A lesson in DMEK graft sizing after previous failed PK
Click Here to Manage Email Alerts
This column is a part of the ongoing series “I wish I hadn’t done that.” Past submissions can be read here.
The old, small, failed, penetrating keratoplasty is the bane of my existence.
Source: Jack S. Parker, MD, PhD
Nowadays, we virtually never replace one of these decompensated transplants with a new penetrating graft because the results of endothelial keratoplasty, particularly DMEK, are so good.
When performing a Descemet's membrane endothelial keratoplasty for an eye with a failed previous penetrating keratoplasty, numerous studies have shown that the donor DMEK graft should be sized smaller than (or equal to) the old penetrating graft. Normally, this sizing obligation is a nonissue because the typical PK averages 8 mm in diameter and because most DMEK grafts are 8 mm themselves. Occasionally, however, a small-diameter (7 mm, or less) PK turns up, and in these cases, the surgeon is forced to choose between using a very small DMEK graft or ignoring the well-established sizing guidelines.
Theoretically, the sizing guidelines exist because DMEK grafts that exceed the size of the previous PK, and therefore which overlap the previous graft-recipient interface, are significantly more likely to detach postoperatively. This interface is notoriously grossly irregular, and it stands to reason that slapping a DMEK graft on the back would result in poor adhesion.
Recently, however, I encountered a decompensated 7-mm PK performed 30 years previously. I was amazed by how surprisingly regular the posterior surface appeared, without any gaps, shelves or tissue spurs at the donor-recipient interface. I was so impressed, in fact, that I made the decision to ignore the well-established sizing guidelines calling for a tiny DMEK and to proceed with a “normal” diameter graft.
The surgery was entirely routine. A large descemetorrhexis performed outside of the previous PK interface was initiated and stripped using an inverted Sinskey hook under air. A single continuous sheet of Descemet’s membrane was removed from the recipient eye, which was peeled without interruption from the area outside of the graft, through the graft area itself, seemingly confirming the smooth, continuous nature of the posterior corneal surface. A 10-mm DMEK graft was injected into the anterior chamber, unfolded and lifted to the posterior corneal surface atop an air bubble.
Day 1 postoperatively, it was immediately evident that the graft was not well situated. Large detachments were present everywhere, emanating from the DMEK edges and extending centrally. Subsequently, the patient underwent three re-bubbling procedures, two in the office and one in the operating room, without success (Figure 1). Three months postoperatively, the patient was lost to follow-up, discouraged no doubt by our inability to induce his graft to adhere. Clearly, not using the tiny DMEK was a mistake.
Preoperatively, the diameter of the recipient PK can be gauged in the clinic by using the measurement on the slit beam, which can precisely measure up to 8 mm. To minimize the risk of intractable graft detachment, we should have used a DMEK 0.5 mm smaller than the recipient prior PK; in this case, a 6.5-mm DMEK, to complement the existing 7-mm graft. Undersizing the DMEK by 0.5 mm relative to the PK provides some cushion in case the DMEK graft is not perfectly centered and also permits positioning the graft in the most favorable possible location, ie, away from any tissue shelves or spurs.
Lately, I have been wondering whether we should use as large a graft as possible in DMEK, particularly when the indication for surgery is bullous keratopathy rather than Fuchs’ dystrophy. A 10-mm graft for instance has almost 40% more surface area than an 8-mm graft, and when treating an eye depleted of cells, you might think that every last little bit helps. However, in this case, I learned the painful lesson that in eyes with PKs, less is more.
- For more information:
- Jack S. Parker, MD, PhD, can be reached at Parker Cornea, 3745 Corporate Woods Drive, Vestavia Hills, AL 35242; email: jack.parker@gmail.com.
Collapse
Read more about
Click Here to Manage Email Alerts