Since its invention in 1991, optical coherence tomography (OCT) has revolutionized the clinical examination of the anterior and posterior segments of the eye.1
During the past 5 years, several research groups developed devices that allow the use of OCT intraoperatively (iOCT) to gain instant feedback of ophthalmic surgical maneuvers in the anterior segment2 and the posterior segment.3 This experimental technology has been described in various surgical settings, including full-thickness macular hole (FTMH),4 optic-pit maculopathy,5 epiretinal membrane,3 and retinopathy of prematurity.6
The PIONEER study (a prospective, multi-surgeon, single-center study conducted at Cole Eye Institute, Cleveland) evaluated the usefulness of iOCT with a custom-made, microscope-mounted, portable spectral-domain OCT (SD-OCT) system.7 A major challenge for iOCT is the smooth integration into this intricate process. Binder et al. first described a prototype iOCT system with a Cirrus HD-OCT system (Carl Zeiss Meditec, Dublin, CA) adapted to the optical pathway of a Zeiss OPMI VISU 200 surgical microscope (Oberkochen, Germany) in 2011.8 However, in this publication, Binder noted that “a weakness of this intrasurgical SD-OCT system is the complexity of handling it during surgery” — a problem that is inherent for all experimental iOCT systems. In the meantime, the first commercially available iOCT, the Rescan 700 (Zeiss, Oberkochen, Germany), received the CE mark in Europe. It allows imaging of anterior and posterior segment structures.
In this retrospective, single-center, pilot case series study, the utility of this new iOCT system in a routine clinical operating room setting during anterior and posterior segment ophthalmic surgery was evaluated.
Patients and Methods
The study was approved by the local ethics committee in Zurich and was conducted according to the Declaration of Helsinki. All patients agreed on retrospective evaluation of the imaging data.
The Rescan 700 is a SD-OCT system fully integrated into the surgical microscope OPMI Lumera 700 (Zeiss, Oberkochen, Germany). For retinal imaging, the built-in, noncontact, wide-angle system Resight (Zeiss, Oberkochen, Germany) was used. The OPMI Lumera 700 is connected to the CALLISTO eye 3.2 control monitor and video-recording system (Zeiss, Oberkochen, Germany), which allows full control of the iOCT system.
All iOCT imaging during the surgical procedures was recorded as video. Prior to surgery, the surgeons did not receive any commands regarding which anatomical structures should be imaged or regarding the supposed iOCT imaging time. Thus, the total iOCT imaging time in each surgery was the choice of the responsible surgeon. After surgery, all surgeons were systematically interviewed about the performed procedure and asked whether the iOCT provided relevant additional information, as well as whether this information led to altered decision-making. If the answer to either question was “yes,” the surgeons were asked to specify their answer and substantiate it based on the iOCT video recordings. Only comprehensible and documented findings were counted as either additional information or altered decision-making. The total imaging time was determined with the help of the recorded video files. Beyond that, the health records and surgical reports of the patients were reviewed for information on the age of the patient, preoperative diagnoses, surgical procedures, and supplementary information on previous OCT imaging and iOCT imaging. In addition, the possibility for OCT imaging under different dyes and tamponades was studied.
The iOCT-imaging system was used in 40 consecutive surgical cases. All 40 eyes — 18 right eyes and 22 left eyes — were included in the analysis. The mean age of the patients was 70.3 years (standard deviation [SD] = 9.4 years). With regard to the lens status of these 40 eyes, nine were phakic, nine were pseudophakic, and none were aphakic; 22 eyes underwent concurrent phacoemulsification with pars plana vitrectomy (PPV). As surgical intervention, a total of nine exclusive anterior segment procedures, nine exclusive posterior segment procedures, and 22 combined procedures were performed.
Additional Information and Altered Decision-Making in Exclusive Anterior Segment Procedures
The most common indications for surgery in the nine exclusive anterior segment procedures were cataract (n = 6; 66.6%) and glaucoma (n = 2; 22.2%), which were treated by cataract extraction with IOL implantation (CE/IOL) or trabeculectomy (TE), respectively. The other remaining surgery was an explantation of a dislocated IOL followed by an iris-claw lens implantation. Of these nine eyes, six were phakic, three were pseudophakic, and none were aphakic.
The average iOCT imaging time per anterior segment surgery was 1 minute, 57 seconds. In only two of these nine surgeries did the responsible surgeon believe there was a gain of additional information postoperatively using the iOCT imaging. In CE/IOL procedures, it was possible to evaluate the corneal stromal edema following the hydration of the clear cornea incision, and in TE procedures, the surgeon was able to gain detailed anatomical information of the surgically induced structural alterations during surgery. However, neither of the two cases of additional information was evaluated as altered decision-making by the surgeon.
Intriguingly, in the 22 combined (anterior and posterior) surgical procedures, none of the gained additional information or altered decision-making by iOCT was attributed to imaging of the anterior segment.
Qualitative Observations in Exclusive Anterior Segment Procedures
As stated above, it is possible to visualize and evaluate the cornea, including the corneal edema induced by incision hydration. In TE procedures, the exact postsurgical anatomy of the sclera, Schlemm’s canal, trabecular meshwork, and anterior chamber is assessable using iOCT. Furthermore, it was possible to visualize the iris, the anterior and posterior lens capsules, the cortex, and nucleus of the lens; however, these findings were not classified as additional information, since they were also visible with the surgical microscope.
We found the employed injector/cartridge set (AT. Shooter A2-2000/ACM2 1.5 mm; Zeiss, Oberkochen, Germany) and viscoelastic (ProVisc; Alcon, Hünenberg, Switzerland) are transparent and therefore fully iOCT-compatible. Thus, it was possible to visualize all major steps of CE/IOL surgery in detail: the clear cornea incision, capsulorhexis, phacoemulsification, lens implantation, and incision hydration. Some interesting observations could also be made; among them is the instant swelling of the cortical lens fibers upon capsulorhexis, which results in a sudden increase of reflectivity. Also, in some patients, the posterior lens capsule was sprinkled with agglomerations and irregularities.
Additional Information and Altered Decision-Making in Posterior Segment Procedures and Combined Procedures
In the nine exclusive posterior segment procedures, the most common indications for surgery were the presence of an epiretinal membrane (ERM; n = 3; 33.3%), a full-thickness macular hole (FTMH; n = 2; 22.2%), a retinal detachment (RD; n = 2; 22.2%), and/or a lamellar macular hole (LMH; n = 1; 11.1%). Very similarly, in the 22 combined (anterior and posterior) surgical procedures, the most common indications for surgery were, apart from cataract (n = 22; 100%), the presence of an ERM (n = 15; 68.1%), a FTMH (n = 3; 13.6%), a LMH (n = 3; 13.6%), and/or a RD (n = 3; 13.6%). The average iOCT imaging time per surgery in the nine exclusive anterior segment procedures and 22 combined surgical procedures was 2 minutes, 8 seconds, and 3 minutes, 2 seconds, respectively.
The frequency of additional information gained by iOCT, as reported by the surgeons, was highly similar among the exclusive posterior segment procedures (n = 7; 77.7 %) and combined surgical procedures (n = 16; 72.7 %). The top four reasons were: evaluation of the ERM or ILM prior to peeling, reevaluation of the retina after the peeling procedures, identification of vitreous remnants, and intraoperative assessment of a FTMH or LMH. However, there was a surprisingly strong difference in the reported frequency of altered decision-making due to iOCT-imaging. In the nine exclusive posterior segment procedures, altered decision-making was only reported in one case (11.1%), whereas in the 22 combined surgical procedures, altered decision-making was reported in 12 cases (54.5 %). All cases of altered decision-making in combined procedures were solely attributed to posterior segment imaging. The top three findings were: identification of cleavage sites for ERM or ILM peeling (Figure 1), removal of otherwise clinically undetectable vitreous remnants, and influence on the choice of the appropriate gas or silicon oil tamponade.
(A) Delineation of a proper cleavage plane before initiation of peeling maneuver during 23-gauge pars plana vitrectomy for epiretinal membrane (ERM). (B) Immediate reevaluation of the retinal surface right after peeling of the ERM.
Qualitative Observations in Posterior Segment Procedures and Combined Procedures
The ability of the iOCT to enhance imaging and the compatibility with the commonly used dyes triamcinolone-acetonide, brilliant blue G, and trypan blue was assessed. Triamcinolone-acetonide (Triesence; Alcon, Hünenberg, Switzerland) had a contrast-enhancing effect on the iOCT-imaging (Figure 4). Staining of the premacular vitreous by triamcinolone facilitated the iOCT examination of premacular and prepapillary vitreous structures and was evaluated as useful in all 15 triamcinolone-acetonide assisted vitrectomy procedures. The mixture of brilliant blue G and trypan blue dyes (Membrane blue dual; D.O.R.C., Zuidland, the Netherlands), which was used for ILM staining in 25 of the 30 vitrectomy procedures, did not influence the iOCT-imaging and are, therefore, fully compatible. The latter finding is important, since the resolution of the Rescan 700 is not sufficient to make an intact ILM visible (Figure 2). Similarly, the common vitreous tamponade media air, sulfur hexafluoride (SF 6), perfluorocarbon liquid (PFCL), and silicone oil had no influence on the iOCT resolution and are thus also fully compatible. iOCTimaging proved to be especially useful in the two patients with PFCL application during RD repair, since the iOCT image was fully in focus, whereas the optical view through the surgical microscope upon the surface of the retina was altered. Interestingly, it was also possible to visualize the retina using the Rescan 700 iOCT without the use of indirect optics for posterior segment visualization. However, the OCT quality was lower with this approach.
Triamcinolone-assisted vitrectomy with visualization of the bursa premacularis during surgery in an eye with attached posterior hyaloid.
The internal limiting membrane is only visible in a peeled state before complete removal from the retinal surface. The resolution of the Rescan 700 is not sufficient to eliminate chromovitrectomy.
In the five surgeries for RD, an edema and accordion-like microarchitectural alteration within the outer layers of the retina due to the swelling of these layers was detectable (Figure 3). Both of these irregularities persisted even after reattachment of the retina. On the first postoperative day, those alterations were already indiscernible using regular OCT-imaging.
Immediate detection of macula-on or -off situation during surgery. (A) Accordion-shaped microarchitectural alteration of the outer layers of the retina and retinal edema. (B) The microarchitectural alteration persists after reattachment of the retina.
Lastly, the surgeons were able to evaluate the sites of sclerotomy and trocar systems using iOCT. Strikingly, in all 12 cases, in which the sites of sclerotomy were imaged after removal of the trocar, the sclerotomy channel was gapping and remained widely open. The closure and leak tightness was unexpectedly only provided by very superficial scleral layers and the conjunctiva.
Time Burden and Adverse Events
For all three types of surgery, the average imaging time was highly variable, with an average of 117 seconds (SD = 105 seconds) for exclusive anterior segment surgery and an average of 128 seconds (SD = 70 seconds) for exclusive posterior segment surgery. The average imaging time for the combined surgical procedures amounted to 182 seconds (SD = 140 seconds). A trend during the course of the study, indicating either a learning curve or a change in attitude toward iOCT imaging, was not discernible. No adverse events were recorded related to the iOCT imaging, neither due to the expanded time of anesthesia, nor to any direct effect on ocular tissue.
Safety and Time Burden
The time burden in our study was, on average, 2 minutes, 36 seconds — significantly lower than the 5 minutes reported in the PIONEER study,9 which in contrast used a portable OCT mounted on a microscope. The reduction in time burden may translate to fewer anesthesia-related complications.9 The near-infrared spectrum laser of the iOCT seems to be safe for ocular tissues.10 In contrast, excessive exposure to operating microscope light was shown to be phototoxic to retinal tissue and associated with postoperative dry-eye syndrome.11–14
Utility of iOCT in Anterior Segment Surgical Procedures
In our study, the iOCT did not show any significant advantage in anterior segment surgical procedures. The surgeons in our study only reported two instances of additional information gain and no case of altered decision-making. In contrast, Ehlers et al. demonstrated an additional value of iOCT imaging for anterior segment surgical procedures.9 However, the reported 40% of altered understanding/decision-making were reported for Descemet’s stripping automated endothelial keratoplasty and deep anterior lamellar keratoplasty procedures, which are not included in the cohort of this study.9
Utility of iOCT in Posterior Segment Surgical Procedures
The iOCT was found to be beneficial in posterior segment and combined surgical procedures. The surgeons in this study reported that iOCT imaging provided additional information in 23 (74.1%) of the 31 posterior or combined procedures, which resulted in 13 (41.9%) cases with altered decision-making. Interestingly, the latter result is close to the 40% reported by Ehlers et al. in the PIONEER study.9
The difference in the rate of altered decision-making between the exclusive posterior segment procedures (11.1%) and combined surgical procedures (54.5%) can be partially attributed to the differences in the preoperative lens status. In the exclusive posterior segment procedures, the eyes were pseudophakic (66.6%) or phakic without cataract (33.3%). In contrast, in the combined surgical procedures, all eyes (100%) had cataract. Consequently, the preoperative diagnostic procedures might have been impaired in these patients. In one case, the surgeon performed an intraoperative intravitreal injection of ranibizumab (Lucentis; Genentech, South San Francisco, CA) based on iOCT imaging following CE/IOL. It is reasonable to assume, therefore, that iOCT imaging under full anesthesia might facilitate OCT imaging in some patients, such as young children or patients who are handicapped.
The iOCT was classified as very useful by our surgeons in all procedures that involved a vitrectomy. Staining of the premacular vitreous by triamcinolone was especially found to facilitate iOCT examination of the vitreous by having a contrast-enhancing effect.15,16 For MH surgery and ILM peeling, the iOCT was evaluated by our surgeons as very useful, which is in line with most other reports on this matter.3,8,17–21 Usually, iOCT imaging was performed prior to and following the ILM peeling. However, due to the resolution, the ILM is only visible if it has irregularities or if the edge of the partially peeled membrane sticks out. The Rescan 700 system is thus not capable of eliminating dye-assisted ILM peeling (chromovit-rectomy). In ERM peeling procedures, the iOCT was evaluated as useful, as well,3,8,18,20,21 especially since it is possible to detect unstained ERMs and potential cleavage planes.
RD surgery may benefit most from iOCT imaging, especially when utilizing PFCL, since the optical image is altered by PFCL. Additionally, an accordion-like microarchitectural alteration within the outer layers of the retina and persistent subretinal fluid were detectable using iOCT imaging in RD surgery, which subsided within the first day post-surgery.22,23 Whether these microarchitectural changes occur in all cases of repair of RD and are of prognostic value remains an open question.
Another example of an iOCT-exclusive finding is the site of sclerotomy following the removal of the cannula/trocar. Interestingly, at first the former trocar channel in the sclera remains widely open, and the leak tightness of the sclerotomy site is only due to closure of superficial layers of the sclera and conjunctiva. The caliber of the remaining channel might be a predisposing risk factor for postoperative endophthalmitis.
The iOCT, as shown by the PIONEER-study, the preliminary results of the DISCOVER-study, and this study, has the potential to enhance the quality of ophthalmic surgery in the near future.9,24 Randomized, controlled trials will be needed to assess the effect of iOCT on the surgical outcome.