Although survival rates for paediatric cancer have improved greatly during the past four decades, outcomes for children diagnosed with metastatic solid tumours have barely changed. These patients and their families are in desperate need of novel therapeutic approaches. Chemotherapeutic options and approaches might have come to their limit. Surgical resection of metastatic disease is largely unstudied and unproven with the exception of isolated recurrent pulmonary lesions for osteosarcoma, which can sometimes be cured with surgery alone. Radiation therapy is often used with demonstrated benefit to pulmonary metastases, although it is usually given at subtherapeutic doses because of whole lung radiotolerance.
Hepatoblastoma is the most common paediatric liver tumour in the first decade of life and is typically diagnosed in patients younger than 3 years. Surgical resection of the primary tumour is necessary as part of curative therapy whether by primary or delayed resection, including the use of orthotopic liver transplantation. Unfortunately only about a third of newly diagnosed patients are amenable to resection at diagnosis.1 Neoadjuvant chemotherapy can render tumours in many of these patients resectable, but typically a quarter to a third of such patients might never get to resection. About a quarter of newly diagnosed patients present with metastatic disease and typically have had survival rates less than 30%.1 Whether these metastatic patients have better outcomes if their pulmonary disease is eradicated by chemotherapy or by surgical resection has never been formally studied and is not clear or known.
In The Lancet Oncology, József Zsiros and his SIOPEL colleagues2 are to be congratulated for their report of improvement in outcome of patients with high-risk hepatoblastoma, many of whom had pulmonary metastases. 3-year event-free survival was 76% (95% CI 65—87) in the 62 patients studied (77%, 95% CI 63—90, in the 39 patients with metastasis). This finding is perhaps the most substantial improvement in survival that has been described for children with solid tumours and metastatic disease in decades. Because cisplatin is thought to be the most active drug in this disease, the investigators used a novel, well designed therapeutic schema that incorporated weekly dose-dense cisplatin chemotherapy—revisiting an old basic tenet of chemotherapy administration that has never been fully evaluated in paediatric malignancies.3
The use of dose intensification of chemotherapy has had some success in adult malignancies but has been used rarely in paediatric cancers.4 Interval compression of chemotherapy has been shown to be effective in acute myeloid leukemia and Ewing’s sarcoma, although with substantial costs and burden of care, and is now being explored for rhabdomyosarcoma.5, 6Trials in patients with osteosarcoma have shown positive benefits in some trials but no effect in others.7 A previous hepatoblastoma trial from the Children’s Oncology Group (P9645)8 attempted to intensify platinum delivered therapy by alternating cisplatin and carboplatin on an every 2 week schedule. Ultimately, this regimen was inferior to the standard control group, perhaps because of decreased dose intensity of cisplatin.
The SIOPEL-4 trial seems to be a success. 60 (98%) of 61 evaluable patients had at least a partial response from preoperative chemotherapy. 19 of 20 with metastatic disease who had a complete response to chemotherapy alone remain disease-free after surgery, with only one patient having a recurrence. In view of these striking results, these data need to be validated in a large trial. Zsiros and colleagues correctly point out the limitations of this study being a single arm trial that compared outcomes with historical results and therefore needing further study before the approach can be adopted as standard of care. Although the investigators state that their regimen is feasible, the study was not truly designed with parameters to answer this question. The only statistical criterion the study met was that it did not exceed the number of expected serious adverse events, although 18 occurred, which seems quite generous for this sample size. 97% of patients had grade 3 or 4 haematological toxicity. This result is consistent with some studies in adults that have used a similar approach and might be regarded as acceptable.7 However, febrile neutropenia occurred in 71% of patients and four patients had toxic deaths (two from infection, one from surgical bleeding, and one with tumour bleeding). More than 50% of patients had with significant hearing loss. Since ototoxocity is difficult to measure, notoriously under-reported, and can progress over time, the question remains as to what hearing function these very young patients will have in the long run and whether this loss is an acceptable price to pay for the survival rates observed. Lastly, during the three cycles of weekly cisplatin therapy, patients typically had average cumulative delays of 10 days (range −4 to 63 days), which resulted in dose reductions in 9% of the initial chemotherapy block A cycles. Such delays might render a regimen not feasible. Interestingly, irrespective of delays in therapy, all patients did well. This outcome somewhat contradicts the study conclusion that dose intensity was the reason for improvement, since the outstanding results were even seen in patients who had delays and reductions in therapy (unless all patients exceeded the dose-density threshold).
Importantly, the entire international paediatric oncology community, including Europe, Japan, and South and North America, are now working together to create unified approaches to liver tumour classification and treatment and there are plans for a single forthcoming trial. The results here suggest that the SIOPEL-4 regimen should be compared in a randomised trial further evaluating the toxicities of this design. As exciting as these results are, they emphasise the urgent need for new drugs that hold the promise of a cure with acceptable long-term side-effects.