Reflections on “Hot” Blind Spots: Lessons from Research on Aging Macula Disorder and GlaucomaThe Weisenfeld Lecture

The Association for Research in Ophthalmology (ARO) was founded in 1928 by Conrad Berens, (Fig. 1) a New York ophthalmologist, to strengthen the scientific basis of ophthalmology. ¹ Initially, only clinicians with some scientific expertise attended the half-day meetings that were paired with the American Academy conventions. The fledgling ARO became more viable in the years after World War II which, despite its terrible consequences, spurred the development of visual sciences. In 1970, ARO changed to the Association for Research in Vision and Ophthalmology (ARVO), and thus ARVO seems to be, with 83 years, in the same age range in which the disorders that I am going to discuss have their highest incidence. 

Figure 1.

 

Conrad Berens, 1889–1963.

What would Conrad Berens think of this ARVO meeting? Would he be impressed by its 13 scientific sections and its nearly 7000 abstracts or puzzled by the overlap in posters on certain topics as observed from these abstracts? He might be pleased by the fact that 44% of the 6000 ARVO members are non-MDs or be disappointed because only 2.3% of all ophthalmologists world-wide are members of ARVO. Would he get the impression that for some, vision research seems a goal in itself, instead of a search for better cures for eye disorders? Would he consider the recent separation of the Visual Section from ARVO's main meeting a bright idea or the sequelae of blind spots in its board members? 

In this lecture, I will describe my transformation, perhaps somewhat similar to that of Berens', from pure clinician to clinical researcher in ophthalmology, a field that has undergone a striking metamorphosis over the past 40 years. I will deal more at length with aging macula disorder (AMD) and open-angle glaucoma (OAG). I will analyze what might bewilder Berens about our ophthalmology advancements and will conclude my reflections by mentioning some human interactions and scientific factors that may influence progress in vision research. 

My Start in Ophthalmology

When I began my residency at the Amsterdam University Hospital in 1971, general ophthalmology was concerned mainly with prescribing glasses, routinely checking eyes for cataract and glaucoma, and seeing patients with (acute) complaints. There were four subspecialties: orthoptics and difficult squint cases, retinal detachment surgery, corneal surgery, and neuro-ophthalmology. The university clinic had no specialized clinics, and residents might see in 1 day patients with dry eyes, foreign bodies, diabetic retinopathy, retinal detachments, AMD, glaucoma, intraocular tumors, Graves's orbitopathy, or hereditary disorders like retinitis pigmentosa or aniridia. Residents learned cataract, glaucoma, and squint surgery, and junior staff did a fellowship in retinal detachment or corneal surgery. To prepare for a locum in Pakistan, where I had to operate alone on 25 to 40 cataract patients daily, I had to master the fast but difficult left and right-handed incision with a Von Graefe knife, so as to open with one cut one third of the corneal limbus. Routinely, residents were taught a beveled incision for intracapsular extractions; later came extracapsular surgery without and with multiple types of implant lenses and eventually phacosurgery. Operations for OAG changed from iridencleisis to trabeculectomies, first without and later with mitomycin or filtering devices. Retinal detachment surgery moved from various kinds of intrascleral buckles to external sponges and tires, later followed by silicon oil use. I was among the pioneers in The Netherlands performing trans pars plana vitrectomy, operating with equipment that was very primitive by current standards and had no internal light source. ² A study tour of six vitrectomy centers in the United States brought me in contact with Steve Charles, Don Gass, Bob Machemer, Ron Michaels, Gholam Peyman, and Steve Ryan. We performed retinal light coagulation with the Xenon “Jumbo” coagulator from Essen, Germany, and this was soon followed by all kinds of laser sources. While seeing 30 to 40 retinal patients with two ophthalmologists in the course of one afternoon around 1980, we also injected 10 to 15 patients with fluorescein and read their angiograms afterward. How stimulating the discussions were about these in the International Fluorescein Angiography Club! During my residency and early staff years, there was hardly any structural research, and the focus was on tertiary patient care. Visual requirements for all kinds of jobs and professions stimulated me finally, around 1990, to enter the field of ergophthalmology. ³ The dictum: “Chaotic action is preferable to orderly inaction” ⁴ is applicable to my activities during these years. 

Most innovations in ophthalmic practice since those years have been technology driven. Gradual improvements occurred in surgical microscopes, the refinement of instruments and suturing materials, laser equipment for retinal coagulation and refractive surgery, autorefraction, ionizing irradiation for malignancies, imaging with scanning laser ophthalmoscopes and optical coherent tomography, and, more recently, intraocular pharmacotherapy and even gene therapy. From these technological innovations sprouted a large number of subspecialties in ophthalmology, from the original 4 to more than 25. Is there a similarity here with Marxism, which asserts that the means of production induce their own organizational framework? I will consider whether this subspecialization is a blessing after elaborating on how I entered research. 

Tumbling into Research

At medical school I had to wait twice for 9 months because there were no internships available. There was a vacancy in an ear, nose, and throat (ENT) research laboratory, and after I had worked there for 1.5 years, the ENT chairman suggested that I do a PhD on the neurophysiology of cervical proprioceptive nystagmus. I spent my spare time in the last period of medical school and during military service in the ENT laboratory. I concluded, however, that ophthalmology seemed more interesting than ENT. I remember at that time regretting not having chosen a topic in ophthalmology for my thesis, ⁵ which would have been an advantage in my new specialty. I saw only three patients with nystagmus after a cervical injury in my life! Approximately 10 years later, I realized, however, that overcoming the hardships of finishing a neurophysiological thesis also contributed to my work in ophthalmology. It was not just that, as thesis supervisor, I remembered the ups and downs that most PhD students encounter during their experiments or writing sessions. It also broadened my area of interest in ophthalmology and facilitated communicating with researchers in neurophysiology and other areas. 

In 1973, research was not a high priority in most academic clinics. Therefore my ophthalmology chairman Robert Crone was glad to have a resident with a PhD and suggested that I start research on botulinum toxin for patients with squint who had repeatedly been operated on. It so happens that the blood supply to the anterior eye segment runs through the extraocular muscles and too many squint operations might jeopardize the eye circulation and result in blindness. It took me 4 years, concurrent with my ophthalmology training, before I injected botulinum toxin, in about 1977, in squint patients and in cases of oscillopsia due to acquired nystagmus. ⁶ First, this was an exciting time. Although it was not a stipulation of the institutional review committee, I routinely confirmed the biological activity of every vial of botulinum toxin from an already calibrated batch, by administering a lethal dose (LD₅₀) test to mice after I injected the patient. The LD₅₀ could not be performed before the patient was injected, because the diluted toxin would have lost its effectiveness by the time test results were available. This, literary “secondary prevention” was intended to warn the intensive care doctor that a problem might arise. Fortunately, we never had a serious adverse reaction. Independently, Alan Scott from the Smith Kettlewell Institute in San Francisco was working on botulinum and around 1980, he was licensed by the U.S. Food and Drug Administration to use botulinum toxin (Scott A, personal communication, 2010). He published his classic papers on botulinum, ^7,8 and we nearly failed to do so, ⁶ because of an unexpectedly small number of squint patients. Most of them were, like their orthoptists, not eager to have these injections of a “potentially deadly” poison. Our toxin was, however, very useful for many years for the ENT specialists and neurologists who had no other free-of-charge cure for hemifacial spasm ⁹ or essential blepharospasm. ¹⁰  

During my ophthalmology residency and fellowship in retinal surgery with an on average daily workload of 10 clinic hours, I hardly published at all. Basic, nonclinical researchers are rarely aware of the theoretical and skill requirements of clinical training and sometimes are disdainful of nonpublishing ophthalmologists. I did, however, enjoy working on genetic disorders with a very stimulating personality, Willem Delleman, our ophthalmogeneticist. After my thesis and the botulinum adventure, it was a welcome and rewarding experience to sort out a problem and document it on paper, introducing some variety into the inevitably routine nature of clinical tasks. ^{11 –15}  

The 1973 oil crisis and 1980 economic crisis resulted in severe budgetary cuts, which hit universities in The Netherlands. The university and hospital boards became more aware of their primary tasks: teaching, research, and tertiary care medicine. Structured PhD thesis programs became better administered, and, by the time I became chairman in Rotterdam in 1982, clinical research by the happy few was swapped for structural research programs in medical schools. We soon started AMD research on postmortem eyes with the department of pathology. ¹⁶ Because of the uncertainties in the treatment of ocular melanoma patients in this referral center for South-Western Holland, we set up a research line for ocular melanomas that is still operating today. ^{17 –19} Finally, we joined The Rotterdam Study with longitudinal epidemiologic research on cardiovascular, neurologic, and endocrine disorders and focused on AMD ²⁰ and OAG. ²¹  

Changing Conditions for AMD Research

Contrary to my casual start in nystagmus and botulinum research, launching AMD research was motivated by a strong feeling of futility and a sense that the system had failed patients who had lost their sight to this disorder. They often became depressed on hearing their diagnosis of aging macula disorder (exactly the same as age-related macular degeneration, and also abbreviated AMD). Over the past years I often met nonargued resistance from reviewers and editors against my efforts to improve the name AMD. ^{22 –24} The name age-related macular degeneration was coined in 1969 ²⁵ and in the United States in 1985 ²⁶ as a substitute for senile macular degeneration. It is remarkable that Junius and Kuhnt ²⁷ unintentionally sowed the seeds for this controversy in their classic (1926) monograph on AMD by using half the time the word “disorder” and half the time “degeneration” to describe AMD. In the epilogue Junius wrote (Kuhnt had died 4 months earlier): “Die Bezeichnung ‘Scheibenförmige Erkrankung der Netzhautmitte’ (Degeneratio maculae luteae disciformis) für die von uns auf Grund neuer Beobachtungen geschilderte Krankheitsgruppe mag etwas farblos erscheinen, hat aber auf Grund der gegebenen Erläuterungen so lange Sinn und Berechtigung für die Charakterisierung dieses scharf abgrenzbaren, mit keinem anderen Leiden der Netzhaut zu identifizierenden und daher selbständigen Krankheitsbildes, bis ein treffender Name dafür gefunden ist. ” (“The description ‘disciform disorder of the retinal center' for the disease entity that we described based on new observations, may seem somewhat colorless but is useful and justified based on our examples to characterize this well defined clinical picture that cannot be mistaken for other retinal disorders, as long as not a more striking name has been found.) I prefer to call AMD aging macula disorder instead of age-related macular degeneration, for the following reasons: “Age-related” does not differentiate between the existing juvenile, adolescent or adult-onset maculopathies, whereas “aging” implies a development occurring later in life. A few whitish deposits in the retina, so-called drusen, are considered to be hallmarks of early AMD, but should we call these drusen a “degeneration” in the absence of visual defects? Until today, the cause of drusen or late AMD remains unknown. ²² Degeneration is a vague word with several, but all negative, meanings. It is unclear whether malfunctioning retinal pigment epithelium (RPE) or a hemorrhage from a subretinal neovascular membrane can be defined as a degeneration according to valid definitions in pathology. In their text, Junius and Kuhnt wrongly translated “Erkrankung” in the Latin “Degeneratio” instead of in “Morbus. ” Is it because they wanted to get rid of the “colorless disorder” that they swapped “Erkrankung” for “Entartung” in the title of their monograph? Is it necessary to mention to a patient with early AMD with a few drusen and no visual complaints for years to come that there is a degenerative disease? Good doctors should try not to alarm their patients when there is no absolute necessity and in particular when there is no cure for a disorder like dry late AMD. Is there pressure from pharmaceutical companies to make these patients anxious from early on about a degeneration instead of a disorder to find more targets for their nutritional supplements or other therapies? The notion of having a degenerative disease leading to blindness may be harmful to a patient, as the following case shows. A 55-year-old woman minister had metamorphopsia, seeing straight lines as if they were curved, in one eye. Her optician and later her ophthalmologist told her that she had AMD. Soon she became severely depressed with uncontrollable crying spells when conducting wedding or burial ceremonies. Convinced that she would become blind soon, she resigned her ministry. Luckily, there was no need for this. After ophthalmoscopy and biomicroscopy I rejected the diagnosis of AMD because of some slight preretinal fibrosis in one eye with a favorable prognosis and no drusen in either eye. She jumped up and much to my bewilderment (and that of her husband) said: “Now I am going to embrace you!” She suited the action to the word, and later resumed her ministry. 

Around 1985, returning to AMD research, I started studying the AMD literature searching for the origin of drusen. Wedl, an Austrian pathologist, seems to have been the first to describe drusen ²⁸ in 1854 (Fig. 2), and soon thereafter, Donders ²⁹ (Fig. 3) from Utrecht wondered if they originated from the nuclei of the RPE cells! Speculations about the origin of drusen flourished, and approximately 20 years later, Rudnew ³⁰ thought that drusen originated from leukocytes. Is it not surprising that we still do not know, 150 years later why drusen start and where they originate from? In about 1980, Gass ³¹ and Sarks ^32,33 combined clinical and pathological observations in AMD cases. In conjunction with the department of pathology, we attempted to elaborate on their findings and to discover the constituents of basal laminar deposits, an earlier sign of AMD than drusen, in postmortem eyes. ¹⁶ We would need the biochemistry department to advise about the chemical composition of drusen in a more sophisticated way than the earlier methods of immersion of histology slides in acids or lyes. ²⁹ However, the biochemist had no microanalytical equipment and needed at least 200 to 300 mg of drusen material, which I could not provide. In our intensive search for the drusen constituents, we could hardly determine what they were. We did find in AMD eyes glycosaminoglycans in basal laminar deposits between the cell membrane and the basement membrane of the RPE ³⁴ and increased expression of vascular endothelial growth factor in the RPE. ³⁵ Inhibition of this growth factor now plays a crucial role in the therapy of late, wet AMD. I was excited by crucial publications on drusen composition that appeared around 2000. It became clear that drusen are mainly formed by proteins such as immunoglobulins, complement components and regulators, acute phase proteins, serum and intracellular proteins, and some lipids. ^{36 –39} A second breakthrough for AMD patients and researchers was the discovery of the CFH gene for AMD ^{40 –42} in 2005, resulting in high expectations for future therapies. ⁴³ It has become increasingly clear that AMD is a multifactorial disorder in which genes, ⁴⁴ lifestyle (smoking and diet), ⁴⁵ and environmental factors all may play a part. ⁴⁶  

Figure 2.

 

First drawings by Wedl of drusen, called hyalin masses and fat globules.

Figure 3.

 

Franciscus Cornelius Donders, 1818–1869.

Advances in Open-Angle Glaucoma Research

In the field of OAG, our second focus of interest, it was, once again, 19th century scientists who contributed seminal concepts. In 1861, Haffmans, a PhD student of Donders, coined the concept of simple glaucoma, similar to what we now call OAG, as a separate entity, quite distinct from inflammatory glaucoma. The latter we now would call acute angle-closure glaucoma. ⁴⁷ This simple glaucoma had the following features: elevated intraocular pressure (IOP), cupping of the optic disc, and visual field loss. ⁴⁷ We still use more or less the same definition. Donders ⁴⁸ invented a tonometer in 1863 to measure the IOP. Von Graefe identified at that time several multigenerational families in Berlin with glaucoma, showing earlier onset of the disease in younger generations. ⁴⁹ Many hypotheses on OAG were formulated, including vascular insufficiency, ^50,51 autonomous nerve dysfunction, mechanical forces due to rubbing of the optic nerve axons along the cribriform plate or bulging of the cribriform plate, oxidative stress in optic nerve axons, ⁵² and subnormal cerebrospinal fluid pressure. ⁵³ Apoptosis of one or more of the 25 different types of retinal ganglion cells ⁵⁴ was considered; deficient neuroprotection, for which cure more than 500 compounds have been tested in vain ⁵⁵ ; and trabecular meshwork changes as an explanation for elevated IOP. ⁵⁶ Why an elevated IOP leads to retinal neuronal loss is still unanswered in OAG and even more so in the so-called normotensive OAG. Fortunately, there has been recent support for the 140-year-old hypothesis that lowering the IOP slows glaucomatous visual field loss. ⁵⁷ In fact, lowering the IOP is still the only therapy of proven effectiveness. However, despite substantial improvements in IOP-lowering therapies, after 5 to 10 years 50% of the OAG patients show progression of visual field loss. ^58,59 Extensive international research failed to find major genes responsible for OAG, ^60,61 despite finding genes involved in early-onset monogenic glaucoma and some secondary late glaucomas. ⁶² Re-examining DNA from well-documented families with next-generation sequencing techniques and with technologies enabling detection of translational defects, errors in epigenetic signaling, and regulation by microRNAs is an approach that may become helpful in finding genes for OAG. There is certainly progress in our AMD and OAG research, but why is it so maddeningly slow in crucial aspects at a time when thousands of manuscripts in ophthalmology are published and several new journals start annually? 

How Should “Blind Spots” Be Handled in Research?

Blind spots in our understanding need as close a scrutiny as that used in delaying the progression of blind spots or scotomas that are caused in our patients by AMD and OAG. Factors limiting progress must be addressed. Numerous factors may delay research, and so I will mention just a few. We could divide them into factors related to human interactions (Table 1) and structural factors intrinsic to fundamental complexities in nature. 

Table 1.

 

Human Causes of Slackening Research and Possible Cures

Causes	Cures
Insufficient cooperation between researchers and research groups	Less emphasis on bibliometry for career purposes.
	No limits on number of co-authors in journals.
	Better rewarding of grants for (inter) national cooperation between top groups.
	Less secrecy about new findings, obtained with public funding, prior to publication.
	Discouragement of “me too” and/or double publications.
	Fair and better balanced remuneration for basic and clinical research.
“Fashionable” instead of innovative grant system	Accepting that grant outcomes do not always meet expectations.
	Stop considering negative study results as implicit failures.
	A less excessive preliminary data requirement to avoid killing of innovation.
	Harder study of “what interests you the most in the most undisciplined, irreverent and original manner possible.“ 63
Isolation by overspecialization	Creation of multidisciplinary teams focusing on specific topics.
	Less jargon or protectionism of one's own “specialty”
The Matthew Effect*	Investment in younger, as yet not well known, researchers 64 –66

*  The Matthew Effect in scientific sociology ^{64 –66} is derived from the following words in Matthew's Gospel: “For to all those who have, more will be given, and they will have an abundance; but from those who have nothing, even what they have will be taken away.“ ⁶⁴

That better cooperation may increase numbers, and thus statistical power can be demonstrated by the four new loci found in 8309 participants with late-onset Alzheimer's disease from nine cohorts. ⁶⁷ There are, of course, differences between the various research fields. A theoretical mathematician told me that they would become suspicious about the input of each co-author in a paper with three or four authors. The opposite can be found in the Physical Review Letters where the number of co-authors may vary from three to 2000! ⁶⁸ Instead of promoting multidisciplinary teams to speed up research, absolutely essential for solving problems such as the translation from genes to protein or disease mechanisms, some journal editors are restricting the number of co-authors to six or eight! 

Reasons of job security may induce us to walk along well-trodden paths. The 2010 physics Nobel Prize winner Andre Geim, stressed the importance of having a wide variety of interests and going off the track. He stated, “There is a research style where your professor assigns a project to you and you will remain on that track for the remainder of your life, from your scientific cradle to your scientific tomb. Boorring. There is a very small chance that you find anything new close to that track. I like to nose around. Knowledge and equipment are like Lego blocks. The more you gather, the more possibilities you get to build something new.” His last advice was, “Work hard. The more hours you make, the higher your chance of a stroke of luck.” ⁶⁹  

In a review system where the author or grant applicant is known to the (unknown) reviewer, well-known senior researchers may obtain funds to the disadvantage of (sometimes smarter) junior scientists who may be more open to innovation—the Matthew Effect (see Table 1). My first publication was a typical example. Who would accept a manuscript from an unknown person? After several rejections, we submitted to a new journal and our work appeared in volume 1 of a now respected neurologic journal. ⁷⁰  

Differences in the pathophysiology of disorders and in the complexity of the aging process may reveal structural dissimilarities between various illnesses. Diseases may greatly differ in their accessibility by novel research methods. AMD and OAG give striking evidence of such extremes. Recent genome-wide association studies estimated that 60% of the genes responsible for AMD but only 5% of those associated with OAG have been identified. Despite efforts by excellent geneticists on extensive families with OAG, ⁶⁰ I have to come to the embarrassing conclusion that I, as an ophthalmologist with more than 40 years in practice, hardly know what OAG is. Is OAG the result of interactions between many frequently occurring genes, each of them having its own small effect? A review on genomewide association scans pointed out that complex diseases have risk alleles with small effects that often remain undetected even in thousands of cases, because of insufficient statistical power. ⁷¹ Or is what we now call OAG, essentially a collection of 10 to 20 different disorders and is that the reason that we do not find genes for this disease for which familial occurrence is well documented? ^49,72,73 Is OAG in this respect similar to amyotrophic lateral sclerosis (ALS), until 20 years ago considered to be one neurologic disease, and after genetic testing found to be an assembly of 13 genes and loci? ⁷⁴ The “painfully slow progress” in unraveling the origin of ALS, also first described 150 years ago, has picked up momentum since 2009. ⁷⁴ I do hope that progress in understanding OAG will do the same soon. Another common feature between these two disorders is that the OPTN gene is involved in both ALS and OAG. ⁷⁴  

Up to now, I have discussed interactions in the clinic and between basic and clinical research. Have we paid enough attention to the setting in which we tried to do our best? In trying to optimize our ophthalmology research system we have delved deeper and subspecialized more, resulting in groups with experts in, for example, drusen composition. Within the context of a system that wants to solve, after 150 years, the origin of drusen, this is great, but medical research has a wider goal. This has even elicited from “real scientists” the remark that medicine is not a real science, because by definition it is applied to curing people. When a patient after searching the internet asks me, “Doctor, how does this drusen stuff help me?” honesty requires me to be modest about its “benefits” to him. Several times, I promised funding agencies in sincere ignorance that once a specific question was solved, it would be a “big” step forward. But apart from the botulinum toxin, ⁶ reduced melanoma-related mortality, ⁷⁵ and dietary secondary prevention for patients with early AMD, ⁴⁵ I do not seem to have contributed much to curative research over the past 40 years. 

Might it be that our explosive subspecialization has reduced our sensitivity to the needs of our patients? To accelerate overall clinical performance, we need to focus on the problems of the patient who is attached to the eye we want to treat, and be aware of the impact of our high-tech interventions on his vision and health. Some ophthalmologists in high-throughput intervention clinics seem to have swapped part of their Hippocratic Oath, stating “primarily do not do your patient any harm,” for the “hypocritical oath,” “how to make as much money as possible, ” regardless of real benefit to the patient. ⁷⁶ By the way, the Hippocratic Oath demonstrates that, for ages, some doctors have had difficulty in choosing between the best for their patients and themselves. Are we sufficiently aware that a system that seems to function in a limited way, may prove to be at fault in a larger context? The “simplest” example is the successful cataract surgery on a high-risk patient who dies 3 days later of pulmonary embolism because of medical neglect. When a top oncology clinic spends 40,000$/month on a new “cure” for a patient who wants to leave the hospital because there is no support for her anxieties at night, is that system working well? In a larger context, our politicians need our help in deciding about health care for an aging population. Perhaps, basic and clinical research should shift its focus to the effectiveness of primary or secondary prevention for common needs versus high expenditure for clinical or pharmacologic interventions, resulting in minimal extension of life or enhancement of its quality. We need to realize that our clinical and research systems may have to change in the coming decades. 

What Are Our Greatest Challenges?

Even though there are presently thousands of serious, eager, and capable researchers, Conrad Berens might be disappointed by our progress in understanding the causes of AMD, known for 137 years, ^77,78 and for the even longer known OAG. ⁴⁷ Might this limited progress be due to the monocausal, linear way of thinking, inculcated from high school on? The riddles we have to solve appear to be more and more complex than we anticipated. From youth on, we solved problems by demonstrating relations between two variables. It is like an algebraic equation with two unknowns. You find one and thus the other one becomes clear. But a complex system is like an algebraic equation with three or more mutually dependent unknowns. We need “polyfactorial logic” to solve this. We become more and more aware that many biological processes and disorders are actually complex disorders that are the result of the integrated action of many different and interdependent actors. In complex systems, all factors are actors in every circumstance, and all factors are related. In most of our experiments, we try to circumvent the complexity issue by creating linear conditions within a specified context that we think will prove associations or even monocausality. This monocausality is, however, strictly dependent on our scientific paradigm(s). It is customary to remove possible confounders identified by univariate analysis from a complex model because of insufficient statistical power, before using multivariable logistic regression. Adjustment for confounders might also remove valuable information from the model and jeopardize the conclusions. Do you recognize this when in English a confounder is also called a “third variable” or when we read “a well-planned experimental study design eliminates the worst confounding factors”? In many natural systems the dependencies of the parameters are not simply linear and there are more than two parameters. That is, it is a multiparameter nonlinear system. Often one uses statistical methods to find an approximating linear system. This process may result in the elimination of valuable information, and as a result the approximation is not good. Often the original system is such that no linear approximation is good. “Linearity kills complexity.” ⁷⁹  

Let me give you some examples of complexity. We can consider a long-term relationship between a man and a woman to be a linear one. Even given the complexity of human nature, this can result in a stable relationship that adapts in a sensible and often predictable way to personal, intrarelational, and environmental changes. However, when this bilateral relationship develops into a triangular one, the relationship will become much more unstable and unpredictable. ⁷⁹ A clinical example of linear thinking is the single case report describing a woman whose diabetic retinopathy regressed after a severe hemorrhage during childbirth resulted in a malfunctioning pituitary gland. ^80,81 This report resulted in a wave of hypophysectomies for proliferative diabetic retinopathy, often with poor outcomes and high mortality, apart from those due to the diabetes. ⁸² Another example is the concept of one gene → one disease that made us believe that once the gene is found, a cure for the disease is nearby. In practice, one starts looking for coding mechanisms for structural proteins or identifiable regulatory processes. However, this leaves out thousands of genes and epigenetic factors involved in regulation of transcription, RNA splicing, regulation of chromatin structure, and so on. It took 20 years after the cystic fibrosis gene was identified for therapies to start to emerge. Five years after the CFH findings, ^{40 –42} we are not sure yet about its mechanisms. Recently the DICER1 gene association with dry late AMD was identified, and it was proposed that the disease was due to insufficient breakdown of long, double-stranded Alum RNA by Dicer. ⁸³ It is unknown why Dicer levels are reduced in the RPE and whether other tissue-specific transcription factors are missing as well. ⁸⁴ The molecular basis of Alum RNA detoxification by antisense oligonucleotide injections has to be solved before clinical therapy can be contemplated. ⁸⁴  

At present, we only can solve the intricacies of a complex system when we include all continuously changing cofactors to find a diagnosis or even a cure. When we do not take all these into account, we may bog down. Our biggest challenge is that we have not yet developed a sufficiently complex mode of thinking to solve the polyfactorial problems that we are confronted with. And if the context changes too and thereby becomes a variable or actor in itself, the predictions will become even less accurate. Mathematicians and physicists are starting to think about how to create valid logic to solve complex problems. Sufficient computer storage capacity is available, but the power to calculate the outcomes of these multifactorial problems seems lacking. Perhaps future quantum computers with serial as well as parallel calculation capacities will do this. Meanwhile, I will have to live with the “blind spots” originating from my cerebral cortex with regard to the pathophysiology of AMD and OAG until new thought-provoking, complex research models are developed. 

I have tried to convey my joys and concerns about ophthalmic research to you. Our world probably is as complex as it was in 1500 when Erasmus lived in Rotterdam or in 1850 when Donders worked in Utrecht. This brilliant MD with wide scientific interests proved at age 26 that the principle of “conservation of energy” applies to life processes. One year later, 12 years before Darwin published The Origin of Species Donders contested that species were created each by a separate and independent act. He claimed that all forms of life were molded in their separate forms by the continuous operation of natural laws acting throughout the ages. He became professor of forensic medicine, anthropology, general biology, and, coincidentally, of ophthalmology. Fortunately, Donders needed money to provide a living for his wife and child. As a biologist and physiologist, he could not sufficiently support them, and thus he took on translating Ruete's German Textbook of Ophthalmology. Just by doing so, he became enthralled with the beautiful field of ophthalmology. A piece of unsolicited advice for young scientists might be: Try to follow your heart when you are at major crossroads in your scientific life. My thesis supervisor was the ENT chairman who was disappointed when I told him that I had chosen ophthalmology above ENT. At that time I could only vaguely explain this preference by quoting The Book of Genesis: “God saw all that He had made, and behold, it was very good.” Apparently smelling or hearing were not involved! Try to be flexible and do not restrict yourself to one scientific toy that may be outdated before you know it. See for example the impact of digitizing on the photography and film industry! A former 10,000$ analog Hasselblad camera is now available for 200$! Be open to cooperation with good multidisciplinary teams and focus on disciplines needed for reaching a certain goal. Let us go for quality and not quantity of papers in order not to pollute our scientific environment, and do not shun papers with say 100 coauthors, provided they all contributed their part. Even a so-called “nonscientific” clinician who brings in well-documented cases in a study has earned a place as a contributor to that paper. Try as a clinician to combine clinical work with research. This intellectual cross-fertilization is very stimulating. When a breakthrough occurs in solving complexity model calculations, an even brighter 21st century lies before us than even now, where it is so stimulating to see that ophthalmology is in the forefront of genetics. ^{85 –87}  

It remains for me to thank the persons who nominated me for the Weisenfeld award, the selection committee members for their decision, and the ARVO Board for following their advice. I want to share this award with my past and present collaborators, from whom I learned so much. Many thanks to Astrid Fletcher who with Paul Mitchell embellished my life, while introducing me, in a nearly unrecognizable way and to the audience for their patience. Above all, thanks to my wife and children who were the greatest variables, if you like, the “confounders” in my life and who allowed me to cultivate so many “hot” blind spots. 

Footnotes

 Disclosure: P.T.V.M. de Jong, None

I would like to warmly thank Nic M. H. van Dijk, PhD; Marij J. M. de Jong-Koomen, MSc; Martinus F. Niermeijer, MD, PhD; Jos M. Koomen, PhD; Astrid E. Fletcher, PhD; Philip G. Breen, BA, HDE; Dick van Norren, PhD; Bouke C. H. de Jong, MD, PhD; Ype P. de Jong, MD, PhD; Paul M. B.Vitanyi, PhD; and Jos E. M. A. Tielens, MA, for their valuable discussions and critical reviews of the manuscript.