Pneumatic Vitreolysis: The Little Bubble that Could. Interview with Clement Chan, MD

Eric Nudleman, MD, PhD
UCSD Shiley Eye Institute
Editor, RETINA Roundup

It has been more than 20 years since Clement Chan, MD first reported the use of C3F8 for the treatment of vitreomacular traction1.  In that series of 19 eyes, 18 (94.7%) developed a PVD within 2-9 weeks following gas injection.  In this month’s issue of RETINA, Chan et. al. report an updated series with remarkably consistent results2.  Here, 50 eyes with VMT were treated by the same technique with 86% PVD induction, mostly within the first 3 weeks.  For the subset of eyes with stage 2 macular holes, both studies showed impressive results with this simple intervention – half of the holes closed in 1995, and two-thirds closed in 2017.  These data have been corroborated with independent results from other groups in recent years.  For example, Steinle et al. published a series of 30 eyes of which PVD was induced in 83%3.  These results are approximately twice as effective as published series using ocriplasmin.  We contacted Dr. Chan to get his perspective.  I’m sure you will agree that his responses below are entertaining and enlightening.  Enjoy!

Dr. Chan, how did you first get the idea to attempt pneumatic vitreolysis, more than 20 years ago?

chan_headshotDuring my formative years to become a retinal surgeon, Rob Wendel (My senior fellow 6 months ahead of me in my fellowship program at Wash U in St. Louis) and his senior associate, Neil Kelly in Sacramento had just developed their vitrectomy technique for repair of idiopathic macular holes (MH). It was a revolutionary advancement that became a paradigm shift in the way we manage MH. However, their procedure can only be performed in the OR. I started thinking whether there is a procedure that can induce a PVD and possibly close select MH in an outpatient setting for patients who are not candidates for the OR, or who do not want to go to the OR.  At that time, Dr. Gass had published his detailed descriptions of impending MH (stages 1A and 1B) after enunciating his famous classification of full-thickness MH (stages 2 to 4). Around the same time, Serge de Bustros of Wilmer had organized a multicenter study to investigate potential benefits of vitrectomy for treating symptomatic impending MH, but the results of that study were inconclusive. I was bothered by the lack of a widely accepted outpatient procedure for treating patients with symptomatic impending MH or early full-thickness MH. Just as I completed my fellowship in 1986, Paul Tornambe (my mentor in my residency program) and his former mentor, George Hilton, had concluded their multicenter study in pneumatic retinopexy for repairing select retinal detachments. I began to think whether a similar outpatient procedure can be performed for resolving select impending MH and small MH. This unsolved problem continued to disturb me from time to time, as I started clinical retinal practice in California. A few years later, my interest on this subject was particularly rekindled when a patient presented to my office with vision loss due to VMT and a narrow stage-2 MH in her only seeing left eye. This 59 year-old woman had lost complete vision in her fellow right eye due to a prior CRAO. To complicate matters, she lived alone in an isolated location in Yucca Valley, which was at 4 thousand feet elevation above sea level. Therefore, I was reluctant to perform a standard PPV with placement of a large gas bubble for her MH. In addition, she refused a PPV with silicone oil, as she was worried about enduring postoperative care by herself at the isolated location of her mountain home. I began to explore alternative options again in earnest for this condition.

A few days later, an inspiration suddenly came to me during my morning exercise.  I recalled that investigators in Robert Machemer’s lab and also Stephen Ryan’s lab used an expansile gas bubble to induce a PVD in animal eyes4,5. It was a Eureka moment for me. Further search of the literature also revealed that a number of investigators had successfully employed gas tamponade without a  vitrectomy to close MH in highly myopic  eyes with liquefied vitreous and with or without a retinal detachment in an outpatient office setting6-9. It dawned on me that perhaps a small expanding gas bubble is capable of inducing a VMT release and also close select MH. I decided to try this procedure on my monocular patient with her isolated home located at a high altitude. Given her macular hole status, I decided on a long-acting C3F8 gas bubble to maximize the chance of its closure. Her baseline VA was NLP, OD and 20/70, OS. I injected 0.30 mL of C3F8 gas. To my immense delight and almost disbelief, she developed a beautiful Weiss ring, indicating achievement of a PVD at Week 2. She was compliant with face-down positioning, and the MH was closed at Week 4 with VA recovering to 20/40. By week 6, VA was 20/20, OS. Since that time, I have followed her on an annual basis for the last 20 plus years.

To this day, she continues to exhibit MH closure without progression of her mild cataract.  Her VA remains at the 20/20 to 20/25 range.  Needless to say, I have become her hero forever, since this procedure has allowed her to maintain her independence at the isolated location of her mountain home for more than 2 decades. Given the spectacular success of this initial case, I was emboldened to perform PVL on more patients. Soon, I accumulated my initial cae series of PVL on 19 eyes (13 stage-1B impending MH and 6 full-thickness MH). I was successful in inducing VMT release on 95% of these eyes and also closed 50% of the full-thickness MH. Subsequently, I made a podium presentation on my pilot study at the annual AAO meeting in Chicago in 1993 and also published its results in Ophthalmology in 19951.  That was the beginning.

Illustrative case from Dr. Chan’s original series.

Have you been performing this procedure routinely for VMT and small macular holes over the last 20 plus years?

After my initial case series, I have performed PVL on an intermittent basis throughout the years. The initial presentation of my pilot study in 1993 and its publication in 1995 was met with enthusiasm and interest from some retinal colleagues, but it also generated much skepticism from many other colleagues. Those were the days before the availability of OCT to clearly document vitreoretinal interface changes, so that I had to use B-scan images with much lower resolution to demonstrate the achievement of a PVD in my cases. But the greatest criticism was the fear of an expanding intraocular gas bubble inducing unwanted vitreoretinal traction, leading to a high incidence of retinal tears and detachment. It generated sufficient controversy and uncertainty among many colleagues that the interest soon died down. Nevertheless, I knew that PVL is highly capable of inducing a PVD, and therefore, I continued to perform this procedure on select eyes that I thought were good candidates for PVL after my initial case series. Following the landmark publication of the MIVI-TRUST Trial in the New England Journal of Medicine and FDA’s approval of ocriplasmin in 2012, it appeared that high-tech pharmacologic therapy would be a more elegant way of inducing vitreolysis than PVL, and I thought then that the future of PVL was doomed and its days were numbered.  So, I too jumped on the ocriplasmin bandwagon for a period of time. To my surprise, multiple anecdotal case reports of adverse events associated with ocriplasmin soon appeared, so that a voluntary moratorium of its use was made by many retinal clinicians in the US. Despite the subsequent OASIS Trial and ORBIT Registry demonstrating the safety profile of ocriplasmin, the use of ocriplasmin remains limited in the US by its high cost and the persisting conviction among many retinal clinicians in its adverse potential. Therefore, there has been increasing interest in an alternative procedure for vitreolysis in recent years. Given its low cost and sufficiently high success rate for releasing select focal VMT, PVL is a good candidate to fill this niche. Thus, the increased frequency of my performance of PVL in recent years reflect the general sentiment of my retinal colleagues.

Can you describe your technique?

My technique is quite simple.  It is performed in the office setting with the application of focal anesthesia.  I perform a pre-injection paracentesis to prevent excessive post-injection IOP rise and shallowing of AC commonly encountered with a long-acting expansile gas such as C3F8, with a potential of 4X expansion, making AC tap difficult after gas injection.  After sterile prepping, I inject 0.3 mL of C3F8 gas at either ST or IT pars plana.  The majority of eyes tolerate 0.3 mL of C3F8. For an unusually small eye, the volume may be reduced to 0.2 ml of gas.

 Do you require any positioning?

For eyes with VMT-only and without a MH, face-down positioning is not required, although all patients who have received gas injections should avoid the supine position due to risk of pupillary block glaucoma and cataract worsening.  Face–down position is recommended for patients with advanced VMT associated with only a thin inner retinal flap at high risk of rupturing through and becoming a stage-2 MH.  Partial face-down position for at least 4 days is required for patients with a small stage-2 MH.  PVL does not work for large MH (>200 to 250 microns).

Does phakic status affect your decision to perform pneumatic vitreolysis?

PVL is performed regardless of lens status.  But in fact, prior studies (including ocriplasmin studies) have shown that it is easier to induce VMT-release in phakic than pseudophakic eyes—— I know, it seems to be counter-intuitive, but we found this to be true for PVL cases as well.

Have you attempted using air or SF6 initially, and then C3F8 if it fails?  Are the outcomes as good for C3F8 if another gas has already been tried?

 Yes, SF6 and air have been tried, but the success rates for VMT-release associated with air or SF6 gas are much lower than with C3F8.  Steinle et al and Day et al have studied this subject, showing approximately 50% success in VMT-release utilizing Sf6 gas3,10.  In contrast C3F8 gas yields 70 to 80% successes in VMT release2,3.

Air is expected to yield an even lower success rate than SF6.  In our expanded case series, we have had several eyes that received more than one gas injection (double bubbles and triple bubbles), ultimately resulting in both successful PVD as well as MH closure. Thus, repeat gas injections do not jeopardize a successful outcome. On the contrary, repeat gas injections (touch-up procedures) can improve the success rate of PVL.  Although the same gas (C3F8) was injected each time in our “double-bubble” and “triple bubble” cases, there is no logical reason to expect a poorer final outcome utilizing a shorter-acting gas first followed by a longer-acting gas. However, if one decides to use air or SF6 first, one needs to forewarn the patient that there is greater than 50% chance that a second injection is needed.

By the way, the ability to repeat gas injection is an important advantage of PVL.  There is very little added cost for repeating gas injection to enhance success in comparison to repeating injection of an expensive pharmacologic agent, such as ocriplasmin.

Have you had any complications?  If so, what were they?  (Can you estimate the incidence of tears and detachments in your experience?)

There are no perfect surgical procedures. All procedures are associated with certain complications. It is a matter of whether the benefits outweigh the complications.  The benefits associated with PVL far weigh the associated risks.  In our recent expanded retrospective case series, adverse events occurred in 7% of treated eyes.  They included 2 eyes with retinal tears closed with laser, 1 eye with initial stage-1B impending MH becoming full-thickness MH resolved with PPV, and 1 eye with initial impending MH developing RD resolved with PPV.  Final VA was 20/40, 20/40, 20/30, and 20/70, respectively.

Many others are now validating your results.  Why do you think this has taken so many years to catch on?

All good ideas have to withstand the test of time before they are proven and become widely accepted treatment paradigms.  With few exceptions, every innovative idea that is against conventional wisdom is initially met with skepticism, rejection, and challenges.  Any idea that is too easily accepted without going through the trial by fire runs the risk of a short lifespan and eventual abandonment.  It has been said (viz Yale Fisher) that in ophthalmology, it has taken approximately 20 years for each new ground-breaking idea to go through a process of controversy, challenges, and testing before taking root and becoming widely accepted with  a paradigm shift. Take the example of phacoemulsification for cataract surgery. It took approximately 20 years from the time that Charles Kelman first introduced the concept of phacoemulsification before his wisdom was fully appreciated and it became the standard method of cataract removal in modern-day ophthalmology.  It took at least that long for vitrectomy to be well-accepted for vitreoretinal surgery  by the vitreoretinal community from the time that Robert Machemer first proposed the idea.

When PVL was first introduced for clinical use in 1993, OCT was not available. The timing was off. It took the development of high-quality OCT imaging and also multiple other factors and events outlined above before widespread interest in PVL is generated, and then multiple investigators duplicated similarly successful results with PVL for treatment of VMT.  In conclusion, for PVL, the time is NOW!


  1. Chan CK, Wessels IF, Friedrichsen EJ. Treatment of idiopathic macular holes by induced posterior vitreous detachment. Ophthalmology 1995;102:757–767.
  2. Chan CK, Crosson JN, Mein CE, Daher N. PNEUMATIC VITREOLYSIS FOR RELIEF OF VITREOMACULAR TRACTION. Retina (Philadelphia, Pa) 2017;37:1820–1831.
  4. Thresher RJ, Ehrenberg M, Machemer R. Gas-mediated vitreous compression: an experimental alternative to mechanized vitrectomy. Graefes Arch Clin Exp Ophthalmol 1984;221:192–198.
  5. Miller B, Lean JS, Miller H, Ryan SJ. Intravitreal expanding gas bubble. A morphologic study in the rabbit eye. Arch Ophthalmol 1984;102:1708–1711.
  6. Blodi CF, Folk JC. Treatment of macular hole retinal detachments with intravitreal gas. Am J Ophthalmol 1984;98:811.
  7. Miyake Y. A Simplified Method of Treating Retinal Detachment With Macular Hole. Arch Ophthalmol 1986;104:1234–1236.
  8. Blankenship GW, Ibanez-Langlois S. Treatment of myopic macular hole and detachment. Intravitreal gas exchange. Ophthalmology 1987;94:333–336.
  9. Menchini U, Scialdone A, Visconti C, Brancato R. Pneumoretinopexy in the treatment of retinal detachment with macular hole. Int Ophthalmol 1988;12:213–215.