Interstitial laser hyperthermia: a new approach to local destructionof tumours

A C Steger, W R Lees, K Walmsley, S G Bown

AbstractThe use of local hyperthermia to treat cancer of theinternal organs has been limited by the difficulty ofcontrolling delivery of heat and limiting the effects tothe tumour, but this can be overcome by using laserlight transmitted through thin flexible fibres. Laserenergy was delivered to tumours through fibresinserted percutaneously through needles directlyinto the centre of the tumour area. Ultrasoundscanning was used to locate the tumour, position thefibres correctly within the tumour, and monitor thedevelopment of thermal necrosis in real time duringlaser exposure and through the subsequent period ofhealing. Five patients were treated (one with atumour of the breast, one with a subcutaneoussecondary tumour, one with a recurrent pancreatictumour, and two with secondary tumours in theliver). Tumour necrosis was found on ultrasono-graphy or computed tomography in all, and therewere no immediate or delayed complications. In onepatient the size of the isolated secondary tumour inthe liver had not increased over 10 months, and hesubsequently showed no other evidence of residualcancer.To develop this technique careful studies are

essential to ensure that in every case the extent ofthermal necrosis produced by absorption of the laserlight can be matched to the full extent of the tumourbeing treated and that there is always sufficientadjacent normal tissue to ensure safe healing. Thesepreliminary results suggest that this simple techniquecan be applied safely and effectively to commontumours in humans; more extensive trials in a rangeof cancers of solid organs are warranted.

IntroductionHyperthermia has for many years been investigated

as a way of destroying diseased tissue. The tempera-tures reached during conventional treatment by hyper-thermia induced by microwave or radiofrequencyapplicators (41-440C) 2 have been used in treating avariety of superficial and deep seated tumours, as haswhole body hyperthermia3; varying degrees of re-mission have been achieved. These techniques, how-ever, are not precise in their control of energy deliveredto the tumour.

Interstitial treatment delivers treatment directly intothe centre of the target lesion, minimising the effectson surrounding normal areas. Radioactive isotopegrains have been placed through needles into tumoursof the pancreas4 and liver.5 Interstitial hyperthermiainduced by lasers by inserting the light conductingquartz fibre into the tumour was first described in1983.6 Experimental work showed that areas of thermalnecrosis ofpredictable extent (up to 1 6 cm in diameter)could be produced in the livers of rats with verv lowpowers from a neodymium yttrium aluminium garnet(Nd YAG) laser (0 5-2 W). The lesions heal by fibrosisand regeneration.7 With ultrasound scanning the de-velopment of these thermal changes can be visualisedin real time during laser irradiation, their resolution

during healing followed, and the appearances matchedwith histological changes and the tissue temperaturesreached.' In this pilot clinical study we applied thistechnique to four types of tumour in five patients.

MethodThe five patients were referred as no other treatment

was considered appropriate or they had refused alter-natives. All had given informed consent. All tumourswere accessible for percutaneous insertion of a flexiblefibre for delivery of laser energy. The nature of thelesion being treated was confirmed histologically.Sedation and analgesia (diazepam 5-10 mg, pethidine50 mg intravenously) were given to the patientsundergoing an intra-abdominal application togetherwith prophylactic antibiotics (flucloxacillin 500 mgwith added gentamicin 80 mg for the intra-abdominalapplications). Local anaesthesia (1% plain lignocaine)was given at the site of needle insertion.The laser fibres (0-2-0-4 mm core diameter) were

sterilised in glutaraldehyde and rinsed with saline.Each fibre was inserted by passing it down a 0 9-1 2mm needle, the tip of which had been placed at anappropriate point within the tumour under ultrasoundcontrol. Once in the correct position, the fibre tip wasadvanced 3-5 mm from the needle to be in directcontact with tissue. When several treatment pointswere used the distance between these was <1 5 cm,the maximum separation at which destruction of allintervening tissue could be ensured.'

Both continuous wave (Flexilase, Living Tech-nology, Glasgow) and pulsed (Lumonics MS 35 LD,Rugby) Nd YAG lasers were used with a mean powersetting of 1-5 W for 500-670 s at each site. The twolasers gave equivalent thermal necrosis.9 In case 5 a1 x 4 Star coupler (Canstar, North York, Ontario) wasused to allow delivery of energy at four sites simul-taneously; for technical reasons this could only be donewith the continuous wave laser.

Case I-A 57 year old man had an anterior resectionfor a rectal carcinoma with resection of two smallhepatic metastases in March 1987. No other depositswere detected at that time. In July 1988 a follow upultrasound scan showed a 35 mm diameter lesion in theright lobe of the liver, positioned near the inferior venacava and the right and left hepatic veins, whichprecluded surgical removal. Four sites were treatedwith the laser using a single fibre. He reported no painand was discharged well the following day. Eightweeks later ultrasonography showed a hypoechoicring surrounding a central hyperechoic area (fig l(a)).The initial biopsy had confirmed secondary adeno-carcinoma, whereas the repeat biopsy showedonly necrotic tissue. To maximise the chance ofablating the entire lesion the outer part of the tumourwas treated at three sites (fig l(b)). Two furthertreatments were carried out six and 10 months later asultrasound scans suggested regrowth, but at 10 monthsafter the original treatment he remained well, thelesion in his liver was no larger than it was originally,and he had no other evidence of residual cancer.

BMJ VOLUME 299 5 AUGUST 1989

National Medical LaserCentre, University Collegeand Middlesex School ofMedicine, LondonWC1E 6JJA C Steger, FRCSGLAS,research fellowS G Bown, MD, director

Department of Radiology,Middlesex and UniversityCollege Hospitals,London WIN 8AAW R Lees, FRCR, consultantradiologistK Walmsley, FRCR,consultant radiologist

Correspondence to: MrSteger.

BSrAfledj 1989;299:362-5

362

on 1 April 2022 by guest. P

rotected by copyright.http://w

ww

.bmj.com

/B

MJ: first published as 10.1136/bm

j.299.6695.362 on 5 August 1989. D

ownloaded from

the breast refused all forms of conventional treatment,including surgery, radiotherapy, and tamoxifen, over aperiod of two years. The tumour size was monitoredand its volume calculated with ultrasonic data. Whenthe tumour grew rapidly she was offered and acceptedinterstitial laser treatment. Three sites were treated.She declined any sedation and had only local anaesthe-sia but experienced no pain or discomfort. Twomonths later a second treatment was instigated in anattempt to destroy residual tumour. Figure 3 showstumour growth in relation to treatment. After a furthertwo months she was found to have distant metastases;she refused further treatment.

Laser4 treatment

3-

E

0

E

FIG 1-Case 1: (a) tumourimmediately afterfirst lasertreatment; (b) tumour (t)immediately after second lasertreatment, showing hypoechoicrim (r) between tumour andsurrounding liver

Case 2-Eighteen months after a distal pancreatec-tomy a 53 year old woman developed severe centralabdominal pain. Ultrasonography showed a recurrence

of the tumour 2 cm in diameter in the residual pancreasand further surgery was thought to be inappropriate.Three sites were treated. A computed tomographyscan 14 days later showed evidence of oedema in theretroperitoneum and an increase in the amount ofcentral necrosis in the tumour (fig 2). The serum

amylase activity rose from 30 IU before treat-ment to 70 IU after 24 hours and then fell to 35 IU at 72hours (normal range 0-300 IU). Though ultrasono-graphy and computed tomography showed thermaldamage to the tumour, her main symptom, pain, was

not improved. She died six months later with dis-seminated disease.

Case 3-A 51 year old woman with a carcinoma of

Weeks of observationFIG 3-Case 3: growth of volume of tumour as measured byultrasound, showing response to laser treatment

Case 4-Five years after a radical partial gastrectomyfor carcinoma a 73 year old man developed a malignantpleural effusion that required drainage. Secondarytumours (12 x 5 mm and 8 x 5 mm) developed in theskin at the site of drainage of the chest and recurreddespite two surgical removals. These were treated attwo sites for the larger tumour and one for the smaller.Three months later the smaller tumour was no longerdetectable and the larger one was smaller (8x4 mm)and firmer. Nine months after the original treatmentthe pleural effusion had recurred, but the skin nodulesremained unchanged.

Case 5-Five months after an anterior resection forcarcinoma of the rectum a 70 year old man had a

FIG 2-Case 2: (a) arrowindicates tumour in head ofpancreas; (b) tumour 14 daysafter laser treatment. Size ofcentral necrotic area hasincreased, and halo hasdeveloped, suggesting oedema

BMJ VOLUME 299 5 AUGUST 1989

20 40 60 80 100 120

363

on 1 April 2022 by guest. P

rotected by copyright.http://w

ww

.bmj.com

/B

MJ: first published as 10.1136/bm

j.299.6695.362 on 5 August 1989. D

ownloaded from

FIG 4-Case 5: (a) patient positioned fi)r treatmentwith needles and fibres in position; (b) ultrasonicappearatnce of tumoutr after laser treatment, outlinedin two dimensiotns (x . . x and + . +); (c)ultrasonic appearance of turmour (+ . . . +) 10 daysafter laser treatment, show1ig clear margin (arrowed)between treated tumour and surrounding liver

Hartmann's operation for local recurrence, and fivemonths after this ultrasonography showed a singlesecondary tumour in the liver (5 x 5 cm). Four needleswere positioned under ultrasound guidance with -0-1-5 cm between the tips in a square (fig 4(a)). Thefour output fibres of the coupler were fed down theneedles, and laser treatment was applied. The needlesand fibres were then drawn back 2 cm, the samespacing between the fibre tips confirmed, and thetreatment repeated. Figure 4(b) shows the treatedarea (9x7.7 cm). The patient was discharged well48 hours later. Ultrasound examination after 10 daysshowed a well demarcated lesion measuring 4-6 x 5 *6cm(fig 4(c)) with appearances suggestive of necroticchange. Computed tomography two months latershowed an area of necrosis 5 cm in diameter with welldefined margins in the treated area, but also detectedanother tumour deposit.

DiscussionOne of the main problems with the techniques for

inducing hyperthermia that are currently used is todeliver and localise the heat to the target area. Conse-quently, most trials have been limited to skin tumours.Deep seated tumours present the problems of avoidingdamage to adjacent normal areas and ensuring adequatetemperatures throughout the volume of diseased tissuebeing heated.

Interstitial laser hyperthermia at low power over-comes many of these problems. Our case historiesindicate that the technique is feasible and safe. Thelaser fibre is thin and can easily be passed down thin

needles into either superficial or deep seated lesionswith minimal damage to the normal tissue throughwhich it passes, making this method of inducinghyperthermia potentially applicable to solid tumoursin many parts of the body. In addition, ultrasoundscanning can not only define the target lesion andthe position of the fibre but can also monitor thethermal changes in the treated area in real time, whichgives an independent method of assessing the extent ofnecrosis, an advantage not seen before in clinicalhyperthermia.We are not claiming cures of the patients reported

here, but tumours were at least partially destroyed inall cases and thermal damage healed safely. Growth ofthe two tumours in the liver was reduced.The Nd YAG laser has been used in Japan at higher

powers (5-15 W) at open surgery with a sapphire lightdiffuser tip on the laser fibre to treat hepatomas.Reduction in serum (t fetoprotein concentration and oftumour size on computed tomography followed,"' butthe technique was much cruder than the low powermethod we describe. A device to use even higherpowers (up to 100 W) with the fibre within a handpiecethat also contains a water cooling circulation for the tipof the fibre has been developed in France." Neither ofthese techniques is as suitable for percutaneous appli-cation as the one described here.

Another interstitial technique described recently isthe use of95% ethyl alcohol injected into small hepatictumours under computed tomography and ultrasoundguidance. 12 '1 The changes of necrosis seen on follow upultrasonography were similar to those that we haveseen, and necrosis in 70-90% of the tumour volume was

BMJ VOLUME 299 5 AUGUST 1989364

on 1 April 2022 by guest. P

rotected by copyright.http://w

ww

.bmj.com

/B

MJ: first published as 10.1136/bm

j.299.6695.362 on 5 August 1989. D

ownloaded from

seen on further ultrasonography and in biopsy speci-mens. Several treatments were required for eachtumour, and injecting alcohol was often associatedwith considerable pain, whereas our patients did notreport pain. These reports did not mention changesseen on ultrasound scans during or immediately afterinjection, which we found useful in laser treatment.The most important advantage of the laser is its

precision. It is unlikely that it will ever be possible topredict the extent of necrosis around a site at whichabsolute alcohol has been injected with an accuracycomparable to that already possible with the lasertechnique.

In conclusion, interstitial laser hyperthermia isfeasible and seems to be safe. A multiple fibresystem makes it feasible to treat tumours of clinicallyrelevant size in the centre of solid organs. The realchallenge for the future will be to develop diagnostictechniques that disclose exactly how far individualtumours extend in a wider range of organs (unlike thewell defined tumours treated in this pilot study) and toestablish the conditions of laser treatment that givecomplete tumour ablation with safe healing. Thiscombination of technologies may be valuable fortreating otherwise untreatable tumours in a range ofsolid organs and for the primary treatment of smallneoplasms such as tumours of the prostate and adrenalglands.

We thank Mr R C G Russell, Mr P Hawley, Mr W Slack,and the late Professor C G Clark for referring these patients

and for permission to report these results. We also thank Dr TN Mills, Mr P Hill, and Miss L A Potter of the department ofmedical physics for their help. Mr Steger was supported byLiving Technology Ltd, Glasgow, and Dr Bown by thespecial medical development on lasers from the Departmentof Health and by the Imperial Cancer Research Fund.

1 Storm KF, Kaiser L-R, Goodnight JE, ei al. *Fhermotherapy for melanomametastascs in liser. Cancer 1982;49: 1243-s.

2 Lindholm C-E, K'ellan E, Nilson P, Hertzman S. Microwave indtuccdhy-perthermia and radiotherapy in human superficial tumotirs clinicalresults with a comparative study of combined treatment versus radiothcrapyalone. lnr,7 s'p1erpthermia 1987;3:393-41 1.

3 Milligan AJ. Whole body hypcrthertmia iinductioni tcchniqtucs. Cancer Res1984;44 (10 Suppl):4869-72.

4 Shiplc WU, Nardi GL, Cohen AM, Cliftoin Ling C. Iodine-'"' implant andcxtcrnal bceam irradiation in patienits with localized pancreatic carciinoma.Cancer 19X0;45:709-14.

5 Dritclilo A, (irant EG, Harter KW;, Holt RWY, Rustigi SN, Rodgers JE.Intcrstitial radiation therapy for hcpatic metastases: sonographic guidancefor applicator placement. Am,7 Radtol 1986;164:275-8.

6 Bown SG. Phototherapy of tumors World.7 Surg 1983;7:700-9.7 Matthewson K, Coleridge-Smith 1, O'Sullisan JP, Northfield PIC, Bown SG.

Biological effects of intrahepatic Nd-YAG lascr photocoagulation in rats.Gastrosnicrologv 987;93:550-7

8 Stcger AC, Bown SG, Clarke C(G Intcrstitial lascr hypcrthermia-stttdics in thenortmal liver. fBrj Surg 1988;75:59X

9 Mlatthewson K, Coleridge-Smith 1P, Northlfield I,, Bowil SG. Cotoparisonl ofcontintUoUs uave attd puIlsed excitationi for interstitial Nd-YA( iidttdccdhvprthermia. lasers inn Medical Sciunce 1986;1 197-201.

10 Hashimoto 1). Ultrasonography guidcd lasers and sphcric lasers. In: RicmannJF, Ell C, cds. Lasers in gastrocnicrologs' Gec)rg Thiemc V'crlag Inc,Stuittgart: Thicmc P'uLblishicrs, 1989:134-8.

11 Godlewski (i, Sambtc P, Eledjam JJ, Pignodel C, Ould-Said A, Bourgeois JM.A new device ftr inducing deep localised vaptrisation in liver with the Nd-YAG laser. lasers inMedical Science 19XX;3: 111-7.

12 Shiina S, Yasuda H, Aluto H, et al. P'ercutaneous ethanol iijectioin in thetreatment of liser neoplasms. Amj7 Radiol 1987;149:949-52.

13 Livraghi T, Fcsti Ml, Monti F, Salmi A, Vcttori C. US-guidcd pcrCutatlcousalcohol iIjectioIn of small hepatic aind abdomiial tuimiors. Radiology1986;161 :3(9-12.

(A.4septsd31 .IlaY 1919

Departments ofRheumatology and ClinicalPharmacology, StBartholomew's Hospital,London ECIA 7BEPeter Fisher, FFHOM, visitintgrheumatologistAlison Greenwood, SRN,clinical metrologistE C Huskisson, FRCP, head ofrheumatolopv departmentPaul Turner, FRCIP, projfessorof clinical pharmtacology

Laboratoires Boiron, 69110Ste Foy les Lyon, FrancePhilippe Belon, MD, researchdirector

Correspondence to: DrFisher.

BrAfedj 1989;299:365-6

Effect of homoeopathictreatment on fibrositis (primaryfibromyalgia)

Peter Fisher, Alison Greenwood, E C Huskisson,Paul Turner, Philippe Belon

In scientific research negative results are often moredifficult to interpret than positive ones, as was shownby a clinical trial in which the homoeopathic medicineRhus toxicodendron 6x was compared with a placeboand fenoprofen in the treatment of osteoarthritis. Thehomoeopathic medicine was found to be ineffectivewhereas fenoprofen gave an improvement.' Therewere two interpretations: either the effects of homoeo-pathy are only a placebo effect-that is, a true negative-or the result was a false negative one because offlaws in the design. Another trial had previouslysuggested that homoeopathy was effective in rheuma-toid arthritis.We designed a trial to clarify these results by

overcoming the methodological criticisms while retain-ing a rigorous design. The main problem in designingclinical trials of homoeopathy is that prescriptions arebased on criteria such as the pattern of symptoms aswell as the diagnosis. A clinical trial based solely ondiagnosis is therefore inappropriate. In a pilot studywe had shown that R toxicodenidron 6c was the mostcommonly indicated homoeopathic medicine forfibrositis in our patients, being indicated in 42%.

Patients, methods, and resultsWe used the diagnostic criteria of Yunus et al for

fibrositis.' Only patients with this syndrome, in whomthe homoeopathic medicine R toxicodendron 6c waspositively indicated were entered into the study.Thirty patients meeting the admission criteria were

recruited in the rheumatology clinic of this hospital.The clinical characteristics of the patients were similarto those of other reported series regarding age, sexdistribution, duration of symptoms, modalities, andnumber of tender points. The trial was doubleblind, placebo controlled, and of crossover design.After entry there was no further contact between thehomoeopathic doctor and the patient until the treat-ment was finished. The clinical metrologist dispensedthe treatment and performed the assessments andanalyses blind. Patients received active treatment andan identical placebo for one month each in randomsequence. The dose was two tablets sucked three timesdaily.The active preparation was R toxicodendron 6c

(Boiron) prepared from a tincture of the leaves ofpoison oak diluted 1:99 in ethanol and then vigorouslyshaken. This process was repeated six times to give the6c potency-a dilution of 102 of the tincture. This wasthen put up on 125 mg lactose tablets (2% volume perweight). Preparation was as specified in the Frenchnational pharmacopoeia. The placebo was identicallactose tablets to which only pharmaceutical ethanolhad been added (2% volume per weight). Blind testingof the active and placebo preparations for a battery ofdrugs yielded negative results. Assessment comprisedthe number of tender spots, 10 cm visual analoguescales of pain and sleep, and overall assessment.Comparison was made between values at the end ofactive and placebo treatment periods.The patients did better in all variables when they

took active treatment rather than placebo. The numberof tender spots was reduced by about a quarter(p<0005). We reduced subjective data to nominalmeasurements (worse or better). If the null hypothesiswere correct the direction of change after placeboand active treatment would be randomly distributed.Analysis showed a significant difference in favour ofthe homoeopathic medicine (table). Overall assess-ment also showed a preference for the active treatment,which was not significant.

BMJ VOLUME 299 5 AUGUST 1989 365

on 1 April 2022 by guest. P

rotected by copyright.http://w

ww

.bmj.com

/B

MJ: first published as 10.1136/bm

j.299.6695.362 on 5 August 1989. D

ownloaded from