Giessen 2019 – Day 3

Jump Direct to Video

Saturday – 31st August 2019

Session 8 – Secondary Cancer and Survival in Retinoblastoma

Retinoblastoma and second primary malignancies: a Dutch overview and update

Annette Moll

Amsterdam University Medical Centers, location VUMC

Due to their genetic predisposition (RB1 mutation), the heritable retinoblastoma (Rb) patient is at risk for the development of second primary malignancy (SPM). Throughout their lives, survivors of heritable retinoblastoma are susceptible to various but discrete types of SPMs. This risk of developing SPMs is increased even further by therapy like external beam irradiation and this in combination with chemotherapy. Survival from Rb exceeds 95% in high and middle income countries. While no significant changes in mortality were found for non-heritable Rb survivors, due to SPMs heritable Rb survivors in the Netherlands have a significantly increased standardized mortality ratio (SMR) of 12.8 (95% CI: 9.6-16.5). Consequently, SPMs are now the leading cause of death in patients with heritable retinoblastoma. Patients with SPMs that survive after treatment are at risk for third, fourth and fifth malignancies. The risk of a third primary malignancy was increased 8-fold in Dutch patients that have suffered an SPM (SIR=8.5 (95% CI: 3.7-16.7) ) and the Absolute Excess Risk (AER) increased to 202 excess malignancies per 10,000 person yearsThe Dutch nationwide retinoblastoma cohort has recently been updated (1945-2017) and now contains 910 Rb patients of whom 524 non-heritable and 386 heritable. Among the heritable patients 98 SPMs were diagnosed with a SIR of 10.4 (95% CI: 8.4–12.7) for all malignancies combined and an AER of 85.5. With extended follow-up the risk for sarcomas of soft tissue (SIR=149.6 (95% CI 91.4-231.0) ) and bone and joint (SIR=188 (95% CI 107.7-305.8) ) remains high in this group. When looking at therapy, radiotherapy (SIR 11,9 (95% CI 8.9 – 15.6) ) and radiotherapy combined with chemotherapy (SIR 20.8 (95% CI 13.5 – 30.7) ) show a significantly higher risk for all malignancies compared with local therapy and/or enucleation (SIR 4.7 (95% CI 2.7 – 7.5) ) while the risk from sole chemotherapy (SIR 11.3 (95% CI 1.6 – 48.2) ) is based on two events alone. The next step is to pool our data with other international retinoblastoma cohorts in the IRISC (International Retinoblastoma and Second Cancer) consortium in order to have sufficient power to better identify factors responsible for SPM development.


Conflict of interest disclosure: none

Study of genetic predisposition to uveal melanoma

Mohamed Abdel-Rahman1

Sample, Klarke1; Pilarski, Robert1; Grossel, Timothy1; Kinnamon, Daniel1; Boru, Getachew1; Massengill, B. James1; Schoenfield, Lynn1; Kelly, Ben2; Gordon, David2; Johansson, Peter3; Marino, Meghan J.4; Singh, Arun4; Walsh, Tom5; Walsh, Silvia Casadei5; Davidorf, Frederick H.6; White, Peter2; Tischkowitz, Marc7; King, Mary-Claire5; Hayward, Nicholas3; Cebulla, Colleen1

1. The Ohio State University, Columbus, OH, USA
2. Nationwide Children’s Hospital, Columbus, OH, USA
3. QIMR Berghofer Medical Research Institute, Brisbane, QLD, Australia
4. Cole Eye Institute, Cleveland Clinic, Cleveland, OH, USA
5. University Of Washington, , Seattle, WA, USA
6. Havener Eye Institute, Columbus, OH, USA
7. East Anglian Medical Genetics Service, Cambridge University Hospitals, Cambridge, UK

Purpose: The aim of this study was to identify germline mutations in cancer susceptibility genes in uveal melanoma (UM) patients with strong personal and/or family history of cancer.
Methods: A total of 164 UM probands, including 33 with familial UM, were studied. Germline mutation and large deletion in
BAP1 were assessed by direct sequencing and multiplex ligation probe amplification (MLPA), respectively. Whole exome sequencing and cancer gene panel testing were carried out on patients with no detectable BAP1 alterations. Functional validation was carried out by combination of genomic and proteomic studies.

Results: We identified actionable pathogenic variants in nine known hereditary cancer predisposition genes BAP1, PALB2, MLH1, MSH6, CHEK2, SMARCE1, ATM, BRCA1 and CTNNA1 in 18 patients, including 9/33 (27.3%) with familial UM and 9/131 (6.9%) of UM patients with strong personal and/or family history of cancer. Seven patients had pathogenic variants in BAP1, one had a whole BAP1 gene deletion, three had pathogenic variants in CHEK2; two in PALB2; while variants in the other genes each occurred in one patient. Biallelic inactivation of BAP1, PALB2 and MLH1 was observed in tumors from the respective patients. The frequencies of pathogenic variants in BAP1, PALB2, MLH1 and SMARCE1 in UM patients were significantly higher than the observed frequencies in non-cancer controls [p= 5.7 -18, OR: 940 (198-4460); p=0.03, OR: 7.6 (1.9-31); p=0.048, OR: 21.9 (2.9-167); p= 0.0012, OR: 1770 (108-28984), respectively].

Conclusions: Together with published data the study provides definitive evidence of gene/disease association of germline BAP1 with predisposition to UM, as well as, moderate evidence for germline mutation in PALB2 and MLH1. It also identifies several other candidate susceptibility genes. The results suggest locus heterogeneity in predisposition to UM. Genetic testing for hereditary predisposition to cancer is warranted in UM patients with strong personal and/or family history of cancers.

Conflict of interest disclosure: none

Prognostic information for mosaic and high penetrant carriers of RB1 mutations

M. Ashwin Reddy1
Butt, Mussa1; Hinds, Anne-Marie1; Duncan, Catriona1; Price, Elizabeth2; Sagoo, Mandeep1; Onadim, Zerrin1
1. Barts Health NHS Trust, London, United Kingdom
2. Moorfields Eye Hospital NHS Foundation Trust, London, United Kingdom

Purpose: The role of mosaicism in Retinoblastoma (Rb) is increasingly being recognised as molecular genetic tests become more sensitive. Previous studies have been based upon disease eye ratio and not shown striking differences between mosaic and heterozygous RB1 mutation carriers. Providing information regarding laterality and tumor number for mosaic and heterozygous RB1 mutation carriers would be helpful for clinicians and parents.
Methods: A retrospective analysis of mosaic and heterozygous RB1 mutation carriers (low penetrant (LP) and high penetrant (HP) ) from 1992 to 2017 was conducted. Tumor number per eye was assessed in patients classified with mainly A, B and C tumors using the International Intraocular Retinoblastoma Classification system. Patients with D or E group eyes were assessed based upon age at diagnosis.

Results: Data were analysed for 107 patients, 64 were full germline familial patients (53 HP and 11 LP) and 43 mosaic patients. 25% of high penetrant patients were unilateral at presentation and 9 of 13 (69%) developed tumors in their previously unaffected eye. 72% of mosaic patients were unilateral and only 1 of 31 (3%) developed tumors in their unaffected eye. Age at diagnosis was higher in mosaic patients (median 16 months range 2-117) than HP patients (median 7 range 2-33) (p<0.001). Tumor number per eye was lower in mosaic patients (median 1.5 tumors range 1-6) than highly penetrant patients (median 3 range 1-8) (p=0.009). There were only 3 gaugeable eyes regarding tumor number with LP.

Conclusion: This is the first study to provide prognostic information in the form of tumor number. Children with mosaicism have fewer tumors in eyes with Rb compared to HP carriers and are more likely to remain unilateral.

Conflict of interest disclosure: none

Outcomes of RB1 Gene Testing from blood samples of 113 Retinoblastoma survivors and their families (398 in total) collected on a single day at Aravind Eye Hospital, Coimbatore, India

Parag Shah1
Vanniarajan, A.2; Narendran, Venkatapathy1
1. Aravind Eye Hospital, Coimbatore, India
2. Aravind Medical Research Foundation, Madurai, India

Purpose: To report the outcomes of RB1 gene testing of blood samples from 113 retinoblastoma survivors and their family members (398 in total), collected on a single day at Aravind Eye Hospital, Coimbatore, India.

Methods: During the world retinoblastoma week celebration, blood samples of 113 retinoblastoma survivors along with their immediate family members were collected on a single day on 13th May 2017 after taking an informed consent and documenting a detailed family history along with pedigree chart. Ethics committee approval was taken. Overall 398 blood samples were collected of which 113 samples were of known retinoblastoma survivors and 285 of their family members. DNA was extracted and stored at -200 C after initial quality check. RB1 gene mutation analysis was performed by Sanger Sequencing. Large deletions and duplications were analyzed by Multiplex Ligation dependent Probe Amplification. Genetic counseling was given to all the cases two years later on 18th May 2019 during this year’s world retinoblastoma week celebration.

Results: Of the 113 cases, 65 were unilateral and 48 were bilateral. Totally 60 germline mutations were identified. Inherited germline mutations were identified in all the affected family members in 5 families with known clinical history of retinoblastoma, in 12 patients (5 bilateral and 7 unilateral), mutations were identified in a family member other than proband, however they did not had RB. De novo germline mutation were identified in 34 patients (29 bilateral and 5 unilateral). In 9 families, only one parent was available for analysis and hence those 9 germline mutations could not be categorized as inherited or de novo. 9 germinal mutations were not know whether it is de novo or inherited (8 bilateral and 1 unilateral). Two novel nonsense mutations causing truncated RB1 protein was identified in our patients.

Conclusion: Overall, mutations were identified in 47/48 (98%) and 13/65 (20%) in bilateral and unilateral patients respectively, which indicates the highest mutation detection rate. The need for continuous surveillance in patients and family member with RB1 mutations were explained. Reduced risk of retinoblastoma in siblings and offspring was a greater relief for families (57%) with no RB1 mutations.

Conflict of interest disclosure: none

Type of RB1 mutation and age at diagnosis of familial retinoblastoma screened from birth

Milo van Hoefen Wijsard1
Fabius, Armida2; Charlotte, Dommering2; Maka, Erika3; Vishnevskia-Dai, Vicktoria4; Kebudi, Rejin5; Hadjistilianou, Doris6; Carol, Shields6; Munier, Francis7; Lumbroso, Livia8; Kivelä, Tero9; Moll, Annette2
1. Amsterdam UMC, Location VUMC, Amsterdam, Netherlands
2. Amsterdam UMC Amsterdam, Netherlands
3. Semmelweiss University, Budapest, Hungary
4. Goldschleger Eye Institute, sheba Medical Center, Tel-Aviv, Israel
5. Istanbul University, Istanbul, Turkey
6. University Of Siena, Siena, Italy
7. Wills Eye Hospital, Philadelphia, PA, USA
8. Jules-Gonin Eye Hospital, Lausanne, Switzerland
9. Institut Curie, Paris, France
10. University Of Helsinki, Helsinki, Finland

Purpose: Currently, the guidelines for screening children at risk for familial retinoblastoma (Rb) are a balance between early detection and the least exposure to potentially harmful and expensive diagnostics (general anesthesia). If RB1 mutation type influences the time of Rb tumor presentation, personalized screening based on germline mutation could reduce unnecessary exposure to anesthesia. The aim of this study is to explore the genotype-phenotype relation between the germline RB1 mutations and the age at detection of the first Rb tumor.
Methods: All patients at risk for familial Rb who were screened from birth and with a known RB1 gene mutation from nine study centers were included in this study. Patients were categorized by mutational type based on its effect on the RB1 protein and its presumed penetrance based on previous literature: Truncating, Altered Splicing, Large Deletions, Missense and regulatory variant mutations. We compared median time to a first Rb tumor between the groups using survival analyses and Log-Rank testing.

Results: One-hundred-forty-one (141) patients from 115 families were included. After a median follow-up of 6.6 years, 128 patients developed at least one Rb tumor while 13 patients did not develop any Rb tumor. In Truncating mutations (median time to a first Rb tumor ‘MTFT’ (95%CI) = 19 days (10-28) ) and Large Deletions (MTFT = 20 days (11-29) ) first Rb tumors occur sooner and more often than in Altered Splicing (MTFT =65 days (0-202) ) and Missense and regulatory variant mutations (MTFT = 220 days (0-726) ) (Log-Rank; p<0.03). Although the difference in time to a first Rb tumor is significant the majority of first tumors still present in the first 2 months.

Conclusion: Low penetrance RB1 mutations not only lead to less cases of Rb but the first Rb tumor also occurs significant later in time. Still half of first Rb tumors are diagnosed within the first two months, this underscores the importance of early and regular screening in familial Rb, even in patients with a low penetrance mutation.

Conflict of interest disclosure: none

Session 9 – Patients in Focus

What could be the EYE-EYE role in Rare Eye Diseases care in Europe?

Helene Dollfus
Cargo, Center For Rare Eye Diseases, Hôpitaux Universitaires De Strasbourg, Institut De Génétique Médicale D’alsace (IGMA) , Strasbourg, France

ERN-EYE, dedicated to Rare Eye Diseases (RED), constituted by 29 healthcare-providers from 13 Member States, with important interactions with patient groups (European Patient Advocacy Groups noticeably), is organized to cover all RED conditions in four thematic groups: retinal RED, neuro-ophthalmology RED, paediatric ophthalmology RED & anterior segment RED. Moreover, six transversal working groups are addressing issues common to the four main themes. The objectives of ERN-EYE are wide, mainly patient-centered, with patients’ diagnosis and care improvement across EU, registry strategy, education and training program and development of guidelines noticeably. The working group dedicated to paediatric Rare Eye Diseases is currently working on all these topics.

The creation of a virtual Clinic, EyeClin, to better diagnose and treat patients is the cornerstone of ERN-EYE. EyeClin was built thanks to a Clinical Patient Management System (CPMS) common for all ERNs and provided by the EC, in addition to a customized eye-dataset to fit ERN-EYE needs specifically. All experts registered had declared their field of expertise to help users to request their help. Each specialty is clearly represented in the system and the paediatric group organizes specific regular meeting to discuss difficult cases. The launching of EyeClin bring expertise to a large number of RED-affected EU citizens and stimulate their participation to initiatives generated or recognized by ERN-EYE such as nourishing registries, empower research, stimulate trials.

As we are in the middle of the 5 first years of the projects, a lot of initiatives are currently being developed. All our communication is centralized on a dedicated website

Conflict of interest disclosure: none

Achieving meaningful patient research partnership: Development of the Canadian Retinoblastoma Research Advisory Board

Helen Dimaras
Moses, Catherine1; Flegg, Kaitlyn1; Ristevski, Ivana1
The Hospital For Sick Children, Toronto, ON, Canada

Introduction: Research on retinoblastoma survivors indicates that the disease has lifelong effects on daily life regardless of prognosis. Our prior research shows that retinoblastoma survivors and their families keenly interested in being involved in retinoblastoma research. In response to this, we developed The Canadian Retinoblastoma Research Advisory Board (CRRAB) as a way to generate meaningful and sustained involvement of patients in research.

Methods: CRRAB was created in 2016. It is composed of a general membership including people affected by retinoblastoma, clinicians, allied healthcare providers, researchers, engagement experts and policymakers. Led by a steering committee elected at an annual general meeting, three smaller working groups meet monthly by teleconference to lead tasks related to patient engagement, research advisory and priority setting/research development. To inform implementation of CRRAB and its activities, we conducted focus groups to uncover patient knowledge, experiences and preferences about research. Discussions were transcribed verbatim and analysed by thematic analysis.

Results: CRRAB resulted in the development of: a prospective ‘Research Registry’ of people willing to engage in research; the launch of a ‘Retinoblastoma Champions Program’; social media channels to disseminate research findings and opportunities; and a ‘Top 10’ list of jointly identified research priorities. Findings indicated that patients viewed their experiential knowledge as valuable to improving care and directing research and distinct from theoretical knowledge of the cancer held by clinicians. Patients recognized gaps in their theoretical knowledge of retinoblastoma, and wished this education to be provided from trusted sources. Patients were willing and motivated to engage in research as more than study subjects (a role many had experience with), but noted barriers (i.e. time, compensation, training).

Conclusions: CRRAB serves as an effective medium to generate new knowledge from the patient perspective, and translate this information into sustainable long-term patient engagement. We consider the findings suggestive of a novel role for research engagement: addressing patient educational needs through the establishment of trusted connections with researchers. The findings will be used to refine CRRAB activities to provide research engagement opportunities that meet the needs of patients.

Conflict of interest disclosure: none

DEPICT HEALTH “full view for life” for circle of care including families will empower research

Brenda Gallie

The Hospital For Sick Children, Toronto, ON, Canada

 Purpose: To reach the 9,000 children newly diagnosed with retinoblastoma each year (globally 70% die, with resources and knowledge >95% survive), we aim to facilitate collaborative care and research through a retinoblastoma-specific database and make retinoblastoma a zero death cancer. The paucity of quality evidence for retinoblastoma care due to disease complexities confounds conventional clinical trials.

Methods: DEPICT HEALTH v1 facilitated clinical care and research in one center for years. The core of DEPICT HEALTH (now hosted globally using high-performance cloud computing) is a language-independent timeline graphically displaying data (See Figure).

Results: The International Retinoblastoma Consortium (IRbC) established a governance process for DEPICT HEALTH that includes Steering, Operations and working Committees, modeling participation of every collaborating Center of Excellence. Centers join with service agreements; children/parents provide consent for use of DEPICT HEALTH for clinical communication across their care Center, enroll others (family, doctors) to view, and separately consent for Open research on their child’s coded data. Each Center receives their own patients’ data for research; other Centers and researchers will apply to IRbC for collective coded data for ethics approved projects and be encouraged to collaborate. Artificial Intelligence/machine learning on DEPICT HEALTH Real World Data can improve speed and quality of clinical research.

Conclusions/Significance: The DEPICT HEALTH model is uniquely suited to enhance retinoblastoma standard clinical care by supporting multidisciplinary teams and facilitate innovative future research by access to point-of-care data from children everywhere.

Conflict of interest disclosure: none

The PRO RETINA patient registry

Franz Badura
ProRetina Deutschland e.V.

The PRO RETINA patient registry aims to minimize the timespan between recruiting patients and conducting a new clinical study. The goal of this patient registry is to encourage as many affected people as possible to provide clinical and genetically data. Therefore, the patient registry can help to establish new therapeutic concepts and options for rare inherited retinal diseases. Due to the lack of therapy perspectives, many patients do not contact a physican, as this does not appear promising to them. Medical centres, clinical trial centres and affected persons will be contacted and receive information about the patient registry.

Registerd patients, who are interested in clinical studies, will obtain informations via the patient registry. Therefore, the patient data must be entered in the corresponding database for evaluation according to age, disease and genetic defect. The evaluated data are compared with databases of clinical studies. Suitable patients will be informed and advised about the possible clinical study.

Conflict of interest disclosure: none

Leave no one behind –Patient’s perspective

Christina Fasser
Retina International

Patients with inherited retinal degenerations dreamed for years to get therapies that would slow down or stop or even cure the disease. After many years of basic research, a huge amount of knowledge was achieved and the complexity of this disease, respectively of this group of disease became evident. Since 2005 the number of clinical trials increased exponentially and thus the hope of many patients and their families around the world. Gene therapy is a reality today for a few patients with a very rare form of IRD. With upcoming therapies and treatment trials patients have to face new challenges. So far in many countries genotyping was not considered to be necessary to diagnose an IRD, or even was unavailable. Patients that had received their diagnosis 10 years ago or earlier may have a different diagnosis or better name of their disease that they would get today. The descriptive name of the disease is today often replaced by the genotype or the disease is positioned into a new category: Early onset RP might be today called LCA, LCA10, LCA2, CEP209 or RPE65 thus adding to the patient’s confusion. Therefore, it is challenging for patients to know whether a clinical trial or treatment option is addressing their own disease. Apart from diagnosis, patients have to consider very new questions: When should I be treated, should I let treat both eyes, what is the impact should there be a severe adverse effect, what happens if there will be a better treatment on the horizon?

What are we to expect from a treatment? What would be for me the best outcome and what would be the worst? These are questions that have to be discussed among patients in our community as well together with the specialists. However, with the excitement of new therapies, we should not forget all those, for which no treatment is available or even worse, where neither the diagnostic capacities nor the financial means are available to access treatments that would be available. We should not forget that auxiliary aids, low vision and rehabilitation are extremely valuable tools to help to master life with visual impairment or blindness.

The new therapies on the market are paving the way for future new treatment options and our society has to seek for a positive way to make them available also to those that had not the luck to be born in the highly industrialised world by addressing the disparities. Politicians world-wide have agreed on the sustainable goals 2030 with a clear commitment to the wellbeing of all and access to the health system for all. Let us together challenge politicians to answer to this goal by supporting people with inherited retinal degenerative diseases and let them have access to diagnosis, treatment and rehabilitation: In short to have equal access to the civil society.

Conflict of interest disclosure: none

François Lecture 2019

Francois Lecture 2019 – Introduction

Biomarkers in IRDs: scientifically valid – clinically relevant

Birgit Lorenz
Department of Ophthalmology, Justus Liebig-University Giessen, Giessen, Germany

Purpose – Background: The development of expensive therapies for inherited retinal degenerations (IRDs) requires at the same time the identification of novel biomarkers to quantify therapeutic effects unequivocally, and to demonstrate patient relevant improvements.
Methods: Comparative evaluation of morphological and functional biomarkers measured with sophisticated scientific set-ups and patient relevant parameters.

Results: IRDs are genetically and clinically extremely heterogeneous entities complicated further by their progressive nature. As rod and cone pathways may be affected with a different time course and at different degrees of severity, biomarkers measuring rod and cone distribution separately in their spatial distribution have been shown to be of particular value. Specific imaging methods visualize the morphology and specific metabolic features at high resolution, and allow correlation with functional parameters. Translation into patient relevant improvements by specific therapies can be difficult as some results may be statistically significant but not necessarily patient relevant. Novel functional tests such as sophisticated mobility tests with varying contrast and light levels have been shown to help close this gap. All methods have limitations as to application in infancy and early childhood.

Conclusion: Great advances have been made to show therapeutic effects that are relevant to patients. Future research should focus on the development of sensitive methods in very young children as some of the therapies are considered to be most effective at a very early age.

Conflict of interest disclosure: none

Session 10 – Understanding Treatment Effects from Natural History Studies

Natural history of the progression of atrophy secondary to Stargardt Disease (ProgStar) study

Hendrik Scholl
Institute Of Molecular And Clinical Ophthalmology Basel (IOB), Basel, Switzerland

Stargardt disease (STGD1; OMIM: 248200) is the most common juvenile macular dystrophy. It is inherited as an autosomal-recessive trait associated with mutations in the ABCA4 gene, and more than 1000 disease-associated variants have been reported. Clinically, STGD1 is characterized by fundus flecks in the retinal pigment epithelium (RPE) and by macular atrophic lesions. Visual acuity (VA) and central visual fields deteriorate progressively commonly leading to legal blindness in adulthood. Currently there is no approved treatment for STGD1. The international multicenter Progression of Atrophy Secondary to Stargardt Disease (ProgStar) study aimed to understand the natural history of disease progression to help determine appropriate outcome measures for future treatment trials. The primary aim was to assess the yearly rate of progression of STGD1 using the growth or the development of atrophic lesions as measured by fundus autofluorescence (FAF) imaging. Secondary aims include to assess the yearly rate of progression of STGD 1 using spectral-domain optical coherence tomography (SD-OCT) to measure the rates of retinal thinning and the loss of photoreceptors; to assess the yearly rate of loss of retinal sensitivity as measured by microperimetry (MP); to assess the yearly rate of VA changes; to correlate the presence and progression of morphological abnormalities in FAF and SD-OCT images with visual function as measured by MP and VA; to perform exploratory analysis of factors associated with STGD1, such as participants’ use of vitamin A supplementation and mutations in the ABCA4 gene.

It was found that the mean rate of best-corrected VA loss was clinically small at 0.55 letters on a standard ETDRS chart per year during two years. Patients were stratified by baseline VA and found to lose 1.3 letters/year (p=0.07) when starting with no visual impairment, to lose 1.9 letters/year (p<0.001) when starting with only mild visual impairment, to lose 0.6 letters/year (p=0.002) when starting with moderate visual impairment and to gain 0.7 letters/year (p=0.02) when starting with severe visual impairment at baseline. Eyes that showed abnormal FAF in the fovea deteriorated faster than eyes with normal FAF at baseline (p<0.001). The rate of VA change was not significantly associated with genotype group, age at baseline, sex, or race/ethnicity.

Conflict of interest disclosure: none

Long-term follow-up, phenotypic and genetic spectrum of patients with juvenile X-linked retinoschisis in the Netherlands

Leo Hahn
Wesseling, Nieneke L.1; Van Schooneveld, Mary J.1; Van Genderen, Maria M.2; Florijn, Ralph J.1; Ten Brink, Jacoline B.1; Van Den Born, L. Ingeborgh3; Meester-Smoor, Magda A.4; Ossewaarde-Van Norel, Jeanette5; Thiadens, Alberta A.4; Klaver, Caroline C.4; Hoyng, Carel B.5; Bergen, Arthur A.1; J.F. Boon, Camiel1
1. Amsterdam UMC, Amsterdam, Netherlands
2. Bartiméus, Zeist, Netherlands
3. The Rotterdam Eye Hospital And The Rotterdam Ophthalmic Institute, Rotterdam, Netherlands
4. Erasmus Medical Center, Rotterdam, Netherlands
5. University Medical Center Utrecht, Utrecht, Netherlands
6. Radboud University Medical Center, Nijmegen, Netherlands

Purpose: Detailed genetic and clinical characterization of juvenile X-linked retinoschisis (XLRS) is essential to provide an accurate prognosis, and to identify clinical endpoints and the optimal intervention window for (gene) therapy. The aim of this retrospective observational clinical study was to describe the genotypic and phenotypic spectrum, and long-term clinical course in a large cohort of XLRS patients in the Netherlands.

Methods: Medical records of 173 affected males from 60 presumably unrelated families were reviewed for medical history, symptoms, best-corrected visual acuity (BCVA), ophthalmoscopy, visual field, full-field electroretinography and retinal imaging (fundus photography, spectral-domain optical coherence tomography, fundus autofluorescence). A causative RS1 mutation was identified in all families.

Results: Twenty-seven different RS1 mutations were identified in families, including 22 missense (81%), 3 frameshift (11%), 1 splice site (4%) mutation and 1 exon deletion (4%). The c.214G>A mutation (33%) and a deletion of exon 3 (12%) were most frequently detected in families, representing 42% and 19% of the total patient cohort, respectively. The median age at last examination was 30 years (interquartile range [IQR] 16-46 years). Follow-up data were available for 120 patients, with a median follow-up time of 13 years (IQR 6-21 years). BCVA ranged from no light perception to 0.0 logarithm of the minimum angle of resolution (LogMAR) (median 0.68 LogMAR [IQR 0.40-0.91]). BCVA of the better eye was not significantly different at last follow-up visit when compared to the first visit (median 0.49 LogMAR [IQR 0.40-0.70] and 0.48 LogMAR [IQR 0.30-0.70] respectively, p=0.069). Older age correlated significantly with lower mean BCVA, as well as lower mean central macular thickness (Spearman’s r -0.47 p<0.001 and r -0.46 p=0.001 respectively). No significant difference was found in mean BCVA between patients with truncating and non-truncating RS1 variants (p=0.744).

Conclusion/Significance: A wide range of BCVA was seen in different age categories. Although there is a correlation between age and BCVA, BCVA remained relatively stable during follow-up in this study, suggesting a slow decline of BCVA. Our study demonstrates no genotype-phenotype correlation. Trends in BCVA suggest an optimal intervention window for therapy in the first decades of life.

Conflict of interest disclosure: none

The disease course of rhodopsin (RHO)-associated retinitis pigmentosa (RP): a follow-up study

Xuan-Thanh-An Nguyen1
Talib, Mays1; Van Cauwenberg, Caroline2; Van Schooneveld, Mary3; Wijnholds, Jan1; Schalij-Delfos, Nicoline1; Dagnelie, Gislin4; Van Genderen, Maria5; De Baere, Elfride2; Meester-Smoor, Magda6; Thiadens, Alberta6; Hoyng, Carel7; Klaver, Caroline6; Van Den Born, Ingeborgh8; Bergen, Arthur9; Leroy, Bart2; Boon, Camiel1
1. Leiden University Medical Center, Leiden, Netherlands
2. Ghent University Hospital, Ghent, Belgium
3. Amsterdam University Medical Center, Amsterdam, Netherlands
4. Wilmer Eye Institute, Baltimore, MD, USA
5. University Medical Center Utrecht, Utrecht, Netherlands
6. Erasmus University Medical Center, Rotterdam, Netherlands
7. Radboud University Medical Center, Nijmegen, Netherlands
8. Rotterdam Eye Hospital, Rotterdam, Netherlands
9. Amsterdam University Medical Center, Amsterdam, Netherlands

Purpose: In view of upcoming gene therapy trials, more insight into the natural disease progression in RHO-associated RP is required. A more detailed clinical disease profile will aid in the selection of eligible candidates, and the establishment of appropriate clinical endpoints for future gene therapy trials. In this retrospective cohort study, we provide a description of the clinical variability and the natural disease course in patients with RHO-associated RP in a longitudinal cohort.

Methods: We reviewed the medical records of patients with RHO-associated retinitis pigmentosa for symptoms, best-corrected visual acuity (BCVA), slit-lamp examination, fundus photography, Goldmann kinetic perimetry (V4e and I4e isopters), full-field electroretinography, spectral-domain optical coherence tomography and fundus autofluorescence imaging.
Results: One-hundred patients with RHO mutations were included for analysis. Mutations were located in the extracellular (n=64; 64%), transmembrane (n=20; 20%), or cytoplasmic (n=16; 16%) domains. The median age at which patients reached mild visual impairment (0.3 = BCVA < 0.5) was 72 years. First occurrences of low vision were observed from the 3rd decade onwards, whereas severe visual impairment and blindness were seen from the 6th decade onwards. Mutations in the cytoplasmic domain were associated with a worse age-adjusted BCVA (p=0.033), but did not alter the BCVA decline rate of 2% per year (p<0.001). Mutations in the extracellular domain were associated with a larger age-adjusted seeing retinal area (I4e) than those in the cytoplasmic (p=0.006) or transmembrane domains (p=0.006), but did not affect the yearly decline rate of the I4e seeing retinal area (5%; p<0.001). Visual function parameters correlated robustly with the thickness of the foveal photoreceptor-retinal pigment epithelium complex (p<0.001). Cystoid maculopathy (CME) was reported in 34/56 patients (61%).
Conclusions: RHO mutations result in a relatively mild form of retinitis pigmentosa. Mutations in the cytoplasmic domain are associated with worse visual function. The relatively late occurrence of blindness and slow rates of BCVA decline indicate a broad therapeutic window, but also warrants further exploration of alternative functional and structural parameters to assess treatment efficacy in upcoming clinical trials.

Conflict of interest disclosure: none

Further evaluation of a simple perimetric approach to the differential diagnosis between blue cone monochromacy (BCM) and achromatopsia (ACHM)

Alfonso Senatore1, 2
Skalak, Cindy1; Kheir, Wajiha1; Iannaccone, Alessandro1
1. Duke Eye Center, Durham, NC, USA
2. Department of Ophthalmology – “G. D’Annunzio” University, Chieti-Pescara, Italy

Purpose: Without the aid of sophisticated functional studies and molecular genetic testing, differentiating BCM and ACHM is challenging, especially in isolated male cases. Consistent with the fact that S-cones are spared in BCM, it has been shown that light-adapted chromatic (600 nm and 440 nm) automated perimetry is differentially affected in BCM (Luo et al. PLoS One 2015) and not in ACHM, but these techniques are available only in highly specialized Centers. Thus, we sought to develop a simpler perimetric approach to BCM/ACHM differential diagnosis.

Methods: Out of 25 molecularly confirmed cases, 10 subjects [4 ACHM (all CNGB3 gene deletions or mutations, 10-70 yo) and 6 BCM patients (3 OPN1LW/OPN1MW gene cluster or LCR deletions, and one with the p.C203R point mutation, 9-72 yo)] were tested. Automated perimetry was performed with a Humphrey 30-2 FASTPAC test, using fovea on, fluctuation and fixation/gaze control off options. Size-V stimuli were presented with dilated pupils, after at least 3 min of background adaptation. To probe mainly S-cones, standard SWAP was used (blue stimuli on a bright yellow background). To probe mainly L-cones, an identical test was performed on a standard white background with red stimuli, custom-selected from the “change parameters” menu.

Results: Each BCM subject consistently exhibited markedly elevated to non-detectable thresholds to red-on-white stimuli and a range of normal to mildly elevated thresholds to blueon-yellow stimuli, whereas each ACHM subject showed markedly elevated to non-detectable thresholds to both tests. Each pair of perimetries was completed within 45 to 50 minutes on both eyes, adaptation time included (about 10-12 min of active testing time per eye).

Conclusions: Our perimetric method allows for accurate differentiation between BCM and ACHM, and can be easily implemented in a routine eye care setting with minimal customization of standard testing routines. We plan to extend further our studies on this method in these congenital cone disorders to estimate test-retest variability, and investigate possible trends for changes in these chromatic thresholds over time, so as to determine if this simple testing method may also be a useful outcome measure in gene therapy trials.

Conflict of interest disclosure: none

Longitudinal natural history study in patients with Retinitis Pigmentosa in preparation for gene therapy clinical trials

Francesco Testa1
Di Iorio, Valentina1; Brunetti, Raffaella1; Banfi, Sandro2; Simonelli, Francesca1
1. Eye Clinic, Multidisciplinary Department Of Medical, Surgical And Dental Sciences, University Of Campania “luigi Vanvitelli”, Naples, Italy
2. Department Of Precision Medicine, University Of Campania “luigi Vanvitelli”, Naples, Italy

Purpose: To evaluate natural history of Retinitis pigmentosa caused by mutations in several disease genes.

Methods: we retrospectively evaluated 139 patients with clinical diagnosis of RP and mutations in the following genes: – USH2A (27 patients); RHO (35 patients); CRB1 (22 patients); RP1 (20 patients); EYS (13 patients); RPGR (9 patients); PDE6A (8 patients); (5 patients). Clinical investigation included best-corrected visual acuity (BCVA), Goldmann visual field (GVF), fundus photography, and electroretinography. Longitudinal analysis was performed over a median follow-up time of 3 years.

Results: At the study baseline, at a mean age of 33.7 ± 16.0 years, the patients showed a BCVA of 0.6 ± 0.8 logMAR in both eyes. They showed, on average, a GVF area with III4e target size of 2,069 ± 2,312°2 and of 2,138 ± 2,259°2, respectively in right and left eyes. Our patients reached legal blindness based on BCVA (i.e., >1.0 logMAR in the better eye) at a median age of 55.8 ± 1.4 years, RP1 patients reached it 13 years later whereas CRB1 18 years earlier. Our patients reached legal blindness based on GVF (i.e. III4e area <314°2) at a median age of 31.8 ± 1.0 years, PDE6A, EYS and USH2A patients reached it about 5 years later whereas CRB1 and RPGR about 7 years earlier.

Conclusions: The results of our natural history study confirmed the heterogeneity of clinical phenotype and natural history in RP patients according to the causative genes. For that reason, natural history data are strongly required in order to design appropriately gene therapy clinical trials.

Conflict of interest disclosure: The following conflict(s) of interest must be disclosed:
Francesco Testa and Francesca Simonelli received honoraria or consultation fees from Sanofi (FS & FT), Spark Therapeutics (FS), and Bayer (FT)

Session 11 – Gene and Cell based Therapies

Photoreceptor transplantation into the mammalian retina

Marius Ader
Center For Regenerative Therapies Dresden, TU Dresden, Dresden, Germany

Human vision depends on light sensing photoreceptors in the retina and their degeneration results in permanent vision impairment and blindness. In mammals, photoreceptors cannot regenerate from endogenous cell sources and, therefore, strategies are currently explored using pre-clinical animal models to regain visual function via photoreceptor transplantation. While photoreceptor replacement represents a promising approach in late-stage retinal degenerative diseases, we also observed the transfer of cytoplasmic material between donor and host photoreceptors, that might represent a potential new route for retinal therapy development. Furthermore, 3D retinal organoids derived from pluripotent stem cells have been established for the generation of high numbers of transplantable mouse and human photoreceptors. In my talk I will present our recent data about the generation of stem cell-derived human cone photoreceptors, sorting-systems for photoreceptor enrichment, and their transplantation into mouse models of retinal degeneration.

Conflict of interest disclosure: none

DNA repair mechanisms in photoreceptors

Knut Stieger
Department of Ophthalmology, Experimental Ophthalmology, Justus Liebig-University, Giessen, Germany

Purpose: The advent of genomic editing technologies has opened new possibilities to target genes at endogenous genomic loci. However, gene editing requires maximum precision of DNA repair to avoid de-novo mutations at the target site. Currently, there is insufficient knowledge of the DNA repair machinery in postmitotic neurons, specifically in rod and cone photoreceptors. Consequently, efficient therapeutic DNA repair has so far not been achieved. In order to shed further light on the status of DNA repair mechanisms in photoreceptors, we studied the gene expression profile of DNA repair proteins and characterized DNA repair mechanisms following double strand break (DSB) induction in photoreceptors in vivo.

Methods: Expression profile and localization of several DNA repair proteins, including Ku80, yH2AX, 53bp1, Lig 1, 3 and 4 were studied in vivo in the retinae of several different model systems. Additionally, we have developed a mouse model (RPGR-KI mouse containing an homing endonuclease I-SceI site) and studied DNA repair events following DSB induction via AAV mediated gene transfer in vivo.

Results: We have observed unexpected localization of DNA repair proteins in retinal neurons, most importantly concerning Ku80, a key protein in the non-homologous end-joining (NHEJ) pathway. In vivo induction of DSBs at the I-SceI site following AAV mediated gene transfer resulted in sequence changes at the target site in up to 20% of cases, most often representing indel formations following NHEJ.

Conclusion: Unexpected localization of DNA proteins in photoreceptors indicates the presence of non-canonical DNA repair pathway mechanisms, which may explain the reported DNA repair deficiency in murine rod photoreceptors. However, frequency and character of sequence changes after DSB repair in vivo indicate that efficient repair of DSBs by NHEJ is possible.

Conflict of interest disclosure: none

DNA repair after ISce-I mediated DSB in photoreceptors and RPE cells following AAV mediated gene transfer in vivo

Brigitte Müller1
Janise-Libawski, Annabella1; Götzmann, David1; Vössing, Cornelia2; Fühler, Bärbel1; Neidhardt, John2; Lorenz, Birgit3; Stieger, Knut1
1. Department of Ophthalmology, Experimental Ophthalmology, Justus Liebig-University, Giessen, Germany
2. Faculty Of Medicine And Health Sciences, Oldenburg, Germany
3. Department of Ophthalmology, Justus Liebig-University, Giessen, Germany

Purpose: Genome editing represents a potentially powerful tool to treat inherited retinal disorders. However, DNA sequence alterations during the repair process after DNA double strand break (DSB) induction are manifold and uncontrolled so far. This represents a serious safety issue in the development of therapeutic strategies. The purpose of this study was to characterize the DNA sequence alterations following a DSB repair on the X chromosome in photoreceptors and RPE cells after AAV mediated gene transfer in a mouse model for XLRP.

Methods: Two month old mice ((B6J.SV129-Rpgrtm1stie) were used. This mouse line contains an Isce-I target site downstream of the RPGR-ORF15 gene on the X chromosome. A total of 4 different AAV-vectors were applied. Fifty-six days after subretinal injection, eyes were harvested, retinae isolated, and the GFP-positive cell population enriched by FACS. PCR fragments were subjected to T7 surveyor assay and subsequently Sanger sequenced to detect DNA sequence changes at the target site.

Results: GFP expression was observed in all injected eyes. Following Sanger sequencing of PCR clones, we observed a high frequency of small deletions and single nucleotide substitutions (between 10 and 30% of all clones), most often located in the 5’ region of the target site. Small insertions were outnumbered by deletions and substitutions. Larger DNA sequence modifications were also found, but to a much lesser extent. In retinae injected with an all-in-one vector containing the template in addition to the nuclease, we detected replacement of the I-SceI site by the HindIII site with low frequency, indicating the presence of HDR at the target site in addition to more frequent NHEJ events.

Conclusion: We successfully induced in vivo genome editing in photoreceptors and RPE cells following AAV-mediated gene transfer. Both, NHEJ and HDR were detectable. These data represent the basis for further studies regarding the occurrence of DNA sequence changes at target sites in retinal neurons in vivo.

Conflict of interest disclosure: none

A Bioluminescence Resonance Energy Transfer based Sensor for the precise determination of non-homologous end joining DNA repair events

Tobias Wimmer
Bonthu, Dileep; Stumpf, Constanze; Klekamp, Philip; Moeschl, Vincent; Kapygen, Riina; Thiel, Christian; Lorenz, Birgit; Stieger, Knut
Department of Ophthalmology, Experimental Ophthalmology, Justus Liebig-University, Giessen, Germany

Purpose: Endonuclease based, therapeutic genome editing is a powerful strategy to treat inherited retinal degenerations. However, specificity and efficacy analysis are crucial in early development. Therefore, we generated a bioluminescence resonance energy transfer (BRET) based sensor for measuring the activity of endonucleases in inducing DNA double strand breaks (DSBs) and quantifying DNA repair through non-homologous end joining (NHEJ).

Methods: The NHEJ sensor, which is expressed as a single molecule, consists of a luciferase domain and a GFP2 domain with a shuttle-cloning box inserted for the integration of any given endonuclease target sequence. The luciferase as an energy donor is acting as an internal standard with which the disruption of the reading frame with premature stop codons of the green fluorescent energy acceptor by indel formation can be measured. This results in a change of the BRET ratio, which corresponds to the endonuclease and NHEJ activity. The sensor was tested with different endonucleases (I-SceI, spCas9) in different cell lines (HEK293-T, ARPE19, etc.) in vitro.

Results: The sensor showed an endonuclease induced reduction of the BRET ratio in all cell lines tested. The measured reading output is endonuclease concentration, but not NHEJ sensor concentration dependent. The sensor/endonuclease system shows outstanding specificity. Furthermore, the NHEJ sensor was validated with a FACS based reporter system successfully and applied to measure NHEJ pathway protein knockouts.

Conclusion: The NHEJ sensor is a useful tool to answer important questions regarding endonuclease activity, NHEJ activity in a variety of different possible cell lines in vitro before the translation to in vivo experiments. This would shorten the development of strategies to treat inherited, retinal diseases.

Conflict of interest disclosure: none

Giessen 2019 – Day 3 Photo Gallery