DISCUSSION
The results of this study suggest that Th-131I-MIBG administered shortly before HD CT BuMel and ASCT is a feasible, safe and effective regimen in patients with advanced NB. To our knowledge, this cohort is the largest described in the literature on HR-NB patients treated with this combination.
Treatment of high-risk NB is currently a challenge for pediatric oncologists. Indeed, despite the improvement obtained with recent protocols, event-free survival (EFS) remains unsatisfactory. Data from the Italian registry [3] on the outcome of NB patients over the years showed a steady improvement in OS in patients with lower stages of disease; in stage 4 patients, by contrast, after a significant increase from the 1980s to the 1990s (from 6.7% to 20.6%), OS remained stable in the early 2000s (29.3 %). More recently, data from the randomized HR-NBL1/SIOPEN trial showed a 3-year rate of EFS in HR-NB (stage 4) ranging from 33 to 49% according to the HD CT regimen (CEM vs BuMel), and indicated that BuMel was the better of the two conditioning regimens in stage 4 [2] . Moreover, the 5-year overall survival rate of HR-NB patients in the Children’s Oncology Group between 2005 and 2010 was 50% [17] . These poor outcomes mean that it is mandatory to explore new therapeutic approaches, since the benefit of intensified chemotherapy seems to be approaching the end of its potential.
Th-131I-MIBG has proved to be an effective and reliable targeted therapy in relapsed or newly diagnosed NB, with an overall response rate of about 30% in several studies[17-24] , the majority of which have concerned small series of patients treated with various schedules of Th-131I-MIBG, either alone or combined with various myeloablative chemotherapies. The rationale of Th-131I-MIBG lies in the propensity of NB cells to internalize 131I-MIBG within the cytoplasm and mitochondria through norepinephrine transporter channels, thereby delivering a lethal radiation dose to the tumor while limiting the damage to healthy tissues [23, 24] . Moreover, its activity seems to be increased by combination with other agents.
Previous studies have focused on the safety of combining Th-131I-MIBG and myeloablative treatments. Gaze et al.[25] reported the feasibility and good tolerability of Th-131I-MIBG combined with melphalan and total-body irradiation in 5 children with advanced NB. In larger numbers of patients, other groups [7, 26] found that the addition of Th-131I-MIBG to a CEM regimen did not affect the toxicity profile of CEM or bone marrow reconstitution after ASCT. Moreover, Lee et al. reported the results of incorporating 131I-MIBG treatment into tandem therapy with ASCT and chemotherapy in which CEC (carboplatin + etoposide + cyclophosphamide) and TM (thiotepa + melphalan) were used as the first and second HDCT regimens, respectively) [27] .
The combination of Th-131I-MIBG with HD CT BuMel has previously been reported by French et al. [28] . In 8 children with refractory NB who received HD BuMel 6-8 weeks after Th-131I-MIBG, no significant impact on treatment-related toxicity was seen and 5 of 7 evaluable patients with refractory disease achieved either a complete or partial response. More recently, in a case series of 9 patients,Ferry et al. [29]confirmed the safety of Th-131I-MIBG (up to 24 mCi/kg) and topotecan combined with HD CT BuMel, administered 2 months later.
In our study, a cohort of 28 children with advanced HR-NB, either newly diagnosed or relapsed, received Th-131I-MIBG (median dose 8.5mCi/Kg) shortly (median 17 days) before HD CT BuMel because of poor response to treatment. This combination proved feasible and well-tolerated, since the toxicities observed were no greater than would be expected after HD BuMel[2] and no toxic death occurred. Th-131I-MIBG did not seem to influence the time to engraftment or organ toxicities.
VOD/SOS was diagnosed in only 2 patients (7.1 %), who did not present any significantly different features. This incidence is considerably lower than that reported after Busulphan in the HR-NBL1/SIOPEN trial (18%) [2] but similar to that seen among NB patients in the Italian Pediatric Hemato-Oncology Association Group study on the incidence of VOD/SOS [30] .
Hypothyroidism was observed in more than one third of patients and can therefore be regarded as a high-incidence late complication after these treatments. This could be due to an increased risk of thyroid failure following high-dose chemotherapy with busulphan and melphalan in patients previously treated with Th-131I-MIBG, despite the administration of thyroid-blocking treatment before this latter.
That adding Th-131I-MIBG before HD CT can improve disease response or survival in HR-NB patients has already been hypothesized in previous studies [7-9, 21, 28, 29] . However, the retrospective nature of our study, which did not include a control group of patients with the same features who did not receive Th-131I-MIBG, prevents us from drawing any definitive conclusions regarding the effectiveness of this approach. Nevertheless, considering the dismal prognosis of the patients included, the overall disease response (67.8%) and the OS probability observed (53% and 41% at 3 and 5 years, respectively), we may suggest that the combination of Th-131I-MIBG given 2-3 weeks before HD BuMel and ASCT is a potentially valid approach in advanced NB patients with MIBG-avid disease.
Despite these encouraging results, our study has some limitations, such as its retrospective, single-center nature, which may have introduced a selection bias, thereby reducing the reliability of our data. Nevertheless, the main strength of this study lies in the number of patients treated by means of the same approach; this aspect is particularly valuable in a subgroup of patients with a very poor prognosis who require the collaboration of a multidisciplinary team, in that it yields useful information on the clinical effect of this treatment.
Another limitation concerns the relatively low median activity of131I-MIBG administered in these patients, in comparison with more recent therapeutic approaches [31,32] ; this may have reduced its potential antitumor effect. However, the doses of 131I-MIBG in our patients were in line with, or slightly higher than, those that had proved effective in previous studies in which 131I-MIBG was not combined with chemotherapy) [33].
Recent data on the positive role of immunotherapy with antibody anti-GD2 in improving survival in HR-NB [31, 32] could suggest that Th-131I-MIBG administration shortly before HD BuMel should be included in a sequential approach with immunotherapy in HR-NB patients with MIBG-avid disease. For this purpose, it would be advisable to centralize poorly responsive patients with MIBG-avid disease in centers with proven expertise in the administration of Th-131I-MIBG and in the management of HD CT and ASCT. Moreover, these centers should share protocols for treatment, thyroid blockade, patient surveillance and the monitoring of residual radioactivity.
Randomized prospective studies are needed in order to confirm the feasibility and the impact of the above-described strategy on the long-term outcome of HR-NB patients and to define both the131I-MIBG activity with the highest efficacy/toxicity ratio and the best timing for combination with high-dose chemotherapy. The next SIOPEN phase II protocol VERITAS (ClinicalTrials.gov Identifier: NCT03165292 ), which will compare topotecan plus Th-131I-MIBG versus HD CT with Thiotepa before HD CT BuMel in very HR NB patients, may provide useful information on these issues.