Phenotype Matters: Why Case Report Quality Determines Diagnosis in Rare Diseases

In rare diseases, diagnostic success is often determined by… a few lines of well-collected, structured patient clinical data. Despite the increasing popularity of the genotype-first approach, precise phenotype description remains essential for accurate diagnosis and implementation of appropriate treatment in patients with rare diseases. The latest paper by Fabre A. et al., "Poorly described phenotypes add to the misfortune of rare diseases," published in 2025 in the European Journal of Medical Genetics, highlights the systemic problem of cursory documentation of symptoms and formulates practical recommendations to improve phenotypic descriptions in patients with rare diseases.

It's tempting to think that a single pathogenic variant explains a patient's entire history. However, Fabre and co-authors remind us that a phenotype isn't a list of traits taken out of context , but a patient profile that changes with age and the natural history of the disease, so the quality of the medical records is crucial.

The history of defining Alagille syndrome demonstrates the long development of a canonical phenotype. The authors point out that early reports focused on liver and cardiac symptoms. Only later were ophthalmological changes, characteristic facial features, and so-called butterfly circles added. A clear, practical definition was not formulated until 1987, after years of accumulating descriptions. This is a good example of how, without complete, consistent clinical records, the disease remains incomplete, and the diagnostic path is longer than it could have been.

With the advent of NGS (next-generation sequencing) and the genotype-first approach , we risk narrowing the phenotype to the most common manifestations, omitting rare, delayed, or negative (absent) symptoms. The authors emphasize a simple yet fundamental principle that must be kept in mind when dealing with rare diseases: absence of evidence is not evidence of absence .

Statistics can be ruthless in rare diseases. Fabre and co-authors calculate that if a symptom occurs in 5% of patients and is 50 times more common than in the general population, approximately 72 complete case reports are needed to have a >95% probability of the same symptom being recorded at least twice. This assumption is optimistic, however, and assumes that every present symptom will be described.

The reality, however, is worse. Positive associations are published far more frequently in the literature than carefully documented absence of a trait. It's no wonder that important phenotypic elements sometimes take decades to become canonized, as in the case of Microvillus inclusion disease (MVID), where the impact of the PFIC-like phenotype or MYO5B gene variants was not described until many years after the first reports.

In smaller hospitals and clinics, where patients with rare diseases appear sporadically, a reliable, structured phenotype description may be the only way to further diagnosis and, consequently, appropriate treatment. An overly general entry, such as "dysmorphic face," "abnormal teeth," or "delay," loses its informative value and does not help diagnosticians look deeper.

Fabre and co-authors emphasize that using standards in descriptions based on the Human Phenotype Ontology (HPO) , recording both the presence and absence of features, and updating the description over time increases the chance of accurate diagnosis and thus shortens the patient's diagnostic odyssey . The authors add that inaccurate case reports cost rare disease patients years of life without a diagnosis.

The proposals are pragmatic and feasible. First, they propose a return to publishing complete case reports in high- impact journals in a standardized, indexed case report format.

Secondly, they draw attention to the need to create a universal checklist based on HPO terminology, covering all organ systems, with mandatory marking of both abnormalities and correct findings as well as missing data , so that it is possible to conclude about the absence of symptoms.

Third, they point to the need for a publicly available database that would aggregate data from published cases and allow for updates over time . At the same time, the authors warn against the "lantern effect" in large registries: if classic manifestations are primarily recorded, these will be the most frequently seen. Even extensive networks such as the Global Alagille Alliance (over 1,400 patients from 32 countries) are limited in this regard. Electronic medical records can help, but only if there is a habit of also noting the absence of a given symptom in the medical history .

From the perspective of the Ministry of Health's announcement, which introduced a plan for rare diseases , yes, but on one condition... In a July interview with Prawo.pl, Deputy Minister Urszula Demkow confirmed that the Polish Registry of Rare Diseases will collect information on symptoms and treatment effectiveness, among other things, creating a knowledge bank for doctors and decision-makers, and that work on it is nearing completion. This means that the clinical component has been established at the conceptual level . Following Fabre's considerations, time will tell whether the Polish registry will actually take into account the phenotype so as not to fall into the trap of superficial descriptions, and whether it will support shortening the diagnostic path for patients with rare diseases.

The authors of the publication, paraphrasing Albert Camus's "misnaming of things adds to the misery of the world," argue that incomplete descriptions of rare diseases increase the "misery" of patients. Therefore , thorough disease description remains an effective weapon in establishing a diagnosis. A better phenotype means better genotype matching, shorter diagnostic time, and ultimately, appropriate therapy.

It also provides better material for national registries, which clinicians use to create recommendations and provide payers with more accurate monitoring of therapy effects and a lower risk of missed diagnoses.