Re: How NHS investment in proton beam therapy is coming to fruition
The article by Matthew Limb (1) presents the view that an assessment of the potential benefits from expanded indications for proton therapy requires outcomes from randomised clinical trials (RCTs), being the “gold standard” methodology for assessment of therapeutic benefit from wide range of medical interventions. However in the case of proton therapy, where in most scenarios advantages in radiation dose distribution are employed to reduce the incidence or severity of long-term morbidity, running trials where the end-point may be evident decades later may well be problematic. In the case of head and neck cancer, as discussed in Limb’s article, the end-point of radiation-induced xerostomia becomes evident within the early years after treatment. On the other hand when the aim is to reduce the risk of late cardiac events decades after breast or chest wall irradiation for left sided breast cancer when the internal mammary lymph nodes also require irradiation, RCTs may not necessarily provide the only or even the optimum means of evaluating potential benefits. Although an RCT is running in North America it will be many years before results will be mature. There may be a good case for extrapolating from the data published by Darby et al (2) which reports the risk of late cardiac events in relation to the mean cardiac dose from conventional radiotherapy for which there is no low-dose threshold.
RCTs may also miss potential benefits for sub-groups, and may not be the most appropriate for picking up the rare, very late stochastic events such as radiation-induced cancers for which proton therapy is aimed at risk reduction.
The Dutch health authorities have relied heavily on a model-based approach, and have predicted an eventual requirement for proton therapy of up to 10% of irradiated patients. This model-based approach (3) involves the analysis of comparative proton and conventional x-ray plans, which in several clinical scenarios, particularly head and neck cancer, can model the benefits in terms of reduction in Normal Tissue Complication Probability (NTCP). This approach is now being explored by other groups and in additional clinical scenarios, including breast cancer (4).
It is important to remember that the current NHS England indications for proton therapy are not based on the outcomes from randomised trials. Probably the strongest consensus for the benefits of proton therapy is for the treatment of children. However the evidence comprises mainly single institution case series and planning comparisons with extrapolation of likely reduction in late effects from historic dose/response data.
Although in a number of scenarios there is a strong case for evaluating the true benefit of proton therapy using RCTs, this is not the only approach being employed internationally. The UK health administrations should employ a holistic approach to reviewing the evidence for a range of cancers with not only RCTs, but where appropriate, planning studies and model-based approaches. Continued collection of “real-world” outcome data on an international basis is also essential. Furthermore it is also important to conduct further research into the impact of health economic parameters, taking into account the cost of treating long-term complications of therapy.
1. Limb M. Where will the NHS focus its proton beam therapy plans? Br Med J 2019; 364: 253-294
2. Darby SC, Ewertz M, McGale P, et al. Risk of Ischemic Heart Disease in Women after Radiotherapy for Breast Cancer. New England Journal of Medicine 2013; 368: 987-998.
3. Langendyjk JA, Lambin P, De Ruysscher D, et al. Selection of patients for radiotherapy with protons aiming at reduction of side effects: The model-based approach. Radiotherapy and Oncology 2013; 107: 267-273
4. Vast ME, Vredeveld EJ, Credoe HM, et al. Whole breast proton irradiation for maximal reduction of heart dose in breast cancer patients. Breast Cancer Res Treat 2014; 148: 33-39
Competing interests: Clinical Advisor to Proton Partners International