Noonan syndrome is a relatively common genetic syndrome occurring in an estimated 1 in 1,000–2,500 live births (1). Mutations in the Ras/mitogen-activated protein kinase (RAS/MAPK) pathway, which regulates cell growth, differentiation, senescence, and death, have been identified in Noonan syndrome. Multiple giant cell lesions (MGCLs), also known as central giant cell lesions or giant cell granulomas, are non-neoplastic lesions characterized by a proliferation of granulation tissue containing multinucleated giant cells embedded in a fibrous stroma (2). MGCL is a rare but typical complication of a dysregulated RAS/MAPK pathway, including Noonan syndrome, termed NL/MGCLS, and shares phenotypic characteristics of cherubism (3, 4). The underlying genetic mutations and sequelae distinguish the two. Mutations in PTPN11 (5, 6) and SOS1 (2, 7, 8) have been reported in people with NL/MGCLS, whereas mutations in SH3BP2 are associated with cherubism (9). Giant cell lesions in cherubism tend to spontaneously resolve, whereas those observed in NL/MGCLS can have aggressive signs and symptoms (10–12). MGCLs most commonly occur in the jaw and joints (pigmented villonodular synovitis) (2).

MGCLs, as opposed to giant cell tumors of the bone, are benign but locally aggressive tumors that cause osteolytic destruction of bone (13). Multiple medical treatments have been explored with variable efficacy (11, 12). Surgical resection, similar to the tumor itself, often leads to disfigurement, especially in cases with jaw involvement. Enucleation and curettage has also been used but is associated with a high recurrence rate and does not completely avoid the morbidity associated with resection (14). An effective medical treatment option would be preferential. Both MGCLs and giant cell tumors are composed of osteoclast-like giant cells that express the receptor activator of nuclear factor kappa B (RANK) and mononuclear stromal cells that express RANK ligand (RANKL) (15), leading to osteoclast activation (16). Denosumab, a RANKL monoclonal antibody that inhibits osteolytic activation, is approved for malignant giant cell tumors in adults (16). Safety and efficacy of denosumab in MGCLs is limited to case reports, which have demonstrated successful avoidance of surgery and radiologic improvement with denosumab treatment. Only one case in an adult with Noonan syndrome has been published to date (17). Further data are needed, particularly in the pediatric population, to understand the efficacy and potential side effects on the growing skeleton.

This retrospective case series describes the safety and efficacy of denosumab treatment in four children (3, 17, 8, and 13 years old) with NL/MGCLs of the jaw. The diagnosis of Noonan syndrome was based on clinical phenotype in three of the patients, whereas one patient was recognized only after genetic testing for MGCLs. All patients had symptomatic lesions affecting tooth positioning, chewing, and/or facial disfigurement. Surgical and medical options were discussed, including surgical curettage, calcitonin, bisphosphonate, interferon, and denosumab. After discussions between the care team and families, the decision was made to try denosumab. Baseline labs and Dual-energy X-ray absorptiometry (DXA) were obtained. Weight-based dosing was calculated by using the approved adult dose (120 mg per month) and dividing by an average adult weight of 70 kg. Legal guardians of all patients signed written informed consent for inclusion in this report. Baseline characteristics of the four patients are shown in Table 1. NL/MGCLs improved grossly and radiographically within 6 months (Figure 1). Complications of treatment included joint pain, hypocalcemia, and hypercalcemia when initiating and discontinuing denosumab, respectively (Figure 2).

There is significant phenotypic heterogeneity in Noonan syndrome among our patients. Cardinal features include short stature, cardiovascular defects, broad or webbed neck, pectus deformity, developmental delay, cryptorchidism, and characteristic facies (18). In our cohort, three of the four patients had an established diagnosis of Noonan's syndrome prior to development of MGCLs, whereas the Noonan phenotype was not overtly obvious and diagnosis was not made until presentation with MGCLs in one of the four. Although rare, giant cell tumors of the jaw, which can arise from dysregulation of RAS/MAPK pathways, should also be considered part of the phenotypic presentation of Noonan syndrome (7). Given the differing clinical courses and potential additional systems that may be affected, genetic testing should be considered in children presenting with giant cell tumors of the jaw with specific testing for mutations in PTPN11, SOS1, and SH3BP2 (2, 5–9).

Giant cell tumors are composed of osteoclast-like giant cells that have formed secondary to stromal cell expression of RANKL. Denosumab, a monoclonal antibody against RANKL, has been approved by the United States Food and Drug Administration and by the European Medicines Agency following a clinical study evaluating the efficacy for malignant giant cell tumors in skeletally mature adults and adolescents (19). Given the unknown and unstudied effects of denosumab on growing bones, the use of denosumab in children has been limited. To date, we have found reports of 17 children and adolescents in the literature who have been treated with denosumab for various indications, including osteogenesis imperfecta type VI (20), juvenile Paget's disease (21), giant cell tumors (22–26), fibrous dysplasia (27), osteoglyphonic dysplasia (28), cherubism (17), and non-syndromic giant cell lesions of the jaw (17, 26). The use of denosumab in NL/MGCL has been reported in one adult (17). The safety and efficacy in non-skeletally mature people with NL/MGCL have not previously been reported.

All of our patients showed resolution of symptoms associated with their mandibular NL/MGCLs, likely because of the regression of MGCLs within the mandible. In all cases, the MRI, CT scans, and dental orthopantograms showed regression of the MGCLs and improvement in the radiographic appearance of mandibular bone. As shown in Figure 2, there was a significant increase in the radiodensity of the mandibular bone. The improved radiographic appearance of these lesions is suggestive of normalization of mandibular bony architecture; however, this cannot be concluded without a biopsy of the newly formed bone within the previous lesions. We achieved very similar results with denosumab in our patients with NL/MGCLs relative to patients with non-syndromic central giant cell lesions (17, 26). Using denosumab, we were able to avoid drastic surgical interventions, which risk extensive facial disfigurement, damage to developing teeth, injury to inferior alveolar nerve, and altered growth of the mandible. Although imaging studies showed improvement of the bone density in the lesions, the bony expansion caused by these lesions did not reduce. Some patients may require recontouring of the facial bone to provide improved esthetics.

Observed side effects were similar as previously described, including hypocalcemia during treatment and rebound hypercalcemia with discontinuation (17, 20–27). Namely, three of the four patients experienced hypocalcemia with the initiation of denosumab, noted within 1 week after medication administration, which was corrected with calcium supplementation. One patient who had non-adherence to calcium supplementation had more severe protracted hypocalcemia, which improved with the addition of calcitriol. None required hospitalization for hypocalcemia. All four cases experienced rebound hypercalcemia when discontinuing denosumab. The timing of hypercalcemia occurred between 67 and 80 days after the last monthly denosumab dose (Figure 2A). Two of four patients were treated with additional doses of denosumab, which quickly resolved the hypercalcemia, but recurred again 57–90 days later. All four patients required treatment with a bisphosphonate. Three had significant enough hypercalcemia to warrant hospitalization and treatment with bisphosphonate infusion, whereas hypercalcemia was prevented in one case by using alendronate orally preemptively beginning about 50 days after the last denosumab dose and continuing for 1 full month. One case had further hypercalcemia after zoledronate, which was managed with alendronate 70 mg weekly orally. No patient developed osteonecrosis of the jaw.

As the effect of denosumab on the growing skeleton remains unknown, extreme caution was utilized in monitoring for potential skeletal side effects. All four cases were skeletally immature during the denosumab treatment course. Height z-score remained stable (Case 1 and 3) or increased (Case 2 and 4). Height z-score increase (+1 standard deviation score) in Case 2 was attributed to concurrent treatment with growth hormone. At baseline, BMD assessed by DXA was low in two of four cases, and although within the normal range, below the 50th percentile for the other two cases. BMD increased in all patients but did not exceed the normal range. All four cases experienced extremity/joint pain coinciding with hypercalcemia. One case also reported extremity/joint pain during treatment with denosumab that was most intense in the week following injections and improved within 10 days. X-rays were obtained in three of the four cases and were all notable for metaphyseal sclerosis and widening of the growth plates, similar to the bone changes observed in children treated with bisphosphonates (29) (Supplemental Figure 1). Noonan syndrome is also associated with pigmented villonodular synovitis, a benign proliferative disorder of the large synovial joints or of the tendon sheath of small joints. Presenting symptoms including painless swelling of the joints (30). The resolution of symptoms within a few days of each case did not warrant tissue biopsy to evaluate for this but may be a unique risk factor in the NL/MGCL population.

Osteonecrosis of the jaw (ONJ) has been noted to occur in up to 1% of adults receiving denosumab treatment for giant cell tumors (19), and one case has been described in the literature describing ONJ in a child (25). To reduce risk of ONJ, all patients were counseled on good dental hygiene. They were instructed to continue with routine dental care, including brushing and flossing twice a day, and encouraged to set up regular checkups with their dentist for routine cleanings and necessary care. In this case series, none of the patients developed ONJ, even with the targeted lesions being within the jaw. There is suspicion that the risk of ONJ may be associated with a cumulative dose. In the Uday case, the patient who developed ONJ had an unresectable sacral giant cell tumor that was treated with denosumab, 120 mg loading, then monthly for a total of 46 doses, with a total treatment dose of 5,520 mg. In comparison, cases 1–4 received a total denosumab dose of 460, 1,080, 210, and 390 mg, respectively, which is significantly less. In NL/MGCLs, dosing denosumab using a weight-based dose may help reduce risk of adverse effects while still maintaining efficacy.

We report that a weight-adjusted denosumab dose improved NL/MGCLs. There is limited information in the literature regarding the medical treatment of NL-MGCLs, and there are no randomized clinical trials comparing or showing efficacy of specific medical treatment for NL-MGCLs. In this case series, denosumab treatment has shown excellent results. Alternatively intra-lesional injection of corticosteroids, systemic calcitonin, and interferon has been reported for the treatment of central giant cell granulomas of the jaw; but, they have shown variable results (31, 32). More recently, oral imatinib has been used for treatment of isolated central giant cell lesions and those associated with cherubism (33); however, the data are limited to case reports and small series. The patients with NL/MGCLs in this study have shown excellent response in regression of the tumor with minimal side effects using denosumab. Strengths of the study include usage of a uniform dose among cases for a minimum of 6 months. Limitations of the study include limited sample size, lack of control for comparison, and variation in monitoring for safety and efficacy. Despite the limitations, all four patients demonstrated a significant improvement in NL/MGCLs and were able to avoid further surgery.

Interdisciplinary care, including genetics, oral surgery, and endocrinology, was an important factor in monitoring both safety and efficacy. To avoid hypocalcemia while on denosumab, laboratory data support the optimization of 25-OH vitamin D status prior to and supplementation with calcium if history elicits a calcium deficient diet. At the cessation of denosumab therapy, hypercalcemia occurred within a finite window of 10–11 weeks after the last dose with delay of hypercalcemia noted with preemptive treatment strategies. Based on our collective knowledge gained in these four cases, a proposed treatment regimen for further long-term studies to evaluate safety and efficacy of denosumab in NL/MGCLs would include monthly subcutaneous denosumab 1.7 mg/kg for 6 months. Serum calcium should be monitored periodically throughout treatment and at least 4 months after cessation. Monitoring urine calcium to creatinine ratio may be useful after the cessation of therapy as well as hypercalcemia appears to occur after renal calcium excretion capacity is exceeded. Administration of a bisphosphonate (either orally or intravenously) 8–10 weeks after cessation of denosumab may help prevent rebound hypercalcemia. On-going monitoring of NL/MGCLs continues and is necessary to assess for disease recurrence.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by Johns Hopkins Institutional Review Board, Baltimore, MD 21205. Written informed consent to participate in this study was provided by the participants' legal guardian/next of kin. Written informed consent was obtained from the minor(s)' legal guardian/next of kin for the publication of any potentially identifiable images or data included in this article.

KF, BS, YY, and JLC wrote the manuscript. KF, BS, JC, and JLC collected the data, KF and SK organized and data, and KF, YY, JC, JN, MB, and JLC followed up the case. MB and JLC supervised the management and follow up of the case. All authors revised and approved the final manuscript and agreed to be accountable for the content of the work.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.