Updates to gene-disease classifications and inheritance patterns for porphyrias.

The UROS gene has only been reported in relation to a single porphyria entity with a loss-of-function mechanism: congenital erythropoietic porphyria (CEP). Therefore, no lumping and splitting considerations were necessary. UROS (uroporphyrinogen-III synthase) is a cytosolic enzyme catalyzing the fourth step of the heme synthesis pathway. It catalyzes the cyclization of the linear substrate hydroxymethylbilane into uroporphyrinogen III. When UROS activity is reduced, hydroxymethylbilane can spontaneously cyclize into a uroporphyrinogen I isomer. Although this isomer, like uroporphyrinogen III, is a substrate for the next enzyme in the pathway, UROD (Uroporphyrinogen decarboxylase), the product, coproporphyrinogen I is not a substrate for the subsequent enzyme, CPOX (coproporphyrinogen-III oxidase), in heme synthesis. As a result, the non-enzymatically produced uroporphyrinogen I isomer represents a dead end in the pathway [ 54 ]. Deficiencies in UROS present as a phenotypic spectrum ranging from mild to severe, with clinical manifestations that differ primarily in degree of severity and correlate with the level of residual enzyme activity. Symptoms present as blistering skin lesions, hypo- or hyperpigmentation of the affected skin, and recurrent skin infections which collectively classify this condition as a blistering cutaneous porphyria [ 54 ]. The only known cure for severe disease is bone marrow transplantation which is required as early as possible after birth, to improve the chances of its effectiveness [ 2 , 11 , 51 ]. Mild to moderate forms are characterized by later onset which may result from mild symptoms starting in childhood and taking time to be properly diagnosed. The symptoms of anemia and transfusion dependence are absent for these cases. Adult-onset cases are ultrarare and may be due to onset of a clonal bone marrow disorder in which there is expansion of a clone of erythroblasts that contain a UROS variant, which may be inherited or somatic. Onset may precede the recognition of the bone marrow disorder by some years. Biochemically, UROS driven porphyria requires a greater loss of enzyme activity for clinical symptoms to manifest compared with semidominant porphyrias. Whereas severe symptoms in semidominant porphyrias typically occur when residual enzyme activity falls below ~25% of wild-type levels, severe UROS-related porphyria requires biallelic (homozygous or compound heterozygous) variants that nearly abolish enzyme function, typically reducing residual activity to <5% of wild-type. The term congenital erythropoietic porphyria (CEP) has historically been used to describe the phenotypic manifestations resulting from variants in the UROS gene. The word “congenital” indicates that the condition is present from birth, regardless of whether the cause is environmental or genetic. “Erythropoietic” denotes where red blood cell precursors are created in the bone marrow, which signifies the primary site of cytotoxic intermediate production. We have opted to drop the term “congenital”, as the diagnosis is defined by the presence of disease-causing UROS variants and to use “congenital” would be redundant. To follow the explicit dyadic naming approach and be consistent with the other pathway names, we requested UROS -related phototoxic porphyria be added as a synonym to the current porphyria entity (MONDO: 0009902) ( Fig. 3 ). The UROD gene has been reported in relation to two porphyria entities: porphyria cutanea tarda (PCT) and hepatoerythropoietic porphyria (HEP) ( Table 1 ). Recent literature reviews [ 2 , 11 , 52 ] mention both the ultra-rare HEP and the most common porphyria of all, PCT. PCT results from tissue-specific inhibition of hepatic UROD activity to less than ~25% of normal leading to chronic blistering photosensitivity ( Table 2 ). During adult life, environmental factors (e.g., alcohol, smoking, estrogens, hepatitis C, or HIV infection) can trigger the formation of a uroporphyrinogen decarboxylase (UROD) inhibitor via oxidation of uroporphyrinogen in hepatocytes. Processes that generate this oxidation event include CYP450 enzyme activity, oxidative stress, and excess hepatic iron. Heterozygous UROD variants with low penetrance are predisposing factors in Type II (familial) PCT but are absent in Type I (sporadic) PCT. But additional genetic factors (particularly HFE and certain CYP variants) may contribute to development of either type of this iron-related condition. Indeed, variants in HFE that predispose to iron overload are reported in >50% of PCT patients with active disease [ 54 ]. Hepatoerythropoietic porphyria is due to biallelic UROD variants, is ultrarare, and the diverse environmental and genetic susceptibility factors found in PCT are usually absent. It usually presents in childhood, has an erythropoietic component, and resembles UROS-related phototoxic porphyria clinically. The UROD enzyme carries out a four-step decarboxylase reaction on its eight-carbon uroporphyrinogen substrate resulting in the product, coproporphyrinogen, which is the substrate for the next step in the heme biosynthetic pathway ( Fig. 1 ) [ 55 ]. The protein is a homodimer so heterozygotes could produce three types of dimeric proteins: wildtype-wildtype; mutant-wildtype; and mutant-mutant. Described deleterious variants include null, splicing, and missense variants [ 56 ]. We looked at six reports [ 21 , 22 , 57 – 59 ] of HEP with compound heterozygosity or homozygosity ( Table 3 ) to see if the variants are clustered or specific to a region of the protein, but no pattern was found. One variant p.Gly281Glu has been reported to cause both HEP and PCT [ 60 ]. The different HEP variants exhibit a range of enzymatic activity in vitro ( Table 3 ) where activity correlates loosely with reported disease severity. The lack of a more direct correlation is unsurprising given the known role of environmental and other genetic factors in this porphyria. Recent reviews [ 2 , 11 , 55 , 61 ] also report no other functions for this enzyme. In summary, no evidence for the existence of two distinct molecular functions for this protein. Therefore, our analysis favored lumping these porphyria entities together. Recent reviews did not provide a penetrance estimate for Type II PCT but it is clearly very low, even if not well estimated. Additionally, there was no evidence reported for pleiotropy. However, there is evidence of variable expressivity as environmental factors play a huge role in risk for PCT. It has been estimated that 80% of the patients diagnosed with PCT are Type I (no UROD genetic predisposition) while 20% have a predisposing genetic factor in UROD [ 62 ]. A single variant can only decrease enzymatic function by 50%, but for cutaneous symptoms to present, enzyme activity must drop below 25% of wildtype activity ( Table 2 ). Severe disease is typically seen when activity is between 15–20% of WT ( Table 2 ). This means that all patients with monoallelic deficiency in UROD have additional environmental susceptibility factors leading to symptoms. Other important evidence collected during this curation relevant to lumping and splitting, is that three reports of individuals, biallelic for variants in UROD and thus diagnosed with HEP, were very mildly affected with symptoms more similar to PCT (see cases 1–3 in Table 3 ) [ 20 ] [ 21 ] [ 22 ]. Together this evidence is consistent with the disease worsening with lower enzyme activity where this results from some combination of pathogenic variants in UROD or HFE (causes iron overloading) or other genes, alongside other environmental effectors. While cases diagnosed as PCT are distinguished from HEP by their episodic nature, adult onset, and association with lower penetrance, the biallelic and monoallelic forms share the symptom of chronic blistering photosensitivity ( Fig. 2 ). Although hepatic cirrhosis is uncommon in both conditions, porphyrin accumulation can lead to some degree of liver damage. When cirrhosis does occur in PCT, it can usually be largely attributed to associated risk factors such as chronic alcohol use or hepatitis C infection. PCT is also the most treatable of the porphyrias and complete remission is expected with treatment and removal of triggering factors, but this is not the case with HEP. Affected individuals harboring biallelic variants are distinguished in that they present in infancy or childhood and have a more severe phenotype. Because the enzyme deficiency in HEP is permanent rather than caused by environmental factors, the condition does not respond to the treatments that are effective in PCT. HEP exhibits an autosomal recessive inheritance pattern, whereas PCT follows an autosomal dominant inheritance pattern. However, as discussed above, enzymatic activity must be reduced more than 50% for porphyria symptoms to present. Thus, autosomal dominant cases have compounding environmental factors. Furthermore, some biallelic cases have very mild presentations more similar to PCT than HEP [ 63 , 64 ]. This strongly supports a continuum of porphyria favoring lumping. The presence of biallelic variants and earlier onset, as seen in HEP, which sometimes lead to a more severe porphyria phenotype due to additional loss of enzymatic activity, is not considered sufficient to justify retaining the split between HEP and PCT. Instead, and similar to other porphyria genes, the evidence points toward a continuum of disease in favor of lumping. The General IEM GCEP, using the lumping and splitting criteria, found the molecular mechanism (UROD loss-of-function) to be consistent between the HEP cases with biallelic variants and the PCT cases with monoallelic variants. Therefore, cases caused by inherited UROD variants have been curated as a single porphyria entity with a semidominant mode of inheritance. Porphyria cutanea tarda (PCT) has been a clinical term since 1937, when Jan Gösta Waldenström described a porphyria characterized by blistering cutaneous symptoms and liver involvement. While the Latin terms “cutanea” (relating to skin) and “tarda” (late) reflect porphyria phenotypes, their meaning has become less clear due to decreased fluency in Latin. The term “cutanea tarda,” is also insufficiently differentiating as multiple porphyrias present with later onset cutaneous symptoms, and it does not encompass the rare, early-onset cases of hepatoerythropoietic porphyria (HEP). To resolve these issues, we took the explicit dyadic naming approach, incorporating UROD -related as a genetic descriptor and dropping “cutanea tarda”. This resulted in the term UROD -related phototoxic porphyria (MONDO:0100498, OMIM# 176100) which encompasses cases presenting with earlier onset and more severe cutaneous symptoms, as well as monoallelic cases. The existing clinical subtypes: porphyria cutanea tarda (MONDO:0015104) and hepatoerythropoietic porphyria (MONDO:0019799) are positioned in MONDO as child terms to this novel porphyria term to avoid disrupting their continued use in the clinic ( Fig. 3 ). The CPOX gene was first reported in relation to hereditary coproporphyria (HCP) in 1994 [ 50 ]. Since that time, rare biallelic cases have been reported, with two cases diagnosed as homozygous HCP [ 65 ], [ 66 ], [ 67 ] and other cases diagnosed as harderoporphyria [ 68 ], [ 69 ] ( Table 4 ). The OMIM database recognizes coproporphyria and harderoporphyria assertions for CPOX ( Table 1 ). Recent literature reviews also mentioned both HCP and HD-HCP, and discussed a molecular distinction for harderoporphyria [ 2 ], [ 11 ]. However, there was some inconsistency in how these porphyrias were referred to. Review [ 2 ] mentioned harderoporphyria as a variant of HCP while review [ 69 ] referred to harderoporphyria as homozygous dominant hereditary coproporphyria (HD-HCP). With this information in hand, we turned to examine the molecular mechanism, phenotypic variability, and inheritance pattern to better inform our lumping and splitting decision. The CPOX gene encodes the sixth enzyme of the heme biosynthesis pathway, coproporphyrinogen oxidase (CPO) ( Fig. 1 ). CPO catalyzes the conversion of coproporphyrinogen III to an intermediate, harderoporphyrinogen, which is subsequently converted to protoporphyrinogen IX [ 70 ]. CPO functions without a cofactor or a metal ion, but requires molecular oxygen as an electron acceptor [ 71 ]. More than 20 variants associated with HCP have been mapped onto the protein structure; they are spread throughout the gene and a high proportion are missense [ 71 ], [ 72 ]. Important to lumping and splitting decisions, one variant p.Arg331Trp produced neonatal symptoms when biallelic and was diagnosed as homozygous HCP, yet also caused typical HCP when monoallelic, thereby demonstrating that both are following the same loss-of-function mechanism [ 72 ]. Variants reported to cause harderoporphyria impair the enzyme’s ability to bind the harderoporphyrin intermediate, disrupting the second decarboxylation step. For example, the p.Lys404Glu variant affects a type 1 β -turn introducing an electrostatic repulsion causing the intermediate to be prematurely released. Studies have found residues 401–405 are crucial for preventing premature release of the harderoporphyrin intermediate [ 71 ]. Consideration of phenotypic variability for harderoporphyria cases should determine whether this molecular distinction meets the required criteria for these diseases to be retained as distinct entities. We needed to assess biallelic cases for phenotypic distinctions. Are there different phenotypes between general missense variants and the variants causing early release of the harderoporphyrin intermediate? To explore this, we created table 4 listing the reported biallelic cases. Given that p.His327Arg is now recognized to affect harderoporphyrin release [ 69 ] we cataloged only one case of homozygous HCP where both alleles are not known to affect harderoporphyrin release. Interestingly, this case also presented with an acute neurovisceral attack, which usually only occurs in monoallelic cases [ 11 ]. Another case (#2 in Table 4 ) was diagnosed with homozygous HCP, but the other allele impairs harderoporphyrin release. Another case (#3 in Table 4 ) was diagnosed with harderoporphyria but lacks genetic information. Finally, case #6 is diagnosed with harderoporphyria but one allele contains a splice site variant leading to total loss of exon 6. These findings led us to conclude that it would be hard to make a case for a phenotypic distinction between these groups: homozygous HCP (one clear case) and harderoporphyrin cases (three clear cases). What we can state at present is that patients with biallelic variants affecting harderoporphyrin release exhibit symptoms as neonates, including hemolytic anemia and severe jaundice. Biochemically, these patients have prominent accumulation of harderoporphyrin in feces and some in erythrocytes. Variant scientists should add notes on these variants indicating that they specifically affect harderoporphyrin release. For monoallelic cases, affected individuals mainly experience acute neurovisceral attacks with adolescent or adult onset, characterized by severe abdominal pain as well as acute motor neuropathy and other neurological symptoms ( Fig. 2 ). These attacks are similar to other forms of acute hepatic porphyria in that they occur with incomplete penetrance (90% of heterozygotes never experience an acute attack) across the carrier population and can have various triggers such as fasting, the use of certain medications, hormonal changes accompanying the menstrual cycle, or infections. The known environmental triggers exacerbate the underlying insufficiency by increasing flux through the heme biosynthesis pathway enhancing accumulation of the precursor to coproporphyrin. Repeated attacks can result in chronic liver injury, and some cases show blistering photosensitivity. As noted for other heme synthesis genes, CPOX variants show variable expressivity, meaning that within a family, people with the same variant can experience different levels of symptoms. In summary, biallelic harderoporphyria and monoallelic HCP share some phenotypic overlap, but differ substantially in age of onset and severity ( Fig. 2 ). Consistent with this, the level of CPOX activity in lymphocytes from heterozygous patients are 50% of controls ( Table 2 ) [ 72 ] while those from biallelic cases, are approximately 2–18% of control levels ( Table 2 & 4 ), providing a basis for their greater severity. Overall, HCP, HD-HCP, and harderoporphyria follow a loss-of-function mechanism where porphyria severity tracks residual enzyme activity. Accordingly, we favored lumping these two disease entities into a single gene curation for the purpose of variant curation. Harderoporphyria and HD-HCP cases harbor biallelic CPOX variants (autosomal recessive inheritance), while HCP cases harbor a monoallelic CPOX variant (autosomal dominant). Although the difference in inheritance pattern could support splitting these into separate gene curations, distinctions based solely on age of onset or severity are insufficient justification. The General IEM GCEP found the molecular mechanism (CPOX loss-of-function) to be consistent between the HCP cases with monoallelic variants and the harderoporphyria and HD-HCP cases with biallelic variants. In addition, the metabolic and phenotypic differences between the diverse cases appeared to represent a spectrum of porphyria that differed based on age of onset and severity. Therefore, cases caused by inherited CPOX variants have been lumped into a single porphyria entity with a semidominant mode of inheritance. Hereditary coproporphyria (HCP), HD-HCP, and harderoporphyria have been the historical terms for referencing porphyria caused by variants in CPOX . The name “hereditary coproporphyria” derives from coproporphyrin III (the reactant) which has the highest rate of accumulation in mutational impairment of CPOX . However, additional heme synthetic pathway intermediates, including delta-aminolevulinic acid (ALA), porphobilinogen (PBG), and uroporphyrinogen (URO), also accumulate in affected individuals ( Fig. 1 )[ 72 ]. Harderoporphyrin represents an intermediate in the enzymatic reaction. Given our decision to consolidate the prior entities (HCP and harderoporphyria), the revised terminology must encompass the full phenotypic spectrum of CPOX -related disease. Following the explicit dyadic naming approach, we created CPOX -related neurovisceral and phototoxic porphyria (MONDO:0800180) ( Fig. 3 ) to unify HCP and harderoporphyria and identify this as one of the four neurovisceral porphyrias which also have chronic blistering photosensitivity. We believe “neurovisceral” will help clinicians more easily recall and remember the four acute hepatic porphyrias which have similar porphyria presentations. The term “hereditary” has been omitted, as genetic inheritance is inherently implied in the diagnosis of a Mendelian disorder. Additionally, specific metabolite references were removed, as multiple intermediates accumulate in this porphyria. This revised terminology reflects the molecular and clinical complexity of CPOX -related neurovisceral and phototoxic porphyria. The existing phenotypic subtypes hereditary coproporphyria (MONDO:0007369, OMIM# 121300) and harderoporphyria (MONDO:0030048, OMIM# 618892) are positioned as child terms to the lumped entity within MONDO. Please note that despite the lumping of monoallelic and biallelic forms under a broader parent term to facilitate variant curation, the preserved child terms continue to have utility based on the key biochemical and clinical differences that distinguish them, including the additional erythropoietic component that forms a unique and important feature of the biallelic cases. Two porphyria entities with differing modes of inheritance ( Table 1 ) have been reported for the gene PPOX [ 73 ]. Autosomal dominant variegate porphyria is the typical form of the disease where affected heterozygous individuals are at risk for an acute attack, which usually presents with abdominal pain, post puberty onset of photosensitivity, or both. Autosomal recessive variegate porphyria is the ultra-rare form of the disease with only 11 cases having been reported as of 2018 [ 73 , 74 ]. We found disparity in the literature regarding how these porphyria entities are discussed. Some publications have described the gene-disease relationship for PPOX as autosomal dominant with incomplete penetrance [ 74 , 75 ]. Other papers have described the gene-disease relationship as autosomal dominant without mention of the ultra-rare recessive form [ 76 ]. A 2018 review article discussed this porphyria as a single entity, variegate porphyria, but used terms like “homozygous dominant” and “ultra rare autosomal recessive” to distinguish the more severe clinical manifestation of the porphyria from the more typical [ 75 ]. Recognizing that these porphyrias were inconsistently referenced in the literature led us to place greater emphasis on other criteria, such as the molecular mechanism, for our consideration of lumping and splitting. There is substantial evidence supporting a single molecular mechanism of disease for both reported forms of variegate porphyria. The gene PPOX encodes step 7 of the pathway ( Fig. 1 ). The enzyme protoporphyrinogen oxidase converts protoporphyrinogen-IX to protoporphyrin-IX. This enzyme is localized to the mitochondria and requires the cofactor FAD. Clinically observed variants have been mapped to the PPOX structure and separated into the following different categories: affecting FAD binding, affecting substrate binding, affecting the hydrophobic core, affecting protein surface interactions and affecting protein secondary structure [ 76 ]. In curation of genetic evidence for PPOX , we noted the severe biallelic variants occur in all categories, except affecting protein secondary structure [ 76 ]. This demonstrates that variants for the more severe form of this porphyria are not clustering to a particular part of the protein as is typically the case for justification of splitting separate disease entities in a single gene. The molecular mechanism is loss-of-function for both disease entities. Recent reviews for PPOX and the heme synthetic pathway do not provide or mention evidence of pleiotropy for this gene [ 2 , 11 , 18 , 52 , 76 ]. Consistent with this we did not find evidence that PPOX participates in different biological pathways besides the heme synthesis pathway. We also found no evidence of distinct phenotypic patterns exclusively linked to specific molecular mechanisms differentiating the severe and typical form of this porphyria. However, there are system level distinct biochemical and clinical features that depend on whether an individual is mono- or biallelic for a PPOX variant. Biallelic pathogenic variants, which severely reduce enzyme activity (<25% of normal), result in earlier onset with greater severity [ 77 ] and are associated with these specific symptoms: brachydactyly, clinodactyly, intellectual disability, nystagmus, myopia, and growth retardation ( Fig. 2 ). Symptoms specific for the monoallelic form (not mentioned for biallelic cases) include abdominal pain, constipation, vomiting, and muscular paralysis. The overlapping symptoms are chronic blistering photosensitivity, neuropathy, and psychosis ( Fig. 2 ). Individuals who are heterozygous for disease-causing variants are at risk for an acute attack, usually presenting with abdominal pain, post puberty onset of photosensitivity, or both ( Fig. 2 ). The presence or absence of attacks is determined by triggers (certain drugs, alcohol, low-carb diets, etc.) that further enhance the accumulation of pathway metabolites. Individuals with >75% of wildtype enzymatic activity are typically unaffected without any symptoms, while individuals carrying biallelic variants that reduce activity to <20% have severe porphyria ( Table 2 ) [ 78 ]. Penetrance for variegate porphyria has been calculated using female and male patient data from the Finnish porphyria registry. The mean penetrance was 40% for variegate porphyria, but the penetrance was notably higher in females at 44%, which is in accordance with the observation that 90% of symptomatic individuals are female [ 79 ]. Taken together the lack of evidence for pleiotropy and the observation of low penetrance indicate a phenotypic spectrum where severity correlates with enzymatic activity, supporting a decision to lump these two porphryia entities. Inheritance pattern was our fourth consideration for lumping and splitting for PPOX . The two asserted disease entities do have different inheritance patterns, autosomal recessive for the more severe type, and autosomal dominant for the more typical type [ 73 ]. As mentioned previously, differences in inheritance pattern provide support to split into two separate gene curations when differences exceed greater severity and age of onset. In this case, because the two forms of variegate porphyria share the same molecular mechanism, have overlapping symptoms, and differ clinically primarily in severity and age of onset, they can be grouped together under one classification. Based on ClinGen lumping and splitting criteria, two gene-disease entities for variegate porphyria were lumped together into one porphyria with a semidominant inheritance pattern. Both previous porphyria entities shared the name “variegate porphyria,” making renaming seemingly straightforward. However, while “porphyria” remains an essential component of the nomenclature, whether we should retain “variegate” was unclear according to the recommendations for disease naming. The term “variegate” comes from the Latin verb varigare, meaning “varied” or “diverse,” and was initially chosen to reflect the diversity of symptoms associated with the condition. However, this variability is not unique to PPOX; all four hepatic porphyrias are linked to highly variable porphyria presentations. Furthermore, decreased literacy in Latin makes a straightforward reading of the term uncommon, leading to our decision to drop the term “variegate”. This porphyria exhibits symptom overlap with other neurovisceral porphyrias yet also has chronic blistering photosensitivity, therefore, we created the new MONDO parent term: PPOX -related neurovisceral and phototoxic porphyria (MONDO:0700383) ( Fig. 3 ). The existing subtypes variegate porphyria (MONDO:0008297, OMIM# 176200) and variegate porphyria, childhood-onset (MONDO:0957577, OMIM# 620483) have been repositioned within MONDO as child terms to the newly created term, to avoid disrupting their continued utility in the clinic. The gene FECH which codes for the enzyme ferrochelatase has been reported in association with a single porphyria entity, erythropoietic protoporphyria (EPP). Therefore, no lumping and splitting considerations were necessary. Ferrochelatase catalyzes the eighth and final step in heme biosynthesis incorporating iron into the protoporphyrin ring. However, this porphyria was initially mischaracterized with an autosomal dominant inheritance pattern [ 80 ]. This is because the classical clinical presentation is caused by a loss-of-function variant in trans with an intronic hypomorphic allele NM_000140.5 (FECH):c.315–48T>C (this variant has a frequency of ~10% prevalence in Caucasians, ~18% in the Latino/Admixed American populations, and ~34% prevalence in East Asia (China and Japan))[ 81 ]. This common allele causes aberrant splicing of exon 3 resulting in an additional 63 base pairs to exon 3 which causes nonsense mediated decay of the transcript. Combined with a null allele in trans, the common NM_000140.5 (FECH):c.315–48T>C variant causes enzyme levels to drop below 35%. Below this level, clinical symptoms of acute non-blistering photosensitivity to sunlight begin to appear. These symptoms of phototoxicity and hepatoxicity are caused by protoporphyrin accumulation in the plasma, skin, feces, and liver. The lower the functional FECH activity, typically the more severe porphyria (cases with two loss-of-function alleles in trans (ultrarare) report severe porphyria often with liver complications). Before the implications of this intronic variant were discovered, the inheritance pattern had been incorrectly considered to be autosomal dominant. Now, the common intronic hypomorphic FECH allele (c.315–48T>C) often causes confusion in genetic testing reports, which often report it is a pathogenic variant, which is strictly true. However, this variant is i) common in the normal population, ii) it alone does not cause porphyria, and iii) only causes porphyia when combined with a null allele in trans. Erythropoietic protoporphyria (EPP) or Erythropoietic protoporphyria 1 has traditionally been used to describe the porphyria associated with variants in the FECH gene ( Table 1 ), which leads to the accumulation of protoporphyrin IX in erythroid cells. The term “erythropoietic” refers to the primary site of accumulation, while “proto” is used to denote the specific porphyrin intermediate (protoporphyrin IX) that accumulates in this condition. To align with the nomenclature used for the rest of the pathway, and following the explicit dyadic naming approach, we requested the name FECH -related phototoxic protoporphyria be added as a synonym to the existing MONDO entity (MONDO:0008319).

The ALAS2 gene has been reported in relation to two distinct disease entities by OMIM [ 33 ], and recent literature reviews ( Table 1 ) [ 33 ]. The ALAS2 gene was first reported in relation to X-linked erythropoietic protoporphyria in 2008, with the publication of eight affected families of diverse ancestries all carrying deletions in the C-terminal region of ALAS2 [ 34 ]. Prior to this, ALAS2 had an already established gene-disease relationship to X-linked sideroblastic anemia (XLSA). ALAS2 encodes a mitochondrial enzyme, delta-aminolevulinate synthase 2, that contains an autoregulatory C-terminal extension. Frameshift indel variants that either delete or elongate this C-terminal extension lead to protoporphyria by disrupting its autoinhibitory function [ 35 ]. These variants result in a gain-of-function phenotype, enhancing enzyme activity and increasing total protoporphyrin accumulation beyond the demand for heme production [ 34 ]. This gain-of-function mechanism is distinct from X-linked sideroblastic anemia (XLSA), which arises from loss-of-function variants in ALAS2 , leading to insufficient heme synthesis [ 36 ]. The mutational landscape also differs between these two disorders: XLSA-associated variants typically occur within exons 5–11 but may include missense variants in the C-terminal extension. In contrast, variants for XLP are indels that delete or elongate the autoregulatory C-terminal region. ALAS2 nicely showcases how different variants within the same protein can give rise to distinct diseases: one due to a loss-of-function mechanism and the other due to a gain-of-function mechanism. This was a clear case of a need to retain a split for these disease entities. Individuals affected with XLP frequently present in early childhood with severe, non-blistering cutaneous photosensitivity and a marked accumulation of erythrocyte protoporphyrin. In both XLP and XLSA, ferrochelatase activity is normal. However, the biochemical profiles differ significantly. In XLP, overproduction of protoporphyrin leads to accumulation of both metal-free protoporphyrin and zinc protoporphyrin, with the metal-free protoporphyrin fraction dominating the total. This metal-free fraction is primarily responsible for photosensitivity. Patients with XLP may also exhibit decreased iron stores, abnormal transferrin saturation, hyperbilirubinemia, elevated liver enzymes, and other signs of liver involvement. In contrast, individuals with XLSA present with sideroblastic anemia and iron overload due to ineffective erythropoiesis, but without photosensitivity. While zinc protoporphyrin is elevated in XLSA, total protoporphyrin levels are not typically increased to the same extent as in XLP, and metal-free protoporphyrin does not accumulate. Consequently, the phenotypic differences between the two groups of cases appear to represent separate disease entities rather than a spectrum of disease. Both XLSA and XLP follow an X-linked inheritance pattern. Although XLP was initially reported as X-linked dominant with 100% penetrance in males and females, this designation has since been revised. For females affected by XLP, phenotypic manifestation is influenced by the amount of wildtype X-chromosomal inactivation. Heterozygotes where the wildtype allele was not inactivated did not show clinical symptoms, and only minor increases in erythrocyte zinc and metal-free protoporphyrin concentrations [ 37 ]. Thus, while the inheritance patterns of XLSA and XLP are not fundamentally different, differences in assertion criteria, molecular mechanisms, and phenotypic variability supported the decision to classify them as separate disease entities. Therefore, we categorized cases caused by ALAS2 variants into two distinct, separately curated disease entities: X-linked erythropoietic protoporphyria (MONDO:0010420, OMIM #300752) and X-linked sideroblastic anemia 1 (MONDO:0020721, OMIM #300751) ( Fig. 3 ). This classification reflects distinct molecular mechanisms: gain-of-function in X-linked erythropoietic protoporphyria vs. loss-of-function in X-linked sideroblastic anemia. As part of the broader effort to standardize porphyria nomenclature across gene curations using an explicit dyadic format that starts with the name of the gene, the group requested an update of MONDO:0010420 to provide ALAS2 -related phototoxic protoporphyria as a synonym to the existing porphyria names. One of the motivations for this update was to facilitate future ALAS2 variant curations under this more standardized entity. The term “protoporphyria” has become accepted to include diseases in which there is photosensitivity due to elevations in circulating levels of erythrocyte metal-free protoporphyrin. The term “phototoxic” was specifically chosen because the accumulated protoporphyrin IX in circulating erythrocytes triggers phototoxicity, primarily through damage to endothelial cells in the skin upon light exposure. The ALAD gene has only been reported in relation to a single porphyria entity with a loss-of-function mechanism although different disease names have been asserted ( Table 1 ) [ 38 , 39 ]. Therefore, no lumping and splitting considerations were necessary, but nomenclature updates were. ALAD catalyzes the second step in the heme synthesis pathway ( Fig. 1 ). The enzyme is primarily octameric but exists in equilibrium with some hexamers (lower activity) and dimers (inactive). The octameric structure is built from a properly assembled dimer. pH, metals (particularly lead), and drugs can all affect the dimer domain orientation and thereby the quaternary structure of the enzyme. The enzyme binds two molecules of 5-aminolevulinate per subunit and condenses them to form porphobilinogen [ 40 , 41 ]. The porphyria associated with deleterious variants in this gene is extremely rare, characterized by neurovisceral attacks without phototoxic manifestations, and is one of the four acute hepatic porphyrias. Loss of ALAD activity causes primarily the accumulation of the potentially neurotoxic porphyrin precursor 5-aminolevulinic acid (ALA), which is a linear non-fluorescent molecule. Our complete gene-disease curation can be found by searching for the gene name on www.clinicalgenome.org . The previous disease name in MONDO for this porphyria was porphyria due to ALAD dehydratase deficiency (MONDO:0013000) while the name in OMIM was acute hepatic porphyria ( Table 1 ). Common names used in the clinic are: ALAD porphyria or ALAD deficiency porphyria. 5-Aminolevulinic Acid Dehydratase Deficient Porphyria and δ-Aminolevulinate dehydratase (deficiency) porphyria were also found in the literature ( Table 1 ). The prior description in MONDO contained a typographical error and referenced ALAD as “DALAD”, an acronym that is rarely used in the literature. Accordingly, in collaboration with MONDO we revised the errors 11 , referenced ALAD with the approved gene symbol, and requested MONDO include the name ALAD -related neurovisceral porphyria as a synonym. As noted in the introduction, we think it is particularly important for OMIM to update the disease name for this porphyria to either ALAD deficiency porphyria or ALAD-related neurovisceral porphyria to avoid the confusing overlap that currently exists with “acute hepatic porphyria” which is more often used as an umbrella term to unify the neurovisceral porphyrias. Deleterious variants in the HMBS gene have been reported in relation to the disease acute intermittent porphyria (AIP)[ 42 ]. Most individuals with classic AIP exhibit reduced HMBS enzymatic activity in erythrocytes, although a subset (5–10% of patients) [ 18 ] does not, resulting in a second disease assertion for this gene: acute intermittent porphyria nonerythroid variant. Rare, severe cases of AIP that harbor biallelic HMBS variants have also been reported [ 43 , 44 ]. These cases generally have onset in early childhood, present with more severe neurological phenotypes including developmental abnormalities of the brain, ataxia, dysarthria, and leukoencephalopathy, and can have ocular issues such as cataracts and optic nerve hypoplasia. Some studies have called these severe recessively inherited forms homozygous dominant acute intermittent porphyria (HD-AIP) [ 43 ]. OMIM recognized these autosomal recessively inherited forms as separate entities of porphyria-related encephalopathy and porphyria-related leukoencephalopathy ( Table 1 ) [ 45 ]. Because these homozygous cases are extremely rare, the literature describing them is limited, and the terminology (particularly “porphyria-related encephalopathy and porphyria-related leukoencephalopathy”) does not represent current consensus. In current practice, the field most commonly refers to these conditions as HD-AIP [ 11 ]. Notably, one recent literature review treated AIP as a single entity without specifying subtypes [ 45 ]. As recent literature did not consistently mention all these asserted porphyria entities, we anticipated lumping would be necessary, and turned to additional criteria to guide our decision on how to curate this disease. Review of the molecular mechanism for three asserted subtypes of AIP (acute intermittent porphyria, porphyria-related leukoencephalopathy, porphyria-related encephalopathy) provided strong evidence for lumping into a single disease entity. More than 500 disease-causing variants have been identified in the HMBS gene, showcasing significant mutational heterogeneity. These include missense, nonsense, deletions, insertions, and splice site variants [ 46 ]. Disease-causing variants are spread across all three domains of the protein, yet no statistical significance was found for an association between the mutational position and HMBS enzymatic activity [ 46 ] [ 47 ]. These variants cause decreased or complete loss of enzyme function, and no evidence was found for a gain-of-function mechanism. Important for lumping and splitting decisions, variants are not specific between these asserted porphyrias. For example, the NM_000190.4 (HMBS):c.499C>T(p.Arg167Trp) variant was seen in both biallelic cases [ 44 ] and monoallelic cases [ 48 ]. Also variants NM_000190.4 (HMBS):c.104C>T (p.Thr35Met) and NM_000190.4 (HMBS):c.518G>A (p.Arg173Gln) have been seen in biallelic [ 49 ] and monoallelic cases. Therefore, we saw no evidence for distinct molecular mechanisms between these three asserted porphyrias supporting a decision to lump them as a single entity. Regarding the fourth entity, acute intermittent porphyria nonerythroid variant, its molecular mechanism was loss-of-function but involved variants in exon 1, which is not expressed in erythrocytes, resulting in normal HMBS activity in erythrocytes. This is because there is an erythroid-specific promoter downstream of the housekeeping HMBS promoter and when used it produces a transcript that contains only exons 2–15. In contrast, the housekeeping promoter produces a transcript containing exons 1 and 3–15. So deleterious variants occurring within exon 1 or affecting the splicing of exon 1 to exon 3 do not impact the erythrocyte isozyme, but do impact the more broadly expressed housekeeping isozyme [ 49 , 50 ]. Five of the heterozygous probands we scored exhibited this less common nonerythroid subtype and harbored variants disrupting the splice site of the isoform-specific intron 1 [ 50 , 51 ]. Although these individuals harbor a unique subset of deleterious variants, further examination of phenotypic variability was necessary to determine whether this clinical presentation is sufficiently unique to warrant separate disease entities. Individuals with pathogenic monoallelic variants in the HMBS gene have been reported across a diverse range of ethnicities, with the majority remaining unaffected throughout their lifetimes. Using large population databases, the prevalence of disease-causing HMBS variants was estimated to be 1 in ~200,000 individuals of European ancestry [ 11 ]. These numbers suggest that the penetrance of AIP is ~1% among heterozygotes. However, penetrance within affected families is much higher, around 20–30%, which suggests that there are unknown modifying genes or other environmental factors [ 2 , 11 , 17 , 51 ]. Cases diagnosed as acute intermittent porphyria (the vast majority) are distinguished from HD-AIP (the extremely rare biallelic forms) by their episodic nature and adolescent / adult onset. Shared phenotypic overlap between biallelic and monoallelic cases is reduced HMBS activity, weakness, neuropathy, paresthesia, and seizures ( Fig. 2 ). The rare biallelic forms are phenotypically distinguished from monoallelic forms by early onset and increased neurological manifestation. Disease manifestation correlates with the degree of residual HMBS activity, and <4% of WT enzyme function is typical for the rare biallelic form ( Table 2 ). Thus, current evidence supports a model in which phenotypic variability reflects quantitative differences in loss of the same protein function rather than distinct molecular mechanisms. Regarding whether the nonerythroid porphyria entity can be lumped with these other entities, we did not find evidence of distinct clinical features or outcomes in recent reviews of porphyria [ 2 , 11 , 18 , 52 ], of HMBS [ 46 ], or of the nonerythroid subtype [ 53 ]. Together this information supports lumping of the nonerythroid disease variant with the other disease entities. However, this does not preclude a variant scientist from adding variant-specific comments on whether or not the erythroid transcript is involved. Indeed, for variants affecting exon 1, variant scientists should include this information so that it appears in clinical genetic testing reports. Such comments will help explain why an individual (or family member) carrying the variant shows normal erythrocyte HMBS enzyme activity despite having a pathogenic HMBS variant. While classical presentation of AIP has an autosomal dominant inheritance pattern, the handful of severe cases, which present within the first years of life, have autosomal recessive inheritance [ 43 , 44 ]. To keep the disease entities separated based on inheritance pattern, phenotypic features should extend beyond differences in severity or age of onset. While the systemic biochemical consequences in these ultrarare biallelic cases do differ substantially from those observed in typical monoallelic presentations, there are also multiple shared variants between biallelic and monoallelic cases. This evidence supports a single molecular mechanism, favoring the lumping of the ultra-rare autosomal recessively inherited disease entities with the classical autosomal dominantly inherited cases into a single disease entity with a semidominant inheritance pattern. Per criteria outlined by the ClinGen Lumping & Splitting Working Group, the molecular mechanism, HMBS loss-of-function, is consistent across cases with biallelic and monoallelic variants. Phenotypic differences between biallelic and monoallelic patients align with varying degrees of enzyme activity loss (see Table 2 ). Consequently, we lumped into a single porphyria entity for variant curation purposes. Historically, acute intermittent porphyria (AIP) has been the term used to describe porphyria caused by variants in the HMBS gene. The term “acute” reflects the sudden and severe onset of symptoms, while “intermittent” indicates the condition’s low penetrance and episodic nature, often triggered by estrogen/progesterone, oral contraceptives, alcohol, drugs, stress, or infections. As highlighted in the introduction, the other acute hepatic porphyrias exhibit significant symptom overlap. To highlight this commonality, we adopted the term “neurovisceral” across these conditions which is already well used across the literature. We then revised the nomenclature to emphasize differences via genetic etiology rather than relying on symptom-based descriptors that lack specificity due to overlap with other porphyrias. Given our lumping decision the terms “acute” and “intermittent” are no longer appropriate and were dropped. We created the following new parent term in MONDO: HMBS -related neurovisceral porphyria (MONDO:0700382). The recognized phenotypic subtypes of acute intermittent porphyria (MONDO:0008294, OMIM# 176000), acute intermittent porphyria nonerythroid variant (MONDO:0700384, OMIM# 176000), porphyria-related encephalopathy (MONDO:0958224, OMIM# 620704), and porphyria-related leukoencephalopathy (MONDO:0958226, OMIM# 620711) are child terms to the lumped entity ( Fig. 3 ).

ClinGen is a collaborative effort funded by the National Human Genome Research Institute to build a central genomic knowledge base that will improve patient care [ 24 ]. The global initiative aggregates genomic and health data around four types of curation activity: gene-disease validity, variant pathogenicity, dosage sensitivity, and clinical actionability. ClinGen is an ecosystem of working groups and expert panels, composed of individuals holding clinical, diagnostic and/or research expertise. The group that conducted the work in this publication is the General Inborn Errors of Metabolism Gene Curation Expert Panel (IEM GCEP) ( https://clinicalgenome.org/affiliation/40097/ ). The IEM GCEP classifies the clinical validity of genes in relation to IEM following ClinGen guidelines [ 26 ], and exists as a collaboration of members with varied backgrounds as specialists in the diagnosis, treatment, and/or research of IEM [ 27 ]. The ClinGen Curated Disease Entity Working Group is a collaboration of the Monarch Disease Ontology (MONDO), Online Inheritance in Man (OMIM), and ClinGen framework experts. Our team sought this working group’s assistance during curation of the heme synthetic pathway, given that we updated several porphyria monogenic disease entities. Gene-disease curations for the heme synthetic pathway were carried out according to the ClinGen Gene-Disease Validity Standard Operating Procedures (SOP v8 ( ALAS2 ), SOP v9 ( HMBS, UROD, CPOX, FECH ), SOP v10 ( ALAD, PPOX ), SOP v11 ( UROS )) [ 28 , 29 ]. This curation process is designed to aggregate up-to-date information about a monogenic gene-disease relationship into a standard framework. The classification that results from this work is published online with full transparency behind the decisions to score each piece of evidence. This evidence summary is organized so that it can be easily interpreted by doctors and clinicians. This process includes but is not limited to 1) lumping and splitting the disease and mode of inheritance assertions found in the literature for the gene-disease relationship, 2) collecting genetic and experimental evidence, 3) evaluating and storing the evidence, 4) putting the curation to review by an expert panel, and 5) publication of the final classification to clinicalgenome.org . ClinGen curations that do not reach a definitive classification are updated on a regular schedule. ClinGen guidance considers four distinct criteria for lumping and splitting gene-disease entities: assertions of gene-disease relationship, molecular mechanism, phenotypic variability, and inheritance pattern [ 30 ]. The first criterion pertains to whether multiple assertions for distinct gene-disease relationships have been made in the literature or by catalog authorities (including OMIM, Monarch Initiative, and Orphanet) for human disease. If multiple assertions of disease are present for a gene, then steps must be taken by the curator to determine whether these assertions deserve separate curations or should be lumped into a single gene-disease curation. A decision to retain a split or create a split (in situations where only one assertion has been made) is recommended when pre-curation evaluation uncovers two distinct molecular mechanisms underlie the two disease entities. Splitting is not recommended for subtypes of phenotypic features amongst affected individuals as following this pattern could result in ad infinitum curations, since all genomes are unique and may present with slightly different corresponding phenotypes [ 30 ]. The second criterion in determining whether to lump or split is molecular mechanism. Common differences in the molecular mechanism for disease-causing variants are gain-of-function or loss-of-function. A single gene may have more than one biological function which can lead to distinct phenotypes when a subclass of variants disrupts a particular region of the protein that is primarily responsible for only one of those functions. In other instances, variants may occur in specific transcripts when alternative splicing is involved. To justify a split based on molecular mechanism, distinct molecular mechanisms should be evident in both genetic and biochemical data. At the protein level, functional assays can be used to test for distinct mechanisms when variants are found to cluster in a particular functional domain of a protein. Molecular mechanism is the primary criteria when considering whether to lump or split, as this is the key aspect affecting subsequent variant classification [ 30 ]. The third criterion, phenotypic variability, describes the degree to which phenotypes vary across affected individuals. This variability encompasses pleiotropy, variable expressivity, and its quantitative measure of penetrance. In pleiotropy, phenotypic variation results from two distinct functions of a single gene product and causes variability among individuals harboring variants that affect the gene product differently. The protein may participate in two different biological pathways or have multiple distinct biological roles. Variable expressivity refers to varying degrees of severity of a genetic deficiency among individuals with very similar genotypes, such as individuals with the same genotype at the primary locus. To determine whether pleiotropy, which would favor splitting, or variable expressivity, which would favor lumping, is present, curators often analyze phenotypic features within a single family and then compare this to the phenotype set from unrelated individuals. If the variation amongst family members is as significant as the variation amongst unrelated individuals this suggests variable expressivity often due to additional genetic or environmental modifiers, and lumping is favored [ 30 ]. Penetrance quantifies the proportion of individuals with a given genotype who actually show manifestation of the associated phenotype. Penetrance is an additional important consideration for porphyrias, as monoallelic pathogenic variants in often do not manifest to clinical disease. Inheritance pattern is the final lumping and splitting criterion according to the ClinGen guidelines. Disease entities with autosomal dominant and autosomal recessive modes of inheritance are often split into different gene curations. However, gene curation groups have increasingly recognized that some cases are better suited for lumping into a single semidominant gene curation. This is especially true when autosomal dominant and autosomal recessive cases are found within the same families and caused by the same variants. In these situations, an autosomal recessive inheritance pattern for one disease entity and an autosomal dominant inheritance pattern for the other may be an indication that they share an underlying loss-of-function mechanism and represent two ends of a single disease spectrum. In such case, subsequent variant curation for diagnostic purposes is greatly simplified by lumping both into a semidominant mode of inheritance, which is recommended when biallelic variants cause a more severe form of the same disease. Whether to split or retain a split based on inheritance pattern should depend on whether there are clearly distinguishable phenotypic features or distinct clinical management, such as with the ATM gene, which is associated with susceptibility to breast cancer (autosomal dominant) as well as ataxia telangiectasia (autosomal recessive). If the disease exists on a continuum, where severity or age of onset correlates with zygosity, then evidence favors lumping which connects all deleterious variants to the same disease entity. This approach accurately reflects the underlying biology: at the protein level, pathogenic variants (no matter where they are) disrupt the same core function rather than falling into distinct subclasses that affect different molecular roles (one subclass impairing a protein’s role in transport and another impairing that protein’s unique role in signaling). This simplifies future variant curation (variants only need to be curated to a single disease entity) and simplifies variant interpretation in new patients (pre-existing variant information is not split between two disease entities that have no meaningful molecular distinction) [ 30 ]. Together this improves the capability of genetics to be used as first line diagnostics for identifying and treating porphyria. For this work, the decision to lump or split was based on balancing these criteria and reviewing them during regular expert panel calls. The most appropriate mode of inheritance (MOI) for the disease was determined based on the specifications given in the Gene-Disease Validity Standard Operating Procedures [ 28 ]. Genetic evidence was collected by searching HGNC, PubMed, and Google Scholar for gene and disease names. Key papers establishing the gene-disease relationship were identified and curated. An effort was made to collect as much clinical data as practical. Human Phenotype Ontology (HPO) terms were scored for every curation. Experimental evidence typically included enzyme activity assays, non-human model organisms, and a crystal structure of the human protein for the heme synthetic pathway genes. Genetic and experimental evidence is available for the respective gene curations on www.clinicalgenome.org , see Table 1 for direct links. All scoring of genetic and experimental evidence reflected guidelines of the Gene-Disease Validity Standard Operating Procedures and was reviewed and approved by the expert panel for each individual gene. Our scoring is visible under the summary for the respective gene curations on www.clinicalgenome.org , see Table 1 for direct links. Renaming of the porphyrias followed the explicit dyadic strategy that incorporates both the gene symbol and a phenotypically descriptive term, as outlined in [ 25 ]. The explicit dyadic approach is necessary due to advancements in molecular understanding of disease, driven by increased sequencing, which often reveals the need to reorganize some gene-disease relationships and subsequently assign new disease names and inheritance patterns. We worked with ClinGen’s Curated Disease Entity Working Group which includes representatives from MONDO and OMIM. MONDO provides a hierarchical ontology to standardize and harmonize disease definitions globally, enabling the aggregation of diseases into broader categorical groupings [ 31 ]. OMIM, maintained by Johns Hopkins University School of Medicine under the direction of Dr. Ada Hamosh, serves as a comprehensive catalog of human genes and genetic phenotypes, emphasizing genotype-phenotype relationships [ 32 ]. ClinGen links gene-disease relationships to MONDO terms leveraging its ontological structure to ensure hierarchical consistency. Specific to the porphyria pathway, we found the first part of the explicit dyadic strategy (choosing the gene symbol) to be relatively straightforward. However, choosing the phenotypic label was more challenging. Primarily, the goal of the phenotypic label is to convey semantic information that helps an individual recognize the disease. Another goal is to retain historical aspects of disease names [ 25 ]. In our case, although the historical names are well recognized in the clinical setting, the names were often too narrow based on our lumping decisions to be retained. Accordingly, any novel lumped terms we created were set by Monarch Disease Ontology (MONDO) as a parent term under which existing phenotype terms are maintained as child terms to avoid disrupting their continued clinical utility.

Decades of research have characterized the porphyrias caused by deleterious variants in the genes of the heme biosynthesis pathway. Clinically there are three major groupings: acute hepatic or neurovisceral porphyria, blistering cutaneous porphyria, and non-blistering cutaneous porphyria. All porphyrias caused by deficiencies in these eight of the nine enzymes exhibit variable phenotypic severity dependent on residual enzyme activity. Three neurovisceral porphyrias ( HMBS-, CPOX-, PPOX-related ) and one blistering cutaneous porphyria ( UROD-related ) have both an episodic adult-onset form of the porphyria typically caused by monoallelic loss-of-function variants, and a severe childhood-onset form of the porphyria caused by biallelic loss-of-function variants. While the two ends of these phenotypic spectrums are clinically described using separate terms based on severity, age of onset and differences in treatment, we found that the same loss-of-function mechanism at the protein level underlies both forms of these porphyrias. This led us to group evidence (monoallelic and biallelic cases) under a single gene-disease entity with a semidominant mode of inheritance for the purposes of gene and variant curation since variants can cause porphyria in both the monoallelic and biallelic state. This lumping required the creation of gene-first explicit dyadic names to serve as parent terms that encompass the full spectrum of the porphyria. These updates do not disrupt the use of the historical porphyria names in the clinic, which continue to exist in MONDO and are linked to the broader explicit dyadic names as child terms. However, since the existing clinical terms do not include gene names and in some cases contain unnecessary differentiating adjectives, the porphyria field may consider updates to simplified gene-based terms for these subtypes in the future. Only one gene in the heme synthesis pathway, ALAS2 , is associated with two distinct molecular mechanisms of disease. C-terminal gain-of-function variants in ALAS2 cause upregulation of the heme synthesis pathway and accumulation of protoporphyrin IX. Loss-of-function variants in ALAS2 cause X-linked sideroblastic anemia. Our expert panel retained this split and curated only the gain-of-function X-linked erythropoietic protoporphyria variants as evidence for ALAS2-related phototoxic protoporphyria. The remaining porphyrias resulting from heme synthesis pathway deficiencies are caused by biallelic loss-of-function variants in ALAD , UROS , and FECH . Following curation of each of these genes, the panel created gene-first explicit dyadic terms to serve as synonyms to the existing terms already in clinical use. Standardizing gene-first nomenclature for all heme synthesis pathway-related porphyrias offers clear advantages in the current era, where genetic testing is widely available, routinely recommended, and increasingly employed as a first-line diagnostic tool. Embedding the gene names directly into the nomenclature emphasizes the distinct genetic etiologies of these disorders. Eliminating nonstandard adjectives and instead using standardized phenotypic descriptors “neurovisceral” and “phototoxic” helps point out the shared clinical features. These updates make it easier to mentally group and track these related conditions. We feel this is particularly important given that porphyrias are rare diseases, patients often present first to non-experts, and neurovisceral porphyrias as well as EPP and XLP have historically been misdiagnosed. In conclusion, our update of heme synthesis pathway gene-disease entities and nomenclature paves the way for variant curation efforts and enhances clarity and consistency in communication among variant scientists, curators, genetic counselors, clinicians, and patients. This will support improved diagnostic precision in porphyria. Limitations of this study include that this project focused on a subset of the genes that cause porphyria, prioritizing those in the heme synthesis pathway. Additionally, this work does not include digenic or polygenic forms of porphyria that are not amenable to monogenic gene-disease curation. Modifier alleles for predisposition to porphyria will likely be most accurately curated as risk alleles [ 19 ]. Additionally, distinguishing true low-penetrance single variants from cases driven by cumulative variant burden looks to also be an important part of correctly being able to genetically describe these diseases.

Production of heme generates a prosthetic group necessary for proteins including hemoglobin, myoglobin, CYP450s, nitric oxide synthase, catalase, and peroxidase, which perform critical functions across a range of cell types in the human body [ 1 , 2 ]. Heme is multifunctional in the cell acting in chemical catalysis, signaling, transporter function and even in microRNA processing [ 3 – 6 ]. Formation of the metalloporphyrin heme requires eight steps of enzymatic catalysis. The first step and final three steps of heme synthesis take place in the mitochondria while the other four steps occur in the cytoplasm [ 2 ]. Although essential intermediates, porphyrins are also cytotoxic [ 7 – 9 ], and when heme synthesis is impaired the intermediates of the pathway can accumulate leading to toxicity and clinical porphyria [ 2 ]. Heme synthesis occurs to some degree in all cell types [ 10 ] but primarily in erythroid progenitor cells in the bone marrow, where heme is incorporated into hemoglobin to facilitate oxygen transport [ 11 ]. Another major location for heme synthesis is the liver, where production is needed for the activities of the CYP450 enzymes [ 11 ]. These two sites of heme origin, hepatic and erythropoietic, are also the primary locations of porphyrin precursor overproduction [ 11 ]. Clinically, the primary sites of toxicity are the neurological system and/or skin. Phenotypically porphyrias have been classified into three groups: (1) acute hepatic, (2) blistering cutaneous, and (3) non-blistering cutaneous [ 11 ]. Acute hepatic porphyrias present with sudden onset of a neurovisceral attack characterized by severe abdominal pain, peripheral motor neuropathy, hepatic dysfunction, and central nervous system involvement [ 2 , 11 ]. In contrast, cutaneous porphyrias are defined by either chronic blistering or acute non-blistering photosensitivity [ 2 , 11 ]. Now let us consider how these clinical phenotypes arise from the underlying biochemistry. Acute hepatic porphyrias result when a trigger (drugs, fasting, or steroid hormones) induces ALAS1 in the liver [ 11 ]. ALAS is the first and rate-limiting enzyme of the heme biosynthetic pathway, has two isoforms (ALAS1 which is ubiquitously expressed and ALAS2 which is erythroid specific), and catalyzes the condensation of glycine and succinyl-CoA to form 5-aminolevulinic acid (ALA) ( Fig. 1 ) [ 2 ]. In the context of an inherited deficiency of an enzyme more distal in the pathway, induction of ALAS1 leads to the accumulation of potentially neurotoxic ALA and porphobilinogen (PBG) [ 7 , 12 , 13 ]. Deficiencies of four of the enzymes in the heme synthesis pathway are associated with this clinical pattern of acute neurovisceral attacks: ALAD (step 2), HMBS (or porphobilinogen deaminase [ PBGD ]; step 3), CPOX (step 6), and PPOX (step 7) [ 11 ]. Defects in the early enzymes (step 2 and 3) result in the highest accumulations of ALA and PBG, consistent with their prominent neurovisceral manifestations. However, defects in the later enzymes (steps 6 and 7) also result in increased ALA and PBG ( Fig. 1 ). In the liver, ALAS1 is negatively regulated by the free heme pool. Drugs and hormones that induce hepatic CYP450 enzymes increase heme utilization, thereby depleting the free heme pool which relieves this negative regulation. Additionally, ALAS1 is directly induced by drugs and hormones through activation of nuclear transcription factors. PBG also inhibits ALAD in vitro [ 12 ], suggesting a feed-forward mechanism where even modest accumulation of PBG further suppresses ALAD activity, leading to ALA accumulation. Thus, the regulation architecture of the heme synthesis pathway has been shown to play a key role in why partial deficiencies of ALAD, HMBS, CPOX, and PPOX manifest as acute hepatic porphyria following a trigger [ 11 , 12 ]. Chronic blistering cutaneous porphyrias are clinically characterized by fragile skin, skin blistering and scarring, hyperpigmentation, hypertrichosis, alopecia and photosensitivity. Defects in uroporphyrinogen III synthase (UROS; step 4) or uroporphyrinogen decarboxylase (UROD; step 5) cause the most canonical chronic blistering photosensitivity. Impairment of UROS prevents the proper enzymatic cyclization of hydroxymethylbilane (HMB) to uroporphyrinogen III. In the absence of efficient UROS activity, HMB undergoes spontaneous, non-enzymatic cyclization to form uroporphyrinogen I, a dead-end, but photoreactive intermediate [ 14 ]. Similarly, defects in UROD lead to accumulation of fluorescent porphyrin intermediates. These fluorescent porphyrins absorb visible light and generate reactive oxygen species, resulting in the blistering characteristic of these cutaneous porphyrias [ 15 ]. Defects downstream of this like CPOX (step 6) and PPOX (step 7), which are primarily categorized as neurovisceral porphyrias, cause some accumulation of these photoreactive intermediates which may also lead to chronic blistering photosensitivity. Non-blistering cutaneous manifestations define the protoporphyrias, which result from accumulation of the final porphyrin precursor prior to heme formation. Two primary defects lead to this phenotype: loss of autoinhibition of ALAS2 (step 1, the rate-limiting step in heme production expressed exclusively in erythroid cells) due to C-terminal gain-of-function variants, which increases pathway flux, and deficiency of ferrochelatase (FECH; step 8), which impairs conversion of protoporphyrin IX to heme. In both cases, excess protoporphyrin IX accumulates in erythrocytes, plasma, and the skin. Unlike earlier porphyrins, protoporphyrin IX is poorly water-soluble and localizes to the dermis, where light exposure induces rapid phototoxic reactions that cause painful, non-blistering photosensitivity [ 15 , 16 ]. Diagnosing the four acute hepatic porphyrias (AHPs), has proven challenging, with an average delay of 15 years from symptom onset to accurate diagnosis [ 16 , 17 ]. This prolonged diagnostic timeline stems from several factors. First the symptoms (abdominal pain) are episodic (with patients often experiencing only a few attacks over a lifetime) and nonspecific, overlapping with far more common conditions like irritable bowel syndrome, appendicitis, or endometriosis. Second, diagnostic guidelines for presenting complaints like abdominal pain seldom mention porphyria, and the porphyrias’ rarity contributes to a low index of suspicion among providers. Additionally, biochemical testing is confusing due to a lack of harmonization of appropriate panels across laboratories and varying test menus. One common error is ordering only plasma or urinary porphyrins instead of the appropriate porphyrin precursor test, urinary ALA and PBG [ 18 ]. As a result, many patients with AHP undergo multiple unnecessary procedures (appendectomies, cholecystectomies, or exploratory surgeries) during this diagnostic odyssey [ 18 ]. Genetic testing as a first-line approach for patients with nonspecific symptoms, such as severe recurrent abdominal pain, offers a promising strategy to overcome the longstanding diagnostic delays in porphyrias. For example, identification of a pathogenic variant or even a rare VUS variant in HMBS given a presentation of recurrent abdominal pain would prompt immediate referral to a porphyria specialist who would be better equipped to order the appropriate biochemical test and interpret the biochemical results in relation to the patient’s molecular genotype and clinical presentation. Additionally, detection of risk alleles through carrier screening or broader “curiosity” genomic testing would enable at-risk individuals to avoid known precipitating factors (certain medications or hormonal therapies) and facilitate rapid recognition and management of acute attacks should they occur. Realizing this potential, however, requires important foundational work: clear gene–disease relationships and high quality variant curation. ClinGen, an NIH-funded program, serves as the authoritative central resource for defining the clinical relevance of genes and variants to support precision medicine. Its gene-disease validity curation process systematically evaluates the strength of evidence linking variation in a specific gene to a monogenic disease. Once gene-disease relationships are clear, variant curation across clinical testing laboratories can become more standardized. Currently, clinical diagnostic laboratories already apply the ACMG/AMP framework to classify variants as Pathogenic, Likely Pathogenic, Uncertain Significance, Likely Benign, or Benign. It will be very valuable to have a ClinGen Variant Curation Expert Panel (VCEP) add gene and disease specific modifications to these criteria for greater standardization and expert consensus on variants. Expert classification will be especially important for the porphyrias, which involve low-penetrance alleles that will benefit from assessment as risk alleles [ 19 ]. Additionally, for the porphyrias, distinguishing true low-penetrance single variants from cases driven by cumulative variant burden looks to be an important discussion. Using genetics as an initial diagnostic tool may help narrow down the differential conditions and thereby reduce unnecessary testing or medical procedures. While this is a long-term goal rather than an immediate reality, this represents the most achievable way to improve porphyria diagnostics, as relying on training non-experts to recognize these rare diseases and all the biochemical testing nuances is a more challenging effort. During our gene-disease validity curation for the heme synthesis pathway, we identified problems with porphyria gene-disease relationships and nomenclature that are hurdles to genetic diagnostics: (1) The asserted phenotypic subtypes, such as acute intermittent porphyria (AIP) and homozygous dominant AIP (HD-AIP), are clinically distinct in terms of severity and age of onset. However, they share overlapping sets of causative variants, indicating that they do not arise from distinct molecular mechanisms. Here, we define “molecular mechanism” as the primary effect on the cellular function of a protein. When deleterious variants affect a protein that performs only one main function, these variants typically produce a consistent effect which most commonly is loss-of-function. (2) While disease subtyping is necessary for treatment, having two terms associated with one molecular (i.e. loss-of-function) mechanism hinders attribution of pathogenicity to individual variants. (3) Existing porphyria names obscure shared features of acute hepatic or neurovisceral porphyria. (4) The AHP term which typically refers to the four deficiencies of ALAD (step 2), HMBS (step 3), CPOX; (step 6), and PPOX; (step 7) lacks genetic precision and currently is the exact same term OMIM ascribes for the porphyria due to pathogenic variants in ALAD ( Table 1 ). Each problem is described in more detail in the respective paragraph below. Multiple disease entities have been reported in association with a single heme synthetic pathway gene. Specifically for genes HMBS (or porphobilinogen deaminase [ PBGD ]), UROD , CPOX , and PPOX there are at least two distinct porphyria entities reported per gene ( Table 1 and Fig. 3 ). In our curations of these genes, we found no variant subtype distinctions to warrant this subtyping on a molecular level, although clinically there are differences between disease subtypes in severity and age of onset. However, the same variants were described in both autosomal and autosomal recessive disease subtypes (see HMBS and CPOX for specific variant examples). We also identified intermediate cases which do not clearly fall into disease subtype bins. For example, an individual with biallelic hypomorphic UROD variants exhibited a phenotype resembling a heterozygote with a null allele but was binned to the porphyria entity corresponding to severe early onset (see cases 1–3 in Table 3 ) [ 20 ] [ 21 ] [ 22 ]. A more common example of how current porphyria subtyping is ineffective involves heterozygous pathogenic HMBS variants, where penetrance according to population databases is ~1% (with no clinical evaluation of carriers) or 20–30% in clinically evaluated families [ 22 – 24 ]. These low measures of penetrance occur because a causative variant alone is insufficient, but must be coupled with environmental triggers (e.g., drugs, fasting) to induce an acute attack. The existing porphyria terms would diagnose these heterozygous individuals with autosomal dominant AIP. We corrected this oversimplification by lumping non-molecularly distinct disease subtypes and using the inheritance pattern of semidominant, which indicates that disease severity scales with allele dosage. The creation of the lumped term as a parent term allows for more inclusive variant interpretation requirements and continued diagnostic use of these subtype terms in the clinic ( Fig. 3 ). The lack of molecular distinctions among the multiple porphyria entities would have necessitated redundant variant curation, as identical variants could cause both porphyria subtypes depending on whether they are present in the biallelic or monoallelic state. This redundancy complicates variant classification, a crucial process for improving porphyria diagnostics and management, by dividing the biallelic and monogenic case evidence for the same variant into separate curations. To address these issues, as described below, we applied ClinGen’s framework for lumping and splitting gene-disease relationships for the purpose of establishing the most appropriate monogenic disease entities [ 23 – 25 ]. Porphyria nomenclature is based on phenotypic descriptive terms rather than genetic etiology. Disease-causing variants in any enzyme of the heme synthetic pathway disturb the pathway beyond the immediately preceding step, causing overlapping intermediate excretion patterns across the various enzyme deficiencies ( Fig. 1 ) [ 16 ]. This results in highly similar clinical symptoms especially for the acute hepatic or neurovisceral porphyrias caused by variants in ALAD, HMBS, CPOX , and PPOX where the potentially neurotoxic ALA metabolite accumulates ( Fig. 1 ). However, the porphyria descriptors, “acute hepatic porphyria”, “acute intermittent porphyria”, “coproporphyria”, and “variegate porphyria”, have unnecessary name differences, where the differentiating adjectives or names obscure clinical similarities and do not highlight the underlying genetic distinctions ( Table 1 ). The umbrella term acute hepatic porphyria (AHP) emerged to unify the neurovisceral porphyrias, emphasizing their similar clinical presentation [ 17 ]. While this term has improved recognition of these porphyrias among providers, its lack of molecular specificity renders it unsuitable for genetic reports, where gene-level precision is mandatory. The term “acute hepatic porphyria” also overlaps with OMIM’s name for the specific porphyria subtype due to pathogenic variants in ALAD (OMIM #612740) 10 This dual usage of “acute hepatic porphyria” as both a specific molecular diagnosis and a category label for a subset of porphyrias introduces significant risk of miscommunication. A patient diagnosed with “acute hepatic porphyria” clinically without molecular confirmation, may be interpreted as carrying a pathogenic variant in ALAD. Such ambiguity can lead to confusion of providers and patients. During our curation, we found individuals referred to as having acute hepatic porphyria, but because the term has double meaning, it was unclear whether they had variants in HMBS, PPOX, CPOX , or ALAD . It is essential for gene-disease validity and variant pathogenicity curation that each monogenic disease entity has a unique name to identify and distinguish it. This name and its description should convey accurate information about the disease, aiding in the recognition of symptoms and supporting quick diagnosis and treatment. For these reasons, ClinGen gene curation expert panels (GCEPs) have increasingly adopted an explicit dyadic strategy for naming disease entities, which was developed in collaboration with representatives from Monarch Disease Ontology (MONDO) and Online Inheritance in Man (OMIM) [ 25 ]. This explicit dyadic approach involves including the gene and a phenotypically distinguishing term, meaning the name should incorporate the gene label and the associated phenotype. Accordingly, we revised the MONDO nomenclature for monogenic heme-synthesis disease entities using an explicit dyadic nomenclature to increase specificity and clarity and resolve the aforementioned problems with the historical disease names. Our work provides justification for additional interest groups to update the nomenclature around the porphyrias.