1979, British Journal of Radiology, 52, 523-535
VOLUME 52 NUMBER 619
JULY 1979
The British Journal of Radiology New horizons in radiation oncology Presidential address delivered on April 13,1978 at The British Institute of Radiology Joint meeting and Annual Congress held at the RIBA, London W1 By J. F. Fowler, D.Sc, Ph.D., F. Inst.P. Gray Laboratory of the Cancer Research Campaign, Mount Vernon Hospital, Northwood, Middlesex H A6 2RN {Received January, 1979)
The scientific foundations of radiotherapy are more strongly based than for any other type of cancer treatment except possibly surgical removal of a small lump. Nowadays advances in the laboratory are put into effect in the clinic more rapidly than ever. This is illustrated by the progressively shorter timescales of the clinical trials in hyperbaric oxygen, then neutron therapy, and now hypoxic-cell radiosensitizers. Each pre-clinical research and gestation time was shorter than the last. And yet throughout the 28 years of my association with the subject people have been saying "radiotherapy will be finished in a few years' time: cancer will be prevented by vaccines or cured by immunotherapy". However, the major advances in cancer treatment during this period have come from radiotherapy and more recently from chemotherapy and from the combination of both with surgery. This seems likely to continue in the foreseeable future, as long as patients present with a palpable lump. No viruses have been identified as causing human cancer, although it is suspected that a limited number of types of cancer may be so caused. No immunological effect has been demonstrated on the "take" or growth-rate of tumours arising spontaneously in man or in experimental animals. It is becoming more widely recognized that valid experimental animal tumours must not be those caused by strong carcinogens or by viruses, or tumours transplanted out of their animal strain of origin, because these carry an unrealistically high immune response as a laboratory artefact. The 80-year-old art of radiotherapy is aided as a rational speciality by the quantitative nature of its supporting scientific disciplines, radiobiology and radiation physics. Radiobiology has been called the
"pharmacology of radiotherapy" and it provides a framework for teaching and for thinking about future improvements in radiotherapy. It has its own "four R's": repair, repopulation, redistribution (around the phases of the cell cycle) and reoxygenation (in tumours). Physical dose distributions
The major gains in radiation treatment of cancer came first from physics, with better depth doses, sparing of skin, and avoidance of heavy doses from absorption in bone, as supervoltage machines were introduced in the 1950s. At present the two approaches of physics and radiobiology are "growing points" in radiotherapy from which a great deal of stimulating new work is coming. It is sometimes said that the physics objective—to increase the ratio of dose absorbed in a target volume to that outside it—has been fully achieved and that the law of diminishing returns operates in that approach. We have not reached that point yet. The argument ignores the fact that each advance in physical technology has so far found radiotherapists willing and able to develop improved ways of treating certain types of cancer. It is unlikely that no gains will result from further improvements in physical dose distribution in tissues. There are several ways of achieving these improvements, including the apparently ultimate in high energy physics technology such as beams of negative pions or accelerated neon particles. Less expensive alternatives include proton beams, betatrons, 20 MeV electron-photon linacs, three-dimensional treatment planning, shrinking-field techniques, implanted needles or wires, after-loading techniques and computer systems.
523
VOL.
I
52, No. 619 J. F. Fowler
Diagnostic and prognostic detection techniques Improvement in physical dose distribution demands improvements in diagnostic accuracy for precise localization of tumour deposits. Here computerized tomographic scanners can help in certain sites, together with grey scale and real-time ultrasound, as reviewed in the 1978 Annual Congress (P. Morley, MacDonald, Husband, Merrick, Kirkpatrick, Van Waes and Shonfield) and improved X-ray film or ion-multiplying detectors. Developments in diagnostic techniques play an important part in improving treatment. These methods will play a further role as they become used, in conjunction with radiobiology and plasma "marker" methods, to track the success or failure of individual treatments. The development of new methods Figure 1 summarizes the growing points in radiotherapy (Fowler, 1977a). Some are derived from ADVANCES
IN
RADIOTHERAPY
1)Better physics dose distributions:— dose -». depth (cm) 2) Optimum Fraction Intervals: 3)Hyperthermia + X-rays: 4)Chemotherapy t X- rays: 5)Hypoxic cells: (a) Hyperbaric Oxygen (b) Neutrons
physics, some from chemistry and some from radiobiology. All of them are already undergoing clinical trials. They have been described by G. E. Adams as the five "highs" of radiotherapy: (1) hyperfractionation; (2) hyperbaric oxygen; (3) high LET; (4) hyperthermia; (5) hypodermics (chemotherapy) combined with radiotherapy. The first three of these are primarily concerned with treatment of the primary tumour. It is obvious however that as chemotherapy (or any other method of dealing with distant metastases) becomes more successful, so also must the control of the primary tumour be improved by better methods than we use conventionally today. One of the best examples of the improvement in treatment of cancer by radiotherapy is shown by the progress is treating Hodgkin's disease (Table I). In the 1940's this disease was regarded as inevitably fatal—a self-fulfilling prophecy, as Kaplan (1962) pointed out. The use of large-volume lymph-system irradiation with 250 kV X rays raised the five-year survival rate from 5% to 35%. When mega voltage beams became available in the 1950s, their physical advantages enabled the cure rate to be raised to about 70%. It was only possible to make use of the physical advantages because of the tremendous improvements in knowledge about pathways of spread in the body, obtained by sustained collaboration between surgeons and pathologists, and by developments in diagnostic radiology, especially lymphangiography. The contribution of radiobiology was to demonsstrate that lymphoid tissues, including lymphoma cells, were more radiosensitive than many other types of mammalian cells, because they have a smaller shoulder on the survival curve and a fairly steep slope; so that radiotherapy was indeed a method worth taking time and trouble to develop. The radiotherapists demonstrated that a high proportion of these patients could be cured, and that, with carefully developed techniques of beam-definition and local shielding, normal tissue complications could be kept to an acceptably low proportion. This was an admirable example of collaboration between determined and scientifically-minded radiotherapists, TABLE I
(c) Pi mesons
IMPROVEMENT IN 5-YEAR SURVIVAL FROM HODGKIN'S DISEASE
(d) Hypoxic cell Radio sensitizers
FIG.1.
Some of the present growing points in radiotherapy (Fowler, 1977a). 524
Date
Treatment
1940 1960 1970
Considered inevitably fatal 250 kVX rays Supervoltage mantle with exploratory surgery and lymphangiography Supervoltage plus chemotherapy
1974
5-year survival 5% 35% 73% 79%
JULY 1979
New horizons in radiation oncology. Presidential address, April 13, 1978 physicists, surgeons, pathologists and diagnostic radiologists. But there is yet more collaboration to relate: the addition of quadruple chemotherapy has raised the five-year survival rate to nearly 80%. Clinical trials The sequence of advances in treating Hodgkin's disease, taking place in steps over some 20 to 30 years, seems to be a likely pattern for improvement in treating other types of cancer. It requires willing collaboration between specialities, the gradual collection and careful analysis of pathological and clinical data, and above all the determination to improve results. It also requires patience, as does much else in cancer research. The difference between 5% and 79% survival rate is easy to demonstrate with statistical significance but it occurred at an average rate of only 3% per year. Since an improvement of about 20% requires several hundred patients before it can be substantiated (Boag et al., 1971), no "proof" that clinical progress was being made could be expected in less than six or seven years. This remains true for future methods. It is necessary to pay more attention to the reliable collection of data on results of cancer treatment. Every patient treated should contribute in some way to the pool of data. Real improvements in clinical results could then be recognized at the earliest opportunity and false claims would not receive undue publicity. FRACTIONATION IN RADIOTHERAPY
It has long been suggested that one of the reasons for the successes of multiple-small-dose radiotherapy might be that cells in the tumour are less capable of repairing sublethal injury than cells in normal tissues, although reviewers have not agreed on this (Withers, 1970; Field, 1970; Phillips, 1972; Denekamp and Stewart, 1979). One reason for this was suggested by Littbrand and Revesz (1969) on the basis of loss of repair capability if cells are deeply anoxic, with an oxygen tension as low as 5 ppm (Pettersen et ah, 1973). However, whether such low levels of hypoxia exist in vivo is not known. Hill and Bush (1976) have suggested slower repair in mouse tumours than in normal tissues, so that at three or four hours after a dose of X rays, cells in normal tissues would have repaired their sublethal radiation injury but cells in tumours would not. These straws in the wind have led to the clinical use of two or three small doses of X rays on each treatment day in a number of centres including Houston, Boston, Vancouver, Stockholm, Goteborg, Portsmouth, Rome and Bologna. No adverse reports have yet been
made although we shall have to wait the usual six or seven years after starting before definitive reports can appear. The studies have been under way for more than five years in some of the centres and preliminary results have been reported. For example Littbrand from Stockholm, using 84 fractions of 100 rad each (given at three per day over six weeks with a gap in the middle) has reported better local control of bladder carcinoma and better survival of the patients than with the conventional 32x200 rad treatment (Littbrand, 1975; 1978) although these results must be regarded as still preliminary. Douglas in Vancouver (Douglas, 1977) has reported an increased mean survival time for brain tumour patients treated with 60 fractions given as 3 X 100 rad per day over four weeks, as compared with a historical control group of previous patients given 20 fractions in the same overall time. During the last week of treatment he used smaller sized fields. Turesson and Notter (1976) in Goteborg have shown that acceptable levels of skin reaction can be obtained using three doses per day and that the late skin reactions that follow are only a little greater, not significantly so, than the usual expectation (Turesson, 1978). (This does not seem to be true for either two fractions per week or for very extended schedules exceeding seven weeks overall time, when late reactions are appreciably greater than might have been expected from the early ones.) The schedules described above make use of several X-ray doses per day and are called superfractionation or hyperfractionation. One reason for employing this approach is the assumption that less repair of X-ray damage occurs in tumours than in normal tissues. What can radiobiology contribute to this idea? Evidence has been obtained that the capacity to repair X-ray injury in some types of mouse tumour at 24 hours is less than in normal tissues and that this might result from a steeper initial slope to the survival curve for tumour cells (Denekamp and Stewart, 1979). However, this is not a uniform finding in all types of mouse tumour studied. Further work is required on the physiological and biochemical conditions which lead to less or slower repair. At the other extreme, where only one or two radiation doses may be given each week, recent clinical evidence has been reported in the literature that two fractions per week cause more damage to normal tissues than would be expected either from the usual NSD formula or from doses yielding local control of tumours as good as that from conventional daily fractionation (Bates and Peters, 1975; Withers
525
VOL.
52, No. 619 J. F. Fowler
et al., 1978). The implication leads in the same direction as the previous paragraphs: towards a preference for multiple small fractions. Superfractionation means that the overall time need not be extended and might even be shortened. One further remark concerning overall time in fractionation schedules is worth noting. Where the limiting radiation damage is late injury, these tissues must be turning over very slowly, or their damage would be expressed sooner. It would therefore make little difference to injury in these tissues whether the overall time is two weeks or ten, if proliferation is the only factor which requires more dose when the overall time is lengthened. They will not proliferate in any case during the radiation treatment time (Denekamp, 1975; Denekamp and Fowler, 1977). However, the phenomenon of slow repair (Field and Hornsey, 1976; 1977) explains in a different way why lengthening the overall time also causes a reduction in late injury. This reduction will be smaller than for acute skin reactions, because slow repair occurs mostly in the first two weeks after starting irradiation whereas accelerated proliferation in acutely reacting tissues occurs after this time (Denekamp, 1973). By the same argument, shortening the overall time (to say one-half) will not increase late injury as much as it will increase acute reactions. Therefore when late injury is what limits the radiation dose level, there should be more flexibility in choice of overall time than the traditional "rules of thumb" (often based on acute skin reactions) have allowed. This is important when we consider below the suggestions to shorten overall treatment times provided that hypoxic cells can be effectively dealt with. METHODS OF DEALING WITH HYPOXIC CELLS
The much greater resistance of hypoxic cells to X rays has been recognized for many years as one possible reason for the failure to control tumours locally (Gray et al., 1953). The presence of hypoxic cells in both animal and human tumours is accepted on histological grounds (Thomlinson and Gray, 1955) but the continued presence of hypoxic cells throughout a course of radiotherapy has been questioned because of possible reoxygenation in the tumour (Thomlinson, 1969; Van Putten and Kailman, 1968). Hyperbaric oxygen The recent results of the MRC hyperbaric oxygen (HBO) trials have shown unequivocally that hypoxic cells are present in human tumours, at least of some
types, and are a problem even in treatment schedules using 25-30 small fractions of X rays. Table II summarizes these results from the use of HBO in cancer of the uterine cervix (Watson et al., 1968). The results for local control at five years are significantly better for HBO. Table III shows similar results for head and neck cancer (Henk et al., 1977; Henk and Smith, 1977). No improvement was seen, however, for Ca bladder when HBO was used. In both the cervix and head and neck trials the increase in local control was about 20%. Because of distant metastases the increase in survival was smaller than the increase in local control but was just statistically significant (P
VOLUME 52 NUMBER 619
JULY 1979
The British Journal of Radiology New horizons in radiation oncology Presidential address delivered on April 13,1978 at The British Institute of Radiology Joint meeting and Annual Congress held at the RIBA, London W1 By J. F. Fowler, D.Sc, Ph.D., F. Inst.P. Gray Laboratory of the Cancer Research Campaign, Mount Vernon Hospital, Northwood, Middlesex H A6 2RN {Received January, 1979)
The scientific foundations of radiotherapy are more strongly based than for any other type of cancer treatment except possibly surgical removal of a small lump. Nowadays advances in the laboratory are put into effect in the clinic more rapidly than ever. This is illustrated by the progressively shorter timescales of the clinical trials in hyperbaric oxygen, then neutron therapy, and now hypoxic-cell radiosensitizers. Each pre-clinical research and gestation time was shorter than the last. And yet throughout the 28 years of my association with the subject people have been saying "radiotherapy will be finished in a few years' time: cancer will be prevented by vaccines or cured by immunotherapy". However, the major advances in cancer treatment during this period have come from radiotherapy and more recently from chemotherapy and from the combination of both with surgery. This seems likely to continue in the foreseeable future, as long as patients present with a palpable lump. No viruses have been identified as causing human cancer, although it is suspected that a limited number of types of cancer may be so caused. No immunological effect has been demonstrated on the "take" or growth-rate of tumours arising spontaneously in man or in experimental animals. It is becoming more widely recognized that valid experimental animal tumours must not be those caused by strong carcinogens or by viruses, or tumours transplanted out of their animal strain of origin, because these carry an unrealistically high immune response as a laboratory artefact. The 80-year-old art of radiotherapy is aided as a rational speciality by the quantitative nature of its supporting scientific disciplines, radiobiology and radiation physics. Radiobiology has been called the
"pharmacology of radiotherapy" and it provides a framework for teaching and for thinking about future improvements in radiotherapy. It has its own "four R's": repair, repopulation, redistribution (around the phases of the cell cycle) and reoxygenation (in tumours). Physical dose distributions
The major gains in radiation treatment of cancer came first from physics, with better depth doses, sparing of skin, and avoidance of heavy doses from absorption in bone, as supervoltage machines were introduced in the 1950s. At present the two approaches of physics and radiobiology are "growing points" in radiotherapy from which a great deal of stimulating new work is coming. It is sometimes said that the physics objective—to increase the ratio of dose absorbed in a target volume to that outside it—has been fully achieved and that the law of diminishing returns operates in that approach. We have not reached that point yet. The argument ignores the fact that each advance in physical technology has so far found radiotherapists willing and able to develop improved ways of treating certain types of cancer. It is unlikely that no gains will result from further improvements in physical dose distribution in tissues. There are several ways of achieving these improvements, including the apparently ultimate in high energy physics technology such as beams of negative pions or accelerated neon particles. Less expensive alternatives include proton beams, betatrons, 20 MeV electron-photon linacs, three-dimensional treatment planning, shrinking-field techniques, implanted needles or wires, after-loading techniques and computer systems.
523
VOL.
I
52, No. 619 J. F. Fowler
Diagnostic and prognostic detection techniques Improvement in physical dose distribution demands improvements in diagnostic accuracy for precise localization of tumour deposits. Here computerized tomographic scanners can help in certain sites, together with grey scale and real-time ultrasound, as reviewed in the 1978 Annual Congress (P. Morley, MacDonald, Husband, Merrick, Kirkpatrick, Van Waes and Shonfield) and improved X-ray film or ion-multiplying detectors. Developments in diagnostic techniques play an important part in improving treatment. These methods will play a further role as they become used, in conjunction with radiobiology and plasma "marker" methods, to track the success or failure of individual treatments. The development of new methods Figure 1 summarizes the growing points in radiotherapy (Fowler, 1977a). Some are derived from ADVANCES
IN
RADIOTHERAPY
1)Better physics dose distributions:— dose -». depth (cm) 2) Optimum Fraction Intervals: 3)Hyperthermia + X-rays: 4)Chemotherapy t X- rays: 5)Hypoxic cells: (a) Hyperbaric Oxygen (b) Neutrons
physics, some from chemistry and some from radiobiology. All of them are already undergoing clinical trials. They have been described by G. E. Adams as the five "highs" of radiotherapy: (1) hyperfractionation; (2) hyperbaric oxygen; (3) high LET; (4) hyperthermia; (5) hypodermics (chemotherapy) combined with radiotherapy. The first three of these are primarily concerned with treatment of the primary tumour. It is obvious however that as chemotherapy (or any other method of dealing with distant metastases) becomes more successful, so also must the control of the primary tumour be improved by better methods than we use conventionally today. One of the best examples of the improvement in treatment of cancer by radiotherapy is shown by the progress is treating Hodgkin's disease (Table I). In the 1940's this disease was regarded as inevitably fatal—a self-fulfilling prophecy, as Kaplan (1962) pointed out. The use of large-volume lymph-system irradiation with 250 kV X rays raised the five-year survival rate from 5% to 35%. When mega voltage beams became available in the 1950s, their physical advantages enabled the cure rate to be raised to about 70%. It was only possible to make use of the physical advantages because of the tremendous improvements in knowledge about pathways of spread in the body, obtained by sustained collaboration between surgeons and pathologists, and by developments in diagnostic radiology, especially lymphangiography. The contribution of radiobiology was to demonsstrate that lymphoid tissues, including lymphoma cells, were more radiosensitive than many other types of mammalian cells, because they have a smaller shoulder on the survival curve and a fairly steep slope; so that radiotherapy was indeed a method worth taking time and trouble to develop. The radiotherapists demonstrated that a high proportion of these patients could be cured, and that, with carefully developed techniques of beam-definition and local shielding, normal tissue complications could be kept to an acceptably low proportion. This was an admirable example of collaboration between determined and scientifically-minded radiotherapists, TABLE I
(c) Pi mesons
IMPROVEMENT IN 5-YEAR SURVIVAL FROM HODGKIN'S DISEASE
(d) Hypoxic cell Radio sensitizers
FIG.1.
Some of the present growing points in radiotherapy (Fowler, 1977a). 524
Date
Treatment
1940 1960 1970
Considered inevitably fatal 250 kVX rays Supervoltage mantle with exploratory surgery and lymphangiography Supervoltage plus chemotherapy
1974
5-year survival 5% 35% 73% 79%
JULY 1979
New horizons in radiation oncology. Presidential address, April 13, 1978 physicists, surgeons, pathologists and diagnostic radiologists. But there is yet more collaboration to relate: the addition of quadruple chemotherapy has raised the five-year survival rate to nearly 80%. Clinical trials The sequence of advances in treating Hodgkin's disease, taking place in steps over some 20 to 30 years, seems to be a likely pattern for improvement in treating other types of cancer. It requires willing collaboration between specialities, the gradual collection and careful analysis of pathological and clinical data, and above all the determination to improve results. It also requires patience, as does much else in cancer research. The difference between 5% and 79% survival rate is easy to demonstrate with statistical significance but it occurred at an average rate of only 3% per year. Since an improvement of about 20% requires several hundred patients before it can be substantiated (Boag et al., 1971), no "proof" that clinical progress was being made could be expected in less than six or seven years. This remains true for future methods. It is necessary to pay more attention to the reliable collection of data on results of cancer treatment. Every patient treated should contribute in some way to the pool of data. Real improvements in clinical results could then be recognized at the earliest opportunity and false claims would not receive undue publicity. FRACTIONATION IN RADIOTHERAPY
It has long been suggested that one of the reasons for the successes of multiple-small-dose radiotherapy might be that cells in the tumour are less capable of repairing sublethal injury than cells in normal tissues, although reviewers have not agreed on this (Withers, 1970; Field, 1970; Phillips, 1972; Denekamp and Stewart, 1979). One reason for this was suggested by Littbrand and Revesz (1969) on the basis of loss of repair capability if cells are deeply anoxic, with an oxygen tension as low as 5 ppm (Pettersen et ah, 1973). However, whether such low levels of hypoxia exist in vivo is not known. Hill and Bush (1976) have suggested slower repair in mouse tumours than in normal tissues, so that at three or four hours after a dose of X rays, cells in normal tissues would have repaired their sublethal radiation injury but cells in tumours would not. These straws in the wind have led to the clinical use of two or three small doses of X rays on each treatment day in a number of centres including Houston, Boston, Vancouver, Stockholm, Goteborg, Portsmouth, Rome and Bologna. No adverse reports have yet been
made although we shall have to wait the usual six or seven years after starting before definitive reports can appear. The studies have been under way for more than five years in some of the centres and preliminary results have been reported. For example Littbrand from Stockholm, using 84 fractions of 100 rad each (given at three per day over six weeks with a gap in the middle) has reported better local control of bladder carcinoma and better survival of the patients than with the conventional 32x200 rad treatment (Littbrand, 1975; 1978) although these results must be regarded as still preliminary. Douglas in Vancouver (Douglas, 1977) has reported an increased mean survival time for brain tumour patients treated with 60 fractions given as 3 X 100 rad per day over four weeks, as compared with a historical control group of previous patients given 20 fractions in the same overall time. During the last week of treatment he used smaller sized fields. Turesson and Notter (1976) in Goteborg have shown that acceptable levels of skin reaction can be obtained using three doses per day and that the late skin reactions that follow are only a little greater, not significantly so, than the usual expectation (Turesson, 1978). (This does not seem to be true for either two fractions per week or for very extended schedules exceeding seven weeks overall time, when late reactions are appreciably greater than might have been expected from the early ones.) The schedules described above make use of several X-ray doses per day and are called superfractionation or hyperfractionation. One reason for employing this approach is the assumption that less repair of X-ray damage occurs in tumours than in normal tissues. What can radiobiology contribute to this idea? Evidence has been obtained that the capacity to repair X-ray injury in some types of mouse tumour at 24 hours is less than in normal tissues and that this might result from a steeper initial slope to the survival curve for tumour cells (Denekamp and Stewart, 1979). However, this is not a uniform finding in all types of mouse tumour studied. Further work is required on the physiological and biochemical conditions which lead to less or slower repair. At the other extreme, where only one or two radiation doses may be given each week, recent clinical evidence has been reported in the literature that two fractions per week cause more damage to normal tissues than would be expected either from the usual NSD formula or from doses yielding local control of tumours as good as that from conventional daily fractionation (Bates and Peters, 1975; Withers
525
VOL.
52, No. 619 J. F. Fowler
et al., 1978). The implication leads in the same direction as the previous paragraphs: towards a preference for multiple small fractions. Superfractionation means that the overall time need not be extended and might even be shortened. One further remark concerning overall time in fractionation schedules is worth noting. Where the limiting radiation damage is late injury, these tissues must be turning over very slowly, or their damage would be expressed sooner. It would therefore make little difference to injury in these tissues whether the overall time is two weeks or ten, if proliferation is the only factor which requires more dose when the overall time is lengthened. They will not proliferate in any case during the radiation treatment time (Denekamp, 1975; Denekamp and Fowler, 1977). However, the phenomenon of slow repair (Field and Hornsey, 1976; 1977) explains in a different way why lengthening the overall time also causes a reduction in late injury. This reduction will be smaller than for acute skin reactions, because slow repair occurs mostly in the first two weeks after starting irradiation whereas accelerated proliferation in acutely reacting tissues occurs after this time (Denekamp, 1973). By the same argument, shortening the overall time (to say one-half) will not increase late injury as much as it will increase acute reactions. Therefore when late injury is what limits the radiation dose level, there should be more flexibility in choice of overall time than the traditional "rules of thumb" (often based on acute skin reactions) have allowed. This is important when we consider below the suggestions to shorten overall treatment times provided that hypoxic cells can be effectively dealt with. METHODS OF DEALING WITH HYPOXIC CELLS
The much greater resistance of hypoxic cells to X rays has been recognized for many years as one possible reason for the failure to control tumours locally (Gray et al., 1953). The presence of hypoxic cells in both animal and human tumours is accepted on histological grounds (Thomlinson and Gray, 1955) but the continued presence of hypoxic cells throughout a course of radiotherapy has been questioned because of possible reoxygenation in the tumour (Thomlinson, 1969; Van Putten and Kailman, 1968). Hyperbaric oxygen The recent results of the MRC hyperbaric oxygen (HBO) trials have shown unequivocally that hypoxic cells are present in human tumours, at least of some
types, and are a problem even in treatment schedules using 25-30 small fractions of X rays. Table II summarizes these results from the use of HBO in cancer of the uterine cervix (Watson et al., 1968). The results for local control at five years are significantly better for HBO. Table III shows similar results for head and neck cancer (Henk et al., 1977; Henk and Smith, 1977). No improvement was seen, however, for Ca bladder when HBO was used. In both the cervix and head and neck trials the increase in local control was about 20%. Because of distant metastases the increase in survival was smaller than the increase in local control but was just statistically significant (P
