1975, British Journal of Radiology, 48, 652-655
The radiosensitivity of endometrial carcinoma By G. F. G. O. De Muelenaere, M.B., Ch.B., M.Med.(Rad.T.), M.D. Department of Radiotherapy, H. F. Verwoerd Hospital and University of Pretoria, Private Bag X169, Pretoria 0001, South Africa {Received July, 1974) ABSTRACT
An analysis of all cases (121) of endometrial carcinoma treated by pre-operative radium in the uterine cavity revealed that the menopausal status of a patient affects the radiosensitivity of her tumour. It was found that if the patient is more than ten years post-menopausal at time of diagnosis her carcinoma will be more resistant to radiation. The technique used by the authors is adequate in eradicating a considerable percentage (68 per cent) of endometrial carcinomas.
(3) the length of the uterine cavity; (4) the grade of differentiation; (5) the histological stage; (6) the patient material; and these are discussed below. TECHNIQUE
Our applicators have been developed over a period of one year. Our skill in applying them may have been lacking at first. The technique may have been deficient in that not enough applicators were inserted at first, which may have influenced the radiation dose. I have therefore divided the cases into an early and late group, which consist of 50 and 71 cases respectively, the early group having been treated in the first five years. Table I shows the marked difference in these two groups. Our technique probably did improve over the years, although there still are a significant number of patients with residual tumour in the latter group.
Operable patients with endometrial carcinoma are treated in this series by a single intracavitary radium application, followed six weeks later by total hysterectomy and bilateral salpingo-oophorectomy. The applicators and the technique had been fully described (Savage and Millin, 1963; de Muelenaere and Fichardt, 1973). They consist of four small applicators, containing 10 mg of radium each, which are inserted into the cornua. A long central stalk containing 60 mg of radium is inserted last, and should be in the endocervical canal. No radium at all is left in the vagina. The applicators are based on both the Manchester and Heyman applicators. When the uterine cavity is small not all the small applicators are inserted, but the cavity should always be full without stretching it, and the central stalk must always be in the endocervical canal. A dose of 8,000 rads is reached in 72 hours, 2 cm lateral to the radium in a single application. Between 1960 and the end of 1972, 121 patients with histologically proven endometrial cancer have been treated in this way. Thorough histological examination of the hysterectomy specimen confirmed that the tumour was totally destroyed in 82 (68 per cent). An analysis of all these cases has been done to find a possible common factor that could explain why some tumours had not been destroyed. It is of course possible that some tumours are inherently radioresistant, but an attempt was made to see if an external factor might be implicated. Several factors could have been responsible, namely: (1) difference in technique; (2) the presence of myomas;
MYOMAS
According to Table II the presence of myomas, as detected in the hysterectomy specimen, did not influence the ability of radium to destroy the tumour. TABLE I N o RESIDUAL CARCINOMA
Early group Late group Total
28/ 50 = 56-0% 54/ 71 =76-0°,, 82/121=67-8% 0-02>P>001
TABLE II PRESENCE OF MYOMAS
Tumour free group Residual tumour group
Read at XIII International Congress of Radiology, Madrid, October 1973. 652
18/82 = 22-0% 10/39 = 25-6%
0-70>P>0-50
AUGUST
1975
The radiosensitivity of endometrial carcinoma TABLE III
TABLE VII
LENGTH OF UTERINE CAVITY—TUMOUR FREE: 1
MYOMETRIAL INFILTRATION
(Cases with residual turmour)
40/55 = 72-7% 18/28 = 64-3% 12/21=57-1% 12/17 = 70-6%
10cm Unknown
None Inner third Middle third Outer third Unknown
0-80>P>0-70
6/39 = 19/39 = 7/39 = 3/39 = 4/39 =
15-4% 48-7% 17-9% 7-7% 10-3%
TABLE IV
TABLE VIII
LENGTH OF UTERINE CAVITY—TUMOUR FREE: 2
AGE OF PATIENTS—TUMOUR FREE
10 cm 10 cm
58/83 = 70-0% 12/21 = 57-1%
40-49 years 50—59 years 60—69 years 70-79 years 80+ years
0-30>P>0-20
6/7 = 85-7% 30/38 = 78-9% 33/51=64-7% 11/21 = 52-4% 2/4 = 50-0%
TABLE V
0-20>P>010
DIFFERENTIATION OF TUMOUR—TUMOUR FREE
Well differentiated Moderately differentiated Poorly differentiated Adenoacanthoma Unknown
TABLE IX
44/67 = 65 .70/
MENOPAUSAL STATUS
' /o
8/14 = 57 .11 0/ ,'0 16/21=76 •9°/ .70/ ' /o 8/12 = 66 .70/ 6/7 = 85 ' /o
TABLE VI CERVICAL INFILTRATION—TUMOUR FREE
Stage I Stage II
15/18 = 83 .70/
Pre-menopausal < 5 years postmenopausal 5—9 years post-menopausal 10—14 years post-menopausal 15-19 years post-menopausal > 20 years postmenopausal
0-95>P>0-90
78/112 = 69-6% 6/9 = 66-7%
TUMOUR FREE : 1
J /o
17/20 = 85 •0% 11/13 = 84 •6% 14/24 = 58 .10/ J /o
8/17 = 47 •1 0/ 1
/o 0
17/29 = 58 •ft '
° /O
0-05>P>0-02 agreement with that given in the literature (Lewis, Stallworthy and Cowdell, 1970).
0-90>P>0-80 LENGTH OF UTERINE CAVITY
It seems as if the length of the uterine cavity did influence the results, as seen in Table III. Unfortunately, for a large number the length was not noted in the records. Because the difference is not statistically significant, I compared only two groups, i.e. all the cases where the cavity was less than, and more than 10 cm in length. Although the chances are still one in four that this could be a chance finding, the longer uteri do worse (Table IV). GRADE OF DIFFERENTIATION
Table V indicates that the differentiation of the tumour has relatively little effect on its radiosensitivity. The small difference is not statistically significant. The incidence of undifferentiated tumours is in
STAGE
Table VI shows that cervical infiltration (stage II tumour) is no hindrance to the destruction of endometrial carcinoma. When analysing those cases with residual tumour, it is clear that even carcinomas that did not penetrate the myometrium can be radioresistant (Table VII). The number of cases with residual carcinoma is similar to that given in the literature for a whole series where no pre-operative destruction of tumour had occurred (Lewis et al., 1970). It may thus be submitted that a significant number of deeply penetrating tumours are destroyed, and that our dose to the serosal surface of the uterus is adequate. PATIENT MATERIAL
No difference could be found in the patients whose uteri were clear of cancer, and those where
653
Vol.. 48, No. 572 G. F. G. O. De Muelenaere TABLE X MENOPAUSAL STATUS
TABLE XIII
TUMOUR FREE : 2
< 10 years postmenopausal > 10 years postmenopausal
HIGH OESTROGENIC INDEX—TUMOUR FREE
10 years post-menopausal
43/51=83-2% 39/70 = 55-7%
6/6 =100% 8/16 = 50% 0-10>P>0-05
P 10 years postmenopausal 11/25 = 44-0% 28/45 = 62-2% and that the menopausal factor was much more important than the uterine length. I must presume that P002 than and more than ten years menopausal, is due to an endocrinal factor. The literature does show that TABLE XII progesterone can influence the radiosensitivity of UTERINE CAVITY OVER 10 CM—TUMOUR FREE endometrial carcinoma (Bonte, Decoster and Ide, 1970) but progesterone cannot still be excreted up to 10 years post-menopausal 2/9 = 22-2% who had a high oestrogenic index in the vaginal smear, once again shows the importance of the 002>P>001 menopausal status, but not of the presence of oestrogen as seen in the vaginal smear (Table XIII). residual carcinoma was found, as far as the inciIt has previously been suggested that the worse dence of diabetes mellitus, hypertension and prognosis in older patients who have exclusively cardiovascular disease, and obesity are concerned. been treated radiotherapeutically, may be due to a The age of the patient was, however, an important lack of steroids (Nilson and Roller, 1969). Boronow factor, as can be seen in Table VIII. The tumours (1966) has also shown that the menopause is a prognostic factor in patients with endometrial carcinoma. in older patients seem to be more radioresistant. There was no difference in incidence of differenI propose that this difference in radiosensitivity tiation or of length of uterine cavity in the different may be due to the sudden decrease of the levels of age cohorts so that these factors are not involved. luteinising hormone or follicle stimulating hormone, The menopause was correlated to the destruction which may possibly be excreted for a considerable of tumour, and revealed the very significant figures period after the menopause. We are at present inshown in Table IX. The patients have been grouped vestigating these factors. according to the time after the menopause in those In 1966 Mussey and Malkasian suggested that who are post-menopausal. Table X shows that ten progesterone might sensitize endometrial carcinoyears after the menopause is the critical period. This ma to radiation. This seems to have been confirmed is statistically highly significant. by the work of Bonte et al. (1970). In the report To ascertain whether this difference remained, the by Boyd, Pollard and Blaikley (1973) no difference patients were again divided in an early and late in the incidence of residual tumour at hysterectomy group (Table XI). This reveals that the results in after pre-operative radium insertion was found both groups improved, probably due to technical between patients treated with progesterone at the skill. It can also be seen that the early group has a time of the radium application and controls. more favourable number of patients less than ten We intend to prime all our patients who are years post-menopausal, and should have done better. more than ten years post-menopausal with proThis confirms that our technique was important gesterone for a time before the radium is inserted. MENOPAUSAL STATUS
TUMOUR FREE." 3
654
AUGUST 1975
The radiosensitivity of endometrial carcinoma This will elucidate whether or not progesterone is a radiosensitizer in endometrial carcinoma. Some work has been done on the concomitant use of progesterone in the treatment of primary endometrial carcinoma, but the studies seem to be aimed at improving the prognosis, more than at determining the radiosensitization (Lewis et al., 1967; Boyd et al., 1973). Bonte et al. (1970) did not correlate the incidence of residual tumour with the menopausal status of their patients. CONCLUSION
1. The most important factor influencing the radiosensitivity of endometrial carcinoma is the menopausal status of the patients, and therefore probably the hormonal milieu in the patient. 2. Our technique has improved over the years, and is effective in the presence of myomas and cervical or myometrial infiltration. 3. The depth of the uterine cavity is important only in those patients who are more than ten years post-menopausal. 4. Histological differentiation of endometrial carcinoma does not influence its radiosensitivity.
ment of Gynaecology and Obstetrics, and Dr. Kenny, Superintendent, H. F. Verwoerd Hospital, for access to their records. REFERENCES BONTE, J., DECOSTER, J. M., and IDE, P., 1970. Radiosensi-
tization of endometrial adenocarcinoma by means of medroxyprogesterone. Cancer, 25, 907-910. BOYD, I. E., POLLARD, W., and BLAIKLEY, J. B., 1973.
Pre-operative intramuscular progestogen in the treatment of endometrial carcinoma. Journal of Obstetrics and Gynaecology of the British Commonwealth, 80, 360-363. BORONOW, R. C , 1966. Carcinoma of the corpus: treatment at M.D. Anderson Hospital. In Cancer of the Uterus and Ovary, 1969 (Year Book Medical Publishers, Inc , Chicago). DE MUELENAERE, G. F. G. O., and FICHARDT, T., 1973.
Carcinoma of the corpus uteri: the Cinderella of cancer therapy. South African Medical Journal, 47, 245-255. LEWIS, G. C , NADLER, S. H., BROSS, I. D. J., and HACK,
N. H., 1967. Adjuvant chemotherapy for cancer of the corpus uteri: preliminary report. Obstetrics and Gynecology, 29, 797-802. LEWIS, B. V., STALL WORTHY, J. A., and COWDELL, R., 1970.
Adenocarcinoma of the body of the uterus. Journal of Obstetrics and Gynaecology of the British Commonwealth, 77, 343-348. MUSSEY, E., and MALKASIAN, G. D., 1966. Progestogen
treatment of recurrent carcinoma of the endometrium. American Journal of Obstetrics and Gvnecology, 94, 78— 85. NILSEN, P. A., and ROLLER, O., 1969. Carcinoma of the
endometrium in Norway 1957-1960 with special reference to treatment results. American Journal of Obstetrics and Gynecology, 105, 1099-1109. SAVAGE, D. J., and MILLIN, J. C. B., 1963. Technical
ACKNOWLEDGMENTS
aspects of the Pretoria approach to the use of radium in the treatment of cancer of the uterus. South African Medical Journal, 37, 703-707.
I express my gratitude to Professor T. Fichardt, former director of the Department of Radiotherapy, for his advice. My thanks are due to Professor Evans of the Depart-
Book review An Atlas of Normal Roentgen Variants that may Simulate Disease. By Theodore E. Keats, pp. 351, 1973. Year Book Medical Publishers, Chicago, Distributed by Lloyd-Luke (Medical Books) Ltd. £19-25. This is a first-class atlas. The title puts the common problem. Radiologists in training need help with three times as many queries on normal variants as on abnormal lesions, suggests the foreword. The radiological borderland between the normal and pathological is difficult, fascinating, and of course important. Overdiagnosis, followed by further investigation or even treatment, clearly does no good at all. The radiologist can render his patient a considerable service by being able to say with informed confidence: "this may look odd, but it is a normal variant, and no harm will come of it". By way of a major textbook, we already have Kohler's classic on skeletal variants. Trying to encompass the whole field of human variants in a 350-page atlas might appear foolish or impossible. However, Dr. Keats has produced a highly successful book by restricting himself just to those
variants that simulate disease. Thus, there is no need to catalogue every possible congenital oddity, and this makes for a very practical and manageable volume. This is an atlas in the strict sense, with no narrative paragraphs. There are brief captions to each picture, very much to the point, and often giving a useful key reference. The illustrations are splendid. It is natural that two-thirds of the book should be devoted to bone, with the skull particularly prominent. These are the areas that we all sweat over most: is there a fracture or not? Pre- and post-F.R.C.R. radiologists will find this a useful companion when looking at casualty films. The atlas is well organized by anatomical territory, so that it is easy to arrive at the correct pages. The last part of the work covers soft tissues, including the gastrointestinal and urinary tracts. Wisely there is no attempt to deal with specialized fields like neuroradiology or angiocardiography. Should one even quibble about irritating inconsistency in right/left projection of radiographs? This is a unique, wanted and beautiful book. T. SHERWOOD.
655
The radiosensitivity of endometrial carcinoma By G. F. G. O. De Muelenaere, M.B., Ch.B., M.Med.(Rad.T.), M.D. Department of Radiotherapy, H. F. Verwoerd Hospital and University of Pretoria, Private Bag X169, Pretoria 0001, South Africa {Received July, 1974) ABSTRACT
An analysis of all cases (121) of endometrial carcinoma treated by pre-operative radium in the uterine cavity revealed that the menopausal status of a patient affects the radiosensitivity of her tumour. It was found that if the patient is more than ten years post-menopausal at time of diagnosis her carcinoma will be more resistant to radiation. The technique used by the authors is adequate in eradicating a considerable percentage (68 per cent) of endometrial carcinomas.
(3) the length of the uterine cavity; (4) the grade of differentiation; (5) the histological stage; (6) the patient material; and these are discussed below. TECHNIQUE
Our applicators have been developed over a period of one year. Our skill in applying them may have been lacking at first. The technique may have been deficient in that not enough applicators were inserted at first, which may have influenced the radiation dose. I have therefore divided the cases into an early and late group, which consist of 50 and 71 cases respectively, the early group having been treated in the first five years. Table I shows the marked difference in these two groups. Our technique probably did improve over the years, although there still are a significant number of patients with residual tumour in the latter group.
Operable patients with endometrial carcinoma are treated in this series by a single intracavitary radium application, followed six weeks later by total hysterectomy and bilateral salpingo-oophorectomy. The applicators and the technique had been fully described (Savage and Millin, 1963; de Muelenaere and Fichardt, 1973). They consist of four small applicators, containing 10 mg of radium each, which are inserted into the cornua. A long central stalk containing 60 mg of radium is inserted last, and should be in the endocervical canal. No radium at all is left in the vagina. The applicators are based on both the Manchester and Heyman applicators. When the uterine cavity is small not all the small applicators are inserted, but the cavity should always be full without stretching it, and the central stalk must always be in the endocervical canal. A dose of 8,000 rads is reached in 72 hours, 2 cm lateral to the radium in a single application. Between 1960 and the end of 1972, 121 patients with histologically proven endometrial cancer have been treated in this way. Thorough histological examination of the hysterectomy specimen confirmed that the tumour was totally destroyed in 82 (68 per cent). An analysis of all these cases has been done to find a possible common factor that could explain why some tumours had not been destroyed. It is of course possible that some tumours are inherently radioresistant, but an attempt was made to see if an external factor might be implicated. Several factors could have been responsible, namely: (1) difference in technique; (2) the presence of myomas;
MYOMAS
According to Table II the presence of myomas, as detected in the hysterectomy specimen, did not influence the ability of radium to destroy the tumour. TABLE I N o RESIDUAL CARCINOMA
Early group Late group Total
28/ 50 = 56-0% 54/ 71 =76-0°,, 82/121=67-8% 0-02>P>001
TABLE II PRESENCE OF MYOMAS
Tumour free group Residual tumour group
Read at XIII International Congress of Radiology, Madrid, October 1973. 652
18/82 = 22-0% 10/39 = 25-6%
0-70>P>0-50
AUGUST
1975
The radiosensitivity of endometrial carcinoma TABLE III
TABLE VII
LENGTH OF UTERINE CAVITY—TUMOUR FREE: 1
MYOMETRIAL INFILTRATION
(Cases with residual turmour)
40/55 = 72-7% 18/28 = 64-3% 12/21=57-1% 12/17 = 70-6%
10cm Unknown
None Inner third Middle third Outer third Unknown
0-80>P>0-70
6/39 = 19/39 = 7/39 = 3/39 = 4/39 =
15-4% 48-7% 17-9% 7-7% 10-3%
TABLE IV
TABLE VIII
LENGTH OF UTERINE CAVITY—TUMOUR FREE: 2
AGE OF PATIENTS—TUMOUR FREE
10 cm 10 cm
58/83 = 70-0% 12/21 = 57-1%
40-49 years 50—59 years 60—69 years 70-79 years 80+ years
0-30>P>0-20
6/7 = 85-7% 30/38 = 78-9% 33/51=64-7% 11/21 = 52-4% 2/4 = 50-0%
TABLE V
0-20>P>010
DIFFERENTIATION OF TUMOUR—TUMOUR FREE
Well differentiated Moderately differentiated Poorly differentiated Adenoacanthoma Unknown
TABLE IX
44/67 = 65 .70/
MENOPAUSAL STATUS
' /o
8/14 = 57 .11 0/ ,'0 16/21=76 •9°/ .70/ ' /o 8/12 = 66 .70/ 6/7 = 85 ' /o
TABLE VI CERVICAL INFILTRATION—TUMOUR FREE
Stage I Stage II
15/18 = 83 .70/
Pre-menopausal < 5 years postmenopausal 5—9 years post-menopausal 10—14 years post-menopausal 15-19 years post-menopausal > 20 years postmenopausal
0-95>P>0-90
78/112 = 69-6% 6/9 = 66-7%
TUMOUR FREE : 1
J /o
17/20 = 85 •0% 11/13 = 84 •6% 14/24 = 58 .10/ J /o
8/17 = 47 •1 0/ 1
/o 0
17/29 = 58 •ft '
° /O
0-05>P>0-02 agreement with that given in the literature (Lewis, Stallworthy and Cowdell, 1970).
0-90>P>0-80 LENGTH OF UTERINE CAVITY
It seems as if the length of the uterine cavity did influence the results, as seen in Table III. Unfortunately, for a large number the length was not noted in the records. Because the difference is not statistically significant, I compared only two groups, i.e. all the cases where the cavity was less than, and more than 10 cm in length. Although the chances are still one in four that this could be a chance finding, the longer uteri do worse (Table IV). GRADE OF DIFFERENTIATION
Table V indicates that the differentiation of the tumour has relatively little effect on its radiosensitivity. The small difference is not statistically significant. The incidence of undifferentiated tumours is in
STAGE
Table VI shows that cervical infiltration (stage II tumour) is no hindrance to the destruction of endometrial carcinoma. When analysing those cases with residual tumour, it is clear that even carcinomas that did not penetrate the myometrium can be radioresistant (Table VII). The number of cases with residual carcinoma is similar to that given in the literature for a whole series where no pre-operative destruction of tumour had occurred (Lewis et al., 1970). It may thus be submitted that a significant number of deeply penetrating tumours are destroyed, and that our dose to the serosal surface of the uterus is adequate. PATIENT MATERIAL
No difference could be found in the patients whose uteri were clear of cancer, and those where
653
Vol.. 48, No. 572 G. F. G. O. De Muelenaere TABLE X MENOPAUSAL STATUS
TABLE XIII
TUMOUR FREE : 2
< 10 years postmenopausal > 10 years postmenopausal
HIGH OESTROGENIC INDEX—TUMOUR FREE
10 years post-menopausal
43/51=83-2% 39/70 = 55-7%
6/6 =100% 8/16 = 50% 0-10>P>0-05
P 10 years postmenopausal 11/25 = 44-0% 28/45 = 62-2% and that the menopausal factor was much more important than the uterine length. I must presume that P002 than and more than ten years menopausal, is due to an endocrinal factor. The literature does show that TABLE XII progesterone can influence the radiosensitivity of UTERINE CAVITY OVER 10 CM—TUMOUR FREE endometrial carcinoma (Bonte, Decoster and Ide, 1970) but progesterone cannot still be excreted up to 10 years post-menopausal 2/9 = 22-2% who had a high oestrogenic index in the vaginal smear, once again shows the importance of the 002>P>001 menopausal status, but not of the presence of oestrogen as seen in the vaginal smear (Table XIII). residual carcinoma was found, as far as the inciIt has previously been suggested that the worse dence of diabetes mellitus, hypertension and prognosis in older patients who have exclusively cardiovascular disease, and obesity are concerned. been treated radiotherapeutically, may be due to a The age of the patient was, however, an important lack of steroids (Nilson and Roller, 1969). Boronow factor, as can be seen in Table VIII. The tumours (1966) has also shown that the menopause is a prognostic factor in patients with endometrial carcinoma. in older patients seem to be more radioresistant. There was no difference in incidence of differenI propose that this difference in radiosensitivity tiation or of length of uterine cavity in the different may be due to the sudden decrease of the levels of age cohorts so that these factors are not involved. luteinising hormone or follicle stimulating hormone, The menopause was correlated to the destruction which may possibly be excreted for a considerable of tumour, and revealed the very significant figures period after the menopause. We are at present inshown in Table IX. The patients have been grouped vestigating these factors. according to the time after the menopause in those In 1966 Mussey and Malkasian suggested that who are post-menopausal. Table X shows that ten progesterone might sensitize endometrial carcinoyears after the menopause is the critical period. This ma to radiation. This seems to have been confirmed is statistically highly significant. by the work of Bonte et al. (1970). In the report To ascertain whether this difference remained, the by Boyd, Pollard and Blaikley (1973) no difference patients were again divided in an early and late in the incidence of residual tumour at hysterectomy group (Table XI). This reveals that the results in after pre-operative radium insertion was found both groups improved, probably due to technical between patients treated with progesterone at the skill. It can also be seen that the early group has a time of the radium application and controls. more favourable number of patients less than ten We intend to prime all our patients who are years post-menopausal, and should have done better. more than ten years post-menopausal with proThis confirms that our technique was important gesterone for a time before the radium is inserted. MENOPAUSAL STATUS
TUMOUR FREE." 3
654
AUGUST 1975
The radiosensitivity of endometrial carcinoma This will elucidate whether or not progesterone is a radiosensitizer in endometrial carcinoma. Some work has been done on the concomitant use of progesterone in the treatment of primary endometrial carcinoma, but the studies seem to be aimed at improving the prognosis, more than at determining the radiosensitization (Lewis et al., 1967; Boyd et al., 1973). Bonte et al. (1970) did not correlate the incidence of residual tumour with the menopausal status of their patients. CONCLUSION
1. The most important factor influencing the radiosensitivity of endometrial carcinoma is the menopausal status of the patients, and therefore probably the hormonal milieu in the patient. 2. Our technique has improved over the years, and is effective in the presence of myomas and cervical or myometrial infiltration. 3. The depth of the uterine cavity is important only in those patients who are more than ten years post-menopausal. 4. Histological differentiation of endometrial carcinoma does not influence its radiosensitivity.
ment of Gynaecology and Obstetrics, and Dr. Kenny, Superintendent, H. F. Verwoerd Hospital, for access to their records. REFERENCES BONTE, J., DECOSTER, J. M., and IDE, P., 1970. Radiosensi-
tization of endometrial adenocarcinoma by means of medroxyprogesterone. Cancer, 25, 907-910. BOYD, I. E., POLLARD, W., and BLAIKLEY, J. B., 1973.
Pre-operative intramuscular progestogen in the treatment of endometrial carcinoma. Journal of Obstetrics and Gynaecology of the British Commonwealth, 80, 360-363. BORONOW, R. C , 1966. Carcinoma of the corpus: treatment at M.D. Anderson Hospital. In Cancer of the Uterus and Ovary, 1969 (Year Book Medical Publishers, Inc , Chicago). DE MUELENAERE, G. F. G. O., and FICHARDT, T., 1973.
Carcinoma of the corpus uteri: the Cinderella of cancer therapy. South African Medical Journal, 47, 245-255. LEWIS, G. C , NADLER, S. H., BROSS, I. D. J., and HACK,
N. H., 1967. Adjuvant chemotherapy for cancer of the corpus uteri: preliminary report. Obstetrics and Gynecology, 29, 797-802. LEWIS, B. V., STALL WORTHY, J. A., and COWDELL, R., 1970.
Adenocarcinoma of the body of the uterus. Journal of Obstetrics and Gynaecology of the British Commonwealth, 77, 343-348. MUSSEY, E., and MALKASIAN, G. D., 1966. Progestogen
treatment of recurrent carcinoma of the endometrium. American Journal of Obstetrics and Gvnecology, 94, 78— 85. NILSEN, P. A., and ROLLER, O., 1969. Carcinoma of the
endometrium in Norway 1957-1960 with special reference to treatment results. American Journal of Obstetrics and Gynecology, 105, 1099-1109. SAVAGE, D. J., and MILLIN, J. C. B., 1963. Technical
ACKNOWLEDGMENTS
aspects of the Pretoria approach to the use of radium in the treatment of cancer of the uterus. South African Medical Journal, 37, 703-707.
I express my gratitude to Professor T. Fichardt, former director of the Department of Radiotherapy, for his advice. My thanks are due to Professor Evans of the Depart-
Book review An Atlas of Normal Roentgen Variants that may Simulate Disease. By Theodore E. Keats, pp. 351, 1973. Year Book Medical Publishers, Chicago, Distributed by Lloyd-Luke (Medical Books) Ltd. £19-25. This is a first-class atlas. The title puts the common problem. Radiologists in training need help with three times as many queries on normal variants as on abnormal lesions, suggests the foreword. The radiological borderland between the normal and pathological is difficult, fascinating, and of course important. Overdiagnosis, followed by further investigation or even treatment, clearly does no good at all. The radiologist can render his patient a considerable service by being able to say with informed confidence: "this may look odd, but it is a normal variant, and no harm will come of it". By way of a major textbook, we already have Kohler's classic on skeletal variants. Trying to encompass the whole field of human variants in a 350-page atlas might appear foolish or impossible. However, Dr. Keats has produced a highly successful book by restricting himself just to those
variants that simulate disease. Thus, there is no need to catalogue every possible congenital oddity, and this makes for a very practical and manageable volume. This is an atlas in the strict sense, with no narrative paragraphs. There are brief captions to each picture, very much to the point, and often giving a useful key reference. The illustrations are splendid. It is natural that two-thirds of the book should be devoted to bone, with the skull particularly prominent. These are the areas that we all sweat over most: is there a fracture or not? Pre- and post-F.R.C.R. radiologists will find this a useful companion when looking at casualty films. The atlas is well organized by anatomical territory, so that it is easy to arrive at the correct pages. The last part of the work covers soft tissues, including the gastrointestinal and urinary tracts. Wisely there is no attempt to deal with specialized fields like neuroradiology or angiocardiography. Should one even quibble about irritating inconsistency in right/left projection of radiographs? This is a unique, wanted and beautiful book. T. SHERWOOD.
655
