Microkeratome-Assisted Lamellar

Keratoplasty for the Surgical Treatment

of Keratoconus
Massimo Busin, MD,1 Luca Zambianchi, MD,1 Robert C. Arffa, MD2

Purpose: To evaluate the visual and refractive results of microkeratome-assisted lamellar keratoplasty (LK)
performed on keratoconus patients intolerant to spectacles and contact lenses.
Design: Prospective, noncomparative, interventional study.
Participants: A microkeratome-assisted LK procedure was performed on 50 eyes of 50 keratoconus
patients. All patients were spectacle and contact lens intolerant.
Intervention: All patients included in this study underwent a standard surgical procedure involving removal
of a lamella (9 mm in diameter cut with the 250-␮m microkeratome head) from the recipient cornea by means of
a hand-driven microkeratome and suturing of a donor lamella (0.5 mm smaller in diameter than the removed
corneal lamella, cut with the 350-␮m microkeratome head) obtained from a cornea mounted on an artificial
anterior chamber. Each patient was examined preoperatively and at different postoperative times (1 and 6
months and 1, 2, 3, and 4 years).
Main Outcome Measures: Uncorrected visual acuity (UCVA), best spectacle-corrected visual acuity (BSCVA),
1-year best contact lens– corrected visual acuity (BCLCVA), refraction, and computerized analysis of corneal topog-
raphy.
Results: After suture removal was completed, both UCVA and best-corrected visual acuity were significantly
improved over properative values at all examination times. One year postoperatively, when follow-up was still
available for all patients, UCVA was better than 20/200 in 8 of 50 (16%) patients and BSCVA was ⱖ20/40 in 44
of 50 (88%) patients, whereas BCLCVA was ⱖ20/40 in all 50 patients. Refractive astigmatism within 4 diopters
was seen in 43 of 50 (86%) patients. Corneal topographic patterns were classified as regularly astigmatic in 39
of 50 (78%) patients. The 1-year values did not change substantially at later postoperative examination times.
Complications included preparation of donor grafts of poor quality that needed to be discarded (8 cases [16%]),
irregular astigmatism of various degrees (11 cases [22%]), high-degree astigmatism requiring secondary inter-
vention (6 cases [12%]), epithelial interface ingrowth (1 case [2%]), and cataract formation (1 case [2%]).
Conclusions: Microkeratome-assisted LK can be performed on corneas with moderate to advanced keratoco-
nus with a minimal corneal thickness of ⬎380 ␮m. The procedure is relatively simple, may be standardized in most
of its parts, and does not involve time-consuming maneuvers. All complications recorded did not threaten vision and
were dealt with successfully. Our results indicate that microkeratome-assisted LK is as efficacious as conventional
penetrating keratoplasty for the surgical treatment of keratoconus. However, the time necessary to achieve stable
results is considerably shorter. Ophthalmology 2005;112:987–997 © 2005 by the American Academy of Ophthal-
mology.

Although the outcome of penetrating keratoplasty (PK) for
keratoconus is usually satisfactory,1–7 the unnecessary sub-
stitution of healthy recipient endothelium exposes patients
to the risk of immunologic rejection for the rest of their life.
To preserve the recipient endothelium, several authors
Originally received: October 14, 2004.
Accepted: January 5, 2005. Manuscript no. 2004-186.
1
Department of Ophthalmology, Villa Serena Hospital, Forlì, Italy.
2
University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania.
Presented in part at: American Academy of Ophthalmology Annual Meet-
ing, November 15–18, 2003; Anaheim, California.
None of the authors has any financial interest to disclose.
Correspondence and reprint requests to Prof Massimo Busin, Via Sisa 33,
47100 Forlì, Italy. E-mail: mbusin@alinet.it.
have employed different techniques of lamellar keratoplasty
(LK).8 –18 The goal of all these types of LK is to remove most
of the stromal tissue (possibly up to Descemet’s membrane),
preserve the recipient endothelium, and suture in place a donor
lamellar graft thick enough to restore normal corneal contour.
However, hand dissection is a difficult, painstaking pro-
cedure that is rarely as precise as required. As a result,
microperforation and macroperforation may occur10,18 and
prompt the surgeon to convert to PK. More often, some
corneal stroma is left attached to Descemet’s membrane,
and the uneven quality of the resulting stromal surface can
induce later scar formation, which in turn may strongly
affect visual acuity (VA).19 –21 In addition, lack of standard-
ization and reproducibility remains a major disadvantage of
the procedures involving hand dissection.

© 2005 by the American Academy of Ophthalmology ISSN 0161-6420/05/$–see front matter 987
Published by Elsevier Inc. doi:10.1016/j.ophtha.2005.01.024
 Ophthalmology Volume 112, Number 6, June 2005

To eliminate these problems, some surgeons have used
an excimer laser to remove the recipient corneal stroma22,23:
although the quality of the surface of the recipient stromal
bed obtained is excellent, ablation of a large amount of
corneal tissue requires several minutes and involves high
energy levels, possibly inducing endothelial damage. In
addition, high costs are involved, and logistics are certainly
far from optimal, as the excimer laser and operating room
are rarely adjacent, thus increasing the risk of contamination
and other intraoperative complications.
LASIK has proved that corneal dissection performed by
means of a microkeratome produces a very regular surface,
the optical quality of which is compatible with 20/20 vision.
Microkeratomes are easy to use and allow safe and rela-
tively reproducible dissection of corneal stroma. Moving
from this concept, a system dedicated to therapeutic LK has
been recently developed and marketed (ALTK [Automated
Lamellar Therapeutic Keratoplasty] system, Moria, Paris,
France). The ALTK system includes both a microkeratome
with interchangeable heads (to cut corneal lamellas of var-
ious thickness both from the recipient and from the donor)
and an artificial anterior chamber, on which a donor cor-
neoscleral rim can be mounted to prepare the lamellar
graft.24 As opposed to electrically driven microkeratomes,
the entire ALTK system can be autoclaved, assuring sterility
of the procedures performed.
We have investigated the feasibility of this system to
perform LK on 50 keratoconus patients, intolerant to spec-
tacles and contact lenses, who were included in a prospec-
tive study conducted at our facility. The rationale for the
present study differed somewhat from that of most studies
published to date. Microkeratome-assisted LK was not
aimed at simply removing the central diseased corneal tis-
sue and exchanging it with a healthy graft. Its purpose was
to reshape the ectatic cornea as a whole, by suturing under
tension a donor lamella thicker and smaller than the one
removed from the recipient, thus flattening the cone and
restoring normal corneal shape (Fig 1). In this report, we
analyze the results of this prospective study obtained to
date.
Patients and Methods
Fifty consecutive eyes of 50 keratoconus patients who could not
have their vision corrected adequately with spectacles or contact
lenses (poor VA and/or short time of contact lens tolerance) were
included in a nonmasked noncontrolled prospective clinical trial
aimed at assessing the effect of a standard microkeratome-assisted
LK procedure. All procedures were performed by the same sur-
geon (MB) between September 1999 and January 2002 at the same
institution, using the surgical technique described in detail below.
Patients were operated on regardless of cone steepness and degree
of surface irregularity. The only exclusion criteria for recruiting
the 50 patients were the presence of a preoperative pachymetric
reading lower than 380 ␮m at any of the locations tested (centrally; at
the cone apex; and at 3-o’clock, 6-o’clock, 9-o’clock, and 12 o’clock,
1–2 mm from the cone apex) and/or corneal opacities extending
beyond the anterior half of corneal thickness.
Preoperatively as well as 1 and 6 months, 1 year, 2 years, 3
years, and 4 years after surgery, each patient (with the exception of
4 whose data were collected by the referring physician) underwent

Figure 1. Microkeratome-assisted lamellar keratoplasty for keratoconus. In the schematic representation, a superficial lamella 250 ␮m in thickness and
9.0 mm in diameter is removed from the recipient cornea (a) and substituted with a donor lamellar graft 300 ␮m in thickness and 8.5 mm in diameter
sutured under tension (b). The clinical side pictures of the cornea illustrate the change in corneal shape from the preoperative ectatic configuration (c)
to the postoperative flattened aspect (d).

988
 Busin et al 䡠 Microkeratome-Assisted Lamellar Keratoplasty for Keratoconus

a complete eye examination by 1 of 2 investigators (MB or LZ), Surgical Technique

including uncorrected VA (UCVA) and best spectacle-corrected
VA (BSCVA), refraction, slit-lamp examination, and computer-
ized analysis of corneal topography (Eyesys, Houston, Texas). A
paired Student’s t test was used to determine the significance of
changes in refractive values at different postoperative examination
times. In addition, to assess better the clinical relevance of even-
tual irregular astigmatism, best hard contact lens– corrected VA
was recorded in each patient 1 year after surgery. All complica-
tions and secondary interventions were recorded.
A detailed informed consent form was signed by all keratoconus
patients undergoing microkeratome-assisted LK. Sedation by 3 cm3
of IV droperidol immediately before anesthetic injection was ad-
ministered in all cases. Local anesthesia was achieved with a
peribulbar injection of 10 cm3 of naropine 10%.
Each patient was prepared and draped in the usual fashion.
Several drops of a 5% povidone–iodine solution were instilled in
the inferior fornix, and an Olivieri locking lid speculum was
inserted to keep the eye wide open. A radial marker stained with

Figure 2. Surgical technique. a, Recipient cornea with 16 radial marks to facilitate suture placement. b, Microkeratome-assisted removal of a lamella
(9.0-mm diameter, 250 ␮m thick) from the recipient cornea. c, Measurement of the size of the recipient bed. d, Microkeratome-assisted preparation of
a donor lamella (9.0-mm diameter, 350 ␮m thick) on the artificial anterior chamber. e, Donor graft sutured in position with 16 interrupted 10-0 nylon
stitches (notice the tension in the stitches causing an evident tissue roll). f, Clear, well-adapted lamellar graft 1 day after lamellar keratoplasty performed
in a 32-year-old keratoconus patient.

989
 Ophthalmology Volume 112, Number 6, June 2005
gentian violet was used to obtain 16 radial marks on the recipient
cornea (Fig 2a). Then a suction ring was applied to the eye and the
intraocular pressure increased to over 65 mmHg. Balanced salt
solution (BSS; Alcon, Fort Worth, TX) was instilled on the corneal
surface, and a hand-driven microkeratome (LSK1, Moria) was
advanced in the track (Fig 2b) until the anterior corneal lamella
was completely severed from the underlying recipient stroma. To
facilitate the use of the ALTK system in eyes with regular curva-
ture between 38 and 46 diopters (D), the manufacturer had devel-
oped a nomogram correlating the diameter of the lamella to be
obtained with keratometric readings of the recipient cornea and the
use of 4 different suction rings (namely, ⫺1, 0, ⫹1, and ⫹2).
However, based on previous experience with our initial cases, we
found that in keratoconic eyes a lamella 9 mm in diameter could be
removed practically in all cases by using the suction ring 0 on
corneas with average keratometric readings up to 55 D and the
suction ring ⫹1 on corneas with average keratometric readings of
⬎55 D. During surgery, maximal care was taken to sweep the
microkeratome across the cornea slowly, thus letting the instru-
ment safely engage the recipient tissue and avoiding formation of
buttonholes. In all cases, the 250-␮m microkeratome head was
used.
The diameter of the excised lamella was measured using a
caliper (Fig 2c). Then the same microkeratome with a 350-␮m
head was employed to prepare the lamellar graft from the donor
cornea, which had been mounted on the artificial anterior chamber
of the ALTK system (Fig 2d). All donor corneas had been pre-
served in storage medium (Optisol, Chiron, Irvine, CA) at 4° C,
except for 5 that had been preserved at room temperature in tissue
culture medium. The diameter of the donor graft was set to be
8.5 mm using the adjustments on the artificial anterior chamber.
The quality of the donor tissue (peripheral contour, diameter, and
surface smoothness) was checked under the operating microscope,
and if it was found to be unsatisfactory, a new graft was prepared.
Figure 3. Two years after lamellar keratoplasty performed in a 48-year-old patient, the graft is clear and the interface barely visible at slit-lamp
examination.
990
Finally, the lamellar graft was sutured in place under tension
by means of 16 interrupted 10-0 nylon sutures (Fig 2e). The
knots were buried, and tobramycin and dexamethasone phos-
phate were injected subconjunctivally and given topically. At
the end of surgery, the speculum was removed, and the eye was
patched.
On the first postoperative day, the patch was removed; tobra-
mycin and dexamethasone phosphate eyedrops were administered
6 times daily and tapered off over a 12-week period.
Three months after surgery, every second interrupted suture
was removed, trying to include those responsible for major graft
distortion, as indicated by corneal topography analysis. Over the
next 3 months, all the remaining sutures were selectively removed,
trying to achieve as regular a corneal curvature as possible. In all
patients, data recorded at the 6-month examination were collected
after suture removal had been completed.
Secondary intervention to reduce postoperative astigmatism
was never performed earlier than 9 months after the initial LK
procedue—that is, when at least 3 months had elapsed after com-
plete suture removal. Relaxing incisions were performed based on
corneal topography by simply opening the peripheral annular scar
under control of intraoperative qualitative keratometry until a
regular spherical corneal curvature was obtained. After the adhe-
sions were severed, a cyclodyalisis spatula was inserted into the
interface and moved across to make sure that no tissue connection
was present centrally between the recipient cornea and overlying
graft. No compressive sutures were used here, as no overcorrection
was intended; the purpose of the relaxing incision was simply to
allow the recipient cornea and the overlying graft to slide on one
another and adjust into a more relaxed configuration, thus reducing
the curvature of the steepest areas.
Penetrating keratoplasty surgery was performed in a standard
fashion. The host bed was trephined leaving the lamellar graft in
place. An 8.0-mm donor button was sutured in a 7.75-mm recipient
 Busin et al 䡠 Microkeratome-Assisted Lamellar Keratoplasty for Keratoconus
bed with a double running 10-0 nylon suture without removing the
small peripheral annulus of donor lamellar tissue (about 0.5 mm in
width) from the previous LK surgery.
In one case, interface debridement was necessary to remove
epithelial cysts located paracentrally and interfering with vision.
To access the interface, the surgical wound was opened for about
10° in the steeper corneal meridian, as described above for per-
forming relaxing incisions. A blunt spatula was used to mechanically
remove the epithelium, and the interface was irrigated first with a 4%
solution of cocaine, then with BSS. The wound was left unsutured.
Phacoemulsification was performed through a 3-mm scleral
tunnel centered on the steepest meridian. A foldable acrylic in-
traocular lens was implanted in the capsular bag. No suture was
applied.
Results
At the time of this review, all 50 patients had completed the 1-year
follow-up, and 49 of 50 had completed the 2-year examination.
Three-year follow-up was available in 24 patients, and 4-year
follow-up in 8. Twenty-six patients were male, and 24 were
female. Their age ranged from 21 to 68 (average, 33.9).
With the exception of preparation of the donor grafts (see
below), surgery was uneventful in all patients. All corneas were
clear on the first postoperative day (Fig 2f), and reepithelialization
was completed within 2 weeks from surgery, except for 3 patients
who needed an additional week and required the use of therapeutic
contact lenses. All corneas remained clear over the time period
considered in this study (Fig 3).
Visual Acuity
Preoperative UCVA was 20/200 or less in all cases. Figure 4
Figure 4. Distribution of uncorrected visual acuity at different examination times.
summarizes the change in UCVA recorded at different postoper-
ative examination times. One month after LK, only 2 of 50 patients
(4%) could see better than 20/200. The number of patients expe-
riencing postoperative UCVA better than 20/200 rose to 6 of 50
(12%) at 6 months, 8 of 50 (16%) at 1 year, 9 of 49 (18.4%) at
2 years, 4 of 24 (16.7%) at 3 years, and 1 of 8 (12.5%) at 4 years.
The improvement in UCVA over preoperative values was always
statistically significant (P⬎0.001), with the exception of the
1-month examination time.
Preoperatively, BSCVA was worse than 20/40 in all patients
included in this series. As long as sutures were in place, only 1
patient could see 20/40 with spectacle correction, and no signifi-
cant improvement in BSCVA was observed. After suture removal
was completed, BSCVA was significantly better than preopera-
tively at all examination times (P⬍0.001). The number of patients
with BSCVA equal to or better than 20/40 was 43 of 50 (86%) at
6 months, 44 of 50 (88%) at 1 year, 46 of 49 (93.9%) at 2 years,
22 of 24 (91.7%) at 3 years, and 7 of 8 (87.5%) at 4 years. These
data are summarized in Figure 5.
Visual acuity tested with refraction over a trial hard contact lens
1 year after LK showed additional improvement in vision (Fig 6). At
this examination time, all 50 patients had a best contact lens– cor-
rected visual acuity of ⱖ20/40, which was significantly better than
BSCVA (P⬍0.0001). After the trial, 9 patients elected to wear contact
lenses.
Refraction
The average spherical equivalent (SE) increased from a preoper-
ative value (mean ⫾ standard deviation [SD]) of ⫺5.16⫾4.24 D to
⫹4.89⫾4.02 D 1 month after LK (P⬍0.0001). After suture re-
moval, a myopic shift was seen. The average value decreased to
⫺3.47⫾2.28 D at 6 months and ⫺3.86⫾3.14 D at 1 year. Both
myopic values were significantly higher than that recorded at 1
991
 Ophthalmology Volume 112, Number 6, June 2005
Figure 5. Distribution of best spectacle-corrected visual acuity at different postoperative examination times.
month (P⬍0.0001) but significantly lower than the preoperative
value (P⬍0.001). No substantial changes were seen at later post-
operative examinations.
Refractive cylinder within 4 D was measured in only 3 of 50
(6%) patients preoperatively and in 5 of 50 (10%) patients 1 month
after LK. This number increased to 39 of 50 (78%) at 6 months, 43
of 50 (86%) at 1 year, 44 of 49 (89.8%) at 2 years, 21 of 24
(87.5%) at 3 years, and 6 of 8 (75%) at 4 years after LK. These
data are summarized in Figure 7.
Corneal Topography
Corneal topography illustrates the flattening of the central ectatic
cornea achieved by LK surgery (Figure 8). The mean value (⫾ SD)
of average simulated keratometric readings obtained from corneal
topography decreased significantly from preoperative 54.5⫾6.9 D
to 40.7⫾5.3 D 1 month after LK (P⬍0.0001). After suture re-
moval was completed, corneal curvature increased significantly
relative to presuture removal values (P⬍0.0001). The mean values
of average simulated keratometric readings were 48.5⫾3.1 D at 6
months, 47.9⫾3.6 D at 1 year, 48.2⫾3.5 D at 2 years, 47.3⫾3.2 D
at 3 years, and 48.8⫾1.0 D at 4 years.
Computerized analysis of corneal topography showed that,
preoperatively, cones with a broad base were present in 28 of 50
(56%) patients, whereas in the remaining 22 patients (44%) the
cones were of the nipple shape.
Postoperative regular astigmatism with symmetric or asymmet-
ric bowtie patterns was seen in only 10 of 50 (20%) patients 1
month after LK. This number increased substantially to 38 of 50
(76%) at 6 months, 39 of 50 (78%) at 1 year, 40 of 49 (81.6%) at
2 years, 20 of 24 (83.3%) at 3 years, and 6 of 8 (75%) at 4 years.
992
Complications
No intraoperative complication was experienced in any phase of
the LK procedures, with the exception of the preparation of the
donor grafts. Microkeratome-assisted dissection had to be repeated
on a spare cornea in 8 patients, as the first graft obtained had to be
discarded because of either wrong size (6 cases) or wrong size and
irregular peripheral contour (2 cases).
Postoperatively, delayed reepithelialization (beyond 15 days
after LK) of the corneal surface was seen in 3 patients but was
succesfully managed with therapeutic contact lenses in all cases.
Astigmatism was the most common postoperative complication
of LK. Six of 7 patients with high-degree astigmatism (ⱖ4 D),
both of the regular (3 cases) and of the irregular (3 cases) type, and
BSCVA lower than 20/40 underwent relaxing incisions. The re-
maining patient underwent PK surgery after an unsuccessful trial
with hard gas-permeable contact lenses. Nine patients with irreg-
ular astigmatism of lower degree (3 cases) or high-degree myopia
(6 cases) elected to wear hard gas-permeable contact lenses. Other
complications included epithelial interface ingrowth requiring sur-
gical debridement (1 case) and cataract formation necessitating
phacoemulsification (1 case). All secondary surgical procedures
were performed between 9 months and 1 year after surgery, and
their postoperative course was uneventful.
Discussion
Keratoconus patients often suffer from progressive reduc-
tion of VA. The development of corneal ectasia causing
irregular astigmatism is the main reason, but in later stages,
 Busin et al 䡠 Microkeratome-Assisted Lamellar Keratoplasty for Keratoconus
Figure 6. Comparison between best spectacle-corrected visual acuity (BSCVA) and contact lens– corrected visual acuity (CLCVA) 1 year after surgery.
central corneal scarring may contribute substantially. In
most of these patients, hard contact lenses cannot be fitted
successfully any longer, and surgery becomes the only
option to restore useful vision.
Surgery for keratoconus is intended to achieve a curva-
ture compatible with good spectacle and/or contact lens
vision and to remove central corneal opacities, if present.
Tissue removal in itself, however, is not therapeutic beyond
removal of central scars, as the still unknown problem
leading to keratoconus formation is probably not limited to
the ectatic area. It is therefore conceptually wrong to try to
eradicate the entire cone by substituting the diseased cornea
with healthy donor tissue. In fact, relatively often this is
actually impossible, as the cone extends far into the corneal
periphery. Remodeling of the corneal contour is what is
sought, rather, and to achieve this, various surgical options
have been developed over the past decades.1–23
A large number of studies have shown that PK surgery
performed in keratoconic eyes usually leads to very satis-
factory results in terms of VA, postoperative refractive
error, and endothelial survival.1–7 Most surgeons usually
combine full-thickness removal of the central cornea (usu-
ally comprising an area 7.5 to 8.5 mm in diameter) with
cone cauterization6,25,26 or variations of the graft– host dis-
parity,25–30 with the purpose of collapsing the ectatic cor-
neal dome and/or pulling the peripheral cornea towards the
center. However, a variable number of keratoconus patients
experience over time one or more endothelial rejection
episodes, causing graft decompensation in up to 9% of
eyes.2,7,31–33 It must also be noted that keratoconus is a
bilateral condition, often requiring bilateral surgery, and
evidence has shown that the risk of rejection may be higher
for the second eye.34
In view of this, several authors have employed lamellar
stromal grafts, as both inlays and onlays, to treat keratoco-
nus, thus avoiding the unnecessary transplantation of
healthy endothelium.8 –23 All types of LK have in common
additional advantages over PK surgery: quality donor en-
dothelium is not required, reducing the problem of tissue
availability; the procedure is extraocular, and intraoperative
complications are minimized, especially those threatening
vision (i.e., endophthalmitis and expulsive hemorrhage);
and steroidal therapy can be discontinued much earlier in
the postoperative course than it is usually after PK, thus
practically eliminating the risk of secondary cataract or
glaucoma. Nevertheless, to date PK has remained the pro-
cedure of choice for the surgical treatment of keratoconus.
The main reason for this is that vision after LK is often
unsatisfactory, leading to a best-corrected VA of 20/40 to
20/50 at best.19 –21 The poor quality of hand-dissected sur-
faces in both donor and recipient corneas was held respon-
sible for postoperative scar tissue formation, increased light
scattering, and consequent limited vision. More recently,
several authors have claimed that visual results of LK for
keratoconus can be improved if hand dissection is carried
out very deep in the recipient cornea.11–18,22,23 Postopera-
993
 Ophthalmology Volume 112, Number 6, June 2005

Figure 7. Distribution of refractive astigmatism at different postoperative examination times. n.m. ⫽ not measured.

tive healing between recipient bed and donor full-thickness
grafts deprived of their endothelial layer should be minor
and would not cause stromal scarring capable of affecting
vision. According to the reports published to date, similar
results can be obtained both when Descemet’s membrane
is bared and when a thin layer of deep stroma is left in
place.18
However, hand dissection is a painstaking procedure,
requiring a substantially longer surgical time than standard
PK. The accuracy of stromal removal is strongly dependent
on the surgeon’s surgical skills, making the procedure
hardly reproducible. Moreover, the learning curve is quite
demanding, and intraoperative complications, often requir-
ing conversion to PK, are relatively frequent, even in expe-
rienced hands.10,18 As a result, most keratoplasty surgeons
have been unwilling to abandone a well-established tech-
nique such as PK in favor of a time-consuming method with
an unknown long-term outcome, such as LK. To simplify
and standardize LK, excimer laser ablation has been used in
the preparation of the recipient bed. Eckhardt et al22 and
Buratto et al23 have shown encouraging results in small
series, but concerns regarding the amount of energy neces-
sary for the ablation and the related possible endothelial
damage still remain. In addition, to complete the entire
procedure, either the patient must be moved from the laser
suite to the surgical room, increasing the risk of contami-

Figure 8. Corneal map of a central keratoconus (a) and the postoperative result obtained after microkeratome-assisted lamellar keratoplasty (b).
Flattening of the cone and regularization of corneal shape are evident.

994
 Busin et al 䡠 Microkeratome-Assisted Lamellar Keratoplasty for Keratoconus
nation, or the 2 stages should be performed separately in 2
successive sessions, which is unacceptable to the majority
of patients.
As opposed to PK for keratoconus, the use of LK not
only to remove diseased tissue but also to reshape the
cornea as a whole has not been extensively addressed. In
most studies published to date, the lamellar graft either had
the same size as the recipient bed or was oversized8 –18; as
a variation, some authors have advocated burying the edge
of a 1.0- to 1.5-mm oversized graft in a peripheral pocket of
the recipient cornea.23 These LK techniques do not achieve
substantial cone flattening, and the effect of the procedure is
limited only on the center of the host cornea, making it
indicated only in patients with moderate keratoconus and
central location of the ectasia. In 1979, Kaufman and Werblin
introduced epikeratophakia, an onlay type of LK employing
undersized stromal grafts for the treatment of keratoconus.35
In this procedure, an 8.5-mm donor lamella was sutured in
an annular superficial keratectomy, 9.0 mm in diameter,
prepared in the host cornea. The central recipient cornea
was left untouched, with the exception of removal of the
epithelial layer. As a consquence of graft undersizing, su-
tures were placed under a considerable amount of tension,
and flattening of even advanced cones was achieved, as
demonstrated in a multicentric prospective study.36 The
degree of ectasia was not a limiting factor, but as the entire
recipient stroma was left in place, the cornea had to be free
of central scars, and therefore, the number of very steep
cones operated on was relatively small.
The technique of LK for keratoconus presented in this
article was developed to try to overcome all the problems
discussed above.
Microkeratome-assisted dissection is easy to perform,
and shortens surgical time considerably. Similar to what has
been reported by others,24 reproducibility of the excision in
our series was relatively good.
The diameter of the host bed obtained was 9 mm, and its
contour was very regular in all cases, despite the wide range
of preoperative corneal curvatures and degrees of surface
irregularity. From previous experience, we had realized that
sweeping the hand-driven microkeratome across the very
steep corneas very slowly was essential in preventing for-
mation of buttonholes. We also found that the diameter of
the corneal lamella removed could be kept accurate by using
only 2 suction rings—that is, the ring 0 for corneas with an
average simulated keratometric reading up to 55 D, and the
ring ⫹1 for corneas with an average simulated keratometric
reading above 55 D. All recipient beds prepared in our
series presented a very smooth surface free of any scar
tissue.
Preparation of the donor graft from the corneas mounted
on the artificial anterior chamber was complicated in 8
cases: in these procedures the diameter of the graft obtained
was other than 8.5 mm, and in 2 of 8, the contour was
irregular. The pressure in the system is critical to be able to
applanate the cornea properly and cut it to the desired
diameter. In our series, the pressure was too low in 1 case
and was completely lost due to leakage while cutting in 2
cases (the scleral rim was too small and fixation was not
good); as a result, the microkeratome did not indent the
tissue sufficiently and cut a smaller graft, which also had an
irregular contour in the cases when pressure was lost. On the
contrary, in 5 cases increasing the pressure beyond 70 to 80
mmHg made the cornea slide out of position and bulge
excessively, resulting in a graft larger than expected.
When performing microkeratome-assisted LK, spare do-
nor tissue should always be available. This usually would
not represent a major hindrance to the procedure, as almost
all corneas not suitable for PK can be used.
We employed a head with a 250-␮m slit to cut the
recipient bed and one with a 350-␮m slit to cut the donor
cornea, with the purpose of restoring normal corneal thick-
ness. However, the final intended increase of 100-␮m in
corneal thickness represents more a theoretical value, as
many factors may affect it. First, published data have shown
that all microkeratome heads cut somewhat more thickly
than expected,24 mainly because corneal tissue is squeezed
through the slit while being dissected. The amount of vari-
ability in thickness obtained seems to increase proportion-
ally to the width of the slit used for dissection.24 In addition,
donor corneas preserved at 4° C, like the tissue used for
most procedures of this series, are swollen and dehydrate
after transplantation, becoming considerably thinner than
they were at the time of preparation. Therefore, even LK
techniques that do not employ microkeratome dissection
cannot allow precise titration of postoperative corneal thick-
ness. The importance of this issue for the final result re-
mains to be demonstrated. Refractive surgery has shown
that the cornea can tolerate both reduction and increase of
its thickness within a relatively wide range of values, with-
out being affected in its function. To date, no correlation has
ever been reported between postoperative corneal thickness
and refractive results, rate of recurrence, or other parameters
of keratoconus patients undergoing LK.
Visual and refractive results recorded in our study com-
pare favorably with those reported after other types of
LK.8 –23 In particular, the percentage of patients enjoying a
BSCVA equal to or better than 20/40 ranged from 86% as
early as 6 months after surgery to 93.9% 2 years postoper-
atively. As opposed to other authors,37 we did not see any
correlation between patient age and BSCVA in our series.
Visual results after PK do not differ substantially from those
of our series1–7; nevertheless, visual rehabilitation is typi-
cally longer after PK, as stable refraction is achieved only
after suture removal is completed, usually later than 12
months, as opposed to 6 months in our LK patients.
The average SE was initially myopic, shifted to hyper-
opic at 1 month, and finally reversed back to myopic at all
following examination times, indicating that release of ten-
sion after complete suture removal reversed to some extent
the initial effect of the procedure. That these shifts were due
to changes in corneal curvature was confirmed by topogra-
phy. Recently, to avoid this regression of effect, before
performing the microkeratome cut we cauterized the central
superficial cornea in 2 patients who were not part of this
study. Similar to what we have experienced in PK patients,6
cone cauterization has resulted in a final less myopic SE
(within 1 D).
Astigmatism beyond 4 D was seen relatively seldom
after completion of suture removal, the percentage varying
995
 Ophthalmology Volume 112, Number 6, June 2005
between 10% and 14% 1 to 3 years after surgery. It must be
noted that 6 patients with high-degree astigmatism under-
went surgical correction by means of relaxing incisions,
whereas 1 patient elected to receive a PK procedure. Al-
though irregular astigmatism did not greatly affect vision, 9
patients preferring their contact lens– corrected visual acuity
were fitted with contact lenses. These percentages of high-
degree and/or irregular postkeratoplasty astigmatism do not
differ substantially from the data reported in other LK or PK
series.1–23 However, when comparing visual and refractive
results of our prospective study, one should consider that
unsatisfied patients in other retrospective studies may have
been lost to follow-up.
Vision-threatening complications were not seen in our
study. Surgery other than that required to correct astigma-
tism was necessary in only one patient, a 62-year-old who
developed a visually significant nuclear cataract and under-
went uneventful phacoemulsification. This, again, compares
favorably, especially with the results of PK surgery for
keratoconus.1–7
In summary, the results of microkeratome-assisted LK col-
lected to date encourage further experience with this technique.
It shares all the advantages of other LK techniques over con-
ventional PK. However, as opposed to other types of LK, the
procedure does not require particular surgical skills and can be
fairly standardized in most of its parts, and it seems that
complications can be easily managed. In addition, the optical
quality of surfaces cut with the microkeratome is certainly
superior to that of surfaces prepared with hand dissection.
Surgeons familiar with PK may be willing to try a technique
that has a short learning curve, is not time consuming, and does
not add particular difficulties to their routine.
References
1. Paglen PG, Fine M, Abbott RL, Webster RG Jr. The prognosis
for keratoplasty in keratoconus. Ophthalmology 1982;89:651– 4.
2. Epstein RJ, Seedor JA, Dreizen NG, et al. Penetrating kera-
toplasty for herpes simplex keratitis and keratoconus. Allo-
graft rejection and survival. Ophthalmology 1987;94:935– 44.
3. Troutman RC, Lawless MA. Penetrating keratoplasty for ker-
atoconus. Cornea 1987;6:298 –305.
4. Kirkness CM, Ficker LA, Steele AD, Rice NS. The success of
penetrating keratoplasty for keratoconus. Eye 1990;4:673– 88.
5. Sharif KW, Casey TA. Penetrating keratoplasty for keratoconus:
complications and long-term success. Br J Ophthalmol 1991;75:
142– 6.
6. Busin M, Zambianchi L, Franceschelli F, et al. Intraoperative
cauterization of the cornea can reduce postkeratoplasty refrac-
tive error in patients with keratoconus. Ophthalmology 1998;
105:1524 –9, discussion 1529 –30.
7. Ing JJ, Ing HH, Nelson LR, et al. Ten-year postoperative
results of penetrating keratoplasty. Ophthalmology 1998;105:
1855– 65.
8. Anwar M. Technique in lamellar keratoplasty. Trans Ophthal-
mol Soc U K 1974;94:163–71.
9. Archila EA. Deep lamellar keratoplasty dissection of host tissue
with intrastromal air injection. Cornea 1984 –5;3:217– 8.
10. Sugita J, Kondo J. Deep lamellar keratoplasty with complete
removal of pathologic stroma for vision improvement. Br J
Ophthalmol 1997;81:184 – 8.
996
11. Tsubota K, Kaido M, Monden Y, et al. A new surgical
technique for deep lamellar keratoplasty with single running
suture adjustment. Am J Ophthalmol 1998;126:1– 8.
12. Melles GR, Rietveld FJ, Beekhuis WH, Binder PS. A tech-
nique to visualize corneal incision and lamellar dissection
during surgery. Cornea 1999;18:80 – 6.
13. Manche EE, Holland GN, Maloney RK. Deep lamellar kera-
toplasty using viscoelastic dissection. Arch Ophthalmol 1999;
111:1561–5.
14. Melles GR, Lander F, Rietveld FJ, et al. A new surgical
technique for deep, stromal, anterior lamellar keratoplasty.
Br J Ophthalmol 1999;83:327–33.
15. Melles GR, Remeijer L, Geerards AJ, Beekhuis WH. A quick
surgical technique for deep, anterior lamellar keratoplasty
using visco-dissection. Cornea 2000;19:427–32.
16. Amayem AF, Anwar M. Fluid lamellar keratoplasty in kera-
toconus. Ophthalmology 2000;107:76 –9, discussion 80.
17. Anwar M, Teichmann KD. Big-bubble technique to bare De-
scemet’s membrane in anterior lamellar keratoplasty. J Cata-
ract Refract Surg 2002;28:398 – 403.
18. Anwar M, Teichmann KD. Deep lamellar keratoplasty: surgi-
cal techniques for anterior lamellar keratoplasty with and
without baring of Descemet’s membrane. Cornea 2002;21:
374 – 83.
19. Soong HK, Katz DG, Farjo AA, et al. Central lamellar kera-
toplasty for optical indications. Cornea 1999;18:249 –56.
20. Panda A, Bageshwar LM, Ray M, et al. Deep lamellar kera-
toplasty versus penetrating keratoplasty for corneal lesions.
Cornea 1999;18:172–5.
21. Saini JS, Jain AK, Sukhija J, Saroha V. Indications and
outcome of optical partial thickness lamellar keratoplasty.
Cornea 2003;22:111–3.
22. Eckhardt HB, Hütz WW, Heinrich AW, Kaiser WE. Lamel-
lierende Keratoplastik mit dem Excimerlaser. Erste klinische
Ergebnisse. Ophthalmologe 1996;93:242– 6.
23. Buratto L, Belloni S, Valeri R. Excimer laser lamellar kera-
toplasty of augmented thickness for keratoconus. J Refract
Surg 1998;14:517–25.
24. Springs CL, Joseph MA, Odom JV, Wiley LA. Predictability
of donor lamellar graft diameter and thickness in an artificial
anterior chamber system. Cornea 2002;21:696 –9.
25. Girard LJ, Esnaola N, Rao R, et al. Use of grafts smaller than
the opening for keratoconic myopia and astigmatism. A pro-
spective study. J Cataract Refract Surg 1992;18:380 – 4.
26. Kaufman HE, et al, eds; Ryan SJ Jr, consultant. Color Atlas
of Ophthalmic Surgery. Corneal and Refractive Surgery.
Philadelphia: Lippincott; 1992:22.
27. Goble RR, Hardman Lea SJ, Falcon MG. The use of the same
size host and donor trephine in penetrating keratoplasty for
keratoconus. Eye 1994;8:311– 4.
28. Wilson SE, Bourne WM. Effect of recipient-donor trephine
size disparity on refractive error in keratoconus. Ophthalmol-
ogy 1989;96:299 –305.
29. Javadi MA, Mohammadi MJ, Mirdeghan SA, Sajjadi SH. A
comparison between donor-recipient corneal size and its effect
on the ultimate refractive error induced in keratoconus. Cor-
nea 1993;12:401–5.
30. Girard LJ, Eguez I, Esnaola N, et al. Effect of penetrating
keratoplasty using grafts of various sizes on keratoconic my-
opia and astigmatism. J Cataract Refract Surg 1988;14:541–7.
31. Williams KA, Roder D, Esterman A, et al. Factors predictive
of corneal graft survival. Report from the Australian Corneal
Graft Registry. Ophthalmology 1992;99:403–14.
32. Bishop VL, Robinson LP, Wechsler AW, Billson FA. Corneal
graft survival: a retrospective Australian study. Aust N Z J
Ophthalmol 1986;14:133– 8.
 Busin et al 䡠 Microkeratome-Assisted Lamellar Keratoplasty for Keratoconus

33. Price FW Jr, Whitson WE, Marks RG. Graft survival in four
common groups of patients undergoing penetrating kerato-
plasty. Ophthalmology 1991;98:322– 8.
34. Donshik PC, Cavanagh HD, Boruchoff SA, Dohlman CH.
Effect of bilateral and unilateral grafts on the incidence of
rejections in keratoconus. Am J Ophthalmol 1979;87:823– 6.
35. Kaufman HE, Werblin TP. Epikeratophakia for the treatment
of keratoconus. Am J Ophthalmol 1982;93:342–7.
36. McDonald MB, Kaufman HE, Durrie DS, et al. Epikeratopha-
kia for keratoconus. The nationwide study. Arch Ophthalmol
1986;104:1294 –300.
37. Benson WH, Goosey CB, Prager TC, Goosey JD. Visual
improvement as a function of time after lamellar keratoplasty
for keratoconus. Am J Ophthalmol 1993;116:207–11.

997