N.G. Burnet

Normal tissue reactions to radiotherapy: towards tailoring treatment dose by genotype

A key challenge in radiotherapy is to maximize radiation doses to cancer cells while minimizing damage to surrounding healthy tissue. As severe toxicity in a minority of patients limits the doses that can be safely given to the majority, there is interest in developing a test to measure an individuals radiosensitivity before treatment. Variation in sensitivity to radiation is an inherited genetic trait and recent progress in genotyping raises the possibility of genome-wide studies to characterize genetic profiles that predict patient response to radiotherapy.

Years of life lost (YLL) from cancer is an important measure of population burden — and should be considered when allocating research funds

Recently, cancer mortality has been compared to research spending by the National Cancer Research Institute (NCRI), whose research budget is approximately £250 million. The analysis shows a mis-match between mortality and research spending. As well as crude mortality rates, other measures of cancer burden should be considered because they contribute additional information. ‘Years of life lost’ (YLL) summed over each individual dying after a diagnosis of cancer represents a population-based mortality indicator of the impact of that disease on society. Years of life lost divided by the number of deaths for each cancer site produces an additional statistic, the average years of life lost (AYLL), which is a measure of the burden of cancer to the individual patient. For 17 cancer sites where data are available, four tumour sites have a rather large difference in mortality, comparing YLL to crude mortality. Years of life lost shows the population burden from cancers of the ovary, cervix, and CNS to be rather larger than suggested by crude mortality, despite screening programmes for cervix cancer. Using YLL, the underprovision of funding for lung cancer research is similar to that reported using percentage mortality. Breast cancer and leukaemia receive a relatively higher research spend than the population burden of these cancers, and the spending on leukaemia is quite extreme. Prostate cancer has a low per cent YLL but attracts a moderate amount of research spending. The use of AYLL as an indicator of individual cancer burden considerably changes the ranking of the mortality from different tumours. The mean AYLL is 12.5 years. Prostate cancer has the lowest AYLL, only 6.1 years; brain tumour patients have the highest, at just over 20 years. Comparing AYLL to research spending suggests four ‘Cinderella’ cancer sites with high individual cancer burden but low research spending: CNS tumours, cervix and kidney cancers, and melanoma. Breast cancer and leukaemia have roughly average AYLL but a considerable excess of research spending. YLL emphasises the discrepancy between research spending and mortality, and may be helpful for decisions concerning research support. Avearage years of life lost measures the burden to individual patients and may be helpful where individuals’ needs are relevant, such as palliative care. As well as crude mortality, more subtle and comprehensive calculations of mortality statistics would be useful in debates on research funding and public health issues.

Diffusion tensor imaging of brain tumours at 3 T: a potential tool for assessing white matter tract invasion?

AIM To determine whether diffusion tensor imaging (DTI) of brain tumours can demonstrate abnormalities distal to hyperintensities on T2-weighted images, and possibly relate these to tumour grade. MATERIALS AND METHODS Twenty patients with histologically confirmed supratentorial tumours, both gliomas (high and low grade) and metastases, were imaged at 3T using T2-weighted and DTI sequences. Regions of interest (ROI) were drawn within the tumour, in white matter at various distances from the tumour and in areas of abnormality on DTI that appeared normal on T2-weighted images. The relative anisotropy index (RAI)-a measure of white matter organization, was calculated for these ROI. RESULTS The abnormality on DTI was larger than that seen on T2-weighted images in 10/13 patients (77%) with high-grade gliomas. New abnormalities were seen in the contralateral white matter in 4/13 (30%) of these cases. In these high-grade tumours the RAI in areas of white matter disruption with normal appearance on T2-weighted images was reduced (0.19+/-0.04). Even excluding patients with previous radiotherapy this difference remains significant. In all non high-grade tumours (WHO grade II gliomas and metastases) the tumour extent on DTI was identical to the abnormalities shown on T2-weighted imaging and RAI measurements were not reduced (0.3+/-0.04). CONCLUSIONS Subtle white matter disruption can be identified using DTI in patients with high-grade gliomas. Such disruption is not identified in association with metastases or low-grade gliomas despite these tumours producing significant mass effect and oedema. We suggest the changes in DTI may be due to tumour infiltration and that the DTI may provide a useful method of detecting occult white matter invasion by gliomas.

Independent validation of genes and polymorphisms reported to be associated with radiation toxicity: a prospective analysis study

BACKGROUND Several studies have reported associations between radiation toxicity and single nucleotide polymorphisms (SNPs) in candidate genes. Few associations have been tested in independent validation studies. This prospective study aimed to validate reported associations between genotype and radiation toxicity in a large independent dataset. METHODS 92 (of 98 attempted) SNPs in 46 genes were successfully genotyped in 1613 patients: 976 received adjuvant breast radiotherapy in the Cambridge breast IMRT trial (ISRCTN21474421, n=942) or in a prospective study of breast toxicity at the Christie Hospital, Manchester, UK (n=34). A further 637 received radical prostate radiotherapy in the MRC RT01 multicentre trial (ISRCTN47772397, n=224) or in the Conventional or Hypofractionated High Dose Intensity Modulated Radiotherapy for Prostate Cancer (CHHiP) trial (ISRCTN97182923, n=413). Late toxicity was assessed 2 years after radiotherapy with a validated photographic technique (patients with breast cancer only), clinical assessment, and patient questionnaires. Association tests of genotype with overall radiation toxicity score and individual endpoints were undertaken in univariate and multivariable analyses. At a type I error rate adjusted for multiple testing, this study had 99% power to detect a SNP, with minor allele frequency of 0·35, associated with a per allele odds ratio of 2·2. FINDINGS None of the previously reported associations were confirmed by this study, after adjustment for multiple comparisons. The p value distribution of the SNPs tested against overall toxicity score was not different from that expected by chance. INTERPRETATION We did not replicate previously reported late toxicity associations, suggesting that we can essentially exclude the hypothesis that published SNPs individually exert a clinically relevant effect. Continued recruitment of patients into studies within the Radiogenomics Consortium is essential so that sufficiently powered studies can be done and methodological challenges addressed. FUNDING Cancer Research UK, The Royal College of Radiologists, Addenbrookes Charitable Trust, Breast Cancer Campaign, Cambridge National Institute of Health Research (NIHR) Biomedical Research Centre, Experimental Cancer Medicine Centre, East Midlands Innovation, the National Cancer Institute, Joseph Mitchell Trust, Royal Marsden NHS Foundation Trust, Institute of Cancer Research NIHR Biomedical Research Centre for Cancer.

Prediction of normal-tissue tolerance to radiotherapy from in-vitro cellular radiation sensitivity.

The success of radiotherapy depends on the total radiation dose, which is limited by the tolerance of surrounding normal tissues. Since there is substantial variation among patients in normal-tissue radiosensitivity, we have tested the hypothesis that in-vitro cellular radiosensitivity is correlated with in-vitro normal-tissue responses. We exposed skin fibroblast cell lines from six radiation-treated patients to various doses of radiation and measured the proportions surviving. There was a strong relation between fibroblast sensitivity in vitro and normal-tissue reactions, especially acute effects. Assessment of radiosensitivity could lead to improved tumour cure rates by enabling radiation doses to be tailored to the individual.

Randomized Controlled Trial of Intensity-Modulated Radiotherapy for Early Breast Cancer: 5-Year Results Confirm Superior Overall Cosmesis

PURPOSE There are few randomized controlled trial data to confirm that improved homogeneity with simple intensity-modulated radiotherapy (IMRT) decreases late breast tissue toxicity. The Cambridge Breast IMRT trial investigated this hypothesis, and the 5-year results are reported. PATIENTS AND METHODS Standard tangential plans of 1,145 trial patients were analyzed; 815 patients had inhomogeneous plans (≥ 2 cm(3) receiving 107% of prescribed dose: 40 Gy in 15 fractions over 3 weeks) and were randomly assigned to standard radiotherapy (RT) or replanned with simple IMRT; 330 patients with satisfactory dose homogeneity were treated with standard RT and underwent the same follow-up as the randomly assigned patients. Breast tissue toxicities were assessed at 5 years using validated methods: photographic assessment (overall cosmesis and breast shrinkage compared with baseline pre-RT photographs) and clinical assessment (telangiectasia, induration, edema, and pigmentation). Comparisons between different groups were analyzed using polychotomous logistic regression. RESULTS On univariate analysis, compared with standard RT, fewer patients in the simple IMRT group developed suboptimal overall cosmesis (odds ratio [OR], 0.68; 95% CI, 0.48 to 0.96; P = .027) and skin telangiectasia (OR, 0.58; 95% CI, 0.36 to 0.92; P = .021). No evidence of difference was seen for breast shrinkage, breast edema, tumor bed induration, or pigmentation. The benefit of IMRT was maintained on multivariate analysis for both overall cosmesis (P = .038) and skin telangiectasia (P = .031). CONCLUSION Improved dose homogeneity with simple IMRT translates into superior overall cosmesis and reduces the risk of skin telangiectasia. These results are practice changing and should encourage centers still using two-dimensional RT to implement simple breast IMRT.

DESCRIBING PATIENTS' NORMAL TISSUE REACTIONS: CONCERNING THE POSSIBILITY OF INDIVIDUALISING RADIOTHERAPY DOSE PRESCRIPTIONS BASED ON POTENTIAL PREDICTIVE ASSAYS OF NORMAL TISSUE RADIOSENSITIVITY

Clinical radiotherapeutic doses are limited by the tolerance of normal tissues. Patients given a standard treatment exhibit a range of normal tissue reactions, and a better understanding of this individual variation might allow for individualisation of radiotherapeutic prescriptions, with consequent improvement in the therapeutic ratio. At present, there is no simple way to describe normal tissue reactions, which hampers communication between clinic and laboratory and between groups from different centres. There is also no method for comparing the severity of reactions in different normal tissues. This arises largely because there is no definition of a “normal” reaction, an “extreme” reaction or the particular term “over‐reactor” (OR). This report proposes definitions for these terms, as well as a simple terminology for describing normal tissue reactions in patients having radiotherapy. The “normal” range represents the individual variation in normal tissue reactions amongst large numbers of patients treated in the same way which is within clinically acceptable limits. The term “OR” is applied to an individual whose reaction is more severe than the normal range but also implies that this forced a major change in the radiotherapeutic prescription or that the reactions were very severe or fatal. A “severe OR” would develop serious problems with a typical radical dose, while an “extreme OR” would have such difficulties at a much lower dose. To describe the normal range, a numerical scale is suggested, from 1 to 5, resistant to sensitive. The term “highly radiosensitive” (HR) is suggested for category 5. An “informal” relative scale, as suggested here, is quick and simple. It should allow comparison between different hospitals, compensate for differences in radiotherapeutic dose and technique and allow comparison of reactions between different anatomical sites. It should be adequate for discriminating patients at the extremes of the normal range from those at the centre. It is hoped that the definitions and terminology proposed here will aid communication in the field of predictive testing of normal tissue radiosensitivity. Int. J. Cancer (Pred. Oncol.) 79:606–613, 1998.

A randomised controlled trial of forward-planned radiotherapy (IMRT) for early breast cancer: baseline characteristics and dosimetry results.

BACKGROUND AND PURPOSE This large trial was designed to investigate whether correction of dose inhomogeneities using intensity-modulated radiotherapy (IMRT) reduces late toxicity and improves quality of life in patients with early breast cancer. This paper reports baseline characteristics of trial participants and dosimetry results. MATERIALS AND METHODS Standard tangential plans of 1145 trials were analysed. Patients with inhomogeneous plans, defined by ICRU recommendations, were randomised to forward-planned IMRT or standard radiotherapy. RESULTS Twenty-nine percentage of patients had adequate dosimetry with standard 2D radiotherapy. In the randomised patients, the decreases in mean volumes receiving greater than 107% (Vol>107) and less than 95% (Vol<95) of the prescribed dose in the IMRT compared with the control group were 34.0 cm(3) (95% CI 26.4-41.6; P<0.0001) and 48.1 cm(3) (95% CI 34.4-61.9; P<0.0001), respectively. In this study, 90% of patients who had a breast separation greater > or = 21 cm had Vol>107>2 cm(3) on standard radiotherapy plans. CONCLUSION This large trial, in which patients with all breast sizes were eligible, confirmed that breast dosimetry can be significantly improved with a simple method of forward-planned IMRT and has little impact on radiotherapy resources. It is shown that patients with larger breasts are more likely to have dose inhomogeneities and breast separation gives some indication of this likelihood. Photographic assessment of patients at 2 years after radiotherapy, as the next part of this randomised controlled trial, will show whether these results for IMRT translate into improved cosmetic outcome in patients with early breast cancer. This would provide impetus for the widespread adoption of 3D planning and IMRT.

A genome wide association study (GWAS) providing evidence of an association between common genetic variants and late radiotherapy toxicity

BACKGROUND AND PURPOSE This study was designed to identify common single nucleotide polymorphisms (SNPs) associated with toxicity 2years after radiotherapy. MATERIALS AND METHODS A genome wide association study was performed in 1850 patients from the RAPPER study: 1217 received adjuvant breast radiotherapy and 633 had radical prostate radiotherapy. Genotype associations with both overall and individual endpoints of toxicity were tested via univariable and multivariable regression. Replication of potentially associated SNPs was carried out in three independent patient cohorts who had radiotherapy for prostate (516 RADIOGEN and 862 Gene-PARE) or breast (355 LeND) cancer. RESULTS Quantile-quantile plots show more associations at the P<5×10(-7) level than expected by chance (164 vs. 9 for the prostate cases and 29 vs. 4 for breast cases), providing evidence that common genetic variants are associated with risk of toxicity. Strongest associations were for individual endpoints rather than an overall measure of toxicity in all patients. However, in general, significant associations were not validated at a nominal 0.05 level in the replication cohorts. CONCLUSIONS This largest GWAS to date provides evidence of true association between common genetic variants and toxicity. Associations with toxicity appeared to be tumour site-specific. Future GWAS require higher statistical power, in particular in the validation stage, to test clinically relevant effect sizes of SNP associations with individual endpoints, but the required sample sizes are achievable.

Guidelines for preclinical and early phase clinical assessment of novel radiosensitisers.

SUMMARY There is a growing appreciation of the potential value of combining novel molecularly-targeted drugs with radiotherapy or chemoradiotherapy. Such approaches have the potential to improve locoregional disease control and cure rates across a diverse range of tumour types. In this report, we outline a rational framework for developing novel drug–radiation combinations. In doing so, we make recommendations regarding the core preclinical data sets that are required to serve as justification for studies in humans and describe potential clinical trial designs that may be adopted by investigators. Radiotherapy (RT) has a pivotal role in the management of many tumours, such that B50% of cancer patients will receive RT during the course of their illnesses and 40% of those cured of cancer will have received RT as part of their treatment (http://info.cancerresearchuk. org/). Indications for prescribing RT include (i) definitive, curative (radical) treatment; (ii) adjuvant therapy following surgery in an attempt to eradicate microscopic (or rarely macroscopic) residual disease; and (iii) as palliative treatment to ameliorate cancerrelated symptoms. In the majority of situations, RT is a highly localised treatment that targets defined volumes of tissue that are known (or suspected) to contain cancer cells. Wide-field hemibody (Bashir et al, 2008) or total body (Adkins and DiPersio, 2008) irradiation techniques are used rarely in very specific indications, such as metastatic bone disease or ‘conditioning’ before transplantation in haematological malignancies, respectively. In addition, the latest developments in stereotactic body RT (SBRT) techniques have also resulted in protocols that aim to treat oligometastatic disease spread through an organ (or more than