1983, The British Journal of Radiology, 56, 851-857

NOVEMBER

1983

The incidence and management of osteoradionecrosis of the jaws following head and neck radiotherapy
By *Frank Coffin, F.R.C.S., F.D.S. Royal Marsden Hospital, Fulham Road, London SW3

{Received March 1983 and in revised form May 1983)

ABSTRACT

Only 22 cases of osteoradionecrosis (ORN) of the jaws so severe as to require resection have been reported in 2,853 cases of head and neck tumours receiving radiation in doses capable of causing ORN between 1.1.1970 and 31.12.1981. But for surgical interference, including dental extractions both preand post-therapy, the incidence would have been even lower. The management of the irradiated patient and of established necrosis is discussed.

The problem of necrosis of tissues after curative doses of radiation has dogged the use of radiotherapy since its inception. Necrosis of soft tissues has largely been controlled by improvements in techniques but in the head and neck region jaw-bone necrosis still occurs, often long after the original lesion has been dealt with successfully. It was realised very early by the original radium users that damage to the bone, usually as a result of the extraction of teeth, would precipitate bone necrosis. Thus it became the accepted practice to remove all teeth before treatment began, although, as experience was gained both with radium and so-called deep X-ray machines, it was agreed that teeth could be removed with little risk up to two months after therapy had finished. The dose to the tissues at this time was in the region of 4000 rad (40 Gy) (Del Regato, 1939). With the introduction of supervoltage therapy (by telecobalt machines and linear accelerators), increased doses could be applied to the tumour without risk of skin breakdown. This enabled hitherto untreatable tumours to be treated, or tumours hitherto insufficiently treated to be dealt with more adequately. Experience with the new machines showed that whilst treatment was more effective with the increased dosage60-70 Gythe risks to the bone increased, so that it was suggested (Coffin, 1964) that even pretreatment extractions were best avoided. Since that time a more and more conservative view has been taken and, over the last ten years, extractions have been avoided almost completely both pre- and post-therapy.
MANAGEMENT OF ORAL CONDITIONS

Emphasis is laid on the careful control of dental caries by the patient and his general dental practitioner, with the strict understanding that, should interference *Present address: 53 New Cavendish Street, London Wl

with the bone in any way be considered desirable, the patient should be reassessed and the procedure carried out, if agreed, at the hospital. The general dental practitioners are very content with this arrangement provided they are put in the picture early. A general dental practitioner assistant in the department is of the greatest help for those patients without their own practitioner. The patients themselves are of course the key to the situation, in that they must clearly understand the need for continual obsessional dental care and the necessity to avoid extractions. They must be disabused of the idea that the "prohibition of extractions" lapses with time. Although the method by which the general dental practitioner restores his patient's teeth is in his own hands, if advice is sought from the hospital (because radiation-induced decreased salivation tends to cause persistently recurring caries), it is suggested that very elaborate and costly crowns and bridges are best avoided. Root treatments are resorted to frequently, and strangely, the inability of the irradiated tissues to develop inflammation results in a very much lower incidence of apical abscesses than one would expect from teeth so repeatedly attacked by caries. Should infection occur or extractions be absolutely necessary, these should not be delegated to junior staff but be dealt with by the most skilful operator available. Tetracycline 250 mg q.d.s. is given before extraction for seven days. This is reduced to t.d.s. for 14 days and continued at b.d. for as long as the wound or infection is present. The same regime is used for established bone necrosis and goes on until the sequestrum is shed or removed. Such extractions, it must again be stressed, are not undertaken lightly, it being better to wait until teeth are so loose as to be removed by the patient with his own fingers, or in danger of being inhaled, and even then a necrosis may occur. Certainly any suggestion of a socalled "surgical" extraction is ill advised, both pre- and post-therapy. It only adds a surgical insult to radiation injury. This very conservative attitude is appreciated by the modern patient who strongly resists the removal of teeth because of future radiotherapy, and may well refuse the radiotherapy in order to preserve his
851

VOL.

56, No. 671 Frank Coffin eventually a large sequestrum has to be removed due to a pathological fracture taking place at the junction of the dead and living bone. On several occasions a pathological fracture without displacement has been nursed through the long period of sequestration to a fibrous union with no infection remaining. Monilial infections are exceedingly rare, providing the patient is obsessional about cleaning his mouth, and his general dental care. By far the best mouthbath for both general health of the mouth, and inhibition of caries and monilial infections, is dilute chlorhexidine. 5 ml of 0.2% w/v aqueous chlorhexidine is mixed with 200 ml (about \ pint or \ litre) warm water. One mouthful of this mixture is held in the mouth for 1 minute after every meal and once in between i.e. 7-8 times a day. Stronger solutions tend to "burn". Unfortunately the makers of the commercially available 0.2% w/v solution advise that their solution be used undiluted, and thus patients tend not to use it at all. If it is decided to attack apparently healthy bone in any way (e.g., for a soft tissue recurrence) the preextraction regime is followed. So slow is the bone turnover in the irradiated jaw that firstly, the full extent of necrosis may not declare itself, on a radiograph for instance, for a year or more, although the tooth socket itself fails to heal in the usual few days. The mucosa does not close over it, and leaves an infected painful cavity with bare bone clearly apparent, a situation mimicking the common "localised osteitis" known as a "dry socket" in dental parlance, common when undue trauma or manipulation of a tooth is used for removal in an unirradiated mouth. Secondly, it means that broken-down teeth or residual roots give very little trouble. They can be conserved by means of adequate root treatment and dentures with a soft rubber lining can be worn over them. The decision to remove the dead bone surgically should be delayed for as long as possible. The two situations which precipitate intervention are: (a) an unstable pathological fracture with at least one fragment, usually the posterior, being displaced forwards and upwards into the mouth, and which by its sharp edges chafes the tongue, cheek or lips, or (b) the appearance of chronic fistulae from the mouth to the skin (which is threatened with dissolution), and which are not quickly aborted by raising the tetracycline intake to full doses. The removal should be via the mouth and very drastic, extending back from the fracture site to include the whole of the horizontal and vertical rami of the mandible including the coronoid process. As a rule the condyle appears to survive by virtue of its muscular and ligamentous attachments. As the pathological fracture often takes place between completely dead and partially viable bone anteriorly in the canine region, the bulk of the necrosis is behind the canine, so that a much smaller amount of partially viable bone needs removal anteriorly before

dentition. Such patients who have been so threatened are all the more willing to adopt the stringent mouth care advocated. The underlying pathology of the bone necrosis is accepted as a marked decrease in blood supply to the part due to endarteritis obliterans caused by the irradiation. The residual blood supply may well maintain the bone intact, but is unable to supply the extra flow needed for healing, growth, repair, or resistance to infection. Thus children's jaw growth may be arrested, or slow healing of extraction wounds occur, allowing ingress of infection and thereby the tissue death. The dead part is then shed by inflammation occurring in the healthy tissue at the periphery of the dead tissue. The line of demarcation between dead or dying tissue and that with a blood supply adequate for repair is ill defined and will only declare itself with time. Although related to the fields applied, the necrosis will extend much further along the course of the obliterated blood supply than might be expected. For example, in a mandible which has been irradiated during the treatment of a carcinoma of the posterior third of the tongue, because the reduced inferior alveolar vessel joins up with no vessels of comparable size and most of its terminal capillaries have been obliterated, any bone necrosis will extend well forward of the fields, possibly as far as the incisors or canine in the situation quoted, and the line of separation will be in the mental region rather than the edge of the fields in the molar area. The rationale for the use of tetracycline as the antibiotic of choice is the fact that it is chelated by the calcium of the bone, being incorporated into the bone crystals, and has been used to study bone growth by this ability. It is given by mouth, between meals, with water, and not with milk, antacids or iron, as these combine with the tetracycline in the stomach and prevent absorption into the blood stream. The last dose at night should be about one hour before retiring so that the tablet does not lodge in the oesophagus where the low pH of its solution may cause oesophageal ulceration (Channer & Hollander, 1981). As it is taken up by the bone and released as the normal turnover of bone takes place, there is always some antibiotic available should the patient miss the occasional dose. This drug can be used blindly without bacterial control. The amount of analgesic needed is markedly reduced, within 24 hours. For an established necrosis seen for the first time the pre-extraction regime is followed, but the final b.d. dosage is continued as long as dead bone is in place, which may be several years, increasing the dose to t.d.s. for a week to control the occasional flare in inflammation. Over the last 20 years very little trouble has been found with these long courses, providing the dose is lowered as suggested. By this method the patient has the use of his jaw functionally and cosmetically for several years, even if

852

NOVEMBER 1983

Osteoradionecrosis of the jaws viable bone is reached. However, the precise extent is always in doubt, and usually a small sequestrum remains or forms as a result of the surgery, and this is best allowed to separate itself. Some guidance to the extent needed is given by the ease with which the periosteum is stripped. In the affected area it comes off very easily if not already free; it is thus turned back until the normal resistance to the periosteal elevator is felt, when something like healthy bone is present. At operation, on occasion, when a patient has been receiving tetracycline for some time, the living bone contains this antibiotic and appears greenish in colour. Wood's light may make it fluoresce. The dead bone remains white, not having taken up the tetracycline. This process of gradual decay, fracture and separation may take several years, during which control by the method suggested gives the patient full function and appearance, with little pain. A varying degree of trismus may be an unfortunate discomfort. There can be no possibility of a free bone graft "taking" if inserted into the debilitated soft tissue bed. The only way such a reconstruction is possible is by providing a totally new bed with a completely fresh blood supply to the part by some form of pedicle graft, taken from unirradiated tissue. There are many forms of such pedicles. Such repair can be a task formidable both to the surgeon and patient, may not be successful and will not in any case restore full function and appearance. From time to time enquiry is made by colleagues as to the management of the problem at the Royal Marsden Hospital. Doubt as to the efficacy of the plan is displayed, principally by radiotherapists, but a recent questionnaire by an oral surgeon, troubled by what he felt was an unduly high incidence of bone necrosis in his catchment area, prompted us to look into our own figures more closely, our impression being that the incidence was low. lower alveoli; paranasal sinuses; salivary glands; and some extensive laryngeal carcinomata with high cervical nodes. The patients came from no precise catchment area or ethnic group. Besides British nationals in the home counties, a fair number were from farther afield in the British Isles, some expatriates returned to England for treatment, and a good proportion were from the Mediterranean countries, Africa and the Far East. More than half of the patients were secondary referrals, having had some form of primary treatment elsewhere, and were returned to their referring point for follow-up, from whence we obtained follow-up information in addition to that derived from an occasional visit from the patient. All clinicians dealing with the patients at risk were asked to identify personally their own cases, and the records of patients obtained from the computer were examined by the author. Because the interval between precipitating trauma and presentation of the necrosis was sometimes long, there were times when no cases were being actively treated. Such a period occurred towards the end of 1980, but in the following few months several cases presented. Examination of the cases shows that bone necrosis develops in two forms. A. Minor, as a series of small sequestra which separate spontaneously after varying periods of weeks or months. Typically these occur after dental extractions have been undertaken in the maxilla, or where a radioactive implant has been used by itself for the tongue or floor of mouth, and thus has been in close proximity to the inner side of the mandible, even where no surgical trauma has been inflicted upon the mandible. Where a lesion of the floor of mouth or tongue has had a full course of external beam therapy and a small local residuum is dealt with by an implant, then it is likely that a thin (3-4 mm) flake of inner mandibular THE ROYAL MARSDEN SERIES cortex will die and be shed after a period of time which It was decided to collate all head and neck tumours may extend to years. The degree of this death is very treated at the Marsden's two hospitals and try to find similar to that caused by 3 x 2 minutes freeze/thaw the incidence of bone necrosis, the degree and cycles of the N2O cryoprobe. A similar piece of bone precipitating causes, and thus the degree of efficacy of will separate should the cryoprobe be used instead of an our management methods. implant after a course of external irradiation. From 1st January 1970 to 31st December 1981, 8218 These small areas can be seen clinically but cannot be head and neck cases were treated at the Royal Marsden demonstrated radiologically even by the best intra-oral Hospitals. Certain of these cases were excluded from and occlusal films; just as small degrees of involvement consideration as the jaws were not irradiated, or were of the inner table of the mandible by carcinoma of the treated to such low doses as to incur little risk to the floor of the mouth cannot be demonstrated radiolobone, i.e.: gically, and have to be assessed clinically. brain; thyroid; most laryngeal carcinomas; eye and B. Major, where a necrosis occurs of such an extent orbit; basal and squamous carcinomata of the skin; and as to involve the entire thickness of the jaw, and a such other lesions which were treated by surgery alone. pathological fracture is inevitable. It can be easily seen This left 2853 cases to be considered under the following clinically as bare bone, and demonstrated by almost any headings: form of radiology by virtue of the gross destruction of nasopharynx; pharynx (including tonsils, faucial pillars tissue. This is confined almost entirely to bone composed of and soft palate); tongue and floor of mouth; upper and
853

VOL.

56, No. 671 Frank Coffin

mainly cortical structure such as the mandible or malar. Only one major case in the maxilla occurred in this series, where the whole hard palate and premaxilla sequestrated. This will be discussed later. The freedom of the maxilla from major necrosis would seem to be due to its structure by thin plates of bone covered by very vascular membranes and facial tissues. Necroses of the minor type cause only easilycontrolled discomfort. The granulations around the separating sequestra can be mistaken for recurrence of tumour, but can be monitored by simple biopsy. Occasionally the larger flakes take a long time to separate but again the discomfort of separative inflammation can be controlled by long-term low-dose tetracycline. Thus the form of radionecrosis that is of real concern to the patient and his attendants is the major type and is confined completely to the mandible in our series with the one exception mentioned. Twenty-two such cases were found.

B. Surgery during irradiation (1 patient) Patient Bl, with a carcinoma of the nasal septum and premaxilla, had a dental clearance of all his upper teeth soon after commencing a full course of external irradiation. The sockets were very slow to heal and three months after the end of irradiation almost the whole bony palate necrosed and had to be removed. This is the only massive necrosis of the maxilla in the series. C. Surgery after irradiation (8 patients) Eight patients had major surgical interference with the mandible after full courses of external irradiation to carcinomas of the tonsil or faucial pillar. In spite of the usual precise instructions to the patient personally and written advice to the patient's dental surgeon about the inadvisability of dental extractions, patient Cl in Teheran and patient C2 in Ireland allowed themselves to have a lower wisdom tooth removed from the irradiated area. It was possible to nurse patient C2 with a pathological fracture over a period of two years with continuous low-dose tetracycline until he developed a fibrous union with full painless function. The patient Cl in Teheran, now back in England, has recently had a large sequestrectomy involving the whole of one side of his mandible. Patient C3 in Ireland had his mandible divided through the irradiated area to aid visualisation for a block dissection of the neck. The mandible was wired together at the end of the operation, but was removed four months later for necrosis. Patient C4 had upper and lower molar teeth removed from the irradiated area in Greece, thus inducing a necrosis of the mandible necessitating surgical sequestration. Patient C5 presented with a major mandibular necrosis after two operations on his left parotid salivary gland, followed by a course of electrons in Germany and later extraction of teeth there. It was not possible to save his mandible, which eventually sequestrated and had to be removed. He later died from pulmonary metastases. Patient C6 had a carcinoma of the infra-surface of the tongue treated by 51 Gy by external beam, followed by 70 Gy as a radium implant. Over the next ten years she twice underwent deep biopsy to bone because of suspected recurrence, but these happily proved to be non-malignant. In the eleventh year she developed a massive necrosis of the horizontal ramus with multiple fistulae externally. Two years later still, after sequestrectomy, she is well, healed and comfortable. Patients C7 and C8, both from India, both with carcinoma of the tongue and treated by external irradiation (one supplemented by an iridium implant to 28 Gy), both complained of persistent severe pain in the region of the healed lesion. In both, a natural tooth, present in the area, was only extracted at the Royal Marsden Hospital when it became so loose as to be in
854

RESULTS

When the case histories of these 22 major necroses were examined they showed certain definite patterns: A. Surgery before irradiation B. Surgery during irradiation C. Surgery after irradiation D. Irradiation alone. A. Surgery before irradiation (9 patients) Patients Al and A2 each had difficult lower wisdom teeth removed in preparation for external beam therapy of 65 Gy to a carcinoma of the floor of mouth and tongue. Bone necrosis began about one year after first presentation to the referring clinicians. Patient A3 had a full dental clearance before three courses of VBM (vinblastine, bleomycin and methotrexate) (Price et al, 1975) as a preparation for external beam therapy of 66 Gy to a carcinoma of the oropharynx. The necrosis presented one year after biopsy by the referring clinician. Five patients (A4, A5, A6, A7, A8) had what appears to have been a biopsy-removal of a carcinoma of the floor of the mouth, combined with surgical removal of teeth in the area of the tumour, together with varying amounts of alveolar bone. On referral each needed external beam therapy in the region of 65 Gy to residual tumour. Necrosis began some 3-4 months after completion of the radiotherapy. Patient A9 orginally had facial pain misdiagnosed by her dental surgeon who removed a first lower molar for it. The pain did not improve, and was found to be due to a carcinoma of the alveolus by the oral surgeon to whom she was referred. In spite of the relatively small dose of 40 Gy, she developed a large area of necrosis three months after finishing therapy, for which she later had a hemimandibulectomy.

NOVEMBER 1983

Osteoradionecrosis of the jaws danger of inhalation. Both developed necrosis and pathological fracture in spite of persistent tetracycline in an attempt to tide them over. One patient died of general debilitation soon after the fracture developed, the other is doing well after a massive sequestrectomy, and his trismus is improving. D. Irradiation alone (4 patients) Patient Dl presented with an established necrosis of the right mandible, having been treated in Chile. No information is to hand regarding dosage save that the patient says the treatment extended over nine weeks. The mandible is still in continuity after three years, although incredibly fragile, in spite of gross loss of bone by separation of many small sequestra, and removal of molars with fingers as they become dangerously loose. Continuous low-dose tetracycline keeps him pain-free. He has recently developed a second primary in the left faucial region which has been dealt with surgically. Patient D2, after three iridium implants of 45 Gy each, over 18 months developed necrosis of the full thickness bone in the region of the implants. Three years after first presenting she died. Patient D3, edentulous for ten years before treatment by 67 Gy for a carcinoma of the left linguo-tonsillar fold, developed an enlarging area of necrosis of the left mandible and marked trismus. Several small sequestra have separated and he has recently developed anaesthesia of his left lip which strongly suggests a sofar undisplaced fracture. He is kept reasonably comfortable with continuous low-dose tetracycline. No obvious trauma had been given to the mandible in this case, but the trismus has raised the possibility of a recurrence, making it impossible to see the treated area without an anaesthetic. Because of the difficulty in opening his jaws it is possible that pressure by the blades of the Ferguson's gag on his edentulous mandible may have been the precipitating trauma. Patient D4 presented with a mass in the right submandibular region and paraesthesia of 1st, 2nd and 3rd divisions of the Vth cranial nerve. Examination under anaesthetic showed the cause to be an adenocarcinoma of the nasopharynx. She was given 53 Gy by external beam but developed a recurrence a year later, proved by biopsy. Three years after this biopsy, a pathological fracture of the mandible occurred through an area of necrosis.
DISCUSSION

The occurrence of necrosis of the mandible to any extent can cause morbidity and suffering second only to the development of a recurrence, metastasis or second tumour in the area. Many clinicians have reported upon their experience and expressed their concern about the subject, with particular reference to the part that dental extractions and other surgical interference plays in its causation. Bedwinck (1976), discussing 381 patients with

squamous carcinoma of the oral cavity, naso- and oropharynx, reported 54 cases of ORN, 35 of which were directly related to dental extractions. Of the first 203 patients, 12 needed resection, but after a more conservative approach about extractions had been adopted, only 5 of 178 patients needed resection. He reported spontaneous ORN as absent when the dose was below 60 Gy, uncommon under 70 Gy, but rising to 9% if the dose was over 70 Gy. Carl et al (1972) reported two cases of ORN in 49 patients and advocated meticulous dental care. Cheng and Wang (1974) reported 13 cases of ORN needing mandibulectomy in 76 cases of tonsillar carcinoma. They were all heavy smokers and drinkers and received 60 Gy or more. Coffin (1964) first had doubts about the need for routine pre-therapy extractions but gave no figures. Cook (1966) suggested that it is probably better to extract teeth considered doubtful prior to therapy, and accept the risk of doing this as less than the risk of posttherapy extractions. Daley et al (1972) found, out of 304 patients, 66 with ORN, of which 46 needed resection, and these would appear to be comparable with the "major" cases discussed in this Royal Marsden Hospital series. Del Regato (1939), in a very early discussion of irradiation caries and bone necrosis, advised pretherapy extractions; the larger the irradiation planned, the more drastic should the extractions be. Not many clinicians would agree with this view today. Horiot (1981) had 2% ORN in a series of 528 patients. He advocated very careful oral hygiene and very careful extraction techniques, allowing 10 days after extraction before therapy started and using a nasogastric tube for feeding during this time. McComb (1962) reported 16 resections in 204 patients with carcinoma of the tongue, floor of mouth, and lower gum. He concluded that it is justifiable to give a large enough dose to cure the disease, accepting the risk of necrosis. He felt that all teeth that would require extraction within one year should be removed prior to therapy. Morrish et al (1981) reported eight patients needing resection for ORN out of 100 patients treated. The dosage used in this series was higher than usual, extending to 95 Gy in some cases. Murray (1980) reported an incidence of 19% necrosis of the mandible in 404 patients treated by external irradiation and by local implants. This was reduced to 7.8% by adopting a more conservative approach to pretherapy extractions. Paterson (1963) advised that most teeth be retained, and adopted a very conservative attitude towards extractions. In the context of this paper the matter hinges upon two main factors. Firstly, the dose of radiation needed to cure the disease. This is a concern of the therapist, based on his judgement and experience. A palliative dose is unlikely to be large enough to risk ORN, whilst individual
855

VOL.

56, No. 671

Frank Coffin patient tolerance is a consideration that can modify intended doses to a marked degree. Secondly, the period which can be allowed to elapse between extractions and the need to commence curative radiotherapy depends to a great extent upon the speed with which a tooth socket heals. This subject has been examined by many authors, mostly in animals which can be sacrificed in order to enable the jaws to be sectioned for histology. However Amler (1969; Amler et al, 1960) suggested that minute quantities of osteoid are apparent at 7 days in humans, increasing slowly to fill two-thirds of the socket at 40 days, and that replacement of granulation tissue by connective tissue takes 20 days. Epithelialisation was not complete in some cases until the 35th day. Radiologically, 100 days were needed to arrive at a degree of radiopacity identical to the surrounding alveolus. Simpson (1958, 1969) gave much the samefiguresbut commented that surgical flaps did not have such a beneficial effect as is usually expected, and questioned whether the use of the burr deposited debris into the socket to delay healing. Very few radiotherapists would be willing to wait anything like two to three months before commencing treatment even if their patients, knowing the diagnosis, allowed them to. Moreover many more patients today than ever before are concerned to retain their teeth at all costs, and are prepared to refuse treatment in some cases if extractions are pressed. They are therefore willing to accept the rigorous self-care and regular professional examinations and treatment if found necessary. A very clear and precise explanation of the effects of radiation can be given in the planning period or first week of treatment. By the second week this forecast of the changes in salivary flow, mucositis and soreness will have begun to come true, and this will be particularly effective in convincing the patient that the advice he is receiving is genuine, and make the acceptance of the obsessional dental care much easier. Radiotherapists often forget this aspect of care in their concern for the technicalities of their own modality, but the dental consultant is particularly well equipped to provide it. A good schedule of consultations is as follows: the first during the planning period, a second about the middle of treatment when the prophesied morbidity has begun to be felt, and the instructions will have begun to be appreciated. At the end of treatment the patient can be seen again before dismissal, when his radiation reaction is at its height, and he will be most receptive to the instructions about his future care. This is a good time to give the patient's general dental practitioner helpful advice and information, particularly in regard to the need to avoid dental extractions, or other surgical interference to the bone. A slightly more relaxed attitude can be adopted with the maxilla than with the mandible, but it is felt that even here surgery should be done at the treatment centre at which irradiation was given. If this is not possible the operating surgeon should have full knowledge and technical information from the centre where treatment was given.
CONCLUSIONS

This series has shown that with the individual radiotherapeutic dosages and techniques used by the therapists at this hospital occurrence of gross radionecrosis of the jaws can be reduced to a minimum. This can be done by a strict avoidance of any surgical interference with the bone of the jaws, certainly after a full course of therapy, and probably, for about three months before therapy begins. This implies that biopsy of the usual squamous carcinoma of the region should also avoid interfering with the bone, provided sufficient tissue is obtained to make a diagnosis. True tumours of bone per se are a completely different problem in that their management follows a different course. Once a radical course of radiotherapy has been agreed, commenced orfinished,any surgery to the bone should be designed to remove the irradiated bone completely. Alveolectomies or similar partial removals and dental extractions are almost certain to lead to ORN. The obsessional dental care necessitated by these precepts has been dealt with at length.
ACKNOWLEDGMENTS

I have to thank all the radiotherapists, past and present, at the Royal Marsden Hospital for referring their patients, and for accepting my handling of the dental aspects of head and neck tumours, and in particular, Dr. M. Lederman and Dr. V. M. Dalley for their continuing confidence in a method which tended to be against the general thought in therapy circles. Miss M. Vollman, my chief assistant for some twenty years, has dealt with all the routine care not undertaken by the patient's own general dental practitioner, as have many other assistants, for shorter periods, who are now consultants themselves. Dr. Eileen Busby has given me tremendous help with the script, which has been tirelessly typed and retyped by Miss Eve Lister.
NOTE

Since the end of 1981 two recent patients have had courses of irradiation of 66 Gy for lesions of the left floor of mouth and tongue. This was followed by excision of a residuum including a partial thickness only of the mandible in order to maintain its continuity, and repair, one by a forehead flap, the other by a deltopectoral flap. Both cases have developed a massive necrosis of the partially resected bone, but in only one so far has it fractured, necessitating sequestrectomy. The other maintains continuity but a sequestrum extends from the right lower canine into the left ascending ramus. A third patient has developed osteoradionecrosis from 87 to ft at least, with a pathological fracture at the site of a partial resection and extraction of three teeth,

856

NOVEMBER 1983

Osteoradionecrosis of the jaws

followed by a full course of external radiotherapy to a residuum found on histology. These three cases exemplify the argument against improper surgical interference.
REFERENCES AMLER, M. H., 1969. Time sequence of tissue regeneration in human extraction wounds. Oral Surgery, Oral Medicine, Oral Pathology, 27, 307-318.
AMLER, M. H., JOHNSON, P. L. & SALMON, I., 1960.

DALEY, T. E., DRANE, J. B. & MCCOMB,

W.

S.,

1972.

Histological and histochemical investigation of human alveolar socket healing in undisturbed extraction wounds. Journal of the American Dental Association, 61, 32-AA. BEDWINCK, M. B., 1976. Osteoradionecrosis in patients treated with definitive radiotherapy for squamous ca. of oral cavity, naso and oropharynx. Radiology, 119, 665-667.
CARL, W., SCHAAF, N. G. & CHEN, TAK YEE, 1972. Oral care

Management of problems of teeth and jaw in patients undergoing irradiation. American Journal of Surgery, 124, 539-542. DEL REGATO, J. A., 1939. Dental lesions observed after roentgen therapy in cancer of the buccal cavity, pharynx and larynx. American Journal of Roentgenology, 42, 404-410. HORIOT, J. C , 1981. Systematic dental management in H & N irradiation. International Journal of Radiation Oncology Biology & Physics, 7, 1025-1029. MCCOMB, W. S., 1962. Necrosis in treatment of intra-oral cancer in radiation therapy. American Journal of Roentgenology, 87, 431-440.
MORRISH, B., CHAN, E., SILVERMAN, S., MEYER, J., Fu, K. &

of irradiated patients. Cancer, 30, 448-453.

oesophageal 1359-1360.

ulceration.

British

Medical

Journal, 282,

GREENSPAN, D., 1981. Osteoradionecrosis in patients treated for head and neck carcinoma. Cancer, 47, 1980-1983. MURRAY, C. G., 1980. Radiation necroses of the mandible. International Journal of Radiation Oncology Biology & Physics, 6, 543-548, 549-553. PATERSON, R., 1963. The Treatment of Malignant Disease by Radiotherapy. 2nd edit. (Williams & Wilkins, Baltimore), pp. 226-227.
PRICE, L. A., HILL, B., CALVERT, A. H., SHAW, H. J. &

CHENG, M. D. & WANG, M., 1974. Osteoradionecrosis of

mandible resulting from external megavoltage radiation therapy. Radiology, 112, 685-689. COFFIN, F., 1964. Cancer and the dental surgeon. British Dental Journal, 116, 191-202. COOK, H. P., 1966. Tooth extraction and radiation of the mouth. British Dental Journal, 120, 372-376.

HUGHES, K. B., 1975. A kinetically based multiple drug treatment for advanced head and neck cancers. British Medical Journal, iii, 10-12. SIMPSON, H. E., 1958. Healing of extraction wounds. British Dental Journal, 104, 99-100. 1969. Healing of extraction wounds, ibid, 126, 550-557.

857