Penetrating Keratoplasty in Keratoconus


  • Despite the increasing popularity of deep anterior lamellar keratoplasty (DALK), penetrating keratoplasty (PK) remains an important modality in the surgical management of advanced keratoconus, particularly in cases of deep stromal scarring and/or history of corneal hydrops.

  • PK provides excellent graft survival and acceptable visual outcomes. However, visual rehabilitation can be prolonged, owing to high postoperative astigmatism or anisometropia.

  • Femtosecond laser for trephination in PK theoretically enables better apposition, faster wound healing, and patterns of trephination not achievable with conventional trephines.

  • DALK eliminates the risk of endothelial graft rejection, resulting in improved long-term graft survival, comparable visual outcomes, reduced endothelial cell loss, and decreased need for postoperative steroids. The procedure is, however, more technically challenging, and interface haze remains a concern.


Historically, for almost a century, penetrating keratoplasty (PK), in which the entire thickness of the cornea is replaced by a donor corneal tissue, has been the gold standard surgical approach for the treatment of advanced keratoconus. , In contemporary practice, however, deep anterior lamellar keratoplasty (DALK)—in which the superficial corneal layers are removed, leaving only the host Descemet layer and endothelium intact—is increasingly becoming the preferred primary surgical option, currently representing 10% to 20% of all transplants for keratoconus. , PK, nonetheless, has a significant role in the surgical management of advanced keratoconus, especially where there is deep stromal scarring or a history of corneal hydrops, and in the less common scenario of coexisting endothelial dysfunction, such as Fuchs endothelial dystrophy. ,


Longitudinal data show that keratoconus is one of the most common indications for PK. In a review of data from the years 1980 to 2014, keratoconus was the leading indication for PK in many regions of the globe (Europe, Australia, the Middle East, Africa, and South America). In North America, keratoconus represented the third most common indication for PK (approximately 14% of cases), following pseudophakic bullous keratopathy/aphakic bullous keratopathy and regrafts. Along with regional variabilities in the incidence of keratoconus, this difference can also be attributed partially to the availability of less invasive surgical options in the developed world.

Although PK is the most common surgical option in advanced keratoconus, most patients do not ultimately require penetrating surgery. A study of keratoconus subjects followed for 48 years reported that fewer than 20% of patients underwent PK. Another study of 1065 keratoconus subjects found that only 12% required PK over an 8-year follow-up period. Reported indications for PK include the presence of corneal scarring, best-corrected visual acuity (BCVA) worse than 20/40 with contact lens correction, keratometry steeper than 55 diopters (D), corneal astigmatism greater than 10 D, early age of keratoconus development, poor contact lens tolerance, and nonclearing corneal hydrops.


PK techniques for the treatment of keratoconus have evolved over the years. However, a recent review concluded that there was no evidence of superiority for any specific technique. The pattern of graft suturing (interrupted, single, combined interrupted and running, or double running) does not appear to influence ultimate BCVA. , The effect of graft sizing, although the subject of controversy, is likely modest. The type of mechanical trephine used has also not been shown to influence visual outcome, although the use of femtosecond laser trephination may enable better wound healing and faster visual rehabilitation with earlier suture removal (see section later).

Graft sizing has been controversial, with various studies reporting slightly better (or worse) results with oversized versus same-sized grafts. Postkeratoplasty myopia can be reduced by using donor tissue that is the same size as the recipient trephination. Some have suggested undersizing the donor to further flatten the postoperative corneal contour. These maneuvers have to be undertaken with caution, since reducing donor size in an eye with a relatively short axial length may result in significant postoperative hyperopia. In addition, if undersizing the donor, the surgeon must take extreme care in coaptation of the wound to avoid leaks. Moreover, the flattened corneal contour can complicate postoperative contact lens fitting.

“Recurrent” keratoconus has been observed at a rate of 12% over 25 years. Recurrence in the donor graft is commonly the result of an error in technique, related to incomplete excision of the cone. The iron (Fleischer) ring, found at the base of the cone, can be used as a reference when planning graft size and should be excised fully.


Postoperative suture adjustment may play an important role in visual rehabilitation following PK. The timing of first suture removal is debated but can be initiated when wound healing is evidenced (typically at 3–6 months postoperatively). If corneal astigmatism is satisfactory with sutures in place (typically equal or lower than 3.00 D), sutures can remain unless there are other indications for removal, such as suture breakage, infection, vascularization, marked perisutural fibrosis, patient discomfort, or graft rejection. Loose interrupted sutures provide no structural support and are a risk for graft infection. They should, therefore, be removed. This can generally be done without endangering wound integrity ( Fig. 32.1 ).

Fig. 32.1

(A) Clear penetrating graft status post suture removal. (B) Anterior segment of a penetrating graft in situ. (Image courtesy Luis Izquierdo.)

Postoperative suture removal can cause large and unpredictable swings in the degree of astigmatism, regardless of the type of suture used. In the interrupted suture technique, one can attempt to titrate the astigmatic effect through sequential suture removal. Alternatively, adjusting the tension of a continuous suture at the slit lamp, in the first 48 to 72 hours postkeratoplasty, can be employed to reduce corneal astigmatism, albeit with the inherent risk of suture breakage. Once all sutures have been removed, however, the measured astigmatism tends to remain relatively stable.


PK provides excellent outcomes in terms of graft clarity, exceeding 90% at 1 year. Graft survival following PK for keratoconus is also quite high, with a survival rate of 97% at 5 years, 89% to 92% at 10 years, and 49% at 20 years. , This does suggest, however, that a young keratoconic patient with a PK may require more than one graft over a lifetime. Subsequent grafts may have a shorter survival than the first graft. Risk factors for PK failure in this setting include previous graft failure, glaucoma, synechia formation, corneal neovascularization, aphakia, and pseudophakia.

The definition of successful PK continues to evolve. Beyond maintaining a clear cornea, a successful refractive result also needs to be considered. PK offers good vision in the long term but visual rehabilitation may be slow, owing to high postoperative astigmatism or anisometropia. After PK for advanced keratoconus, the final uncorrected visual acuity (UCVA) ranges from 20/50 to 20/100. , , Contact lenses can improve BCVA to 20/25 or better, with 67% to 96% of patients seeing at least 20/40 at 1 year postoperatively. , , , , ,

The visual gains post-PK may occasionally recede over time owing to progressive donor-recipient misalignment or recurrence of keratoconus, with resultant irregular astigmatism in the graft. This period of gradual worsening vision tends to begin around 10 years after first suture removal. Moreover, recurrence of keratoconus, occurring in 12% at 25 years, can also confound long-term visual outcomes.

There are significant challenges in the postoperative management of the keratoconus patient undergoing PK. These include suture and wound-healing problems, progression of disease in the recipient rim, allograft reaction in a generally younger population, and persistent high or irregular astigmatism. In addition, all penetrating grafts are at risk of dehiscence with trauma.

Although primary graft failure following PK is quite rare, allograft rejection affects 13% to 31% of eyes within the first 3 years after surgery, with a mean onset of 8 to 15 months. Risk factors for graft rejection include larger-sized grafts, the number of previous corneal transplants, and the presence of peripheral corneal neovascularization.

Ultimately, patients must be counseled on the goals of transplantation and realistic expectations for postoperative visual acuity, as well as about the modalities for correction. Patients should be advised that they will most likely require spectacles or contact lenses to attain best vision, and it often takes at least 1 year or longer to achieve stable BCVA.


Manual trephination in PK is vertical, with little variation in wound construction. In conventional PKP, when the donor corneal tissue is punched from the endothelial surface, it produces a graft with an anterior diameter of 0.2 to 0.3 mm smaller than the posterior diameter. , Moreover, vacuum trephines can lead to incomplete/oblique cuts. This may result in an irregular donor-host junction, anteroposterior and rotational misalignment of the donor-host corneal interfaces, and slow and asymmetric postoperative wound healing. This, in turn, can result in regular and/or irregular astigmatism that may influence final visual outcome.

The introduction of the femtosecond laser in PK for trephination of both donor and recipient tissues theoretically enables better apposition and faster wound healing. Femtosecond laser incisions produce precise and reproducible trephination of the donor and recipient buttons, avoiding the aforementioned disparities produced by the manual techniques. Moreover, femtosecond laser incisions allow for same-sized donor and graft tissues, resulting in improved natural alignment, a watertight seal, less suture tension, and less induced astigmatism.

The femtosecond laser also allows for patterns and angles of incisions not achievable with conventional trephines. The variations in wound configurations include top hat, mushroom, zigzag, and christmas tree, among others. , These shaped incisions have been associated with a stronger wound profile, presumed faster donor-recipient junction wound healing, earlier suture removal, less surgically induced corneal astigmatism, and earlier visual rehabilitation.

Laboratory investigations have demonstrated significantly higher resistance to wound leakage in laser-fashioned incisions compared with the traditional vertical trephination (240.69 vs. 38.11 mmHg, respectively). Ignacio et al. also observed that femtosecond-created incisions in a top-hat shape led to a seven-fold increase in resistance to wound leakage and less astigmatism.

Clinically, femtosecond-assisted PK has been associated with better visual outcomes, including earlier stabilization of UCVA and BCVA, and lower astigmatism compared with conventional PK. More rapid wound healing, as evidenced by fibrosis along incisions and suture tracks, enabled earlier suture removal.


DALK has become the preferred surgical procedure for keratoconic eyes that are free of deep corneal scarring or hydrops. DALK involves a staged dissection of the stroma down to the level of the Descemet membrane followed by transplantation of an endothelium-denuded donor tissue. However, it is technically more challenging and time-consuming than PK, and several techniques have been introduced, including Anwar’s big bubble technique and injection of intracameral air, to improve surgical success.

In DALK, the host Descemet membrane and endothelium are preserved. This eliminates the risk of endothelial immunogenic graft rejection, although stromal and epithelial rejection can still occur. In turn, DALK entails improved long-term graft survival, reduced endothelial cell loss, and a decreased need for postoperative steroid treatment with its associated complications (cataract, glaucoma). Mitigated risk of rejection also enables for larger grafts. Moreover, given that the procedure remains extraocular, the risk of endophthalmitis is theoretically lessened.

Graft survival after DALK is better than that of penetrating graft PK. In the largest reported series, an average DALK graft survival rate of 99.3% was observed over a mean follow-up period of 4.5 years. Visual and refractive outcomes comparing DALK to PK are similar, specifically in cases of DALK where minimal (equal or lower than 20 microns) or no residual host stroma remains. , Retention of thicker (greater than 30 microns) stromal tissue in DALK can result in significant haze at the graft-host tissue interface and reduce long-term visual outcomes. Interface haze has been reported in 0.7% of cases. Indeed, a single-center comparative study demonstrated equivalent or better visual acuity in eyes following DALK using the big bubble technique when compared with PK. However, visual outcomes were inferior when DALK was performed via manual pre-Descemet dissection.

A recent Cochrane meta-analysis was conducted comparing outcomes in PK versus DALK in keratoconus patients. Only two randomized controlled trials, both conducted in Iran, were included, with a total of 111 participants. Although the quality of evidence was limited, the data indicated that rejection rates were lower in DALK than in PK (odds ratio: 0.33, 95% confidence interval: 0.14–0.81). , , No evidence was found to support a difference in outcomes with regard to postoperative BCVA, graft survival, final UCVA, or keratometric outcomes. The authors felt that there was insufficient evidence to determine which technique led to superior outcomes.

Reinhart et al. also reviewed the literature comparing DALK and PK for any indication (not specifically for keratoconus) and found the two techniques to be comparable in terms of visual and refractive measurements. Shi et al. found greater improvement in BCVA with PK, but no significant differences in refractive measurements, and a significantly lower likelihood of rejection with DALK. In terms of postoperative astigmatism, no significant differences have been observed between eyes undergoing PK compared with DALK, with mean astigmatism of 4.00 to 5.00 D. This is not surprising given that both techniques have similar suturing requirements for securing the donor tissue.

Overall, DALK appears to deliver visual and refractive outcomes comparable to those of PK. With the preservation of host endothelial cells, DALK is associated with lower rejection rates and better graft survival. Conversely, the more technically demanding and time-consuming DALK procedure has contributed to a slower uptake than might have otherwise been expected.


PK remains an important surgical modality for advanced keratoconus, offering excellent graft survival and reasonable visual outcomes. The evolving role of femtosecond laser in PK is promising but warrants further investigation.


Only gold members can continue reading. Log In or Register to continue

Stay updated, free articles. Join our Telegram channel

Oct 30, 2022 | Posted by in OPHTHALMOLOGY | Comments Off on Penetrating Keratoplasty in Keratoconus

Full access? Get Clinical Tree

Get Clinical Tree app for offline access