# 13 - 505 Novel Approaches to Diseases of Unknown Etiology ## 505 Novel Approaches to Diseases of Unknown Etiology Levine B, Kroemer G: Biological functions of autophagy genes: A disease perspective. Cell 176:11, 2019. Mallucci GR et al: Developing therapies for neurodegenerative disor­ ders: Insights from protein aggregation and cellular stress responses. Annu Rev Cell Dev Biol 36:165, 2020. Song J et al: Quality control of the mitochondrial proteome. Nat Rev Mol Cell Biol 22:54, 2021. David R. Adams, Camilo Toro, Joseph Loscalzo Novel Approaches to Diseases of Unknown Etiology THE UNDIAGNOSED DISEASE STATE The term disease, etymologically meaning “lack of ease” or the presence of discomfort, is defined as an abnormal state that negatively affects the structure or function of all or part of an organism and that is not due to any immediate external injury. When referring to a person experienc­ ing a disease, the word patient is used in its original, meaning “the one who endures suffering.” These terms are well suited to patients with undiagnosed diseases. A patient with an undiagnosed disease is one for whom a medical diagnosis is not discerned after reasonable efforts utilizing established methods and procedures. Multiple factors may contribute to a failure to reach a diagnosis (Table 505-1). Patients who are affected by an undiagnosed disease for a protracted period exist in an undiagnosed state, which presents characteristic challenges for the patients, their families, and their medical providers. ■ ■THE MEANING AND CONTEXT OF A DIAGNOSIS A diagnosis often entails hierarchical levels of information specific­ ity with varying levels of relevance to the users (consumers) of such information (e.g.) patients and their families, health care providers, government health care agencies, insurers, epidemiologists, genetic counselors, pharmacologists, biologists). As an example, a diagnosis of Parkinson’s disease in an adult is based on the progressive emer­ gence of signs and symptoms of bradykinesia, rigidity, asymmetric rest tremor, and postural instability (clinical diagnosis), which are typically responsive to the administration of L-dopa (a therapeutic response biomarker). Together, these are cardinal features of striatonigral degen­ eration (a mechanistic diagnosis), a process associated with neuronal α-synuclein deposition and Lewy body pathology (histopathologic diagnosis) often based on a genetic susceptibility conferred by muta­ tions in genes such as synuclein (SYNCA, a molecular diagnosis) and likely influenced by environmental exposures (e.g., manganese or other neurotoxins). With ongoing advances in medical science and technology, the standard for what constitutes a reasonable diagnosis continues to evolve toward higher levels of specificity. For example, the utility of an antisense oligonucleotide therapy may be restricted to a specific subset of mutations associated with a given, monogenic, heritable disease. Efforts to adopt the principles of precision medicine include a growing emphasis on the context of disease within the genomic landscape, envi­ ronment, social factors, medical history, nutrition, and the microbiome of any given individual. Examples include cancer susceptibility, geneti­ cally determined idiosyncratic reactions to medications, and unique pathogen susceptibilities in patients with certain immune deficiencies. ■ ■UNDIAGNOSED RARE DISEASES Most chronically undiagnosed diseases are rare. While individual rare diseases have a low prevalence, they are numerous in aggregate. It is TABLE 505-1  Factors Contributing to the Presence of an Undiagnosed Disease FACTOR EXAMPLE Misleading information False-negative and false-positive test results Rare disorder Many inherited disorders have only been identified in a few individuals. For example, sialuria, a well-understood disorder of sialic acid metabolism, has been reported in 10 individuals (OMIM 269921). Unusual causes of common diseases, including atypical course of illness Insulin-dependent diabetes mellitus may be the presenting feature for the relatively rare autoimmune polyendocrinopathy syndrome, type I (OMIM 240300). Presence of multiple disorders (blended phenotypes) For an example, see PubMed ID 24863970. Lack of characteristic symptoms of known disease Diseases are commonly ascertained via cardinal signs or symptoms leading to incomplete ascertainment of all possible disease presentations. For instance, not all persons with Marfan’s syndrome are tall relative to other family members. For progressive diseases, pathognomic signs and symptoms may be missing in early stages of disease. New disease No prior knowledge or record of such disease Incorrect affected status assignments in family history A heritable disorder may be inappropriately excluded if family history information is incorrect. Primary disease manifestations obscured by other factors Maladaptive behavior, medication effects, and secondary disease manifestations may obscure signs and symptoms of a primary disorder. CHAPTER 505 Disease not expected in region or population Cystic fibrosis in persons of African ancestry, sickle cell disease in persons of northern European ancestry; infectious agents with marked geographical incidence patterns Diseases thought to be eradicated Poliomyelitis Novel Approaches to Diseases of Unknown Etiology Diseases occurring in unexpected time of life Parkinson’s disease in children, lysosomal storage disease in adults Malingering Feigned disease features intended to achieve secondary gain (Munchausen syndrome) Rare disease mechanisms Transmitted or sporadic prion disease, unusual zoonotic diseases Abbreviation: OMIM, Online Mendelian Inheritance in Man. estimated that >6000 rare diseases affect millions of people throughout the world. Estimates of aggregate population prevalence range from 6 to 10%. Many rare diseases have a genetic basis and onset in child­ hood. As the cloud of uncertainty inherent in the undiagnosed disease state is removed, new disease-specific counseling, therapies, resources, community engagement, and advocacy opportunities become possible. ■ ■THE EFFECT OF THE UNDIAGNOSED DISEASE STATE ON THE PATIENT Patients with an undiagnosed disease are frequently driven to under­ stand the basic nature of their ailment (what, when, where, how, etc.). Individuals, families, physicians, and society, however, might have a wide range of tolerance to the uncertainties associated with the undi­ agnosed disease state. Being undiagnosed has profound detrimental effects. Patients can go undiagnosed for decades, leading to personal and family uncertainty, high levels of stress, decreased productiv­ ity, limited accessibility to disease-specific counseling and resources, decreased quality of life, and excess utilization of medical services. APPROACH TO CHALLENGING DISEASES OF UNKNOWN ETIOLOGY Approaches to a patient with an undiagnosed disease can be separated into two categories. The first is a new assessment by a consultant, new provider, or diagnostic referral center. The second is periodic reassess­ ment by an existing provider for a patient who remains undiagnosed. TABLE 505-2  Essential Records for Undiagnosed Disease Patients 1. Any narrative summaries that detail the course of the illness 2. Copies of original test results with names, dates, testing circumstances, normal ranges, and test facility information 3. Electronic copies of imaging studies 4. Consultation notes 5. Hospitalization intake and discharge summaries 6. Accurate family history accounts and family relations Optional but potentially useful records include: 1. Photographs and/or videos of disease manifestations 2. Longitudinal data (growth charts, symptom logs, serial lab measurements) 3. Data or specimens that could be reanalyzed, including pathology specimens and genomic sequencing of raw data ■ ■COMPREHENSIVE DATA COLLECTION A potentially time-consuming but critical initial step is the gathering of all available medical data. Essential records are listed in Table 505-2. The overall goal of data collection is a full understanding of the course of the disease and a verification of critical data elements used for diagnostic decision-making. Incorrect or partial second-hand accounts of prior test results contribute substantively to incorrect or missed diagnoses. Analysis of the collected data allows for reconstruction of the process by which previous disease presentation, diagnostic thought processes, and test interpretation led to the current understanding of a patient’s illness. Unintentional obfuscation of the history and findings can result from missing records, incomplete recall by the patient and fragmentation, and propagation of information (and misinformation) in the medical record. Optimally, the presence and character of key features of the illness will be reinforced by perspectives derived from multiple evaluations. PART 20 Emerging Topics in Clinical Medicine ■ ■VALIDATION OF SUBJECTIVE AND OBJECTIVE FINDINGS Teasing apart the layers of a patient’s presentation often uncovers a variety of adaptive (and maladaptive) coping strategies. Some are idiosyncratic to the disease state (e.g., sun avoidance in a patient with xeroderma pigmentosum), whereas others are driven by psychoso­ cial factors and could become primary drivers of the phenotype. It is important to consider, however, that patients believed to have “functional” or “somatoform” disorders may have unrecognized underlying ill­ nesses, e.g., nonepileptic events (pseudo-seizures), frequently have concurrent bona fide epileptic events. Careful consideration of clinical phenomenology and associated findings on physical examination and ancillary investigations may provide clarity, affirmation, and effective redirection. Distinct clinical, radiographic, and laboratory abnormali­ ties provide entry points to the generation of a differential diagnosis and could become effective biomarkers of disease progression and response to interventions. Testing Strategies and New Technology  The historical exclu­ sion of a diagnostic hypothesis may be based on testing that is no lon­ ger state of the art. For example, congenital disorders of glycosylation (CDG) were historically diagnosed using transferrin isoelectric focus­ ing. It was subsequently found that the diagnosis of many CDG types required mass spectrometric and molecular approaches. The initial assessment of a patient with an undiagnosed disease should include a reassessment of the diagnostic logic and data used in past decisionmaking. In the absence of concrete diagnostic leads, the use of broad scope screening tools may prove beneficial in generating meaningful diagnostic hypotheses (Table 505-3). In some cases, newer testing options may be difficult to obtain and/or be costly. Prior probability of disease and available resources will factor in determining whether new diagnostic testing is practical. Molecular Approaches, Including Genomics  The availability and variety of clinical molecular modalities have transformed diag­ nostic testing in many settings. These advances have arisen from both TABLE 505-3  Clinically Available Tests with Notable Utility for Undiagnosed Cases TEST TARGET PHENOTYPES RATIONALE Single nucleotide polymorphism microarray and/or karyotype Dysmorphic features, cognitive impairment, neurodevelopmental disorders Genomic structure abnormalities may be missed by other testing Exome or genome sequencing Any undiagnosed disease that is chronic and not clearly acquired Tests a broad range of potentially unconsidered diagnostic entities Lysosomal storage diseases (LSDs), molecular or enzymatic tests. urine organic acids, urinary glycosaminoglycans (GAGs), oxysterols Progressive neurologic disorders, psychiatric disorders Some LSDs have nonspecific presentations, and adultonset cases are often missed Congenital disorders of glycosylation, Apo CIII and N-glycan mass spectrometry Pediatric-onset disorders, cognitive impairment, neurologic phenotypes Large group of disorders; phenotypes for many still being characterized Biochemical disorders, ammonia, serum polyols, urine purines and pyrimidines, plasma amino acids, very-longchain fatty acids Neurologic phenotypes, especially with waxing and waning course, selective speech involvement, or patients with unusual selfselected diets Metabolic disorders may have nonspecific symptoms, and adultonset cases are often missed Mitochondrial sequencing and mitochondrial depletion studies; biochemical screening with serum lactate, blood pyruvate, plasma amino acids, and GDF-15 Complex multisystem disorders with neurometabolic, endocrine, and gastrointestinal symptoms, muscle dysfunction, and waxing and waning or progressive course Large group of disorders with a wide range of presentations; yield is improved by studies in affected tissue (e.g., liver or muscle) Cerebrospinal fluid (CSF) studies including amino acids (AAs), lactate, pterins, methyltetrahydrofolate (MTHF), or special CSF flow studies Synthetic neurotransmitter defects in patients with unexplained fluctuating encephalopathy/ movement disorders or patients with atypical neuroinflammatory syndromes Patterns of profiles point to particular enzymatic deficits in neurotransmitter synthesis or characterization of unique immunologic profiles of inflammatory central nervous system diseases testing scope, such as exome-wide, genome-wide, and transcriptome (RNA sequencing [RNA-Seq]) sequencing, and new medical knowl­ edge, such as new disease-gene associations and molecular interaction networking (network medicine). Complementary screening tools, such as metabolomics, show diagnostic promise, particularly when combined with sequencing data to generate a fuller picture of disease manifestations. Simultaneous consideration of multiple data types can provide a means of appreciating overlapping and reinforcing evidence, with the potential to inform both hypothesis-driven and agnostic approaches to diagnosis. HYPOTHESIS-DRIVEN MOLECULAR TESTING  Hypothesis-driven test­ ing implies that a defined set of heritable (or potentially heritable) dis­ orders is the principal impetus for testing. Selection of a targeted gene sequencing panel, ideally augmented with structural variant detection, may allow for improved sensitivity, lower cost, and fewer unrelated (secondary) findings relative to full exome or genome sequencing stud­ ies. In the setting of an initial undiagnosed disease evaluation, prior sequencing panels may not include recently discovered genes. Testing with an updated panel or targeted sequencing of a newer gene is an option for consideration. In some cases, sequencing panels are gener­ ated by selective reporting of relevant genes within an exome dataset. In such cases, it may be possible to expand the analyses to new genes of interest without additional sequencing. AGNOSTIC MOLECULAR TESTING  Agnostic testing typically uses data from a broad testing platform, such as exome or genome sequencing, and considers all detectable diagnoses, even those with a low pretest probability of being present. This approach can also generate hypoth­ eses for potentially new disease-gene associations. Analysis of the sequencing data typically includes an unrestricted search throughout the entire human genome or exome space. DNA sequence variants with potential medical relevance are identified first by bioinformatic characteristics, including known association with disease, predicted importance for protein function, interspecies conservation, popula­ tion frequency, and an evolving list of other associated or functional factors. The list of candidate variants is then subject to expert review (i.e., curation). The interpretation of test results in this setting is highly influenced by both the adequacy of communication between the clini­ cal and testing teams and the information content of the data sources used to annotate each of the thousands of variants generated in the course of sequencing. There is a rapid proliferation of new testing platforms and analytical tools with the potential to contribute to solving undiagnosed diseases, but it remains challenging to judge their broad utility. While awaiting systematic validation and practice standards, novel techniques may be considered in special cases where a diagnostic hypothesis is closely aligned with the type of data generated by a specific testing strategy (Table 505-4). ■ ■PERIODIC REEVALUATION The cornerstone for the care of a patient in an undiagnosed disease state is a plan for periodic reevaluation. The Undiagnosed Diseases Network (UDN), a 10-year National Institutes of Health–sponsored national program, was specifically designed to evaluate undiagnosed patients. The overall diagnostic rate of the UDN, including periodic reevaluation of early enrollees, was reported as ~30%. This finding TABLE 505-4  Emerging or Special Testing Strategies and Related Diagnostic Questions AVAILABLE CLINICALLYa TESTING STRATEGY RELATED DIAGNOSTIC QUESTION Transcriptomics, RNA-Seq Relevance of splice, regulatory, and other noncoding variants; correlated changes in gene expression within pathways Yes Metabolomics Hypothesis generation via nontargeted approaches, correlated pathway changes, correlation with molecular findings Yes Epigenetics Diseases known or suspected to be caused by methylation or parent-of-origin effects Yes Transcriptional profiling Search for profile particular to certain disease states, e.g., interferon-inducible gene panels (interferon signature) in certain autoinflammatory disorders Some Specialized, diseasespecific testing Prion-related diseases, metabolic diseases, and many other assays Some Functional validation Model organisms, cell biology, and other approaches to validating a hypothesized gene-disease association No Metagenomics Search for molecular fingerprints of other organisms (e.g., infectious agents) within human samples Yes Long-read sequencing technology Accurate resolution of low-complexity regions of the human genome (repeat expansion disorders) and complex genome structural rearrangements No Deep sequencing Accurate resolution of low levels of mosaicism Some Optical genome mapping High-resolution chromosomal structure Yes aAvailable clinical tests are often a small subset of approaches available via research collaboration. Clinical testing offerings are evolving rapidly and should be reassessed periodically. illustrates the fact that many affected individuals remain in an undi­ agnosed state for a protracted period. For a medical provider, the care of an undiagnosed patient includes a program for symptomatic care, support related to the undiagnosed state itself, and plans for a regular reevaluation strategy seeking new insights into the diagnosis by follow­ ing its time trajectory. Reevaluation is guided by emerging knowledge in the field, disease progression, and the development of new signs and symptoms. The appearance of a similar disease in a sibling or close relative may provide critical insight. Communication with the patient is an essential component. Many individuals with an undiagnosed dis­ ease report feeling abandoned by their providers once initial diagnostic ideas have been exhausted. Providers themselves may feel discouraged in being unable to provide a diagnosis. The institution and discussion of a well-defined plan for periodic reassessment and communication can help reinforce the patient-provider relationship and set reasonable expectations. Reevaluation of Differential Diagnosis  The key to success for a planned reevaluation visit is preparation. The problem and differential diagnosis lists should be subject to careful, evidence-based review. New or resolved clinical features may add or remove diagnostic consider­ ations. The passage of time may result in the emergence of distinct new phenotypic manifestations that serve as new clues in the formulation of a definitive diagnosis. Special consideration should be given to the effects of reaching maturity and aging. The establishment of a pheno­ type as being static versus progressive has prognostic value. Careful documentation of the rationale for including or excluding individual disorders will streamline the process for both future reevaluations and the need for consultants. Concurrent development of common diseases should be thoughtfully considered as a possible component of the primary undiagnosed condition. For example, emergence of insulindependent diabetes mellitus in an undiagnosed patient with a complex phenotype could be a feature of the rare autoimmune polyendocrinopa­ thy associated with mutations in the autoimmune regulator gene AIRE. CHAPTER 505 Novel Approaches to Diseases of Unknown Etiology New Literature  Keeping abreast of current literature is an impor­ tant and challenging activity for all medical providers as the body of medical knowledge continues to grow exponentially. For undiagnosed diseases, newly reported disorders and disease-gene associations are an important source of diagnostic resolution. Literature search tools such as PubMed can be augmented by online resources that connect clinical signs and symptoms (phenotypes) to disorders. For example, using the search terms “cardiomyopathy arthropathy diabetes hyperpigmenta­ tion” in the Online Mendelian Inheritance in Man website (https:// omim.org) produces a list of disorders that includes hemochromatosis. In the context of an undiagnosed disease, this type of phenotypedriven approach can be used to search for new, relevant publications and disorders. Tools for search automation continue to be developed in both open-source and commercial settings. The success of these approaches is augmented by iterative application, ideally as part of formal, periodic reevaluation of the undiagnosed patient. ■ ■GENOMICS The use of medical testing based on the determination of DNA sequence and structure (sometimes referred to as molecular testing) has proliferated in recent years. A wide variety of approaches are available to the clinician, from single-gene sequencing to exome or genome sequencing. RNA sequencing and optical genome mapping have recently become available as clinical tests. Many reviews of this topic are available (see Marwaha et al., 2022, in “Further Reading”). Consultation with colleagues trained in genetics can be useful when developing an optimal testing approach. In some cases, genetic testing results may already exist in the medi­ cal record during the initial evaluation of an undiagnosed patient. This is increasingly true for younger patients; exome and genome sequenc­ ing are being used earlier, and with increasing frequency, for complex diagnostic challenges. Reanalysis of previously obtained exome and genome data should start with consideration of both the age and qual­ ity of the study and the reported patient phenotype at the time of the study report. For sequence results generated in a clinical laboratory, a discussion between the provider and the laboratory director often answers important questions about recommended next steps. The dis­ cussion should touch on how technologic advances have affected the utility of the older data and whether the laboratory offers reanalysis of the data. At a minimum, the provider, the testing laboratory, or an identified subspecialist should review previously reported DNA vari­ ants of unknown significance considering interval reports about the gene in question. More advanced reanalysis strategies are emerging and may be offered by the testing laboratory. Some laboratories offer release of raw DNA sequencing data to their patients on request. The utility of raw data varies and depends on the identification of bioinformatics collaborators willing to reanalyze the data. Sequencing data obtained as part of a research study may not be suitable for clinical diagnostic purposes. In practice, raw and research-generated sequencing data are most useful when a collaborating researcher can be identified. When considering a new sequencing test, the inclusion of the par­ ents and siblings of the proband has the potential to provide enormous value in some situations. Discussion of an optimal approach with an expert colleague or the testing laboratory is encouraged. ■ ■EXPOSOME In many cases, a detailed occupational and environmental expo­ sure history should be obtained. Some rare disease phenotypes are pathognomonic of specific toxicant exposures (e.g., mesothelioma and asbestos exposure, clear cell adenocarcinoma of the vagina and intrauterine diethylstilbestrol [DES] exposure, chloracne and exposure to halogenated aromatic hydrocarbons). For the most part, however, chemical toxicant exposures do not produce unique phenotypes. Rather, chemical exposures operate in conjunction with lifestyle factors (e.g., smoking, alcohol intake, and nutritional status), dif­ ferential host susceptibility (determined by age, sex, comorbidities, genetics, etc.), and nonchemical stressors (e.g., psychosocial stress) to produce (1) common, readily diagnosed medical diseases (e.g., asthma); (2) unusual or nonspecific phenotypes (e.g., erethism and metallic mercury exposure); or (3) atypical presentations of otherwise well-characterized disease states, initially considered an undiagnosed disease (e.g., manganese-induced parkinsonism). The nonspecificity typical of chemical-induced disease risk is further complicated by lack of exposure biomarkers for many common environmental toxicants (e.g., volatile organics), the short half-life of some contaminants (e.g., arsenic), and the possibility of decades long latency between exposure and disease onset (e.g., chemical carcinogenesis or dietary exposures to specific biochemical risk factors for atherothrombosis). In addition, we live in an era in which new chemicals are introduced into consumer products and the environment at a pace well beyond our capacity to characterize their toxicity. Within this context, one of the most power­ ful tools for ascertaining chemical-related disease risk is a systematic exposure history. Although there are no standardized instruments for this purpose, there are published guidelines to implement exposure assessments (Goldman and Peters, 1981; see “Further Reading” below). These include a multistep approach to exposure assessment including a job history; a review of exposures at work and at home or via hobbies and recreation; ascertainment of any temporal relationship of symp­ toms or disease onset to work, home, or recreational activities; and the food frequency questionnaire. If this screening identifies a potential exposure or exposures of concern with respect to patient symptoms and phenotype, a second step of evaluation involves a more detailed history to identify specific suspect agents, options for quantitative environmental exposure assessment (e.g., household tap water sam­ pling, review of workplace Material Safety Data Sheets [MSDSs]) and biomonitoring, and etiologic plausibility for at least some aspects of the patient’s phenotype. PART 20 Emerging Topics in Clinical Medicine The traditional approach to focused external exposure assessment proposed above does not, however, provide an integrated, quantitative measure of all exposures over the life course, an exposure characteriza­ tion of particular interest for the risk of chronic diseases such as cancer or atherothrombosis. The exposome has been proposed as a promising means for capturing the totality of human exposure over a lifetime (analogous to the totality of genetic exposure assessed via genomic analyses), including not only external chemical or dietary/foodome (Barabasi et al., 2020; see “Further Reading” below) exposures but also internal (e.g., metabolic, hormonal, microbiome) influences and psychosocial factors. However, techniques for measuring the expo­ some are in relatively early stages of development, are limited by the substantial variability in human exposure experience, and have not yet been designed to capture complex combinations commonly encoun­ tered in environmental or occupational exposure settings (Peters et al., 2012; Wild, 2012; Brunekreef 2013; Barabasi et al., 2020; see “Further Reading” below). This important element of assessing patients with undiagnosed disease is, however, evolving rapidly and offers the prom­ ise of becoming a more formal part of the evaluation of many patients with undiagnosed disease. ■ ■ENGAGEMENT OF RESEARCH APPROACHES Establishing a research collaboration for a patient with undiagnosed disease can be both challenging and rewarding. Time and effort resources are likely to limit this approach to a subset of patients with particularly compelling clinical presentations and a strong hypothesis about disease causation. The process must include early and detailed communication with the patient. Several approaches may be consid­ ered. Undiagnosed disease medicine, as a focus of specialized study, investment, and infrastructure, has proliferated around the world in the last decade, with numerous centers providing support for evaluat­ ing qualifying individuals and families. Leveraging Phenotypic and Genotypic Similarities  For a patient with a rare or undiagnosed disease and distinct presenting fea­ tures, finding similarly affected individuals adds substantial benefits. It can encourage research, provide a community for affected patients, and improve the chances of finding commonalities in pathogenesis and therapeutic strategies. Phenotypic aggregation may also allow the patient to connect with consortia invested in related medical presenta­ tions. Examples include organizations dedicated to the study of related diseases such as leukodystrophies, autoinflammatory disorders, and even undiagnosed diseases; NORD, the National Organization for Rare Disorders (rarediseases.org), can be a useful starting point. The Office of Rare Disease Research within the National Center for Advancing Translational Science supports consortia under the Rare Disease Clini­ cal Research Network program. Building patient cohorts may also be based on specific biological mechanisms or pathways, for example, the United Mitochondrial Disease Foundation. Data Sharing  The proliferation of DNA sequencing technology and the subsequent generation of many DNA variants of unknown clinical significance have prompted the creation of data-sharing resources spe­ cifically designed to match similar cases submitted by clinicians and researchers around the world. For example, a clinical exome report may identify variants in a gene with a potential but unproven relation­ ship to the patient’s presenting illness. The clinician could enter the gene name into a gene-matching database, and if the same gene name had been already entered by a different submitter, the database would flag a match and send contact information to both submitters. The matching procedure has the potential to identify additional cases of an ultrarare or newly described condition, while avoiding the shar­ ing of the patient’s personal health information. Embellishments of this approach involve inclusion of phenotypic features, data entry by families, and specific details of sequence variants. Example systems include GeneMatcher (the most populated database for single-gene submissions) and DECIPHER (which adds expended utility for larger structural variants). As an illustration of their utilization at the time of this chapter’s publication, GeneMatcher has single-gene submissions for almost 70% of all known protein coding genes. Collaboration  Collaborations around undiagnosed disease patients may take many forms. Studies focusing on related medical conditions can sometimes be identified using the https://clinicaltrials. gov website, which lists many U.S. and non-U.S. clinical studies. Data­ bases of clinical information (e.g., this textbook, GeneReviews) can be used to identify subject matter experts for related conditions. Such No diagnosis after comprehensive evaluation Symptomatic care, consider empiric treatments Consultative assessment Iterative assessment Comprehensive review and re-evaluation of evidence for prior conclusions Review new records, signs, symptoms, and environmental history Testing to validate key results and evaluate new diagnostic hypotheses Consider new literature and availability of new or updated testing strategies Consider hypothesisgenerating tests including genomic sequencing Reformulate differential diagnosis DIAGNOSIS? Yes No Document reasoning and supporting evidence Consider collaboration to explore hypotheses Work with patient to define concrete follow-up plan FIGURE 505-1  Approach to the patient with an undiagnosed disease. experts can be queried about ongoing studies. In some cases, a willing­ ness to work with consenting families to provide biological specimens can open additional avenues for collaboration. ■ ■CHALLENGES Data Portability  Obtaining specimens, data, and records for a chronically undiagnosed patient can be time-consuming and challenging. Families may be charged fees for obtaining copies of old studies. Although continuing advances in record access are occurring, families should be encouraged to collect and maintain an updated col­ lection of medical records. These should include copies of consultation notes, original laboratory results, and radiology studies (the latter pref­ erably in electronic form). These record collections are useful for con­ sultation, second opinions, and transitions between primary providers. Managing Illness Behaviors, Expectations, and Secondary Manifestations  Patients with undiagnosed diseases may present in any stage of the grieving process. Coping with uncertainty, loss of abilities, work, relationships, autonomy, and financial security com­ pound the primary manifestation of the disease. Patients may have a wide range of expectations about the possible benefits of achieving a diagnosis, including successful therapy. Patients of reproductive age may find that their greatest uncertainty surrounds the potential herita­ bility of their disorder, its effects on future reproductive decisions, and the potential risk it may represent to their children and living relatives. These factors may be equally or more disabling than the primary illness and require an individualized and multidisciplinary approach. CONCLUSION Chronically undiagnosed diseases present a complex challenge to patients, medical providers, and society at large. Development of a comprehensive plan for evaluation, reevaluation, and support requires a substantial investment of time and effort (Fig. 505-1). Achieving an accurate diagnosis removes at least one level of uncer­ tainty and allows for disease-specific counseling, therapies, resources, community engagement, and advocacy opportunities otherwise not afforded to undiagnosed patients. CHAPTER 505 ■ ■FURTHER READING Barabasi AL et al: The unmapped chemical complexity of our diet. Nat Food 1:33, 2020. Brunekreef B: Commentary: Exposure science, the exposome, and Novel Approaches to Diseases of Unknown Etiology public health. Environ Mol Mutagen 54:596, 2013. Goldman RH, Peters JM: The occupational and environmental health history. JAMA 246:2831, 1981. Lee CE et al: Rare genetic diseases: Nature’s experiments on human development. iScience 23:101123, 2020. Marwaha S et al: A guide for the diagnosis of rare and un-diagnosed disease: Beyond the exome. Genome Med 14:23, 2022. Peters A et al: Understanding the link between environmental expo­ sures and health: Does the exposome promise too much? J Epidemiol Community Health 66:103, 2012. Splinter K et al: Effect of genetic diagnosis on patients with previously undiagnosed disease. N Engl J Med 379:2131, 2018. Wild CP: The exposome: From concept to utility. Review. Int J Epide­ miol 41:24, 2012. This page intentionally left blank