to y22may1991 500 35 3 49 0070(0-013) l9june199150021 0 27 0-043(0-010) used iridium-192 sources...

Occupational and Environmental Medicine 1994;51:713-718

Death of a classified worker probably caused byoverexposure to y radiation

D C Lloyd, A A Edwards, E J Fitzsimons, C D Evans, R Railton, P Jeffrey, T G Williams,A D White, M Ikeya, H Sumitomo

National RadiologicalProtection BoardD C LloydA A EdwardsMonklands Hospital,AirdrieE J FitzsimonsC D EvansR RailtonField OperationsDivision Health andSafety Executive,GlasgowP JefferyT G WilliamsUniversity Hospital,CardiffA D WhiteUniversity ofOsakaM IkeyaH SumitomoCorrespondence to:D C Lloyd, NationalRadiological ProtectionBoard, Chilton, Didcot,Oxfordshire OX 11 ORQ.Accepted 15 July 1994

AbstractThis paper describes the case of an

industrial radiographer who was seri-ously overexposed to y radiation. Theexact circumstances of his exposure were

not established but it was concluded thathe was repeatedly irradiated probably toa total average whole body dose of at least10 Gy over several years. Also, a muchlarger dose to a hand required its partialamputation. He developed myelodyspla-sia, which progressed to acute myeloidleukaemia from which he died.Karyotypic examination of the leukaemicblasts showed changes very similar -tothose associated with secondaryleukaemia that may develop after radioor chemotherapy. The paper describeshis medical case- history, the investiga-tion of his workplace, and the attempts toestimate his radiation dose by chromoso-mal analysis of blood lymphocytes andelectron spin resonance of dental enameland bone.

(Occup Environ Med 1994;5l:713-718)

Keywords: industrial radiographer, y radiation, over-

exposure

Figure 1 Fingers of the right hand in November 1989 showing ulceration of the indexfinger, scattered small keratoses, sclerosis with loss of wrinkling, and nails with ridging andragged margins.

Serious occupationally related exposure toradiation of classified workers in the UnitedKingdom is rare. Good workplace standards,partly reflecting legislative encouragement,'2have helped to lessen risks although somespecific problem areas remain. This paperdescribes the diagnosis, treatment, and doseestimates performed on an industrial radio-grapher who had a hand burn and subse-quently developed a neoplasm from which hedied. Both conditions were most probablycaused by overexposure to ionising radiation.

Case historyThe patient first experienced redness andswelling of his right index finger in 1984. Hedid not consult a doctor and the symptomsresolved without treatment. In May 1988 heconsulted his general practitioner when thehand swelled again. The appearance was der-matitic with deformity of the nails for whichhe received griseofulvin but without benefit.The hand failed to heal and in October 1988he was seen by a consultant dermatologist.The right index finger showed sclerosis andtelangiectasia with three small crusted ulcersat the base, middle, and tip. The last ulcerinvolved the nail bed, with its partial destruc-tion. The remaining nails of both hands werealso abnormal with linear streaks and a raggedfree margin. There were several keratoticlesions mainly affecting the right hand. Inview of his professional use of hand heldsources, radiation was assumed to be thecause. A skin biopsy from the distal part of theright index finger showed keratin thickeningwith parakeratosis and considerable acantho-sis with irregular downgrowths of the basallayer. Individual cell keratinisation was seenhere, but no atypia. These features are consis-tent with late radiation dermatitis. No defini-tive evidence of malignancy was seen.Flucloxacillin, topical gentamicin, and potas-sium permanganate baths helped for a while.The man then presented at a hospital outpa-tients' clinic in November 1989 with lesionson both hands; the right being considerablyworse. The right index finger had deterioratedwith ulceration over the terminal phalanx,such that its bone and the distal joint wereexposed (fig 1). By February 1990 the fingerwas heavily infected. Terminalisation of thefinger was unsuccessful and amputation wascarried out with good results. The distal pha-lanx of the third finger was also removed.

Apart from a mild macrocytosis (mean cor-puscular volume 99 fl) routine blood screen-

713

on August 4, 2021 by guest. P

rotected by copyright.http://oem

.bmj.com

/O

ccup Environ M

ed: first published as 10.1136/oem.51.10.713 on 1 O

ctober 1994. Dow

nloaded from

http://oem.bmj.com/

Lloyd, Edwards, Fitzsimons, Evans, Railon, Jeffrey et al

ing was normal in February 1990. A bloodspecimen was also taken for chromosomalanalysis, which confirmed that the patient hadbeen excessively exposed to ionising radiation.He continued to be followed up as an outpa-tient and periodically further blood sampleswere sent for chromosomal analysis and rou-tine screening. In early 1991 the nail changesand keratoses were still present in the lefthand. Mild sclerosis was also slowly develop-ing.A blood examination in April 1991 showed

changes that indicated myelodysplasia,haemoglobin 9-5 g/dl, white count 3-8 (blasts0.2) x 109/1, platelets 44 x 109/1. This diag-nosis was confirmed by bone marrow exami-nation. This showed gross dyshaemopoieticactivity. Myelopoiesis showed hypogranula-tion, pelger-huet forms and blasts at 20%.Micromegakaryocytes were abundant anderythropoiesis was megaloblastic. A trephinebiopsy showed intense myelofibrosis withhypercellular islands of abnormal haemopoi-etic activity.

This myelodysplastic state in transforma-tion progressed to acute myeloid leukaemia byOctober 1991. Starting in April 1991 he wasdependent on transfusions at a frequency ofthree units every three weeks. He was treatedwith chemotherapy, mitozantrone 10 mg/M2on days 1 and 2, and a cytosine (1 g/m2) intra-venous infusion twice daily on days 1-3 inMarch 1992. This treatment induced pro-longed aplasia that was complicated by infec-tion. At the time of his death in June 1992aged 61 his marrow continued to show con-siderable hypocellularity.At a post mortem examination a healthy

molar tooth and some specimens ofbone wereremoved for analysis by electron spinresonance.

INVESTIGATION OF WORKPLACE ANDEMPLOYMENTIn March 1990, on learning that the patienthad a possible radiation induced burn, mem-bers of the Health and Safety Executive'sField Operations Division started inquiries.After a one month training course the manhad started industrial radiography in 1974.Before this he had been a printer. He hadworked exclusively in the United Kingdom forone employer, who provided contract radiog-raphy services. Throughout the whole perioduntil he ended radiography in 1990, theman had been based at a large oil and chemi-cals complex where he performed non-destructive tests on metal structures such aspipe welds. From 1974 until early 1983 he

Unstable chromosomal aberrations found in blood lymphocytes

Date of No of cells Centric Excess Dicentrc yeldIsample scored Dicentrics rings acentrics cell (SE)

5 February 1990 2000 215 25 316 0-108 (0 008)30 April 1990 500 35 6 68 0 070 (0-013)17July 1990 500 40 2 78 0-080 (0-014)13 December 1990 500 34 2 41 0-068 (0-013)22 May 1991 500 35 3 49 0 070 (0-013)l9June 1991 500 21 0 27 0-043 (0-010)

used iridium-192 sources ranging from 10 to25 Ci (400-1000 GBq) in torch type contain-ers. In a typical working day he would per-form about 30 radiographs. In 1983 thecompany stopped using these containers andradiography was then done with wind outremotely operated iridium-192 sources.Iridium-192 has a half-life of 74 days and aprincipal yray emission of 320 keV.An audit of the workplace in 1990 showed

that the equipment was in good working orderand documented procedures currently in usewere safe. The man had unremarkable dosesrecorded on his routine personal monitors,which were generally monthly film badgesuntil 1982, when the company changed tothermoluminescence dosemeters. For theperiod January 1986 to March 1988 he hadbeen supplied with fortnightly badges. Hislifetime total recorded dose was 104 mSv,none of which was notional, meaning that hehad always returned his dosemeters for pro-cessing. He had only been issued with badgesthat were supposed to be pinned to the trunkand he invariably wore them at his waist. Hehad never been supplied with fingertipdosemeters. By comparison with the recordsof other long serving employees of the com-pany, his recorded dose was lower than aver-age.The man had been examined medically in

1976 and 1982 under the terms of the 1969Ionising Radiations (Sealed Sources)Regulations3 and no problems had been iden-tified. Subsequently, his health was surveyedaccording to the Ionising RadiationsRegulations, 1985.2 His health and employ-ment records showed no abnormal periods ofabsence due to illness.

Interviews were conducted with theemployer's Radiation Safety Officer,Radiological Protection Adviser, a number ofradiographer colleagues of long standing, andthe man himself. No explanation to accountfor untoward irradiation was forthcoming. Itwas noted that in the period from 1983onwards when wind out sources were used forradiography the man undertook sourcechanges when a decayed source needed to bereplaced. This equipment was owned byanother company.

CHROMOSOMAL ANALYSIS; DOSIMETRYBetween February 1990 and June 1991 sixspecimens of heparinised blood were taken foranalysis of chromosomal damage.Lymphocyte metaphase preparations weremade from 48 hour cultures by a standardtechnique. First in vitro division cells wereexamined for unstable chromosomal aberra-tions; dicentrics, centric rings, and excessacentric fragments (table).

CHROMOSOMAL ANALYSIS; KARYOTYPINGBone marrow and peripheral blood were alsotaken in June 1991 for karyotyping. Both werecultured by standard methods for 24 hourswithout mitogenic stimulation. Slides wereprepared and marrow and blood blast cellswere karyotyped with Giemsa banding.

714



.bmj.com

/O

ccup Environ M


ctober 1994. Dow

nloaded from

http://oem.bmj.com/

Death ofa classified worker probably caused by overexposure to y radiation

ELECTRON SPIN RESONANCEThe distal phalanx of the right hand secondfinger, a piece of right humerus and a healthymolar tooth were removed at post mortemexamination, and stripped of soft tissue byplacing them in a colony of Dermestes maculatusbeetles. The specimens were sent to theUniversity of Osaka for electron spin reso-

nance (ESR) dosimetry, ensuring that theywere not subjected to x rays in transit. Enamelfrom the tooth was carefully removed fromthe dentine by a diamond cutter microsaw.The technique for ESR spectroscopy and

converting the signals to absorbed dose hasbeen described previously for atomic bombsurvivors and residents close to the Chernobylaccident.5 6 In brief, the sample is placed in a

static magnetic field and exposed tomicrowave radiation. The reflectedmicrowaves are detected and, by sweepingthrough the magnetic field, a characteristicdrop then a rise is seen in the microwave signalas a derivative absorption curve to the inten-sity of the magnetic field. This resonance isproduced by a change in the spin state ofunpaired electrons in the particular radical,CO, in hydroxyapatite, being examined. Fortooth enamel and bone the measured carbon-ate radical resonates at a magnetic field ofabout 335 mT in a microwave frequency of9.5 GHz. If the specimen has been exposed toionising radiation increased numbers ofunpaired electrons are produced, which per-sist almost indefinitely, subject to reminerali-sation of the tissue during life. The signalintensity was calibrated against radiation dosewith the additive dose method.

Results and discussionThe first problem with the right hand beganabout one year after the patient stopped work-ing with torch radiographic sources (1984).The patient treated himself and did not seekmedical advice or tell his employer. Fouryears later he sought treatment, so it is possiblethat the two episodes of swelling of the handwere unrelated and coincidental. The second

Figure 2 Vitosonics torch type of radiography source and its shielded carrying container.

episode, which led to the ulceration, and wasdiagnosed as being induced by radiation,could still have been due to the torch sources.A latency of many years (>10) may elapsebetween irradiation of the hand and develop-ment of ulceration or malignant degenerationof the skin, particularly if it is the cumulativeeffect of many exposures.78

Figure 2 shows a Vitosonics torch sourcecontainer that when not in use is housed inthe depleted uranium pot. A typical surfacedose rate from this apparatus when not in useis about 1 mGy h-'. In use the torch is han-dled by the knurled end. The smooth anddarker end contains the iridium-192 sur-rounded by a depleted uranium shield that isspring loaded. This source end is inserted intoa lead "castle"; an applicator, which has beenpreviously attached to the object being radi-ographed. The act of inserting the torchthrough a bayonet fitting, pushes back theuranium shield, and thereby exposes thesource. Typical proper use would produce10-20 mGy/y to the operator. Cases of burnsrelated to the use of such instruments havebeen reported, however, they usually result indamage to the non-dominant hand; whereasthis patient was right handed. These havearisen when the sources were inserted intoapplicators, before fixing the latter to theobjects being radiographed. Another misusethat has been discovered is the practice ofwedging the spring loaded shield back with,for example, a matchstick or sticky tape anddispensing with the use of the applicator.9 Inthis mode the radiographer's body would beexposed to y rays with the highest dose beingreceived by the hand holding the torch. Thismalpractice saves time when a lot of radi-ographs are being carried out on piece workpayments. Such misuse of sources is generalknowledge among industrial radiographers,but it should be emphasised that in the casereported here, misuse was denied.On changing over to use of wind out irid-

ium-192 sources the man occasionally under-took source replacements for anothercompany that owned radiographic equipment.These replacements occurred at the mosttwice yearly and the procedure took about 15minutes. The maximum strength of sourceever transferred was understood to be 22 Ci(814 GBq). In essence, the old source waswound from the source container into a holealong the central axis of a lead or steel shield-ing cylinder. A transverse slit in the cylinderenabled a hacksaw blade to be inserted so thatthe "pig tail" attachment could be severed.The detached source was then of a size suchthat it could be decanted into a torch typecontainer for continued use as a small sourceof ionising radiation. (Before this transfercould be achieved the spent source previouslyin the torch type container had to be put backinto a transport container for return toAmersham.) Done properly it was calculatedthat a source changeover from projection totorch container might entail a hand dose of380 yuGy and a whole body dose of 8 ,uGy.Unattenuated radiation was able to emerge,

715



.bmj.com

/O

ccup Environ M


ctober 1994. Dow

nloaded from

http://oem.bmj.com/

Lloyd, Edwards, Fitzsimons, Evans, Railton, Jeffrey et al

however, from holes in the shielding and if thehand and upper body or head had been insuch a beam possible doses of 140 mGy and 3mGy respectively could have been received.In such circumstances the waist, the normalsite for wearing his personal dosemeter, andmuch of the lower body would be shielded bythe shielding container. Procedures were inplace to prevent exposure to the unattenuatedradiation and there was no evidence that suchexposure took place.As no explanation for the severe lesion to

the hand was given, the first firm corrobora-tion that the patient had been significantlyirradiated, was provided by the chromosomalanalysis in February 1990. The incidence ofaberrations (table) was considerably in excessof that expected from studies of control sub-jects exposed only to background radiation,namely about one dicentric, no rings, andthree acentric fragments/1000 cells.4 Whereasacentric fragments can be caused by a varietyof agents, a principal cause of dicentrics andrings in lymphocytes is ionising radiation. Theonly possibility other than radiation is a shortlist of cytotoxic drugs, which he had certainlynot been prescribed. It was also apparent thatthe exposure was more complex than a singlebrief and very heavy irradiation of a hand. Forsuch an event chromosomal analysis of circu-lating lymphocytes would be unlikely todetect aberrations much in excess of back-ground. The proportion of the body's lym-phocytes that would be within the radiationfield would be considerably diluted by the cir-culation and the few that might be containedin the blood specimen would be unlikely toreach metaphase because of interphase deathand mitotic delay.

Induction of dicentric aberrations has ahighly reproducible relation with radiationdose, so that with reference to an in vitro cali-bration curve, it may be used as a biologicaldosemeter.4 Interpretation of the dicentrics interms of dose requires some knowledge of thetime interval between irradiation and bloodsampling, information on whether exposurewas brief or protracted, single or repeatedevents.Some limited data from humans indicate

that dicentric yields in lymphocytes declinewith a half time of about three years and aftera high (> 1 Gy) exposure, there is a more rapidfall off over the first year.4 There may be someevidence of this from the earlier data in thetable, but it is not clear cut. Exposures lastingfor >30 minutes result in a lower aberrationyield than if the same dose is received briefly.Also, intermittent exposures, even if eachfraction is brief, will similarly result in a loweraberration yield, than if the whole dose isgiven at one time.4With no information on the time course of

exposure, the results (215 dicentrics in 2000cells) from the first blood sample were inter-preted as indicating an average whole bodydose of around 1 Gy if the exposure had beenreceived as a single brief event at some time inthe few months before February 1990. If,however, exposure had been protracted or

intermittent during those months, the esti-mate of dose would rise to about 5 Gy. If itwas assumed that the hand lesion, includingthe original swelling in 1984, was connectedwith more general irradiations that caused theaberrations seen in the lymphocytes, thenmuch older exposure needs to be considered.A single brief exposure 10 years before bloodsampling would lead to an estimate of about 4Gy although it is difficult to envisage thisoccurring without the patient having earlysymptoms of radiation sickness. Many inter-mittent exposures, each being either brief orprotracted, spanning the period 1974 to 1983,when he worked with the torch sources,would lead to a much higher estimate of possi-bly several lOs Gy.An indication that at least some of his

exposure was not recent was that as well asthe aberrations noted in the table each speci-men also contained cells with dicentrics orrings that were missing the usually accompa-nying acentric fragment. In the first specimen,which was taken while he was haematologi-cally normal, there were an additional 49dicentrics and five ring "incomplete" aberra-tions noted in lymphocytes that had under-gone only one in vitro cell cycle. This impliesthat in vivo division had occurred at whichfragments were lost by non-disjunction. Asperipheral T lymphocytes rarely divide andhave long life spans this is suggestive of oldexposure but the findings can not be usedquantitatively.10The distributions of dicentrics among the

scored cells were significantly overdispersedwith respect to the Poisson distribution thatwould characterise a uniform whole body irra-diation.4 For the first specimen the ratio ofvariance to mean was 1 24, u = 7 5, whichprovides evidence of non-uniformity andwould correspond to about half the body hav-ing been irradiated. It is difficult to envisagehow such a result could have come fromrepeated hand exposure where perhaps only1% of the lymphocytes were exposed on eachoccasion. The electron spin resonance analy-sis, done two years later, confirmed that therehad indeed been exposure to more than justthe hands.

In the table it can be seen that from April1990 to May 1991 the observed yield ofdicentric aberrations was constant. It then fellconsiderably in the final sample in June 1991.The last two samples are from the periodwhen myelodysplasia had developed. As wellas the aberrations listed, the penultimate sam-ple also contained a very obvious clone of cellswith 44 chromosomes including a centric ringbut without an accompanying acentric frag-ment. This abnormality had not been foundin the previous samples. In the final sample ofJune 1991 the clone was even more pro-nounced and the reduced yield of dicentricaberrations may well reflect the dilution oflymphocytes by blast cells.

Chromosomal analysis of bone marrow andunstimulated peripheral blood taken in June1991 showed the following karyotype in allcells examined:

716



.bmj.com

/O

ccup Environ M


ctober 1994. Dow

nloaded from

http://oem.bmj.com/

Death ofa classified worker probably caused by overeVposure to y radianon

44,XY,del(5) (qI 3q33),-7,derdel(7),-1 2,der(1 2),der(17p),-20, + ring

This is similar to blast cell karyotypesfound in secondary myelodysplasia arising inpatients previously treated with genotoxicagents for cancer."l Of patients with secondarymyelodysplasia 55-84% progress to acutemyeloid leukaemia and their median survivalis <1 year. Clonal abnormalities have beenreported in 76-97% of cases, with -7 anddel(5) seen in up to 90%, which are consid-ered to be characteristic. Other changes asso-ciated with secondary myelodysplasia or acutemyeloid leukaemia include rearrangementsaffecting 12p and 17p. Unstable configura-tions such as rings and dicentrics are alsocharacteristic. Secondary myelodysplasia oracute myeloid leukaemia have primarily beenassociated with exposure to alkylating agentsand more recently epidophyllotoxin, but irra-diation to a large volume of bone marrow alsoincreases the risk of leukaemia. It has beensuggested that 5q- is particularly associatedwith radiotherapy whereas -7 is more char-acteristic of chemotherapy. The medianlatency of secondary myelodysplasia or acutemyeloid leukaemia related to treatment is4-5 years but with a range of <1->25years.11 1213

Data from Pedersen-Bjergaard et al suggestthat there may be differences between thosepatients who develop secondary myelodyspla-sia or acute myeloid leukaemia 30-60 monthsafter treatment and those who develop iteither before 30 months or after 60 months.'4In the group with a latency of <30 months,patients most often present with myelodyspla-sia rather than acute myeloid leukaemia (88%v 47%), have an abnormal karyotype (100% v66%) with a hypodiploid modal number(100% v 40%) and combined B and C groupchromosome changes (50% v 13%). It isinteresting that this patient presented withmyelodysplasia and a hypodiploid karyotypewith both B and C group chromosome abnor-malities.The tooth and bone specimens examined

post mortem by ESR provided further evi-dence that the man had been excessively irra-diated. The value derived from the tooth was14-5 (2 0) Gy. This is possibly the most reli-able evidence in the sense that once a tooth isformed its enamel is never renewed and so thedose derived from it reflects almost the wholeof the person's life time exposure. As this mandied in his early 60s one would expect thetooth to have received 50-100 mGy from nat-ural background radiation and, unless he hadreceived an exceptional amount of dentalradiography, about another 100 mGy fromthis source. The values obtained from thebone specimens were 14-7 Gy for the fingerdigit and 7 0 Gy for the humerus. It should beremembered that unlike dental enamel themineral material of bone is subject to renewalthroughout life, and hence loss of signal.Nevertheless these values, which representfixed sites in the body, complement the averagedose estimate obtained from the lymphocytes,and confirm that the patient had received very

high exposure that was not confined just tothe hands. The magnitude of the estimateddoses is considerably above the threshold foracute radiation sickness and thus provides fur-ther confirmation of the postulate that he hadbeen exposed on many occasions rather thanas a single event.A computer program (SPIDER-1'5) that

considers radiation related risks and probabil-ity of causation of cancer has been developedat the National Radiological ProtectionBoard. It follows a BEIR-V type risk projec-tion model,'6 modified to be specific to thepopulation of the United Kingdom,'7 by usingUnited Kingdom Office of Population Censusand Surveys cancer statistics and life tables for1985.The algorithms were applied to the case of a

man, dying aged 61, after an average wholebody exposure of 10 Gy, received at a con-stant rate spanning 10 years from the age of40-49. To allow for the time course of irradia-tion a dose and dose rate effectiveness factorof 2 was assumed.'7 With these assumptionsthe annual risk of fatal leukaemia in the agerange 60-65 years was 0-22% compared withan annual background risk of 0-013%. Theresultant probability of causation is

0-220.22+0.013 094

This seems fairly conclusive although it mustbe remembered that this is a probability ratherthan a proof of causality. Apart from theunderlying uncertainties of total dose and itstime course the main limitation of this calcu-lation is that epidemiological and demo-graphic data are applied to the individual. Itassumes that the individual is of average sus-ceptibility to cancer. Despite this the proba-bility of causation approach'8 seems to be alogical and reasonable procedure for assessingthe presumption of malignancy induced byradiation.

ConclusionChromosomal and electron spin resonanceanalyses have provided firm evidence of seri-ous overexposure to penetrating radiation.Although all the evidence suggests that thiswas due to work with industrial y sources ithas not been possible to establish how theexposure occurred. The right hand intermit-tently received an immense dose; possibly 100Gy. This resulted in chronic dermatitis thatprogressed to ulceration and required partialamputation. Much, perhaps all, of the bodywas also exposed to a substantial averageddose, likely to be in excess of 10 Gy, whichwas received intermittently probably spanningseveral years. It is highly probable that thiscaused his fatal myelodysplasia, which wascharacterised by blast cells of which the mor-phology, karyotypes, and progression to acutemyeloid leukaemia bore a striking resem-blance to those reported in patients whodevelop this condition some time after a suc-cessful course of radio and chemotherapy forcancer.

717



.bmj.com

/O

ccup Environ M


ctober 1994. Dow

nloaded from

http://oem.bmj.com/

Lloyd, Edwards, Fitzsimons, Evans, Railton, Jeffrey et al

We thank Mr J D Robb of the National RadiologicalProtection Board for carrying out the probability of causationcalculation.

1 Health and Safety at Work etc Act, 1974. London: HMSO,1974.

2 The Ionising Radiations Regulations, 1985. London: HMSO,1985. (Statutory Instrument No 1333.)

3 Ionising Radiations (Sealed Sources) Regulations, 1969.London: HMSO, 1969. (Statutory Instrument No 808.)

4 Bioloical dosimety: chromosomal aberration analysis for doseassessment. Vienna: International Atomic Energy Agency,1986. (Technical Report No 260.)

5 Ishii H, Ikeya M. An electron spin resonance system for invivo human tooth dosimetry. Japanese Journal ofAppliedPhysics 1990;29:871-5.

6 Ikeya M. New applications of electron spin resonance, dating,dosimetry and microscopy. Singapore: World Scientific,1993.

7 Tilkorn H, Drepper H, Ehring F. Indications for the treat-ment by plastic surgery of the effects of radiation andradiolesions on the skin. Br Radiol 1986;(suppl 19):131-4.

8 Hartwell S, Huger W, Picknell K. Radiation dermatitisand radiogenic neoplasms of the hands. Ann Surg1964;160:828-34.

9 Mullarkey DT. Radiation overexposure of an industrialradiographer. Radiological Protection Bulletin 1974;6:16-7.

10 Lloyd DC, Purrott RJ, Dolphin GW. Chromosome aber-ration dosimetry in a case of overexposure to radiation.Nature 1973;241:69-70.

11 Levine EG, Bloomfield CD. Leukaemia and myelodys-plastic syndromes secondary to drug, radiation and envi-ronmental exposure. Semin Oncol 1992;19:47-84.

12 Johansson B, Mertens F, Heim S, Kristoffersson U,Mitelman F. Cytogenetics of secondary myelodysplasia(sMDS) and acute nonlymphocytic leukaemia (sANLL).EurJ Haematol 1991;47:17-27.

13 Pedersen-Bjergaard J. Radiotherapy-and chemotherapy-induced myelodysplasia and acute myeloid leukaemia.A review. Leuk Res 1992;16:61-5.

14 Pedersen-Bjergaard J, Philip P, Mortersen BT, et al. Acutenonlymphocytic leukaemia, preleukaemia and acutemyeloproliferative syndrome secondary to treatment ofother malignant diseases. Clinical and cytogenetic char-acteristics and results of in vitro culture of bone marrowand HIA typing. Blood 1981;57:712-23.

15 Stokell PJ, Robb JD, Crick MJ, Muirhead CR. SPIDER-1:software for evaluating the detriment associated with radia-tion exposure. Chilton, Oxfordshire: NationalRadiological Protection Board 1993;Report R-261.

16 BEIR-V. Health effems of exposure to low levels of ionis-ing radiation. Fifth report ofthe Committee on the BiologicalEffects of Ionising Radiation. Washington: NationalAcademy of Sciences, National Research Council, 1990.

17 Muirhead CR, Cox R, Stather JW, MacGibbon BH,Edwards AA, Haylock RGE. Estimates of late radiationrisks to the UK popuaion. Documents of the NRPB.Chilton, Oxfordshire: National Radiological ProtectionBoard, 1993. (Vol 4:15-157.)

18 National Council for Radiation Protection. The probabilitythat a particular malignancy may have been caused by aspecific irradiation. Bethesda: NCRP, 1992. (StatementNo 7.)

Correspondence and editorialsOccupational and Environmental Medicine wel- minimum Letters are accepted on thecomes correspondence relating to any of the understanding that they may be subject tomaterial appearing in the journal. Results editorial revision and shortening.from preliminary or small scale studies may The journal also publishes editorials whichalso be published in the correspondence are normally specially commissioned. Thecolumn if this seems appropriate. Letters Editor welcomes suggestions regardingshould be not more than 500 words in length suitable topics; those wishing to submit anand contain a minimum ofreferences. Tables editorial, however, should do so only afterand figures should be kept to an absolute discussion with the Editor.

718



.bmj.com

/O

ccup Environ M


ctober 1994. Dow

nloaded from

http://oem.bmj.com/