According to Weitzman classification patients had
According to Weitzman classification  15 patients had a single ONJ lesion (4A) and 9 patients had multiple lesions (4B) or a “multifocal” lesion (Table 2). Six patients interrupted the treatment with O3 gas application for disease progression (five patients) and for fear of an experimental therapy (one patient). The drop-outs were considered as failure in the Intention to treat (ITT) analysis, and excluded, in the per-protocol (PP) analysis. No patients reported adverse events during O3 insufflations and the days after; moreover no objective or subjective symptoms of intolerance to the O3 gas applications were observed. In 10 patients with the largest and deeper ONJ lesions, O3 gas therapy produced the sequestrum of the necrotic bone. Nine patients required more than ten O3 gas applications due to multiple lesions and/or abundant purulent AG 013736 notwithstanding the concomitant antibiotic therapy. In all these patients surgery was necessary to remove the necrotic bone after sequestrum. Of interest, surgical removal was possible without the resection of healthy mandible edge because of the presence of bone sequestrum. One patient (number 8) received two further insufflations while waiting for the surgery because he had bone sequestrum ready to be removed surgically; however because of sudden disease progression and Performance Status worsening, he was not operated on. Fig. 2 shows the ONJ lesion before starting O3 gas insufflations (picture A), the X-ray before the first therapy with O3 (it is possible to see the extent of necrotic lesion and the involvement of the loop of the mandibular nerve) (B), and the necrotic area post 10 gas insufflations, with the perfect soft tissue tropism, cannot be evaluated in the picture and the mobility of the necrotic area, easily detectable clinically (C). Fig. 2 also shows the post-insufflation X-ray (D), where the complete demarcation of the area of necrotic bone and its particular nearness to the emergence of the mandibular nerve can be observed. In this case, the nerve is not at all involved in removal and the patient will not suffer any loss of sensitivity or paraesthesia. The photograph of removal of the necrotic fragment (E) in the operating theatre can also be seen. After the elevation of the mucoperiosteal, vestibular and lingual flaps, the necrotic fragment can be more easily removed using a periosteal elevator, without having to use tools for cutting bone walls. This allows special softness during the procedure, without causing any surgical traumas to the bone treated with bisphosphonates. Six patients had complete sequestrum or partial (one patient) spontaneous expulsion of the necrotic bone with oral mucosa re-epithelisation, after 4 to 27 insufflations of O3 gas. Fig. 3 shows the patient number 3 before starting O3 gas insufflations, the necrotic bone removed by a pinch and re-epithelizated area after the O3 gas therapy. In the area 47 (A), archaeocyathids is possible to observe the infectious inflammatory damage to the soft tissues, despite the fact that the necrotic lesion is slightly infiltrating just below the mucosal margin. Picture (C) shows the mobile necrotic bone removed without anaesthesia during the seventh application, simply using an anatomical forceps in an outpatient setting. Insufflation treatment was performed as usual after removal. Picture (D) shows area 34 immediately after the removal of the necrotic bone in an outpatient setting. It is worth noting that the mucosal surface is completely different from the condition in regard with both colour and continuity. The area does not present any signs of bleeding as experienced in other cases, the mucous membrane has formed again between the healthy bone and the necrotic area, leading to the expulsion of the sequestrum. No ONJ relapse was observed in any of the 18 patients treated with O3 gas ± surgery, during a mean of follow-up of 18 months (range 1–3 years). After completion of the treatment, all patients were able to eat normally and prothesic dentures were adapted and re-positioned for those patients who had dentures. The patients restarted the therapy only if their Performance Karnosfsky Status was >70; the dental team visited them every 6 months.