Camurati-Engelmann Disease

Progressive diaphyseal dysplasia related to TGFB1 (type 1) and TGFB2 (type 2), including Ribbing disease

Literature and regulatory information update: 1 November 2026.

0. Initial Reference Table

1. Introduction

1.1. Definition

Camurati-Engelmann disease is an autosomal dominant bone dysplasia in which the cortex of the diaphyses of the long bones thickens progressively, by bone apposition on the periosteal and endosteal surfaces. Hyperostosis usually begins in the diaphyses of the femurs and tibias and extends to the fibulas, humeri, ulnas and radii; with progression it may involve the metaphyses, and the skull base is affected in a proportion of patients [3, 12]. The clinical picture is not limited to bone: limb pain, proximal muscle weakness, reduced muscle mass and adipose tissue, fatigability and waddling gait are the complaints that bring patients to the clinic, and anaemia, leucopenia and hepatosplenomegaly occur occasionally [2, 3]. Orphanet defines the entity (ORPHA:1328) as a clinically variable bone dysplasia with hyperostosis of the long bones, skull, spine and pelvis, associated with severe pain in the extremities, broad-based waddling gait, joint contractures, muscle weakness and easy fatigability [11].

Two genetic entities are distinguished. Camurati-Engelmann disease type 1 (CAEND1, OMIM 131300) results from heterozygous TGFB1 variants, which encodes transforming growth factor beta 1 (TGF-β1), and encompasses Ribbing disease, today considered a phenotypic variation of the same disease [2, 3]. Camurati-Engelmann disease type 2 (CAEND2, OMIM 606631) results from heterozygous TGFB2 variants and has a radiological pattern of its own, with striations and epiphyseal sclerosis [6, 9]. The other sclerosing dysplasias, including those related to SOST, LRP4, LRP5, CTSK, ANKH and AMER1, and the osteopetroses enter the differential diagnosis, summarised in section 6.

1.2. Original descriptions

Cockayne described a probable case in 1920, before the publications that gave the disease its name, and at the time the possibility of syphilitic osteitis was discussed [3, 12]. In 1922, Camurati, in Bologna, described a “hereditary symmetrical osteitis” of the lower limbs in a father, a son and other relatives, in a total of four generations, with pain and fusiform swelling of the legs below the knees, and was the first to suggest the hereditary nature of the disease [3, 12]. In 1929, Engelmann, in Vienna, published an isolated case with muscle atrophy and marked bone involvement, which he called “osteopathia hyperostotica (sclerotisans) multiplex infantilis” [1, 3]. The terms “Engelmann disease” and “diaphyseal dysplasia” were used until Neuhauser and colleagues coined, in 1948, the designation “progressive diaphyseal dysplasia”, which highlights the progression of the hyperostosis and the constant involvement of the diaphyses; Gulledge and White proposed in 1951 “progressive diaphyseal hyperostosis”, which did not come into general use [2, 12]. The eponym “Camurati-Engelmann disease” is today the most widely used [12].

In 1961, Lennon and colleagues reviewed the literature and recorded that onset usually occurred before 30 years and often before 10, in all ethnic groups and in both sexes, and that muscle weakness frequently led to the mistaken diagnoses of muscular dystrophy or poliomyelitis [3]. Ribbing described in 1949 a family in which four of six siblings had diaphyseal sclerosis limited to one to four long bones, always with tibial involvement, and named the picture “hereditary multiple diaphyseal sclerosis” [4].

1.3. Molecular genetics milestones

In 2000, Ghadami and colleagues studied two Japanese families with 27 members, 16 of them affected, and obtained a maximum lod score of 7.41 at marker D19S918, with a shared haplotype in a 15.1 cM interval at 19q13.1-q13.3; Janssens and colleagues and Vaughn and colleagues confirmed the linkage, and the second group narrowed the interval to 3.2 cM at 19q13.2 [3]. Because TGFB1 lies within that interval, Kinoshita and colleagues sequenced the gene in seven Japanese families and two of European origin and identified three heterozygous missense variants in exon 4, near the carboxy-terminal end of the latency-associated peptide (LAP), in all nine families; this work was published in September 2000, and that of Janssens and colleagues, with variants in the same domain, in November of that year [3, 16, 17]. In 2001 and 2003, functional studies showed that the variants increase TGF-β1 signalling, through facilitated release of the mature factor or through intracellular accumulation of abnormal protein [18, 19].

In 2002, Nishimura and colleagues described two Japanese girls, aged 17 and 11 years, with the clinical picture of the disease and radiological changes that included striations and metaphyseal widening, without TGFB1 or TGFBR1 variants, and proposed the designation “Camurati-Engelmann disease type II” [5]. In 2024, Wang and colleagues identified by exome sequencing de novo TGFB2 variants in three patients, two of them the girls from 2002, and in 2025 Mumm and colleagues described the first family with autosomal dominant transmission, of Peruvian origin [9, 10]. OMIM updated entry 606631 in September 2025 with these data [6].

1.4. Nomenclature and classification

The 2023 revision of the Nosology of Genetic Skeletal Disorders, with 771 entities and 552 genes, adopted dyadic nomenclature, which links each phenotype to the gene, and designates the disease “TGFB1-related Camurati-Engelmann diaphyseal dysplasia” [2, 20]. In that revision, the sclerosing disorders are divided into group 24 (osteopetrosis and related osteoclastic disorders), due to defective resorption, and group 25 (osteosclerotic disorders), due to excessive bone formation or disturbed remodelling, which includes Camurati-Engelmann disease [21]. The Oxford Desk Reference, following the earlier nomenclature, places it among the disorders with increased bone density and diaphyseal involvement [22]. The atlas of Spranger and colleagues includes Ribbing disease, also called Ribbing disease II, as a mild, late and somewhat asymmetrical form of the same entity; the designation “Ribbing disease I” was used for a multiple epiphyseal dysplasia with flat epiphyses, unrelated to this disease [1].

2. Molecular Genetics and Pathophysiology

2.1. The TGFB1 gene and the protein

TGFB1 lies at 19q13.2 (GRCh38 coordinates 19:41,330,323-41,353,922), has seven exons with long introns and encodes a precursor of 391 amino acids, of which the carboxy-terminal 112 form the mature protein; an arginine-arginine dipeptide precedes the furin cleavage site [7]. The pre-pro-peptide has a signal peptide of 29 amino acids and is cleaved after amino acid 278 into LAP and mature TGF-β1; the LAPs dimerise through disulphide bonds at cysteines 223 and 225 [2]. The mature TGF-β1 homodimer remains non-covalently associated with the LAP homodimer, through residues Ile53-Leu59, in what is called the small latent complex, and the LAP conceals the binding sites for the type II receptor [2]. Most cells secrete a large latent complex, in which cysteine 33 of the LAP binds covalently to latent TGF-β binding protein (LTBP), which facilitates folding, secretion and targeting to the extracellular matrix; activation of the large complex depends on binding of the amino-terminal domain of LTBP to the matrix [2]. The crystal structure of pro-TGF-β1 shows a ring complex in which a region termed the “straitjacket” surrounds each monomer of the factor and has to be loosened to release it [7]. TGF-β1, TGF-β2 and TGF-β3 signal through the same receptors and through the SMAD proteins, and TGF-β1 is highly expressed in developing cartilage and in endochondral and membranous bone [1, 7].

2.2. Mechanism of type 1 disease

GeneReviews classifies the mechanism as gain of function [2]. Most variants substitute an amino acid at the carboxy-terminal end of the LAP, next to the disulphide bonds between the two LAPs, disturb dimerisation and binding to mature TGF-β1 and increase the release of active TGF-β1 [2]. Saito and colleagues showed, in patient fibroblasts and in insect cells, that the variants p.(Arg218His), p.(Arg218Cys) and p.(Cys225Arg) disrupt the association between the LAP and TGF-β1 and release the mature factor; growth of the patients' fibroblasts was suppressed through TGF-β1, and the suppression was attenuated by neutralising antibodies or by dexamethasone, while proliferation of MG-63 osteoblasts in co-culture accelerated [18]. Janssens and colleagues tested five variants in a TGF-β-dependent luciferase assay and all increased the activity: p.(Arg218Cys), p.(His222Asp) and p.(Cys225Arg) through increased active TGF-β1 in the medium, and the leucine duplication of the signal peptide and p.(Tyr81His) through a different mechanism, with reduced secretion but intracellular accumulation capable of triggering the transcriptional response [19]. Aberrant activation of Rho GTPases has been proposed as part of the effect on bone remodelling [2].

The 2006 review by Janssens and colleagues distinguishes two functional groups: the exon 4 variants destabilise the disulphide bonds and prematurely activate the mature factor, and those of exon 1 alter secretion and retain the protein inside the cell [12]. Narrowing of the medullary canal on the endosteal side and the modelling defect on the periosteal side indicate that osteoclastic resorption and osteoblastic formation are both disturbed, which is consistent with the role of TGF-β1 in stimulating formation and suppressing resorption [12]. The finding of raised type I collagen C-terminal telopeptide (β-CTX) in 91% of patients in a Chinese series and the normalisation of formation and resorption markers under bisphosphonates point in the same direction, towards accelerated remodelling [14, 23].

Mumm and colleagues describe the same variants as “loss of function” defects of the LAP that increase the skeletal action of TGF-β1; this is the same final effect described at another point in the chain, because the LAP loses the ability to keep the factor latent and TGF-β1 gains activity [2, 10].

The reduction in muscle mass (39% of patients) and adipose tissue (21%) does not appear to be secondary to the skeletal involvement: TGF-β1 inhibits myoblast fusion and the expression of myogenic transcription factors, and also inhibits adipogenesis, while dexamethasone reverses the TGF-β inhibition of preadipocyte differentiation [12]. In murine models of fibrillin 1 deficiency and in the mdx mouse, increased TGF-β activity prevents muscle regeneration, and the neutralising antibody or losartan normalise muscle architecture and function [7]. To explain the restriction to the skeleton of a variant in a factor with ubiquitous receptors, Janssens and colleagues proposed that bone cells mainly produce the small latent complex, without LTBP, in which the variant can alter the conformation of the LAP; in the large complex of other tissues, the change would be neutralised [12].

2.3. Spectrum of TGFB1 variants

In 46 families studied up to 2006, 10 different variants were identified in all families but one; seven were missense in exon 4, in the region surrounding the homodimerisation residues Cys223 and Cys225, and accounted for 82.2% of all described variants; residue Arg218 was a hotspot, with 60% of the variants; outside exon 4 there was a 9 base pair duplication in exon 1, which encodes the signal peptide, and two missense variants in exons 1 and 2, at the amino-terminal end of the LAP [12]. GeneReviews attributes about 40% of patients to p.(Arg218Cys) and about 35% to p.(Arg218His) [2]. In 2001, p.(Arg218Cys) was the most frequent variant worldwide, present in 17 of 28 published families [24]. Haplotype analysis in 13 unrelated patients showed at least nine independent mutational events and three sites of clustering, at positions 218, 223 and 225 [7]. The most frequently cited variants, normalised in VariantValidator for the MANE Select transcript NM_000660.7 and for the protein NP_000651.3, are as follows.

There is an inconsistency of notation in a Brazilian report of 2024, which describes “c.653G>A, p.(Arg218Cys)” with the GRCh37 coordinate chr19:41,848,135; in VariantValidator, that coordinate (NC_000019.9:g.41848135G>A) corresponds to c.652C>T p.(Arg218Cys), and c.653G>A corresponds to p.(Arg218His); the coordinate and the protein therefore indicate p.(Arg218Cys) [8, 27]. No large intragenic deletions or duplications have been described, and the gain-of-function mechanism makes deletion testing unnecessary [2].

2.4. TGFB2 and type 2 disease

TGFB2 lies at 1q41 and encodes TGF-β2, also synthesised as a pro-protein with LAP [6]. Wang and colleagues identified, in three patients from three lineages, two heterozygous de novo TGFB2 variants, an in-frame deletion of 9 base pairs and the missense p.(Arg36Gly), both in the “straitjacket” region of the LAP; structural modelling predicted conformational changes with reduced inactivation of TGF-β2, an activity assay confirmed increased TGF-β2/SMAD signalling, and induced pluripotent stem cells from a patient differentiated into osteoblasts ossified more than those of controls [9]. The proposed mechanism is increased activation of TGF-β2 through loss of function of the LAP, and the difference in the pattern of hyperostosis between types 1 and 2 suggests distinct functions of the two ligands in the skeleton [9]. In the Peruvian family, Mumm and colleagues identified NM_003238.6:c.108G>T p.(Arg36Ser) (NC_000001.11:g.218346809G>T in VariantValidator), absent from gnomAD, at a conserved residue homologous to residue 36 of TGFB1, with autosomal dominant cosegregation and classification as likely pathogenic by the criteria of the American College of Medical Genetics and Genomics (ACMG) [8, 10]. The c. notations of the variants of Wang and colleagues were not confirmed in the sources consulted, and the ClinVar entries come from OMIM, without assertion criteria [6].

2.5. LRP6 variant and residual heterogeneity

One family with the clinical and radiological picture of Camurati-Engelmann disease shared no haplotypes at the TGFB1 locus and had no variants in exons 1 to 7, and other patients without a variant suggested heterogeneity [2]. Janssens and colleagues noted that, in a substantial proportion of patients without a variant, there were indications of incorrect diagnosis, through atypical findings or another pattern of inheritance [12]. Pickering and colleagues described a young woman with severe lower limb pain and hyperostosis and sclerosis of the long bones with a rare missense LRP6 variant, whose pathogenic role remains to be assessed [28]. GeneReviews includes this “LRP6-related Camurati-Engelmann-like disease” in the differential diagnosis, with diaphyseal sclerosis of the metacarpals and cranial sclerosis sparing the calvarium [2].

2.6. Inheritance, penetrance and expressivity

Inheritance is autosomal dominant [2]. Penetrance is incomplete: some obligate heterozygotes have normal radiographs, and the exact value is not known [2]. In the review of 100 molecularly confirmed cases, 94% had the typical involvement of the long bones and 74% had symptoms; radiological penetrance was 5 in 9 (56%) with p.(Tyr81His) and 77 in 78 (99%) with the other variants (p < 0.02) [12]. In the family of Wallace and colleagues, the father of the most severely affected patient, an obligate carrier of p.(Arg218His), remained asymptomatic and without radiological hyperostosis in his ninth decade [29].

Expressivity varies widely between families with the same variant and within the same family, and there is no correlation between the nature of the variant and severity [2, 24]. Several authors have described anticipation: Saraiva, in a Portuguese family with 15 affected members in three generations, by age at onset; Campos and colleagues, in a Brazilian family; and the atlas of Spranger and colleagues reports a tendency to earlier manifestation in the younger generation [1, 3, 27]. Janssens and colleagues observed the tendency in at least seven families, but in five of them the asymptomatic parent was only diagnosed after the birth of a severely affected child, which creates the appearance of anticipation; in two other families there was improvement in the following generations, and the nature of the variants does not favour an expansion mechanism [12]. In 1997, Saraiva had proposed that severity depended on sex and on the paternal origin of the allele, through a possible dynamic mutation, a hypothesis that predated identification of the gene [30].

The modifiers have not been identified. The TGFB1 polymorphisms (C-509T, C-800T, T29C, G75C and a TAAA repeat in intron 1; T29C and G75C are historical numbering, = NM_000660.7:c.29C>T p.(Pro10Leu), rs1800470, and NM_000660.7:c.74G>C p.(Arg25Pro), rs1800471) did not explain the variability in two families studied [24, 29]. In a Chinese series of 14 patients, the T allele of rs1800470 (NM_000660.7:c.29C>T p.(Pro10Leu)) was associated with trends towards higher haemoglobin and calcium and lower erythrocyte sedimentation rate (ESR) and high-sensitivity C-reactive protein, without differences in age at onset or in manifestations [31]. In the New Zealand family studied by Hughes and colleagues, an arm paretic since a meningoencephalitis in childhood had far less disease than the contralateral one, which suggests the participation of mechanical or growth factors [32].

2.7. Allelic disorders

Biallelic loss-of-function TGFB1 variants cause a different autosomal recessive disease, with infantile-onset inflammatory bowel disease, immunodeficiency and encephalopathy (OMIM 618213), described in three children from two lineages with epilepsy, cerebral atrophy and posterior leukoencephalopathy; the mutant proteins had impaired secretion, function or stability of the TGF-β1-LAP complex [2, 33]. GeneGuidance records for this gene-disease relationship a ClinGen classification of “limited” (2024) and stresses that a heterozygous loss-of-function carrier does not have Camurati-Engelmann disease [34]. OMIM also lists TGFB1 as a modifier of the pulmonary disease of cystic fibrosis [7].

Heterozygous loss-of-function TGFB2 variants cause Loeys-Dietz syndrome type 4 (OMIM 614816), a heritable aortic disease with root aneurysm and dissection, mitral valve prolapse and mild systemic Marfan-like manifestations, with a generally less severe arterial phenotype and greater non-penetrance than types 1 to 3 [35, 36]. In aortic tissue, molecular loss of function coexists with increased TGF-β signalling [36]. Camurati-Engelmann disease type 2 and Loeys-Dietz syndrome type 4 are thus allelic disorders with opposing mechanisms at the level of the LAP; two of the three patients of Wang and colleagues had mitral valve prolapse with mild regurgitation and one had mild annuloaortic ectasia, whereas the Peruvian siblings had no valvular or aortic root disease on echocardiography [6].

3. Clinical Spectrum and Manifestations by System

3.1. Age at onset and course

The mean age at symptom onset in 306 published patients was 13.4 years, with a range from birth to 76 years [2]. OMIM and Janssens and colleagues place the onset usually in childhood and almost always before 30 years [3, 12]. The atlas of Spranger and colleagues reports usual manifestation after the second year of life and mild forms that only become manifest in young adults [1]. The more recent hospital series, which recruit symptomatic patients, describe lower ages at onset: median of 2.5 years in 18 patients from Beijing, median of 3.0 years in 14 patients from the same centre, 8.8 ± 5.5 years (4–14) in four Korean patients and 3 to 7 years in four Turkish children [13, 21, 23, 31]. These figures reflect the mode of recruitment and not the age at onset of all carriers.

In the severe forms, the disease becomes manifest in early childhood with delayed walking, clumsy, broad-based gait, poor weight gain, reduced muscle mass, weakness and pain, which may lead to marked disability; there is sometimes spontaneous improvement during or after adolescence, and in other patients the disease remains active in adulthood, with progressive cranial nerve compression [1]. Fracture healing, life expectancy and adult height are normal, according to the atlas [1]. Mildly affected patients may have no symptoms and be diagnosed incidentally [1].

The study by Hughes and colleagues, in a large New Zealand family with p.(Arg218His), changed the perception of the natural history: only one member had the classic onset in childhood, with leg pain and altered gait; eight reported leg pain in adolescence which, by the beginning of the third decade, had disappeared or persisted at a low level, and two of these eight later developed cranial nerve palsies [32]. The extent and activity of the disease in the long bones, measured by scintigraphy, decreased with age (p < 0.025); in the older members, the scan could be quiescent despite florid radiological changes, and sequential scans in two members showed less activity on the more recent examination [32]. The authors conclude that the natural history may be benign, that activity subsides in adults and that the severe forms with childhood onset or cranial nerve involvement may be in the minority [32]. The 45-year follow-up of a patient described at 28 years shows the other extreme: the disease progressed from the diaphyses to the metaphyses, the epiphyses, the articular surfaces of the femoral heads and acetabula, the spine and the hands, with progression in the skull as well [37].

3.2. Frequency of manifestations

Sources for the table: GeneReviews [2], Janssens and colleagues [12], Li and colleagues [23], Liang and colleagues [31], Peláez and colleagues [38] and Carlson and colleagues [39]. The series differ in recruitment (published literature, genotyped families, referral centres), and the percentages are not directly comparable. Orphanet reports hearing loss in fewer than 20% of patients, a value close to that of Janssens and colleagues and below the GeneReviews range [2, 11, 12].

3.3. Musculoskeletal system

Bone pain is constant, dull and most intense in the lower limbs; it worsens with physical activity, stress and cold, and many patients have intermittent episodes of severe, disabling pain, with swelling, erythema and warmth of the limbs [2]. Intensity ranges from mild, requiring no treatment, to severe, requiring opioids, and it may diminish in adulthood [2]. The widened diaphyses may be palpable and tender: 52% of patients report tenderness on palpation [2]. Weakness and reduced muscle mass predominate in the pelvic girdle and thighs, with difficulty rising from a seated position and a positive Gowers' sign in some children [2, 40]. The weakness is not necessarily progressive [3]. The gait is waddling and broad-based, and some patients have late onset of walking [2].

The habitus is slender, sometimes marfanoid, with a low body mass index [2, 12]. In the Korean series, the mean body mass index was −2.4 ± 2.5 standard deviations at presentation; in the Turkish series, three of four children had an asthenic habitus, with indices of −8.71, −3.55 and −1.8 standard deviations; in one patient of the series of Janssens and colleagues, at 15 years, weight was 27 kg, with no subcutaneous fat, muscle hypotrophy and walking limited to 20–50 metres [12, 13, 21]. In eight Korean patients, all those who presented with a motor disorder before 6 years had a body mass index below 18.4 kg/m² [40].

Joint contractures occur in 43% of patients, and lumbar hyperlordosis, kyphosis, scoliosis, coxa valga, dislocation of the radial head, genu valgum, hallux valgus and flat feet may arise [2]. In the Korean series of Kim and colleagues, one patient needed corrective osteotomy for genu valgum at 18 years and another surgical correction of a severe thoracolumbar scoliosis (T3–L3) at 35 years [13]. Bone mineral density measured at the hip and femoral neck is increased, but strength measured by microindentation was below normal in three siblings, and high-resolution peripheral quantitative computed tomography (HR-pQCT) in nine Chinese patients showed lower volumetric density, altered microarchitecture and lower estimated strength at the distal radius and tibia, compared with controls [2, 23]. In the 12 Beijing patients with densitometry, two had a Z-score above 2.5 at the femoral neck and five had a Z-score below −2.5 at the neck and/or lumbar spine [23]. The Turkish series describes osteopenia in early childhood and osteosclerosis only in late childhood [21]. GeneReviews records no increased susceptibility to fractures and reports that healing may be slow, whereas the atlas of Spranger and colleagues describes normal healing [1, 2]. One British patient had had fractures after falls at 2, 7, 15 and 21 years [25].

3.4. Skull, cranial nerves and nervous system

Cranial hyperostosis begins at the base of the anterior and middle fossae and often includes the frontal bone [2]. In the review of 306 patients from the Mayo Clinic and the literature, 173 (56.5%) had radiologically confirmed skull base hyperostosis, but fewer than a quarter had symptoms; the commonest manifestation was hearing loss, followed by visual disturbance, vertigo, facial weakness, symptomatic brainstem compression, facial hypoaesthesia and hyposmia [39]. Wallace and colleagues estimated cranial involvement in 61% of patients in the literature [29]. Cranial nerve deficits occur in 38% of patients, mainly hearing loss, visual disturbance and facial palsy [2].

Hearing loss may be conductive, through narrowing of the external auditory canal, bony invasion of the ossicles or narrowing of the oval and round windows, or sensorineural, through narrowing of the internal auditory canal and compression of the cochlear nerve or its vessels; it may also be a consequence of an attempt at facial nerve decompression [2]. Tinnitus and vertigo have been described [2]. Orbital involvement causes blurred vision, proptosis, papilloedema, epiphora, glaucoma, globe subluxation and retinal detachment [2, 13]. In 2026 two cases of visual impairment were published: a 29-year-old Chinese woman with bilateral proptosis, orbital apex crowding, narrowing of the optic canals and superior visual field defects, with a reduction of the mean retinal nerve fibre layer thickness to 66 µm after one year; and a 21-year-old British woman with presumed disease, bilateral choroidal osteomas and papilloedema with an opening pressure of 33 cm H₂O, in whom the visual loss resulted from intracranial hypertension and choroidal pathology [41, 42].

Hyperostosis of the calvarium may cause intracranial hypertension, chronic and sometimes severe headache and frontal bossing [2]. Chronic intracranial hypertension caused, in a 57-year-old patient, a bone defect with a transethmoidal meningoencephalocele, and the hyperostosis may recur after surgical decompression [2]. Clonus, sensory loss, dysarthria, dysphagia, cerebellar ataxia and incontinence have rarely been described [2].

3.5. Endocrine system and growth

Hypopituitarism has been described, with short stature, delayed puberty and hypocortisolism, attributed to a small sella turcica from sclerosis of the skull base, to pituitary compression from intracranial hypertension or to a direct effect of TGF-β1 on the pituitary; hypogonadism may also result from scarcity of adipose tissue or from a gonadal effect of TGF-β1 [2, 43]. In a 20-year-old Indian man, with height of −3.6 and weight of −4.3 standard deviations and a bone age of 13 years, there was secondary hypogonadism (luteinising hormone 0.1 mIU/ml, follicle-stimulating hormone 1.0 mIU/ml, testosterone 0.087 nmol/l), growth hormone deficiency (insulin-like growth factor 1 of 120 ng/ml; peak of 7 ng/ml on insulin-induced hypoglycaemia), secondary hypocortisolism (cortisol 105 nmol/l) and a partially empty sella on magnetic resonance imaging [43]. Puberty was delayed in 8 of 11 evaluable Chinese patients, and 4 of 16 had short stature [23]. Hypothyroidism is rare, of pituitary or autoimmune origin [2]. A Japanese patient with type 2 had menarche at 20 years and a Peruvian patient had premature ovarian insufficiency at 33 [6].

3.6. Haematological and systemic manifestations

Anaemia, attributed to narrowing of the medullary cavity, leucopenia, hepatosplenomegaly and raised ESR have been described as systemic manifestations since 1982, in a patient who also had Raynaud's phenomenon and periungual infarcts [2, 44]. The Oxford Desk Reference includes anaemia among the features that point towards the diagnosis [22]. Inflammatory markers were raised in more than 60% of the 14 patients of Liang and colleagues, with a median ESR of 1.40 times the upper limit of normal and median high-sensitivity C-reactive protein of 1.71 times, both negatively correlated with haemoglobin, calcium and creatinine [31]. In the Korean series, the mean ESR was 46.5 ± 22.2 mm/h [13]. Less common are atrophic skin, palmoplantar hyperhidrosis, delayed dental eruption and extensive caries [2]. An osteoblastoma of the distal radius and an enchondroma of the fifth finger have been described in patients with a clinical diagnosis without an identified TGFB1 variant [2].

3.7. Aorta and heart

The GeneReviews authors know of an adolescent with the disease who died from dissection of a dilated ascending aorta; it is not known whether aortic dilatation is a rare manifestation of the disease, and the authors know of no other cases, but, because the mechanism shares the increased TGF-β signalling of Marfan and Loeys-Dietz syndromes, they recommend that the incidence of aortic disease be reassessed, with no current recommendation for routine aortic assessment [2]. In type 2, two Japanese patients had mitral valve prolapse with mild regurgitation, and one of them mild annuloaortic ectasia [6].

3.8. Quality of life

In a series of 15 unselected patients from a centre in Barcelona, the physical component of the SF-12 questionnaire was 29.8 (standard deviation 7.5), against 50.0 (10.0) in the United States reference population (p < 0.0001), and the mental component was 48.8 (11.23), without a difference; fatigue was reported by all, limb pain by 93.3% and muscle pain by 86.7% [38].

3.9. Ribbing disease

According to Seeger and colleagues, up to 1996 13 cases of Ribbing disease had been described in the English-language literature: 10 were siblings from three families and 3 were isolated; eight had involvement of both tibias only and the rest combinations of tibia, femur, fibula and radius; 7 presented with pain and 3 with swelling followed by pain, and 3 were asymptomatic, detected in the study of siblings [4]. Seeger and colleagues argued that Ribbing disease was distinct: it appeared after puberty, was unilateral or asymmetrical and asynchronously bilateral, was limited to the long bones, had no gait or neurological abnormalities or anaemia, and histology showed only osteoblastic activity with progressive obstruction of the Haversian systems, whereas in Camurati-Engelmann disease there is osteoblastic and osteoclastic activity [4]. Makita and colleagues described a three-generation Japanese family with 12 patients, seven with the complete disease and five with segmental, asymmetrical diaphyseal sclerosis without symptoms, resembling Ribbing disease, and proposed that the two represented the same disease; the identification of TGFB1 variants in patients with the Ribbing phenotype confirmed this [2, 45]. On imaging, Ribbing disease may mimic a stress fracture, a chronic infection, a bone-forming tumour or a metabolic or endocrine disorder [4].

3.10. Type 2 disease

The two Japanese patients described by Nishimura and colleagues had a marfanoid habitus, waddling gait, weakness, severe leg pain, flexion contractures of the hips and knees, delayed sexual development and raised alkaline phosphatase and ESR; radiology showed diaphyseal cortical thickening, metaphyseal widening, coarse and thick trabeculae in the long and short tubular bones, striations in the spine, pelvis and long bones and cranial sclerosis limited to the petromastoid regions [5]. Reassessed at 30 and 38 years, the hyperostosis also involved the epiphyses; the elder had occasional joint pain, walked and climbed stairs unaided and had a healthy child at 37 years [6]. The third patient of Wang and colleagues, a girl of Moroccan Jewish origin, had a low-energy femoral fracture at 7 months and pain and weakness from 2.5 years, with dolichostenomelia, muscle atrophy and a normal skull [6]. In the Peruvian family, the disease began in early childhood with disabling pain of the hips and lower limbs and hyperostosis of the whole long bone, more severe in the men: the son was carried to school and walked with support, the father was in a wheelchair at 70 years, and the siblings, at 45 and 49 years, had proximal weakness and a Trendelenburg gait; radiographs showed mild thickening of the calvarium, maxillary and mandibular hyperostosis, enlarged ribs and clavicles and hyperostosis of the limbs and pelvis [6, 10]. The sister's pain had catamenial exacerbations and improved greatly during pregnancy [6].

3.11. Pregnancy

One patient who improved with corticosteroids had less pain and more strength during pregnancy, which allowed treatment to be stopped; scintigraphy a few hours after the second delivery showed less uptake than before the pregnancy and than six weeks after delivery [2]. The same improvement was described in the Peruvian patient with type 2 [6].

4. Clinical Recognition and Warning Signs

4.1. When to suspect

GeneReviews recommends suspecting the disease in the presence of proximal muscle weakness, easy fatigability, bone pain mainly in the lower limbs and waddling gait [2]. Orphanet adds the combination of proximal weakness with hyperostosis of one or more long bones on radiography [11]. In its approach to increased bone density, the Oxford Desk Reference places Camurati-Engelmann disease in the age groups from 2 to 12 years and from 12 years to adulthood and lists, as points of the history, muscle weakness and pain, visual and hearing problems, headache and fractures, and, as points of the examination, height, head circumference, the cranial nerves, fundoscopy for papilloedema and pubertal development, which may be compromised by pituitary compression [22].

In the young child, the presentation is mainly motor: delayed walking, waddling gait, difficulty running or climbing stairs and Gowers' sign, between 4 and 6 years in the Korean series, or from 24 months in a Turkish boy [26, 40]. In the adolescent and young adult, lower limb pain predominates, which may begin at 15, at 20 or even at 54 years [40]. In adults, the disease may present with hearing loss, visual disturbance, proptosis or headache with intracranial hypertension, and the Mayo Clinic recommends including it in the differential of any thickening of the skull base with cranial neuropathies or intracranial hypertension [39]. A tibial diaphyseal lesion may be found incidentally in an asymptomatic child [40].

4.2. Frequent diagnostic errors

Muscle weakness and gait lead to initial diagnoses of muscular dystrophy, polymyositis or poliomyelitis, and pain and inflammation to diagnoses of rheumatoid arthritis or sarcoidosis [2, 3]. In the Brazilian family of Campos and colleagues, the youngest son was first investigated for spinal muscular atrophy, and the TGFB1 variant emerged on a muscular dystrophy panel [27]. The Turkish boy of Alkaya and colleagues was treated as transient synovitis before radiography [26]. The British patient of Hassan and colleagues was diagnosed with juvenile Paget disease and had the molecular diagnosis ten years later [25]. In the family of Wallace and colleagues, the symptomatic relatives initially received several different diagnoses [29]. A plain radiograph of the long bones, in a patient with leg pain and proximal weakness, is the investigation that avoids these pathways [2, 46].

4.3. Warning and emergency signs

The following are warning signs: headache with vomiting, papilloedema or visual loss, which require exclusion of intracranial hypertension; progressive visual loss or visual field defects, which suggest optic nerve compression in the optic canal; facial palsy; progressive hearing loss; and signs of brainstem compression [2, 39, 41, 42]. In a child with intracranial hypertension refractory to medical treatment and papilloedema, hemicraniectomy was needed [47]. Severe acute pain, with swelling, erythema and warmth of a limb, usually corresponds to a flare of disease activity, assessed with ESR and scintigraphy [2]. There is no newborn screening for this disease.

5. Diagnosis: Criteria, Tests and Algorithms

5.1. Criteria

There are no published consensus clinical diagnostic criteria [2]. The clinical diagnosis is established in a proband with the characteristic radiological findings; if radiology is inconclusive, it is established by the identification of a heterozygous pathogenic or likely pathogenic TGFB1 variant [2]. Under the standards of the ACMG and the Association for Molecular Pathology (AMP), pathogenic and likely pathogenic variants have the same diagnostic value, and a TGFB1 variant of uncertain significance (VUS) neither establishes nor excludes the diagnosis [2, 48].

5.2. Radiology

In the long bones, hyperostosis begins in the diaphyses, may progress to the metaphyses and rarely involves the epiphyses; periosteal involvement produces irregular cortical thickening and increased diameter, and endosteal sclerosis narrows the medullary canal; involvement is usually symmetrical, but may be asymmetrical [2]. The atlas of Spranger and colleagues describes cortical thickening and patchy sclerosis of the diaphyses, sparing the metaphyses and epiphyses, with fusiform widening of the femurs and of the proximal two thirds of the tibias and fibulas, coarse trabeculation, minimal involvement of the short tubular bones and preservation of the distal ulna; the metaphyses, epiphyses and pelvis may be involved in older patients with advanced disease [1]. In the skull, hyperostosis begins at the base of the anterior and middle fossae; less often it involves the calvarium; there is mild sclerosis of the posterior vertebral arches, ribs, clavicles and the parts of the flat bones that correspond to diaphyses [1, 2]. Computed tomography (CT) of the head shows narrowing of the optic canals, orbital fissures and internal auditory canals and loss of the diploë [12, 13]. In the Turkish series, progression included Erlenmeyer flask deformity [21].

Bone scintigraphy shows increased uptake in the diaphyses and skull in 74% of the patients studied and may be positive before sclerosis is visible on radiography, which makes it useful at an early stage [12]. It has, however, limitations: in adults the scan may be quiescent despite marked radiological changes, and isolated uptake at the skull base is frequent in the normal population and is not a useful marker [12, 32]. The use of whole-body magnetic resonance imaging to identify the affected areas has been described [2, 49].

5.3. Laboratory

Bone biochemistry is often normal [22]. ESR and scintigraphy serve to assess disease activity during pain flares [2]. Alkaline phosphatase was raised in 41% (7/17) of the Beijing patients, and β-CTX in 91% (10/11) [23]. Formation and resorption markers (osteocalcin, procollagen type 1 N-terminal propeptide or P1NP, C-terminal telopeptide) were markedly raised in patients treated with bisphosphonates and normalised under treatment, which led to the proposal of using them as surrogate indices of activity [14]. ESR and C-reactive protein are raised in more than half of patients and fall with glucocorticoids [13, 31]. The full blood count assesses anaemia, leucopenia and thrombocytopenia [2, 22].

5.4. Molecular tests and their limitations

The options are TGFB1 sequencing alone, a multigene skeletal dysplasia panel including the genes of the differential diagnosis or, when the phenotype is atypical, exome or genome sequencing [2]. Sequencing detects a variant in more than 90% of probands, mainly missense in exon 4; no intragenic deletions or duplications have been described, and targeted deletion and duplication analysis of TGFB1 is not indicated [2]. This proportion refers to probands with an established clinical diagnosis and does not equate to the sensitivity of the test in practice, which depends on patient selection. A negative TGFB1 result does not exclude the disease: there is type 2 due to TGFB2, the variant described in LRP6 and families without a known cause, and a signal peptide duplication may not be recognised on first analysis [2, 9, 25, 28]. In the patient of Hassan and colleagues, a skeletal dysplasia panel by genome sequencing was initially reported as negative, and the duplication c.31_48dup was only recognised on reanalysis of the VUS with additional clinical information [25]. When the molecular cause is not found, GeneReviews recommends storing the proband's DNA [2].

The report should state the gene, the reference transcript with version (NM_000660.7 for TGFB1; NM_003238.6 for TGFB2), the variant in Human Genome Variation Society (HGVS) notation at the coding DNA and protein level, the zygosity and the ACMG/AMP classification [8, 48]. In simplex cases, study of the parents and, if necessary, confirmation of paternity and maternity allow the variant to be classified as de novo [2].

5.5. Assessment after diagnosis

Source for the table: GeneReviews [2].

5.6. Algorithm

The diagnostic pathway can be summarised in four steps. First, in the presence of lower limb pain with proximal weakness, fatigability or waddling gait, in a child, adolescent or young adult, request radiographs of the long bones and skull [2, 22]. Second, if there is bilateral periosteal and endosteal diaphyseal cortical thickening, with the metaphyses and epiphyses spared, the clinical diagnosis is established and is confirmed by TGFB1 sequencing; if radiology is atypical (striations, metaphyseal widening, early epiphyseal involvement, bowed bones, severe anaemia) or inconclusive, request a sclerosing dysplasia panel including TGFB1, TGFB2, SOST, LRP5, LRP6, TBXAS1, TNFRSF11B, COL1A1 and FAM111A, or the exome [2, 9]. Third, if TGFB1 and TGFB2 are negative with a typical phenotype, reanalyse the signal peptide regions and the VUS, consider the genome and store DNA [2, 25]. Fourth, complete the initial assessment of table 5.5 and study at-risk relatives [2].

5.7. Prenatal and preimplantation diagnosis

Once the familial variant has been identified, prenatal diagnosis and preimplantation genetic testing are possible, but, because of the reduced penetrance and the variability, the result does not predict the age at onset, the severity, the type of symptoms or the speed of progression [2, 12]. Fetal ultrasound has no described role in this disease; in the class of disorders with increased bone density, ultrasound generally detects changes only in the second or third trimester [22].

6. Differential Diagnosis

Few diseases share the clinical and radiological picture of Camurati-Engelmann disease, and the diagnosis is usually made by physical examination and radiological study of the skeleton [2, 11]. The differential has two levels: the other dysplasias with diaphyseal or cranial sclerosis, which are distinguished by radiology, age, extraskeletal signs and the gene; and the non-skeletal diseases with pain and proximal weakness, which are excluded with a radiograph.

AD, autosomal dominant; AR, autosomal recessive.

The differential diagnosis with craniometaphyseal dysplasia rests on metaphyseal widening, which develops in that disease but not in Camurati-Engelmann disease, and on the lesser cranial involvement of the latter [1, 54]. Craniodiaphyseal dysplasia is distinguished by the more massive cranial hyperostosis and by the absence of irregular cortical thickening of the long bones, and Worth endosteal hyperostosis, van Buchem disease and sclerosteosis by the absence of irregular periosteal thickening [1, 50]. Ghosal haematodiaphyseal dysplasia shares the pain, the walking difficulty and the diaphyseal sclerosis, and in some patients the medullary involvement, but runs with severe anaemia or pancytopenia [21, 51]. Osteoectasia with hyperphosphatasia has bowed bones and short stature, and the density of the tubular bones is not increased as in Camurati-Engelmann disease [52]. Infantile cortical hyperostosis begins in infancy, is asymmetrical, runs with fever and irritability and regresses, whereas Camurati-Engelmann disease appears after infancy, is generalised, more endosteal and associated with muscle weakness [53]. Pachydermoperiostosis becomes manifest at puberty, without weakness, and with less marked thickening of the mid-diaphyseal portions [56]. Hardcastle syndrome has no cranial sclerosis [55]. Orphanet also includes generalised cortical hyperostosis [11].

Ribbing disease should be considered in the presence of unilateral or asymmetrical bilateral tibial diaphyseal sclerosis in an adolescent or adult, and its differential diagnosis includes stress fracture, chronic infection and bone-forming tumour [4]. Type 2 disease should be considered when radiology shows striations and epiphyseal sclerosis, a pattern that brings it close to osteopathia striata with cranial sclerosis and to generalised hyperostosis with striations [5, 6]. In a young child, the initial osteopenia described in the Turkish series may delay the suspicion of a sclerosing disease [21].

7. Treatment and Risk Management Strategies

7.1. Principles and level of evidence

There is no curative treatment and no approved disease-modifying treatment, and no clinical practice guidelines have been published for Camurati-Engelmann disease [2, 11]. A search of ClinicalTrials.gov, in November 2026, found no registered studies for this disease [57]. The evidence on medicines is limited to case reports, small series and the review of 24 families by Janssens and colleagues, without randomised trials or comparisons between treatments [12, 34]. The aims of treatment are control of pain, preservation of strength and ambulation, prevention or treatment of the cranial complications and correction of the endocrine deficits. Management is multidisciplinary and includes physical and rehabilitation medicine, orthopaedics with experience in skeletal dysplasias, otorhinolaryngology, ophthalmology, neurosurgery, endocrinology and genetics [2]. None of the medicines used has an approved indication for this disease in any jurisdiction; all the described uses are off-label [34, 58].

7.2. Glucocorticoids

Glucocorticoids are the treatment with the greatest accumulated experience since the first descriptions by Royer and colleagues (1967), Allen and colleagues (1970) and Lindstrom (1974); Minford and colleagues (1981) recorded, besides pain relief, partial regression of the radiological changes [3]. Reports describe reduction of pain and weakness, improvement in gait, exercise tolerance and flexion contractures, correction of anaemia and hepatosplenomegaly and delay or prevention of hyperostosis, including that of the skull and nerve compression [2]. Wallace and colleagues argued for a trial of corticosteroid therapy in all symptomatic patients [29]. The mechanism is twofold: in bone, glucocorticoids increase apoptosis of osteoblasts and osteocytes, stimulate osteoclastogenesis and reduce intestinal calcium absorption, which counters excessive bone formation; and they divert TGF-β binding to receptors without signalling capacity and inhibit TGF-β-dependent transcription [12]. Because the improvement in pain can be very rapid, it probably does not depend on osteoblastic suppression alone [12]. Histology after corticosteroid therapy showed more resorption and secondary remodelling; sclerosis decreased on radiographs and scintigraphic uptake decreased in some patients, but did not regress in others [2].

Doses. GeneReviews, the reference for dosing in this disease, proposes: in patients with severe symptoms, a prednisolone bolus of 1.0–2.0 mg/kg/day followed by rapid tapering to the lowest tolerated alternate-day dose; in less symptomatic patients, starting with 0.5–1.0 mg/kg on alternate days; higher doses may help in acute pain flares, and some patients may stop treatment in quiescent phases [2]. Janssens and colleagues consider 1 mg/kg/day a good starting dose, which should be reduced in long-term treatment [12]. In comparison with the label, the Portuguese summary of product characteristics (SmPC) of Lepicortinolo® (prednisolone, 5 and 20 mg tablets) states usual paediatric doses of around 1 mg/kg/day and, in adults, of 10 to 100 mg/day, with no known maximum dose, with gradual tapering and the lowest effective dose in long treatments; Camurati-Engelmann disease is not among the indications [59]. The reports illustrate the variety of doses used: a child on prednisone 30 mg/day tapered to 10 mg on alternate days over a year improved in weight, appetite and gait, but treatment was stopped because of aggressive behaviour; a Tanzanian patient obtained relief with prednisolone 15 mg on alternate days for a month and tramadol 50 mg twice daily; an Indian patient improved partially with prednisolone 30 mg/day and developed Cushing syndrome [12, 46, 60].

Deflazacort. Deflazacort, a prednisolone derivative, was used at a dose of 1.2 mg/kg/day in a 5-year-old boy, with clinical and radiological improvement in 12 months without adverse effects, and at an initial dose of 1–1.2 mg/kg/day in four Korean patients aged 11 to 45 years, with relief of acute pain in 1 to 3 months and a fall in ESR [13, 61]. Janssens and colleagues consider it a possibly safer alternative, with comparable effect [12]. The Portuguese SmPC (Deflazacorte Alter, 6 and 30 mg) states for children 0.25 to 1.5 mg/kg/day and for adults up to 120 mg/day in acute disease and maintenance usually no higher than 18 mg/day, with a minimum effective dose of 3 mg/day [62]. The doses published in the disease lie within the paediatric range of the SmPC.

Adverse effects and monitoring. GeneReviews recommends monthly follow-up after initiation, tapering to the lowest tolerated dose to limit the risk of osteoporosis and vertebral compression fractures, and measurement of blood pressure at each visit, because of the risk of hypertension [2]. The disease does not appear to increase spinal density, so corticosteroid therapy may cause vertebral osteoporosis; annual densitometry is indicated in treated patients [2]. In the series of Janssens and colleagues, two patients from a family treated long term had vertebral osteoporosis, whereas two untreated patients from another family had increased vertebral density; two other patients became cushingoid, and prolonged corticosteroid therapy compromises linear growth in children [12]. In a Brazilian family, corticosteroid therapy caused weight gain, acanthosis and fluid retention [27]. In the Korean series, three of four patients had worsening pain or headache when withdrawal of the corticosteroid was attempted after two years [13].

7.3. Losartan

Losartan, an angiotensin II type 1 receptor antagonist, reduces TGF-β1 signalling and the expression of TGF-β receptors types 1 and 2; in the murine model of Marfan syndrome, it prevented aortic aneurysm, and in mice with fibrillin 1 deficiency and in the mdx mouse it normalised muscle regeneration [7, 63, 64]. Its use in Camurati-Engelmann disease is based on that rationale and on case reports [2].

Ayyavoo and colleagues treated a prepubertal 9-year-old girl with losartan 0.75 mg/kg/day, increased over 12 weeks to 1.0 mg/kg/day; at assessments at 9, 17 and 38 months, the severe, disabling pain had disappeared, walking capacity and physical activity had increased and lean and fat mass had improved, with normal full blood count and liver and renal function [65]. Simsek-Kiper and colleagues treated a 13-year-old boy with p.(Arg218Cys) with 50 mg/day for 18 months: pain intensity and frequency decreased in a pain diary and strength improved, allowing him to walk and climb stairs again, without evident adverse effects [63]. In the Korean series, losartan 0.8–1.2 mg/kg/day was used as maintenance in combination with deflazacort; after two years, one patient controlled the pain with losartan alone, and two had transient hypotension and dizziness that resolved with dose reduction; in an 11-year-old patient, scintigraphy after one year showed lower diaphyseal uptake [13]. In an 18-year-old Chinese woman with hypogonadotrophic hypogonadism, the combination of prednisone (doses between 3.75 and 30 mg/day over 28 months) with losartan 50 mg/day, calcium and vitamin D improved bone mass at the spine and hip and fat mass, initiated puberty and lowered the ESR from 64 to 26 mm/h [66]. In a 27-year-old Indian patient, refractory to prednisolone and alendronate, losartan allowed the corticosteroid to be reduced and lowered CTX from 1617 to 375 pg/ml and P1NP from 163 to 50 ng/ml [60]. There was also absence of response: in a French patient, losartan failed as had the conventional treatments, and in a Brazilian boy it did not relieve the pain [27, 67]. GeneReviews concludes that losartan may be a useful adjuvant to reduce the need for corticosteroids and that long-term studies of the benefits and risks are lacking [2].

Doses and comparison with the label. In the absence of a guideline dose, the reference doses are those published in the disease itself: 0.75 to 1.2 mg/kg/day in children and 50 mg/day in adolescents and adults [13, 63, 65, 66]. For comparison, the United States Food and Drug Administration (FDA) label for Cozaar® (losartan potassium; NDA 020386, initial approval on 14 April 1995, revision of 20 January 2026) states, in paediatric hypertension from 6 years of age, 0.7 mg/kg once daily up to 50 mg, with adjustment according to blood pressure; doses above 1.4 mg/kg/day or 100 mg/day have not been studied in children, and the medicine is not recommended before 6 years of age or with a glomerular filtration rate below 30 ml/min/1.73 m²; in adults, the usual dose is 50 mg/day, up to 100 mg [58]. The Portuguese SmPC of a generic (Losartan Alter 50 mg) states, from 6 to 18 years, 25 mg/day between 20 and 50 kg (exceptionally up to 50 mg) and 50 mg/day above 50 kg (exceptionally up to 100 mg) [68]. The doses used in the disease are within the limits studied in the label, but the indication is different and the use is off-label [34].

Adverse effects and precautions. According to the label, hypotension may occur, especially with volume depletion, deterioration of renal function and hyperkalaemia, so renal function and potassium are monitored; the commonest effects were dizziness (3% against 2% with placebo), upper respiratory tract infection, nasal congestion and back pain [58]. Baseline blood pressure should be measured before starting losartan and monitored at each visit; hypotension is managed with dose reduction and fall prevention [2]. Losartan carries a warning of fetal toxicity: it should be stopped as soon as a pregnancy is detected, it is contraindicated in the second and third trimesters and it is not recommended in the first trimester or during breastfeeding [58, 68].

7.4. Bisphosphonates

The role of bisphosphonates is controversial. GeneReviews places them among the treatments to be considered with caution: pamidronate did not improve symptoms in four patients, and clodronate increased pain in one patient and had no effect in another [2]. Janssens and colleagues describe weekly pamidronate for five weeks without improvement in one patient, three months of intravenous pamidronate without benefit in two patients and an increase in alkaline phosphatase with etidronate in another, and considered it unlikely that these antiresorptives had a future in the disease [12]. Castro and colleagues, in Campinas, obtained with bisphosphonates no improvement beyond that of corticosteroids in two women with severe disease [69]. Orphanet, in a 2013 synthesis, considers non-steroidal anti-inflammatory drugs and bisphosphonates ineffective, and Carlson and colleagues found no benefit of corticosteroids or bisphosphonates on skull base symptoms [11, 39].

Later reports describe favourable responses. Baroncelli and colleagues treated two patients from an Albanian family with p.(Glu169Lys): the 19-year-old female proband, who had had only partial relief with indometacin 100 mg twice daily and with prednisone 2 mg/kg/day for two months, received zoledronic acid 0.020 mg/kg in 50 ml of normal saline over 40 minutes, every 4 months, for a total of seven cycles over 2.2 years, and the pain fell from 9 to 1 on a scale of 0 to 10, with improvement in gait and bone lesions and a rise in the lumbar Z-score to 0.8; the 4-year-old male proband (13.1 kg; body mass index of 11.8, Z −4.6) did not improve with neridronate 1 mg/kg every 4 months, but improved with zoledronic acid 0.015–0.018 mg/kg every 4 to 6 months, with pain from 10 to 2 and weight gain to 17.7 kg, although the radiological hyperostosis persisted [14]. Each infusion was followed by intravenous calcium gluconate and oral calcium carbonate for 10 days, with vitamin D 1000 IU/day; both had an acute-phase reaction (mild fever, myalgia, arthralgia and worsening pain) with the first infusion, and the remodelling markers normalised [14]. A 20-year-old man, followed at a German centre, had immediate pain relief after a single dose of zoledronic acid [2, 70]. In the Turkish series, pamidronate relieved pain in two children with a Z-score below −2.5 and severe pain [21]. In an Indian boy with hypopituitarism, zoledronate was part of a combined treatment with losartan and prednisolone [43], and in a Chinese woman with compressive optic neuropathy zoledronate combined with methotrexate and hormone therapy improved pain and markers after six months [41]. In type 2 disease, aminobisphosphonates appeared useful [10].

In 2025, Hassan and colleagues reported a discordant response in a British man: oral alendronate led to almost complete resolution of the musculoskeletal symptoms in six weeks, with a partial fall in CTX from 1.7 to 1.03 µg/l at nine months, whereas intravenous zoledronic acid worsened the pain and hearing loss for two months; the return to alendronate again improved the symptoms within 12 weeks [25]. The authors speculate that rapid remodelling alters the skeletal retention of the medicine and suggest monitoring the remodelling markers [25]. The alendronate dose was not published; the FDA label for Fosamax Plus D® states, in adult osteoporosis, 70 mg once weekly, taken fasting with water at least 30 minutes before any food, without lying down for the following 30 minutes, and the medicine is not indicated in paediatric age; in a trial in children with severe osteogenesis imperfecta, 5 mg/day below 40 kg and 10 mg/day from 40 kg were used, without a reduction in fractures, with no maximum paediatric dose defined in the label [71]. This dose is an extrapolation from another indication.

The increased bone mineral density in the long bones led some clinicians not to start bisphosphonates, but density may be reduced at the spine and at trabecular sites, especially in patients treated with corticosteroids [2, 23, 25]. The available data do not allow the patients who respond to be identified, nor bisphosphonates to be compared with corticosteroids or with losartan.

7.5. Other medicines

Non-steroidal anti-inflammatory drugs relieve pain, but do not alter the bone lesions [12]. Intranasal calcitonin relieved pain in one patient, but in another it had no effect and was stopped [2, 12]. Infliximab, an anti-tumour necrosis factor alpha antibody, relieved the pain of a corticosteroid-dependent Portuguese patient from Coimbra, treated for ulcerative colitis; it is the only report with this medicine [2, 72]. Methotrexate was used in combination with zoledronate in a patient with orbital involvement [41]. Janssens and colleagues listed, as theoretical possibilities for sequestering TGF-β1, decorin, biglycan, α2-macroglobulin, soluble betaglycan, fetuin, anti-TGF-β antibodies and the soluble type II receptor, and warned of the risk of systemic effects, preferring local application to bone and muscle; none of these approaches has been tested in patients [12].

7.6. Treatment of the manifestations

Sources for the table: GeneReviews [2], Janssens and colleagues [12], Kim and colleagues [13] and Lei and colleagues [41].

7.7. Surgery

Reaming of the medullary canal and osteotomy relieved persistent bone pain and deformities in selected patients [1, 2, 12]. Because the disease is progressive, symptoms may return [12]. At the skull base, Carlson and colleagues consider wide surgical decompression the only effective option in patients with progressive symptoms, because corticosteroids and bisphosphonates do not improve these symptoms; surgery is technically more difficult because of the thick, sclerotic bone, the loss of anatomical landmarks and the reduced supratentorial space, and patients should be informed of the increased risk and of the possibility of new bone deposition with recurrence [39]. Decompression of the internal auditory canal by the retrosigmoid approach and facial nerve decompression have been described in isolated patients, but facial nerve decompression may cause sensorineural hearing loss [2]. Craniectomy relieved intracranial hypertension and headache, and the recurrence was treated with radical craniectomy and cranioplasty with titanium mesh [2]. In 2024 the case was published of a child with refractory intracranial hypertension and papilloedema treated with unilateral hemicraniectomy and titanium cranioplasty, with long-term resolution of the symptoms [47]. Orbital decompression was used to free the optic nerve in one patient [12], and was considered in a patient in 2026 with progression of the visual field defects [41].

In hearing, bilateral myringotomy improves conductive hearing loss due to serous otitis; a cochlear implant improved hearing in a patient with conductive hearing loss and patent internal auditory canals, but narrowing of the internal auditory canal and the absence of a functioning cochlear nerve, both possible in this disease, contraindicate implantation [2]. In intracranial hypertension with papilloedema, acetazolamide and furosemide have been used, and the visual loss may have associated choroidal causes that require specific ophthalmological treatment [42].

7.8. Endocrinology and nutrition

Hypopituitarism is treated with hormone replacement: in an Indian boy, with levothyroxine and testosterone, maintaining the glucocorticoids, which also covered the hypocortisolism [43]. The combination of prednisone and losartan allowed puberty to be initiated in an 18-year-old woman with hypogonadotrophic hypogonadism, which the authors attribute to improved adipogenesis [66]. Vitamin D deficiency should be corrected, as in patients treated with bisphosphonates or with corticosteroids [14, 66].

7.9. Agents and circumstances to avoid

Excess phosphate should be avoided: cellulose phosphate worsened hypocalcaemia and proximal myopathy in one patient [2]. Reduction and withdrawal of corticosteroids after prolonged treatment should be gradual, because suppression of the hypothalamic-pituitary-adrenal axis depends on the dose and duration and relative adrenal insufficiency may persist for months; in this disease, withdrawal worsened pain and headache in three of four Korean patients [13, 59]. Losartan is contraindicated in the second and third trimesters of pregnancy [58, 68].

7.10. Pregnancy, perioperative care and transition

Pregnancy improved pain and strength in a patient treated with corticosteroids, who was able to stop them, and improved the pain of a patient with type 2 [2, 6]. Losartan should be replaced before conception or stopped as soon as pregnancy is detected [58, 68]. The prednisolone SmPC does not recommend its use in pregnancy and breastfeeding unless the expected benefit outweighs the risk, and newborns of treated mothers should be observed for signs of adrenal insufficiency [59]. Two patients with type 2 had pregnancies, one spontaneous at 37 years and the other by in vitro fertilisation with a donor after premature ovarian insufficiency [6].

Perioperatively, skull base hyperostosis, narrowing of the foramina and intracranial hypertension require planning with CT, and skull base surgery carries an increased risk [39]. Patients on corticosteroid therapy, or who stopped it a few months earlier, should receive a rapid-acting glucocorticoid before, during and after surgery or other intense stress, as the prednisolone SmPC indicates [59]. The sources consulted do not describe anaesthetic difficulties specific to this disease.

Transition to adult medicine should maintain auditory, ophthalmological and neurological surveillance, because the cranial complications may appear after adolescence even when limb pain improves: in the family of Hughes and colleagues, two of the eight members with pain limited to adolescence later developed cranial nerve palsies [32]. Family planning and genetic counselling should be offered to affected or at-risk young adults [2].

7.11. Regulatory status

The sources consulted do not include FDA labels for the glucocorticoids or for zoledronic acid, and losartan has no centralised authorisation from the European Medicines Agency; no label mentions Camurati-Engelmann disease [34].

7.12. Published evidence on treatment

None of these works is a clinical trial: they are reports and series without a control group, without blinding and with subjective outcomes, in a disease with spontaneous remissions and attenuation with age, which prevents improvement from being attributed with confidence to the treatment [32]. Selective publication of favourable responses is likely, and reports of absence of response are few. It is not possible to compare the treatments with one another from these data.

7.13. Developments of the last 12 months

Between November 2025 and November 2026 no clinical trials or authorisation applications were registered for this disease [57]. The literature published in that period consists of case reports and series: the discordant response to alendronate and zoledronate and the identification of the largest TGFB1 signal peptide duplication through reanalysis of a genome sequencing panel (report published in 2025) [25]; the first case with bilateral choroidal osteomas and papilloedema (report published in July 2026) [42]; and a patient with compressive optic neuropathy treated with zoledronate, methotrexate and hormone therapy (report published in 2026) [41]. In 2025 there were also published the Peruvian family with type 2 disease, the Spanish quality of life series, the Turkish cohort of osteosclerotic disorders and the audit of the Indian registry of rare metabolic bone diseases, which molecularly confirmed cases of Camurati-Engelmann disease [10, 21, 38, 74]. None of these publications alters the GeneReviews management recommendations.

7.14. Recommendations and divergences between sources

The sources agree on corticosteroid therapy as first-line symptomatic treatment, on tapering to the lowest effective dose and on losartan as a corticosteroid-sparing adjuvant [2, 12, 13]. They diverge on bisphosphonates: Orphanet (2013) considers them ineffective, Janssens and colleagues (2006) unpromising and GeneReviews (2023) to be considered with caution, whereas reports from 2017 to 2025 describe responses to zoledronate, pamidronate and alendronate [2, 11, 12, 14, 21, 25]. They also diverge on the skull base: GeneReviews accepts that corticosteroids may delay cranial involvement, and Carlson and colleagues find no medical benefit for the symptoms already established [2, 39].

8. Reproductive Options and Family Counselling

8.1. Recurrence risks

Type 1 disease is transmitted in an autosomal dominant manner [2]. Many patients have an affected parent, and others have a de novo variant, in an unknown proportion [2, 11]. De novo variants have been confirmed, for example, in a young Chinese woman and in a British man, and the three TGFB2 variants described in 2024 were all de novo [9, 25, 66].

The family history may appear negative because the disease was not recognised in relatives, because a parent died before symptom onset or because of reduced penetrance; an apparently isolated case is therefore only confirmed with radiographs of the parents and/or molecular study [2]. The family of Janssens and colleagues with p.(Tyr81His) illustrates the problem: the mother and maternal grandfather of a severely affected girl had been diagnosed as carriers by scintigraphy and radiology, but the variant segregated in the paternal branch, in a father and a paternal grandmother with normal radiographs [12]. The study by Clybouw and colleagues had proposed scintigraphy for counselling in sporadic cases, but isolated uptake at the skull base is frequent in the normal population [3, 12]. Ribbing disease was also linked to autosomal recessive inheritance on the basis of sibships without a parent being studied, and the later identification of heterozygous TGFB1 variants in these patients supports dominant inheritance [2, 4].

8.2. Study of relatives

Assessment of at-risk relatives allows early diagnosis, avoids diagnostic errors and allows limb pain to be treated; it is done by targeted molecular testing when the familial variant is known, or by radiographs looking for hyperostosis when it is not [2]. The Oxford Desk Reference notes that clinical and radiological examination can be informative in the dominant forms [22]. Because the disease has treatable manifestations in childhood, GeneReviews includes at-risk children in this assessment, the aim of which is early diagnosis and treatment of pain [2].

8.3. Prenatal and preimplantation diagnosis

Once the familial variant has been identified, prenatal diagnosis and preimplantation genetic testing are possible, but the result does not predict the age at onset, the severity, the type of symptoms or the speed of progression [2]. Janssens and colleagues stress that a healthy carrier may have a severely affected child and that a severely affected parent may have a child with mild disease; in the series of 24 families, only one parent requested prenatal diagnosis, after a thorough discussion in the genetics clinic [12]. Views on the use of prenatal diagnosis differ between professionals and between families, and the ideal moment to discuss the risk and the reproductive options is before pregnancy [2].

8.4. Type 2 disease and allelic disorders

In type 2 disease, the Peruvian family confirmed autosomal dominant transmission, and the risk for the children of an affected person is 50% [6, 10]. Counselling for TGFB2 variants should distinguish the LAP variants associated with Camurati-Engelmann disease type 2 from the loss-of-function variants associated with Loeys-Dietz syndrome type 4, which have different cardiovascular implications [9, 36]. The parents of a child with the autosomal recessive disease due to biallelic TGFB1 loss of function are heterozygous carriers, without risk of Camurati-Engelmann disease [34].

9. Special Considerations: Population Diversity and Epidemiology

9.1. Prevalence

The prevalence is unknown [2, 11]. Orphanet assigns a prevalence class below 1 per 1 000 000 and records more than 300 published cases; GeneReviews reports more than 300 published patients, and Carlson and colleagues gathered 306 in 2010 [2, 11, 39]. Ayyavoo and colleagues counted about 250 described cases in 2014, and Klemm and colleagues about 300 in 2023 [65, 70]. These figures count published patients and do not constitute population estimates; because many carriers have mild disease or no manifestations, the true frequency is probably higher than the published cases suggest [32]. No prevalence estimates were found for type 2 disease, with six published patients [9, 10].

9.2. Geographical distribution

The disease has been described in families from Europe, Asia, Africa, the Americas, Australia and Oceania, in various ethnic groups and with equal distribution between the sexes [11, 12]. There are no founder variants: haplotype analysis showed independent mutational events at the same positions, and p.(Arg218Cys) is the most frequent in several populations, including the four Turkish children of Uludağ Alkaya and colleagues [7, 21, 24].

In the United States, the most frequently cited series are those of the Mayo Clinic (12 patients assessed between 1968 and 2008) and the four-generation family of Wallace and colleagues [29, 39]. In Europe, the Belgian review of 24 families gathered patients from several countries, and there are recent series in Spain (15 patients, Barcelona) and Turkey (4 children, Istanbul), besides Italian and British families and French cases [12, 21, 38, 67]. In Asia, the largest series are Chinese (18 and 14 patients from Peking Union Medical College Hospital, probably overlapping), Korean (4 and 8 patients) and Japanese, and the Indian registry of rare metabolic bone diseases (218 patients between 2010 and 2024, with sclerosing disorders in 13.7%) included cases with molecular confirmation [13, 23, 31, 40, 45, 74]. In Africa, the first case with molecular confirmation south of the Sahara was described in Tanzania in 2018 [46]. Type 2 disease has been described in Japan, in a family of Moroccan Jewish origin and in Peru [6].

9.3. Portugal and Brazil

In Portugal, Saraiva described in 1997, from the genetics clinic of the Hospital Pediátrico de Coimbra, a three-generation family with 18 affected members, six of whom were studied in detail, with highly variable expressivity and a good symptomatic response to corticosteroids in one patient; in 2000, the same author described anticipation of the age at onset in a large family with 15 affected members in three generations [30, 75]. Saraiva was a co-author of the European molecular study of Campos-Xavier and colleagues, which included European families with p.(Arg218His), p.(Arg218Cys) and p.(Cys225Arg) [24]. In 2017, a team from Coimbra described pain relief with infliximab in a patient with disease of early childhood onset and ulcerative colitis [72]. In Brazil, two women with severe disease were described in Campinas (2005), the use of whole-body magnetic resonance imaging in Rio de Janeiro (2020) and a family with three affected members in Goiânia (2024) [27, 49, 69]. The sources consulted contain no frequency estimates for either country.

9.4. European Reference Network and organisations

The relevant European network is ERN BOND, the European Reference Network on Rare Bone Diseases [15]. GeneReviews lists as resources the International Skeletal Dysplasia Registry of the University of California, Los Angeles, the American Society for Deaf Children, the National Association of the Deaf and Face Equality International [2]. The Oxford Desk Reference refers to patient organisations for skeletal dysplasias and to the Contact a Family network in the United Kingdom [22]. The sources consulted did not identify patient organisations specific to this disease in Portugal or Brazil.

10. Illustrative Clinical Case

Hassan and colleagues, from Bristol, Oxford and London, published in 2025 the case of a man who presented in 2014, at 34 years of age, with generalised pain in the limbs and spine, arthralgia and muscle weakness since adolescence [25]. He had had fractures after falls at 2, 7, 15 and 21 years, had developed bilateral sensorineural hearing loss in his third decade and had no relevant family history; a previous vitamin D deficiency had been corrected without improvement in the symptoms [25]. On examination he was slender (173 cm, 53.8 kg, body mass index of 18 kg/m²), used hearing aids and had reduced proximal muscle bulk in the lower limbs [25]. Alkaline phosphatase was slightly raised (155 U/l; reference 20–120), with normal calcium, phosphate, parathyroid hormone and 25-hydroxyvitamin D and normal creatine kinase (37 U/l); the remodelling markers were markedly raised, with P1NP of 179 µg/l (reference 20–76) and CTX of 1.7 µg/l (reference 0.1–0.5) [25]. Radiographs showed a diffuse coarse sclerotic pattern with marked diaphyseal expansion, and densitometry had Z-scores of +4.1 at the lumbar spine, +3.0 at the total hip and +5.3 at the femoral neck; muscle biopsy was inconclusive [25].

On the basis of the radiographs and the hearing loss, juvenile Paget disease was diagnosed, but sequencing of the known monogenic causes was negative, and bisphosphonates were not started because of the high mineral density [25]. In 2020, pain and hearing worsened; alkaline phosphatase had risen to 411 U/l, the audiogram showed severe, progressive bilateral sensorineural hearing loss, and cranial CT showed diffuse thickening of the skull with almost complete obliteration of both internal auditory canals [25]. With alendronate, the musculoskeletal symptoms resolved almost completely within six weeks, and CTX fell to 1.03 µg/l at nine months [25]. In 2022, with recurrence of the pain, he changed to intravenous zoledronic acid, and in the week after the first dose the pain and hearing loss worsened, for two months; once alendronate was resumed, the symptoms improved within 12 weeks [25].

In 2024, with more weakness, instability and fatigue, he had a skeletal dysplasia panel by genome sequencing (panel R104 of the NHS genomic service), initially reported without causal variants; reanalysis with additional clinical information and review of the VUS identified the heterozygous in-frame duplication NM_000660.7:c.31_48dup, p.(Leu11_Leu16dup), in exon 1 of TGFB1, absent in both parents and therefore de novo [8, 25]. The duplication adds six residues (LLLPLL) to the hydrophobic region of the signal peptide, which goes from 14 to 20 amino acids, and is the largest duplication described in this region; the authors accept that it may impair signal peptide cleavage and reduce secretion, with intracellular accumulation capable of activating signalling, as shown for the smaller duplications [19, 25]. Reassessment of the history, biochemistry and imaging was considered compatible with Camurati-Engelmann disease [25]. Figure 1 of the article, consulted, shows the radiographs of the pelvis and knee with a coarse trabecular pattern and cortical expansion, a model of the protein with the duplication, a diagram of the normal and mutant protein and the chromatograms of the proband and the parents (https://pmc.ncbi.nlm.nih.gov/articles/PMC12671855/figure/keaf441-F1/) [25].

The case brings together several characteristic aspects of the disease: the diagnostic delay, with an initial diagnosis of another sclerosing dysplasia; the hearing loss from obliteration of the internal auditory canals; the fractures, which GeneReviews does not record as a feature of the disease; the high bone mineral density, which does not exclude the usefulness of antiresorptives; and the need to reanalyse negative results [2, 25]. The therapeutic approach described, with alendronate, does not correspond to the GeneReviews recommendations, which favour corticosteroids and losartan and place bisphosphonates among the treatments to be considered with caution, and the discordant response to the two bisphosphonates is an observation in a single patient [2, 25].

11. Prognosis and Long-Term Follow-Up

11.1. Prognosis

Life expectancy and adult height are normal according to the atlas of Spranger and colleagues, and pain may diminish in adulthood [1, 2]. The New Zealand family of Hughes and colleagues showed that scintigraphic activity decreases with age and that most carriers had pain limited to adolescence, but also that cranial nerve palsies may appear later [32]. In other patients, the disease remains active and progresses: the 45-year follow-up of Grey and colleagues showed extension to the metaphyses, epiphyses, hip joints, spine and hands, and the most severely affected patient of Wallace and colleagues went, over 30 years, from mild cranial hyperostosis to severe thickening with nerve compression [29, 37]. Orphanet, in a 2013 synthesis, describes the disease as progressive and of poor prognosis, with quality of life compromised by pain and reduced mobility; this assessment mainly reflects patients diagnosed through symptoms and diverges from the family data of Hughes and colleagues [11, 32]. The physical component of quality of life is reduced and the mental component preserved [38]. The serious complications are intracranial hypertension, visual loss from optic nerve compression or papilloedema, severe hearing loss and hypopituitarism; an adolescent died from dissection of a dilated ascending aorta, without it being known whether this complication is related to the disease [2, 39, 42, 43]. In type 2 disease, the Peruvian family shows a disabling course, with one patient in a wheelchair at 70 years [6].

11.2. Surveillance

The annual and per-visit intervals are those proposed by GeneReviews, without studies validating them [2]. In the mild forms, Carlson and colleagues recommend serial examination, audiometry and radiography [39]. Routine aortic assessment is not recommended; the fatal case of aortic dissection and the patients with mitral valve prolapse and annuloaortic ectasia in type 2 disease justify attention to cardiovascular signs [2, 6].

12. Summary and Final Considerations

Camurati-Engelmann disease is a rare autosomal dominant bone dysplasia in which variants of the LAP of TGFB1 (type 1) or of TGFB2 (type 2) increase the availability of the active factor and cause cortical hyperostosis of the diaphyses, sclerosis of the skull base, limb pain, proximal weakness, reduced muscle mass and adipose tissue and waddling gait [2, 9]. Ribbing disease is a mild, late and asymmetrical form of the same disease [2]. Penetrance is incomplete and expressivity highly variable, even within the same family, and the Arg218 variants account for about three quarters of patients [2, 12].

The diagnosis rests on radiography of the long bones and skull and is confirmed by TGFB1 sequencing, which detects a variant in more than 90% of patients with a typical phenotype; a negative result leads to study of TGFB2 and to reanalysis of the signal peptide [2, 25]. Diagnostic delay is frequent, with initial diagnoses of neuromuscular or inflammatory diseases, and a plain radiograph resolves most of those pathways [2, 27].

Treatment is symptomatic and based on case reports: prednisolone or deflazacort at the lowest effective dose, losartan as an adjuvant, bisphosphonates in selected patients, rehabilitation and surgery for the cranial and orthopaedic complications; there are no registered trials and no medicines approved for the disease [2, 13, 14, 57]. Annual surveillance of hearing, vision, neurological function and the full blood count, and surveillance at each visit of growth, puberty and blood pressure, are directed at the complications that most affect the prognosis [2]. The natural history is more benign than the first series suggested, with attenuation of activity in adulthood in many carriers, but a minority have disabling disease or severe cranial complications [32, 39]. Genetic counselling should explain the 50% risk for children and the impossibility of predicting severity from the genotype [2, 12].

13. References

1. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Camurati-Engelmann disease (MIM 131300). In: Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford: Oxford University Press; 2018. p. 639–643.
2. Wallace SE, Wilcox WR. Camurati-Engelmann disease. 2004 Jun 25 [updated 2023 Aug 31]. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews®. Seattle (WA): University of Washington, Seattle; 1993–2026. https://www.ncbi.nlm.nih.gov/books/NBK1156/
3. Online Mendelian Inheritance in Man, OMIM®. Camurati-Engelmann disease 1; CAEND1. MIM Number: 131300. Baltimore (MD): Johns Hopkins University; last edited 2025 Sep 17. https://www.omim.org/entry/131300
4. Online Mendelian Inheritance in Man, OMIM®. Ribbing disease. MIM Number: 601477. Baltimore (MD): Johns Hopkins University; last edited 2019 Jul 24. https://www.omim.org/entry/601477
5. Nishimura G, Nishimura H, Tanaka Y, Makita Y, Ikegawa S, Ghadami M, et al. Camurati-Engelmann disease type II: progressive diaphyseal dysplasia with striations of the bones. Am J Med Genet. 2002;107(1):5-11. doi:10.1002/ajmg.10079. PMID: 11807860.
6. Online Mendelian Inheritance in Man, OMIM®. Camurati-Engelmann disease 2; CAEND2. MIM Number: 606631. Baltimore (MD): Johns Hopkins University; last edited 2025 Sep 17. https://www.omim.org/entry/606631
7. Online Mendelian Inheritance in Man, OMIM®. Transforming growth factor, beta-1; TGFB1. MIM Number: 190180. Baltimore (MD): Johns Hopkins University. https://www.omim.org/entry/190180
8. VariantValidator (GRCh38, MANE Select transcripts). Manchester: University of Manchester; accessed 1 Nov 2026. https://variantvalidator.org/
9. Wang Z, Kometani M, Zeitlin L, Wilnai Y, Kinoshita A, Yoshiura KI, et al. Heterozygous mutations in the straitjacket region of the latency-associated peptide domain of TGFB2 cause Camurati-Engelmann disease type II. J Hum Genet. 2024;69(11):599-605. doi:10.1038/s10038-024-01274-1. PMID: 39014191.
10. Mumm S, Paz-Ibarra JL, Campeau PM, Garrido-Carrasco E, Baker JC, Pino-Nina E, et al. Transforming growth factor, beta-2 gene mutation causes autosomal dominant Camurati-Engelmann disease, type 2 (OMIM % 606631). Bone. 2025;197:117477. doi:10.1016/j.bone.2025.117477. PMID: 40204055.
11. Orphanet. Camurati-Engelmann disease (ORPHA:1328). https://www.orpha.net/en/disease/detail/1328
12. Janssens K, Vanhoenacker F, Bonduelle M, Verbruggen L, Van Maldergem L, Ralston S, et al. Camurati-Engelmann disease: review of the clinical, radiological, and molecular data of 24 families and implications for diagnosis and treatment. J Med Genet. 2006;43(1):1-11. doi:10.1136/jmg.2005.033522. PMID: 15894597. PMC2564495.
13. Kim YM, Kang E, Choi JH, Kim GH, Yoo HW, Lee BH. Clinical characteristics and treatment outcomes in Camurati-Engelmann disease: A case series. Medicine (Baltimore). 2018;97(14):e0309. doi:10.1097/MD.0000000000010309. PMID: 29620655. PMC5902284.
14. Baroncelli GI, Ferretti E, Pini CM, Toschi B, Consolini R, Bertelloni S. Significant Improvement of Clinical Symptoms, Bone Lesions, and Bone Turnover after Long-Term Zoledronic Acid Treatment in Patients with a Severe Form of Camurati-Engelmann Disease. Mol Syndromol. 2017;8(6):294-302. doi:10.1159/000479859. PMID: 29230158. PMC5701277.
15. ERN BOND — European Reference Network on Rare Bone Diseases. https://ernbond.eu/
16. Kinoshita A, Saito T, Tomita H, Makita Y, Yoshida K, Ghadami M, et al. Domain-specific mutations in TGFB1 result in Camurati-Engelmann disease. Nat Genet. 2000;26(1):19-20. doi:10.1038/79128. PMID: 10973241.
17. Janssens K, Gershoni-Baruch R, Guañabens N, Migone N, Ralston S, Bonduelle M, et al. Mutations in the gene encoding the latency-associated peptide of TGF-beta 1 cause Camurati-Engelmann disease. Nat Genet. 2000;26(3):273-5. doi:10.1038/81563. PMID: 11062463.
18. Saito T, Kinoshita A, Yoshiura Ki, Makita Y, Wakui K, Honke K, et al. Domain-specific mutations of a transforming growth factor (TGF)-beta 1 latency-associated peptide cause Camurati-Engelmann disease because of the formation of a constitutively active form of TGF-beta 1. J Biol Chem. 2001;276(15):11469-72. doi:10.1074/jbc.C000859200. PMID: 11278244.
19. Janssens K, ten Dijke P, Ralston SH, Bergmann C, Van Hul W. Transforming growth factor-beta 1 mutations in Camurati-Engelmann disease lead to increased signaling by altering either activation or secretion of the mutant protein. J Biol Chem. 2003;278(9):7718-24. doi:10.1074/jbc.M208857200. PMID: 12493741.
20. Unger S, Ferreira CR, Mortier GR, Ali H, Bertola DR, Calder A, et al. Nosology of genetic skeletal disorders: 2023 revision. Am J Med Genet A. 2023;191(5):1164-1209. doi:10.1002/ajmg.a.63132. PMID: 36779427. PMC10081954.
21. Uludağ Alkaya D, Usluer E, Alp Ünkar Z, Şeker A, Adaletli İ, Güneş N, et al. Insights into Natural History, Phenotypic, and Molecular Spectrum in a Large Cohort of Osteosclerotic Disorders. Calcif Tissue Int. 2025;116(1):59. doi:10.1007/s00223-025-01366-w. PMID: 40198394. PMC11978542.
22. Firth HV, Hurst JA. Increased bone density. In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 198–199.
23. Li Q, Zhao Z, Wu B, Pang Q, Cui L, Zhang L, et al. Alteration of Bone Density, Microarchitecture, and Strength in Patients with Camurati-Engelmann Disease: Assessed by HR-pQCT. J Bone Miner Res. 2022;37(1):78-86. doi:10.1002/jbmr.4436. PMID: 34490910.
24. Campos-Xavier B, Saraiva JM, Savarirayan R, Verloes A, Feingold J, Faivre L, et al. Phenotypic variability at the TGF-beta1 locus in Camurati-Engelmann disease. Hum Genet. 2001;109(6):653-8. doi:10.1007/s00439-001-0644-8. PMID: 11810278.
25. Hassan N, Gregson CL, Oldridge M, Lester T, Duncan EL, Tobias JH. Identification of a novel TGFB1 variant in a patient with Camurati-Engelmann disease responsive to alendronate. Rheumatology (Oxford). 2025;64(12):6427-6429. doi:10.1093/rheumatology/keaf441. PMID: 40811023. PMC12671855.
26. Alkaya A, Yıldız AE, Bağlan E, Özdel S. Unveiling the uncommon: diagnostic journey of camurati-engelmann disease in a pediatric patient. Pediatr Rheumatol Online J. 2024;22(1):89. doi:10.1186/s12969-024-01016-9. PMID: 39379987. PMC11460205.
27. Campos T, Uchoa E, Santos V, Zatarin R, Benício R, Gomes C, et al. Phenotypic Variability in Camurati-Engelmann Disease: A Case Report of a Family with the c.653G>A Pathogenic Variant in the TGFB1 Gene. Genes (Basel). 2024;15(11). doi:10.3390/genes15111354. PMID: 39596554. PMC11593567.
28. Pickering ME, Ltaief-Boudrigua A, Feurer E, Collet C, Chapurlat R. A new LRP6 variant and Camurati-Engelmann-like disease. Bone. 2021;143:115706. doi:10.1016/j.bone.2020.115706. PMID: 33164853.
29. Wallace SE, Lachman RS, Mekikian PB, Bui KK, Wilcox WR. Marked phenotypic variability in progressive diaphyseal dysplasia (Camurati-Engelmann disease): report of a four-generation pedigree, identification of a mutation in TGFB1, and review. Am J Med Genet A. 2004;129A(3):235-47. doi:10.1002/ajmg.a.30148. PMID: 15326622.
30. Saraiva JM. Progressive diaphyseal dysplasia: a three-generation family with markedly variable expressivity. Am J Med Genet. 1997;71(3):348-52. doi:10.1002/(sici)1096-8628(19970822)71:3<348::aid-ajmg17>3.0.co;2-k. PMID: 9268107.
31. Liang H, Jiajue R, Qi W, Liu W, Chi Y, Jiang Y, et al. Clinical characteristics and the influence of rs1800470 in patients with Camurati-Engelmann disease. Front Endocrinol (Lausanne). 2022;13:1041061. doi:10.3389/fendo.2022.1041061. PMID: 36339419. PMC9631481.
32. Hughes P, Hassan I, Que L, Mead P, Lee JH, Love DR, et al. Observations on the Natural History of Camurati-Engelmann Disease. J Bone Miner Res. 2019;34(5):875-882. doi:10.1002/jbmr.3670. PMID: 30690794.
33. Kotlarz D, Marquardt B, Barøy T, Lee WS, Konnikova L, Hollizeck S, et al. Human TGF-β1 deficiency causes severe inflammatory bowel disease and encephalopathy. Nat Genet. 2018;50(3):344-348. doi:10.1038/s41588-018-0063-6. PMID: 29483653. PMC6309869.
34. GeneGuidance. TGFB1 — clinical curation of the gene (version of 2026-09-26).
35. Firth HV, Hurst JA. Marfan syndrome. In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 484–489.
36. Mulder BJM, van de Laar IMBH, De Backer J. Heritable thoracic aortic diseases: syndromal and isolated (F)TAAD. In: Baars HF, van der Smagt JJ, Doevendans PAFM, van Tintelen JP, editors. Clinical Cardiogenetics. 3rd ed. Cham: Springer; 2020. p. 309–343. https://doi.org/10.1007/978-3-030-45457-9_19
37. Grey AC, Wallace R, Crone M. Engelmann's disease: a 45-year follow-up. J Bone Joint Surg Br. 1996;78(3):488-91. PMID: 8636193.
38. Peláez P, Peris P, Guañabens N. Health-related quality of life in patients with Camurati-Engelmann disease: A case series study. Med Clin (Barc). 2025;165(4):107043. doi:10.1016/j.medcli.2025.107043. PMID: 40513308.
39. Carlson ML, Beatty CW, Neff BA, Link MJ, Driscoll CL. Skull base manifestations of Camurati-Engelmann disease. Arch Otolaryngol Head Neck Surg. 2010;136(6):566-75. doi:10.1001/archoto.2010.68. PMID: 20566907.
40. Yuldashev AJ, Shin CH, Kim YS, Jang WY, Park MS, Chae JH, et al. Orthopedic Manifestations of Type I Camurati-Engelmann Disease. Clin Orthop Surg. 2017;9(1):109-115. doi:10.4055/cios.2017.9.1.109. PMID: 28261436. PMC5334020.
41. Lei C, He Z, Liang Y, Li Y, Zhao SX, Qiao J, et al. Camurati-Engelmann disease with bilateral proptosis and optic neuropathy: a case report and literature review. Orbit. 2026;45(4):557-564. doi:10.1080/01676830.2026.2650802. PMID: 41949641.
42. Rao A, Cristescu I, Quhill H, Harley U, Ramalingam S, Tsermoulas G, et al. Case Report: Vision-threatening manifestations of choroidal osteomas and papilloedema in presumed Camurati-Engelmann disease. Front Neurol. 2026;17:1831878. doi:10.3389/fneur.2026.1831878. PMID: 42582186. PMC13457156.
43. Das L, Dhiman V, Dutta P, Sood A, Prakash M, Kaur S, et al. Camurati-Engelmann Disease Complicated by Hypopituitarism: Management Challenges and Literature Review of Outcomes With Bisphosphonates. AACE Clin Case Rep. 2022;8(2):58-64. doi:10.1016/j.aace.2021.10.002. PMID: 35415221. PMC8984528.
44. Crisp AJ, Brenton DP. Engelmann's disease of bone--a systemic disorder? Ann Rheum Dis. 1982;41(2):183-8. doi:10.1136/ard.41.2.183. PMID: 7073346. PMC1000905.
45. Makita Y, Nishimura G, Ikegawa S, Ishii T, Ito Y, Okuno A. Intrafamilial phenotypic variability in Engelmann disease (ED): are ED and Ribbing disease the same entity?. Am J Med Genet. 2000;91(2):153-6. doi:10.1002/(sici)1096-8628(20000313)91:2<153::aid-ajmg15>3.0.co;2-u. PMID: 10748417.
46. Mwasamwaja AO, Mkwizu EW, Shao ER, Kalambo CF, Lyaruu I, Hamel BC. Camurati-Engelmann disease: a case report from sub-Saharan Africa. Oxf Med Case Reports. 2018;2018(7):omy036. doi:10.1093/omcr/omy036. PMID: 30034812. PMC6049015.
47. Guillen Arguello R, Sader N, McComb JG. Unilateral hemicraniectomy with titanium cranioplasty for the treatment of high intracranial pressure in a pediatric patient with Camurati-Engelmann disease: illustrative case. J Neurosurg Case Lessons. 2024;8(25). doi:10.3171/CASE24590. PMID: 39680872. PMC11665146.
48. Richards S, Aziz N, Bale S, Bick D, Das S, Gastier-Foster J, et al. Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology. Genet Med. 2015;17(5):405-24. doi:10.1038/gim.2015.30. PMID: 25741868. PMC4544753.
49. Corrêa DG, Canella C, de Queiroz Campos Araújo AP, Mendonça S, Costa FM. Whole-Body Magnetic Resonance Imaging in Camurati-Engelmann Disease. JBMR Plus. 2020;4(4):e10357. doi:10.1002/jbm4.10357. PMID: 32258971. PMC7117840.
50. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Craniodiaphyseal dysplasia (MIM 122860, 218300). In: Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford: Oxford University Press; 2018. p. 663–665.
51. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Ghosal hematodiaphyseal dysplasia (MIM 231095). In: Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford: Oxford University Press; 2018. p. 644–646.
52. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Osteoectasia with hyperphosphatasia (MIM 239000). In: Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford: Oxford University Press; 2018. p. 629–633.
53. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Infantile cortical hyperostosis (MIM 114000). In: Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford: Oxford University Press; 2018. p. 623–626.
54. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Craniometaphyseal dysplasia (MIM 123000, 218400). In: Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford: Oxford University Press; 2018. p. 666–669.
55. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Diaphyseal medullary stenosis with bone malignancy (MIM 112250). In: Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford: Oxford University Press; 2018. p. 685–688.
56. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Pachydermoperiostosis, autosomal dominant (MIM 167100). In: Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford: Oxford University Press; 2018. p. 656–659.
57. ClinicalTrials.gov. Search for “Camurati” and “diaphyseal dysplasia” (API v2). Bethesda (MD): National Library of Medicine; accessed 1 Nov 2026. https://clinicaltrials.gov/
58. Organon LLC. COZAAR (losartan potassium) tablets, for oral use. Prescribing information. NDA 020386 (initial approval 1995 Apr 14); effective revision 2026 Jan 20. Silver Spring (MD): U.S. Food and Drug Administration. https://www.accessdata.fda.gov/scripts/cder/daf/index.cfm?event=BasicSearch.process&ApplNo=020386
59. Angelini Pharma Portugal. Lepicortinolo 20 mg tablet (prednisolone). Summary of Product Characteristics, approved 2022 Apr 22. Lisboa: INFARMED; Infomed med_id 4972. https://extranet.infarmed.pt/INFOMED-fo/
60. Agarwal K, Rajan R, Paul J, Cherian KE, Kapoor N, Paul TV. Losartan as a Steroid-Sparing Adjunct in a Patient With Features of Refractory Camurati-Engelmann Disease. AACE Clin Case Rep. 2022;8(2):54-57. doi:10.1016/j.aace.2021.08.002. PMID: 35415233. PMC8984203.
61. Baş F, Darendeliler F, Petorak I, Sadikoğlu B, Bilir A, Bundak R, et al. Deflazacort treatment in progressive diaphyseal dysplasia (Camurati-Engelmann disease). J Paediatr Child Health. 1999;35(4):401-5. doi:10.1046/j.1440-1754.1999.00381.x. PMID: 10457303.
62. Alter, S.A. Deflazacorte Alter 6 mg and 30 mg tablets. Summary of Product Characteristics, approved 2022 Jan 28. Lisboa: INFARMED; Infomed med_id 46271. https://extranet.infarmed.pt/INFOMED-fo/
63. Simsek-Kiper PO, Dikoglu E, Campos-Xavier B, Utine GE, Bonafe L, Unger S, et al. Positive effects of an angiotensin II type 1 receptor antagonist in Camurati-Engelmann disease: a single case observation. Am J Med Genet A. 2014;164A(10):2667-71. doi:10.1002/ajmg.a.36692. PMID: 25099136.
64. Habashi JP, Judge DP, Holm TM, Cohn RD, Loeys BL, Cooper TK, et al. Losartan, an AT1 antagonist, prevents aortic aneurysm in a mouse model of Marfan syndrome. Science. 2006;312(5770):117-21. doi:10.1126/science.1124287. PMID: 16601194. PMC1482474.
65. Ayyavoo A, Derraik JG, Cutfield WS, Hofman PL. Elimination of pain and improvement of exercise capacity in Camurati-Engelmann disease with losartan. J Clin Endocrinol Metab. 2014;99(11):3978-82. doi:10.1210/jc.2014-2025. PMID: 25140400.
66. Cui L, Li Q, Guan W, Yu W, Li X, Xia W, et al. Improvement of Bone Health and Initiation of Puberty Development in Camurati-Engelmann Disease With Glucocorticoid and Losartan Treatment: A Case Report and Review of Literature. Front Endocrinol (Lausanne). 2022;13:882144. doi:10.3389/fendo.2022.882144. PMID: 35784539. PMC9247158.
67. Combier A, Palazzo E, Forien M, Gardette A, Dieudé P, Ottaviani S. Failure of conventional treatment and losartan in Camurati-Engelmann disease: A case report. Joint Bone Spine. 2018;85(5):649-650. doi:10.1016/j.jbspin.2018.01.015. PMID: 29452301.
68. Alter, S.A. Losartan Alter 50 mg and 100 mg film-coated tablets. Summary of Product Characteristics. Lisboa: INFARMED; Infomed med_id 39041. https://extranet.infarmed.pt/INFOMED-fo/
69. Castro GR, Appenzeller S, Marques-Neto JF, Bértolo MB, Samara AM, Coimbra I. Camurati-Engelmann disease: failure of response to bisphosphonates: report of two cases. Clin Rheumatol. 2005;24(4):398-401. doi:10.1007/s10067-004-1056-7. PMID: 15660289.
70. Klemm P, Aykara I, Lange U. Camurati-Engelmann Disease: A Case-Based Review About an Ultrarare Bone Dysplasia. Eur J Rheumatol. 2023;10(1):34-38. doi:10.5152/eurjrheum.2023.21115. PMID: 36880809. PMC10152113.
71. Organon LLC. FOSAMAX PLUS D (alendronate sodium/cholecalciferol) tablets. Prescribing information. NDA 021762; effective revision 2026 Feb 3. Silver Spring (MD): U.S. Food and Drug Administration. https://www.accessdata.fda.gov/scripts/cder/daf/index.cfm?event=BasicSearch.process&ApplNo=021762
72. Moreira S, Cunha B, Jesus NP, Santos L. Pain improvement in Camurati-Engelmann disease after anti-TNFα therapy. BMJ Case Rep. 2017;2017. doi:10.1136/bcr-2017-221760. PMID: 29184006. PMC5720334.
73. Agência Nacional de Vigilância Sanitária (ANVISA). Open data — registered medicines (DADOS_ABERTOS_MEDICAMENTOS.csv). Brasília: ANVISA; accessed 1 Nov 2026. https://dados.anvisa.gov.br/dados/DADOS_ABERTOS_MEDICAMENTOS.csv
74. Sangar M, Das L, Kaur S, Dhiman V, Bhadada SK. Clinical Characteristics and Management of Rare Metabolic Bone Diseases: An Audit of the Rare Metabolic Bone Disease Registry of India. Calcif Tissue Int. 2025;116(1):119. doi:10.1007/s00223-025-01423-4. PMID: 40914927.
75. Saraiva JM. Anticipation in progressive diaphyseal dysplasia. J Med Genet. 2000;37(5):394-5. doi:10.1136/jmg.37.5.394. PMID: 10905898. PMC1734592.