Apoptosis and Response to Radiation: Implications for Radiation Therapy

Tumor growth is the result of two opposing processes--cell division and cell loss. As long as division outpaces loss, tumors will continue to grow. The form of "active" cell death called apoptosis is now known to be controlled by specific genes, and it is hoped that manipulating the expression of these genes could shift the balance in favor of cell loss.

The review by Dr. Meyn covers promising early results using gene therapy techniques to administer p53 (a tumor-suppressor protein involved in apoptosis and growth regulation) or bcl-2X_s (an inhibitor of bcl-2 that, itself, inhibits apoptosis). These results support the idea that promoting apoptosis may be a viable way of controlling tumor growth.

Dr. Meyn was actively involved in apoptosis research as it relates to radiation damage before it became fashionable. He points out that almost 40 years ago, radiobiologists labeled this form of cell death as "interphase cell death" to distinguish it from the mitotic cell death that occurred as radiation damaged cells attempted to divide. What was not appreciated until relatively recently is that rapid interphase death may be prevented by a simple gene mutation or by overexpression of certain other gene products. Such genetic changes are often present in solid tumors.

An important, valid point made in Dr. Meyn's review is that "time" is the critical element needed to define the significance of apoptosis in radiotherapy. He makes the distinction between apoptosis as a primary mode of cell death occurring within a few hours of irradiation and "secondary" apoptosis observed at much longer times (24 to 96 hours) after irradiation. Clearly, damaged cells that fail to undergo apoptosis within the first few hours after irradiation can still undergo reproductive cell death. However, Dr. Meyn argues (and is probably right) that the manner in which those cells die--whether by secondary apoptosis, necrosis, or simple senescence--is largely immaterial for practical purposes.

Radiosensitivity of Cells Capable of Primary Apoptosis

Dr. Meyn's second point is that cells capable of undergoing primary apoptosis are more radiosensitive. Again, results from a number of studies using cells in culture support this position. He could probably go further and add that while it is possible to reduce radiation cell killing by preventing primary apoptosis, it may be impossible to prevent secondary apoptosis. Secondary apoptosis seems to be a response to lethal (irreparable) damage. Since most cells capable of primary apoptosis are of hematologic or lymphoid origin, the relevance of apoptosis to the response of most solid tumors has been questioned.

Dr. Meyn approaches this question by summarizing a great deal of work carried out in over a dozen solid murine tumors. Squamous cell carcinomas, hepatocarcinomas, and fibrosarcomas did not undergo apoptosis in response to radiation. However, the apoptotic response of adenocarcinomas and lymphomas peaked at about 4 hours after irradiation, suggesting that this was primary apoptosis by his definition. In these studies, Dr. Meyn found that less radiation was required to control tumors with significant amounts of radiation-induced apoptosis--the expected result if cells capable of under going primary apoptosis are more radiosensitive.

Perhaps more interesting, the spontaneous apoptotic index correlated with tumor curability in the same series of murine tumours. Dr. Meyn argues that if tumors contain more apoptotic cells, there must be fewer clonogens to kill, and the tumor should be controlled with lower radiation doses. This observation has important implications, since it suggests that the ability to undergo apoptosis could have a greater influence on outcome than other properties of solid tumors known to affect response to radiotherapy, such as hypoxia and tumor growth kinetics.

Clinical Results Less Clear-Cut

However, results from the clinic are not as clear-cut. Dr. Meyn reports that some studies show a correlation between high pretreatment apoptotic index and survival/response, while others do not. Gordon Steel's "bucket and tap" analogy is useful in thinking about this situation.[1] Steel illustrates the relationships among cell production by mitosis (the tap dripping tumor cells into a bucket), cell loss by apoptosis or necrosis (holes in the bucket that release the cells), and tumor growth (the accumulation of cells in the bucket). Influencing the speed of the dripping tap, or the size of the holes in the bucket, will alter tumor volume. However, if there are more holes in the bucket (ie, a higher apoptotic fraction), mitosis will have to increase to maintain tumor growth. This could result in a higher growth fraction and, perhaps, a lower likelihood of curing some tumors with radiation.

Alternatively, tumor cell types with a high incidence of spontaneous apoptosis could be dying because they do not tolerate a nutrient-poor (hypoxic?) tumor microenvironment. Such tumors should be more radiocurable.

Unfortunately, resolving these issues is not easy because measurements of apoptotic index in vivo are difficult to perform and interpret. Apoptotic index varies with time, radiation dose, and fractionation schedule. As pointed out in an article by Potten[2] in reference to apoptosis in the gut and by Dewey et al[3] in relation to solid tumors, the apoptotic yield measured in vivo rarely relates to the yield of reproductively sterilized cells. This is perhaps not surprising considering that the duration of the apoptotic process (ie, how rapidly apoptotic cells develop, how long they survive in vivo) is largely unknown, as is the contribution of other cell types within the tumor.

The inability to directly correlate reproductive cell death with apoptotic fraction is clearly a limitation in defining the kinetics and significance of the process in vivo. However, this should not be a detriment to designing effective strategies to promote apoptosis in tumors that have lost this capacity.