Denosumab Treatment Does Not Halt Progression of Bone Lesions in Multicentric Carpotarsal Osteolysis Syndrome

Abstract

Here we report the use of denosumab, a monoclonal antibody against receptor activator of nuclear factor κB ligand (RANKL), as monotherapy for multicentric carpotarsal osteolysis syndrome (MCTO) in an 11.5-year-old male with a heterozygous missense mutation in MAFB (c.206C>T; p.Ser69Leu). We treated the subject with 0.5 mg/kg denosumab every 60-90 days for 47 months and monitored bone and mineral metabolism, kidney function, joint range of motion (ROM), and bone and joint morphology. Serum markers of bone turnover reduced rapidly, bone density increased, and renal function remained normal. Nevertheless, MCTO-related osteolysis and joint immobility progressed during denosumab treatment. Symptomatic hypercalcemia and protracted hypercalciuria occurred during weaning and after discontinuation of denosumab and required treatment with zoledronate. When expressed in vitro, the c.206C>T; p.Ser69Leu variant had increased protein stability and produced greater transactivation of a luciferase reporter under the control of the PTH gene promoter than did wild-type MafB. Based on our experience and that of others, denosumab does not appear to be efficacious for MCTO and carries a high risk of rebound hypercalcemia and/or hypercalciuria after drug discontinuation. © 2023 The Authors. JBMR Plus published by Wiley Periodicals LLC on behalf of American Society for Bone and Mineral Research.

Keywords: DENOSUMAB; MULTICENTRIC CARPOTARSAL OSTEOLYSIS SYNDROME; MafB; RANKL‐INHIBITOR.

Fig. 1

MAFB structure and associated MCTO mutations. (A) The upper panel shows the domain structure of human MafB protein, which comprises two main domains, an N‐terminal TAD and a C‐terminal DBD. The latter consists of an extended homology region (red), basic region (green), and leucine zipper (orange). The MAFB mutations reported to date in patients with MCTO have been exclusively clustered within the phosphorylation region (codons 54–71) in the TAD. Two regions representing histidine repeats (130–138 and 161–167) are shown in gray. (B) Amino acid alignments of the GSK3 motif among MAF orthologues. The priming Ser70 and four tandemly arranged phosphorylatable serine/threonine residues (Ser66, Thr62, Thr58, and Ser54) within the GSK3 recognition region at the N‐terminal TAD of MafB are noted. Amino acids that are replaced in MCTO patients are above red circles, with the affected codon in the patient described in this report above a blue circle. Mutations in homologous residues of the TAD of closely related genes encoding the large MAF proteins NRL (black boxes), MafA (green boxes), and Maf (blue boxes) are also shown that cause autosomal dominant retinitis pigmentosa, familial insulinomatosis, and Aymé‐Gripp syndrome, respectively. DBD = DNA‐binding domain; TAD = transactivation domain.

Fig. 2

Images of left wrist. (A). Frontal radiographs of left wrist demonstrate multiple erosions in the distal radius, carpus, and proximal metacarpals with progressive carpal fusion (*) and loss of joint space. In posttreatment, overgrown distal radius and ulna epiphyses and multiple growth recovery lines are visible. (B) T1 signal‐weighted MR images of left wrist demonstrate to advantage numerous erosions in the radius (thick black arrows), ulna, carpals (thin black arrows), and metacarpal bases (white arrow). By 2017, the lunate and triquetrum have fused. By 2020, very little carpal joint space remains. MR = magnetic resonance.

Fig. 3

Denosumab dosing schedule. Horizontal bar shows time in weeks, colors change with change in: dosing interval or weight‐based dosing. Denosumab dosing (mg/kg) denoted below the bars. Arrows above the bars point to time of change (weeks at which change was made are included). *Indicates denosumab was 1 month late. #Indicates zoledronic acid treatment (0.125 mg/kg).

Fig. 4

Wrist and elbow range of motion. Range of motion was measured with a digital goniometer and is shown over time (weeks). Mean of measurements taken in triplicate is shown; error bars (standard deviation). Wrist flexion (A), wrist extension (B), elbow flexion (C): • = left (line with circle), right (line with square), normal (straight line). elbow extension (D): • = left (line with circle), right (line with square), normal (line with triangle).

Fig. 5

Images of left elbow. (A). Frontal radiographs of left elbow demonstrate multiple progressive erosions in the humerus and ulna (black arrows) as well as the radius (white arrows) with progressive deformity of the distal humerus and overgrowth of the subluxed radial head. (B) (a, b, c, d) T1 signal‐weighted MR images of left elbow demonstrate multiple progressive erosions in the humerus and ulna (blue arrows) as well as the radius (orange arrows) in the coronal plane (a, b). By 2020, the patient had a fixed flexion deformity that prevented acquisition in standard planes (c, d). (e) Post‐contrast T1 signal‐weighted image with fat saturation shows synovitis (arrow) and radiocapitellar subluxation. MR = magnetic resonance.

Fig. 6

Biochemical markers. (A) Markers of bone formation: P1NP (reference range 22–105 ng/L); BSAP (reference range 48.4–155.5 until 1/16/19, then 27.8–210.9 ng/L); marker of bone resorption: CTX (reference range 553–2071 pg/mL until 1/16/19, then 485–2468). (B) PTH (reference range 12–65 pg/mL); serum calcium (reference range 8.5–10.6 mg/dL); urine calcium/urine creatinine (reference range less than 0.2 mg calcium/mg creatinine). BSAP = bone‐specific alkaline phosphatase; CTX = c‐telopeptide; P1NP = procollagen N‐terminal peptide; PTH = parathyroid hormone.

Fig. 7

Bone densitometry. Changes in bone density of L₁–L₄ by DXA from baseline to month 48. Bone mineral density is shown on the left axis and the Ht Z adjusted Z score is shown on the right axis. Treatments are shown above the graphs. Ht = height.

Fig. 8

In vitro studies of MafB. Biochemical studies (A) MafB protein levels were determined by immunoblot analysis of lysates from HEK293T transfected with the indicated expression vectors after 1–18 hours of CHX treatment. The blot shown is representative of three different independent experiments. (B) Levels of MafB were calculated as the ratio of MafB protein to β‐actin after reprobing the membrane in A and scanning images as described in Methods. MafB half‐life is based on linear regression analysis of transformed y values. (C) MafB‐dependent MARE‐specific transcriptional activity in HEK293T cells transfected with the indicated MafB expression vectors or an EV plus GCM2. MARE‐specific luciferase activity is determined as the ratio of PTH promoter‐Luc reporter activity relative to that produced a Renilla‐luc reporter, expressed as fold stimulation over PTH‐luc only. n = 5, *p < 0.05 for WT versus EV and **p < 0.001 for S69L versus EV and WT. CHX = cycloheximide; EV = empty vector; WT = wild‐type.