The Melles Technique

C. Maya Tong, BSc, MD; Jack Parker, MD, PhD; Korine van Dijk, BOptom, PhD; Rénuka S. Birbal, MD; and Gerrit R. J. Melles, MD, PhD

Deep anterior lamellar keratoplasty (DALK) is designed to exchange healthy donor corneal stroma for recipient pathologic corneal stroma. Unlike its predecessor, penetrating keratoplasty (PK), DALK leaves the recipient Descemet’s membrane (DM) and endothelium intact. Broadly speaking, there are 2 basic, competing, surgical strategies for DALK: those relying on injection and those relying on manual dissection. Melles Manual DALK (MM-DALK) falls into the latter category; it is unique in that, unlike other manual techniques, MM-DALK facilitates precise, reliable dissection down to the level of the DM by use of a phenomenon known as the air-endothelial reflex.^1,2

BACKGROUND

Anterior lamellar keratoplasty (ALK) was first developed over 150 years ago.³ Early attempts at the operation, where the recipient cornea was dissected near the DM, were made as early as the 1950s by Hallermann.⁴ While structurally and technically successful, these early efforts were largely regarded, both optically and functionally, as failures, because the visual results of the early anterior lamellar transplants were often far inferior to those achieved by penetrating grafts.⁵ The explanation for this underperformance was that total removal of the recipient stroma, down to the DM, was difficult to achieve by the usual means of manual dissection from the corneal surface. The layers of stroma inevitably left behind would interact negatively with the donor tissue, producing haze at the graft interface and spoiling the transplant’s optical performance.6 Therefore, major attempts at ALK were largely abandoned for decades until new methods for more complete recipient stromal removal could be discovered.

Important technical improvements were made in the early 1990s, when Amayem and Anwar ⁷ published that injection of saline into the deep corneal stroma could often result in a cleavage plane just above the DM. Several years later, Anwar and Teichmann⁸ refined the technique utilizing air instead of saline, effectively developing the big bubble strategy for DALK. Around the same time, it was also independently discovered that deep injections of viscoelastic could yield similar results.⁹ Nevertheless, a major problem with these injection techniques was their reliability; big bubbles were not always achieved, and, even worse, approximately 10% to 50% of attempts resulted in inadvertent DM perforation secondary to rapid expansion of the injected material.^9–16 This unpredictability in outcome drove the search for a new technique that would permit a controlled separation of the recipient stroma from its DM, and gave rise to the development of the MM-DALK method.

THE MELLES MANUAL DALK SURGICAL TECHNIQUE

The MM-DALK operation is based on a single observation and relies on a specialized set of instruments. The observation is this: if the recipient anterior chamber is completely filled with air, when instruments are inserted into the peripheral cornea, a reflection of the tip of those instruments appears. The deeper the instrument is pressed into the cornea, the closer the reflection appears to the instrument, such that, when the 2 are nearly touching, a stromal dissection level just anterior to the posterior corneal surface has been achieved (Figure 7-1). This air-endothelial reflex provides an immediate, ever-present indication of the depth of the ongoing dissection and thereby enables a controlled, precise manual separation of the host stroma away from the underlying DM.^1,2 After a superior peritomy is performed and a 5-mm corneoscleral tunnel is fashioned using a crescent blade, the anterior chamber is filled with air (Figure 7-2). A specialized set of 3 spatulas (Melles spatula set; DORC International BV) is then employed. These spatulas have red, white, and blue handles corresponding to their increasing lengths and curvatures. They are used in sequence to bluntly dissect through the recipient cornea, from limbus-to-limbus, 360° around, guided by the air-endothelial reflex (see Figure 7-2).^1,2

After this stromal dissection is completed, the air is removed from the anterior chamber and viscoelastic is injected into the dissected stromal pocket. The anterior corneal surface is trephined away and the donor lamellar graft is sutured into position (see Figure 7-2).^1,2

PREOPERATIVE CONSIDERATIONS AND SURGICAL INDICATIONS

The most common indications for MM-DALK are advanced keratoconus, or a patient with a history of corneal hydrops treated conventionally, which is typically regarded as a contraindication to big bubble DALK due to an increased risk of DM perforation.^17–21 The Jacob modification of pre-Descemetic DALK has been described for primary management of acute hydrops to prevent the scarring that is generally seen following conventional management.^22,23 Faster visual rehabilitation through a single staged surgery and decreased contact lens dependence are other advantages, as are visual, optical, topographic, biomechanical, and structural improvement in these advanced cases of keratoconus.

Other less common indications include post-herpetic/traumatic corneal scarring and a variety of corneal stromal dystrophies.^24–26 Unquestionably, the health of the corneal endothelium is a main criterion for deciding if a DALK procedure is indicated. The best surgical candidates are adults with 2 severely visually impaired eyes, because psychological surveys have demonstrated that patient satisfaction with corneal replacement for keratoconus depends more on the operated eye becoming the better seeing eye, rather than the absolute level of visual improvement the surgery yields.^27–29 Few studies of pediatric DALK (of any technique) have been conducted, but patient ability and willingness to cooperate are generally correlated with operative success.^30,31

MM-DALK may be regarded as a minimally invasive procedure, because nearly all its steps are extraocular, and do not involve manipulations within the globe.^1,2 As a result, MM-DALK may be performed with only retrobulbar anesthesia. Obtaining a soft eye may be the single most important pre-surgical step, as the manual stromal dissection is made substantially more difficult by significant posterior pressure, which raises the chances of inadvertent DM perforation.^2,26 Therefore, an effective retrobulbar block, followed by digital ocular massage and a Honan balloon, are strongly advised. Intraoperatively, the patient should be placed in a proper anti-Trendelenberg position, and careful attention should be paid to the tightness of the speculum to ensure that excessive pressure is not being applied to the globe.

The visual results of MM-DALK likely depend on whether deep stromal dissection is successfully achieved, but, given that, may equal those achieved by big bubble DALK and conventional PK, with an average resultant best spectacle-corrected visual acuity of 20/40 (0.5).32–39 The endothelial cell density of MM-DALK–operated eyes tends to be higher than that found with big bubble DALK, suggesting that rapid bubble expansion may be somehow traumatic to the recipient corneal endothelium.^40,41

Compared to PK, DALK generally results in more postoperative myopia, because the resultant corneas are often steeper by, on average, 2 diopters (D); this may be because DALK avoids the intraoperative collapse of the anterior chamber and angle that is entailed by PK.^42–44 The degree of myopia may, however, be decreased by using same size grafts. Whereas, after PK, approximately 10% of operated eyes may require rigid contact lens fitting to achieve satisfactory visual results, no such figures have been published for big bubble or MM-DALK.^45–50

Theoretically, DALK may permit earlier suture removal compared to PK and, therefore, more rapid visual stability. This claim has been occasionally extended to argue that DALK performed with the assistance of the femtosecond laser for shaping the edges of the donor and recipient tissues into interlocking configurations may be conducive to even greater stability and earlier suture removal compared to big bubble and MM-DALK.^51–53 However, the only study on the subject to specifically examine the issue failed to corroborate this hypothesis.⁵⁴ Another limitation of the femtosecond laser for DALK is that the device has demonstrated a surprising inability to perform adequately deep dissections in the recipient cornea. Specifically, deep laser cutting depths have been shown to generate ridges in the bed of the recipient cornea, and these ridges may undermine the optical performance of the eye compared to the dissections produced by big bubble and MM-DALK methods.^55,56

One criticism of both PK and DALK is that neither operation tends to prevent disease recurrence; that is, resumed corneal ectasia in the transplanted tissue or at the graft-host junction.^57–65 Infiltration of the donor tissue with pathologic recipient keratocytes is one proposed mechanism for such recurrences.⁶⁶ If this is correct, then DALK may actually carry a somewhat higher risk than PK, as more of the recipient cornea is left in-situ, unexcised, in proportion to the depth and diameter of the achieved dissection.

SURGICAL AND POST-SURGICAL COMPLICATIONS

DALK’s most common complication is intraoperative DM perforation, which may occur in approximately 10% of eyes operated with both the MM- and big bubble DALK techniques.^{8,9,11,12,37,67} In the latter, perforation may frequently necessitate conversion to PK because large surface incisions into the recipient cornea have already been created; the intrastromal air injection is usually performed after partial-thickness surface trephination.^68,69 In contrast, with MM-DALK, perforation may be managed expectantly by aborting the operation and permitting the cornea to heal, thereby enabling another attempt at DALK in the future.

Another commonly discussed complication after DALK is graft failure; this remains a highly controversial topic today, mostly secondary to disputes over terminology. Specifically, graft failure originally referred to corneal decompensation stemming from donor endothelial cell death or depletion.70 This type of graft failure is exceedingly rare with DALK. The average life expectancy of an anterior lamellar transplant following uncomplicated surgery has been estimated at approximately 50 years, compared to approximately 18 years following PK.⁷¹ However, it has recently also become popular to use the term graft failure to describe an optically undesirable result following otherwise successful DALK surgery. These underperforming grafts may, for example, feature dense interface haze that limits the visual performance of the eye that received the transplant.⁷² Some studies have decided to classify these optically disappointing DALKs as failed grafts; however, it should be noted that graft failure in this sense is not the same as it has been conventionally understood. Moreover, it means special caution must be applied when comparing the rates or incidences of graft failure between PK and DALK, as different outcomes may be lumped under the same broad term. Therefore, it may be advisable to distinguish between cellular and optical graft failure, and to distinguish clearly what is meant in each instance.⁷³

No comparative studies yet exist between MM-DALK and big bubble DALK for allograft reaction or cellular graft failure. However, as cell densities tend to be higher after mm- compared to big bubble DALK, and the occurrences of inadvertent DM perforation may tend to be lower, graft survival may be longer in the manual method.^38,39,69

MELLES MANUAL DALK ALTERNATIVE AND THE FUTURE OF ANTERIOR LAMELLAR KERATOPLASTY

Although DALK confers a number of advantages over conventional PK for many patients, significant, intrinsic drawbacks to the operation nevertheless remain. These include the necessity to make large surface incisions in the recipient cornea, the requirement for long-standing corneal sutures, the use of cellular donor material, the ongoing risk for allograft reaction and graft rejection, and the demand for significant extended follow-up examinations.^74–77 All of these postoperative risks and obligations are entailed by the inherent nature of DALK itself, and, therefore, are not likely to be resolved by any future advances in operative technique, instrumentation, or medication. If they are to be definitively solved or avoided, some entirely new surgical technique may be necessary.

It has been recently shown that alternative techniques, such as UV–cross-linking (UV-CXL), may be changing the role of DALK for anterior corneal diseases. The introduction of UV-CXL in Europe effectively slowed or halted the progression of keratoconus, and may have contributed to a 25% decrease in DALKs being performed in the years following its nationwide.^78,79 The frequency of DALK in the United States, on the other hand, remained stable over the last 10 years.⁸⁰ However, with the approval of UV-CXL by the US Food and Drug Administration in April 2016 for treatment of patients with progressive keratoconus and post-LASIK ectasia, the United States may also experience a decrease in frequency of DALK in the coming years.

Stay updated, free articles. Join our Telegram channel

Join