411

BREAST CANCER AND THE

S BLOOD-GROUP SYSTEM

P. MOROSINI

M. N. JONES

E. G. LEE

M. P. VESSEY

Radcliffe Infirmary, Oxford

O. P. HEINONEN

Division of Epidemiology and Preventive Medicine,National Pensions Institute, Helsinki, Finland

H. R. NEVANLINNA

Finnish Red Cross Blood Transfusion Service, Helsinki

U. L. M. SVINHUFVUD

Roentgen Diagnostic Department of Meilahti Hospital,University Central Hospital, Helsinki

Summary In contrast with a report from Boston,U.S.A., no association between the S

blood-group system and breast cancer has been foundin data collected in Oxford and Helsinki.

Introduction

A REPORT from Boston, U.S.A., has suggested thatthere may be an association between the S blood-groupsystem and breast cancer. Two independent series ofbreast-cancer patients and control subjects were

studied. In the first, 61 % of 82 breast-cancer patientswere of the ss genotype, in comparison with only 46%of 472 controls. In the second, the correspondingfigures were 56% of 89 breast-cancer patients andonly 41 % of 102 controls. There was also a suggestionthat the association was stronger among those in whombreast cancer was diagnosed before the age of 50 years.We thought that the association should be studied

in other countries, and our data were collected inOxford and Helsinki. Information was collected aboutthe MN and the S blood-group systems, because theseare known to be closely linked. a

Materials and Methods

Oxford.-Blood-samples were collected from 100 breast-cancer patients attending follow-up clinics at the RadcliffeInfirmary and Churchill Hospital and from 100 femalecontrol subjects. Only white women born in the UnitedKingdom were included. The control subjects fell intotwo broad categories: 19 normal women, mainly nurses,and 81 patients with a wide variety of diseases other thanbreast cancer. All blood-samples were tested with anti-M,anti-N, anti-S, and anti-s sera within 48 hours of collectionby M. N. J., who did not know which samples came frombreast-cancer patients and which from controls. The

collection of samples was arranged so that roughly equalnumbers of breast-cancer samples and control sampleswere tested in each batch.

Helsinki.-Blood-samples were taken from 199 patientswith breast cancer, seen in the outpatient department ofthe Radiotherapy Clinic, Helsinki University CentralHospital. This hospital treats patients from all areas ofFinland. Blood-samples were tested on the day of collectionwith anti-M, anti-N, anti-S, and anti-s sera in the centrallaboratory of the Finnish Red Cross blood transfusionservice. The laboratory staff conducting the tests wereunaware of the hypothesis under investigation. Data for5536 young men, representing a sample of 2-5 per 1000 ofthe rural population of Finland, were used for controlpurposes. Blood-samples from these young men, collectedduring 1964-67 as part of a different investigation, werealso tested in the central laboratory of the Finnish RedCross blood transfusion service, but with anti-M, anti-Nand anti-S sera only.

Results

Table i shows the S and MN genotypes of thebreast-cancer patients and controls from Oxford andHelsinki. There is no suggestion in the data from

TABLE II-DISTRIBUTION OF S GENOTYPES IN PATIENTS WITH

BREAST CANCER SUBDIVIDED INTO DIAGNOSIS BEFORE OR AFTER

THE MENOPAUSE

* In the absence of exact information about age at menopause, patientsin whom breast cancer was diagnosed before the age of 50 yearshave been taken as premenopausal cases.

Patients with Swedish names and those having one or both parentsborn in a Swedish-speaking area are not included.

either country that the genotype ss is over-representedamong patients with breast cancer; the point estimateof relative risk in the presence of ss is 0-92 both inOxford and in Helsinki, with 95 % confidence limits of0-72-1-19 in the combined series. Indeed, agreementbetween the data for breast-cancer patients and con-trols is extremely close in all the subclasses shown intable I. The Finnish data were further analysed afterexcluding all breast-cancer patients with Swedish

TABLE I-DISTRIBUTION OF MNS GENOTYPES IN PATIENTS WITH BREAST CANCER AND CONTROL SUBJECTS

412

names, together with those having either one or bothparents born in a Swedish-speaking area. This pro-duced no change in the findings.

Table 11 shows the S genotypes of the breast-cancerpatients from both countries, subdivided according towhether the disease was first diagnosed before or afterthe menopause. There is no indication of any signifi-cant association between the genotype distribution andthe age at which cancer was diagnosed.

Discussion

Our findings in Oxford and Helsinki do not agreewith those reported from Boston 1 ; we have not foundany association between the S blood-group system andbreast cancer. There are two possible explanations forthis result. The first, which we consider to be themore likely, is that the American findings represent anextreme chance event. The second is that the differen-ces between the findings are real and reflect theinfluence of different setiological factors in Boston

compared with Oxford and Helsinki.We thank Dr. G. Wiernik and Dr. L. R. M. Holsti for allowing

us to study patients with breast cancer under their care.

REFERENCES

1. Boston Collaborative Drug Surveillance Program. Lancet, 1971,i, 301.

2. Race, R. R., Sanger, R. Blood Groups in Man. London, 1968.

Preliminary Communication

EFFECTIVE RENIN ACTIVITY IN PLASMAOF CHILDREN WITH KWASHIORKOR

E. E. KRITZINGER E. KANENGONI

J. J. JONESGodfrey Huggins School of Medicine, Salisbury, Rhodesia

Summary Effective plasma-renin activity hasbeen compared in patients with

kwashiorkor and in healthy children. Plasma-renin

activity was higher in children with kwashiorkor, andthis increased activity may account for some of thephysiological changes associated with protein-caloriemalnutrition.

INTRODUCTION

CEDHMA is the principal sign of kwashiorkor andrepresents a general state of overhydration resultingfrom a failure to balance water consumption andexcretion. The role of the renin-angiotensin systemin this disease has been investigated.

PATIENTS AND METHODS

Two groups of patients were studied at Harare Hospital:group B was 100 children admitted during the periodJanuary to November, 1970, and group R was 26 childrenadmitted in March, April, and May, 1971. 20 children

attending the well-baby clinics, serving the same community,formed a healthy contrast group. Effective plasma-reninactivity (R.A.) was assayed biologically 3 after fuller’s earthextraction 4 for group B, and by radioimmune assay 5 forgroup R and the healthy contrast group. In both assaysthe plasma pH was adjusted to 5-5 with N hydrochloric acidbefore incubation, and in group B the renin substrate

MEDIANS AND CORRESPONDING 10TH AND 90TH PERCENTILES FOREFFECTIVE PLASMA-RENIN ACTIVITY AND HORMONE CONCENTRA-

TIONS

* Indicates a significant (p < 0-05) difference from the value on admissionin patients who survived (Wilcoxon test for ranked data).

concentration was determined by incubation with 5 Gold-blatt milliunits per ml. of human renin.

In group R and the contrast group serum insulin andgrowth hormone were estimated by radioimmune assay *’’and plasma-cortisol by competitive protein binding. 8

RESULTS

The results appear in the accompanying table.

Although the radioimmune assay gives values thatare threefold larger than the bioassay values,9 thesame essential differences were found by both methods.R.A. was significantly higher in patients with kwashior-kor (group R) than in children from the well-babyclinics, and in both groups (R and B) R.A. on admissionto hospital was significantly higher in patients whosubsequently died than in those who later recovered.Radioimmune values for the healthy children accordwell with an earlier report.1o

DISCUSSION

R.A. showed no correlation with plasma-albumin 11or with plasma or urine electrolytes or osmolality.Significant correlations were found between R.A. andserum-bilirubin (r==+0’55, P<0-01) and betweenR.A. and liver enlargement (r=+0-43, p<O’05),suggesting that increased R.A. is due to reduced

hepatic clearance associated with the characteristicliver damage. a

Increased R.A. could be responsible for aldosterone 13and vasopressin 14 secretion and for the renal changes-reduced blood-flow, inability to excrete a salt orwater load-that occur -in kwashiorkor. 2 Thirstresulting from renin release could also contribute tofluid retention in protein-calorie malnutrition, as hasbeen suggested for psychogenic malnutrition. 15These kwashiorkor patients showed characteristic

alterations in cortisol, growth hormone, and insulinthat have previously been described 16-18; none ofthese changes seemed to be related to the increasein R.A.

We thank Dr. E. P. Forbes, Dr. J. H. M. Axton, Prof. A. C.Kendall and Prof. 0. D. Fisher for allowing us to study theirpatients; the Secretary for Health (Rhodesia) for access to