David J. Brenner

Cancer risks attributable to low doses of ionizing radiation: Assessing what we really know

High doses of ionizing radiation clearly produce deleterious consequences in humans, including, but not exclusively, cancer induction. At very low radiation doses the situation is much less clear, but the risks of low-dose radiation are of societal importance in relation to issues as varied as screening tests for cancer, the future of nuclear power, occupational radiation exposure, frequent-flyer risks, manned space exploration, and radiological terrorism. We review the difficulties involved in quantifying the risks of low-dose radiation and address two specific questions. First, what is the lowest dose of x- or γ-radiation for which good evidence exists of increased cancer risks in humans? The epidemiological data suggest that it is ≈10–50 mSv for an acute exposure and ≈50–100 mSv for a protracted exposure. Second, what is the most appropriate way to extrapolate such cancer risk estimates to still lower doses? Given that it is supported by experimentally grounded, quantifiable, biophysical arguments, a linear extrapolation of cancer risks from intermediate to very low doses currently appears to be the most appropriate methodology. This linearity assumption is not necessarily the most conservative approach, and it is likely that it will result in an underestimate of some radiation-induced cancer risks and an overestimate of others.

Fractionation and protraction for radiotherapy of prostate carcinoma

PURPOSE To investigate whether current fractionation and brachytherapy protraction schemes for the treatment of prostatic cancer with radiation are optimal, or could be improved. METHODS AND MATERIALS We analyzed two mature data sets on radiotherapeutic tumor control for prostate cancer, one using EBRT and the other permanent seed implants, to extract the sensitivity to changes in fractionation of prostatic tumors. The standard linear-quadratic model was used for the analysis. RESULTS Prostatic cancers appear significantly more sensitive to changes in fractionation than most other cancers. The estimated alpha/beta value is 1.5 Gy [0.8, 2.2]. This result is not too surprising as there is a documented relationship between cellular proliferative status and sensitivity to changes in fractionation, and prostatic tumors contain exceptionally low proportions of proliferating cells. CONCLUSIONS High dose rate (HDR) brachytherapy would be a highly appropriate modality for treating prostate cancer. Appropriately designed HDR brachytherapy regimens would be expected to be as efficacious as low dose rate, but with added advantages of logistic convenience and more reliable dose distributions. Similarly, external beam treatments for prostate cancer can be designed using larger doses per fraction; appropriately designed hypofractionation schemes would be expected to maintain current levels of tumor control and late sequelae, but with reduced acute morbidity, together with the logistic and financial advantages of fewer numbers of fractions.

Direct evidence that prostate tumors show high sensitivity to fractionation (low α/β ratio), similar to late-responding normal tissue

Abstract Purpose : A direct approach to the question of whether prostate tumors have an atypically high sensitivity to fractionation (low α/β ratio), more typical of the surrounding late-responding normal tissue. Methods and Materials : Earlier estimates of α/β for prostate cancer have relied on comparing results from external beam radiotherapy (EBRT) and brachytherapy, an approach with significant pitfalls due to the many differences between the treatments. To circumvent this, we analyze recent data from a single EBRT + high-dose-rate (HDR) brachytherapy protocol, in which the brachytherapy was given in either 2 or 3 implants, and at various doses. For the analysis, standard models of tumor cure based on Poisson statistics were used in conjunction with the linear-quadratic formalism. Biochemical control at 3 years was the clinical endpoint. Patients were matched between the 3 HDR vs. 2 HDR implants by clinical stage, pretreatment prostate-specific antigen (PSA), Gleason score, length of follow-up, and age. Results : The estimated value of α/β from the current analysis of 1.2 Gy (95% CI: 0.03, 4.1 Gy) is consistent with previous estimates for prostate tumor control. This α/β value is considerably less than typical values for tumors (≥8 Gy), and more comparable to values in surrounding late-responding normal tissues. Conclusions : This analysis provides strong supporting evidence that α/β values for prostate tumor control are atypically low, as indicated by previous analyses and radiobiological considerations. If true, hypofractionation or HDR regimens for prostate radiotherapy (with appropriate doses) should produce tumor control and late sequelae that are at least as good or even better than currently achieved, with the added possibility that early sequelae may be reduced.

Second Malignancies in Prostate Carcinoma Patients after Radiotherapy Compared with Surgery

In the treatment of prostate carcinoma, radiotherapy and surgery are common choices of comparable efficacy; thus a realistic comparison of the potential long term sequelae, such as the risk of second malignancy, may be of relevance to treatment choice.

Estimating cancer risks from pediatric CT: going from the qualitative to the quantitative.

these discussions, and this is really our rationale – to get some sort of feel for the absolute risks that may be involved. Let me start with a bit of background. It is clear, and I think everybody in the audience knows this, that pediatric CT is different from adult CT and also from any other sort of radiological exam as our article [1] points out. The organ doses are clearly higher for children than for adults [2]. Pediatric CT is of course increasing in frequency quite rapidly and probably more so than adult CT [3], and as Eric mentioned children are much more sensitive to radiation-induced cancer than adults [4]. I will go through these differences in some detail. First, let’s talk about the organ doses. For a given set of machine parameters (including the mAss), organ doses are larger in a child compared to a (larger) adult. Consider, for example, an organ located on the proximal side of the body relative to the x-ray source. This organ will get roughly the same dose in both adult and child (Recall that dose is energy deposited divided by mass.). As the x-ray source rotates, that same organ will be on the distal side of the body relative to the x-ray source; now that organ is partly shielded by the body tissue proximal to it, reducing the organ dose. But this dose-reducing, partial shielding will be much less for a thin individual, such as a child, compared to a thicker adult. Thus organ doses for children are larger than for adults. Pediatric CT usage is rapidly increasing. The following are some very rough numbers, and it must be said there is still a need for more surveys on pediatric CT usage in the USA. In 1989, around 4% of all CTs were pediatric [5], and this rose to around 6% in 1993. Today that number is about 1 in 10, making an estimated 2.7 million pediatric CT examsper year in this country [3]. This is clearly a large increase in the use of pediatric CT. CTs contribute disproportionately to the overall radiation dose from radiological sources. Perhaps 10% of all diagnostic radiological procedures are CT procedures, but their contribution to the overall collective dose is probably 67% simply because the doses are higher [3, 6]. These numbers are from a fine paper by Mettler and colleagues published recently [3]. Although CTs are not the most common radiological exam, they are the most important to the population in terms of the dose. We also need to think about the issue of multiple CTs. Again quoting data from Mettler [3], 30% of patients who undergo CT have at least 3 scans, 7% have at least 5, and 4% have at least 9. Thus, we need to get a better feel for the average number of CTs that any given individual has, and multiply the dose by that number. Finally, we should again discuss the issue that children are clearly more sensitive to radiation than adults [4]. Figure 1 presents more recent data from the A-bomb survivors which show an even bigger age effect. The graph indicates lifetime cancer mortality risk versus age at time of exposure. There is an order of magnitude increase in risk in children versus adults and a significant sex differential, a factor of 2 difference in sensitivity between girls and boys. In general, the reason for the shape of this curve is twofold. One is that children have more time to express a cancer than do adults. Hopefully, they have their whole lives in front of them [7]. Second, it appears that children are inherently more sensitive to radiation simply because they have more dividing cells and radiation SESSION I: HELICAL CT AND CANCER RISK

Conditions for the equivalence of continuous to pulsed low dose rate brachytherapy

Low dose rate interstitial brachytherapy is extremely useful for those tumors that are accessible for an implant, while the introduction of remote afterloaders has eliminated exposure to nursing personnel. Currently, such machines require an inventory of many sources which are loaded into catheters implanted in the tumor and kept in place during treatment. A significant simplification of such machines would be possible in a pulsed mode, with a single source moving under computer control through the catheters. Assuming that the treatment time and average dose rate are kept unchanged, the question addressed is to find those combinations of radiation pulse widths and frequencies that would be functionally equivalent to a continuous irradiation. The linear-quadratic formalism was used to reanalyze published low dose-rate studies on cells of human origin to obtain 36 parameter sets [alpha, beta, T1/2], where T1/2 is the half time for sublethal damage repair. These data are consistent with those for human tumors. For each parameter set, those combinations of pulse width and frequency were calculated that would yield a functionally equivalent cell survival. For a regimen of 30 Gy in 60 hr, a pulse width of 10 min with a period between pulses of 1 hr would be appropriate for all the cell lines considered. Similar results were found for other possible time/dose combinations. For late effects, a 1-hr period between 10-min pulses might produce up to a 2% increase in late-effect probability, which is probably acceptable for the small volumes irradiated in interstitial brachytherapy.

The Bystander Effect in Radiation Oncogenesis: I. Transformation in C3H 10T½ Cells In Vitro can be Initiated in the Unirradiated Neighbors of Irradiated Cells

Abstract Sawant, S. G., Randers-Pehrson, G., Geard, C. R., Brenner, D. J. and Hall, E. J. The Bystander Effect in Radiation Oncogenesis: I. Transformation in C3H 10T½ Cells In Vitro can be Initiated in the Unirradiated Neighbors of Irradiated Cells. It has long been accepted that radiation-induced genetic effects require that DNA be hit and damaged directly by the radiation. Recently, evidence has accumulated that in cell populations exposed to low doses of α particles, biological effects occur in a larger proportion of cells than are estimated to have been traversed by α particles. The end points observed include chromosome aberrations, mutations and gene expression. The development of a fast single-cell microbeam now makes it possible to expose a precisely known proportion of cells in a population to exactly defined numbers of α particles, and to assay for oncogenic transformation. The single-cell microbeam delivered no, one, two, four or eight α particles through the nuclei of all or just 10% of C3H 10T½ cells. We show that (a) more cells can be inactivated than were actually traversed by α particles and (b) when 10% of the cells on a dish are exposed to α particles, the resulting frequency of induced transformation is not less than that observed when every cell on the dish is exposed to the same number of α particles. These observations constitute evidence suggesting a bystander effect, i.e., that unirradiated cells are responding to damage induced in irradiated cells. This bystander effect in a biological system of relevance to carcinogenesis could have significant implications for risk estimation for low-dose radiation.

What hypofractionated protocols should be tested for prostate cancer

PURPOSE Recent analyses of clinical results have suggested that the fractionation sensitivity of prostate tumors is remarkably high; corresponding point estimates of the alpha/beta ratio for prostate cancer are around 1.5 Gy, much lower than the typical value of 10 Gy for many other tumors. This low alpha/beta value is comparable to, and possibly even lower than, that of the surrounding late-responding normal tissue in rectal mucosa (alpha/beta nominally 3 Gy, but also likely to be in the 4-5 Gy range). This lower alpha/beta ratio for prostate cancer than for the surrounding late-responding normal tissue creates the potential for therapeutic gain. We analyze here possible high-gain/low-risk hypofractionated protocols for prostate cancer to test this suggestion. METHODS AND MATERIALS Using standard linear-quadratic (LQ) modeling, a set of hypofractionated protocols can be designed in which a series of dose steps is given, each step of which keeps the late complications constant in rectal tissues. This is done by adjusting the dose per fraction and total dose to maintain a constant level of late effects. The effect on tumor control is then investigated. The resulting estimates are theoretical, although based on the best current modeling with alpha/beta parameters, which are discussed thoroughly. RESULTS If the alpha/beta value for prostate is less than that for the surrounding late-responding normal tissue, the clinical gains can be rather large. Appropriately designed schedules using around ten large fractions can result in absolute increases of 15% to 20% in biochemical control with no evidence of disease (bNED), with no increase in late sequelae. Early sequelae are predicted to be decreased, provided that overall times are not shortened drastically because of a possible risk of acute or consequential late reactions in the rectum. An overall time not shorter than 5 weeks appears advisable for the hypofractionation schedules considered, pending further clinical trial results. Even if the prostate tumor alpha/beta ratio turns out to be the same (or even slightly larger than) the surrounding late-responding normal tissue, these hypofractionated regimens are estimated to be very unlikely to result in significantly increased late effects. CONCLUSIONS The hypofractionated regimens that we suggest be tested for prostate-cancer radiotherapy show high potential therapeutic gain as well as economic and logistic advantages. They appear to have little potential risk as long as excessively short overall times (<5 weeks) and very small fraction numbers (<5) are avoided. The values of bNED and rectal complications presented are entirely theoretical, being related by LQ modeling to existing clinical data for approximately intermediate-risk prostate cancer patients as discussed in detail.

The radiobiology of radiosurgery: Rationale for different treatment regimes for AVMs and malignancies

Based on basic radiobiological principles, we suggest that the radiosurgery technique of delivering a radiation dose in a single fraction, whilst appropriate for benign brain lesions such as arteriovenous malformations (AVM), is not optimal for treating malignant tumors. Radiosurgery was originally developed to treat benign lesions in the brain, such as AVMs, and has been successfully used for this purpose for over four decades. Recently, the technique has been adopted for treating small primary malignant brain tumors or single metastases. We argue, and derive radio-biological data to support the view that, treating malignant tumors with a single fraction will result in a suboptimal therapeutic ratio between tumor control and late effects, even for small tumors; and that improved therapeutic ratios would be expected if the treatment were fractionated into a small number of fractions. On the other hand, no therapeutic gain is to be expected from fractionating treatment of AVMs. A new generation of noninvasive relocatable stereotactic head frames makes feasible the use of fractionated stereotactic external-beam radiotherapy, and may allow significant benefits over single, radiosurgical, treatments for malignant brain tumors. As stereotactic fractionation/protraction regimes become more widespread, a uniform approach for determining equivalent fractionation schemes becomes important for intercomparing clinical results, and such calculations can be reliably carried out using the linear-quadratic formalism.

The linear-quadratic model is an appropriate methodology for determining isoeffective doses at large doses per fraction.

The tool most commonly used for quantitative predictions of dose/fractionation dependencies in radiotherapy is the mechanistically based linear-quadratic (LQ) model. The LQ formalism is now almost universally used for calculating radiotherapeutic isoeffect doses for different fractionation/protraction schemes. In summary, the LQ model has the following useful properties for predicting isoeffect doses: (1) it is a mechanistic, biologically based model; (2) it has sufficiently few parameters to be practical; (3) most other mechanistic models of cell killing predict the same fractionation dependencies as does the LQ model; (4) it has well-documented predictive properties for fractionation/dose-rate effects in the laboratory; and (5) it is reasonably well validated, experimentally and theoretically, up to about 10 Gy/fraction and would be reasonable for use up to about 18 Gy per fraction. To date, there is no evidence of problems when the LQ model has been applied in the clinic.