I wish I hadn’t done that: I’m in big bubble, I mean trouble

This column is the latest article in the ongoing series "I wish I hadn't done that." Past submissions can be read here.

It might be a completely pointless exercise to leave a large gas bubble in the eye at the end of DMEK or DSAEK surgery.

We know that supine posturing after DMEK does not affect detachment rates (which immediately calls into question, what exactly is the bubble doing?), and we know that the size of the bubble (big vs. small) does not make a difference, either.

That said, it does seem necessary — at the end of surgery — to pressurize the anterior chamber with gas to firmly stick the donor tissue to the posterior surface of the recipient cornea. And further, it appears that postoperative detachments should be treated in the same way, ie, repressurizing to restick.

The big risk of the gas bubble, of course, is pupillary block. The most common preventive mechanism is the creation of a peripheral iridotomy (PI) that sits beyond the meniscus of the bubble to avoid occlusion. Seems simple enough.

Recently, though, I experienced the worst pupillary block after DMEK of my life. The patient was operated with phaco-DMEK for combined cataract and Fuchs dystrophy. The operation was uneventful and entailed the routine creation of a large far-inferior PI made with a diathermy handpiece — my preferred technique because it cauterize as it cuts, reducing the risk for postoperative bleeding (Figure 1).

Immediately postoperatively, the air bubble was seen floating above the inferior PI, and the patient was discharged home. On the first postoperative day, a minor peripheral graft detachment was seen, which, by postoperative day 5, had progressed to occupy 30% of the graft surface area. Therefore, a re-bubbling was performed in the office at the slit lamp, in which the anterior chamber was filled with 90% air via an inferior paracentesis. A week later, the detachment looked better but was not completely resolved, so the patient was re-bubbled a second time in the same way.

Two days later, however, the patient presented — in excruciating pain — with a pupillary block and a pressure of 60 mm Hg. Because this had been going on for 48 hours, even releasing all the air via a paracentesis did not break the adhesions between the iris and the peripheral cornea, necessitating an emergency return to the operating room to physically sweep the iris out of the angle.

Two days of sky-high pressure, plus the additional trauma of repeat surgery, resulted in profound intraocular inflammation. This, of course, caused the graft to fail. At the end of the experience, the patient could say she had surgery, plus two uncomfortable re-bubbling procedures, plus one emergency operation and agonizing pain, all for the benefit of counting fingers vision and the need for repeat DMEK. What went wrong?

When the patient first presented in pupillary block, I was surprised that I could not locate the PI at the slit lamp. Even in the OR when sweeping the iris out of the angle, it was not visible anywhere. Apparently, sometime after the surgery, in between the first re-bubbling and the second, the PI had healed and sealed completely. As a result, when she was re-bubbled a second time, there was no escape valve to prevent pupillary block.

Frankly, I had no idea that this was even possible. But the devastating mistake was in assuming the presence of a patent PI. Before the second re-bubbling, I never bothered to check whether the patient still had a functional iridotomy. It seemed to me obvious that she must because I had made it myself only weeks before and because she had already undergone multiple recent anterior chamber air fills.

However, this patient's postoperative course is a testament to the devastating cost of unwarranted assumptions. I learned the hard way that PIs, even large ones, can most certainly seal rapidly and completely and that you should never discharge a patient after re-bubbling without visualizing both the PI and the meniscus of the bubble floating above. It is worth the extra time, trust me.