- Email: [email protected]
Single 4 Gy re-irradiation for painful bone metastasis following single fraction radiotherapy
Radiotherapy and Oncology 52 (1999) 123±127 www.elsevier.nl/locate/radonline
Single 4 Gy re-irradiation for painful bone metastasis following single fraction radiotherapy Branislav Jeremic a,*, Yuta Shibamoto b, Ivan Igrutinovic a a
b
Department of Oncology, University Hospital, Kragujevac, Yugoslavia Department of Oncology, Institute for Frontier Medical Sciences, Kyoto University, Japan Received 19 May 1999; received in revised form 19 July 1999; accepted 20 July 1999
Abstract Purpose: to investigate effectiveness of a single-fraction of 4 Gy given for re-treatment of bone metastasis after previous single-fraction radiotherapy (RT). Material and Methods: Of 135 patients retreated, 109 patients were retreated because of pain relapsing after 4 Gy (group I, n 34), 6 Gy (group II, n 39), or 8 Gy (group III, n 36), while 26 patients were re-irradiated after initial non-response (group I, n 12; group II, n 8; group III, n 6). Results: Of the 109 patients that were re-irradiated for pain relapse, 80 (74%) patients responded (complete response (CR) 31%;partial response PR) 42%). Among the 26 patients that initially did not respond, there were 12 (46%) responses. Patients with previous CR were more likely to achieve CR than were patients with previous PR (P 0.042). No such ®nding was observed for obtaining PR, which was achieved in 45% each of patients previously having either CR or PR (P 0.99). Patients with previous CR had similar chance to obtain either CR or PR (P 0.65), while previous PR in¯uenced subsequent response in the way of achieving more PRs than CRs (P 0.00054). Combined, these data showed that patients with initial CR were more likely to respond than those with previous PR (85% vs. 67%, P 0.037). There were no difference between the three initial treatment groups regarding the ef®ciency (CR or CR 1 PR) of second RT. Toxicity was low and only gastrointestinal. Conclusions: Single-fraction RT consisting of 4 Gy was effective and little toxic treatment that could be administered after previous singlefraction RT. q 1999 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Single-fraction radiotherapy; Re-irradiation; Bone metastasis; Pain
1. Introduction Several studies performed during the last three decades showed effectiveness of single-fraction radiotherapy (RT) in the treatment of metastatic bone pain [1,5±7,9,12,14±16]. It was also shown that response rates obtained that way are not different from those obtained with protracted regimens [2,13]. This was recently recon®rmed by the results of two large prospective randomised studies [4,11]. Nielsen et al., [11] used both visual analogue scales and a ®vepoint categorical scale to show that the degree of pain relief was not different between the two treatment groups regarding the duration of pain relief, the number of new painful metastatic sites and the need for re-irradiation or toxicity. No matter how patients with painful bone metastasis are
* Corresponding author. Department of Radiotherapy, University Hospital, Hoppe-Seyler-Strasse 3, D-72076 Tuebingen, Germany.
treated, many of them experience pain relapse. There is a scarce data on re-treatment in such patient population, including patients that were retreated after initial nonresponse. Mithall et al. [10] reported on effectiveness of re-irradiation in patients treated either with single-fraction RT up to 10 Gy or with more protracted regimens. No difference was found between single and multifraction retreatments. Other studies also indicated that single-fraction RT may be used ef®ciently after initial single-fraction regimens [6]. In our previous randomised study [8], we showed that 6 Gy and especially 8 Gy are superior to 4 Gy given in a single fraction, regarding both complete response (CR) and overall response rates, onset of pain relief and the duration of pain relief in responders. However, many patients were retreated during that study. In all cases of re-treatment, a single-fraction of 4 Gy was used. We herewith report on the results of re-treatment with a single fraction of 4 Gy after initial RT given as either 4, 6, or 8 Gy.
0167-8140/99/$ - see front matter q 1999 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0167-814 0(99)00108-5
124
B. Jeremic et al. / Radiotherapy and Oncology 52 (1999) 123±127
response rates, duration of response and the time to the ®rst occurrence of any pain relief between the treatment groups were evaluated by the chi-square or t-test. All these statistical analyses were carried out using a computer program HALBAU 4 (Gendaisuugakusha, Kyoto, Japan).
2. Material and methods A total of 135 patients were re-irradiated after previous RT with either 4, 6, or 8 Gy. Characteristics of all re-irradiated patients are given in Table 1. All patients were reirradiated under the same conditions as described in detail in the previous report [8]. Brie¯y, all patients were treated with 6±10 MV photons from linear accelerators. Direct ®elds were used to treat vertebral column and parallel-opposed ®elds were used to treat pelvis, hip, and long bones. Doses were speci®ed at 5-cm depth for spine ®elds and at middepth for parallel-opposed ®elds. Tumour dose of 4 Gy was used in all cases of re-treatment. The initial pre-treatment assessment was made by the patient on the day of treatment planning or performing the re-irradiation. The same pain chart, originally developed at the Royal Marsden Hospital, Sutton, UK [13], was used to assess the response to re-irradiation and the same criteria of response was used as during our prospective study reported previously [8]. In this validated, patient-based method, a 4-point categorical pain scale (none, mild, moderate, severe) was used, with complete response (CR) de®ned as a complete disappearance of pain and a partial response (PR) de®ned as an improvement in pain score by at least one category, with pain still existing. Toxicity was also scored using the RTOG/EORTC toxicity criteria [3]. The differences in patient characteristics,
3. Results One-hundred and thirty-®ve patients were re-irradiated with 4 Gy given in a single fraction. Of these, 26 patients were re-irradiated after previous non-response, while 109 patients were re-irradiated due to pain relapse after response to previous single-fraction RT of either 4, 6, or 8 Gy (Table 1). There were no differences between these subgroups of patients regarding age, sex, primary tumour type or site. Patients previously experiencing response were re-irradiated after longer time intervals (from initial RT) than those re-irradiated for initial non-response. Analysis of response achieved with second RT versus response to ®rst RT is given in Table 2. Patients with previous CR were more likely to achieve CR than were those with partial response (PR) able to do so (P 0.042). No such ®nding was observed with regard to PR, which was achieved in 45% each of the patients previously having either CR or PR (P 0.99). Patients with previous
Table 1 Patient characteristics according to previous response a Characteristic
No. patients re-irradiated Age (at ®rst RT) Sex Time of re-irradiation (weeks) Interval from pain recurrence to re-irradiation (weeks) Primary tumour
Site
Treatment group
a
Previous response NR
PR
CR
CR 1 PR
Range Median M/F Range Median Range Median
26 40±69 61 12/14 9±25 14 ± ±
69 32±70 59 29/40 14±68 20 0±32 4
40 37±70 54 11/29 16±149 47 0±45 2
109 32±70 57 40/69 14±149 34 0±45 3
Breast Prostate Lung Myeloma Kidney Rectum Other Spine C Th L±S Pelvis/Hip Femur Humerus I II III
6 8 6 2 1 3 0 0 6 10 9 1 0 12 8 6
29 16 10 5 2 3 4 2 15 17 31 2 2 23 26 20
14 0 12 4 1 4 5 1 6 14 17 1 1 11 13 16
43 16 22 9 3 7 9 3 21 31 48 3 3 34 39 36
NR no response; PR partial response; CR complete response; RT radiation therapy; C cervical; Th thoracic; L±S lumbosacral.
B. Jeremic et al. / Radiotherapy and Oncology 52 (1999) 123±127
125
Table 2 Response to treatment according to initial response Response to reirradiation
Initial response PR
CR
CR 1 PR
CR
15/69 (22%)
16/40 (40%)
31/109 (28%)
PR
31/69 (45%)
18/40 (45%)
49/109 (45%)
CR 1 PR
46/69 (67%)
34/40 (85%)
80/109 (73%)
Onset of response in responders
Range Median Mean ^ SD
1±10 2 4^2
0±9 2 3^2
0±10 2 3^2
Duration of response in responders (weeks)
Range Median Mean ^ SD
4±36 19 19 ^ 8
8±60 25 26 ^ 13
4±60 20 22 ^ 11
Survival from re-irradiation (weeks)
Range Median Mean ^ SD
5±46 23 24 ^ 10
2±69 26 27 ^ 15
2±69 24 25 ^ 12
CR had a similar chance to obtain either CR or PR (P 0.65), while experiencing previously PR made those patients more likely to achieve PR than CR (P 0.00054). When put together, these data showed that patients with previous CR were more likely to respond than patients with previous PR (85% vs. 67%, P 0.037). The onset of response to second RT in responders was similar between patients with initial CR and those with initial PR (P 0.34), but the duration of response was longer in initial complete responders than initial partial responders (P
0.0068). Survival from re-irradiation was not different between patients having different initial response (CR or PR) (P 0.27). When response was analysed according to the initial treatment (i.e. dose) group in initial responders, there was no difference among the three treatment groups regarding either CR or CR 1 PR rates (Table 3): when irradiated with a single fraction of 4 Gy, patients that experienced initial response (either CR or PR) in all three groups had similar both CR and CR 1 PR rates (P . 0:05). However, the onset
Table 3 Response status according to treatment group in previous (initial) responders a Groups I
Total II
III
CR
8/34 (24%)
11/39 (28%)
12/36 (33%)
31/109 (28%)
CR 1 PR
22/34 (65%)
30/39 (77%)
28/36 (78%)
80/109 (73%)
Onset of response in responders (weeks)
Range Median Mean ^ SD
0±10 5 5^3
1±6 3 3^2
1±6 2 2^1
0±10 2 3^2
Duration of response
Range Median Mean ^ SD
4±60 15 19 ^ 13
10±42 32 29 ^ 10
8±28 17 17 ^ 6
4±60 20 22 ^ 11
Survival from re-irradiation (weeks)
Range Median Mean ^ SD
5±69 27 26 ^ 15
12±48 32 32 ^ 10
2±30 18 18 ^ 7
2±69 24 25 ^ 12
a
a group I vs. II, b group II vs. III, c group I vs. III.
126
B. Jeremic et al. / Radiotherapy and Oncology 52 (1999) 123±127
of response in responders in these three treatment groups was different. It was shortest in group III (previously treated with 8 Gy) and longest in group I (previously treated with 4 Gy) (group I vs. II, P 0.05; II vs. III, P 0.011; I vs. III, P 0.0016, respectively). Also, the duration of response to second (4 Gy) RT was signi®cantly longer in group II than in either group I (P 0.02) or group III (P , 0:001), with no difference between groups I and III (P 0.48). When survival from re-irradiation in previous responders in all three treatment groups was taken into account, patients in group II re-irradiated with 4 Gy lived longer than either those in group I (P 0.046) or group III (P , 0:001), while responding patients previously irradiated with 4 Gy (group I) lived longer than their counterparts in group III (treated previously with 8 Gy) (P 0.0048). Owing to many deaths in ®rst several months after the ®rst single-fraction RT, a small number of patients initially non-responding were re-irradiated (27, 28, and 25% of all re-irradiated patients in the three treatment groups, respectively). Although this subset of patients was not a primary focus of the current study, in all treatment groups similar responses were observed, ranging 33±67% (total, 46%). When analysed per total number of patients per group, the same percentage (4%) of responses were seen in the three treatment groups. Therefore, percentage of the `additional' responders contributing to the overall (cumulative) success rate (when combined to the initial responders in all treatment groups), could not substantially change the overall picture, the 6 Gy and 8 Gy groups still having signi®cant advantage over the 4 Gy group, now in terms of overall success rate. There were 17/26 (65%) pain recurrences in responders to second RT in group I, 18/34 (53%) in group II and 18/32 (56%) in group III (P 0.62). There was no difference in the relapse rate after the re-irradiation between the patients that responded to ®rst irradiation and those who did not respond to ®rst irradiation (data not shown). Pathological fractures were seen in 1/46 (2%), 1/47 (2%), and 1/42 (2%) patients in the three treatment groups, respectively (P 1.0), while spinal cord compression was seen in 1/46 (2%), 1/47 (2%), and 1/42 (2%) (P 1.0) patients in the three treatment groups, respectively. Toxicity of the second RT was mild to moderate (grade 1 or 2). It was gastrointestinal, consisting of nausea and vomiting and diarrhoea. Nausea and vomiting of grade 1 or 2 was seen in 8/46 (17%), 9/47 (19%), and 8/42 (19%) of patients in the three treatment groups, respectively (P 0.97). There was also no difference between the three treatment groups regarding grade 1 or 2 diarrhoea, which was observed in 5/46 (11%), 6/47 (13%), and 5/42 (12%) patients, respectively (P 0.96). No grade . 3 acute toxicity was observed during the follow-up post-second RT. As it was not possible to speci®cally address the issue of haematological toxicity after the ®rst RT due to majority of patients receiving previously some form of chemotherapy, it was not possible to address this issue in cases of re-treatments, too.
4. Discussion Regardless of the type of RT used to ameliorate their pain, a proportion of patients with bone metastasis experience pain relapse. Although some of them are doomed to die relatively soon after the ®rst RT, some may live longer. While some patients continue to take analgesics, additional RT may be administered as well as in those that did not respond to previous RT, provided that the ®rst RT did not cause signi®cant toxicity and that it is anticipated that reirradiation should be effective and not too toxic. In our study, patients that relapsed after initial response represented the majority of patients re-irradiated. The reirradiation with 4 Gy was little toxic in this subgroup of patients. Both CR (24±33%) and overall response (65± 78%) rates observed in the three treatment groups were rewarding and similar among the three treatment groups being quite similar to that observed after the ®rst RT. On the other side, only 26 out of 96 (27%) previously nonresponding patients were re-irradiated, the overall response rate in this subgroup of patients being 46% and toxicity similar to that experienced by previous responders (data not shown). This may help assure the community of radiation oncologists that using ®rst unsuccessful single-fraction RT does not preclude its repeating. It seems that initial response status in¯uenced subsequent response, especially in the CR group, because these patients were more likely to achieve either CR (P 0.042) or CR 1 PR (P 0.037) than those with previous PR. No response to second RT was seen in 14/26 (54%) patients with initial non-response. Why majority of patients are more likely to experience the same type of response after second RT, remains to be investigated in the future, especially since no effect of age, sex, histology, primary tumour location or metastatic site could be demonstrated in this study. A few studies using initial single-fraction RT reported on sporadic re-treatments with either single- or multi-fraction regimens. Price et al. [14] reported on 7 patients that did not respond to initial 4 Gy single regimen; 4 of which were reirradiated with a single fraction of 8 Gy, while 3 patients were re-irradiated with a multifraction regimen. Second RT did not provide signi®cant pain relief in any of these 7 patients. Contrary to that, in 4 patients that relapsed after an initial response to 4 Gy, additional 4 Gy enabled second response in all 4 patients. Although in a small number of patients it shows that response to second RT may be previous-response-dependent and that even `low` single-fraction RT (such as 4 Gy) may be of bene®t in previous responders. In contrast to that, Cole [2] reported that as many as 50% of re-irradiated patients further required analgesics. In a study of Hoskin et al. [6] that evaluated optimal singlefraction RT (4 Gy vs. 8 Gy), 71% of evalauble patients responded to re-treatment of 4 Gy, while 44% responded to re-treatment of 8 Gy. This difference was not a signi®cant one, and did not change overall results, 8 Gy being still more effective during the 12-week assessment period. Similar
B. Jeremic et al. / Radiotherapy and Oncology 52 (1999) 123±127
results were obtained during the current study when a single fraction of 4 Gy was used (RR 74%). Furthermore, when `total' (cumulative) responses during the current study period were taken into account (when responses to second 4 Gy in previous non-responders were added to initial responses seen after ®rst 4 Gy), initial 6 Gy or 8 Gy (without counting `additional` 4% response from initial non-responders in both of these two groups) still had higher overall response rates than that achieved (63% vs. 73% vs. 78% for the groups I±III, respectively). Mithal et al. [10] reported on 57 sites retreated once and 8 retreated twice out of 280 individual treatment sites irradiated initially. A total of 17 out of 23 (74%) patients responded (CR 1 PR) to second RT that used a number of single-fraction regimens (an identical ®nding to our own), which was not signi®cantly inferior to 91% obtained with more protracted regimens. Furthermore, the third (second re-irradiation) RT given again as a single fraction, was successful in 80% of patients. As in the current study, previous CR carried an increased likelihood for achieving second CR, with response being shorter in patients with initial PR than in patients with initial CR. Uppelschoten et al. [15] recently reported on a study with a similar policy to our own that considered pain relapse or insuf®cient palliation for re-irradiation. After long intervals from previous RT, re-irradiation with 6 Gy was given to 18 such patients to obtain a reduction of pain in 13 of them (72%). In conclusion, results of this study con®rmed previous observations that more than 70% of re-irradiated patients experience pain relief after second single-fraction RT. Response was also seen in 46% of patients that were re-irradiated after initial unsuccessful RT. Previous response status seems to govern the outcome of second RT favouring patients with initial CR, that were more likely to experience response after re-irradiation. With low toxicity, similar to that observed during the ®rst RT and the incidence of pathological fractures and spinal cord compressions similar between re-irradiated patients and those undergoing only one single-fraction RT, these may be important ®ndings that may help select patients suitable for different approaches when administering single-fraction RT in the future. Acknowledgements This study was supported in part by the Grant-in-Aid for Scienti®c Research (B) from the Japanese Ministry of
127
Education, Science and Culture (11470190, 11877152, 10557087).
References [1] Barak F, Werner A, Walach N, Horn Y. The palliative ef®cacy of a single high dose of radiation in treatment of symptomatic osseous metastases. Int. J. Radiat. Oncol. Biol. Phys. 1987;13:1233±1235. [2] Cole DJ. A randomised trial of a single treatment versus conventional fractionation in the palliative radiotherapy of painful bone metastases. Clin. Oncol. 1989;1:59±62. [3] Cox JD, Stetz J, Pajak TF. Toxicity criteria of the Radiation Therapy Oncology Group (RTOG) and the European Organisation for the Treatment and Research of Cancer (EORTC). Int. J. Radiat. Oncol. Biol. Phys. 1995;31:1341±1346. [4] Gaze MN, Kelly CG, Kerr GR, et al. Pain relief and quality of life following radiotherapy for bone metastases: a randomised trial of two fractionation schedules. Radiother. Oncol. 1997;45:109±116. [5] Hendrickson FR, Shehata WM, Kirchner AB. Radiation therapy for osseous metastasis. Int. J. Radiat. Oncol. Biol. Phys 1976;1:275± 278. [6] Hoskin PJ, Price P, Easton D, et al. A prospective randomised trial of 4 or 8 Gy single doses in the treatment of metastatic bone pain. Radiother. Oncol. 1992;23:74±78. [7] Jensen NH, Roesdahl K. Single-dose irradiation of bone metastases. Acta Radiol. Ther. Phys. Biol. 1976;15:337±339. [8] Jeremic B, Shibamoto Y, et al. Acimovic. Lj.. A randomised trial of three single-dose radiation therapy regimens in the treatment of metastatic bone pain. Int. J. Radiat. Oncol. Biol. Phys. 1998;42:161± 167. [9] Karstens JH, Schnabel B, Amman J. Management of metastatic bone pain: preliminary results with single fraction (4 Gy) radiotherapy. Onkologie 1989;12:41±42. [10] Mithal NP, Needham PR, Hoskin PJ. Re-treatment with radiotherapy for painful bone metastases. Int. J. Radiat. Oncol. Biol. Phys. 1994;29:1011±1014. [11] Nielsen OS, Bentzen SM, Sandberg E, Gadeberg CC, Timothy AR. Randomised trial of single dose versus fractionated palliative radiotherapy of bone metastases. Radiother. Oncol. 1998;47:233±240. [12] Penn CRM. Single dose and fractionated palliative irradiation for osseous metastases. Clin. Radiol. 1976;27:405±408. [13] Price P, Hoskin PJ, Easton D, Austin D, Palmer SG, Yarnold JR. Prospective randomised trial of single and multifraction radiotherapy schedules in the treatment of painful bone metastases. Radiother. Oncol. 1986;6:247±255. [14] Price P, Hoskin PJ, Easton D, Austin D, Palmer S, Yarnold JR. Low dose single fraction radiotherapy in the treatment of metastatic bone pain: A pilot study. Radiother. Oncol. 1988;12:297±300. [15] Uppelschoten JM, Wanders SL, de Jong JMA. Single-dose radiotherapy (6 Gy) palliation in painful bone metastases. Radiother. Oncol. 1995;36:198±202. [16] Vargha ZO, Glicksman AS, Boland J. Single-dose radiation therapy in the palliation of metastatic disease. Radiology 1969;93:1181± 1184.
Single 4 Gy re-irradiation for painful bone metastasis following single fraction radiotherapy Branislav Jeremic a,*, Yuta Shibamoto b, Ivan Igrutinovic a a
b
Department of Oncology, University Hospital, Kragujevac, Yugoslavia Department of Oncology, Institute for Frontier Medical Sciences, Kyoto University, Japan Received 19 May 1999; received in revised form 19 July 1999; accepted 20 July 1999
Abstract Purpose: to investigate effectiveness of a single-fraction of 4 Gy given for re-treatment of bone metastasis after previous single-fraction radiotherapy (RT). Material and Methods: Of 135 patients retreated, 109 patients were retreated because of pain relapsing after 4 Gy (group I, n 34), 6 Gy (group II, n 39), or 8 Gy (group III, n 36), while 26 patients were re-irradiated after initial non-response (group I, n 12; group II, n 8; group III, n 6). Results: Of the 109 patients that were re-irradiated for pain relapse, 80 (74%) patients responded (complete response (CR) 31%;partial response PR) 42%). Among the 26 patients that initially did not respond, there were 12 (46%) responses. Patients with previous CR were more likely to achieve CR than were patients with previous PR (P 0.042). No such ®nding was observed for obtaining PR, which was achieved in 45% each of patients previously having either CR or PR (P 0.99). Patients with previous CR had similar chance to obtain either CR or PR (P 0.65), while previous PR in¯uenced subsequent response in the way of achieving more PRs than CRs (P 0.00054). Combined, these data showed that patients with initial CR were more likely to respond than those with previous PR (85% vs. 67%, P 0.037). There were no difference between the three initial treatment groups regarding the ef®ciency (CR or CR 1 PR) of second RT. Toxicity was low and only gastrointestinal. Conclusions: Single-fraction RT consisting of 4 Gy was effective and little toxic treatment that could be administered after previous singlefraction RT. q 1999 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Single-fraction radiotherapy; Re-irradiation; Bone metastasis; Pain
1. Introduction Several studies performed during the last three decades showed effectiveness of single-fraction radiotherapy (RT) in the treatment of metastatic bone pain [1,5±7,9,12,14±16]. It was also shown that response rates obtained that way are not different from those obtained with protracted regimens [2,13]. This was recently recon®rmed by the results of two large prospective randomised studies [4,11]. Nielsen et al., [11] used both visual analogue scales and a ®vepoint categorical scale to show that the degree of pain relief was not different between the two treatment groups regarding the duration of pain relief, the number of new painful metastatic sites and the need for re-irradiation or toxicity. No matter how patients with painful bone metastasis are
* Corresponding author. Department of Radiotherapy, University Hospital, Hoppe-Seyler-Strasse 3, D-72076 Tuebingen, Germany.
treated, many of them experience pain relapse. There is a scarce data on re-treatment in such patient population, including patients that were retreated after initial nonresponse. Mithall et al. [10] reported on effectiveness of re-irradiation in patients treated either with single-fraction RT up to 10 Gy or with more protracted regimens. No difference was found between single and multifraction retreatments. Other studies also indicated that single-fraction RT may be used ef®ciently after initial single-fraction regimens [6]. In our previous randomised study [8], we showed that 6 Gy and especially 8 Gy are superior to 4 Gy given in a single fraction, regarding both complete response (CR) and overall response rates, onset of pain relief and the duration of pain relief in responders. However, many patients were retreated during that study. In all cases of re-treatment, a single-fraction of 4 Gy was used. We herewith report on the results of re-treatment with a single fraction of 4 Gy after initial RT given as either 4, 6, or 8 Gy.
0167-8140/99/$ - see front matter q 1999 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0167-814 0(99)00108-5
124
B. Jeremic et al. / Radiotherapy and Oncology 52 (1999) 123±127
response rates, duration of response and the time to the ®rst occurrence of any pain relief between the treatment groups were evaluated by the chi-square or t-test. All these statistical analyses were carried out using a computer program HALBAU 4 (Gendaisuugakusha, Kyoto, Japan).
2. Material and methods A total of 135 patients were re-irradiated after previous RT with either 4, 6, or 8 Gy. Characteristics of all re-irradiated patients are given in Table 1. All patients were reirradiated under the same conditions as described in detail in the previous report [8]. Brie¯y, all patients were treated with 6±10 MV photons from linear accelerators. Direct ®elds were used to treat vertebral column and parallel-opposed ®elds were used to treat pelvis, hip, and long bones. Doses were speci®ed at 5-cm depth for spine ®elds and at middepth for parallel-opposed ®elds. Tumour dose of 4 Gy was used in all cases of re-treatment. The initial pre-treatment assessment was made by the patient on the day of treatment planning or performing the re-irradiation. The same pain chart, originally developed at the Royal Marsden Hospital, Sutton, UK [13], was used to assess the response to re-irradiation and the same criteria of response was used as during our prospective study reported previously [8]. In this validated, patient-based method, a 4-point categorical pain scale (none, mild, moderate, severe) was used, with complete response (CR) de®ned as a complete disappearance of pain and a partial response (PR) de®ned as an improvement in pain score by at least one category, with pain still existing. Toxicity was also scored using the RTOG/EORTC toxicity criteria [3]. The differences in patient characteristics,
3. Results One-hundred and thirty-®ve patients were re-irradiated with 4 Gy given in a single fraction. Of these, 26 patients were re-irradiated after previous non-response, while 109 patients were re-irradiated due to pain relapse after response to previous single-fraction RT of either 4, 6, or 8 Gy (Table 1). There were no differences between these subgroups of patients regarding age, sex, primary tumour type or site. Patients previously experiencing response were re-irradiated after longer time intervals (from initial RT) than those re-irradiated for initial non-response. Analysis of response achieved with second RT versus response to ®rst RT is given in Table 2. Patients with previous CR were more likely to achieve CR than were those with partial response (PR) able to do so (P 0.042). No such ®nding was observed with regard to PR, which was achieved in 45% each of the patients previously having either CR or PR (P 0.99). Patients with previous
Table 1 Patient characteristics according to previous response a Characteristic
No. patients re-irradiated Age (at ®rst RT) Sex Time of re-irradiation (weeks) Interval from pain recurrence to re-irradiation (weeks) Primary tumour
Site
Treatment group
a
Previous response NR
PR
CR
CR 1 PR
Range Median M/F Range Median Range Median
26 40±69 61 12/14 9±25 14 ± ±
69 32±70 59 29/40 14±68 20 0±32 4
40 37±70 54 11/29 16±149 47 0±45 2
109 32±70 57 40/69 14±149 34 0±45 3
Breast Prostate Lung Myeloma Kidney Rectum Other Spine C Th L±S Pelvis/Hip Femur Humerus I II III
6 8 6 2 1 3 0 0 6 10 9 1 0 12 8 6
29 16 10 5 2 3 4 2 15 17 31 2 2 23 26 20
14 0 12 4 1 4 5 1 6 14 17 1 1 11 13 16
43 16 22 9 3 7 9 3 21 31 48 3 3 34 39 36
NR no response; PR partial response; CR complete response; RT radiation therapy; C cervical; Th thoracic; L±S lumbosacral.
B. Jeremic et al. / Radiotherapy and Oncology 52 (1999) 123±127
125
Table 2 Response to treatment according to initial response Response to reirradiation
Initial response PR
CR
CR 1 PR
CR
15/69 (22%)
16/40 (40%)
31/109 (28%)
PR
31/69 (45%)
18/40 (45%)
49/109 (45%)
CR 1 PR
46/69 (67%)
34/40 (85%)
80/109 (73%)
Onset of response in responders
Range Median Mean ^ SD
1±10 2 4^2
0±9 2 3^2
0±10 2 3^2
Duration of response in responders (weeks)
Range Median Mean ^ SD
4±36 19 19 ^ 8
8±60 25 26 ^ 13
4±60 20 22 ^ 11
Survival from re-irradiation (weeks)
Range Median Mean ^ SD
5±46 23 24 ^ 10
2±69 26 27 ^ 15
2±69 24 25 ^ 12
CR had a similar chance to obtain either CR or PR (P 0.65), while experiencing previously PR made those patients more likely to achieve PR than CR (P 0.00054). When put together, these data showed that patients with previous CR were more likely to respond than patients with previous PR (85% vs. 67%, P 0.037). The onset of response to second RT in responders was similar between patients with initial CR and those with initial PR (P 0.34), but the duration of response was longer in initial complete responders than initial partial responders (P
0.0068). Survival from re-irradiation was not different between patients having different initial response (CR or PR) (P 0.27). When response was analysed according to the initial treatment (i.e. dose) group in initial responders, there was no difference among the three treatment groups regarding either CR or CR 1 PR rates (Table 3): when irradiated with a single fraction of 4 Gy, patients that experienced initial response (either CR or PR) in all three groups had similar both CR and CR 1 PR rates (P . 0:05). However, the onset
Table 3 Response status according to treatment group in previous (initial) responders a Groups I
Total II
III
CR
8/34 (24%)
11/39 (28%)
12/36 (33%)
31/109 (28%)
CR 1 PR
22/34 (65%)
30/39 (77%)
28/36 (78%)
80/109 (73%)
Onset of response in responders (weeks)
Range Median Mean ^ SD
0±10 5 5^3
1±6 3 3^2
1±6 2 2^1
0±10 2 3^2
Duration of response
Range Median Mean ^ SD
4±60 15 19 ^ 13
10±42 32 29 ^ 10
8±28 17 17 ^ 6
4±60 20 22 ^ 11
Survival from re-irradiation (weeks)
Range Median Mean ^ SD
5±69 27 26 ^ 15
12±48 32 32 ^ 10
2±30 18 18 ^ 7
2±69 24 25 ^ 12
a
a group I vs. II, b group II vs. III, c group I vs. III.
126
B. Jeremic et al. / Radiotherapy and Oncology 52 (1999) 123±127
of response in responders in these three treatment groups was different. It was shortest in group III (previously treated with 8 Gy) and longest in group I (previously treated with 4 Gy) (group I vs. II, P 0.05; II vs. III, P 0.011; I vs. III, P 0.0016, respectively). Also, the duration of response to second (4 Gy) RT was signi®cantly longer in group II than in either group I (P 0.02) or group III (P , 0:001), with no difference between groups I and III (P 0.48). When survival from re-irradiation in previous responders in all three treatment groups was taken into account, patients in group II re-irradiated with 4 Gy lived longer than either those in group I (P 0.046) or group III (P , 0:001), while responding patients previously irradiated with 4 Gy (group I) lived longer than their counterparts in group III (treated previously with 8 Gy) (P 0.0048). Owing to many deaths in ®rst several months after the ®rst single-fraction RT, a small number of patients initially non-responding were re-irradiated (27, 28, and 25% of all re-irradiated patients in the three treatment groups, respectively). Although this subset of patients was not a primary focus of the current study, in all treatment groups similar responses were observed, ranging 33±67% (total, 46%). When analysed per total number of patients per group, the same percentage (4%) of responses were seen in the three treatment groups. Therefore, percentage of the `additional' responders contributing to the overall (cumulative) success rate (when combined to the initial responders in all treatment groups), could not substantially change the overall picture, the 6 Gy and 8 Gy groups still having signi®cant advantage over the 4 Gy group, now in terms of overall success rate. There were 17/26 (65%) pain recurrences in responders to second RT in group I, 18/34 (53%) in group II and 18/32 (56%) in group III (P 0.62). There was no difference in the relapse rate after the re-irradiation between the patients that responded to ®rst irradiation and those who did not respond to ®rst irradiation (data not shown). Pathological fractures were seen in 1/46 (2%), 1/47 (2%), and 1/42 (2%) patients in the three treatment groups, respectively (P 1.0), while spinal cord compression was seen in 1/46 (2%), 1/47 (2%), and 1/42 (2%) (P 1.0) patients in the three treatment groups, respectively. Toxicity of the second RT was mild to moderate (grade 1 or 2). It was gastrointestinal, consisting of nausea and vomiting and diarrhoea. Nausea and vomiting of grade 1 or 2 was seen in 8/46 (17%), 9/47 (19%), and 8/42 (19%) of patients in the three treatment groups, respectively (P 0.97). There was also no difference between the three treatment groups regarding grade 1 or 2 diarrhoea, which was observed in 5/46 (11%), 6/47 (13%), and 5/42 (12%) patients, respectively (P 0.96). No grade . 3 acute toxicity was observed during the follow-up post-second RT. As it was not possible to speci®cally address the issue of haematological toxicity after the ®rst RT due to majority of patients receiving previously some form of chemotherapy, it was not possible to address this issue in cases of re-treatments, too.
4. Discussion Regardless of the type of RT used to ameliorate their pain, a proportion of patients with bone metastasis experience pain relapse. Although some of them are doomed to die relatively soon after the ®rst RT, some may live longer. While some patients continue to take analgesics, additional RT may be administered as well as in those that did not respond to previous RT, provided that the ®rst RT did not cause signi®cant toxicity and that it is anticipated that reirradiation should be effective and not too toxic. In our study, patients that relapsed after initial response represented the majority of patients re-irradiated. The reirradiation with 4 Gy was little toxic in this subgroup of patients. Both CR (24±33%) and overall response (65± 78%) rates observed in the three treatment groups were rewarding and similar among the three treatment groups being quite similar to that observed after the ®rst RT. On the other side, only 26 out of 96 (27%) previously nonresponding patients were re-irradiated, the overall response rate in this subgroup of patients being 46% and toxicity similar to that experienced by previous responders (data not shown). This may help assure the community of radiation oncologists that using ®rst unsuccessful single-fraction RT does not preclude its repeating. It seems that initial response status in¯uenced subsequent response, especially in the CR group, because these patients were more likely to achieve either CR (P 0.042) or CR 1 PR (P 0.037) than those with previous PR. No response to second RT was seen in 14/26 (54%) patients with initial non-response. Why majority of patients are more likely to experience the same type of response after second RT, remains to be investigated in the future, especially since no effect of age, sex, histology, primary tumour location or metastatic site could be demonstrated in this study. A few studies using initial single-fraction RT reported on sporadic re-treatments with either single- or multi-fraction regimens. Price et al. [14] reported on 7 patients that did not respond to initial 4 Gy single regimen; 4 of which were reirradiated with a single fraction of 8 Gy, while 3 patients were re-irradiated with a multifraction regimen. Second RT did not provide signi®cant pain relief in any of these 7 patients. Contrary to that, in 4 patients that relapsed after an initial response to 4 Gy, additional 4 Gy enabled second response in all 4 patients. Although in a small number of patients it shows that response to second RT may be previous-response-dependent and that even `low` single-fraction RT (such as 4 Gy) may be of bene®t in previous responders. In contrast to that, Cole [2] reported that as many as 50% of re-irradiated patients further required analgesics. In a study of Hoskin et al. [6] that evaluated optimal singlefraction RT (4 Gy vs. 8 Gy), 71% of evalauble patients responded to re-treatment of 4 Gy, while 44% responded to re-treatment of 8 Gy. This difference was not a signi®cant one, and did not change overall results, 8 Gy being still more effective during the 12-week assessment period. Similar
B. Jeremic et al. / Radiotherapy and Oncology 52 (1999) 123±127
results were obtained during the current study when a single fraction of 4 Gy was used (RR 74%). Furthermore, when `total' (cumulative) responses during the current study period were taken into account (when responses to second 4 Gy in previous non-responders were added to initial responses seen after ®rst 4 Gy), initial 6 Gy or 8 Gy (without counting `additional` 4% response from initial non-responders in both of these two groups) still had higher overall response rates than that achieved (63% vs. 73% vs. 78% for the groups I±III, respectively). Mithal et al. [10] reported on 57 sites retreated once and 8 retreated twice out of 280 individual treatment sites irradiated initially. A total of 17 out of 23 (74%) patients responded (CR 1 PR) to second RT that used a number of single-fraction regimens (an identical ®nding to our own), which was not signi®cantly inferior to 91% obtained with more protracted regimens. Furthermore, the third (second re-irradiation) RT given again as a single fraction, was successful in 80% of patients. As in the current study, previous CR carried an increased likelihood for achieving second CR, with response being shorter in patients with initial PR than in patients with initial CR. Uppelschoten et al. [15] recently reported on a study with a similar policy to our own that considered pain relapse or insuf®cient palliation for re-irradiation. After long intervals from previous RT, re-irradiation with 6 Gy was given to 18 such patients to obtain a reduction of pain in 13 of them (72%). In conclusion, results of this study con®rmed previous observations that more than 70% of re-irradiated patients experience pain relief after second single-fraction RT. Response was also seen in 46% of patients that were re-irradiated after initial unsuccessful RT. Previous response status seems to govern the outcome of second RT favouring patients with initial CR, that were more likely to experience response after re-irradiation. With low toxicity, similar to that observed during the ®rst RT and the incidence of pathological fractures and spinal cord compressions similar between re-irradiated patients and those undergoing only one single-fraction RT, these may be important ®ndings that may help select patients suitable for different approaches when administering single-fraction RT in the future. Acknowledgements This study was supported in part by the Grant-in-Aid for Scienti®c Research (B) from the Japanese Ministry of
127
Education, Science and Culture (11470190, 11877152, 10557087).
References [1] Barak F, Werner A, Walach N, Horn Y. The palliative ef®cacy of a single high dose of radiation in treatment of symptomatic osseous metastases. Int. J. Radiat. Oncol. Biol. Phys. 1987;13:1233±1235. [2] Cole DJ. A randomised trial of a single treatment versus conventional fractionation in the palliative radiotherapy of painful bone metastases. Clin. Oncol. 1989;1:59±62. [3] Cox JD, Stetz J, Pajak TF. Toxicity criteria of the Radiation Therapy Oncology Group (RTOG) and the European Organisation for the Treatment and Research of Cancer (EORTC). Int. J. Radiat. Oncol. Biol. Phys. 1995;31:1341±1346. [4] Gaze MN, Kelly CG, Kerr GR, et al. Pain relief and quality of life following radiotherapy for bone metastases: a randomised trial of two fractionation schedules. Radiother. Oncol. 1997;45:109±116. [5] Hendrickson FR, Shehata WM, Kirchner AB. Radiation therapy for osseous metastasis. Int. J. Radiat. Oncol. Biol. Phys 1976;1:275± 278. [6] Hoskin PJ, Price P, Easton D, et al. A prospective randomised trial of 4 or 8 Gy single doses in the treatment of metastatic bone pain. Radiother. Oncol. 1992;23:74±78. [7] Jensen NH, Roesdahl K. Single-dose irradiation of bone metastases. Acta Radiol. Ther. Phys. Biol. 1976;15:337±339. [8] Jeremic B, Shibamoto Y, et al. Acimovic. Lj.. A randomised trial of three single-dose radiation therapy regimens in the treatment of metastatic bone pain. Int. J. Radiat. Oncol. Biol. Phys. 1998;42:161± 167. [9] Karstens JH, Schnabel B, Amman J. Management of metastatic bone pain: preliminary results with single fraction (4 Gy) radiotherapy. Onkologie 1989;12:41±42. [10] Mithal NP, Needham PR, Hoskin PJ. Re-treatment with radiotherapy for painful bone metastases. Int. J. Radiat. Oncol. Biol. Phys. 1994;29:1011±1014. [11] Nielsen OS, Bentzen SM, Sandberg E, Gadeberg CC, Timothy AR. Randomised trial of single dose versus fractionated palliative radiotherapy of bone metastases. Radiother. Oncol. 1998;47:233±240. [12] Penn CRM. Single dose and fractionated palliative irradiation for osseous metastases. Clin. Radiol. 1976;27:405±408. [13] Price P, Hoskin PJ, Easton D, Austin D, Palmer SG, Yarnold JR. Prospective randomised trial of single and multifraction radiotherapy schedules in the treatment of painful bone metastases. Radiother. Oncol. 1986;6:247±255. [14] Price P, Hoskin PJ, Easton D, Austin D, Palmer S, Yarnold JR. Low dose single fraction radiotherapy in the treatment of metastatic bone pain: A pilot study. Radiother. Oncol. 1988;12:297±300. [15] Uppelschoten JM, Wanders SL, de Jong JMA. Single-dose radiotherapy (6 Gy) palliation in painful bone metastases. Radiother. Oncol. 1995;36:198±202. [16] Vargha ZO, Glicksman AS, Boland J. Single-dose radiation therapy in the palliation of metastatic disease. Radiology 1969;93:1181± 1184.