New-generation anticancer drugs and medication-related osteonecrosis of the jaw (MRONJ): review of the literature and single-center experience.

Since its first association to Zoledronate in 2004, the number of drugs which may induce medication-related osteonecrosis of the jaw (MRONJ) increased annually. Nowadays, novel chemotherapy agents, such as immunotherapy or targeted therapy are increasingly used in the clinical practice. Adopted for Metastatic Melanoma (MM) therapy, immunotherapy and targeted therapy agents are now used in a wide and increasing variety of cancers, such as head & neck, urologic, gynecologic and pulmonary neoplasms, especially in regard to bone metastasis control.  We thus investigated literature to understand a possible impact of such novel cancer therapies on MRONJ. A review of the literature searching PubMed and Scopus was performed. We also report of a case of late MRONJ onset, 3 years after administration of ipilimumab. As a result, association of immunotherapy and/or chemotherapy and/or targeted therapy, in sequence or as single therapies, may rarely induce osteonecrosis of the jaw.  Ipilimumab was associated 3 times in literature to MRONJ, both in single therapy and in association with other drugs. In a paper published by our Institute, Ipilimumab produced late effects, inducing MRONJ 3 years after its administration; complete healing was reached with pharmacological therapy. Nivolumab was associated once in literature to MRONJ, both in patients undergoing multiple therapy both in single therapy. Pembrolizumab was reletad twice to MRONJ in literature, once when in association with epacadostat and once with Denosumab. Interestingly, 5 out of 7 reported cases were triggered by periodontal abscess. New-generation anti-cancer drugs, such as immunotherapy and targeted therapy agents, seems at low risk of developing MRONJ. Their field of application is widening, and so is the number of patients who are administered with such drugs. Yet, mechanisms underlying their action are not fully comprehended and, even if they may be considered safe, literature suggests that unexpected events in terms of MRONJ break-out may occur.  Multidisciplinary approach, with dental evaluation prior/during/after chemotherapy, is strongly advised.

complete healing was reached with pharmacological therapy. Nivolumab was associated once in literature to MRONJ, both in patients undergoing multiple therapy both in single therapy. Pembrolizumab was reletad twice to MRONJ in literature, once when in association with epacadostat and once with Denosumab. Interestingly, 5 out of 7 reported cases were triggered by periodontal abscess.
New-generation anti-cancer drugs, such as immunotherapy and targeted therapy agents, seems at low risk of developing MRONJ. Their field of application is widening, and so is the number of patients who are administered with such drugs.
Yet, mechanisms underlying their action are not fully comprehended and, even if they may be considered safe, literature suggests that unexpected events in terms of MRONJ break-out may occur.
Multidisciplinary approach, with dental evaluation prior/during/after chemotherapy, is strongly advised.

Background
Events of nonhealing-exposed bone in the maxillofacial region were initially reported in patients treated with intravenous (IV) bisphosphonates (BP). 1 Pamidronate (Aredia) and zoledronic acid (Zometa)-two IV BPs-were thus labeled as at risk for osteonecrosis of the jaws (ONJ). 2 Consequently, in the following years a warning followed for all BPs to be at risk for ONJ, which was thus renamed as bisphosphonate-related ONJ (BRONJ). 3 During subsequent years, other BPs and medications belonging to other classes have been related to the development of ONJ, including denosumab (humanized monoclonal antibody blocking the acti-vation of receptors for nuclear factor kBligand), bevacizumab (humanized monoclonal antibody), and sunitinib (tyrosine kinase inhibitor). [4][5][6][7][8][9][10] As sporadic events, case reports have highlighted possible association between ONJ and azacitidine, imatinib, everolimus, zivaflibercept, ipilimumab, and tocilizumab. [11][12][13][14][15][16][17][18] As the number of drug class widened during the years, this condition was then more widely re-named as medication-related osteonecrosis of the jaw (MRONJ). 1 Both pathogenesis and associated risk factors of this disease are not fully comprehended. Nowadays, novel chemotherapy agents, such as immunotherapy or targeted therapy are increasingly used in the clinical practice. Adopted for Metastatic Melanoma (MM) therapy, immunotherapy and targeted therapy agents are now used in a wide and increasing variety of cancers, such as head & neck, urologic, gynecologic and pulmonary neoplasms, also in regard to bone metastasis control.
We thus investigated literature to understand a possible impact of such novel cancer therapies on MRONJ and report our experience in managing an unusual case of osteonecrosis onset 3 years after the administration of ipilimumab.

Literature analysis
Review of the present scientific literature was performed on the research engine SCOPUS and PUBMED. She showed no extra-oral sign of swelling nor abscess. Clinically, intraoral inspection showed presence of a nonhealed alveolar socket; the bottom and the walls of the alveolus were clearly visible, made of nonvascularized nonsuppurated bone, surrounded by swollen mucosa (Figure 1).
Qeios, CC-BY 4.0 · Article, April 9, 2021 Qeios ID: 48DL31 · https://doi.org/10.32388/48DL31 3/12   Anamnesis was carefully harvested. The patient had been diagnosed with cutaneous melanoma in 2009. She thus underwent surgical resection of primary tumour. Lung progression of disease was found in 2015, which lead to treatment with ipilimumab (3 mg/kg mg iv, every 3 weeks for 4 cycles). Complete remission of the disease was reached. In 2017, the patient suffered from hepatic progression; as Melanoma was BRAF-mutated, she underwent dabrafenib + trametinib (300 + 2 mg per os/die). The patient thus suffered from G. 3 toxicity (fever),which lead to treatment stopping and replacement with vemurafenib + cobimetinib (vemurafenib: 1920 per os/die for 3 months, then 1440 mg per os/die; cobimetinib: 60 mg per os/die for 3 weeks then 1-week pause). Such treatment was still ongoing at the time her dental problem. She reffered no history of smoking nor head and neck radiation therapy. Ipilimumab was the only medication administered to her that has been related to MRONJ 12,17 . Her MRONJ was thus evaluated as a "stage 2" according to the AAOMS classification, showing "Exposed and necrotic bone(…) with evidence of infection, (…) symptomatic" 1 . She was thus administered with amoxicillin + metronidazole (3 + 1.5 g per os/die) and chlorhexidine 0.2% mouth rinse twice a day; paracetamol (1 g per Qeios, CC-BY 4.0 · Article, April 9, 2021 Qeios ID: 48DL31 · https://doi.org/10.32388/48DL31 5/12 os) was prescribed for pain control.

Literature analysis
Seven papers (6 case reports and 1 case series) did not match the exclusion criteria and could be included in our study 12,17,[19][20][21][22][23] (full-texts are added as supplementary data of the present article). Form the case series 17 , 2 cases were treaceable to Ipilimumab (one in single and one in multiple therapy), but one was already reported in another study 12 Table 2.  have caused MRONJ in single (nivolumab, ipilimumab 12 ) or multiple (epacadostat+pembrolizumab) therapy.
Only one MRONJ case (ipilimumab, single therapy three years before) onset after tooth extraction; the other cases arose were reported to be in relation to periodontal/peri-implant conditions.

Personal experience: case report
At the 2-week follow-up visit, clinical improvement was clearly visible. She referred that a 10 × 5 mm bone sequestrum had been spountaneously ejected after 6 days of therapy and that her symptoms had therefore disappeared; clinically, she still showed incomplete wound healing. Two additional weeks of therapy were thus administered and so the patient reached complete alveolar healing (Figure 3).  No further treatment for MRONJ was prescribed. Patient underwent monthly follow-up. During the 6th month, a new OPT confirmed alveolar healing. (Figure 4).

Discussions
Several new medications have been added to the potential cause of MRONJ drug list during the past years. In addition to well-known medication, as highlighted by our review, MRONJ may be a major adverse reaction to several new-generation anticancer drugs. These drugs may have unexpected mechanisms, being their pharmacodynamic not fully comprehended yet.
Pembrolizumab, a humanized monoclonal antibody targeting the programmed cell death protein 1 (PD-1), has become the standard treatment for metastatic non-small-cell lung cancer (NSCLC) patients since it has been shown to prolong progression-free survival and overall survival of these patients 21 . In one report 21  Epacadostat, which inhibits IDO1 increasing the efficacy and growth of T-Cells and NK cell growth with increased production of IFNg, which inhibits osteoclastogenesis. Therefore, up until now, there is no evidence that Pembrolizumab or Epacadostat may cause MRONJ in single therapy.
Alongside with Pembrolizumab, Nivolumab is an anti-PD-1 checkpoint inhibitor. Our literature research showed two cases of MRONJ possible realted to this drug, one in single therapy 23 and on in multiple therapy 22  Interestingly, in the case report we have described in this paper too 19 , Ipilimumb may have caused MRONJ 3 years after its administration. Ipilimumab is known to have a 14.7-day blood halflife 19 but the real advantage of the drug is in the longterm efficacy, due to the immune responses induced by checkpoint inhibitors. Similiarly like the anticancer effects, side effects such as pruritus, diarrhea, vitiligo, hepatitis, and endocrinopathies may occur many years after administration 12 . It is conceivable that MRONJ may be also a late side effect under certain circumstances, which may have been co-caused by the ongoing target therapy (vemurafenib + cobimetinib) of the patient. The effect of BRAF and MEK inhibitors in BRAFmutant melanoma can lead to an immune-stimulating microenvironment by enhancing expression of immune-stimulating molecules and cytokines, reducing immunosuppressive cell populations, and decreasing immunosuppressive cytokines.
The cell damage to the tumor by the target therapy may have induced a tumor-antigen spreading, restimulating T-cell activity whose response had been increased and modulated by the effect of ipilimumab. Moreover, it has been demonstrated that anti-BRAF therapy enhances the reactivity and cytotoxicity of T cells 19 .The re-activation of such empowered T-cell clones may have lead the patient into a window of time in which she was at risk for MRONJ, similarly to when the patient was on treatment with ipilimumab.

Conclusions
MRONJ pathologic pathways are not fully comprehended. Yet, decreased bone turnover by modulation of osteoclast function, inhibition of angiogenesis, infection, inflammation and soft tissue toxicity have been identified as playing an important role 24 . It is well acquired in scientific literature that T-Cells may interfere with osteoclasts both with cell-to-cell Qeios, CC-BY 4.0 · Article, April 9, 2021 contact and the production of different cytokines (CKs). Proinflammatory CKs are pro-osteoclastogenic, while antiinflammatory CKs inhibit osteoclastogenesis by direct or indirect RANKL inhibition. The immune system is thus involved in bone loss and bone regeneration 21 . Trauma from regular oral activity, oral surgery (eg, tooth extraction) or teeth/periodontal diseases, increasing the demand for bone to mend itself 12 , may result in localized bone necrosis in patients with altered immunologic state due to ongoing/past immunotherapy.
In our literature review, 5 out of 7 MRONJ cases were triggered by periodontal/periimplant abscess/acute inflammatory phase. Periodontal patients are well-known for having altered cytokines productions/immunologic cell function when not treated, as they stay in a local chronic inflammation state, which may also reflect on systemic inflammation state. This may both lead to novel prespective in investigating MRONJ pathologic mechanism, confirmig inflammation as a key moment in its comprehension, both underline the necessity of periodontal surveillance for patients undergoing immunotherapy.
New-generation anti-cancer drugs seem at low risk of developing MRONJ, with isolated cases reported in literature.
Yet, literature suggests that unexpected events in terms of MRONJ break-out may occur. The authors thus recommend caution and strict vigilance in the dental management of patients treated with chemotherapy interfering with immunological function of the patients. Multidisciplinary evaluation is thus strongly advised prior/during/after chemotherapy. The administration of the prophylactic antibiotic protocol (amoxicillin + metronidazole; 3 + 1.5 g per os/die) in case of oral surgery and/or continous dental surveillance may be arranged in accordance between the dental surgeon and the oncologist, with the best possible evaluation of both oral and systemic conditions. Such cooperation may reduce the occurrence of adverse events which, as we have shown in our review, may result in patient's discomfort, pain and, as reported in one paper, major complications such as mandible fracture.
Further studies are needed on a large number of cases, in order to fully comprehend the relation between newgeneration anti-cancer drugs and MRONJ.