Steve Lenier

home

authors

steve-lenier


In a study carried out by investigators at Duke University’s Vision, Imaging and Performance Laboratory,

 Atalie C. Thompson, MD, MPH, and colleagues developed a segmentation-free deep learning (DL) algorithm using the entire circle B-scan image from 

. 

The aim of the investigators was to determine if such an algorithm would be better at detecting glaucomatous damage than using the 

l nerve fiber layer (RNFL) value provided by the machine’s automated segmentation software.



OCT: Illuminating the retina layer by layer

is usually diagnosed by a combined analysis of clinical parameters and risk factors such as age and IOP, tests for the evaluation of structural damage to the optic nerve, and visual function assessment with perimetry.



Spectral-domain OCT (SD-OCT) is the most commonly used test for structural evaluation. It provides objective quantification of damage that has occurred to the optic nerve head and RNFL. 

To obtain quantitative thickness measurements, conventional SD-OCT software applies segmentation algorithms to delineate the RNFL and extract its thickness.

Although this is generally accurate for diagnosing 

, the algorithms may fail in the presence of segmentation errors (reported in 20% to 40% of scans). 

Even without segmentation errors, interpreting conventional SD-OCT RNFL information can be difficult, due to a large number of summary parameters in addition to maps and plots.



Assessing quality of OCT scans affects clinical-decision process 


Deep learning for segmentation-free analysis

Recent advances in artificial intelligence have led to the development of DL algorithms, which can be trained to accurately detect complex patterns in images.

These algorithms can learn to analyze information across an entire SD-OCT image rather than just a segment, and may potentially provide more information about the presence of glaucomatous damage than individual SD-OCT parameters from segmentation. 

Interpreting the whole image may minimize false positives that arise when clinicians assess multiple individual parameters.



AI screening: Allowing physicians to see physicians more clearly from afar

This was a cross-sectional study of 1154 eyes in 635 individuals, and was undertaken to develop a segmentation-free DL algorithm to assess glaucomatous structural damage, using the whole peripapillary SD-OCT B-scan.

According to Thompson, the algorithm’s performance was compared with that of conventional RNFL thickness parameters.


Macular OCT: A must for treating glaucoma patients

A convolutional neural network was trained to discriminate glaucomatous from normal eyes using the SD-OCT circle B-scan without segmentation lines.

The research team found the DL algorithm had a greater area under the receiver operating characteristic curve than RNFL global and sector parameters, and it was more sensitive at 80% and 95% specificity. This appeared to be even more likely in early disease.



The authors noted some limitations. Although they used an independent test dataset for the final assessment of diagnostic accuracy, external validation in populations from other clinical settings is desirable.

In addition, it should be noted that the diagnosis of 

 is not based on the results of a single test but rather on a combined interpretation of information on risk factors, such as age and intraocular pressure, and results of structural and functional tests. 

Therefore, it remains to be seen how incorporating such an algorithm would affect clinical diagnosis when combined with these other pieces of information acquired in clinical practice, as well as for assessing change over time.

Technology assisting surgeons in capturing spectral-domain OCT



The authors concluded that the segmentation-free DL algorithm they developed performed better than global and sectoral RNFL thickness parameters for discriminating glaucomatous from control eyes, especially in cases of preperimetric or early perimetric glaucoma.

They believe that using this DL algorithm in a clinical setting may improve the accuracy and sensitivity of SD-OCT for diagnosing 

, while obviating the need for error-prone segmentation of 

Future studies should investigate how such an approach contributes to diagnostic decisions when combined with other relevant clinical information, such as risk factors and perimetry results. 


e:atalie.thompson@duke.edu
Dr. Thompson is a senior resident in ophthalmology at Duke Eye Center in Durham, North Carolina.


1. Thompson AC, Jammal AA, Berchuck SI, Mariottoni EB, Medeiros FA. Assessment of a segmentation-free deep learning algorithm for diagnosing glaucoma from optical coherence tomography scans. 

 2020;138(4):333-339. doi:10.1001/jamaophthalmol.2019.5983

Articles

Research finds automated segmentation software examines OCT scans

Investigators target algorithm for diagnosing glaucoma

This article was reviewed by N. Douglas Baker, MD



A new study is one of the first head-to-head clinical reviews comparing a minimally invasive 

surgery (MIGS) device with trabeculectomy in patients with primary open-angle glaucoma (POAG).

The MicroShunt (Santen) is a subconjunctival 

 device made from polystyrene-block-isobutylene-block-styrene (SIBS) material. It is 8.5 mm in length, has a 70-μm lumen, and is highly biocompatible. It enters the anterior chamber at the trabecular meshwork level and forms an aqueous lake beneath the conjunctiva and Tenon’s capsule 3 mm to 6 mm posterior to the limbus.



Douglas Baker, MD, and colleagues have presented 1-year results from a 2-year study to assess the safety and effectiveness of stand-alone MicroShunt surgery (Figure 1) versus trabeculectomy in patients with POAG.

The randomized, single-masked study was conducted at 29 sites, 24 of which were in the United States. 

Patients were between the ages of 40 and 85 years with POAG and were on the maximum tolerated amount of 

 medications. 

The exclusion criteria were previous conjunctival incisional surgery, aphakia, angle closure, and secondary open-angle glaucomas such as post trauma, pseudoexfoliation, or pigment dispersion.

Minimally invasive glaucoma shunt delivers for patients

The primary outcome the investigators were looking for was a reduction in IOP of 20% or greater at 1 year, without increasing the number of medications from baseline. 

They measured IOP over time, number of 

 medications, adverse events (AEs), incidence of persistent hypotony, postoperative interventions, and corneal endothelial cell density.

, the conjunctiva and Tenon’s capsule were removed at the limbus as a unit. The mytomycin (0.2 mg/mL)–soaked Weck-Cel sponge was placed on the sclera and beneath the Tenon’s for 2 minutes. The initial incision in the sclera is 3 mm posterior to the limbus, and the blade enters the anterior chamber at the trabecular meshwork level.

Within the overall study population, the primary effectiveness end point was met 53.9% of the time with the MicroShunt and 72.7% of the time with trabeculectomy. 

If the focus is placed on the group with IOP 21 mm Hg or greater, the MicroShunt met the primary end point 63.8% of the time and trabeculectomy 75% of the time.

Looking at the mean IOP over time, starting at baseline at 21.1 mm Hg, at 1 year the mean in patients with the MicroShunt was 14.2 mm Hg and with the trabeculectomy, 11.2 mm Hg. The mean number of 

 medications was 3 at baseline, and after intervention at the end of 1 year, 0.6 medications for the MicroShunt and 0.3 for the trabeculectomy.

Glaucoma treatment alternatives: Thinking outside the box

There was a 5.2% loss in corneal cell density with the MicroShunt, and a 6.9% loss with trabeculectomy. Persistent ocular hypotony occurred 3.0% of the time with the MicroShunt and 9.9% with trabeculectomy. Cataract progression was seen 12.2% of the time in the MicroShunt group and 17.6% in the trabeculectomy group. 

For visual acuity, loss of more than 2 lines of best-corrected visual acuity occurred 11.1% of the time in patients with the MicroShunt and 14.5% in those who had trabeculectomy. Choroidal effusion occurred in 4.8% of those with the MicroShunt and in 7.6% with trabeculectomy.

Postoperative interventions included suture lysis and medication change, occurring in 40.8% of patients with the MicroShunt and in 67.4% of those who received trabeculectomy. There were no reports of endophthalmitis in either group at 1 year.

Microinvasive glaucoma surgery in children: Is there a role?

This was a prospective randomized study conducted in patients with primary open-angle glaucoma taking the maximum-tolerated number of 

 medications. 

Both trabeculectomy and MicroShunt implantation resulted in a reduction in IOP and supplemental glaucoma medications at 1 year.

In the MicroShunt group, 53.9% of patients experienced a 20% or greater decrease in mean diurnal intra-operative pressure from baseline to year 1. 

This increased to 63.8% in the group that had IOP 21 mm Hg or greater when they entered the study. There was a reduction in mean diurnal IOP, from 21.1 mm Hg at baseline to 14.2 at 1 year.

The mean number of medications decreased from 3.1 to 0.6 at 1 year and 71.6% of patients were medication free at 1 year. Incidents of AEs and corneal endothelial cell loss were similar between the 2 groups, and there were no reports of endophthalmitis at 1 year.

Genetic variants linked with onset, progression of POAG

The study provides 1-year data on the MicroShunt from 29 sites in the United States, the Netherlands, Spain, Italy, France, and the United Kingdom. Patients with POAG on maximum-tolerated medical therapy (baseline IOP, 15-40 mm Hg) and no previous history of conjunctival surgery were included. 

In the study, patients with uncontrolled POAG who underwent MicroShunt implantation achieved, on average, IOP of less than 15 mm Hg up to 1 year after surgery.


N. Douglas Baker, MD, is a consultant to Santen.

Study targets results of MIGS device versus trabeculectomy 

This article was reviewed by Donald S. Fong, MD, MPH



Research is focusing on topical medication adherence and visual field progression in open-angle glaucoma (OAG).

At the 2020 meeting of the American Glaucoma Society (AGS), Donald S. Fong, MD, MPH, of Kaiser Permanente Southern California (KPSC), presented a study designed to assess the effect of treatment adherence to topical 

medications on visual field progression, using data from clinical practice. 





Not surprisingly, the study found that poorer medication adherence was significantly associated with a faster rate of visual field decline.

Elevated IOP is the only modifiable risk factor to reduce the risk of progression in OAG, and first-line treatment is typically topical ocular drops to lower IOP.

“Low medication adherence may increase the risk of damage from 

, but there is a lack of quantitative data on the relationship between treatment adherence and glaucomatous visual field progression,” Fong said.



Emerging therapy for open-angle glaucoma 


To address the lack of data, Fong and colleagues examined records from KPSC, an integrated health system that provides comprehensive care to over 4.6 million members.

Fong noted several benefits of utilizing this population.

“In our integrated health system,” he said, “KPSC members receive their prescriptions from a single source, and patients tend to be long-term members of the health plan.”

He explained that this allows for tracking patient refills, and makes long-term followup possible. The population is also representative of the state of California for racial diversity.




For the study, patients meeting inclusion criteria were stratified by severity of baseline visual field measurements, using readings from 6 months before and after the index date. 

For patients that had no visual field measurements in the time frame, their baseline disease severity was designated as “missing.”

Adherence was measured by proportion of days covered, which was determined as the number of days that the patient was supplied with medication, divided by the number of days in the period of interest.

Patients who received multiple medications were considered adherent if 1 or more of their medications was filled, and adherence was reported in categories of low, moderate, and high.



 Pharmacologic treatment an option for open-angle glaucoma



The investigators hypothesized that the interaction between adherence and time would be positive. In other words, for a given follow-up time, patients with better adherence would have better mean deviation (i.e., less severe disease).

“Our study is the first to use real-world data to quantify the impact of non-adherence on 

 progression,” Fong said. “We demonstrated that treated 

 patients who were more adherent showed less loss of visual function. We need to improve patient adherence to medications because higher adherence will delay glaucoma progression and loss to a patient’s quality of life.”

626/851-6107
Dr. Fong has no financial disclosures related to this subject.

Research is seeking benefits, options for open-angle glaucoma patients

Study targets topical medication adherence, visual field progression

 shunt was developed by AqueSys, which led to the company’s acquisition by Allergan in 2015. Not long after, the product was launched in Europe as the Xen transscleral gelatin stent.

The Xen stent is a 6-mm long hydrophilic gelatin stent, with an inner diameter of 45 micrometers.

It is inserted with a preloaded injector, through a corneal incision, and exits into the subconjunctival space to form a bleb. Usually, mitomycin C is injected prior to stent insertion.



Microinvasive glaucoma glaucoma surgery in children: Is there a role?


The insertion can be performed as a standalone procedure or combined with 

.

In the United States, Xen is licensed as a treatment for refractory 

 in cases where previous glaucoma surgery may have failed or in cases that have become nonresponsive to maximum tolerated medical therapy.

In Europe, however, the stent was launched for use as a primary surgical intervention that may prevent or delay the need for trabeculectomy and that can be performed at the same time as 

Hari Jayaram, MD, PhD, of the Glaucoma Service at Moorfields Eye Hospital in London, recently presented the results of a study of 7 years of data on surgical experience at Moorfields.

The majority of the cases were of the second type mentioned above, as a primary intervention, which indicates that Xen is suited to primary open-angle glaucoma (POAG) when drops are not enough or previous medical treatments have failed.


The investigators performed a retrospective review of all ab-interno Xen procedures that had been performed at Moorfields.

In extracting the data, they paid special attention to the patient’s ethnic origin and primary underlying diagnosis, their 

 and medication history over the time of follow-up, whether or not bleb needling had been performed, and whether or not the patient required further glaucoma surgery.

They also looked at the incidence of complications, in particular hypotony and infection, as well as loss of vision.

Failure was determined using a generous composite outcome, including an 

 definition of less than 20% reduction of 

 from the highest recorded pressure in the patient’s medical record, either with or without the use of eye drops.

Outcomes measured included whether or not a patient was using drops, and failure also included reoperation for 

, endophthalmitis, or loss of light perception vision. Importantly, the need for bleb needling was not included among the criteria for failure.



Pharmacologic treatment an option for treating glaucoma

The study looked at Xen insertions performed at Moorfields from May 2012 to June 2019. This included a total of 426 eyes from 375 patients.

Five percent of the patients had undergone prior 

. Almost two-thirds of the patients had an underlying diagnosis of POAG, and about a quarter had an underlying diagnosis of either uveitic glaucoma or another form of secondary open-angle glaucoma.

There was a wide age range of patients receiving the stents, from 16 to 95 years, with a mean age of 67. The patient population was typical of the ethnic diversity seen in London, with around 44% of patients coming from nonwhite British ancestry.

Sixty percent of the cases were performed as standalone procedures and 40% were combined with 

 Mitomycin C was used almost universally. Some surgeons also used anti-VEGF treatments, and long-acting corticosteroid therapies. 



Sustained-release implant offers long-term IOP control, preserved visual function


The graph at left shows the long-term outcomes, using the above-mentioned definition for failure. At 1 year, 64% of the procedures were found to be successful. This reduced to 41% after 4 years.

One interesting thing noted by the investigators is that almost a third of patients failed within the first postoperative year.

Comparing standalone to combined surgeries showed no significant difference in the survival curve of the 2 groups.

The researchers also looked at the impact of ethnicity. Patients of nonwhite British ancestry showed a trend in increase in late failure, but when viewed over the entire time course this did not reach statistical significance.

The authors believe this is in keeping with the understanding that there is a more aggressive wound healing response in these ethnic groups. However, when successful, and further 

 was not required, patients with the Xen achieved a mean 

The dropless future: SLT as a first-line treatment for glaucoma


There was a mean reduction in medication use from 3 drops prior to surgery to about 1.3 drops over long-term follow-up in patients classified as successes.

In terms of achieving drop independence, about 40% of patients who did not require further 

 surgery were drop free beyond a 2-year period. 




Hypotony occurred in about 1 in 5 patients but responded well to conservative management.

There was only 1 case of cleft formation that required surgical intervention. Bleb needling was very common, with 26% of patients needing at least 1 bleb needling procedure within the first 3 months after 

 was the cause of failure in around a third of patients. Over half of these were revision procedures, which tended to be necessary early on in the postoperative course with a median time to reoperation of 4 months. 



Pearls for conducting GATT in open-angle glaucoma


The overall infection rate was 1.6%. This is lower than that reported in the trabeculectomy group in the Tube Versus Trabeculectomy (TVT) Study (NCT00306852), which was 1.9%.

However, it is higher than for all ocular 

 at Moorfields performed in 2018 (0.02%) and higher than the infection rate (0.3%) derived from annual core trabeculectomy outcome audits performed at the hospital.




The authors concluded that when stent insertion is successful, and further revision 

 is not required, the Xen stent can effectively lower pressure and reduce the need for medications.

They noted, however, that surgeons should be aware of potentially high rates of reoperation and the long-term risks of bleb-related infection, and said that refining surgical technique and careful patient selection in the future may help optimize outcomes.



Dr Jayaram has no financial disclosurers related to this content.

Author | Steve Lenier

Articles

Investigators target algorithm for diagnosing glaucoma

Study targets results of MIGS device versus trabeculectomy

Study targets topical medication adherence, visual field progression

Minimally invasive glaucoma shunt delivers for patients