Portal de Periódicos da CAPES

The Omega on Alpha and Beta

2011; Elsevier BV; Volume: 81; Issue: 2 Linguagem: Inglês

10.1016/j.ijrobp.2011.01.011

1879-355X

Eli Glatstein,

Effects of Radiation Exposure

OK, I confess: I have trouble with alpha/beta ratios, and I want to provoke some discussion. The basic idea behind alpha/beta is a ratio of two different types of cell killing, essentially single hit and multiple hit types of radiation. This is what one gets with x-rays and the linear quadratic formula to explain a cell survival curve. That part is relatively understandable and straightforward. The problem I have is when people propose to use alpha/betas for treatment of patients. In this issue of the Journal, Tucker et al. (1Tucker et al.Google Scholar) quantify what they call an alpha/beta for Grade 2 and higher rectal toxicity from RTOG 94-06. They come up with reasonable numbers based on what we think we know. On the other hand, I question whether the derived numbers truly reflect two different mechanisms of cell killing from irradiation, which these investigators presume to be the case. It is well understood that, when we treat patients, we try to take advantage of the different characteristics between tumor cells and normal tissues that are to be irradiated. Any kind of patient treatment plan needs to account for both tumor and normal tissue, because both will be exposed, and both are critical for risk–benefit concepts as they apply to the so-called "therapeutic ratio." (It is a paradox in radiation therapy that what we usually do, in fact, is treat normal tissues to doses that we think are safe on the basis of experience, and accept whatever tumor cytotoxicity that we can achieve with that dose. Such is the impact of primum non nocere). That's what we suggest by "tolerance," although virtually no one defines the time factor, which in reality implies as long as the patient lives (whether 10 days or 50 years). It is also understood that tumor cells have one and only one function, and that is to divide, whereas normal tissues not only have to replace depleted cellular compartments by cell division but also have to function in ways that permit organs to function in an integrated fashion. Organs are comprise of many different types of those normal cells. When one tries to assess alpha/beta in tumors, the original end point was quick and in vitro, i.e., a specific number based on cell survival curves in tissue culture. Trying to get alpha/beta on normal tissues is much more problematic. What is the functional end point to be assessed, exactly how and when? If we take an organ such as the kidney, for example, there are various parenchymal renal tubules; there are glomeruli with endothelial composition, ordinary blood vessels, neurologic tissues, fibroblasts, and others. To quantify the alpha/beta of the kidney directly (as opposed to the inferential approach by Tucker et al.), which of those cell types should be plated? The answer is certainly unclear. In addition, it is likely that the alpha/beta for each one of those cell types would be somewhat different. After all, we are talking about the "laws" of biology, not the immutable laws of physics. Everything we know about biology is that there is a range of normal biological expression for organisms and organs; everything in physiology has a range of what we call "normal," and almost everything is multifactorial. It is customary to talk about the acute side effects of having an alpha/beta of 10 (range said to be 7–20) (2Joiner M.C. Bentzen S.M. Fractionation: The linear-quadratic approach.in: Joiner M. Vander Kogel A. Basic clinical radiobiology. Hodder Arnold, 2009: 102-119Crossref Google Scholar). What is the expected range of alpha/beta for each individual patient, both within "normal" and beyond? To assess alpha/beta ratios both acutely and chronically, one has to presume that the effects or end points are entirely a reflection of cell killing. It is probably true that the acute effects of radiation are a direct expression of cytotoxicity on fast renewal systems, primarily skin, mucous membranes, and bone marrow. By contrast, when we talk about the late effects of radiation therapy, we say that there is a typical alpha/beta of approximately 3 (range said to be 0.5–6) (2Joiner M.C. Bentzen S.M. Fractionation: The linear-quadratic approach.in: Joiner M. Vander Kogel A. Basic clinical radiobiology. Hodder Arnold, 2009: 102-119Crossref Google Scholar). This is talking about chronic effects on normal tissues. One thing is clear, at least to me: the chronic effects of radiation certainly include some degree of cell killing, but direct cell killing does not explain all of the late effects. Radiation must do something to normal cellular functions such as muscle contraction, absorption, protein synthesis, and filtration in an organ such as the kidney, air exchange in the lung, secretion from the endocrine organs, and a variety of other normal cell functions. It is very difficult to conceive of radiation ordinarily improving most of those cellular and organ functions. There must be some compromise, although hopefully (and typically) the compromise is fairly minimal. Nonetheless, if we were to measure the individual cellular functions of some of these organs (which we do not), most of us would expect there to be some kind of decrease in function, although it might not necessarily be important physiologically. The point to emphasize here is that altered cell function does not necessarily mean cell death. Cell and organ function can be compromised to a sublethal level by irradiation. So how does alpha/beta actually apply to late effects in normal tissues? Normal tissue injury from radiation over time may well reflect a linear-quadratic formula mathematically, but that does not necessarily mean that the parameters of that formula refer to two different mechanisms of radiation cell killing. Tucker et al. inherently presume that all of the late damage is from cell killing: They may call their determination an alpha/beta, but that does not necessarily make it an alpha/beta ratio. These are some of the problems that I have with alpha/beta. I really question how anyone can accurately estimate the alpha/beta ratio for normal tissues and organs from any functional assessment(s), with a possible exception of the central nervous system. What Tucker et al. have done is commendable, but I question their interpretation. There are a couple of other points worth mentioning as well. The alpha/beta ratio is considered by many to be some kind of constant number largely because, as mentioned above, the original determination for tumor cells was based on tissue culture work. For normal tissues, this cannot be correct. If we have a late radiation–induced compromise of organ function, it will evolve and worsen over time. Thus, it cannot be a short-term constant but, rather, a long-term variable dependent upon specific time points, measured in years in human beings. These points are clearly emphasized by Joiner and Bentzen (2Joiner M.C. Bentzen S.M. Fractionation: The linear-quadratic approach.in: Joiner M. Vander Kogel A. Basic clinical radiobiology. Hodder Arnold, 2009: 102-119Crossref Google Scholar), but it is my strong impression that most physicians who wish to use alpha/beta to justify major alterations in fractionation overlook their admonition. Finally, there has been a rather clever analysis done by Brenner and Hall (3Brenner D.J. Hall E.J. Fractionation and protraction for radiotherapy of prostate carcinoma.Int J Radiat Oncol Biol Phys. 1999; 43: 1095-1101Abstract Full Text Full Text PDF PubMed Scopus (780) Google Scholar) to determine alpha and beta (separately) in patients treated for prostate cancer with either brachytherapy alone or external beam alone. Thus, an alpha and a beta have been quantified separately but from different patients. This is not a true alpha/beta for any specific patient; what has been quantified is an "average" alpha and an "average" beta. It is not clear, to me at least, what the alpha/beta is for any given patient in terms of normal tissue. We do know that biology has a wide range of variability as an essential element. Heterogeneity and variation are not only the facts of biological life, but also both fascinating and frustrating facts of life in clinical medicine and clinical care. It is my strongest belief that every patient and every tumor are unique. To me, this is the most fundamental element in human oncology. When we perform prospective studies, we group patients together on the basis of shared characteristics, but we neglect the differences because we really have no reliable way of assessing what those specific differences are. Perhaps microarrays and other molecular technologies may change that, but it will take a very long time. There are no easy answers to any of these issues; they are simply far more nuanced than has been generally discussed. Over the years, I have told many trainees that one can be an excellent clinical radiation oncologist and not necessarily know squat about alpha/beta. I believe that remains true today. Estimation of α/β for Late Rectal Toxicity Based on RTOG 94-06International Journal of Radiation Oncology, Biology, PhysicsVol. 81Issue 2PreviewTo estimate α/β, the parameter ratio from the linear-quadratic (LQ) model, for Grade ≥2 late rectal toxicity among patients treated on Radiation Therapy Oncology Group (RTOG) protocol 94-06; and to determine whether correcting the rectal dose–volume histogram (DVH) for differences in dose per fraction, based on the LQ model, significantly improves the fit to these data of the Lyman–Kutcher–Burman (LKB) normal-tissue complication probability (NTCP) model. Full-Text PDF On the Parameters of the Linear-Quadratic Model: In Regard to the Editorial by GlatsteinInternational Journal of Radiation Oncology, Biology, PhysicsVol. 82Issue 4PreviewTo the Editor: The editorial by Glatstein casts some doubt on the validity and the relevance of estimating α/β, especially far late effects (1). Some of his arguments need to be qualified. Full-Text PDF