- Email: [email protected]
PROSTAGLANDINS AND BREAST CANCER
624
(patients aged 36-87) and 16 benign neoplasms (patients aged 16-63 years). Macroscopically normal breast tissue was obtained from 23 of the mastectomy specimens. Samples were cut into pieces weighing approximately firmed 66 carcinomas
PROSTAGLANDINS AND BREAST CANCER A. BENNETT A. M. MCDONALD I. F. STAMFORD
E. M. CHARLIER J. S. SIMPSON T. ZEBRO
Departments of Surgery and Morbid Anatomy, King’s College Hospital Medical School, London SE5 8RX
Summary
Malignant human breast tumours produced more prostaglandin-like material during homogenisation than did benign tumours or normal breast tissue. Greatest synthesis in vitro occurred with tumours associated with bone metastases, and the highest "basal" amounts tended to occur in tumours showing spread histologically. It is suggested that prostaglandins may play a part in tumour spread and growth in bone and that drugs which inhibit prostaglandin synthesis may be valuable therapeutic agents in cancer.
Introduction THE evidence linking prostaglandins (P.G.s.) to the formation of bone metastases is substantial.1.2 We now present full data confirming our earlier reports,1.3 and we tentatively suggest that prostaglandins might be involved in the local spread of tumour.
Patients and Methods
Specimens from 82 women who underwent surgery for a breast lump were taken for pathological examination 5-60 min after excision, and parts of the specimen were removed for prostaglandin studies. Subsequent histological examination con-
Simon, M., Boure, M., Fauchet, R., Genetet, B. Gut, 1976, 17, 332. Walters, J. M., Watt, D. W., Stevens, F. M., McCarthy, C. E. Br. med. J. 1975, iv, 520. 3. Bomford, A., Eddleston, A. L. W. F., Kennedy, L. A., Batchelor, J. R., Williams, R. Lancet, 1977, i, 327. 4. Finch, S. C., Finch, C. A. Medicine, Baltimore, 1955, 34, 381. 5. Grace, N. D., Powell, L. W. Gastroenterology, 1974, 67, 1257. 6. Williams, R. , Smith, P. M., Spicer, E. F. J., Barry, M., Sherlock, S. Q. Jl
1. 2.
Med.
1969, 38, 1.
Powell, L. Postgrad. med. J. 1970, 46, 200. Crosby, W. H. New Engl. J. Med. 1976, 294, 333. Jacobs, A., Miller, F., Worwood, M., Beamish, M. R., Wardrop, C. A. Br. med. J. 1972, iv, 206. 10. Lipschitz, D. A., Cook, J. D., Finch, C. A. New Engl. J. Med. 1974, 290, 7. 8. 9.
1213. 11. Walters, G. O., Miller, F. M., Worwood, M. J. clin. Path. 1973, 26, 770. 12. Letsky, E. A., Miller, F., Worwood, M., Flynn, D. M. ibid. 1974, 27, 652. 13. Wands, J., Rowe, J. A., Mezey, S. E., Waterbury, L. A., Wright, J. R., Halliday, J. W., Isselbacher, K. J., Powell, L. W. New Engl. J. Med. 1976, 294, 302. 14. Feller, E. R., Pont, A., Wands, J. R., Carter, E. A., Foster, G., Kourides, I. A., Isselbacher, K. J. ibid. 1977, 296, 1422. 15. Powell, L. W., Halliday, J. W. ibid. 1976, 294, 1185. 16. Powell, L. W. in The Liver: Normal and Abnormal Functions (edited by F. Becker), part A, p. 129. New York, 1976. 17. Peters, T., Giovanello, T. J., Apt, L., Ross, J. F. J. Lab. clin. Med. 1956,
48, 274. 18. International Committee for Standardisation of Hæmatology,
1971, 20, 451. 19. Halliday, J. W., Gera,
K.
L., Powell, L. W.
Br. J.
Clinica chim. Acta.
Hœmat.
1975, 58,
207.
20. Scheuer, P. J., Williams, R., Muir, A. R. J. Path. Bact. 1962, 21. Powell, L. W. Australas. Ann. Med. 1966, 15, 110. 22. Powell, L. W., Kerr, J. F. R. in Pathobiology Annual Ioachim); p. 317. New York, 1975. 23. Bomford, A., Williams, R. Q. Jl Med. 1976, 45, 611. 24. Crosby, W. H., Likhite, V. V., O’Brien, J. E., Forman, D.
84, 53.
(edited by
J. Am. 1974, 227, 310. 25. Smith, P. M., Studley, F., Williams, R. Lancet, 1976, i, 133. 26. Prieto, J., Barry, M., Sherlock, S. Gastrœnterology, 1975, 68, 525.
5-15mg and washed in Krebs’ solution. Weighed amounts homogenised at room temperature (30 s in a Silverson homogeniser) either in Krebs’ solution/ethanol (1/1) acidified to approximately pH3 with formic acid, to determine "basal" amounts of prostaglandin-like material or in Krebs’ solution gassed with 5% CO2 in O2, which allows synthesis to occur during homogenisation.4 Prostaglandin-like material was then extracted into chloroforms and bioassayed on the rat gastric fundus preparation in the presence of drugs to increase assay sensitivity and selectivity.46 Amounts of material extracted from homogenates in Krebs’ solution have been called "total" since they include "basal" and newly synthesised material. "Synthesis" was calculated by subtracting basal from total values. Results are expressed as ng p.G.E2 equivalents/g fresh tissue (medians and semiquartile ranges). Some extracts from homogenates in Krebs’ solution contained enough biological activity to allow tentative characterisation by alkali hydrolysis (which inactivates P.G.E ’but not P.G.F compounds;’ 32 cases), and All or xi chromatography8,9 on paper impregnated with silica gel9 (29 and 6 cases respectwere
H. L.
med. Ass.
ively). Sections (5 pun) of all tumours were stained with haematoxylin and eosin and examined by a pathologist unaware of the prostaglandin data. Tumour size, histological type, grade of malignancy,’O border of growth, lymph-node involvement, sinus histiocytosis of lymph-nodes," production and type of fibrous tissue, invasion of blood and lymphatic vessels, cellular infiltration around and in the tumour, and predominant celltype in the infiltrate were recorded in the 66 cases of carcinoma. Lesion size, type, amount of fibrous tissue, inflammatory cell infiltration, and duct epithelial proliferation and dilatation were recorded in the 16 benign tumours. Immediately before or after surgery all patients with malignant tumours underwent skeletal scintigraphy following intravenous injection of 5-10 mCi 99mTc-labelled ethane hydroxydiphosphonate.12 Results were compared statistically with the Mann-Whitney U test or the Wilcoxon matched-pairs test; all probability values refer to two-tailed tests.
Results
Total, basal, and synthesised
amounts of prostaglan-
din-like activity were significantly higher in extracts of malignant tumours than in benign tumours or apparently normal breast tissue (table I). All chromatography (homogenates in Krebs’ solution; 29 tumours) showed material running with p.G.E2, P.G.F,,, P.G.E1, p.G.F,,, or P.G.E3/F3,(22, 19, 14, 7, and 6 respectively) or at other Rf values (see figure). In 2 cases the extracted material ran as* a single spot (one case with p.G.E2, another between P.G.E2 and P.G.F,2x). However, most tumours produced 2-5 substances with highest amounts of rat-fundus-contracting activity occurring at the Rf values of P.G.E2 67%; between P.G.E2 and p.G.F2, 51 %; and p.G.E1 25% (see figure). Tumours from patients with scintiscan evidence of bone metastases produced more prostaglandins than did those from scan-negative patients (table n). Chromatography (xi and All, n=35) indicated more p.G.F in tumours associated with bone metastases (P=0-005), but contrary to our preliminary results, alkali hydrolysis failed to show this (P=0.119). Total and basal amounts of prostaglandin-like material tended to be higher in tumours showing invasiveness histologically as judged by the appearance of malignant cells in the blood-vessels, lymphatics, and/or lymph-
625 TABLE I-PRODUCTION OF PROSTAGLANDIN-LIKE MATERIAL
*The values represent ng p.G.E2 equivalents/g tissue, expressed P=significance of differences; b=basal; t=total;s=synthesised.
as
(P.G.) FROM MALIGNANT AND BENIGN TUMOURS AND NORMAI
medians with
(table itt). There was no correlation between total, basal, or synthesised amounts of prostaglandin-like material from malignant tumours with tumour size, histological type, border of growth, grade of malignancy, sinus histiocytosis in lymph-nodes, type -and production nodes
of fibrous tissue, cellular infiltration around and in the tumour, or predominant cell type. Nor did prostaglandin measurements from benign tumours correlate with his-
tology. Discussion
_
Malignant breast tumours produced more prostaglandin-like material during homogenisation than did benign tumours or normal breast tissue, and those producing most in vitro were associated with the highest frequency of bone metastases. These factors might be causally related, but it is not known to what extent homogenisation studies reflect tumour prostaglandin synthesis in vivo. The total prostaglandin-like material extracted from homogenates in Krebs’ solution consists
semiquartile
TISSUE.**
ranges in brackets; number of experiments in italics.
of substances formed from released endogenous precursors, plus basal amounts already present, less any inactivated. Homogenates in acid-ethanol (which inhibits enzyme activity) probably contain basal amounts, together with prostaglandins which may be formed during tumour excision and subsequent processing. It therefore seems reasonable to assess prostaglandin formation by subtracting basal values from the total activity, and to suggest that this reflects synthetic potential in vivo. The many prostaglandin-like substances present in the extracts cannot be firmly identified by thin-layer chromatography. For example, in our system the p.G.E2 metabolites 15-keto p.G.E2 and its 13, 14-dihydro derivative, run with p.G.Fh.3The rat-stomach bioassay is almost equally sensitive to p.G.E1 and P.G.E2, but it is less sensitive to other prostaglandins, which are thereTABLE 11—SYNTHESIS OF PROSTAGLANDIN-LIKE MATERIAL FROM PATIENTS WITH POSITIVE AND NEGATIVE BONE
SCANS*
*Units as in table I. The groups were more clearly separated when activity running with P.G.F,,, F2x, and F3x was multiplied by 50, 10, and 100 respectively (total corrected chromatography) in an attempt to compensate for differences in bioassay sensitivity. In 10 cases, there was only sufficient tumour to measure total P.G., and these are not included. Numbers in italics are numbers of experiments. TABLE III-PROSTAGLANDIN-LIKE MATERIAL IN INVASIVE AND
NON-INVASIVE
/
/
9
22
19
1
14
3
(no of tumours)
Prostaglandin-like activity
extracted from
homogenates
TUMOURS*
of 29
malignant breast tumours. The results are presented as chromatograms (medians and ranges) obtained by All chromatography and assayed against prostaglandin E2. Some activity ran with the standard prostaglandins, E1, E2, E3, Fix, F2x,Y, F3x, as indicated below some columns; other activity ran at different rates (unlabelled columns). The numbers below show how many tumours produced the prostaglandin-like material indicated.
*Units as in table 1. "Invasive" tumours showed histological evidence of tumour spread into blood-vessels, lymphatics, and/or lymph-nodes. All patients with "invasive" tumours had positive bone scans, but so did 2 with "non-invasive" tumours. If these two tumours are reclassified as "invasive", the significance of the differences increases to P=0-03, 0-019, and 0-16 for total, basal, and synthesised amounts respectively. These results are calculated on the tumours in table tt, but one result is omitted because the histological examination was on a needle biopsy, and spread could not be assessed. Numbers in italics are numbers of experiments.
626
fore underestimated when assayed in terms of p.G.E2. The total amounts of particular prostaglandins may be important, but they cannot be calculated without knowing their identity and potency. We attempted to gain a better indication of the actual amount of prostaglandin present by somewhat arbitrarily multiplying the activity running at the Rf values Of P.G.F,,, P.G.F2x, and P.G.F3. by 50, 10 and 100 respectively. This more clearly distinguishes primary tumours associated with and without bone metastases. The ability of malignant breast tumours to produce such large amounts of prostaglandin-like material during homogenisation, particularly when associated with bone metastases, reinforces our previous suggestion that the two findings may be causally related. Studies in animals also support this view. In rabbits VX2 tumours synthesise p.G.E2-like material and release it into the venous blood;13,14 aspirin-like drugs inhibit both prostaglandin synthesis" and the development of bone metastases in rats’6 and rabbits. 14 p.G.E and P.G.F compounds released from human primary breast tumours seem somewhat unlikely to affect bone directly, because almost 70% of an infusion of p.G.E1 is inactivated during a single passage through the lungs.17 Perhaps prostaglandin released from malignant cells present in bone causes resorption and helps establish metastases. Alternatively, prostaglandins might increase bone metastatic spread by influencing other factors, such as primary tumour growth and dispersal. Thus, tumour basal (and possibly therefore total) amounts of prostaglandin tend to correlate with spread of the primary tumour, as judged histologically by the appearance of malignant cells in the blood-vessels, lymphatics, and lymph-nodes. In some animal studies aspirin or like-acting drugs reduced either tumour growth’8,19 or metastatic spread.2O Other work already citedl4,16 showed that the drugs inhibited bone destruction but did not reduce softtissue metastases. The reasons for these tumour and species differences are not clear, and the situation in man has not been determined. Nevertheless, the association of prostaglandins with human breast cancer indicates that aspirin-like drugs might be valuable therapeutically, either by stopping the early bone destruction or
possibly by altering tumour growth and spread generally. Since only initial bone destruction seems to be mediated by osteoclasts,21 it may be that patients with early disease, including those with low tumour-prostaglandin levels, might benefit most from such drugs.
CORRELATION BETWEEN LEUCOCYTE-MIGRATION INHIBITION BY BREAST-CANCER ANTIGENS, MAMMOGRAPHIC FINDINGS, AND OTHER BREAST-CANCER RISK FACTORS LORNE J. BRANDES G. HARDY
M. K. KIERNAN N. A. NELSON
GERALD J. GOLDENBERG
Departments of Medicine and Radiology, University of Manitoba, and Department of Epidemiology and Statistics and Manitoba Institute of Cell Biology, Manitoba Cancer Treatment and Research Foundation, Winnipeg, Manitoba, Canada
study of 62 normal carried out to determine if any correlation between the mean score of six
Summary
A double-blind women was
there
was
weekly leucocyte-migration-inhibition tests (L.M.T.) against breast-cancer antigen, xeromammographic findings, and the presence of other epidemiological risk factors for the development of breast cancer. 36 women had abnormal mammograms, and 11 of these 26 (30·5%) had a mean L.M.T. score ≽10%. 26 women had normal mammograms, and only 1 of these (3·8%) had a mean value ≽10%. This difference was significant (P<0·02). A family history of breast cancer and/or previous surgery for cystic and proliferative disease also correlated significantly with L.M.T. score (P<0·05); the combined presence of epidemiological risk factors and an abnormal mammogram correlated most strongly with L.M.T. score (P<0·01). Although the results may indicate an "immunological risk factor" towards the development of breast cancer, the biological relevance of this observation remains to be proven. Introduction THE
leucocyte-migration-inhibition
test
(L.M.T.)
is
widely used as an in-vitro test of cellular immunity both well-defined antigens1,2 and to crude extracts of malignant tissue including breast cancer. 3-6 Investigations in our laboratory demonstrated significant weekto-week variability of results in all subjects tested with the L.M.T. against both well-defined antigens such as P.P.D.7 and crude hypotonic-saline antigen extracts of breast-cancer tissue (B.C.A.).8 We therefore thought that mean scores of sequential tests (weekly for 6 weeks) to
We thank the Medical Research Council and Cancer Research
Campaign for support, the Upjohn Company for prostaglandins, and many colleagues for providing specimens and helping with tests in patients.
Requests of Surgery, 8RX.
reprints should be addressed to A.B. at the Department King’s College Hospital Medical School, London SE5
for
REFERENCES
Bennett, A., McDonald, A. M., Simpson, J. S., Stamford, I. F. Lancet, 1975, i, 1218. 2. Lancet, 1976, ii, 1063. 3. Bennett, A., Charlier, E. M., McDonald, A. M., Simpson, J. S., Stamford, I. F Prostaglandins, 1976, 11, 461. 4. Bennett, A., Stamford, I. F., Unger, W. G. J. Physiol. Lond. 1973, 229, 349. 5. Unger, W. G., Stamford, I. F., Bennett, A. Nature, 1971, 233, 336. 6. Gilmore, N., Vane, J. R., Wyllie, J. H. ibid. 1968, 218, 1135. References continued at foot of next column 1.
7. Bennett, A., Murray, J. G., Wyllie, J. H. Br. J. Pharmac. 1968, 32, 339. 8. Green, K., Samuelsson, B. J. Lipid Res. 1964, 5, 117. 9. Stamford, I. F., Unger, W. G. J. Physiol, Lond. 1972, 225, 4P. 10. Bloom, H. J. G., Richardson, W. W. Br. J. Cancer, 1957, 11, 359. 11. Black, M. M., Speer, F. D. Surgery Gynec. Obstet. 1958, 106, 163. 12. Citrin, D. L., Bessent, R. G., Tuohy, J. B., Elms, S. T., McGinlay, E, Greig, W. R., Blumgart, L. H. Br. J. Surg. 1974, 61, 330. 13. Voelkel, E. F., Tashjian, A. H., Franklin, R., Wasserman, E., Levine, L
Metabolism, 1975, 24, 973. Galasko, C. S. B., Bennett, A. Nature, 1976, 263, 508. Vane, J. R. Nature new Biol. 1971, 231, 232. Powles, T. J., Clark, S. A., Easty, D. M., Easty, G. C., Neville, A. M. Br. J. Cancer, 1973, 28, 316. 17. Golub, M., Zia, P., Matsuno, M., Horton, R. J. clin. Invest. 1975, 56, 1404. 18. Plescia, O. J., Smith, A. H., Grinwich, K. Proc. natn. Acad. Sci. U.S.A 1975, 72, 1848. 19. Strausser, H. R., Humes, J. L. Int. J. Cancer, 1975, 15, 724. 20. Gasic, G. J., Gasic, T. B., Galanti, N., Johnson, T., Murphy, S. ibid 1973. 11, 704. 21. Galasko, C. S. B. Nature, 1976, 263, 507. 14. 15. 16.
(patients aged 36-87) and 16 benign neoplasms (patients aged 16-63 years). Macroscopically normal breast tissue was obtained from 23 of the mastectomy specimens. Samples were cut into pieces weighing approximately firmed 66 carcinomas
PROSTAGLANDINS AND BREAST CANCER A. BENNETT A. M. MCDONALD I. F. STAMFORD
E. M. CHARLIER J. S. SIMPSON T. ZEBRO
Departments of Surgery and Morbid Anatomy, King’s College Hospital Medical School, London SE5 8RX
Summary
Malignant human breast tumours produced more prostaglandin-like material during homogenisation than did benign tumours or normal breast tissue. Greatest synthesis in vitro occurred with tumours associated with bone metastases, and the highest "basal" amounts tended to occur in tumours showing spread histologically. It is suggested that prostaglandins may play a part in tumour spread and growth in bone and that drugs which inhibit prostaglandin synthesis may be valuable therapeutic agents in cancer.
Introduction THE evidence linking prostaglandins (P.G.s.) to the formation of bone metastases is substantial.1.2 We now present full data confirming our earlier reports,1.3 and we tentatively suggest that prostaglandins might be involved in the local spread of tumour.
Patients and Methods
Specimens from 82 women who underwent surgery for a breast lump were taken for pathological examination 5-60 min after excision, and parts of the specimen were removed for prostaglandin studies. Subsequent histological examination con-
Simon, M., Boure, M., Fauchet, R., Genetet, B. Gut, 1976, 17, 332. Walters, J. M., Watt, D. W., Stevens, F. M., McCarthy, C. E. Br. med. J. 1975, iv, 520. 3. Bomford, A., Eddleston, A. L. W. F., Kennedy, L. A., Batchelor, J. R., Williams, R. Lancet, 1977, i, 327. 4. Finch, S. C., Finch, C. A. Medicine, Baltimore, 1955, 34, 381. 5. Grace, N. D., Powell, L. W. Gastroenterology, 1974, 67, 1257. 6. Williams, R. , Smith, P. M., Spicer, E. F. J., Barry, M., Sherlock, S. Q. Jl
1. 2.
Med.
1969, 38, 1.
Powell, L. Postgrad. med. J. 1970, 46, 200. Crosby, W. H. New Engl. J. Med. 1976, 294, 333. Jacobs, A., Miller, F., Worwood, M., Beamish, M. R., Wardrop, C. A. Br. med. J. 1972, iv, 206. 10. Lipschitz, D. A., Cook, J. D., Finch, C. A. New Engl. J. Med. 1974, 290, 7. 8. 9.
1213. 11. Walters, G. O., Miller, F. M., Worwood, M. J. clin. Path. 1973, 26, 770. 12. Letsky, E. A., Miller, F., Worwood, M., Flynn, D. M. ibid. 1974, 27, 652. 13. Wands, J., Rowe, J. A., Mezey, S. E., Waterbury, L. A., Wright, J. R., Halliday, J. W., Isselbacher, K. J., Powell, L. W. New Engl. J. Med. 1976, 294, 302. 14. Feller, E. R., Pont, A., Wands, J. R., Carter, E. A., Foster, G., Kourides, I. A., Isselbacher, K. J. ibid. 1977, 296, 1422. 15. Powell, L. W., Halliday, J. W. ibid. 1976, 294, 1185. 16. Powell, L. W. in The Liver: Normal and Abnormal Functions (edited by F. Becker), part A, p. 129. New York, 1976. 17. Peters, T., Giovanello, T. J., Apt, L., Ross, J. F. J. Lab. clin. Med. 1956,
48, 274. 18. International Committee for Standardisation of Hæmatology,
1971, 20, 451. 19. Halliday, J. W., Gera,
K.
L., Powell, L. W.
Br. J.
Clinica chim. Acta.
Hœmat.
1975, 58,
207.
20. Scheuer, P. J., Williams, R., Muir, A. R. J. Path. Bact. 1962, 21. Powell, L. W. Australas. Ann. Med. 1966, 15, 110. 22. Powell, L. W., Kerr, J. F. R. in Pathobiology Annual Ioachim); p. 317. New York, 1975. 23. Bomford, A., Williams, R. Q. Jl Med. 1976, 45, 611. 24. Crosby, W. H., Likhite, V. V., O’Brien, J. E., Forman, D.
84, 53.
(edited by
J. Am. 1974, 227, 310. 25. Smith, P. M., Studley, F., Williams, R. Lancet, 1976, i, 133. 26. Prieto, J., Barry, M., Sherlock, S. Gastrœnterology, 1975, 68, 525.
5-15mg and washed in Krebs’ solution. Weighed amounts homogenised at room temperature (30 s in a Silverson homogeniser) either in Krebs’ solution/ethanol (1/1) acidified to approximately pH3 with formic acid, to determine "basal" amounts of prostaglandin-like material or in Krebs’ solution gassed with 5% CO2 in O2, which allows synthesis to occur during homogenisation.4 Prostaglandin-like material was then extracted into chloroforms and bioassayed on the rat gastric fundus preparation in the presence of drugs to increase assay sensitivity and selectivity.46 Amounts of material extracted from homogenates in Krebs’ solution have been called "total" since they include "basal" and newly synthesised material. "Synthesis" was calculated by subtracting basal from total values. Results are expressed as ng p.G.E2 equivalents/g fresh tissue (medians and semiquartile ranges). Some extracts from homogenates in Krebs’ solution contained enough biological activity to allow tentative characterisation by alkali hydrolysis (which inactivates P.G.E ’but not P.G.F compounds;’ 32 cases), and All or xi chromatography8,9 on paper impregnated with silica gel9 (29 and 6 cases respectwere
H. L.
med. Ass.
ively). Sections (5 pun) of all tumours were stained with haematoxylin and eosin and examined by a pathologist unaware of the prostaglandin data. Tumour size, histological type, grade of malignancy,’O border of growth, lymph-node involvement, sinus histiocytosis of lymph-nodes," production and type of fibrous tissue, invasion of blood and lymphatic vessels, cellular infiltration around and in the tumour, and predominant celltype in the infiltrate were recorded in the 66 cases of carcinoma. Lesion size, type, amount of fibrous tissue, inflammatory cell infiltration, and duct epithelial proliferation and dilatation were recorded in the 16 benign tumours. Immediately before or after surgery all patients with malignant tumours underwent skeletal scintigraphy following intravenous injection of 5-10 mCi 99mTc-labelled ethane hydroxydiphosphonate.12 Results were compared statistically with the Mann-Whitney U test or the Wilcoxon matched-pairs test; all probability values refer to two-tailed tests.
Results
Total, basal, and synthesised
amounts of prostaglan-
din-like activity were significantly higher in extracts of malignant tumours than in benign tumours or apparently normal breast tissue (table I). All chromatography (homogenates in Krebs’ solution; 29 tumours) showed material running with p.G.E2, P.G.F,,, P.G.E1, p.G.F,,, or P.G.E3/F3,(22, 19, 14, 7, and 6 respectively) or at other Rf values (see figure). In 2 cases the extracted material ran as* a single spot (one case with p.G.E2, another between P.G.E2 and P.G.F,2x). However, most tumours produced 2-5 substances with highest amounts of rat-fundus-contracting activity occurring at the Rf values of P.G.E2 67%; between P.G.E2 and p.G.F2, 51 %; and p.G.E1 25% (see figure). Tumours from patients with scintiscan evidence of bone metastases produced more prostaglandins than did those from scan-negative patients (table n). Chromatography (xi and All, n=35) indicated more p.G.F in tumours associated with bone metastases (P=0-005), but contrary to our preliminary results, alkali hydrolysis failed to show this (P=0.119). Total and basal amounts of prostaglandin-like material tended to be higher in tumours showing invasiveness histologically as judged by the appearance of malignant cells in the blood-vessels, lymphatics, and/or lymph-
625 TABLE I-PRODUCTION OF PROSTAGLANDIN-LIKE MATERIAL
*The values represent ng p.G.E2 equivalents/g tissue, expressed P=significance of differences; b=basal; t=total;s=synthesised.
as
(P.G.) FROM MALIGNANT AND BENIGN TUMOURS AND NORMAI
medians with
(table itt). There was no correlation between total, basal, or synthesised amounts of prostaglandin-like material from malignant tumours with tumour size, histological type, border of growth, grade of malignancy, sinus histiocytosis in lymph-nodes, type -and production nodes
of fibrous tissue, cellular infiltration around and in the tumour, or predominant cell type. Nor did prostaglandin measurements from benign tumours correlate with his-
tology. Discussion
_
Malignant breast tumours produced more prostaglandin-like material during homogenisation than did benign tumours or normal breast tissue, and those producing most in vitro were associated with the highest frequency of bone metastases. These factors might be causally related, but it is not known to what extent homogenisation studies reflect tumour prostaglandin synthesis in vivo. The total prostaglandin-like material extracted from homogenates in Krebs’ solution consists
semiquartile
TISSUE.**
ranges in brackets; number of experiments in italics.
of substances formed from released endogenous precursors, plus basal amounts already present, less any inactivated. Homogenates in acid-ethanol (which inhibits enzyme activity) probably contain basal amounts, together with prostaglandins which may be formed during tumour excision and subsequent processing. It therefore seems reasonable to assess prostaglandin formation by subtracting basal values from the total activity, and to suggest that this reflects synthetic potential in vivo. The many prostaglandin-like substances present in the extracts cannot be firmly identified by thin-layer chromatography. For example, in our system the p.G.E2 metabolites 15-keto p.G.E2 and its 13, 14-dihydro derivative, run with p.G.Fh.3The rat-stomach bioassay is almost equally sensitive to p.G.E1 and P.G.E2, but it is less sensitive to other prostaglandins, which are thereTABLE 11—SYNTHESIS OF PROSTAGLANDIN-LIKE MATERIAL FROM PATIENTS WITH POSITIVE AND NEGATIVE BONE
SCANS*
*Units as in table I. The groups were more clearly separated when activity running with P.G.F,,, F2x, and F3x was multiplied by 50, 10, and 100 respectively (total corrected chromatography) in an attempt to compensate for differences in bioassay sensitivity. In 10 cases, there was only sufficient tumour to measure total P.G., and these are not included. Numbers in italics are numbers of experiments. TABLE III-PROSTAGLANDIN-LIKE MATERIAL IN INVASIVE AND
NON-INVASIVE
/
/
9
22
19
1
14
3
(no of tumours)
Prostaglandin-like activity
extracted from
homogenates
TUMOURS*
of 29
malignant breast tumours. The results are presented as chromatograms (medians and ranges) obtained by All chromatography and assayed against prostaglandin E2. Some activity ran with the standard prostaglandins, E1, E2, E3, Fix, F2x,Y, F3x, as indicated below some columns; other activity ran at different rates (unlabelled columns). The numbers below show how many tumours produced the prostaglandin-like material indicated.
*Units as in table 1. "Invasive" tumours showed histological evidence of tumour spread into blood-vessels, lymphatics, and/or lymph-nodes. All patients with "invasive" tumours had positive bone scans, but so did 2 with "non-invasive" tumours. If these two tumours are reclassified as "invasive", the significance of the differences increases to P=0-03, 0-019, and 0-16 for total, basal, and synthesised amounts respectively. These results are calculated on the tumours in table tt, but one result is omitted because the histological examination was on a needle biopsy, and spread could not be assessed. Numbers in italics are numbers of experiments.
626
fore underestimated when assayed in terms of p.G.E2. The total amounts of particular prostaglandins may be important, but they cannot be calculated without knowing their identity and potency. We attempted to gain a better indication of the actual amount of prostaglandin present by somewhat arbitrarily multiplying the activity running at the Rf values Of P.G.F,,, P.G.F2x, and P.G.F3. by 50, 10 and 100 respectively. This more clearly distinguishes primary tumours associated with and without bone metastases. The ability of malignant breast tumours to produce such large amounts of prostaglandin-like material during homogenisation, particularly when associated with bone metastases, reinforces our previous suggestion that the two findings may be causally related. Studies in animals also support this view. In rabbits VX2 tumours synthesise p.G.E2-like material and release it into the venous blood;13,14 aspirin-like drugs inhibit both prostaglandin synthesis" and the development of bone metastases in rats’6 and rabbits. 14 p.G.E and P.G.F compounds released from human primary breast tumours seem somewhat unlikely to affect bone directly, because almost 70% of an infusion of p.G.E1 is inactivated during a single passage through the lungs.17 Perhaps prostaglandin released from malignant cells present in bone causes resorption and helps establish metastases. Alternatively, prostaglandins might increase bone metastatic spread by influencing other factors, such as primary tumour growth and dispersal. Thus, tumour basal (and possibly therefore total) amounts of prostaglandin tend to correlate with spread of the primary tumour, as judged histologically by the appearance of malignant cells in the blood-vessels, lymphatics, and lymph-nodes. In some animal studies aspirin or like-acting drugs reduced either tumour growth’8,19 or metastatic spread.2O Other work already citedl4,16 showed that the drugs inhibited bone destruction but did not reduce softtissue metastases. The reasons for these tumour and species differences are not clear, and the situation in man has not been determined. Nevertheless, the association of prostaglandins with human breast cancer indicates that aspirin-like drugs might be valuable therapeutically, either by stopping the early bone destruction or
possibly by altering tumour growth and spread generally. Since only initial bone destruction seems to be mediated by osteoclasts,21 it may be that patients with early disease, including those with low tumour-prostaglandin levels, might benefit most from such drugs.
CORRELATION BETWEEN LEUCOCYTE-MIGRATION INHIBITION BY BREAST-CANCER ANTIGENS, MAMMOGRAPHIC FINDINGS, AND OTHER BREAST-CANCER RISK FACTORS LORNE J. BRANDES G. HARDY
M. K. KIERNAN N. A. NELSON
GERALD J. GOLDENBERG
Departments of Medicine and Radiology, University of Manitoba, and Department of Epidemiology and Statistics and Manitoba Institute of Cell Biology, Manitoba Cancer Treatment and Research Foundation, Winnipeg, Manitoba, Canada
study of 62 normal carried out to determine if any correlation between the mean score of six
Summary
A double-blind women was
there
was
weekly leucocyte-migration-inhibition tests (L.M.T.) against breast-cancer antigen, xeromammographic findings, and the presence of other epidemiological risk factors for the development of breast cancer. 36 women had abnormal mammograms, and 11 of these 26 (30·5%) had a mean L.M.T. score ≽10%. 26 women had normal mammograms, and only 1 of these (3·8%) had a mean value ≽10%. This difference was significant (P<0·02). A family history of breast cancer and/or previous surgery for cystic and proliferative disease also correlated significantly with L.M.T. score (P<0·05); the combined presence of epidemiological risk factors and an abnormal mammogram correlated most strongly with L.M.T. score (P<0·01). Although the results may indicate an "immunological risk factor" towards the development of breast cancer, the biological relevance of this observation remains to be proven. Introduction THE
leucocyte-migration-inhibition
test
(L.M.T.)
is
widely used as an in-vitro test of cellular immunity both well-defined antigens1,2 and to crude extracts of malignant tissue including breast cancer. 3-6 Investigations in our laboratory demonstrated significant weekto-week variability of results in all subjects tested with the L.M.T. against both well-defined antigens such as P.P.D.7 and crude hypotonic-saline antigen extracts of breast-cancer tissue (B.C.A.).8 We therefore thought that mean scores of sequential tests (weekly for 6 weeks) to
We thank the Medical Research Council and Cancer Research
Campaign for support, the Upjohn Company for prostaglandins, and many colleagues for providing specimens and helping with tests in patients.
Requests of Surgery, 8RX.
reprints should be addressed to A.B. at the Department King’s College Hospital Medical School, London SE5
for
REFERENCES
Bennett, A., McDonald, A. M., Simpson, J. S., Stamford, I. F. Lancet, 1975, i, 1218. 2. Lancet, 1976, ii, 1063. 3. Bennett, A., Charlier, E. M., McDonald, A. M., Simpson, J. S., Stamford, I. F Prostaglandins, 1976, 11, 461. 4. Bennett, A., Stamford, I. F., Unger, W. G. J. Physiol. Lond. 1973, 229, 349. 5. Unger, W. G., Stamford, I. F., Bennett, A. Nature, 1971, 233, 336. 6. Gilmore, N., Vane, J. R., Wyllie, J. H. ibid. 1968, 218, 1135. References continued at foot of next column 1.
7. Bennett, A., Murray, J. G., Wyllie, J. H. Br. J. Pharmac. 1968, 32, 339. 8. Green, K., Samuelsson, B. J. Lipid Res. 1964, 5, 117. 9. Stamford, I. F., Unger, W. G. J. Physiol, Lond. 1972, 225, 4P. 10. Bloom, H. J. G., Richardson, W. W. Br. J. Cancer, 1957, 11, 359. 11. Black, M. M., Speer, F. D. Surgery Gynec. Obstet. 1958, 106, 163. 12. Citrin, D. L., Bessent, R. G., Tuohy, J. B., Elms, S. T., McGinlay, E, Greig, W. R., Blumgart, L. H. Br. J. Surg. 1974, 61, 330. 13. Voelkel, E. F., Tashjian, A. H., Franklin, R., Wasserman, E., Levine, L
Metabolism, 1975, 24, 973. Galasko, C. S. B., Bennett, A. Nature, 1976, 263, 508. Vane, J. R. Nature new Biol. 1971, 231, 232. Powles, T. J., Clark, S. A., Easty, D. M., Easty, G. C., Neville, A. M. Br. J. Cancer, 1973, 28, 316. 17. Golub, M., Zia, P., Matsuno, M., Horton, R. J. clin. Invest. 1975, 56, 1404. 18. Plescia, O. J., Smith, A. H., Grinwich, K. Proc. natn. Acad. Sci. U.S.A 1975, 72, 1848. 19. Strausser, H. R., Humes, J. L. Int. J. Cancer, 1975, 15, 724. 20. Gasic, G. J., Gasic, T. B., Galanti, N., Johnson, T., Murphy, S. ibid 1973. 11, 704. 21. Galasko, C. S. B. Nature, 1976, 263, 507. 14. 15. 16.