Spinal Cord Doses in Palliative Lung Radiotherapy Schedules

Transcription

1 Journal of the Egyptian Nat. Cancer Inst., Vol. 8, No., June: -, 00 Spinal Cord Doses in Palliative Lung Radiotherapy Schedules HODA AL-BOOZ, FRCR FFRRCSI M.D.* and CAROL PARTON, Ph.D.** The Departments of Clinical Oncology* and Medical Physics**, Bristol Haematology and Oncology Centre, Bristol, United Kingdom. ABSTRACT Aim: We aim to check the safety of the standard palliative radiotherapy techniques by using the Linear quadratic model for a careful estimation of the doses received by the spinal cord, in all standard palliative lung radiotherapy fields and fractionation. Material and Methods: All patients surveyed at this prospective audit were treated with palliative chest radiotherapy for lung cancer over a period from January to June 005 by different clinical oncology specialists within the department. Radiotherapy field criteria were recorded and compared with the recommended limits of the MRC trial protocols for the dose and fractionation prescribed. Doses delivered to structures off the field central axis were estimated using a standard CT scan of the chest. Dose estimates were made using an SLPLAN planning system. As unexpected spinal cord toxicity has been reported after hypofractionated chest radiotherapy, a sagittal view was used to calculate the isodoses along the length of the spinal cord that could lie within the RT field. Equivalent dose estimates are made using the Linear- Quadratic Equivalent Dose formula (LQED). The relative radiation sensitivity of spinal cord for myelopathy (the a/b dose) cord has been estimated as a/b = Gy. Results: 7Gy in fraction and 39Gy in 3 fraction protocols have spinal cord equivalent doses (using the linear-quadratic model) that lie within the conventional safe limits of 50Gy in fractions for the 0% isodose. However when the dosimetry is modelled for a MV 0cm isocentric linnac in 3 dimensions, and altered separations and air space inhomogeneity are considered, the D-Max doses consistently fall above this limit on our 3 model patients. Conclusion: The 7Gy in fraction and 39Gy in 3 fraction protocol would risk spinal cord damage if the radiotherapist was unaware of the potential spinal cord Correspondence: Dr Hoda Al-Booz, Consultant in Clinical Oncology Bristol Haematology and Oncology Centre, hodabooz@doctors,org.uk doses. Alterative doses are suggested below 5.5 Gy/ fractions (7 days apart) would be most acceptable. Key Words: Spinal cord doses Palliative lung radiotherapy Spinal cord tolerance. INTRODUCTION Palliative radiotherapy is used for more than 80% of cases with lung cancer at diagnoses. The publication of successful United Kingdom Medical Research Council (MRC) randomized trials of hypo-fractionated accelerated radiotherapy has changed practice in the UK (Table ) [-]. The MRC studies have concluded that palliation is just as successful with or fractions of radiotherapy as with prolonged courses and toxicity (mainly radiation oesophagitis) is reduced [,]. There was a trend to survival benefit from longer regimens in patients with good performance status and no significant weight loss [3]. Where long fractionation schedules of to 5 fractions were common as palliation; short courses of or fractions now predominate. The MRC trials showed a good safety profile, but now that the doses and fractionation schedules have been incorporated into standard practice, unexpected spinal cord toxicity has been seen [5]. This observational study was planned to discover why this might have happened and to perform a risk assessment for our department. In this study, we have audited practice in a large UK oncology centre, to see if the patient selection criteria and radiotherapy techniques used in the trials have been maintained in routine clinical practice.

2 Spinal Cord Doses in Palliative Lung Radiotherapy Table (): Randomized trials of hypofractionated radiotherapy for the palliation of non-small cell lung cancer. Reference Doses Patients Palliation Survival Median survival (days) MRC, 99 [] 7Gy/# Vs 30Gy/# MRC, 99 [] Gy/# Vs 7Gy/# Poor PS (WHO -) 0 MRC, 99 [3] 7Gy/# Vs 3Gy/ & 39Gy/3# Good PS (WHO ) Quicker with 7Gy/ # Longer with 39Gy/3# yrs) 70 Rees et al, 997 [] 7Gy/# Vs.5Gy/5# 80 PATIENTS AND METHODS Patients surveyed at retrospective audit were all patients treated with palliative chest radiotherapy for lung cancer over the period from January to April 003. Radiotherapy field arrangements, field size constraints, doses, dose points for calculation, beam quality, fractionation and treatment times were recorded. These were compared with the recommended limits of the MRC trial protocols for the dose and fractionation prescribed. All patients had received conventional isocentric simulator planning in the supine position. Field borders were defined by simulator wire markers, and recorded to radiographic film or digital image. Simulator field borders represent the 50% isodose. Beam arrangements were all anterior-posterior opposed parallel isocentric pairs, using MV photon linear accelerators at 0 cm source to isocentre distance. The typical penumbra of the treatment machine with this beam arrangement is 7-8mm between the 50% and 90% isodose at the isocentre. In some cases, beam shapes were modified from standard square or rectangle using shadow tray mounted lead blocks or multileaf collimators. Doses were calculated at the field centre, mid-plane dose. No corrections for tissue inhomogeneity, nor correction of air, were performed. Plans that exceeded the dose or field size constraints of the MRC trial protocol appropriate for that dose and fractionation regimen were examined further. Doses delivered to structures off the field central axis were estimated using a standard CT scan of the chest. Dose estimates were made using an SLPLAN planning system. As unexpected spinal cord toxicity has been reported after hypofractionated chest radiotherapy, a sagittal view was used to calculate the isodoses along the length of the spinal cord that could lie within the RT field. Equivalent dose estimates were made using the Linear-Quadratic Equivalent Dose formula (LQED) []. The relative radiation sensitivity of spinal cord for myelopathy (the a/b dose) cord has been estimated as a/b = Gy [7]. RESULTS The audit identified sixty-seven patients treated. Five different radiotherapy doses and schedules were used. All had been used in arms of the MRC fractionation trials. Field sizes exceeding 0 cm were identified in 3 of 7 patients (%), and exceeding 00 cm in 5 patients (0.07%). This is summarized in Table (). Chest wall separations varied in all patients. Three patients were chosen to represent the smallest, medium and large separation to be able to represent the data. Spinal cord equivalent doses were estimated for three patient chest CT images. The saggital images are shown in Fig. () (image made up of 3 saggital scans). The three patients differed in the separation at the field centre of patient A= 9.7 cm. Patient B= 0cm and patient C= 3 cm. For each patient a typical palliative anterior-posterior field of 0 cm size was modelled. For each patient, the dose to the spinal cord was estimated at three points, D max, superior, and inferior field margin. Each estimate was tried on 3 different

3 Hoda Al-Booz & Carol Parton 3 chest CT images. The linear quadratic equivalent doses in Gy per fraction at α/ß = for each point are shown in Table (3) for a patient with separation 9.7cm, Table () for a patient with separation 0cm and Table (5) for a patient with separation 3cm. The planning system uses pixel-by-pixel correction for tissue density. This means that the effect of an air gap in the lungs and trachea is accounted for in dose estimates. Table (): Radiotherapy field sizes compared with radiotherapy dose and schedule delivered. Regimen Dose in Gy Overall treatment time Number treated Number who received Rt fields exceeding 0cm (%) Number who received Rt fields exceeding 50cm (%) Number who received Rt fields exceeding 00cm (%) (8%) (9.5%) 7 (30.3%) 3 (3%) (.5%) 5 (.5%) (%) (70%) 5 (50%) (0%) (0%) 7 3 (%) 8 (.8%) 5 (7.%) Total Patient Patient Patient 3 Fig. (): Saggital image of chest for patient, and 3. Doses to the spinal cord estimated at three points from an anterior 0cm field: Superior (A), D max (B) and inferior (C). Table (3): Estimated percentage doses, and spinal cord equivalent doses expressed as equivalent dose in Gy fractions at the three points (D max, Superior, and Inferior) for standard patient with 9.7cm separation. Dose in Gy/ Fraction / 7/ 0/5 3/ 39/ 3 Eq. Dose in Gy with Gy/ fractions. α/ß = (for CNS) D Max. % 0.5 Gy 59. Gy 3.5 Gy 53 Gy 57.5 Gy Superior point (7%) 39. Gy 57.7 Gy 35.3 Gy 5. Gy 55. Gy Inferior point (5%) 38. Gy 5.7 Gy 3. Gy 50.3 Gy 5. Gy

4 Spinal Cord Doses in Palliative Lung Radiotherapy Table (): Estimated percentage doses, and spinal cord equivalent doses expressed as equivalent dose in Gy fractions at the three points (D max, Superior, and Inferior) for standard patient with 0cm separation. Dose in Gy/ Fraction / 7/ 0/5 3/ 39/ 3 Eq. Dose in Gy with Gy/ fractions. α/ß = (for CNS) 3.Gy D Max Gy 55.9Gy 3.Gy 9.Gy 53.9Gy Superior point (98%) 35.8Gy 5.9Gy 3.3Gy.9Gy 50.9Gy Inferior point (97%) 35.5Gy 5.Gy 3Gy.Gy 50.Gy Table (5): Estimated percentage doses, and spinal cord equivalent doses expressed as equivalent dose in Gy fractions at the three points (D max, Superior, and Inferior) for standard patient 3 with 3 cm separation. Dose in Gy/ Fraction / 7/ 0/5 3/ 39/ 3 Eq. Dose in Gy with Gy/ fractions. α/ß = (for CNS) D Max 9.3% 0Gy 59Gy 3Gy 5.3Gy 5.8Gy Superior point (3%) 37.7Gy 55.Gy 3Gy 9.3Gy 53.5Gy Inferior point (%) 37.3Gy 55Gy 33.Gy 8.8Gy 53Gy DISCUSSION The UK MRC trials of altered fractionation radiotherapy to palliate lung cancer symptoms have been successful in changing practice [- ]. Although the aim of treatment is palliation, the median survival of those patients range between and 9 months. For these patients. Late toxicity becomes an important issue. The 7Gy in fraction and 39Gy in 3 fraction schedule has been reported to cause spinal cord toxicity [3]. Our analysis of the schedules used in current practice in a large regional cancer centre suggest that the 7Gy in fraction and 39Gy in 3 fraction protocol would risk spinal cord damage if the radiotherapist was unaware of the potential spinal cord doses. Both schedules have spinal cord equivalent doses (using the linear-quadratic model) that lie within the conventional safe limits of 50Gy in fractions for the 0% isodose [8]. However when the dosimetry is modelled for a MV 0cm isocentric linnac in 3 dimensions, and altered separations and air space inhomogeneity are considered, the D-Max doses consistently fall above this limit on our 3 model patients. The MRC protocol for the 39Gy in 3 fraction schedule advised that posterior spinal cord blocking should be used for the last fractions, to reduce spinal cord dose. No such advice has been given for the 7Gy in fraction protocol. Table (): Randomized trials of palliative radiotherapy regimens. Side effects for the different dose schedules. D: Dysphagia, M: Radiation Myelopathy. Reference Doses Patients Fraction 5 or 3 MRC, 99 7/ Vs 30/ D= 0% M=0.%@ yr.%@ yr D= 0% M=.5%@yr.5%@yr MRC, 99 / Vs 7/ Poor PS (WHO -) D= 3% D= 5% MRC, 99 7/ Vs 39/3 Good PS (WHO ) D= 50% D= 8% Rees et al, 997 7/ Vs 0/5 D= 50% M=0.8% D= 38% M=.3%

5 Hoda Al-Booz & Carol Parton 5 The risk of high D-max areas within the beams increases for a MV lineac with patient separation. The effect appears greater when the air gaps of lung and trachea are also considered as in our model patients. As a result of this risk assessment, we advise caution in the use of 7 Gy in fractions or 39 Gy in 3 fractions without special consideration. The risks appear greater in larger patient with greater separations, in larger fields or where the separations vary in the saggital plane. Many of our patients required large area palliative radiotherapy fields of greater than 0cm (%), 50cm (.8%) and 00cm (7.%). In these high risk situations, the Gy in fraction, or 0 Gy in 5 fraction schedules would be predicted to be safer treatment options. Palliative regimens are delivered in parallel opposing fields. Most departments would use MV Linear accelerators for such treatment. Fig. () shows the normalized relative-axis dose profiles from parallel-opposed beams for a xcm field at source to skin distance (SSD) of 0cm and for patient diameters of 5 to 30 cm in 5cm increments. It indicates that: RELATIVE DOSE PROFILES Co-0 Co-0 5 cm 8 Fig. (): Relative central-axis dose profiles as a function of X-ray energy ( 0 Co or,, and 8 MV) and patient thickness (5,0, and 30 cm). Because of the symmetry, only half of the profile is shown []. cm 30 cm 0 cm Co-0 8 Co-0 8 DEPTH FROM MIDLINE Doses at the surfaces are increased with an increase in separation. It is obvious that separations of more than 8 cm should not be treated with parallel-opposed fields if critical organs are present in the high dose region. Doses along the spinal cord are shown in Table () to be above the TD5/5 for the spinal cord (50 Gy) [8] in the commonly used palliative regimens (7 Gy/ fractions and 39 Gy/3 fractions). Although it is our policy to shield the cord posteriorly in the last fraction for the 39 Gy/3 fractions, it is often missed. Also, it appears that restricting to the acceptable field size (< 0cm ) is also not done. As the spinal cord s tolerance dose has been shown to be sensitive to the length of the cord [9-], such doses are not thought to be safe for use in good performance patients who s survival is thought to be longer than the majority of palliative patients. Radiation Myelopathy has been reported following the use of 7 Gy/ fractions and 39 Gy/3 fractions [5]. The annual risk was found to be small, 0.8% (95% CI 0.%-.%) and.3% (95% CI 0.%-.7%) respectively. The cumulative -year risks were % and.5%. This supports that these regimens should be used very cautiously. However, this complication was only seen 8 months after treatment and the median survival of this group of patients is around months. In conclusion, for those prescribed 39 G/3 fractions, 3 Gy in fractions may be more appropriate to be used although doses are still considered to be high. For the 7 Gy/ fractions (7 days apart), doses are also high along the spinal cord, Table (). REFERENCES - Medical Research Council Lung Cancer Working Party. Inoperable non-small cell lung cancer: a Medical Research Council randomized trial of palliative radiotherapy with two fractions or ten fractions. Br J Cancer. 99; 3: Medical Research Council Lung Cancer Working Party. A Medical Research Council (MRC) randomized trial of palliative radiotherapy with two fractions or a single fraction in patients with inoperable non-small cell lung cancer (NSCLC) and poor performance status. Br J Cancer. 99; 5: Macbeth FR, Bolger JJ, Hopwood P, Bleehen NM, Cartmell J, Girling DJ, et al. Randomized trial of palliative two fractions versus more intensive 3-

6 Spinal Cord Doses in Palliative Lung Radiotherapy fraction radiotherapy for patients with inoperable nonsmall cell lung cancer and good performance status. Medical Research council Lung Cancer Working Party. Clin Oncol. 99; 8: Rees GJG, Devrell CE, Barley VL, Newman HFV. Palliative radiotherapy: two versus five fractions. Clin Oncol. 997; 9: Macbeth FR, Wheldon TE, Girling DJ, Stephens RJ, Machin D, Bleehen NM, et al, Medical Research Council Lung Cancer Working Party, Radiation Myelopathy: estimates of risk in 8 patients in three randomized trials of palliative radiotherapy for nonsmall cell lung cancer. Clin Oncol. 99; 8: Taylor, J.M.G., H.R. Withers. Estimating the parameter in the -component model for cell survival from experimental quantal response data. Radiat. Res. 985; : Barton M. Tables of equivalent dose in Gy fractions: a simple application of the linear quadratic formula. Int J Radiat Oncol. Biol. Phys. 995 Jan5; 3 (): Emami B, Lyman J, Brown A, Coia L, Goitein M, Munzenrider JE, et al. Tolerance of normal tissue to therapeutic irradiation. Int J Radiation Oncol Biol Phys. 99; : Withers H.R., Taylor J.M.G., Maciejewski B. Treatment volume and tissue tolerance. Int. J. Radiat. Oncol. Biol. Phy. 988; : Withers, H.R., Taylor J.M.G. Volume effect in spinal cord. Letter to the Editor. Br. J. Radiol. 988; : Perez CA, Brady LW, eds. Principles and Parctice of radiation oncology, 3 rd ed. Philadelphia: Lippincott- Raven. 998; 8-30.