Chnical Radiology (1986) 37, 267-271 000%9260/86/581267502.00 © 1986 Royal College of Radiologists

Conformation Therapy: A Method of Improving the Tumour Treatment Volume Ratio TERESA TATE, J. A. BRACE, HELEN MORGAN and D. B. L. SKEGGS

Departments of Radiotherapy and Oncology and Medical Physics, The Royal Free Hospital

Restriction of the volume irradiated is a fundamental tenet of radiotherapy. Conventional two-dimensional treatment results in geometrically shaped high-dose regions, the sizes of which are defined by the greatest dimensions of a tumour mass. Conformation therapy permits a tumour to be considered for treatment as multi- ple short segments, and the radiation field to be tailored to fit each segment accurately. This paper demonstrates that this results in a reduction of the treatment volume of between 10% and 80%, depending on the shape of the tumour. Also, it is shown that a greater proportion of the high-dose region is occupied by tumour, thus reducing the volume of normal tissue treated to a high dose. Smaller treatment volumes allow the prescription of tumoricidal doses and improve the tolerance of radiation by patients.

It has long been recognised that the volume of tissue irradiated will affect the response of a patient to radia- tion. This paper demonstrates that the use of conforma- tion (three-dimensional) therapy achieves a substantial reduction in both the 'high dose volume' and the 'treat- ment volume' when compared with a conventional tech- nique. This reduction in treatment volume, because of a closer fit to the tumour, allows a greater length of field to be irradiated with no increase in acute or long-term complications. The term 'treatment volume' is used as defined in the International Commission on Radiation Units and Measurements (I.C.R.U.) Report (1978): 'that volume enclosed by an isodose surface, the value of which is the minimum target absorbed dose'. The 'high dose' volume is that enclosed by the prescribed dose isodose line (Fig. 1).

Ralston Paterson (1948) was the first radiotherapist to write of the 'volume factor', and to develop a theory relating the size of the treatment volume to the dose prescribed. He reached two conclusions which were later confirmed by others. First, the tissue tolerance falls as the volume irradiated increases, and second, the dose required to sterilise a tumour increases with the size of the mass. Paterson also wrote of the concept of 'margi- nal tissues', the adjacent normal tissue which of necessity is irradiated to a significant dose, and stated that an ideal treatment plan has a minimal margin sur- rounding the target volume. This concept has been universally accepted, but few improvements in tech- nique have been directly based upon it. Several authors have published mathematical formulae which demonstr- ate theoretically a relationship between tolerance and volume irradiated (Kirk et al., 1975; Gupta and Ajayi, 1982; Schultheiss et al., 1983), but to date, little has appeared in the literature on the application of these

theories to clinical practice. This study, based on confor- mation therapy, shows that for various tumour sites, significant reduction in the treated volume and in the marginal tissue can be obtained. The methods of calculation are outlined, and the potential for improved tumour response associated with accurate control of the volumes of tissue irradiated is discussed.

MATERIALS AND METHOD

The first step in planning for conformation therapy is to perform a computed tomography (CT) scan of the entire tumour length. The target volume and any critical organs are defined on every CT slice. The structures are either marked directly on hard copies of the scans, or outlined using the scanner's electronic digitiser. The information is transferred via magnetic tape to a Hewlett-Packard mini-computer which produces life- size reconstructions of the target volume in any plane. (Brace et al., 1981). At this stage the treatment tech- nique must be decided. Possible alternatives are adja- cent arcs or multiple longitudinal tracks. The target volume is divided along its length into sections which will allow each portion to be covered by a closely fitting field (Tate et al., 1985) (Fig. 2). A composite of the target shape in each of these portions is constructed from the CT data and is used for planning and for calculation of the tumour volumes. Each portion is plan- ned independently in a conventional manner using a treatment-planning computer. The cross-contributions from these plans are assessed and field weightings are balanced to achieve a homogeneous distribution along the length of the tumour (Davy, 1985). The aim has been to include the entire tumour within the 90% isodose line, although by attempting to spare adjacent struc- tures, this intention has not always been achieved. The treatments are carried out on a TEM MS90 cobalt unit under computer control and usually consist of multiple adjacent arcs. Arcs rather than fixed fields are used, as this method has been found to be the most time-effi- cient, combining machine positional movement with actual treatment. The final beam weightings for each portion are programmed into the treatment-planning computer and summated isodose distributions for each of these sections are produced. The volumes described are measured from these plans. A programme has been written for the treatment-planning computer which will measure the areas enclosed by an isodose line or by a tumour outline. The outlines are traced from the final treatment plan using the computer's electronic digitiser and the area is instantly displayed on the screen. Tumour areas may also be measured from hard copies of the CT scan provided a magnification factor is known.

268 CLINICAL RADIOLOGY

~ T U M O U R

100 =A T min=B

A

100% ISODOSE LINE, DEFINING HIGH DOSE REGION, RECEIVING DAILY PRESCRIBED DOSE

B

TUMOUR MINIMUM ISODOSE LINE, DEFINING TREATMENT VOLUME

C

NORMAL TISSUE CONTAINED WITHIN 100% ISODOSE LINE

Fig. 1 - Diagram of an isodose distribution, from CT data, to show the volumes measured.

LATERAL ~ ~ PROJECTION FUPERIOB

L J :"-

ANT-POST PROJECTION

j ~

/ f

J

• l RIGHT KIDNEY

D LEFT KIDNEY

. . . . LEFT RIGHT

Fig. 2 - Computer projections of a target volume, divided into segments for treatment planning.

Volumes are calculated by simple multiplication of these areas by the length of each section or by the width of the CT slice, and added together to obtain the total tumour and treatment volumes; hence the ratio between the two can be calculated.

As experience of conformation therapy has been gained, the planning technique has evolved into a basic routine with individual adjustments to fit unusually enlarged tumour or nodal masses. Eight patients with

advanced tumours with bulky local or regional disease of varying shape have been chosen as examples for com- parison of tumour and treatment volumes using conven- tional and conformation therapy. They had tumours of the oesophagus, bladder and cervix, and in one case, of the lymphoma. Each case was treated by conform- tion therapy and, in order to demonstrate a contrast, conventional plans were drawn up for each one accord- ing to the standard method used in this department. The

A METHOD OF IMPROVING THE TUMOUR TREATMENT VOLUME RATIO 269

oesophageal tumours, average length 18 cm, were treated using 3 or 4 adjacent arcs; conventionally these tumours are treated with an angled three-field volume with 8 MV X-rays. The pelvic tumours, all with involve- Oesophagus

ment of para-aortic nodes, had an average length of 1 32 cm. Using conformation therapy they were treated 2

by 6 or 7 adjacent arcs. In the calculations these plans 3 have been divided into a para-aortic strip, from the Para-aortic nodes

4 upper border of L1 to the upper border of L5, and a 5 pelvic volume. This division permits the results to be 6 more directly compared with the conventional plans. 7 The standard technique for treating para-aortic nodes is 8

Pelvis with anterior and posterior opposed fields matched to a 4 three field pelvic plan. Case 9, a lymphoma, was treated 5 by 7 adjacent arcs, bifurcated at the lower end (Fig. 2). 6 This plan was compared to a standard inverted 'Y' lower 7

8 mantle treated with 8 MV X-rays. Inverted 'Y'

9

Table 2 - Volume (cc) of normal tissue within daily prescribed dose isodose

Target and case no. Conformation Conventional

122 456 37 393

134 279

11 1813 33 1575

0 26 85.6 189 57.5 765

43 222 0 445 0 304 9 467 8 103

275 see text

RESULTS

Table 1 shows a comparison of the target volume for the three-dimensional and conventional plans for each tumour. A margin was allowed for microscopic exten- sion where appropriate. The conformation volume was assessed from CT images taken at 15 mm intervals, and the conventional volume was measured from simulator films or from a few CT slices in the usual way for two- dimensional planning. These methods result in the con- ventional volume being larger than the conformation volume. This is not because any attempt was made to reduce the target size, but is the result of variation in the tumour shape as seen on CT. The most striking differ- ences are seen when the shape shown on the central CT cut is not representative of the whole and there is taper- ing of the tumour towards the ends of the volume. Case 6 had very bulky para-aortic nodes with an average trans- verse diameter of 11 cm. In the other cases it can be seen that the volumes of the nodes for the three-dimensional and conventional plans are similar as the target is linear. The table also records the length of each target. In Cases 4 to 8, the para-aortic and pelvic volumes were treated in continuity.

Table 1 shows that more precise delineation of the target may result in a smaller volume requiring treat-

Table 1 - The target volume (cc) for conformation and conventional plans

Target and Length (cm) Conformation Conven- case no. tional

Oesophagus 1 22 580 1047 2 13 710 1662 3 19 5 237 812

Para-aortic nodes 4 14 499 548 5 17 407 474 6 16.5 772 899 7 16 513 705 8 13.5 443 600

Pelvis 4 12.5 951 1200 5 18 1568 2900 6 15 1421 2102 7 17.5 1371 1720 8 16 1414 2392

Inverted 'Y' 9 42 2230 3360

ment, but to take advantage of this finding it is necessary to tailor the radiation field to fit the target. Table 2 lists the volume of normal surrounding tissue treated to the prescribed daily dose, usually 200 cGy. This normal tissue has previously been described as matrix normal tissue. (Kramer, 1973). Its volume was calculated by subtraction of the tumour volume from the volume of the 100% prescribed dose isodose line (Fig. 1). There is a greater advantage from the use of the conformational technique when the treatment volume has an irregular tubular shape than when it is more cubic. The fit of the 100% isodose line to the tumour is also seen from these figures, For the standard plans, the conventional tumour volume has been used for the calculation. As the three-dimensional volume is more accurate, and smaller, it is likely that the figures in the conventional column represent an underestimate of the amount of normal tissue treated to a high dose by the conventional plan. Because of the width of the node masses in Cases 6 and 7 it was not possible to use a parallel opposed pair to treat the para-aortic regions in the conventional plan. A planned three-field volume was used in both cases but both gave renal doses at the upper limit of the tolerance dose. It was not possible to measure a conventional volume in Case 9, as full isodose distributions are not produced for mantle treatments. However , an estimate would be between 2500 and 3000 cc.

The proportion of the treatment volume occupied by turnout is represented in Fig. 3. A greater proport ion of the treatment volume is occupied by the tumour when conformation therapy is used, thus reducing the volume of normal tissue irradiated, with the greatest advantage being seen in the para-aortic region. The results show that in all but one case, conformation therapy allowed tailoring of the treatment volume to the shape of the target. In Cases 6 and 7 where the para-aortic nodes were treated using a three-field plan, an advantage is still seen when the target volume is divided into short segments for conformation therapy. In Case 6 the use of arcing conformation therapy resulted in a considerable reduction in the renal dose while not reducing the treat- ment volume. Where appropriate the volume treated by intracavitary 137Cs insertions has been included in the pelvic calculations.

The conformation treatment volume was calculated as a percentage of the treatment volume measured from

270 CLINICAL RADIOLOGY

8 0

o~O

.~ Go

._] 0

~' qO

k-

UA 20

[ - - I C o n I o r m a h o n p l a n I..".:.'.'.'t G o n v e n t t o n a l p l a n

I 2 I

OESOPHAGUS

7

P A R A - A O R T I C S

5

P E L V I S r , y , ,

Fig. 3 - The percentage of the t rea tment volume occupied by tumour for conformat ion and convent ional plans.

120

1 1 0

1 0 0

9 0

8 0

70

6 0

50

4 0

30

20

pI 2

O E S O P H A G U S

m

b

1 4 5 6

P A R A - A O R T I C S

m

8

P E L V I S

F

I i i y u

Fig. 4 - The conformat ion t r ea tmen t volume expressed as a percentage of the conventional t r ea tmen t volume.

the conventional plan (Fig. 4), and demonstrates that the use of conformation therapy results in a smaller irradiated volume. In general this confirms that a grea- ter reduction in volume is achieved when a cylindrical target is treated. Case 6 again shows the larger t reatment volume in the para-aortic region. However , when the treatment volumes of the para-aortic and pelvic regions are considered together , the conformation volume is 76% of the conventional t reatment volume.

The conformation therapy for Case 1, an oesophageal tumour, was multiple fixed fields. This plan resulted in a considerably smaller high-dose volume (Table 2) but little reduction in the t rea tment volume. It is likely that these reductions in t reated volume will also reduce the integral dose. Mayneord (1942) published a formula to estimate the integral dose for static rectangular fields. This formula was applied to the conformation data for Case 1 and a reduction of the integral dose by 50% was demonstrated.

DISCUSSION

Conformation therapy, by allowing an accurate fit of the radiation field to an irregularly shaped target, results

in a considerable reduction in a high-dose volume and also comes closer to Paterson's ideal of a thin rind of 'marginal tissue' around the high-dose volume.

Verification of the observation that tissue tolerance varies with the volume irradiated may be found in studies based on the N.S.D. equation. (Ellis, 1969). The N.S.D. formula, linking the number of fractions and the total treatment time, is based on the isoeffect curve for skin and expresses a tissue tolerance of 1800 rets for a volume of 1000 cc. It has been shown that for smaller volumes of around 350 cc, the maximum that might be used for a laryngeal tumour, the tolerance dose is 1900 rets (Shukovsky, 1970). For a standard treatment of the pelvis, with a volume of about 3000 cc and a dose of 5000 cGy in 5 weeks, a value of only 1520 rets is reached. Restriction of the high-dose volume can be at tempted by different radiation techniques. Interstitial therapy, by employing short treatment distances, has a limited effec- tive range but is suitable for only a small proport ion of tumours. Protons or heavy ions are also able to achieve a significant differential between tumour and the sur- rounding tissue dose by limiting the distribution of the Bragg peak to the tumour volume. It has been calculated that the use of protons to carry out total nodal irradia- tion would reduce the high-dose volume from 21000 to 8000 cc (Lowry, 1973). The two main disadvantages of heavy ion therapy are the inhomogeneity of dose dis- tribution and the cost of beam generation. The use of a computer-controlled megavoltage therapy machine to produce three-dimensional t reatment overcomes both these problems.

The relationship between tumour size and radiation dose required for control was well demonstrated by Fletcher (1973), when he showed, using the examples of the supraglottic larynx and the tonsillar fossa, an increased dose requirement of approximately 400 rets with an increase of tumour diameter of 5 cm. The response of tumours to radiation increases with an increase in dose (Kaplan, 1966), but quite small increases in dose at a crucial level may result in relatively large increases in response (Stewart and Jack- son, 1975).

On many occasions the total volume treated or the volume of normal tissue necessarily included in a con- ventional treatment plan will cause the total dose pre- scribed to be limited, giving priority to the first of Paterson's two observations. The definition of a tumour volume is often a compromise between that tissue with a high probability of involvement and that which it is feasible to include in a conventional treatment field. Indeed assumptions about the proposed treatment tech- nique may influence the shape in which a target is drawn (Goitein, 1982). The use of CT imaging and conforma- tion therapy permits the therapist to consider a tumour as multiple, three-dimensional segments, the length of which is limited only by the smallest possible collimator settings on the treatment unit. This method results in a consistent reduction in the target size and thus in the total tissue irradiated. In these examples the combined treatment volume of the para-aortic and pelvic regions for conformation therapy is only marginally greater than the treatment volume of the pelvis alone using a conven- tional plan. In no case is this more than 400 cc. In turn, the reduction in treatment volume leads to improved tolerance of radiation by the patient. The acute and late reactions to irradiation of the pelvis and para-aortic

A METHOD OF IMPROVING THE TUMOUR TREATMENT VOLUME RATIO 271

regions as a con t inuous vo lume will be repor ted

elsewhere.

CONCLUSION

The technique of con fo rma t ion the rapy reduces the total vo lume of tissue i rradiated. It is ant ic ipated that the use of a l inear accelerator for confo rma t ion therapy will produce a fur ther i m p r o v e m e n t in the reduct ion of t r ea tment volume. Cu r r en t studies will demons t ra te that by t reat ing these small vo lumes to a higher dose it is possible to separate the sigmoid curves of t u m o u r response and compl ica t ion rate (B loomer and He l lman , 1975), and thus achieve a significant increase in response rate with a min ima l rise in the compl ica t ion rate.

Pa terson (1963) c o m m e n t e d : ' O n e canno t have one ' s cake and eat it; one canno t have both subs tant ia l vo lume and top level dosage ' . Because confo rma t ion therapy provides a means of more closely restr ic t ing the high- dose vo lume to the t umour , it permi ts the prescr ip t ion of ' top level dosage ' to a vo lume which previous ly would have been considered too large.

Acknowledgements. This project is largely financed by the Cancer Research Campaign to whom we express grateful thanks. The treat- ment-planning computer used for the calculations is a Rad-8 which has been generously loaned for this developmental work by I.G.E.

REFERENCES

Bloomer, W. D. & Hellman, S. (1975). Normal tissue responses to radiation therapy. New England Medical Journal, 293, 80-83.

Brace, J. A., Davy, T. J., Skeggs, D. B. L. & Williams, H. S. (1981). Conformation therapy at the Royal Free Hospital: a progress report on the Tracking Cobalt Project. British Journal of Radio- logy, 54, 1068-1074.

Davy, T. J. (1985). Physical aspects of conformation therapy using computer controlled tracking units. In Progress in Medical Radia-

tion Physzcs, ed. Orton, C. G., Vol. 2, pp. 63-73. Plenum Press, New York.

Ellis, F. (1969). Dose, time and fractionation: a clinical hypothesis. Clinical Radiology, 20, 1-7.

Fletcher, G. H. (1973). Clinical dose-response curves of human epi- thelial tumours. British Journal of Radiology, 46, 1-12.

Goitein, M. (1982). Limitations of two-dimensional treatment plan- ning programmes. Medical Physics, 9, 580-586.

Gupta, M. K. & Ajayi, D. O. S. (1982). Dependence of tolerance dose on irradiated volume in radiotherapy, and determination of volume corrected time, dose and fractionation factors. Clinical Radiology, 33, 615-620.

International Commission on Radiation Units and Measurements (1978). Dose specification for reporting external beam therapy with photons and electrons. Report 29. Section 2,1,3.

Kaplan, H. S. (1966). Evidence for a tumoricidal dose level in the radiotherapy of Hodgkin's disease. Cancer Research, 26, 1221- 1224.

Kirk, J., Gray, W. M. & Watson, E. R. (1975). Cumulative radiation effect. Part IV. Normalisation of fractionated and continuous therapy: area and volume correction factors. Clinical Radiology, 26, 77-88.

Kramer, S. (1973). The biomedical problem in radiation therapy. In Particle Accelerators in Radiation Therapy, LA5180-C, pp. 6--10. Los Alamos Scientific Laboratory.

Lowry, W. S. (1973). The volume factor in radiotherapy. Proceedings of the XIIIth International Congress of Radiology. Madrid, Octo- ber, 1973.

Mayneord, W. V. (1942). The measurement of radiation for medical purposes. Nature, 149, 600-601.

Paterson, R. (1948). The Treatment of Malignant Disease by Radium and X-rays, p. 14. Edward Arnold, London.

Paterson, R. (1963). The Treatment of Malignant Disease by Radio- therapy, p. 23. Edward Arnold, London.

Schultheiss, T. E., Orton, C. G. & Peck, R. A. (1983). Models in radiotherapy: volume effects. Medical Physics, 10, 410-415.

Shukovsky, L. J. (1970). Dose, time, volume relationships in squamous cell carcinoma of the supra-glottic larynx. American Journal of Roentgenology, 108, 27-29.

Stewart, J. G. & Jackson, A. W. (1975). The steepness of the dose response curve both for tumour cure and normal tissue injury. Laryngoscope, 85, 1107-1111.

Tate, T., Brace, J. A., Davy, T. J., Morgan, H. M., Skeggs, D. B. L. & Tookman, A. J. (1985). Treatment of medulloblastoma using a computer-controlled tracking cobalt unit. Clinical Radiology, 36, 209-212.

25