509

including infections, 1 vaccination, 2 trauma,3 and

pregnancy.4-6 Corticotrophin therapy hastens theresolution of an acute relapse,1. 7.8 though it probablydoes not affect the long-term course,9 nor does it

prevent relapses and progression in chronic multiplesclerosis.10 The association of diffuse cerebral sclerosiswith Addison’s disease 11-13 may also be relevant.

Attempts to demonstrate hypothalamo-hypophysialand adrenocortical abnormalities in multiple sclerosishave, however, been disappointing. Reduced urinary17-ketosteroid levels have been recorded in chronic

cases, but, as Teasdale et al. 14 pointed out, this wasprobably the result of non-specific debilitation. Otherworkers found normal adrenocortical function.15,16Two recent investigations have given slightly con-

flicting results. Teasdale et al. 14 investigated 21ambulant patients, 11 of whom had just had an acuterelapse. Basal levels of urinary and plasma cortico-steroids and the response to corticotrophin were normal,but the rise of plasma-cortisol induced by insulinhypoglycaemia was significantly less than in healthycontrols. Millac et aI.17 have now extended this workto a further 30 ambulant patients, 16 after an acuterelapse, and 20 matched control patients with otherneurological diseases. The basal plasma-cortisol levelwas similar in the two groups, but it was significantlyhigher in multiple-sclerosis patients with a recent

relapse than in the remaining multiple-sclerosispatients. In contrast to the findings of Teasdale et al.,there was no significant difference in the plasma-cortisol response to insulin hypoglycxmia betweenthe multiple-sclerosis patients and neurological con-trols, and in fact the levels tended to be higher inmultiple sclerosis. Patients in acute relapse had asignificantly lower response than the others. Therewas a prolonged rise of plasma-cortisol in response tolysine-vasopressin, compared with the control

neurological patients.All this might make one conclude, rather like Mr.

Bumble, that " statistics is an ass"; but rather itillustrates the pitfalls inherent in the choice of any" control " group. Unfortunately, the relationshipbetween stress and the relapses of multiple sclerosisremains obscure.1. Miller, H., Newell, D. J., Ridley, A. Lancet, 1961, ii, 1120.2. Miller, H., Cendrowski, W., Shapira, K. Br. med. J. 1967, ii, 210.3. Miller, H. Lancet, 1964, i, 848.4. Shapira, K., Poskanzer, D. C., Newell, D. J., Miller, H. Brain,

1966, 89, 419.5. Millar, J. H. D., Allison, R. S., Cheeseman, E. A., Merrett, J. D.

ibid. 1959, 82, 417.6. Leibowitz, U., Antonosky, A., Katz, R., Alter, M. J. Neurol.

Neurosurg. Psychiat. 1967, 30, 354.7. Rawson, M. D., Liversedge, L. A., Goldfarb, G. Lancet, 1966, ii,

1044.8. Alexander, L., Berkeley, A. W., Alexander, A. M. Multiple Sclerosis:

Prognosis and Treatment. Springfield, Illinois, 1961.9. Rawson, M. D., Liversedge, L. A. Lancet, 1969, ii, 222.

10. Millar, H. J., Vas, C. J., Noronha, M. J., Liversedge, L. A.,Rawson, M. D. ibid. 1967, ii, 429.

11. Blaw, M. E., Osterberg, K., Kosak, P., Nelson, E. Archs Neurol.,Chicago, 1964, 11, 626.

12. Dubois, R., Loeb, H., Perier, O., Parmentier, R., Szliwowski, H.Helv. pœdiat. Acta, 1964, 19, 528.

13. Hoefnagel, D., Brun, A., Ingbar, S. H., Goldman, H. J. Neurol.Neurosurg. Psychiat. 1967, 30, 56.

14. Teasdale, G. H., Smith, P. A., Wilkinson, R., Latner, A. L.,Miller, H. Lancet, 1967, i, 64.

15. Garcia-Reyes, J. A., Jenkins, D., Forsham, P. H., Thorn, G. W.Archs Neurol. Psychiat., Chicago, 1952, 68, 776.

16. Alexander, L., Cass, L. J., Enders, M., Sarai, K. Confima neurol.,1966, 28, 1.

17. Millac, P., Cook, D. B., Chase, K. J. Neurol. Neurosurg. Psychiat.1969, 32, 414.

THE COLON AND THE SURGEON

THE thorn of colonic antisepsis has again beenstirred in the flesh of surgeons by the latest findings 1

concerning preoperative suppression of bacteria inthe gut. A combination of 1 g. of neomycin, 100,000units of bacitracin, and 500,000 units of nystatin,given six-hourly for forty-eight hours before operation,caused a " dramatic " fall in the numbers of variousbacteria in the faeces in 14 out of 19 patients. Thebacteria affected were bacteroides, coliforms, strepto-cocci, and clostridia. Although the postoperativesepsis-rate among these patients was 21%, Sellwoodet aLl see this as a significant reduction in comparisonwith 62% in a group of patients who were given 2 g.of phthalylsulphathiazole six-hourly for ninety-sixhours before operations on the colon. In an earlier

investigation at the same hospital, the sepsis-rate ina control group similarly treated with phthalylsulpha-thiazole was 35%. This vast difference in what mightbe expected to be a reasonably constant baseline

points to unmeasurable variables.

A five-day preoperative course of sulphonamidegiven to dogs has been clearly shown to reduce

granuloma and microabscess formation at the anasto-mosis,3 but this preparation has proved of little or noclinical value in the reduction of gross sepsis.1-3 Itsmain effect in fact seems to be an undesirableadherence of fxces to the bowel wall. Preoperativeantibiotics have been tried at great length on manythousands of patients.1-9 The drugs given have

ranged from the combinations used by Sellwood et al.to a single antibiotic-gentamicin, kanamycin, or

cephaloridine. General agreement is lacking, and theviews emerging from these investigations range frommild optimism (but nothing more) to the labelling ofsuch preparation as a positive cause of sepsis.5 Thebowel contents can clearly not be sterilised by anyantibiotic or mixture of antibiotics at present 3,5,6 noris it clear that this is a desirable feat 3-5,10-13 foroperations involving intraperitoneal anastomosis.A clear exception, proved by experiment 14 and

clinically,3,4,8 is the help given by antibiotics when ananastomosis lacks the protection of the peritoneum,such as an anterior resection. Here preoperative anti-bacterial drugs consistently reduce the rate ofbreakdown. This gain seems to be the only surebenefit from this tradition of preparation.Two dangers beset attempts to suppress bacteria in

the colon: one is mentioned in practically every paperon the subject, the other hardly at all. Firstly, staphy-lococci will not live in significant numbers in company

1. Sellwood, R. A., Burn, J. I., Waterworth, P. M., Welbourn, R. B.Br. J. Surg. 1969, 56, 610.

2. Burn, J. I., Sellwood, R. A., Okubadejo, O. A. Postgrad. med. J.1967, 43, suppl. 17.

3. Symposium on Pre- and Post-operative Care of the Colon,American Proctologic Society. Dis. Colon Rectum, 1968, 11, 163.

4. Herter, F. P., Slanetz, C. A. Am. J. Surg. 1967, 113, 165.5. Altemeier, W. A., Todd, J. C., Inge, W. W. Archs Surg., Chicago,

1966, 93, 566.6. Azar, H., Drapanas, T. Am. J. Surg. 1968, 115, 209.7. Smith, W. D., Bacon, H. E. Dis. Colon Rectum, 1967, 10, 322.8. Houghton, G. W. Med. J. Aust. 1968, i, 796.9. Cohn, I. Intestinal Antisepsis. Illinois, 1968.

10. Johnstone F. R. C. Surgery Gynec. Obstet. 1963, 116, 1.11. Vink, M. Br. J. Surg. 1952, 41, 431.12. Cohn, I., Atik, M. Ann. Surg. 1960, 151, 197.13. Cohn, I. Surgery Gynec. Obstet. 1967, 124, 501.14. Ryan, P., Stratford, B., Dixson, S. Aust. N.Z.J. Surg. 1969, 39, 103.

510

with Escherichia, but when E. coli is drasticallyreduced in numbers the staphylococcus thrives. If itis an enterotoxin-producing staphylococcus-andthe drug has not yet been found that will exterminateall species-enterocolitis will strike, in any form froma slight postoperative diarrhoea to a fatal illness.3-5,9The incidence of this complication is probablyrising,6,8 and it is a heavy price to pay for the uncertainbenefits of such " preparation ". The incidencevaries from area to area and at different times in thesame hospital: the figures reported (for clinicallyobvious infection) go as high as 30%.6 The over-

growing staphylococci can cause wound sepsis, andup to a 95% correlation has been observed betweenthe organisms grown from the postoperative stooland those in the wound.3,5,6 The risk of staphylococcalcolonisation is at its height when -antibiotics are givenboth before and after operation. Curiously, it is atthis point that one of the most experienced studentsof this subject, E. J. Poth,3 prefers a sulphonamidepreparation, because, though these drugs permit theovergrowth of Streptococcus facalis, which is non-

pathogenic, they do suppress the overgrowth of" opportunists ". The second, seldom-mentioned,hazard was named in 1952 by Vink,11 when he warnedthat attempted preoperative bowel sterilisation incancer surgery of the colon might be associated witha higher risk of tumour recurrence at the suture line.This risk has been confirmed. 11-13,15

Systemic antibiotics during and after the operationhave perhaps come out better than any other use ofdrugs in colon surgery.5,16 Even here, though, manysurgeons prefer to limit the period of intravenousadministration to three days, again because of the riskof staphylococcal overgrowth.3,5 5 One reply to therisk of overpopulation is to give intravenous methicillinpreoperatively for five days in conjunction withintestinal antisepsis. With this rigorous regimenstaphylococcal infection has been eliminated entirely 3 ;but resistant strains may be just round the corner.

Irrigation of the anastomosis and the peritonealcavity during operation seems useful in certain con-ditions. Poth 3 uses a 1/2°,o solution of neomycin andwithdraws it before closure. A 1-21/2% solution of

noxythiolin cuts the mortality from peritonitis due toperforated diverticula, 17 and its ability to protectanastomoses and wounds during routine surgery ofthe colon is being investigated. Of greater moment inthe results of cancer surgery is the report that thetumour recurrences at the anastomosis encouraged bypreoperative antibiotics can apparently be curbed

by irrigation with low-molecular-weight dextran.13

Many surgeons emphasise that thorough mechanicalcleansing is the foundation of successful colon surgery,and, without it, antibiotics appear to lose theireffect .3-5. 9 Even here, however, a sacred cow mightlurk: since Belsey,18 in replacing the resophagus withcolon (an operation where three anastomoses are at

risk), prefers to operate on a deliberately constipatedcolon, in which fseces can be moved away from thesegments used; his sepsis-rate is unusually low, and15. Floyd, C. E., Corley, R. G., Cohn, I. Am. J. Surg. 1965, 109, 153.16. Herrington, J. L., Lawler, M., Thomas, T. V. Ann. Surg. 1967,

165, 709.17. Browne, M. K., Clin. Trials J. 1967, 4, 673.18. Belsey, R. J. thorac. Surg. 1965, 49, 33.

lately he has also discarded preoperative antibioticsas unhelpful. Even further across the spectrum is thework of J. V. Prohaska, who has performed 300total or partial colectomies for Crohn’s disease andulcerative colitis: he used no preoperative laxatives orenemas (for obvious reasons, working in inflamed,infected, and damaged tissues); nor did he give pre-operative antibacterial drugs, but he did provide threedays of intravenous tetracycline after operation. Outof 300 such operations no sepsis arose in 100consecutive cases, and only 2 instances in the rest.The colon must have its own defence mechanism

against bacterial invasion, and a very efficient oneindeed. The contents of the small bowel are dried

here, segmented, and coated with a lubricant withpossibly antibacterial properties. Enemas, laxatives,and washouts must derange this careful process andbreach the normal barrier, and much work is neededto determine exactly where interference, either

physical or antibacterial, in the lumen or in thecirculation, is damaging rather than beneficial, and why.Wound sepsis alone after colon surgery is a much

more clearly defined problem. In Herter’s series 4 of1042 patients it was commoner than either peritonealinfection or anastomosis leak (except for anterior

resections). There is no doubt at all that the rate ofwound infection has been experimentally and clinicallyreduced to about a tenth when a broad-spectrumantibiotic solution is left in the wound at closure.19-22

Experience with thousands of patients over two

decades thus points away from preoperative inter-ference with the colon. An acceptable sepsis-rate seemsto fall somewhere between 10 and 20%, and this maymost surely be achieved primarily by refinement in sur-gical technique in the following directions: strictavoidance of tension on suture lines with protection ofblood-supply, and the use of iodised sutures 13; andperitoneal irrigation with antibiotic solutions and withfor cancer surgery, low-molecular-weight dextran.

EXTRACORPOREAL FERTILISATION

THE treatment of infertility by in-vitro fertilisationof the ovum and implantation of the fertilised egg inthe patient’s uterus is not so new or so alarming aprospect as recent utterances and criticisms imply.Nor does it seem to raise " difficult ethical problems ",which was the reported response 23 of the Secretaryof the British Medical Association. Surely this treat-ment would simply be an extension of earlier work 24in an area where other aids to conception have longbeen accepted but have not always been successful.Forebodings about " test-tube babies " and " geneticengineering " are unjustified.

PRINTING OF THE LANCET

Publication of The Lancet has again been delayedby an industrial dispute at the journal’s British

printers in Watford.19. Hopson, W. B., Britt, L., Sherman, R. T. J. surg. Res. 1968, 8, 261.20. Nash, A. G., Hugh, T. B. Br. med. J. 1967, i, 471.21. Ryan, E. A. Br. J. Surg. 1967, 54, 324.22. Singleton, A. O., Julian, J. Ann. Surg. 1960, 151, 192.23. Times, Feb. 25, 1970.24. See Lancet, 1969, i, 405.