Dose-Response Modifiers in Radiation Therapy

Importance of Oxygen

In 1909, Gottwald Schwarz, in a simple but elegant experiment, demonstrated that the radiation response of skin was markedly decreased if the blood flow in the irradiated area was reduced by compression. Although he did not state that the phenomenon was the result of a lack of oxygen, his study was probably the first radiobiologically oriented clinical study implicating the importance of environmental parameters in the outcome of radiotherapy. This finding was used to introduce the concept of “kompressionsanämie” by which the skin was made anemic, thereby allowing a higher dose to be given to deeply situated tumors. Following the work of Schwarz, in 1910, Müller reported that tissues in which the blood flow was stimulated by diathermia showed a more prominent response to radiation. This early study not only demonstrated the importance of oxygen supply in radiotherapy but it was also the first clinical approach showing how resistance could be overcome by using hyperthermia. Subsequently, sporadic clinical and experimental observations indicated the importance of sufficient blood supply to secure an adequate radiation response. These observations led Gray et al. in the early 1950s to postulate that oxygen deficiency (hypoxia) was a major source of radiation resistance.

The first clinical indication that hypoxia existed in tumors was made around the same time by Thomlinson and Gray when, from histological observations in carcinoma of the bronchus, they reported seeing viable tumor regions surrounded by vascular stroma from which the tumor cells obtained their nutrients and oxygen. As the tumors grew, the viable regions expanded and areas of necrosis appeared at the center. The thickness of the resulting shell of viable tissue was found to be between 100 and 180 µm, which was within the same range as the calculated diffusion distance for oxygen in respiring tissues. It suggested that as oxygen diffused from the stroma, it was consumed by the cells and, although those beyond the diffusion distance were unable to survive, the cells immediately bordering the necrotic area might be viable yet hypoxic. In 1968, Tannock described an inverted version of the Thomlinson and Gray picture, with functional blood vessels surrounded by cords of viable tumor cells outside of which were areas of necrosis. This “corded” structure, illustrated in Fig. 3.2 , is the more typical picture found in most solid tumors. It arises because the tumor blood vessels, which are derived from the normal tissue vessels by a process of angiogenesis, are inadequate to meet the needs of the rapidly growing tumor cells. This hypoxia is more commonly called chronic hypoxia .

It was also suggested that hypoxia in tumors could be acute in nature. However, it was not until later that Chaplin et al. were able to confirm the existence of acutely hypoxic cells in tumors and demonstrate that these cells were the result of transient stoppages in tumor blood flow (see Fig. 3.2 ). To date, these temporary cessations in blood flow have been observed in mouse and rat tumors as well as human tumor xenografts, with anywhere from around 4% to 8% of the total functional vessels involved, although the exact causes of these stoppages are not known. The current use of chronic or acute to explain hypoxia in tumors is probably an oversimplification of the real situation. Chronic hypoxia generally refers to prolonged and reduced oxygen concentrations that influence radiation response, but there is evidence that oxygen concentrations that are higher, yet below normal physiological levels, are often found. Furthermore, reduced perfusion can be both partial and total. While cells under the former condition would be oxygen deprived, with the latter they would be starved of oxygen and nutrients. As such, their survival and response to therapy would be expected to be different.

Evidence for Hypoxia in Tumors

In experimental tumors, it is not only relatively easy to identify hypoxia but one can also quantitatively estimate the percentage of cells that are hypoxic. Three major techniques are routinely used. These are the paired survival curve, the clamped tumor growth delay, and the clamped tumor control assays. All involve a comparison of the response of tumors when irradiated under either normal air-breathing conditions or when tumors are artificially made hypoxic by clamping. Using these procedures, hypoxia has been directly identified in most animal solid tumors, with the values ranging from less than 1% to well more than 50% of the total viable cell population. None of these procedures can be applied to the clinical situation, however. One therefore must rely on indirect techniques.

Estimating hypoxia in human tumors has generally involved the use of indirect methods. Some of the earliest attempts focused on the vascular supply because it was only via the tumor vasculature that oxygen could be delivered. The endpoints included immunohistochemical estimates of intercapillary distance, vascular density, and distance from tumor cells to the nearest blood vessel ; oxyhemoglobin saturation determined using cryophotometry or noninvasively with near-infrared spectroscopy or magnetic resonance imaging (MRI) ; or measurements of tumor perfusion using MRI, computed tomography (CT), or positron emission tomography (PET). With the finding that hypoxia could upregulate gene/protein expression, it was suggested that endogenous markers could be used to identify hypoxia. The principal markers have included hypoxia-inducible factor 1 (HIF-1), carbonic anhydrase IX (CAIX), the glucose transporters GLUT-1 and GLUT-3, and osteopontin (OPN). Attempts to relate their expression levels with established hypoxia assays have reported mixed results, which is not entirely unexpected since the expression of many endogenous markers can be regulated by factors other than hypoxia. A more reliable hypoxia indicator involves combining endogenous markers in a gene signature. Although such signatures have typically been derived from cell lines exposed to hypoxia versus normoxia, they have been further developed in clinical material through associations with more direct indicators of hypoxia or outcome. More popular techniques for identifying hypoxia involve measurements of the binding of exogenous markers. This can be achieved following immunohistological analysis of biopsied sections using, for example, pimonidazole or EF5. It can also be done noninvasively with PET, single-photon emission computed tomography (SPECT), or MRI analysis of radioactively labeled nitroimidazoles (i.e., [ ¹⁸ F] labeled misonidazole or FAZA; [ ¹²³ I] labeled azomycin arabinoside), or PET imaging of [ ^60–64 Cu]-ATSM.

The most direct method involves determining oxygen partial pressure (PO ₂ ) distributions with polarographic electrodes. How this approach can be used to detect hypoxia and relate the measurements to radiotherapy outcome is illustrated in Fig. 3.3 . In this international multicenter study in head and neck cancer patients, their tumor's PO ₂ was measured before radiation therapy and was found to correlate with overall survival in that those patients with lower tumor oxygenation status did significantly worse.

Probably the best evidence for the existence of hypoxia in human tumors comes from the large number of clinical trials in which hypoxic modification has shown some benefit. The latter situation constitutes a circular argument: if hypoxic modification shows an improvement, then hypoxic clonogenic cells must have been present in tumors. However, it is likely that even tumors with the same histological makeup and of the same type have substantial heterogeneity with respect to the extent of hypoxia. It must be admitted that today, a century after the first clinical description, the importance of hypoxia and its influence on the outcome of radiotherapy is still the subject of substantial debate. We will now discuss in detail how the different hypoxic modifiers have been used to modify the radiation dose response of tumors.

High-Oxygen-Content Gas Breathing

Because the oxygen supply to tumors is insufficient to meet the needs of all of the tumor cells, radiation-resistant hypoxia develops. Therefore, an obvious solution to improving the tumor's radiation response would be to increase the oxygen supply. This has been tried, both experimentally and clinically, by simply allowing the tumor-bearing host to breathe high-oxygen-content gas mixtures before and during irradiation.

Early experimental studies reported that breathing either oxygen or carbogen (95% O ₂ + 5% CO ₂ ) could substantially enhance the response of murine tumors to radiation and that the best effect was generally seen when the gasses were inspired under hyperbaric (typically 3 atmospheres [3 atm]) rather than normobaric conditions. This is not surprising because hyperbaric conditions would be expected to saturate the blood with oxygen more than normobaric conditions. However, later studies indicated that the radiosensitizations produced by normobaric oxygen or carbogen were quite substantial. Because it is quicker and easier to breathe gas under normobaric conditions, the use of cumbersome, expensive, and complex hyperbaric chambers is probably not necessary.

Clinically, the use of high-oxygen-content gas breathing, specifically under hyperbaric conditions, was introduced relatively early by Churchill-Davidson. Most trials were fairly small and suffered from the applications of unconventional fractionation schemes, but it appeared that the effect of hyperbaric oxygen was superior to radiotherapy given in air, especially when few and large fractions were applied. In the large, multicenter clinical trials conducted by the British Medical Research Council ( Table 3.1 ), the results from both uterine cervix and advanced head and neck tumors showed a significant benefit in local tumor control and subsequent survival. The same findings were not observed in bladder cancer nor were they seen in a number of smaller studies. In retrospect, the use of hyperbaric oxygen was stopped somewhat prematurely. This was partly the result of the introduction of hypoxic radiosensitizers and partly because of problems with patients’ compliance. It has been claimed that hyperbaric treatment caused significant suffering, but the discomfort associated with such a treatment must be considered minor compared to the often life-threatening complications associated with chemotherapy, which is used with less restrictive indications.

The use of high-oxygen-content gas breathing under normobaric conditions to radiosensitize human tumors has also been tried clinically, but it failed to show any dramatic improvement. In the most recent study, this may have been the result of size limitation. In previous studies, it may have been caused by the failure to achieve the optimum preirradiation gas breathing time. Experimental studies have shown that the amount of time is critical for the enhancement of radiation damage and that it can vary from tumor to tumor.

Hypoxic Cell Radiosensitizers

An alternative approach to the hypoxia problem is the use of chemical agents that mimic oxygen and preferentially sensitize the resistant population to radiation. The advantage of these drugs over oxygen is that they are not rapidly metabolized by the tumor cells through which they diffuse; thus, the drugs can penetrate farther than oxygen and reach all of the tumor cells. In the early 1960s, researchers found that the efficiency of radiosensitization was directly related to electron affinity, which ultimately led to in vitro studies demonstrating preferential radiosensitization of hypoxic cells by highly electron-affinic nitroaromatic compounds. Several of these compounds were later shown to be effective at enhancing radiation damage in tumors in vivo. As a result, they underwent clinical testing.

The drugs reaching clinical evaluation include metronidazole, misonidazole, benznidazole, desmethylmisonidazole, etanidazole, pimonidazole, nimorazole, ornidazole, sanazole, and doranidazole. Initial clinical studies were done with metronidazole in brain tumors, followed by a boom in clinical trials exploring the potential of misonidazole as a radiosensitizer in the latter part of the 1970s. The results from the multicenter randomized trials are summarized in Table 3.2 . Most of the trials with misonidazole were unable to generate any significant improvement in radiation response, although a benefit was seen in some trials, especially the second Danish Head and Neck Cancer study (DAHANCA 2), which found a highly significant improvement in the stratification subgroup of pharynx tumors but not in the prognostically better glottic carcinomas. The overall impression of the “misonidazole era” was a prolongation of the inconclusive experience from the hyperbaric oxygen trials, namely, that the problems related to hypoxia had not been ruled out indefinitely. Therefore, the search for more efficient or less toxic hypoxic sensitizers continues. Furthermore, the experience from the misonidazole trials has been taken into account to select a more homogeneous tumor population in which hypoxia is more likely to be present.

Results from subsequent randomized trials with other nitroaromatic compounds have been conflicting. The European pimonidazole trial in uterine cervix cancer was disappointing, whereas the two other multicenter trials in head and neck cancer, using etanidazole, showed no benefit. On the other hand, studies with the low toxic drug nimorazole given to patients with supraglottic and pharynx carcinomas (DAHANCA 5) showed a highly significant benefit in terms of improved locoregional tumor control and disease-free survival rates ( Fig. 3.4 ), thereby confirming the result of the DAHANCA 2 study. More recent trials with the 3-nitrotriazole compound, sanazole (AK-2123), in uterine cervical cancer and doranidazole in pancreatic cancer demonstrated significant improvements in both local tumor control and overall survival.

The potential benefit of using hypoxic radiosensitizers to improve radiotherapy is probably best illustrated from a recent meta-analysis of randomized clinical studies in squamous cell carcinoma of the head and neck. These results are summarized in Fig. 3.5 and clearly showed that radiosensitizer modification of tumor hypoxia significantly improved locoregional tumor control and overall survival, with odds ratios of 0.71 and 0.87, respectively. Although the overall observed gains were small (5%–10% for local control and 0%–6% for survival), they are actually relevant. We can conclude that the nonsignificant outcome of most clinical trials (see Table 3.2 ) is not the result of a biological lack of importance of hypoxia, but in most cases is considered a consequence of poor clinical trials methodology with an overly optimistic study design and an expected treatment gain that goes far beyond what is reasonable. Overall, the results with nitroimidazoles add to the general consensus that if a nontoxic hypoxic modification can be applied, then such treatments may certainly be relevant as a baseline therapy together with radiotherapy for cancers such as advanced head and neck cancer. Such a strategy has been adopted in Denmark, where nimorazole has become part of the standard radiotherapy treatment in cancer of the head and neck.