- Email: [email protected]
Mammographic and ultrasonographic comparison between intraoperative radiotherapy (IORT) and conventional external radiotherapy (RT) in limited-stage breast cancer, conservatively treated
European Journal of Radiology 59 (2006) 222–230
Mammographic and ultrasonographic comparison between intraoperative radiotherapy (IORT) and conventional external radiotherapy (RT) in limited-stage breast cancer, conservatively treated Sabino Walter Della Sala a,∗ , Marco Pellegrini a , Daniela Bernardi a , Francesca Franzoso a , Marvi Valentini a , Stefano Di Michele a , Maurizio Centonze a , Salvatore Mussari b a
U.O. di Radiodiagnostica, Ospedale Santa Chiara, Dipartimento di Radiodiagnostica, Azienda Provinciale per i Servizi Sanitari, Trento, Italy b U.O. di Radioterapia, Ospedale Santa Chiara, Trento, Italy Received 9 October 2005; received in revised form 20 February 2006; accepted 3 March 2006
Abstract Objective: Assess mammographic and echographic modifications in mild cases of breast cancer (suitable for conservative surgery) after intraoperatory radio treatment (IORT) as opposed to conventional post-operative radiotherapy (RT). Materials and methods: We report data from 45 patients in each group (IORT and RT). All patients were examined using the same mammographic and ecographic equipment at 6, 12 and 24 months after treatment. We focused on structural alterations, edema and others, and quantified them using pre-established (unbiased) protocols. Both patient examination and subsequent assessment of the results were performed by radiologists with exepertise in breast cancer evaluation. Results: At 6 months, IORT patients showed slightly more pronounced structural distortions and oedema than RT patients; these differences became more apparent at 12 months, with the addition of fat necrosis and/or calcifications. These alterations were evident and consistent under both mammographic and ecographic examination, and became even more pronounced at 24 months. At this stage, RT patients showed minimal alterations of the tissue (apart from normal post-surgical scarring), whereas IORT patients showed virtually no improvement over the preceding 12-month period. Conclusion: We show radiological alterations in post-operative breast cancer are significantly more pronounced in patients treated with IORT as opposed to RT. © 2006 Elsevier Ireland Ltd. All rights reserved. Keywords: Breast cancer; Post-operative radiotherapy; Intraoperative radiotherapy; Mammography; Echography
1. Introduction In medical literature there are currently only few scientific publications dealing with intraoperative radiotherapy (IORT) in early breast cancer and they mainly focus on technical and surgical aspects [1–7]. This is because IORT is a newly introduced technique for conservative treatment of breast cancer. At present 70–80% of breast tumours can be treated with a conservative surgical therapy. In these cases RT represents ∗ Corresponding author at: U.O. di Radiodiagnostica, Ospedale Santa Chiara, Largo Medaglie d’oro, 38100 Trento, Italy. Tel.: +39 0461 903502/2371; fax: +39 0461 902373/3501. E-mail address: [email protected] (S.W. Della Sala).
0720-048X/$ – see front matter © 2006 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ejrad.2006.03.003
an adjuvant therapy, although there is still wide controversy about its duration and length [4,8–12]. Post-operative RT consists of whole breast radiation at 50–55 Gy during each daily session for 4 weeks, followed by 10–16 Gy radiations at the surgical site [4,9,10,12]. IORT involves the use of a single radiation dose delivered directly to the tumour site during surgery either as an actual or palliative treatment [1–3]. There are only few experiences in current medical literature reporting the use of IORT only. In most cases IORT has been used after conservative surgery as a preliminary treatment before standard radiotherapy [1–3]. As far as we know, a bibliography about modifications of mammographic or echographic aspects induced by IORT
S.W. Della Sala et al. / European Journal of Radiology 59 (2006) 222–230
is not yet available. The aim of this study is to analyze the modifications shown by instrumental assessments – such as mammography and echography – following IORT and to compare these modifications with those induced by standard external RT [8–12].
2. Materials and methods From January 2001 to December 2004 we carried out consecutive mammographic and echographic examinations on 45 patients who had undergone conservative surgical treatment of early breast cancer with intraoperative radiotherapy (IORT) as sole radiation treatment modality. Inclusion criteria for this analysis were: patient aged over 45 years; unifocal breast carcinoma up to a diameter of 3 cm; negative receptors and extensive intraductal carcinoma at preliminary biopsy; signed informed consent form. Contemporarily, using the same technique, we studied a group of 45 other patients, affected by limited-breast cancer, who had undergone conservative surgical treatment and treated in the same period with post-operative standard external radiotherapy (RT) (50 Gy plus 10 Gy-boost delivered in 30 fractions at 2 Gy/fraction). For this reason there are size differences between the tumours selected for RT compared to IORT, including a lot of DCIS available only in RT group. In both groups the maximum diameter of tumours was ≤25 mm. The mean age of patients treated with IORT was 62 years (age range: 45–79 years). Table 1 shows detailed characteristics of IORT patients. Two-third of tumours were extremely small (T1a and T1b). Tumours developed more frequently in the upper-external quadrant and the most comTable 1 Main characteristics of the patients (no of cases: 45 for each group) Patients IORT
Patients RT
Tumour size (T) T1a T1b T1c T2
3 28 12 2
6.60% 62.40% 26.60% 4.40%
2 10 26 7
4.40% 22.30% 57.80% 15.50%
Tumour site (quadrant) Upper outer quadrant Upper inner quadrant Lower inner quadrant Lower outer quadrant Central Superior sagittal
20 8 6 2 7 2
44.40% 17.80% 13.50% 4.40% 15.50% 4.40%
29 2 10 1 2 1
64.50% 4.40% 22.30% 2.20% 4.40% 2.20%
Histology Invasive ductal carcinoma Invasive lobular carcinoma DCIS Other
37 6 0 2
82.10% 13.50% 0.00% 4.40%
32 5 6 2
71.10% 11.20% 13.30% 4.40%
Tumour grade (G) G1 G2 G3
17 25 3
37.80% 55.60% 6.60%
12 18 15
26.70% 40% 33.30%
223
monly reported type was the infiltrating ductal carcinoma (in more than 80% of cases). Grading showed that over 50% of tumours were moderately differentiated (G2). The mean age of patients treated with RT was 60 years (age range: 40–78 years). Also for this group, Table 1 shows the main characteristics of patients examined. Over half of the tumours studied had a diameter of 1–2 cm (T1c). Tumours developed more frequently in the upper-external quadrant (more than 60% of cases) also for the RT group and the infiltrating ductal carcinoma was the most commonly reported type (more than 71.1% of cases). Grading of cellular differentiation of cancer showed equal rates for all three grades, with a slight prevalence of G2. In IORT modality, to minimize the irradiation of the thoracic wall and to guarantee the delivery of a full radiation dose to the gland, an attenuator aluminium-lead disk was placed between the gland and the pectoralis muscle. The surrounding tissue was mobilised and temporarily approximated by suture in order to expose the correct part of the gland to the radiation beam and to avoid excessive inhomogeneity in the target volume. Thereafter the skin margins were hitched to the non-traumatic hooks of a plastic ring and gently stretched out of the radiation field. The thickness of the target volume was measured by a needle and a ruler in different points taking into account the average value; excessive inhomogeneities were eventually corrected using bolus material (i.e. wet sterile gauze). In the radiotherapy room a proper applicator diameter was selected and visually placed directly in contact to the breast so that the whole tumour bed and surrounding tissue of approximately 2–3 cm were in the radiation field. The correct position of the attenuator disk was checked by a needle and further tissue-equivalent barrier (wet sterile gauzes) was placed between the skin and the applicator edge to absorb the low energy electrons scattered around by applicator itself. The irradiation was performed using a non-dedicated LINAC (Elekta Precise model) with in-house designed applicator tubes of 5–6 cm in diameter. According to the literature [4,6], three different dose-levels were used: 20 Gy in 5 patients, 22 Gy in 20 patients and 24 Gy in 20 patients. In vivo-dosimetry was performed in 22 patients; for each patient 8 TLD dosimeters were placed in a sterile package on the top of the disk in order to measure the absorbed dose at the maximum depth of the target volume. Using in vivo dosimetry we also defined the appropriate aluminium–lead thickness which minimizes the backscatter factor of the attenuator disk. According to the measured tissue-depth, 34 patients were treated with 8 MeV electrons energy (average breast thickness: 14 mm [minimum]–20 mm [maximum]) and 11 patients with 12 MeV electrons energy (average breast thickness: 24 mm [minimum]–30 mm [maximum]). The dose of electrons was prescribed at the depth of the maximum dose (Dmax ) obtaining an homogeneous coverage of the target even in patients with a residual breast thickness of 25 mm or more. Patients treated with IORT underwent mammographic and echographic examinations during follow-up visits carried out
224
S.W. Della Sala et al. / European Journal of Radiology 59 (2006) 222–230
at our department at 6 and 12 months after the end of treatment and then once a year thereafter. On the contrary, patients treated with RT had their first follow-up visit only 1 year after the end of the treatment. For this reason, all RT patients were previously informed about the features of the study and they were enrolled only after obtaining their consent. Following these procedures, we were able to schedule a mammographic/echographic control visit at 6, 12 and 24 months (timing parameters of this study) from the end of treatment for each patient enrolled in both groups. Instrumental assessments were performed using the same equipment for all patients: an analogue mammography (Performa General Electric Medical System/Instrumentarium, Tuusula, Finland) and a digital ultrasonography (Technos MP Esaote, Genoa, Italy). Each assessment was performed by a qualified breast-screening radiologist working in our Department of Breast Imaging and Mammographic Screening. This was then jointly examined by three qualified breastscreening radiologists working in the same centre. When the radiologists performed the check visit (mammography and ecography) and when the three radiologists reviewed mammographic and echographic findings, they did not know if the patients were treated with IORT or RT. These doctors had previously discussed and agreed upon the indicators (partially common to both assessments) to be considered during their analysis. The indicators considered in mammography were: post-operative architectural distortion, oedema and “other findings”. The indicators considered in echographic examination were: oedema and development of “other echo-structural modifications”, such as the fluid collections, lymphatic stasis, fluid collections related to lymphatic stasis, hypoechoic areas and seromas. Each indicator was also precisely mea-
sured according to seriousness and grade. For example, postoperative architectural distortion was graded from absent to strong distortion, while oedema was graded from negative (−) to highly positive (++++). As for the indicator “other”, i.e. development of other modifications, these were grouped into macro-groups only at the end of the assessment procedure, since it was not possible to predict the alterations observed.
3. Results The analysis of results has been obtained at each control visit (6, 12 and 24 months) including information provided by every single tool (mammography or echography) separately, and then comparing IORT reports versus RT ones. The first mammographic control (at 6 months) did not show substantial differences between the two groups. Analytical results are reported in Tables 2–4 and show that post-operative architectural distortion was more frequent in IORT than RT patients, especially in the more severe grades (Table 2). As for oedema, major differences were not found between the two groups (Table 3). “Other findings” were mainly identified and macro-grouped into masses, fat necrosis, calcifications and fat necrosis associated with calcifications (Table 4). The first echographic control (Table 5) showed that there is a substantial differences in the grade of oedema between the two groups: besides lower grades of oedema are most frequent in RT, while higher grades were most found in the IORT group. “Other echo-structural findings” showed differences in the two groups, which became even more evident
Table 2 Main characteristics of mammography: distorsion (from absent to strong) group IORT vs. RT after 6, 12 and 24 months
S.W. Della Sala et al. / European Journal of Radiology 59 (2006) 222–230
225
Table 3 Main characteristics of mammography: oedema (from negative to highly positive) group IORT vs. RT after 6, 12 and 24 months
in the following control visits (Table 6). Collections (fluidcorpuscled areas; the larger ones were almost completely drained under echographic guidance) were reported more frequently in IORT than RT patients. Also, in the IORT group ecography showed the unusual finding at the surgical site of apparently solid, mainly hypoechoic areas with different morphology, dishomogeneous structure, and with undefined margins (Fig. 1). It was not easy to define the nature of this finding. Therefore, it was necessary to carry out further diagnostic investigations collecting cytological samples under echographic guidance. We classified the cytological samples collected as necrosis areas (both of adipose and parenchymatous components). The onset of seromas in the surgical site (in two cases seromas were drained under echographic guidance with full resolution) and a number of cases of lymphatic stasis (Table 6) was also found to be specific and peculiar to the RT group (Fig. 2). Mammographic control at 12 months (Table 2) confirmed that post-operative architectural distortion was more frequent
in the IORT group, especially in the higher grades and that oedema was persistent in both groups (Table 3). Also, it showed a substantial differentiation in the onset of “other findings” (Table 4). Specifically, IORT group was affected by a major increase in cases compatible with fat necrosis (Fig. 3) with or without associated calcifications (Figs. 4 and 5). Echographic controls at 12 months (Table 6) confirmed differentiation of those echographic reports previously defined as “peculiar” in both groups. Specifically, the number of fluid-corpuscled collections was still clearly high in the IORT group and there was still evidence of hypoechoic areas near the surgical site. This was distinctive of the IORT group only. Comparative assessment between IORT echographic evaluations after 6 and 12 months showed that partial decrease in hypoechoic areas was associated with a parallel increase in fluid-corpuscled collections. This is not an unpredictable occurrence but it is likely related to the evolution of intraparenchymatous changes induced by intraoperative radiotherapy. This is supported by the cytological assessments that showed necrosis cases in the samples collected.
Table 4 Main characteristics of mammography: Other findings group IORT vs. RT after 6, 12 and 24 months IORT
RT
Other findings
6 months
12 months
24 months
6 months
12 months
24 months
Negative Mass Fat necrosis Calcifications Fat necrosis + calcifications
39 1 4 0 1
25 0 12 1 7
24 0 12 1 8
42 0 2 0 1
42 0 2 1 0
42 0 0 3 0
226
S.W. Della Sala et al. / European Journal of Radiology 59 (2006) 222–230
Table 5 Main characteristics of ecography: oedema (from negative to highly positive) group IORT vs. RT after 6, 12 and 24 months
Table 6 Main characteristics of ecography group: other findings group IORT vs. RT after 6, 12 and 24 months Other findings
Negative Collections Lymphatic stasis Collections + lymphatic stasis Hypoechoic areas Seroma
IORT
RT
6 months
12 months
24 months
6 months
12 months
24 months
16 14 0 0 15 0
16 19 0 0 10 0
19 16 0 0 10 0
23 5 10 5 0 2
35 2 8 0 0 0
42 0 3 0 0 0
In other words, some hypoechoic areas, expressing necrotic phenomenons, presented evidence of progressive colliquation (particularly clear when the adipose component was affected), causing the transformation of these areas into fluidcorpuscled collections. A specific feature of the RT group was lymphatic stasis. At 24 months, mammography (Table 2) showed that postoperative architectural distortion was still more frequently found in the IORT group. Moreover, there were still accentuated differences between the two groups in the rate of fat necrosis and fat necrosis with calcifications (Table 4). Despite the small number of cases studied, the observed differences (mammographic other findings at 24 months) were evaluated by the chi-square tests (χ2 ), showing a statistical significance (χ2 = 15.33; p < 0.001). Ecography at 24 months (Table 5) showed an almost complete absence of oedema in both groups. This also confirmed the occurrence noticed during the previous controls carried out on IORT patients: significant number of fluidcorpuscled collections had only slightly decreased since the previous control and of hypoechoic areas at the surgery site,
whose number remained unchanged (Table 6). The RT group still showed cases of lymphatic stasis. However, their number decreased over time. Also for echographic other findings at 24 months, the observed differences, evaluated by the chi-square tests (χ2 ), showed a statistical significance (χ2 = 20.31; p < 0.001).
4. Discussion Before further discussion about the results of this study, it is necessary to provide a general overview about standard external radiotherapy (RT) and intraoperative radiotherapy (IORT). In most cases of breast cancer, suitable for conservative surgery, RT represents an adjuvant therapy. The rate of local recurrences after this procedure is about 5–18% [4]. One of the main causes of the failure of radiation treatment is a tumour inaccurate localization and a partial geographical miss of site: up to 85% of recurrences develop at the surgical site or very close to it. The occurrences of tumours
S.W. Della Sala et al. / European Journal of Radiology 59 (2006) 222–230
227
Fig. 1. IORT. Echographic pictures. (A) Ovoid morphology hypoechoic lesion with undefinited edges and dishomogeneous structure, prevalently solid. (B) Undefinited morphology hypoechoic lesion with irregular edges and dishomogeneous structure, prevalently solid. (C) Predominant vertical developed hypoechoic lesion with irregular edges and dishomogeneous structure, prevalently solid. (D) Ovoid morphology hypoechoic lesion with definited edges and a prevalently liquid structure.
in other areas of the breast, far from the surgical scar, are not considered as recurrences. This is why the common aim was to reach high radiation doses within this narrow volume in order to fully destroy any residual cancer cells [4,9,10,12]. Some recent studies have shown that IORT could prevent the onset of local recurrences in patients affected by grade I and II cancer. This might avoid treatment with external radiations, reducing treatment time and costs and improving the psychological and aesthetic effects [1–4]. These are still preliminary considerations, since a long follow-up period is essential to give full scientific support to the situation previously analyzed [13]. IORT certainly allows shorter exposure to radiations of patients’ skin, lung and subcutaneous soft tissues. It also prevents radiation of the controlateral breast, thus reducing the incidence of radiation-related sequels (e.g. fibrosis and
telangiectasias), enhancing also the aesthetic aspect [1–3]. IORT does not interfere with the oncological criteria of conservative surgery: surgical procedure and resection depth and length do not change and the operating time is just slightly longer [1–4]. Now we would like to analyze the findings of our study. First of all, in medical literature we were not able to find any other scientific publications dealing with this topic from a radiological perspective and literature refers only to post-RT mammographic aspects [14,15]. Secondly, the results of this study might be limited by the small number of cases studied. The most important result was about mammographic and echographic alterations affecting patients undergoing IORT. The study showed how in these patients the alterations became progressively more evident than in patients treated with RT, apart from post-operative structural changes
Fig. 2. RT. Echographic pictures. (A) A little liquid collection in surgical site. (B) Lymphatic stasis.
228
S.W. Della Sala et al. / European Journal of Radiology 59 (2006) 222–230
Fig. 3. IORT. Mammographic pictures. Mammography: medio-lateral (A) and axial (B) projection. In the surgical site a radio-transparent area with definited and thickened edge is shown (fat necrosis).
Fig. 4. IORT. Mammographic pictures. Mammography: medio-lateral (A) and axial (B) projection. In the surgical site a radio-transparent area with calcific edge is shown (fat necrosis). Ecography: same case (C).
S.W. Della Sala et al. / European Journal of Radiology 59 (2006) 222–230
229
Fig. 5. IORT. Mammographic pictures. Mammography: medio-lateral (A) projection. In the surgical site a radio-transparent area with partial calcific edge is shown (fat necrosis). Ecography: same case (B).
reported by mammography and oedema grade (showed both by mammography and echography) that were progressively reducing [16,17]. Specifically, mammography performed on patients treated with IORT showed alterations compatible with fat necrosis phenomena with or without associated calcifications. This result is consistent with the IORT surgical procedure previously explained. IORT involves direct radiation of the surgical site, increasing the chances of developing necrosis. Necrosis was more evident some time after the end of treatment. Recently some authors [18] reported the presence of multiple punctate “pseudocalcifications” at the surgical site in patients treated with IORT; we too found similar findings but in our study calcifications are almost always correlated with the presence of fat necrosis. Also the echographic alterations reported in the IORT group were likely related to what we have just explained. Presence of corpuscled collections at the surgical site is consistent with radio-induced necrosis, and this is why they were also reported in the RT group. Evidence of hypoechoic areas around or next to the surgical site is a peculiar aspect of IORT patients. These findings are not easy to understand since their aspect, morphology and structure resemble local recurrences. More than one cytological assessment was necessary to fully understand these findings, eventually showing cellular alterations compatible with necrosis. Modifications during the time course of these findings that in some cases completely develop into actual fluid-corpuscled collections, helped us understand the situation showed by the echographic assessment in radiobiological terms, expressing the grading of radio-induced necrosis. Radio-induced necrosis did not reduce over time [19]. Echography showing lymphatic stasis in RT patients was partially related to radio-dermatitis [16].
We would like to emphasize again that the small number of cases studied might invalidate this analysis; however, differences showed by mammography and echography in both groups (IORT and RT) were significant as chi-square tests (χ2 ) showed. Finally, it is important to note that during follow-up local recurrences developed neither in the IORT nor in the RT group. 5. Conclusions Data from this study showed that radiological alterations in patients treated with IORT are by all means more significant if compared with the alterations developed in patients treated with RT. Specifically, the different mammographic and echographic aspects are more evident, and somewhat of a more difficult interpretation, some time after the end of the treatment, with no substantial changes also at late follow-up visits. This analysis focused on mammographic and echographic aspects only and did not include any assessment of surgical or radiotherapy techniques used. Also, it is necessary to remember that all reports about the therapeutic efficacy of IORT are still under preliminary analysis, since an adequate follow-up period is essential to give full scientific support to this procedure [1,7,13].
References [1] Gatzemeier W, Orecchia R, Intra M, Veronesi U. Intraoperative radiotherapy (IORT) in treatment of breast carcinoma—a new therapeutic alternative within the scope of breast-saving therapy? Current status and future prospects. Report of experiences from European Institute of Oncology. Strahlenther Onkol 2001;177:30–337.
230
S.W. Della Sala et al. / European Journal of Radiology 59 (2006) 222–230
[2] Veronesi U, Orecchia R, Luini A, et al. A preliminary report of intraoperative radiotherapy (IORT) in limited-stage breast cancers that are conservatively treated. Eur J Cancer 2001;37:2178– 83. [3] Intra M, Gatti G, Luini A, et al. Surgical technique of intraoperative radiotherapy in conservative treatment of limited-stage breast cancer. Arch Surg 2002;137:737–40. [4] Reitsamer R, Peintinger F, Sedlmayer F, et al. Intraoperative radiotherapy given as a boost after breast-conserving surgery in breast cancer patients. Eur J Cancer 2002;38:1607–10. [5] Veronesi U, Gatti G, Luini A, et al. Full-dose intraoperative radiotherapy with electrons during breast-conserving surgery. Arch Surg 2003;138:1253–6. [6] Veronesi U, Gatti G, Luini A, et al. Intraoperative radiation therapy for breast cancer: technical notes. Breast J 2003;9:106–12. [7] Bartelink H. Commentary on the paper “A preliminary report of intraoperative radiotherapy (IORT) in limited-stage breast cancers that are conservatively treated”: a critical review of an innovative approach. Eur J Cancer 2001;37:2143–6. [8] Halberg FE, Shank BM, Haffty BG, et al. Conservative surgery and radiation in the treatment of stage I and II carcinoma of the breast. American college of radiology. ACR appropriateness criteria. Radiology 2000;215:1139–205. [9] Hannoun-Levi JM, Marsiglia H, Garbay JR, Gerard JP. Partial irradiation of the breast: why, how? Cancer Radiother 2003;7: 200–9. [10] Polgar C, Fodor J, Major T, et al. Radiotherapy confined to the tumor bed following breast conserving surgery current status, controversies, and future projects. Strahlenther Onkol 2002;178:597–606.
[11] Taghian AG, Powell SN. The role of radiation therapy for primary breast cancer. Surg Clin North Am 1999;79:1091–115. [12] Vicini FA, Kestin L, Chen P, et al. Limited-field radiation therapy in the management of early stage breast cancer. J Natl Cancer Instit 2003;95:1205–10. [13] Reitsamer R, Peintinger F, Kopp M, et al. Local recurrence rates in breast cancer patients treated with intraoperative electronboost radiotherapy versus post-operative external-beam electronboost irradiation. A sequential intervention study. Strahlenther Onkol 2004;180:38–44. [14] Holli K, Saaristo R, Isola J, Hyoty M, Hakama M. Effect of radiotherapy on the interpretation of routine follow-up mammography after conservative breast surgery: a randomized study. Br J Cancer 1998;78:542–5. [15] Dershaw DD. Breast imaging and the conservative treatment of breast cancer. Radiol Clin North Am 2002;40:501–16. [16] Meric F, Buchholz TA, Mirza NQ, et al. Long-term complications associated with breast-conservation surgery and radiotherapy. Ann Surg Oncol 2002;9:543–9. [17] Wazer DE, Lowther D, Boyle T, et al. Clinically evident fat necrosis in women treated with high-dose-rate brachytherapy alone for early stage breast cancer. Int J Radiat Oncol Biol Phys 2001;50:107–11. [18] Curran S, Khilnani A, Zelefsky M, et al. Evaluation of Mammographic findings after intraoperative radiation (IORT) alone for early stage breast cancer. http://rsna2005.rsna.org/rsna2005/v2005/ conference/event display.cfm?em id=4407750. [19] Boyages J, Bilous M, Barraclough B, Landglands AO. Fat necrosis of the breast following lumpectomy and radiation therapy for early breast cancer. Radiother Oncol 1988;13:69–74.
Mammographic and ultrasonographic comparison between intraoperative radiotherapy (IORT) and conventional external radiotherapy (RT) in limited-stage breast cancer, conservatively treated Sabino Walter Della Sala a,∗ , Marco Pellegrini a , Daniela Bernardi a , Francesca Franzoso a , Marvi Valentini a , Stefano Di Michele a , Maurizio Centonze a , Salvatore Mussari b a
U.O. di Radiodiagnostica, Ospedale Santa Chiara, Dipartimento di Radiodiagnostica, Azienda Provinciale per i Servizi Sanitari, Trento, Italy b U.O. di Radioterapia, Ospedale Santa Chiara, Trento, Italy Received 9 October 2005; received in revised form 20 February 2006; accepted 3 March 2006
Abstract Objective: Assess mammographic and echographic modifications in mild cases of breast cancer (suitable for conservative surgery) after intraoperatory radio treatment (IORT) as opposed to conventional post-operative radiotherapy (RT). Materials and methods: We report data from 45 patients in each group (IORT and RT). All patients were examined using the same mammographic and ecographic equipment at 6, 12 and 24 months after treatment. We focused on structural alterations, edema and others, and quantified them using pre-established (unbiased) protocols. Both patient examination and subsequent assessment of the results were performed by radiologists with exepertise in breast cancer evaluation. Results: At 6 months, IORT patients showed slightly more pronounced structural distortions and oedema than RT patients; these differences became more apparent at 12 months, with the addition of fat necrosis and/or calcifications. These alterations were evident and consistent under both mammographic and ecographic examination, and became even more pronounced at 24 months. At this stage, RT patients showed minimal alterations of the tissue (apart from normal post-surgical scarring), whereas IORT patients showed virtually no improvement over the preceding 12-month period. Conclusion: We show radiological alterations in post-operative breast cancer are significantly more pronounced in patients treated with IORT as opposed to RT. © 2006 Elsevier Ireland Ltd. All rights reserved. Keywords: Breast cancer; Post-operative radiotherapy; Intraoperative radiotherapy; Mammography; Echography
1. Introduction In medical literature there are currently only few scientific publications dealing with intraoperative radiotherapy (IORT) in early breast cancer and they mainly focus on technical and surgical aspects [1–7]. This is because IORT is a newly introduced technique for conservative treatment of breast cancer. At present 70–80% of breast tumours can be treated with a conservative surgical therapy. In these cases RT represents ∗ Corresponding author at: U.O. di Radiodiagnostica, Ospedale Santa Chiara, Largo Medaglie d’oro, 38100 Trento, Italy. Tel.: +39 0461 903502/2371; fax: +39 0461 902373/3501. E-mail address: [email protected] (S.W. Della Sala).
0720-048X/$ – see front matter © 2006 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ejrad.2006.03.003
an adjuvant therapy, although there is still wide controversy about its duration and length [4,8–12]. Post-operative RT consists of whole breast radiation at 50–55 Gy during each daily session for 4 weeks, followed by 10–16 Gy radiations at the surgical site [4,9,10,12]. IORT involves the use of a single radiation dose delivered directly to the tumour site during surgery either as an actual or palliative treatment [1–3]. There are only few experiences in current medical literature reporting the use of IORT only. In most cases IORT has been used after conservative surgery as a preliminary treatment before standard radiotherapy [1–3]. As far as we know, a bibliography about modifications of mammographic or echographic aspects induced by IORT
S.W. Della Sala et al. / European Journal of Radiology 59 (2006) 222–230
is not yet available. The aim of this study is to analyze the modifications shown by instrumental assessments – such as mammography and echography – following IORT and to compare these modifications with those induced by standard external RT [8–12].
2. Materials and methods From January 2001 to December 2004 we carried out consecutive mammographic and echographic examinations on 45 patients who had undergone conservative surgical treatment of early breast cancer with intraoperative radiotherapy (IORT) as sole radiation treatment modality. Inclusion criteria for this analysis were: patient aged over 45 years; unifocal breast carcinoma up to a diameter of 3 cm; negative receptors and extensive intraductal carcinoma at preliminary biopsy; signed informed consent form. Contemporarily, using the same technique, we studied a group of 45 other patients, affected by limited-breast cancer, who had undergone conservative surgical treatment and treated in the same period with post-operative standard external radiotherapy (RT) (50 Gy plus 10 Gy-boost delivered in 30 fractions at 2 Gy/fraction). For this reason there are size differences between the tumours selected for RT compared to IORT, including a lot of DCIS available only in RT group. In both groups the maximum diameter of tumours was ≤25 mm. The mean age of patients treated with IORT was 62 years (age range: 45–79 years). Table 1 shows detailed characteristics of IORT patients. Two-third of tumours were extremely small (T1a and T1b). Tumours developed more frequently in the upper-external quadrant and the most comTable 1 Main characteristics of the patients (no of cases: 45 for each group) Patients IORT
Patients RT
Tumour size (T) T1a T1b T1c T2
3 28 12 2
6.60% 62.40% 26.60% 4.40%
2 10 26 7
4.40% 22.30% 57.80% 15.50%
Tumour site (quadrant) Upper outer quadrant Upper inner quadrant Lower inner quadrant Lower outer quadrant Central Superior sagittal
20 8 6 2 7 2
44.40% 17.80% 13.50% 4.40% 15.50% 4.40%
29 2 10 1 2 1
64.50% 4.40% 22.30% 2.20% 4.40% 2.20%
Histology Invasive ductal carcinoma Invasive lobular carcinoma DCIS Other
37 6 0 2
82.10% 13.50% 0.00% 4.40%
32 5 6 2
71.10% 11.20% 13.30% 4.40%
Tumour grade (G) G1 G2 G3
17 25 3
37.80% 55.60% 6.60%
12 18 15
26.70% 40% 33.30%
223
monly reported type was the infiltrating ductal carcinoma (in more than 80% of cases). Grading showed that over 50% of tumours were moderately differentiated (G2). The mean age of patients treated with RT was 60 years (age range: 40–78 years). Also for this group, Table 1 shows the main characteristics of patients examined. Over half of the tumours studied had a diameter of 1–2 cm (T1c). Tumours developed more frequently in the upper-external quadrant (more than 60% of cases) also for the RT group and the infiltrating ductal carcinoma was the most commonly reported type (more than 71.1% of cases). Grading of cellular differentiation of cancer showed equal rates for all three grades, with a slight prevalence of G2. In IORT modality, to minimize the irradiation of the thoracic wall and to guarantee the delivery of a full radiation dose to the gland, an attenuator aluminium-lead disk was placed between the gland and the pectoralis muscle. The surrounding tissue was mobilised and temporarily approximated by suture in order to expose the correct part of the gland to the radiation beam and to avoid excessive inhomogeneity in the target volume. Thereafter the skin margins were hitched to the non-traumatic hooks of a plastic ring and gently stretched out of the radiation field. The thickness of the target volume was measured by a needle and a ruler in different points taking into account the average value; excessive inhomogeneities were eventually corrected using bolus material (i.e. wet sterile gauze). In the radiotherapy room a proper applicator diameter was selected and visually placed directly in contact to the breast so that the whole tumour bed and surrounding tissue of approximately 2–3 cm were in the radiation field. The correct position of the attenuator disk was checked by a needle and further tissue-equivalent barrier (wet sterile gauzes) was placed between the skin and the applicator edge to absorb the low energy electrons scattered around by applicator itself. The irradiation was performed using a non-dedicated LINAC (Elekta Precise model) with in-house designed applicator tubes of 5–6 cm in diameter. According to the literature [4,6], three different dose-levels were used: 20 Gy in 5 patients, 22 Gy in 20 patients and 24 Gy in 20 patients. In vivo-dosimetry was performed in 22 patients; for each patient 8 TLD dosimeters were placed in a sterile package on the top of the disk in order to measure the absorbed dose at the maximum depth of the target volume. Using in vivo dosimetry we also defined the appropriate aluminium–lead thickness which minimizes the backscatter factor of the attenuator disk. According to the measured tissue-depth, 34 patients were treated with 8 MeV electrons energy (average breast thickness: 14 mm [minimum]–20 mm [maximum]) and 11 patients with 12 MeV electrons energy (average breast thickness: 24 mm [minimum]–30 mm [maximum]). The dose of electrons was prescribed at the depth of the maximum dose (Dmax ) obtaining an homogeneous coverage of the target even in patients with a residual breast thickness of 25 mm or more. Patients treated with IORT underwent mammographic and echographic examinations during follow-up visits carried out
224
S.W. Della Sala et al. / European Journal of Radiology 59 (2006) 222–230
at our department at 6 and 12 months after the end of treatment and then once a year thereafter. On the contrary, patients treated with RT had their first follow-up visit only 1 year after the end of the treatment. For this reason, all RT patients were previously informed about the features of the study and they were enrolled only after obtaining their consent. Following these procedures, we were able to schedule a mammographic/echographic control visit at 6, 12 and 24 months (timing parameters of this study) from the end of treatment for each patient enrolled in both groups. Instrumental assessments were performed using the same equipment for all patients: an analogue mammography (Performa General Electric Medical System/Instrumentarium, Tuusula, Finland) and a digital ultrasonography (Technos MP Esaote, Genoa, Italy). Each assessment was performed by a qualified breast-screening radiologist working in our Department of Breast Imaging and Mammographic Screening. This was then jointly examined by three qualified breastscreening radiologists working in the same centre. When the radiologists performed the check visit (mammography and ecography) and when the three radiologists reviewed mammographic and echographic findings, they did not know if the patients were treated with IORT or RT. These doctors had previously discussed and agreed upon the indicators (partially common to both assessments) to be considered during their analysis. The indicators considered in mammography were: post-operative architectural distortion, oedema and “other findings”. The indicators considered in echographic examination were: oedema and development of “other echo-structural modifications”, such as the fluid collections, lymphatic stasis, fluid collections related to lymphatic stasis, hypoechoic areas and seromas. Each indicator was also precisely mea-
sured according to seriousness and grade. For example, postoperative architectural distortion was graded from absent to strong distortion, while oedema was graded from negative (−) to highly positive (++++). As for the indicator “other”, i.e. development of other modifications, these were grouped into macro-groups only at the end of the assessment procedure, since it was not possible to predict the alterations observed.
3. Results The analysis of results has been obtained at each control visit (6, 12 and 24 months) including information provided by every single tool (mammography or echography) separately, and then comparing IORT reports versus RT ones. The first mammographic control (at 6 months) did not show substantial differences between the two groups. Analytical results are reported in Tables 2–4 and show that post-operative architectural distortion was more frequent in IORT than RT patients, especially in the more severe grades (Table 2). As for oedema, major differences were not found between the two groups (Table 3). “Other findings” were mainly identified and macro-grouped into masses, fat necrosis, calcifications and fat necrosis associated with calcifications (Table 4). The first echographic control (Table 5) showed that there is a substantial differences in the grade of oedema between the two groups: besides lower grades of oedema are most frequent in RT, while higher grades were most found in the IORT group. “Other echo-structural findings” showed differences in the two groups, which became even more evident
Table 2 Main characteristics of mammography: distorsion (from absent to strong) group IORT vs. RT after 6, 12 and 24 months
S.W. Della Sala et al. / European Journal of Radiology 59 (2006) 222–230
225
Table 3 Main characteristics of mammography: oedema (from negative to highly positive) group IORT vs. RT after 6, 12 and 24 months
in the following control visits (Table 6). Collections (fluidcorpuscled areas; the larger ones were almost completely drained under echographic guidance) were reported more frequently in IORT than RT patients. Also, in the IORT group ecography showed the unusual finding at the surgical site of apparently solid, mainly hypoechoic areas with different morphology, dishomogeneous structure, and with undefined margins (Fig. 1). It was not easy to define the nature of this finding. Therefore, it was necessary to carry out further diagnostic investigations collecting cytological samples under echographic guidance. We classified the cytological samples collected as necrosis areas (both of adipose and parenchymatous components). The onset of seromas in the surgical site (in two cases seromas were drained under echographic guidance with full resolution) and a number of cases of lymphatic stasis (Table 6) was also found to be specific and peculiar to the RT group (Fig. 2). Mammographic control at 12 months (Table 2) confirmed that post-operative architectural distortion was more frequent
in the IORT group, especially in the higher grades and that oedema was persistent in both groups (Table 3). Also, it showed a substantial differentiation in the onset of “other findings” (Table 4). Specifically, IORT group was affected by a major increase in cases compatible with fat necrosis (Fig. 3) with or without associated calcifications (Figs. 4 and 5). Echographic controls at 12 months (Table 6) confirmed differentiation of those echographic reports previously defined as “peculiar” in both groups. Specifically, the number of fluid-corpuscled collections was still clearly high in the IORT group and there was still evidence of hypoechoic areas near the surgical site. This was distinctive of the IORT group only. Comparative assessment between IORT echographic evaluations after 6 and 12 months showed that partial decrease in hypoechoic areas was associated with a parallel increase in fluid-corpuscled collections. This is not an unpredictable occurrence but it is likely related to the evolution of intraparenchymatous changes induced by intraoperative radiotherapy. This is supported by the cytological assessments that showed necrosis cases in the samples collected.
Table 4 Main characteristics of mammography: Other findings group IORT vs. RT after 6, 12 and 24 months IORT
RT
Other findings
6 months
12 months
24 months
6 months
12 months
24 months
Negative Mass Fat necrosis Calcifications Fat necrosis + calcifications
39 1 4 0 1
25 0 12 1 7
24 0 12 1 8
42 0 2 0 1
42 0 2 1 0
42 0 0 3 0
226
S.W. Della Sala et al. / European Journal of Radiology 59 (2006) 222–230
Table 5 Main characteristics of ecography: oedema (from negative to highly positive) group IORT vs. RT after 6, 12 and 24 months
Table 6 Main characteristics of ecography group: other findings group IORT vs. RT after 6, 12 and 24 months Other findings
Negative Collections Lymphatic stasis Collections + lymphatic stasis Hypoechoic areas Seroma
IORT
RT
6 months
12 months
24 months
6 months
12 months
24 months
16 14 0 0 15 0
16 19 0 0 10 0
19 16 0 0 10 0
23 5 10 5 0 2
35 2 8 0 0 0
42 0 3 0 0 0
In other words, some hypoechoic areas, expressing necrotic phenomenons, presented evidence of progressive colliquation (particularly clear when the adipose component was affected), causing the transformation of these areas into fluidcorpuscled collections. A specific feature of the RT group was lymphatic stasis. At 24 months, mammography (Table 2) showed that postoperative architectural distortion was still more frequently found in the IORT group. Moreover, there were still accentuated differences between the two groups in the rate of fat necrosis and fat necrosis with calcifications (Table 4). Despite the small number of cases studied, the observed differences (mammographic other findings at 24 months) were evaluated by the chi-square tests (χ2 ), showing a statistical significance (χ2 = 15.33; p < 0.001). Ecography at 24 months (Table 5) showed an almost complete absence of oedema in both groups. This also confirmed the occurrence noticed during the previous controls carried out on IORT patients: significant number of fluidcorpuscled collections had only slightly decreased since the previous control and of hypoechoic areas at the surgery site,
whose number remained unchanged (Table 6). The RT group still showed cases of lymphatic stasis. However, their number decreased over time. Also for echographic other findings at 24 months, the observed differences, evaluated by the chi-square tests (χ2 ), showed a statistical significance (χ2 = 20.31; p < 0.001).
4. Discussion Before further discussion about the results of this study, it is necessary to provide a general overview about standard external radiotherapy (RT) and intraoperative radiotherapy (IORT). In most cases of breast cancer, suitable for conservative surgery, RT represents an adjuvant therapy. The rate of local recurrences after this procedure is about 5–18% [4]. One of the main causes of the failure of radiation treatment is a tumour inaccurate localization and a partial geographical miss of site: up to 85% of recurrences develop at the surgical site or very close to it. The occurrences of tumours
S.W. Della Sala et al. / European Journal of Radiology 59 (2006) 222–230
227
Fig. 1. IORT. Echographic pictures. (A) Ovoid morphology hypoechoic lesion with undefinited edges and dishomogeneous structure, prevalently solid. (B) Undefinited morphology hypoechoic lesion with irregular edges and dishomogeneous structure, prevalently solid. (C) Predominant vertical developed hypoechoic lesion with irregular edges and dishomogeneous structure, prevalently solid. (D) Ovoid morphology hypoechoic lesion with definited edges and a prevalently liquid structure.
in other areas of the breast, far from the surgical scar, are not considered as recurrences. This is why the common aim was to reach high radiation doses within this narrow volume in order to fully destroy any residual cancer cells [4,9,10,12]. Some recent studies have shown that IORT could prevent the onset of local recurrences in patients affected by grade I and II cancer. This might avoid treatment with external radiations, reducing treatment time and costs and improving the psychological and aesthetic effects [1–4]. These are still preliminary considerations, since a long follow-up period is essential to give full scientific support to the situation previously analyzed [13]. IORT certainly allows shorter exposure to radiations of patients’ skin, lung and subcutaneous soft tissues. It also prevents radiation of the controlateral breast, thus reducing the incidence of radiation-related sequels (e.g. fibrosis and
telangiectasias), enhancing also the aesthetic aspect [1–3]. IORT does not interfere with the oncological criteria of conservative surgery: surgical procedure and resection depth and length do not change and the operating time is just slightly longer [1–4]. Now we would like to analyze the findings of our study. First of all, in medical literature we were not able to find any other scientific publications dealing with this topic from a radiological perspective and literature refers only to post-RT mammographic aspects [14,15]. Secondly, the results of this study might be limited by the small number of cases studied. The most important result was about mammographic and echographic alterations affecting patients undergoing IORT. The study showed how in these patients the alterations became progressively more evident than in patients treated with RT, apart from post-operative structural changes
Fig. 2. RT. Echographic pictures. (A) A little liquid collection in surgical site. (B) Lymphatic stasis.
228
S.W. Della Sala et al. / European Journal of Radiology 59 (2006) 222–230
Fig. 3. IORT. Mammographic pictures. Mammography: medio-lateral (A) and axial (B) projection. In the surgical site a radio-transparent area with definited and thickened edge is shown (fat necrosis).
Fig. 4. IORT. Mammographic pictures. Mammography: medio-lateral (A) and axial (B) projection. In the surgical site a radio-transparent area with calcific edge is shown (fat necrosis). Ecography: same case (C).
S.W. Della Sala et al. / European Journal of Radiology 59 (2006) 222–230
229
Fig. 5. IORT. Mammographic pictures. Mammography: medio-lateral (A) projection. In the surgical site a radio-transparent area with partial calcific edge is shown (fat necrosis). Ecography: same case (B).
reported by mammography and oedema grade (showed both by mammography and echography) that were progressively reducing [16,17]. Specifically, mammography performed on patients treated with IORT showed alterations compatible with fat necrosis phenomena with or without associated calcifications. This result is consistent with the IORT surgical procedure previously explained. IORT involves direct radiation of the surgical site, increasing the chances of developing necrosis. Necrosis was more evident some time after the end of treatment. Recently some authors [18] reported the presence of multiple punctate “pseudocalcifications” at the surgical site in patients treated with IORT; we too found similar findings but in our study calcifications are almost always correlated with the presence of fat necrosis. Also the echographic alterations reported in the IORT group were likely related to what we have just explained. Presence of corpuscled collections at the surgical site is consistent with radio-induced necrosis, and this is why they were also reported in the RT group. Evidence of hypoechoic areas around or next to the surgical site is a peculiar aspect of IORT patients. These findings are not easy to understand since their aspect, morphology and structure resemble local recurrences. More than one cytological assessment was necessary to fully understand these findings, eventually showing cellular alterations compatible with necrosis. Modifications during the time course of these findings that in some cases completely develop into actual fluid-corpuscled collections, helped us understand the situation showed by the echographic assessment in radiobiological terms, expressing the grading of radio-induced necrosis. Radio-induced necrosis did not reduce over time [19]. Echography showing lymphatic stasis in RT patients was partially related to radio-dermatitis [16].
We would like to emphasize again that the small number of cases studied might invalidate this analysis; however, differences showed by mammography and echography in both groups (IORT and RT) were significant as chi-square tests (χ2 ) showed. Finally, it is important to note that during follow-up local recurrences developed neither in the IORT nor in the RT group. 5. Conclusions Data from this study showed that radiological alterations in patients treated with IORT are by all means more significant if compared with the alterations developed in patients treated with RT. Specifically, the different mammographic and echographic aspects are more evident, and somewhat of a more difficult interpretation, some time after the end of the treatment, with no substantial changes also at late follow-up visits. This analysis focused on mammographic and echographic aspects only and did not include any assessment of surgical or radiotherapy techniques used. Also, it is necessary to remember that all reports about the therapeutic efficacy of IORT are still under preliminary analysis, since an adequate follow-up period is essential to give full scientific support to this procedure [1,7,13].
References [1] Gatzemeier W, Orecchia R, Intra M, Veronesi U. Intraoperative radiotherapy (IORT) in treatment of breast carcinoma—a new therapeutic alternative within the scope of breast-saving therapy? Current status and future prospects. Report of experiences from European Institute of Oncology. Strahlenther Onkol 2001;177:30–337.
230
S.W. Della Sala et al. / European Journal of Radiology 59 (2006) 222–230
[2] Veronesi U, Orecchia R, Luini A, et al. A preliminary report of intraoperative radiotherapy (IORT) in limited-stage breast cancers that are conservatively treated. Eur J Cancer 2001;37:2178– 83. [3] Intra M, Gatti G, Luini A, et al. Surgical technique of intraoperative radiotherapy in conservative treatment of limited-stage breast cancer. Arch Surg 2002;137:737–40. [4] Reitsamer R, Peintinger F, Sedlmayer F, et al. Intraoperative radiotherapy given as a boost after breast-conserving surgery in breast cancer patients. Eur J Cancer 2002;38:1607–10. [5] Veronesi U, Gatti G, Luini A, et al. Full-dose intraoperative radiotherapy with electrons during breast-conserving surgery. Arch Surg 2003;138:1253–6. [6] Veronesi U, Gatti G, Luini A, et al. Intraoperative radiation therapy for breast cancer: technical notes. Breast J 2003;9:106–12. [7] Bartelink H. Commentary on the paper “A preliminary report of intraoperative radiotherapy (IORT) in limited-stage breast cancers that are conservatively treated”: a critical review of an innovative approach. Eur J Cancer 2001;37:2143–6. [8] Halberg FE, Shank BM, Haffty BG, et al. Conservative surgery and radiation in the treatment of stage I and II carcinoma of the breast. American college of radiology. ACR appropriateness criteria. Radiology 2000;215:1139–205. [9] Hannoun-Levi JM, Marsiglia H, Garbay JR, Gerard JP. Partial irradiation of the breast: why, how? Cancer Radiother 2003;7: 200–9. [10] Polgar C, Fodor J, Major T, et al. Radiotherapy confined to the tumor bed following breast conserving surgery current status, controversies, and future projects. Strahlenther Onkol 2002;178:597–606.
[11] Taghian AG, Powell SN. The role of radiation therapy for primary breast cancer. Surg Clin North Am 1999;79:1091–115. [12] Vicini FA, Kestin L, Chen P, et al. Limited-field radiation therapy in the management of early stage breast cancer. J Natl Cancer Instit 2003;95:1205–10. [13] Reitsamer R, Peintinger F, Kopp M, et al. Local recurrence rates in breast cancer patients treated with intraoperative electronboost radiotherapy versus post-operative external-beam electronboost irradiation. A sequential intervention study. Strahlenther Onkol 2004;180:38–44. [14] Holli K, Saaristo R, Isola J, Hyoty M, Hakama M. Effect of radiotherapy on the interpretation of routine follow-up mammography after conservative breast surgery: a randomized study. Br J Cancer 1998;78:542–5. [15] Dershaw DD. Breast imaging and the conservative treatment of breast cancer. Radiol Clin North Am 2002;40:501–16. [16] Meric F, Buchholz TA, Mirza NQ, et al. Long-term complications associated with breast-conservation surgery and radiotherapy. Ann Surg Oncol 2002;9:543–9. [17] Wazer DE, Lowther D, Boyle T, et al. Clinically evident fat necrosis in women treated with high-dose-rate brachytherapy alone for early stage breast cancer. Int J Radiat Oncol Biol Phys 2001;50:107–11. [18] Curran S, Khilnani A, Zelefsky M, et al. Evaluation of Mammographic findings after intraoperative radiation (IORT) alone for early stage breast cancer. http://rsna2005.rsna.org/rsna2005/v2005/ conference/event display.cfm?em id=4407750. [19] Boyages J, Bilous M, Barraclough B, Landglands AO. Fat necrosis of the breast following lumpectomy and radiation therapy for early breast cancer. Radiother Oncol 1988;13:69–74.