AFTER LASER SCLEROSTOMY
313
FIG. 2. (A) Light micrograph of obliquely-sectioned sclerostomy fistula immediately after surgery. Arrows (- ÷) indicate the depth of the area of collateral thermal damage ensheathing the canal. Z1 (zone 1), the zone of most severe thermal damage exhibiting very intense staining with Toluidine Blue; Z2 (zone 2), the next zone of less severe thermal damage and less intense staining; S, normal sclera; CB, ciliary body. Semithin section (1/¢m). Negative magnification: x 20. Bar, 100/~m. (B) Transmission electron micrograph of the bordering to the lumen (L) zone 1 of the same fistula as in (A): V, vaporization vacuoles; CC, finely granulated mass of coagulated collagen. Negative magnification: x 15 700. Bar. 1 /sm. same person and at the same time on each day (0900 hr). The average of four consecutive measurements was taken as the final value. After the fourth post-operative day, digital pressure was applied to both eyes for 30 sec following the IOP measurement.
Sclerostomy Procedure On the day before surgery, antibiotic eye drops (Spersapolymyxin) were instilled three times at regular intervals in the experimental eye. Preoperatively, miosis was induced with topical 2% pilocarpine to minimize the risk of lens damage. Rabbits were anaesthetized by intramuscular injection of a Xylacin [Rompun: 6 m g ( k g body weight) -1] Ketamin (Narketan; 40 mg kg -1) mixture, and operations performed under sterile conditions. In order to maintain its depth, the anterior chamber was infused with a balanced salt solution (BSS) using a 0.8 mm VenflonPTFE catheter introduced through a paracentesis at the corneal periphery. The conjunctiva was ballooned with cooled (4°C) BSS at the anticipated sclerostomy sites (10 and 2 o'clock) in order to avoid injury by the laser beam; particular care was taken to ensure that the perilimbal region was adequately elevated. The guiding needle enclosing the retracted laser fiber was
inserted 180 ° from the sclerostomy site through a 0"6mm-diameter self-healing paracentesis: it was then advanced parallel to the iris plane and brought into contact with corneo-scleral tissue. The fiber alone was forwarded by gentle manual pressure and pulses of radiation were delivered simultaneously until the fiber tip appeared under the ballooned conjunctiva. The fiber was then retracted into the needle, and the latter withdrawn from the eye. A filtering bleb was formed immediately, thus affirming fistula patency. The lO o'clock sclerostomy was produced first: after stabilization of anterior chamber pressure and depth, the second fistula was created. After gradually decreasing the pressure of the infusion, the latter was discontinued. A drop of gentamicine (Opthagram) and a smear of antibiotic ointment (Neotracin] were placed on the eye and a monoculous made. No steroids were used. Gentamicine and 0"4% oxybuprocain drops were instilled three times daily for 3 days in the operated eye; thereafter no medication was used. Animals were killed by an overdose of pentobarbital administered intravenously. Immediately after death, 0.3 ml of 2.5% glutaraldehyde solution (in 0.1 M sodium cacodylate buffer, pH 7"4) was injected into the anterior chamber of each eye to initiate fixation. Eyes were enucleated and maintained in fixative for
314
M.E. I L I E V ET AL.
FI(;. 3. Transmission electron micrographs of the same fistula as in Fig. 2(A). (A) Longitudinally and obliquely (the bottom lefthand corner) sectioned collagen fibrils in zone 2 of the coagulation sheath. Loss of periodic banding, marked swelling, and minimal or absent interfibrillar space characterize the fibrils in this zone: nevertheless, their orientation within the bundle can still be recognized (arrows). Arrowheads indicate cell remnants. Negative magnification: xlO3OO. Bar. 1/~m. (B) Longitudinally- and cross-sectioned collagen fibrils in the transition zone between zone 2 and the normal sclera: damaged fibrils that have losl their periodic banding alternate with normal, cross-striated ones. Negative magnification: × 20 500. Bar, O' 5 / ~m.
REPAIR RESPON'SE AFTER LASER SCLEROSTOMY
1 hr to harden. They were then dissected (care being taken to avoid tissue dehydration), and zones containing sclerostomy canals as well as control regions were further fixed in 2'5% glutaraldehyde solution overnight. Tissue blocks were washed three times, each for a period of 5 min, in 0' 1 M sodium cacodylate buffer (pH 7"4,340 mosmol), post-fixed in 1% osmium tetroxide solution (340 mosmol) for 2 hr, again rinsed. dehydrated in a graded series of increasing ethanol concentration, and embedded in epoxy resin. Semithin (1/~m) sections were cut on an ultramicrotome and stained with 1% Toluidine Blue for light microscopy. For transmission electron microscopy, ultrathin ( ~ 70 nm) sections were cut, stained with 6% uranyl acetate and lead citrate (after Reynolds), and examined in a Philips EM300 electron microscope,
3. Results
Clinical Course With the exception of one eye, which developed a large intraoperative hyphaema, surgery was performed without complications. Post-operatively, the hyphaemic eye did not produce a filtering bleb, and IOP rose to 35 mmHg; by the sixth day the level had normalized, but this eye was excluded from the final evaluation. On the first post-operative day, all other eyes had restored anterior chambers, well-developed
315
filtering blebs and significantly lowered IOPs. In two cases, no irritation of the anterior eye segment was noted : in all other eyes, a slight or moderate fibrinous exudation was observed within the anterior chamber. A localized anterior synechia developed at the site of paracentesis (for infusion) in four eyes. [nspection of the sclerostomy region revealed a localized nonperforating thermal damage of the iris periphery associated with 14 of the 20 fistulas. No corneal oedema or lens opacification was noted. No symptoms of a deteriorating general status were observed in any of the rabbits during the perioperative period. In all eyes. filtering blebs disappeared within 5 days of surgery although the external fistula was clearly visible under the conjunctiva and IOP remained low. Manual pressure applied for 30 sec daily did not help to prolong the duration of bleb maintenance. On the first post-operative day. IOP was 6-8 mmHg in all eyes (Fig. 1), and throughout the entire observation period for each rabbit, the value for the operated eye was significantly lower than that in the control eye (Student's t-test, P < 0'01). Post-operative IOP levels of the sclerostomy eyes were also significantly lower than their preoperative values (P < 0.01 in all cases).
Morphological Findings For descriptive purposes, the continuous course of
FIG. 4. Transmission electron micrograph of the lumen of sclerostomy fistula 4 days after surgery. Fragments of strongly coagulated collagen (arrowheads) loosened from the canal wall are being phagocytosed. Negative magnification: x 10000. Bar, 1/~m.
316
M.E. I L I E V ET AL.
FIG. 5. (A) Light micrograph {cross section) of the outer opening of a sclerostomy fistula 8 days after surgery. Cells from the episclera (ES) penetrate into the lumen: the integrity of the layer of strongly coagulated collagen (Z1) is undermined. The zone of less severe thermal damage (Z2) exhibits no more high affinity for the stain: it is characterized by a looser tissue structure. and is invaded by cells originating from the surrounding normal sclera (S). Co. conjunctiva. Toluidine Blue: negative
REPAIR RESPON'SE AFTER LASER S C L E R O S T O M Y
317
FIO. 6. Transmission electron micrographs of collagen fibrils [(A) longitudinally-sectioned: (B) cross-sectioned] in zone 2 of the coagulation sheath 6 days after surgery. Periodic banding is observed again: however, numerous electron-loose gaps (arrowheads) between the microfibrils within a fibril suggest that microfibril aggregation is imperfect. In cross-section (B), these gaps render a characteristic (' Swiss cheese') appearance: El, elastic fibrils. Negative magnification: (A) x 32 OOO; (B) × 28 000. Bars, 0"5/zm.
the repair process is subdivided into three (partially overlapping) phases. Phase 1 (days 0-4). Immediately after creation, the transscleral fistula is lined by a n a r r o w layer ( ~ 1 5 2 0 # m in depth) of intensely heat-damaged collagenous tissue w h i c h has a high affinity for the stain toluidine blue [zone 1 in Fig. 2(A)]. Fibrillar structure is completely destroyed and collagenous material appears ultrastructurally as a h o m o g e n e o u s (or finely granulated) electron-dense mass embedded within which are a few cell r e m n a n t s [Fig. 2(B)]. Adjacent to zone 1 is a wider area of less severe (moderate) thermal d a m a g e (zone 2: 8 0 - 9 0 / ~ m in depth) characterized by less intense staining but also by h o m o g e n e o u s appearance in the light microscope [Z2 in Fig. 2(A)]. In electron micrographs of zone 2, fibrillar structures and their orientation within a bundle can still be discerned, but they have u n d e r g o n e
marked swelling and have lost their periodic banding : the interfibrillar space is minimal or absent [Fig. 3(A)], Sclerocytes are destroyed across the whole depth of this zone. In its outer layers [Fig. 3(B)], cross-striated fibrils m a y again be distinguished and mark the transition to normal scleral tissue. During the first 3 - 4 post-operative days, the canal lumen remains cell free. R e m n a n t s of cells and coagulated collagenous tissue begin to loosen from the canal walls. Phase 2 (days 5-7), The first cells are found to penetrate into the canal via its inner and outer openings. These are phagocytes that line the canal walls and begin to digest coagulated material (Fig. 4). Increasing in number, they spread toward the center of the lumen, thus reaching also the loosened particles. Phagocytosing cells near the inner canal opening originate from the iris root and m a y be identified as
magnification: × 33. Bar, 1O0 tlm. (B) The cell network filling the lumen of the fistula in (C) ill higher magnification. Toluidine Blue; negative magnification: x 100. Bar, 20/ml. (C) Light micrograph (cross-section) of the inner opening o[ a sclerostomy fistula 5 days after surgery. Clump cells and fibroblasts stream from the ciliary body (CB) into the canal, line its wall (arrowheads) and occupy the lumen. The sheath of collateral thermal damage (Z1, zone 1, Z2, zone 2) appears cell free. Semithin section (1 inn) stained with Toluidine Blue. Negative magnification: x 33. Bar, l O0/ml.
318
M,E. I L I E V ET AL.
FIG. 7. Light micrograph of cross-sectioned sclerostomy fistula 16 days after surgery. Arrowheads mark the border between normal sclera {S} and the area of original damage (approximately 400/~m in diameter), the latter differing by its loose structure. A rest of coagulated collagen tissue (arrow) is encompassed in a network of cells and blood vessels (asterisks). The border between original lumen and its wall cannot be recognized. Toluidine Blue; negative magnification: x 50. Bar. 100/~m.
clump cells ()wing to their characteristic appearance (high content of pigment granules). As such. they are clearly distinguishable from macrophages originating from the episciera (outer opening of canal). Populations of fibroblasts are observed at both ends of the fistula from where they stream into the lumen (Fig. 5). At the same time. swelling of collagen fibrils in zone 2 of the coagulated tissue sheath decreases, and interfibrillar spaces consequently enlarge: the characteristic periodic banding pattern reappears [Fig. 6(A)]. However. these fibrils differ from normal collagen fibrils by numerous electron loose gaps which, in cross section, render a characteristic ('Swiss cheese') appearance [Fig. 6(B)]. Compared to normal periodicity {640 A), the observed period of banding is decreased {480 A). At this stage, no or very few cells are observed in zone 2 in the light and electron microscope. Phase 3 (days 8-16). Fibroblasts from the surrounding normal sclera invade the outer zone of the coagulation sheath: fibroblasts within the lumen begin to penetrate into the layer of intensely damaged collagen [Fig. 5(A)]: after day 12, its integrity has been completely undermined, remnants of coagulated tissue being surrounded by a network of cells (Fig. 7). The whole fistula is filled with a loose scaffolding of cells. In cross-sections, the original border between the lumen and its wall can no longer be discerned with
clarity. However, the entire area of primary damage (canal and collateral zones together) may be distinguished from the normal tissue by its loose structure (Fig, 7). After day 8. small blood vessels are observed within the lumenal region [Fig. 8(A) and (B)]: the density of this vascular network being m u c h higher than that of the normal sclera. Also found here are polymorphic fibroblasts with a highly developed rough endoplasmatic reticulum and large so-called fibroclasts (Fankhauser et al., 1986: van der Zypen, 1987) containing phagocytosed fragments of mature collagen fibrils [Fig. 8(C) and (D)].
4. Discussion
Although repair is more rapid in rabbits than in h u m a n s (Seetner and Morin, 1979; Skuta and Parish, 1987 : Jaffe et al.. 1988), the data gleaned nonetheless yield valid information on the sequence of events involved in the scarification process. A pigmented species was chosen for this study in order to establish a closer basis for comparison between the experimental model and h u m a n s with respect to uveal pigmentation and radiation-absorbing properties of tissues. The total energy required for perforation of the sclera (2.4 to 4"8]) is comparable with that reported by other
R E P A I R RESPON'SE AFTER LASER S C L E R O S T O M Y
319
Fit;. 8. Micrographs illustrating the repair process in the lmnen of sclerostomy fstulas during phase 3 {8-16 days after surgery). (AI Light micrograph of newly built vessels (arrowheads) in the lumen 10 days alter surgery. \7. an episcleral vein giving sprouts into the lumen. Toluidine Blue: negative magnification: ×132. Bar. 20/ml. (B) Transmission electron micrograph of a venule in the lumen of the same fistula as in A. Arrow, endothelium cell. Negative magnification : × 4600. Bar, 1 /¢m. (C) Transmission electron micrograph of a fibroblast in the lumen 12 days after surgery. Highly developed rough endoplasmatic reticulum (ergastoplasma) implies an active synthesis of collagen precursors. Negative magnification: × 10000. Bar, 1/¢m. (D) Transmission electron micrograph of a fibroclast in the lumen 16 days after surgery. Arrowheads indicate crosssectioned fragments of mature collagen fibrils that have been phagocytosed. Negative magnification: × 20 500. Bar, 0"5/¢m.
320
investigators using a contact Nd:YAG laser in pigmented rabbits (Federman et al., 1987) and is lower than that used in albino rabbits (Higginbotham et al., 1988; Hill et al., 1992). The technique of bringing the fiber tip into contact with the tissue and exerting a gentle pressure while delivering laser pulses (pressure-facilitated laser sclerostomy) is important with the cw-Nd: YAG source (Dfirr et al., 1992). The pressure-facilitated technique compensates for the low water-absorbing properties of Nd:YAG laser light and makes possible the creation of a patent fistula by a few pulses. Dfirr et al. (1992) and England et al. (1994) have shown that the extent of the collateral thermal damage realized by Nd:YAG laser is comparable to the one incurred to the sclera by Ho:YAG and diode laser light despite the difference in the energy required for perforation. Jaffe et al. (1988, 1989) and Wet~el et al. (1990), using argon laser, have reported a similar depth of the sheath of collateral damage around sclerostomy fistulas. Therefore, bearing in mind possible phase-duration variances related to different optical fiber diameters and surgical techniques, the observed course of tissue regeneration may be generally referred to the thermal-mode laser sclerostomies. In one animal, an intraoperative hyphaema accounted for sclerostomy failure. In the other nine eyes, operations could be performed without serious complications. The most frequent side effect was the non-perforating burn of the iris periphery which was most probably due to reflected radiation and seems to be common with this type of laser source (Hill et al., 1992). In terms of clinics, a shortcoming of this side effect of the Nd:YAG irradiation may be seen in provoking of an inflammatory cell reaction in the vicinity of the inner canal opening; a possible advantage might be a decreased incidence of iris prolapse in the inner opening because of a certain flattening of the iris surface (we have observed no iris prolapses in our study). Steroids and antimetabolites were intentionally not used in this study. Filtering blebs disappeared within 5 days of surgery. The time course of bleb failure in our study is comparable to that reported by other authors under similar postoperative medication in rabbits (Jaffe et al., 1988; Karp et al., 1994) and in cynomologous monkeys (Latina et al., 1988). In all our animals, IOP reduction lasted longer than the conjunctival blebs. The same observation has been made also by Higginbotham et al. (1988), Mfiller-Stolzenburg et al. (1991) and Karp et al. (1994). On basis of the morphological findings, we suggest a possible explanation of this phenomenon below. Using the described parameters, pressure-facilitated (contact) Nd:YAG laser sclerostomy leads to the creation of a fistula 2 0 0 - 2 2 0 / ~ m in diameter surrounded by an approximately 100-/tm-thick layer of thermally damaged tissue. The coagulated sheath appears to act as a barrier to cell penetration from the
M. E. I L I E V ET AL.
undamaged sclera during the first 6 - 7 post-operative days. The area of collateral damage may be subdivided into two zones according to the intensity of thermal insult. In the zone abutting on the canal, the collagen fibrils had undergone disintegration into a finely granulated mass of tissue. On the basis of previous studies (van der Zypen, 1985) this morphological manifestation suggests that temperatures as high as 180°C were incurred here. Collagenous tissue which has undergone such thermal insult probably loses the capacity for regeneration since no fibrillar structures were observed in this zone during the post-operative follow-up period. The granular mass was phagocytosed by cells migrating from the iris root and episclera, and after about 2 weeks was completely removed. The tissue next to zone 1 had undergone moderate thermal damage (zone 2, local temperatures in the range of 80-150°C: van der Zypen, 1985). Here, the fibrillar nature of collagen was still recognizable but periodic banding was lost. The course of events observed within this region suggests that in about a week collagen fibrils undergo a process of repolymerization. It might be supposed that, immediately after laser impact, the microfibrils within gross fibrils lose their cross bindings and are displaced laterally to one another, thus causing loss of typical periodicity, but retain their axial orientation. As thermal stress fades in a few days, banding periodicity begins to reappear implying that micro fibrils had reassembled. That the phenomenon observed represents polymerization rather than new synthesis is corroborated by the finding that no cells were observed within zone 2 at this stage. Imperfections in aggregation were evidenced by the presence of numerous gaps between microfibrils (' Swiss cheese' appearance). By the second or third post-operative week, the canal lumen was filled with a framework of proliferating cells and capillaries. The presence of highly developed rough endoplasmatic reticulum (ergastoplasma) in the fibroblasts occupying the lumen and of so called fibroclasts implies that an active turnover of collagen is taking place. The follow-up period of 16 days was too short to provide evidence about the long-term fate of the canal occluding tissue. The major objective of this morphological study was to concentrate on the first postoperative weeks when essential tissue repair reactions are activated and changes are the most intense. The repair response after pressure-facilitated thermal Nd:YAG laser sclerostomy may be summarized as follows: (a) removal of intensely heat-damaged collagenous tissue lining the canal walls by phagocytosis and passive washing out of loosened fragments with the aqueous humor; (b) supposed repolymerization of moderately heat-damaged collagen fibrils; (c) neosynthesis of collagen by fibroblasts migrating into the canal initially from the iris root and episclera and later from the surrounding undamaged
R E P A I R R E S P O N S E AFTER LASER S C L E R O S T O M Y
sclera a n d (d) n e o v a s c u l a r i z a t i o n to form a m u c h denser n e t w o r k of vessels t h a n occurs in the n o r m a l sclera. The clinical discrepancy of persisting low intraocular pressure despite the disappearance of filtering blebs m a y be explained o n the basis of the morphological findings: a l t h o u g h the blebs appeared to be blocked by cells m i g r a t i n g into the o p e n i n g s of the canal, a n d later into its l u m e n , the loose n a t u r e of the occluding cell framework a n d of the s u r r o u n d i n g r e g e n e r a t i n g collagenous tissue, together with the dense n e t w o r k of proliferating capillaries, m i g h t nevertheless allow slow percolation a n d t r a n s p o r t of a q u e o u s h u m o r . F u r t h e r investigations with tracer substances will, however, be necessary to prove this hypothesis.
Acknowledgements The authors would like to thank Ulrich Dtirr and PierreDavid Henchoz (from Meridian AG) for their technical support, and Patricia Huggler for her laboratory expertise. This investigation was supported by grants from the Swiss National Science Foundation (No. 32-36062.92) and the Swiss Commission for the Advancement of Scientific Research (Nos. 2339.1 and 2743.2).
References Brown, R. H., Denham, D. B., Brunner, W. E., Lynch, W. E., Ouigley, H.A. and Parel, ].-M. (1987). Internal sclereetomy for glaucoma filtering surgery with an automated trephine. Arch. Ophthalmol. 105, 133-6. Dfirr, U., Fankhauser, F., England, C., Kwasniewska, S. and van der Zypen, E. (1992). Experimental Nd:YAG diode and Ho:YAG laser sclerostomy performed ab interno and ab externo. Lasers Light Ophthahnol. 5, 1-11. D/irr, U. and Fankhauser, F. (1994). Optical and thermal mechanisms related to the design of laser sclerostomy. German 1. Ophthalmol. 3, 202-11. England, C., van der Zypen, E., Fankhauser, F. and Kwasniewska. S. (1994). Ho:YAG, Nd:YAG and diode laser sclerostomy ab internol: histological and ultrastructural findings in cadavar porcine eyes. Lasers Light Ophthalmol. 6, 183-94. Fankhauser, F., van der Zypen, E. and Kwasniewska, S. (1986). Thermal effects on the trabecular meshwork induced by laser irradiation: clinical implications deduced from ultrastructural studies on the Macaca speciosa monkey. Trans. Ophthalmol. Soc. U.K. 105, 555-61. Federman, ]. L., Wilson, R. P., Fumitaka, A. and Peyman, G.A. (1987). Contact laser: thermal sclerostomy ab interno. Ophthalmic Surg. 18, 726-7. Higginbotham, E.]., Kao, G. and Peyman, G. (1988). Internal sclerostomy with the Nd:YAG contact laser versus thermal sclerostomy in rabbits. Ophthalmology 95, 385-90. Higginbotham, E. ]., Zou, X. and Edward, D. (1990). Internal sclerostomy with the Nd:YAG contact laser versus thermal sclerostomy in cynomolgus monkeys. Lasers Light Ophthalmol. 3, 281-6.
321
Hill, R.A., Ozler, A.S., Baerveldt, G., Viscardi, J.J.. Keates, R.H., Lee, M., Harrington, ]. A. and Berns, M.W. (1992). Ab-interno neodymium:YAG versus erbium:YAG laser sclerostomies in a rabbit model. Ophthalmic Surg. 23, 192L7. Iwach, A. G. and Hoskins, Jr., H. D. (1992). Holmium laser sclerostomy in glaucoma management: a new tool for an old procedure. Bull. Soc. belge Ophtahnol. 244, 187-92. ]affe, G. L., Williams, G. A., Mieler, W. F. and Radius, R. L. (1988). Ab interno sclerostomy with a high-powered argon endolaser. Am J. OphthahnoL 106, 391-6. Jaffe, G.L., Mieler, W.F., Radius, R.L., Kineaid, M.C., Massaro, B. M. and Abrams, G. W. (1989). Ab interno sclerostomy with a high-powered argon endolaser: Clinicopathologic correlation. Arch. Ophthahnol. 107, 1183-5. Kampmeier, ]., Burgass, W., Schfitte. E.. SchrOder, D., Hibst, R. and Steiner, R. (1993). Vergleichende Studie zur Abinterno Sklerostomie mit dem Excimer- und Er:YAGLaser. Ophthahnologe 90, 594-8. Karp, C. L., Higginbotham, E. l., Edward, D.P. and Musch, D. C. (1993). Diode laser surgery: ab interno and ab externo versus conventional surgery in rabbits. Ophthahnology 100, 1567-73. Latina, M., Goode, S., de Kater, A. W., Long. F. H., Deutsch, T. F. and Epstein, D. L. (1988). Experimental ab interno sclerostomies using a pulsed-dye laser. Lasers Surg. Med. 8, 2 3 3 4 0 . M/iller-Stolzenburg, N., yon Haebler, M., Buchwald, H. ]., Bergmann, A. ]., Kruse, H, A. and M/iller, G. ]. (1991). Ab-interno Sklerostomie mit dem Excimerlaser fiber Quarzfaser am Albinokaninchen. Fortschr. Ophthalmol. 88, 762-9. Schuman, ]. S., Stinson, W. G., Hutchinson, B. T., Bellows, A. R., Puliafito, C.A. and Lytle. R. (1993). Holmium laser sclerostomy : success and complications. Ophthalmology 100, 1060-5. Seetner, A. and Morin, ]. D. (1979). Healing of trabeculectomies in rabbits. Can. ]. Ophthahnol. 14, 121-5. Shirato, S., Adachi, M. and Yamashita. H. (1990). Internal sclerostomy with argon contact laser--animal experiment using 5-fiuorouracil. ]pn. ]. Ophthahnol. 34, 381-7. Skuta, G. L. and Parrish II, R. K. (1987). Wound healing in glaucoma filtering surgery. Surv. Ophthahnol. 32, 149-70. Tawakol, M. E., Peyman, G. A., Oncel, M. and Khoobehli, B. (1989). External limbal sclerostomy with contact Nd:YAG laser versus surgical knife. Intern. Ophthahnol. 13, 205-8. van der Zypen, E. (1985). The use of laser in eye surgery: morphological principles. Int. Ophthahnol. Clin. 25, 21-52. van der Zypen, E. (1987). The effects of lasers on outflow structures. In: Glaucoma update Ill, (Ed. Krieglstein, G. K.) Pp. 169-76. Springer-Verlag : Berlin/Heidelberg. Wetzel, W., Schumacher, C., Duncker, G. arid Dreyer, H. P. (1990). Rasterelektronenmikroskopische Befunde nach experimenteller Laser-Sklerostomie ab interno. Fortschr. Ophthalmol. 87, 417-20. Wetzel, W. and Scheu, M. (1993 ). A new application system for laser sclerostomy ab externo. Lasers Light Ophthahnol. 5, 193-8. Wilson, P.R. and ]avitt, ]. C. (1990). Ab interno laser sclerostomy in aphakic patients with glaucoma and chronic inflammation. Am. 1. OphthahnoI. 110, 178-84.