Bruck Syndrome 2 Variant Lacking Congenital Contractures And Involving A Novel Compound Heterozygous PLOD2 Mutation

Abstract

Bruck syndrome (BRKS) is the rare disorder that features congenital joint contractures often with pterygia and subsequent fractures, early on called osteogenesis imperfecta (OI) type XI (OMIM # 610968). Its two forms, BRKS1 (OMIM # 259450) and BRKS2 (OMIM # 609220), reflect autosomal recessive (AR) inheritance of FKBP10 and PLOD2 loss-of-function mutations, respectively. A 10-year-old girl was referred with blue sclera, osteopenia, poorly-healing fragility fractures, Wormian skull bones, cleft soft palate, congenital fusion of cervical vertebrae, progressive scoliosis, bell-shaped thorax, restrictive and reactive pulmonary disease, protrusio acetabuli, short stature, and additional dysmorphic features without joint contractures. Iliac crest biopsy after alendronate treatment that improved her bone density revealed low trabecular connectivity, abundant patchy osteoid, and active bone formation with widely-spaced tetracycline labels. Chromosome 22q11 deletion analysis for velocardiofacial syndrome, COL1A1 and COL1A2 sequencing for prevalent types of OI, and Sanger sequencing of LRP5, PPIB, FKBP10, and IFITM5 for rare pediatric osteoporoses were negative. Copy number microarray excluded a contiguous gene syndrome. Instead, exome sequencing revealed two missense variants in PLOD2 which encodes procollagen-lysine, 2-oxoglutarate 5-dioxygenase 2 (lysyl hydroxylase 2, LH2); exon 8, c.797G>T, p.Gly266Val (paternal), and exon 12, c.1280A>G, p.Asn427Ser (maternal). In the Exome Aggregation Consortium (ExAC) database, low frequency (Gly266Val, 0.0000419) and absence (Asn427Ser) implicated both variants as mutations of PLOD2. The father, mother, and sister (who carried the exon 12 defect) were reportedly well with normal parental DXA findings. BRKS2, characterized by under-hydroxylation of type I collagen telopeptides compromising their crosslinking, has been reported in at least 16 probands/families. Most PLOD2 mutations involve exons 17–19 (of 20 total) encoding the C-terminal domain with LH activity. However, truncating defects (nonsense, frameshift, splice site mutations) are also found throughout PLOD2. In three reports, AR PLOD2 mutations are not associated with congenital contractures. Our patient’s missense defects lie within the central domain of unknown function of PLOD2. In our patient, compound heterozygosity with PLOD2 mutations is associated with a clinical phenotype distinctive from classic BRKS2 indicating that when COL1A1 and COL1A2 mutation testing is negative for OI without congenital contractures or pterygia, atypical BRKS should be considered.

II). Introduction:

Bruck syndrome (BRKS) is the rare autosomal recessive (AR) brittle bone disease that also features congenital joint contractures often with pterygia,⁽¹⁾ and is called osteogenesis imperfecta (OI) type XI.⁽²⁾ Its genetic heterogeneity involves two genes; i) FKBP10 causing BRKS1 (OMIM # 259450)⁽²⁾ that encodes FK506-binding protein 10, a member of a chaperone complex which interacts with lysyl hydroxylase 2 (LH2) and regulates its activity,^(3–9) and ii) PLOD2 causing BRKS2 (OMIM # 609220)⁽²⁾ that encodes procollagen-lysine, 2-oxoglutarate 5-dioxygenase 2 (also called LH2), which hydroxylates collagen lysine residues required for fibril crosslinking.^(7–13) At least 17 different PLOD2 mutations, identified in about 16 probands/families, have been linked with classic BRKS2, however additional patients with PLOD2 mutations have been identified (Table 1).^(7–16) Notably, Puig-Hervas et al.⁽¹¹⁾ in 2012 reported AR PLOD2 mutations in a patient with OI without joint contractures, and in two brothers with either mild BRKS2 or mild OI also without contractures. In 2016, PLOD2 defects were identified in one patient with a mild OI phenotype without contractures.⁽¹⁷⁾. In 2018, two additional patients with PLOD2 mutations without contractures were reported from two families.^(13,15) In one of these families, the older affected sister had congenital contractures and her younger sister, had an OI phenotype with no contractures.⁽¹³⁾

Table 1:

PLOD2 Mutations Causing BRKS2 And OI/Other Skeletal Variants^*

Country	Exon	Hom/Het	Mutation	Diagnosis	Reference
Australia	17	Hom	c.1886 C>T, p.Thr629Ile	BRKS2	van der Slot et al. (2003)(10)
Kurdish	17	Hom	c.1865G>T, p.Gly622Val	BRKS2	″
Turkey	17	Hom	c.1856G>A, p.Arg619His	BRKS2	Ha-Vinh et al. (2004)(8)
Egypt	IVS 12	Hom	c.1358+5 G>A	OI Type III	Puig-Hervas et al. (2012)(11)
Egypt	17	Hom	c.1856G>A, p.Arg619His	BRKS2	″
Spain	17 IVS 18	Comp. Het	c.1864G>T, p.Gly622Cys c.2122–2 A>G	Brothers: mild OI, mild BRKS2	″
Egypt	13a	Hom	c.1559dupC, p.Val523CysfsX7	BRKS2	″
China	14	Comp. Het	c.1624delT, p.Tyr542Thrfs c.1880 T>C, p.Val627Ala	BRKS2	Zhou et al. (2014)(7)
Egypt	17 IVS 12	Hom	c.1358+5 G>A	mild OI	Caparros-Martin et al. (2016)(17)
Egypt	17	Hom	c.1828T>C, p.Trp610Arg	BRKS2	″
China	IVS 4 11	Comp. Het	c.503–2 A>G c.1138 C>T, p.Arg380Cys	BRKS2	Liu et al. (2017)(16 ), Lv et al. (2018)(12)
China	11 17	Comp. Het	c.1153T>C, p.Cys385Arg c.1982 G>A, p.Gly661Asp	BRKS2	Lv et al. (2018)(12)
China	11 18	Comp. Het	c.1138 C>T, p.Arg380Cys c.2038 C>T, p.Arg680X	BRKS2	″
Finland/East Eur./Italy	17 4	Comp. Het	c.1754 A>T, p.Asp585Val c.497C>G, p.Ser166X	Severe kyphomelic dysplasia	Leal et al. (2018)(13)
Brazil	16	Presumed Hom	c.1682 G>A, p.Trp561X	kyphomelic dysplasia	″
Iran	19	Hom	c.2060 A>G, p.His687Arg	BRKS2	″
Palestine	17	Hom	c.1764 G>T, p.Trp588Cys	Sisters: BRKS2, moderate OI	″
India	8	Hom	c.797G>T, p.Gly266Val	BRKS2	Arora et al. (2018)(15)
USA (our patient)	8 12	Comp. Het	c.797G>T, p.Gly266Val c.1280A>G, p.Asn427Ser	OI without contractures	Mumm et al. (2018)(23)
USA	5	Presumed Hom	c.517G>C, p.Ala173Pro	BRKS2	Santana et al. (2019)(14)

^*PLOD2 mutations are listed chronologically based on year of publication.

Mutation numbering is based on transcript NM_182943.2.

Herein, we studied a young woman with progressive brittle bone disease since early childhood without congenital joint contractures or pterygia associated with a unique compound heterozygous PLOD2 mutation.

III). Patient and Methods:

A). Medical history:

This 21-year-old American woman of English, Irish, and German descent was referred at 10 years-of-age in 2007 to our Center for Metabolic Bone Disease and Molecular Research, Shriners Hospitals for Children-St. Louis (SHC-STL) for fracturing with progressive skeletal deformities. She was born prematurely at 28 weeks gestation, required mechanical ventilation initially and supplemental oxygen for ~ 2 months in the neonatal intensive care unit, and underwent closure of a patent ductus arteriosus. At age 1 year, cleft soft palate was repaired. A sphincteroplasty for velopharyngeal insufficiency improved her nasal speech. Early gross motor skills were attained on time. Fragility fractures began at ~ age 3 years with a broken left wrist from minimal trauma. Skeletal dysplasia was then recognized together with failure to thrive. She had dental findings suggestive of dentinogenesis imperfecta featuring small brown deciduous teeth with normal appearing permanent teeth, and congenital fusion of vertebrae C6–C7, C6-hemivertebra, and abnormally contoured ribs. By age 4 years, pain began in her leg and left shoulder and became generalized. At age 5 years, she limped and dragged her left leg with restricted hip rotation. At age 7 years, weakness was associated with normal serum aldolase and creatine kinase activity, rapidly worsening scoliosis, shortness of breath, and nocturnal dyspnea despite normal pulmonary function and sleep studies. Vertebrae C4–T1 were fused surgically. At age 9 years, electromyelography revealed slowed left-sided forearm nerve conduction velocity, and femoral osteotomy was used to treat leg length discrepancy from a femoral neck fracture. Bone scan showed multiple healing fractures, including vertebral compressions. Arthritis was perhaps explained her knee and neck/shoulder stiffness. Sedimentation rate was elevated at 32 mm/hr and a sterile knee aspirate showed inflammatory cells. However, rheumatologic serologies were normal (CRP, ANA, anti-dsDNA, anti-histone, complements, and Lyme titers). Naprosyn treatment seemed helpful and symptoms resolved. Gluten-sensitive enteropathy serology antibodies (IgA, and anti-tissue transglutaminase, anti-gliadin, anti-saccharomyces cerevisiae) and small bowel biopsy gave inconsistent results, and a gluten-free diet began at age 10 yrs. After referral, skeletal pain improved during alendronate therapy (35 mg po weekly for one year, then 45 mg weekly), but opioids were taken intermittently. During alendronate treatment, spinal fusion with rod placement was undertaken without complications, but two femoral fractures had occurred. Before alendronate therapy, DXA bone mineral density (BMD) Z-score was −3.6 at L₁-L₄ (without height correction⁽¹⁸⁾ because of her scoliosis) and then improved to −1.8 after one year of treatment. Familial exudative vitreoretinopathy 4, allelic to osteoporosis pseudoglioma syndrome (OMIM # 259770),⁽²⁾ was suggested by her retinopathy of prematurity and fractures, but follow-up ophthalmological studies were negative for that disorder. Restrictive and reactive lung disease (albuterol prn was begun) included intermittent night-time desaturations. At age 17 years, sleep apnea and fatigue improved with night-time continuous positive airway pressure. Significant hip weakness was explained by the associated deformities and minimal ambulation. She had basilar invagination of her skull, congenital bell-shaped thorax with downward sloping ribs, protrusio acetabuli, persistent osteopenia with recurrent vertebral and long bone fractures, and pain. At age 19 years, total hip Z-score was −2.7. At age 21 years, T-scores were −4.3 for total hip and −1.7 for femoral neck BMD. Sedimentation rate varied between 7–25 mm/hr (Nl, 0–20).

Family history was negative for skeletal or rheumatologic disease and remarkable only for short stature in a maternal great-grandmother (4’10”), maternal great-aunts, and a paternal great-aunt.

Physical exam:

At age 11 years, our patient’s height and arm span were 50th percentile for girls ages 6 and 7 years, respectively. At age 21 years, her length was 138 cm, arm span 156 cm (10th percentile for women), and sitting height 52.3 cm. She had an OI-like voice. There was a triangular face, mild hyperteloric appearance, bulbus nasal tip, short neck, restricted range-of-motion of neck and hips, barrel chest, scoliosis, hypermobile elbows, and leg-length discrepancy (Figure 1).

Figure 1: Patient at age 13 years.

Note, she has no joint contractures. Her right knee is flexed because of hip malformation.

B). Genetic Testing:

Informed written consent for genetic testing was approved by the Human Research Protection Office, Washington University School of Medicine; St. Louis, MO, USA (WUSM). Then, peripheral blood leukocyte DNA was obtained from the patient, her parents, sister, and brother using the Puregene DNA Extraction Kit (Gentra Systems, Minneapolis, MN, USA). In 2006, negative sequencing of COL1A1 and COL1A2 had been performed at Athena Diagnostics (Marlborough, MA, USA). In 2008, negative copy number microarray was reported by Signature Genomics (Spokane, WA, USA). During 2010–11, we performed Sanger sequencing for PPIB and FKBP10 at the National Institute of Child Health and Human Development (NICHD), and for LRP5 [osteoporosis pseudoglioma syndrome (OPPG)]⁽¹⁹⁾ as well as for IFITM5 [OI, type V]⁽¹⁾ at WUSM using primers we designed (sequences and PCR conditions available on request). Then in 2016, whole exome sequencing at the McDonnell Genome Institute, WUSM using Illumina libraries (Illumina, Inc., San Diego, CA, USA) and exome sequencing at 75X coverage (Integrated DNA Technologies, Inc., Skokie, IL, USA) underwent subsequent data analysis at the University of Montreal, Montreal, Quebec, Canada. Data processing, alignment (using a Burrows-Wheeler algorithm, BWA-mem v0.7.5a), variant calling, and annotation were performed with the Genome Analysis Toolkit (GATK) UnifiedGenotyper v2.6–4 tool (https://www.broadinstitute.org/gatk/guide/tooldocs/org_broadinstitute_gatk_tools_walkers_genotyper_UnifiedGenotyper.php). Variant annotation used an in-house pipeline and Annovar v2014-11-12 (http://www.openbioinformatics.org/annovar/). Human and mouse phenotypes for each gene were annotated using Orphanet (https://www.orpha.net/consor/cgi-bin/index.php) and OMIM⁽²⁾ (https://www.omim.org/), and the protein function using Uniprot (https://www.uniprot.org/). Confirmatory Sanger sequencing of the PLOD2 findings and family studies were performed in our research laboratory at WUSM.

C). Serum Multiplex Biomarker Profiling:

Fasting serum obtained from the patient twice, at age 13 years in 2010 and at age 14 years in 2011, was archived at −40°C until 2013 when studied by serum multiplex biomarker profiling (SMBP) of 22 analytes at Amgen, Inc. (Thousand Oaks, CA, USA). At both times, she was intermittently receiving alendronate. The studies used our published methodology^(20,21) in single “batch” assays to investigate many different rare metabolic and dysplastic bone disorders (Supplementary Appendix).⁽²⁰⁾ Results were contrasted to our SMBP values acquired in 2012 for fasting sera from 36 healthy children and adults (12 males and 24 females).^(20,21)

V). Results:

A). Radiological Studies:

Sequential radiographs revealed worsening osteoporosis from ages 4–10 years as well as significant progression of other skeletal abnormalities including scoliosis (Figures 2–4). Worsening osteoporosis seemed independent of immobilization, with vertebral compressions occurring from age 7 years except for some reconstitution during the first years of alendronate treatment. The patient’s skull showed many Wormian bones, osteopenia, basilar impression, and platybasia. Basilar invagination at age 4 years progressed. The cervical spinal canal was wide. After age 10 years, BMD fluctuated with intermittent alendronate treatment, but was always subnormal.

Figure 2: Skull and Lower Limb Radiographic Findings:

Lateral skull radiograph (A) shows multiple Wormian bones (arrows), osteopenia, basilar impression, and platybasia. Lower extremity radiograph (B) at age 15 years shows genu valgum, gracile osteopenic long bones, and orthopedic hardware in the proximal left femur.

Figure 4: Progressive Pelvic Deformity.

Pelvis/proximal femur radiograph at age 9 years (A), 3-D CT reconstruction at age 13 years (B), and radiograph at age 20 years (C) document severe progressive right protrusio acetabuli.

C). Biochemical Findings:

Despite 11 inpatient evaluations at our Research Center between ages 10 and 21 years including counseling by our dieticians, we were unsuccessful in improving her poor dietary calcium intake. In 24-hour urine collections, calcium/Cr ratios ranged ~ 30–70 mg/g with normal sodium excretion, and phosphorus/Cr ratios ranged 0.2–1.3 (Nl, 0.8–1.8). Nevertheless, serum total and ionized calcium, phosphorus, magnesium, alkaline phosphatase, 25(OH)D, PTH, and osteocalcin levels were typically normal.

D). Iliac Crest Histopathology:

Transiliac crest biopsy at age 9.8 years had reportedly shown normal tetracycline double labeling and bone turnover, but half-normal cortical width and low cancellous bone volume without high numbers of osteocytes.

Transapophyseal iliac crest biopsy at age 13 years following three oral courses of tetracycline revealed some areas with abundant osteoid and widely-spaced tetracycline labels indicative of active bone formation (Figure 5A,C). However, the histopathological features were variable in the fragmented specimen, with other areas showing very poor trabecular connectivity and little osteoid or tetracycline labeling (Figure 5B,C). Some osteoclasts were rounded and off of the bone surface, consistent with alendronate treatment (Figure 5A, inset).

Figure 5: Transapophyseal Iliac Crest Biopsy Histology.

A fragmented specimen was obtained at age 13 years. Some fragments showed abundant osteoid on Goldner trichrome-stained sections (A) and widely spaced fluorescent labels (C) indicating active bone formation. Other fragments revealed low trabecular connectivity, minimal osteoid (B), and absence of tetracycline labels (D). Scale bar 200 μm for A,B; 50 μm for inset, 100 μm for C,D.

D). Mutation Analysis:

E). Serum Multiplex Biomarker Profiling:

SMBP of the patient was notable, among the 22 circulating biomarkers studied (Supplementary Appendix; Table 1), only for seemingly elevated adiponectin and TRAcP5 and decreased matrix metalloproteinases 1 and 8 (Figure 6).

Figure 6: Serum Multiplex Biomarker Profiling Abnormalities:

F). Mutation Analyses:

Sanger sequencing of several genes causing OI (COL1A1, COL1A2, PPIB, FKBP10, and IFITM5) and OPPG syndrome (LRP5) was negative. Whole exome sequencing of the genomic DNA of the patient, healthy parents, and two healthy siblings revealed 94 homozygous variants and 259 compound heterozygous variants. Many were filtered out because they were also found in unrelated exomes or in commonly identified genes. Variants in DSPP, encoding dentin sialophosphoprotein 1, were identified in both homozygous and compound heterozygous forms in the proband, however DSPP contains many difficult to sequence, repetitive sequences and our findings perhaps reflected false-positive signals. Notably, mutations in DSPP cause dentinogenesis imperfecta [22], and our patient had brown primary teeth with healthy appearing adult teeth consistent with this complication. Other than the PLOD2 variants, no defects were identified in other candidate genes identified in other candidate genes. Therefore, the two missense variants in PLOD2 validated by Sanger sequencing in exon 8 (c.797G>T, p.Gly266Val) and exon 12 (c.1280A>G, p.Asn427Ser) would explain her AR skeletal disorder (Figure 7)

Figure 7: Patient and Parent Electropherograms.

Electropherograms of PLOD2 exons 8 and 12 show the patient is compound heterozygous for different missense mutations individually carried by her heterozygous parents.

In ExAC, low frequency (Gly266Val, 0.0000419) and absence (Asn427Ser) implicated both PLOD2 variants as pathogenic; i.e., mutations. With Polyphen, the Asn427Ser defect was predicted to be possibly damaging with a score of 0.843 (sensitivity: 0.83; specificity: 0.93).⁽²³⁾ SIFT predicted the substitution at position 427 from Asn to Ser to affect protein function with a score of 0.00.⁽²⁴⁾ Using Polyphen, the Gly266Val mutation was predicted to be probably damaging with a score of 0.998 (sensitivity: 0.27; specificity: 0.99), whereas SIFT predicted the substitution Gly266Val to affect protein function with a score of 0.03.^(23,24)

The father carried the exon 8 mutation (Gly266Val) whereas the mother and sister (not shown) carried the exon 12 (Asn427Ser) defect (Figure 7). All three individuals considered themselves well, as supported by their medical histories. A healthy brother carried neither mutation (not shown).

VI). Discussion:

At about age 3 years, our patient presented with an OI-like disorder featuring osteoporosis, recurrent fractures, and severe progressive scoliosis without large joint contractures or pterygia. However, mutations were not found in the genes most often causing OI (COL1A1 and COL1A2),⁽¹⁾ or in several less common OI genes (PPIB, FKBP10, and IFITM5),⁽¹⁾ or in the gene associated with OPPG (LRP5).⁽¹⁹⁾ PLOD2 was not sequenced early on because she had no contractures or pterygia common in BRKS2. Instead, the etiology of her disorder involving compound heterozygosity for two missense mutations in PLOD2 (one not previously reported in BRKS) was revealed by exome sequencing using DNA from the patient, her parents, and two unaffected siblings.

Autosomal recessive PLOD2 mutations underlie BRKS2 (Table 1, Figure 8).^(2,7–16) At least 17 different PLOD2 mutations representing 16 different probands/families with classic BRKS2 (i.e., with contractures/pterygia) are reported from Afghanistan, Australia, Europe, Middle-East, China, India, and recently in North America (Table 1).^(7–16,25) Most (i.e., 11) BRKS2 probands/families harbored homozygous PLOD2 mutations suggesting consanguinity (or regional mutations), whereas the remainder (i.e., 5) carried compound heterozygous defects (Table 1).

Figure 8: PLOD2 Mutations:

PLOD2 mutations causing BRKS2 and OI variants. The PLOD2 gene structure is shown with intron-exon spacing to scale with the exception of the large intron 1 removed. The exons are numbered 1–19 with the alternatively spliced exon 13a. Mutations shown above the gene diagram are those associated with typical BRKS2, while those below are associated with OI variants or other related skeletal diseases (see Table 1). The PLOD2 mutations found in our patient are enclosed in boxes.

Among the 12 BRKS2 PLOD2 missense mutations, several cluster in exon 17 (Figure 8) encoding the LH domain^{(7–16, 26−38)} where relatively many represent a narrow region of PLOD2 amino acids 610 to 629.⁽²⁸⁾ One defect in PLOD2, His687Arg, involves a Fe²⁺-binding ligand required for LH dimer assembly and LH enzymatic activity.⁽²⁸⁾ Several nonsense, frameshift, or splice site mutations in PLOD2 occur throughout the coding region and terminate translation.^(7–13)

However, in 2012 Puig-Hervas et al.⁽¹¹⁾ reported a patient with OI without contractures and homozygous for a PLOD2 mRNA splice site mutation. They also reported two brothers, one with mild BRKS (with contractures) and the other with mild OI without contractures, both found to be compound heterozygotes for identical PLOD2 defects. In 2016, PLOD2 defects were identified in a patient with a mild OI phenotype without contractures.⁽¹⁷⁾ In 2018, Leal et al.⁽¹³⁾ reported additional individuals with PLOD2 mutations and skeletal disease resembling OI yet without joint contractures. Included were two patients with moderately severe OI, one of whom had an older affected sister with congenital contractures.⁽¹³⁾ Additionally, BRKS1/OI type XI due to FKBP10 mutations also shows variability of phenotype ranging more broadly than BRKS2 and spanning from OI, OI plus contractures, including siblings with the same defect with or without contractures, to a contracture phenotype alone (Kuskokwim Syndrome).^(30,31) BRKS2 clinical variants have now been reported in patients representing multiple geographic regions (Egypt, Europe, Brazil, and Palestine). Our patient is the first identified in North America as a BRKS2 clinical variant without contractures. She represents the fourth report of an OI skeletal phenotype without joint contractures due to PLOD2 mutations. Her PLOD2 mutations are in exons 8 and 12 and presumably lie outside the LH domain. One other “non-classic BRKS2” PLOD2 mutation (Asp585Val) localizes to the dimerization interface.⁽²⁸⁾ In fact, one report,⁽¹³⁾ Trp588Cys, associated with moderate OI without contractures, is nearby. However, homozygosity for our patient’s exon 8 PLOD2 mutations has led, in another patient, to classic BRKS2 with contractures and pterygia.⁽¹⁵⁾ Indeed, a genotypic explanation for the presence or absence of contractures may be difficult to find, as siblings with identical PLOD2 mutations can be discordant for contractures.^(11,13)

SMBP of 22 biomarkers of bone metabolism and remodeling indicated adiponectin was increased in our patient compared to age- and sex-matched controls. In 2003, inverse correlation between adiponectin and BMD was reported by Lenchik et al.,⁽³²⁾ and in 2010 by Reid.⁽³³⁾ In 2011, meta-analysis of four adipokines (adiponectin, leptin, resistin and visfatin) on BMD and fracture risk in healthy adults revealed inverse correlation between circulating adiponectin and BMD at the lumbar spine, total hip, and total body in postmenopausal women and in men,⁽³⁴⁾ and that “adiponectin is the most relevant adipokine negatively associated with BMD, independent of gender and menopausal status”. Now, numerous studies support an inverse relationship between BMD and adiponectin, with adiponectin level as an independent fracture risk factor.⁽³⁵⁾ Of interest, TGFβ1, which has been elevated in other forms of OI,⁽³⁶⁾ was normal.

In conclusion, atypical BRKS2 caused by AR PLOD2 mutations should be considered in pediatric patients with unexplained osteoporosis or OI.

Figure 3: Axial Skeleton Radiographic Findings:

Anteroposterior (A) and lateral (B) spine radiographs show severe scoliosis. Vertebrae are osteopenic with endplate sclerosis reflecting alendronate therapy. Severe protrusio acetabuli is apparent at the unoperated right hip.