Penelope L. Allen

Femtosecond laser-assisted cataract surgery versus standard phacoemulsification cataract surgery: Outcomes and safety in more than 4000 cases at a single center

Purpose To compare the intraoperative complications and safety of femtosecond laser–assisted cataract surgery and conventional phacoemulsification cataract surgery. Setting Single center. Design Prospective consecutive comparative cohort case series. Methods Eyes had femtosecond laser–assisted cataract surgery (study group) or phacoemulsification (control group) by 1 of 5 surgeons. The technique comprised manual corneal incisions and capsulorhexis or laser–assisted anterior capsulotomy, lens fragmentation, corneal incisions, phacoemulsification, and intraocular lens implantation. Results The study group comprised 1852 eyes and the control group, 2228 eyes. Patient demographics were similar between groups. There was a significant improvement in vacuum/docking attempts, surface recognition adjustments, treatment, and vacuum time during the laser procedure in the study group. Anterior capsule tears occurred in 1.84% of eyes in the study group and 0.22% of eyes in the control group (P < .0001). There was no difference in the incidence of anterior capsule tears between the first half and second half of laser‐assisted cases. Anterior capsulotomy tags occurred in 1.62% study group eyes. There was no significant difference in posterior capsule tears between the 2 groups (0.43% versus 0.18%). The incidence of significant intraoperative corneal haze and miosis was higher and the effective phacoemulsification time significantly lower in the study group (P < .001). Conclusions Significant intraoperative complications likely to affect refractive outcomes and patient satisfaction were low overall. The 2 cataract surgery techniques appear to be equally safe. Although anterior capsule tears remain a concern, the safety of femtosecond‐assisted cataract surgery in terms of posterior capsule complications was equal to that of phacoemulsification. Financial Disclosure No author has a financial or proprietary interest in any material or method mentioned.

Intravitreal injections: is there benefit for a theatre setting?

Objective To investigate and compare the rate of endophthalmitis after intravitreal injections performed in an in-office (dedicated procedure room) versus in-theatre setting. Methods A retrospective comparative cohort study was performed of all patients consecutively treated by a single surgeon with intravitreal injection with either ranibizimab or bevacizumab for any recognised clinical indication. All cases received injections between March 2006 and March 2012, during which time all injections were prospectively recorded on an electronic medical record system. A search of the electronic database using a report building system was used to extract the total number of injections into location-specific grouping (ie, in office vs in theatre). Results 12 249 injections were performed over a 6-year period. 3376 of these were performed in the in-office procedure room, compared with 8873 in the operating theatre. Of the 3376 injections performed in office, there were four cases of infective endophthalmitis compared with none of the 8873 injections performed in theatre (p=0.006). In-theatre intravitreal injections were associated with a 13-fold lower risk of endophthalmitis compared to in-office injections. Conclusions The theatre environment is a clinically appropriate location for any intravitreal injection procedures and was associated with a significantly lower risk of infective endophthalmitis in this single-surgeon comparative cohort study.

Effect of femtosecond laser–assisted cataract surgery on the corneal endothelium

Purpose To compare the effect on the corneal endothelium of femtosecond laser–assisted cataract surgery and conventional phacoemulsification cataract surgery. Setting Private clinic, Tasmania, Australia. Design Prospective comparative cohort study. Methods Femtosecond laser–assisted cataract surgery (study group) or conventional phacoemulsification (control group) was performed. The central corneal thickness, central 3.0 mm corneal volume, volume stress index, and central endothelial cell density (ECD) were measured preoperatively and 1 day, 3 weeks, and 6 months postoperatively. Results The study group comprised 405 eyes and the control group, 215 eyes. Postoperative corneal edema was significantly less in the study group at 1 day and 3 weeks. However, the difference was negligible at 6 months. The study group had significant reductions in ECD loss compared with the control group at 3 weeks but not at 6 months (6‐month mean −150 cells/mm2 ± 244 [SD] versus −149 cells/mm2 ± 233). Eyes in the study group with laser‐automated corneal incisions had greater endothelial cell loss at 6 months than eyes in the study group with manual corneal incisions and eyes in the control group (P<.0001). Eyes in the study group with zero effective phaco time and manually created corneal incisions had statistically significantly less endothelial cell loss at 6 months than the other groups (P<.0001). Conclusions Femtosecond laser pretreatment for cataract surgery was associated with a significant reduction in early postoperative corneal edema and endothelial cell loss compared with conventional phacoemulsification; however, the difference diminished with time. Laser‐automated corneal incisions seemed to adversely affect the corneal endothelial cells. Financial Disclosure No author has a financial or proprietary interest in any material or method mentioned.

Anterior chamber flare after femtosecond laser–assisted cataract surgery

Purpose To determine whether postoperative ocular inflammation is less after femtosecond laser–assisted cataract surgery than after conventional phacoemulsification (manual) cataract surgery. Setting Private clinic, Launceston, Tasmania, Australia. Design Prospective consecutive investigator‐masked nonrandomized parallel cohort study. Methods Consecutive cataract patients who had femtosecond laser–assisted cataract surgery or manual cataract surgery by the same surgeon at a single center were assessed. The primary endpoint was postoperative aqueous flare measured by laser flare photometry at 1 day and 4 weeks. Secondary endpoints included retinal thickness measured by optical coherence tomography and slitlamp examination findings at 4 weeks. Results The per‐protocol population comprised 176 patients (100 in laser group; 76 in manual group). Postoperative aqueous flare was significantly greater in the manual cataract surgery group at 1 day (P=.0089) and at 4 weeks (P=.003). There was a significant correlation between effective phacoemulsification time and 1‐day postoperative aqueous flare (r = 0.35, P<.0001). The increase in outer zone thickness measured by optical coherence tomography was less in the laser group (P=.007). Conclusion Anterior segment inflammation was less after femtosecond laser–assisted cataract surgery than after manual cataract surgery, and this appeared to be due to a reduction in phacoemulsification energy. Financial Disclosure No author has a financial or proprietary interest in any material or method mentioned.

A Comparative Cohort Study of Visual Outcomes in Femtosecond Laser-Assisted versus Phacoemulsification Cataract Surgery

PURPOSE To evaluate visual outcomes after femtosecond laser-assisted cataract surgery (LCS) with phacoemulsification cataract surgery (PCS). DESIGN Prospective, multicenter, comparative case series. PARTICIPANTS Consecutive patients undergoing femtosecond LCS or PCS with intraocular lens insertion. METHODS A total of 1876 eyes of 1238 patients (422 male and 772 female) who underwent cataract surgery between January 2012 and June 2014 were included in the study: 1017 eyes from center A and 859 eyes from center B. Cases underwent clinico-socioeconomic selection. Patients with absolute LCS contraindications were assigned to PCS; otherwise, all patients were offered LCS and elected on the basis of their decision to pay (the out-of-pocket cost for LCS). Demographic and postoperative data were collected to determine differences between groups. MAIN OUTCOME MEASURES Six-month postoperative visual and refractive outcomes. Masked subjective refractions were performed 2 to 6 months postoperatively. RESULTS There were 988 eyes in the LCS group and 888 eyes in the PCS group. Baseline best-corrected visual acuity (BCVA) was better in LCS compared with PCS (20/44.0 vs. 20/51.5; P < 0.0003). Preoperative surgical refractive aim differed significantly between groups (LCS -0.28 vs. PCS -0.23; P < 0.0001). More patients who received LCS had Toric lenses implanted compared with PCS (47.4% vs. 34.8%; P < 0.0001). Postoperative BCVA was better after LCS (20/24.5 vs. 20/26.4; P = 0.0003) with a greater proportion of LCS cases achieving BCVA >20/30 (LCS 89.7% vs. PCS 84.2%; P = 0.0006) and 20/40 (LCS 96.6% vs. PCS 93.9%; P = 0.0077). However, PCS cases had more letters gained compared with LCS cases (13.5 vs. 12.5 letters; P = 0.0088), reflecting baseline BCVA differences. Mean absolute error was higher in LCS compared with PCS (0.41 diopters [D] vs. 0.35 D; P < 0.0011). The percentage of eyes within 0.5 D of error from preoperative aim refraction was higher in the PCS group (LCS 72.2% vs. PCS 82.6%; P < 0.0001). CONCLUSIONS Femtosecond LCS did not demonstrate clinically meaningful improvements in visual outcomes over conventional PCS.

Visual outcomes with femtosecond laser-assisted cataract surgery versus conventional cataract surgery in toric IOL insertion

To evaluate the visual outcomes of femtosecond laser‐assisted cataract surgery (LCS) compared with phacoemulsification cataract surgery (PCS) in patients undergoing Toric IOL insertion.

Cystoid macular edema after femtosecond laser-assisted versus phacoemulsification cataract surgery.

Purpose To evaluate the incidence of postoperative clinical cystoid macular edema (CME) associated with femtosecond laser–assisted cataract surgery (Catalys laser system) versus phacoemulsification cataract surgery. Setting Launceston Eye Institute, Launceston, Tasmania, Australia. Design Nonrandomized, single‐surgeon, prospective, comparative cohort case series. Methods Patients who had femtosecond laser–assisted cataract surgery and phacoemulsification cataract surgery between March 2012 and July 2014 were included in the study. The femtosecond laser–assisted cataract surgery group had anterior capsulotomy, lens fragmentation, with or without corneal incisions via femtosecond laser pretreatment. Standard phacoemulsification surgery and foldable acrylic intraocular lens insertion proceeded in all cases. All patients received topical nonsteroidal drops commencing 2 days preoperatively and continuing for 4 weeks postoperatively. The incidence of postoperative clinical CME (confirmed by optical coherence tomography) and comparison between groups were measured. The main outcome measure was the clinical CME rates. Results Of the eyes, 833 had femtosecond laser–assisted cataract surgery amd 458 had standard phacoemulsification cataract surgery. Both groups had similar baseline parameters. There were 7 cases of postoperative CME (0.8%) in the femtosecond laser–assisted cataract surgery group compared to 1 case (0.2%) in the phacoemulsification cataract surgery group, highlighting a trend toward greater cystoid macular edema in the femtosecond laser–assisted cataract surgery group. This correlated with a change in laser treatment speed (due to a software upgrade), suggesting that retinal safety thresholds need further careful analysis. Conclusion Increased CME might be a subthreshold retinal injury safety signal after femtosecond laser pretreatment and warrants further study. Financial Disclosure There are no financial or conflicts of interest for any author.

Whole population incidences of patients presenting with rhegmatogenous retinal detachments within Tasmania, Australia

We report the first whole population study within Australia to investigate the incidence of rhegmatogenous retinal detachment (RRD). All cases of RRD treated in Tasmania were identified between March 2005 and March 2010. During this period, the senior author (BJV) was the only vitreoretinal surgeon in Tasmania; therefore, accurate whole population data could be reported. For analysis of incidence data, a standardized 5-year period with whole year analysis was performed to minimize bias. This study was approved by The Tasmanian Health andMedical Human Research Ethics Committee (REF H0014540) and complies with the tenets of the Declaration of Helsinki. All primary RRDs in Tasmanian residents including non-operative cases were included, and bilateral RRDs were included as separate cases. Preoperative and operative factors were analysed including proliferative vitreoretinopathy (PVR) graded as previously defined by the Silicone Oil Study and subsequent revisions. Exclusion criteria included combined tractional/rhegmatogenous detachments, exudative retinal detachments, prior RRD in the affected eye, retinal tears without evidence of detachment, prior history of posterior segment surgery and a history of penetrating ocular trauma. Subjects not residing in Tasmania but having their detachment managed in Tasmania were also excluded. After exclusion criteria were applied, 231 RRD cases were included for the 5-year-incidence analysis. The overall incidence of primary RRD in Tasmania was 9.7 per 100000 persons per year. This is the first investigation to report the RRD incidence within Australia. Annual incidences, incidences relating to known risk factors and incidence ratios of men versus women are shown in Table 1 with peak incidences relating to risk factors shown in Table 2. The mean age of presentation was 61.6years (+/ 14.2years; range 13–92years). The peak incidence of RRD by age, gender and lens status is demonstrated in Figure 1. Men were more likely to be affected with an incidence ratio of 1.44:1 (P=0.006). The right eye was predominantly affected in 54.0% of cases with a trend to right eye predominance in women (P=0.07). Peak decade of RRD was 60–69years old, with men presenting slightly younger with a mean age of 60.4years compared with 63.5years (P=0.05). Overall, 37.8% of patients were pseudophakic; 9% had a history of blunt trauma, and 8.6%weremyopic greater than 5dioptres. At presentation, the macula was detached in 202 (62.0%) of cases. PVR at presentation was present in 39 eyes (12.0%) with grade B and 26 (8.0%) with grade C PVR. The overall incidence of RRD is similar to that demonstrated bymost studies,which have variably reported incidences between 9 and 13 per 100000. Peak ages, gender distribution, the predominance of right eye detachment and the rates of pseudophakia/trauma/high myopia were also congruent. The higher incidence in men was only partially explained by differences in trauma/pseudophakia/high myopia. Of the 37.4% (n=122) of pseudophakic patients, 11.8% (n=14) provided a history of complicated cataract operation. Given a large Western Australian audit reported a cataract operation complication rate of 1.6%, this supports the hypothesis that complicated cataract surgery increases the risk of RRD. Nine per cent of patients had a background of ocular trauma and presented with a mean age of 12years younger at presentation (P=0.009), confirming trauma as a risk factor for RRD at a younger age. The incidence ratio of trauma-related detachment for men versus women was 1.7:1. A high proportion (13.7%) of phakic patients in this study were myopic greater than 5dioptres, more than double the population rate (4.4%) reported by a Victorian series of 4744 subjects in Victoria, Australia. The incidence ratio of high myopia-related RRD for men versus women was 1.9:1. This cohort was more likely to present without macular detachment (35.7% vs. 61.7%, P=0.006). This has not been reported previously but possibly represents greater awareness of warning symptoms within this group of patients, because of counselling by primary eye services regarding warning symptoms at routine ocular reviews. Patients had a fellow eye detachment in 8.3% of cases (n=28), and 4% had a first degree family history of RRD. Given the incidence of 9.7 per 100000 per year, these proportions are higher than would be expected within the population. Advantages of this study are the well-defined geographic area and population data, covering a whole island. As there was only one vitreoretinal surgeon during the period under investigation, all patients with retinal detachment within the state were

Effect of intravitreal anti-vascular endothelial growth factor agents on blood pressure – a cross-sectional analysis

I ntravitreal injection with anti-vascular endothelial growth factor (VEGF) agents, such as ranibizumab (0.5 mg), bevacizumab (1.25 mg) and aflibercept (2mg), have the theoretical side effect of increasing blood pressure (BP) due to the their downstream effect of decreasing nitric oxide production, thereby inhibiting vasodilation and increasing vascular resistance [1]. When anti-VEGF agents are delivered intravenously, the incidence of new-onset hypertension is documented to be approximately 22.4%, compared with 8.3% for patients receiving placebo [2]. However, the effect of the relatively small doses delivered via the intravitreal route on long-term BP is unknown. Studies have shown that these small doses enter the systemic circulation; however, whether this translates into a clinical effect on BP is yet to be determined [3]. One recent study highlighted that the risk of hospitalization due to stroke or transient ischaemic attack in the month after treatment is higher in patients undergoing treatment with intravitreal ranibizumab; however, underlying issues with BP were not evaluated in the study [4]. The WHO defines the cut-off for safe SBP at 140 mmHg and DBP at 90mmHg [5]. Hypertensive emergency is defined as BP greater than 180 mmHg SBP and 120 mmHg DBP, with evidence of end organ damage. Hypertensive urgency is defined as the same elevation in BP, without evidence of end organ damage. Both are well documented to occur more frequently in patients with a history of hypertension when exposed to an additional inciting factor [6]. We performed a novel cross-sectional, observational study of patients undergoing long-term intravitreal antiVEGF injection for management of choroidal neovascular membrane due to age-related macular degeneration or cystoid macular oedema secondary to central retinal vein occlusion or diabetic retinopathy. Our aim was to highlight the rate of hypertension and hypertensive urgency amongst patients undergoing intravitreal injections with anti-VEGF agents. The first consecutive 251 patients receiving injections from January 2015 to August 2015 at a single centre were selected to participate in the study. Patients received intravitreal injections of ranibizumab, bevacizumab or aflibercept. Informed consent was obtained from the patients. Ethics approval was obtained through the University of Tasmania Human Research Ethics Committee (H1104344). Primary care physicians were contacted and asked to provide the patient’s baseline BP reading (prior to the first intravitreal injection). BP and pulse rate were measured prior to injection and 10min after injection. In addition, BP measurements were taken as patients presented for clinical review, on noninjection days. McNemar’s test was used to investigate the proportion of hypertensive patients (SBP> 140 mmHg or DBP 100 mmHg) at baseline, clinical review and injection. In our cohort, the mean age was 77 years (SD 10.1, range 33–97 years). A total of 251 patients (107 men and 144 women) were involved in the study. Baseline BP readings were collected for 123 patients, and random BP readings on a noninjection day were collected on 52 patients. Of the 251 patients, a total of 54 (22%) had a SBP reading greater than 180 mmHg on the day of the procedure. In addition, 198 (79%) patients were hypertensive (SBP> 140 mmHg) on the day of the procedure compared with 40 (33%) patients with baseline BP readings. Of the total cohort, 151 (60%) patients were diagnosed with preexisting hypertension and were being treated with antihypertensive medication by their primary care physician. Of the 100 patients without a preexisting diagnosis of hypertension, and therefore not on any antihypertensive medication, 72 (72%) had a SBP greater than 140 mmHg on the day of the procedure. There were 83 patients who were normotensive at baseline, with 64 (77%) of these hypertensive (>140 mmHg) on the day of procedure (P< 0.001). The data from our small cross-sectional study highlight that a significant percentage of patients undergoing intravitreal injections with anti-VEGF agents have SBP above the WHO-recommended safe BP level of 140mmHg. According to the Australian diabetes, obesity and lifestyle study conducted in 1999–2000, approximately 28% of people aged over 25 years in the general population within Australia were hypertensive (SBP> 140 mmHg) [7]. Although not directly comparable given our study cohort reflects an older population, our study nonetheless highlights that there are a significant number of patients being treated with anti-VEGF agents who are significantly hypertensive associated with their treatment. We found 79% of patients experienced elevated BP on the day of the procedure compared with 32% at baseline and 36% of patients with elevated BP at random BP readings. Interestingly, a similar proportion of patients without preexisting hypertension had elevated BP on the day of the procedure compared with patients with known preexisting hypertension. A large percentage of patients experience hypertensive urgency on the day of the procedure (22% on day of procedure compared with 8% for random BP reading and 1% prior to the initiation of anti-VEGF therapy). These episodes of dangerously high BP are most likely related to the anxiety associated with the procedure, rather than an effect of the drug. It has long been recognized that autonomic nervous system influences play a fundamental

Femtosecond laser assisted versus standard phacemulsification in toric intraocular lens insertion

We thank the authors, Levitz et al., for their interest in our paper regarding visual outcomes of toric intraocular lens (IOL) implantation in patients undergoing standard phaco-emulsification versus those undergoing additional femtosecond laser pretreatment. Our prospective real-world comparative studywas established in 2012, comparing femtosecond laser with standard phaco-emulsification, prior to publication of the sentinel references providedbyLevitz. Best-corrected visual acuity and letters gained were utilized as part of our study design, as no evidence existed demonstrating comparative safety or efficacy of femtosecond laser pretreatment with phaco-emulsification. Our toric IOL analysis represents a subset analysis from the overall prospective comparative cohorts rather than a studydesigned specifically to assess toric IOLs. We agree with Levitz et al. that 4years down the track, some of these study design measures as presented are too crude regarding providing a full assessment and interpretation of toric IOLs and the influence on themof Femtosecond laser. We are grateful therefore for Levitz et al., encouraging us to revisit our data, re-presenting in Table 1 relevant information, including refractive and uncorrected visual acuity results. As previously identified, there exists selection bias in our cohorts, but together, both cohorts represent all consecutive cataract cases presenting for surgery in a real-world unselected fashion. Our selection bias was firstly clinical, with ~5% of cases not considered suitable for femtosecond laser pretreatment (small pupil, corneal scarring, etc.), all of these (potentially more complicated) cases entering the phacoemulsification cohort. The remaining selection bias was socioeconomic, with all remaining (~95%) patients with cataract self-selecting either femtosecond laser pretreatment or phaco-emulsification based on their ability to pay for the additional femtosecond laser pretreatment procedure. We agree that it is unfortunate that our study design did not look inmore detail prospectively at differences in co-morbidities that may have arisen from such clinico-socioeconomic selection. To do so retrospectively though would be difficult and would introduce observer bias; consequently, we are left with speculation. We are mindful that our results can be interpreted in one of twoways: the Femtosecond laser protagonist seeing the outcome differences as evidence of Femtosecond laser benefit and the phaco-emulsification protagonist seeing the absence of pre-operative to post-operative change differences as evidence of a lack of benefit of femtosecond laser. We do not think that our study design is capable of definitively answering this question. Although we found a difference between cohorts in percentage patients achieving uncorrected visual acuity 6/12 or better, this difference persisted even with best correction (Table 1). Therefore, we think that the absence of significant refractive outcome differences between cohorts (residual cylinder [refractive astigmatism] and MAE) suggests that the answer lies more likely in baseline differences between cohorts (co-morbidities) rather than a real difference between the procedures in refractive outcomes (such as precision of capsulotomy or lens placement might afford from femtosecond laser). As such, we do not think that our results and discussion are incongruous, and based on our results, wedonot think that LCS canbe considered superior to PCS for visual outcomes in toric IOL surgeries. Our enthusiasm for this disruptive technology should be tempered by the lack of well-powered randomized controlled trial results showing significant benefit now 5years after femtosecond laser pretreatment for cataract surgery entering the mainstream marketplace.