0360-3016/92 $5.00 + .oO Copyright 0 1992 Pergamon Press Ltd.
Inr I Radirrrion Oncology Bwl Ph.v.5 Vol. 23. pp. 915-923 Pnnted in the U.S.A. All rights reserved.
??Clinical Original Contribution
LONG-TERM RADIATION COMPLICATIONS FOLLOWING CONSERVATIVE SURGERY (CS) AND RADIATION THERAPY (RT) IN PATIENTS WITH EARLY STAGE BREAST CANCER SUSAN M. PIERCE,M.D.,’ ABRAM RECHT,M.D.,’ TATIANA I. LINGOS,M.D.,’ ANTHONY ABNER, M.D.,’ FRANK VICINI, M.D.,’ BARBARASILVER,B.A.,’ ANDREW HERZOG, M.D.* AND JAY R. HARRIS,M.D.’ ‘Joint Center for Radiation Therapy and Department of Radiation Oncology, Harvard Medical School, Boston, MA; and ‘Department of Neurology, Beth Israel Hospital, Boston, MA
Thefrequenpy of brachial plexopathy, rib fracture, tissue necrosis, pericarditis, and second non-breast malignancies occurring iI the treatment field among 1624 patients with early stage breast cancer treated with conservative surgery and radiation therapy at the Joint Center for Radiation Therapy between 1968 and 1985 is reported. The median follow-up time for survivors was 79 months (range S-233 months). Brachial plexopathy was related to the use of a third field, the use of chemotherapy and the total dose to the axilla. Brachial plexopathy developed in 20 of 1117 women (1.8%) who received supraclavicular irradiation with or without axilhuy irradiation. The median time to its occurrence was 10.5 months (range 1.5-77 mo), and the majority (80%) of cases completely resolved. Among patients treated with a three-field technique, the incidence of brachial plexopathy was 1.3% (13/991) in patients treated with a dose to the axilla of 5 50 Gy, compared with 5.6% (7/126) in women treated with an axilhuy dose of > 50 Gy. The incidence of brachial plexopathy was 4.5% (15/330) among patients receiving chemotherapy, compared with 0.6% (5/787) when chemotherapy was not used (p < 0.0001). Rib fracture was seen in 29 patients (1.8%), at a median time of 12 months following treatment (range l-57). In all cases, the rib fracture healed without intervention. The incidence of rib fracture was 2.2% (28/X300) among patients treated on a 4 MV linear accelerator, compared with 0.4% (l/276) for patients treated on a 6 or 8 MV machine (p = 0.05). Of patients treated on a 4 MV machine, 0.4% (l/279) developed a rib fracture when a whole breast dose of 45 Gy or less was given, 1.4% (10/725) after receiving between 45 and 50 Gy, and 5.7% (17/296) following 50 Gy or higher. Tissue necrosis requiring surgical correction developed in three patients (0.18%) 22,25, and 114 months after treatment. Presumed pericarditis (requiring hospitalization) was seen in 0.4% of women (3/831) who received radiation therapy to the left breast 2, 2, and 11 months after the start of treatment. Three women (0.18%) developed sarcomas in the treatment field at 72,107, and 110 months, for a lo-year actuarial rate of 0.8%. Two of these sarcomas developed in areas of probable match-line overlap. One patient (0.06%) developed an in-field basal cell carcinoma at 42 months. In conclusion, the risk of significant complications following conservative surgery and radiation therapy for early stage breast cancer is low. Small alterations in treatment, such as using a 6 MV machine and limiting the dose to the whole breast and axilla to 50 Gy or lower, may reduce their occurrence. Breast cancer, Conservative surgery, Radiation therapy, Chemotherapy, Complications, Brachial plexopathy, Rib fracture, Tissue necrosis, Radiation-induced second malignancies.
numbers, several years of follow-up and careful documentation. The reported incidence of severe complications following treatment with CS and RT has been 1% to 3% at institutions that use fractionation schemes and radiation technique consistent with current practice (10, 16, 22, 26, 36, 43, 46). This report details the incidence of brachial plexopathy, rib fracture, tissue necrosis, pericarditis, and second nonbreast infield malignancies occurring in patients with early stage breast cancer treated with CS and RT at this insti-
INTRODUCIION
Conservative surgery (CS) and radiation therapy (RT) have been used with increasing frequency in the treatment of early stage breast cancer over the past two decades. Improvements in radiation technique have evolved from observations of the acute effects of radiation on the breast and the subsequent cosmetic result, as well as analysis of local failure rates. However, the accurate definition of the risk of long-term complications requires large patient Presented in part at the 33rd Annual Meeting of the American Society for Therapeutic DC, November 199 1.
Reprint requests to: Susan M. Pierce, M.D., Joint Center for Radiation Therapy, 50 Binney St., Boston, MA 02 115. Accepted for publication 9 March 1992.
Radiology and Oncology, Washington 91s
916
1. J. Radiation Oncology 0 Biology 0 Physics
tution. We have identified an association between these complications and certain treatment factors and suggest alterations in radiation technique which may reduce their occurrence. METHODS
AND MATERIAL
Between 1968 and 1985, 1624 women with Stage I or II breast cancer were treated with CS + RT at the Joint Center for Radiation Therapy (JCRT). Patients with a previous history of malignancy (except non-invasive carcinoma of the cervix and non-melanoma skin cancer) or synchronous bilateral breast cancer were excluded from this study group. The median age at diagnosis was 5 1 years (range, 25-93). The median follow-up time for survivors was 79 months, with a range of 37-233 months (excluding the nine patients lost to follow-up). Surgery consisted of excisional biopsy, defined as gross removal of the tumor without regard to the microscopic margins of resection in 1593 patients (98%) while 3 1 patients (1.9%) had gross residual disease remaining in the breast after surgery. Axillary dissection was performed in 1193 women (73%) at the discretion of the surgeon, and usually was confined to removal of the level I and II lymph nodes. The remaining 43 1 women did not have an axillary dissection. The median age of the women who did not undergo an axillary dissection was 66 years (range. 2793 years), compared 48 years (range, 25-85) in women having an axillary dissection. External beam radiation therapy was given to the entire breast using either tangential fields alone (n = 507) or a three-field technique (n = 1117) as selected by the individual radiation oncologist. The posterior edges of the tangential fields are made coplanar. When used, the supraclavicular or supraclavicular/axillary field was usually angled slightly off the vertical (approximately 10 degrees) to avoid treating the spinal cord. A standard supraclavicular field extended from the sternal notch medially to the border between the distal third and proximal two thirds of the clavicle laterally. The lateral border of the supraclavicular/axillary field usually extended to include two thirds of the humeral head. Matching of the tangential field and this third field was done using the “hanging block” or “corner block” technique in the majority of cases (39, 45). A posterior or en face axillary boost was used in some women. A “hockey stick” directed at the internal mammary lymph nodes was rarely employed. The dose to the whole breast was 45-46 Gy in the majority of women, with a median dose of 46 Gy (range, 25.2-63 Gy), usually given in fractions of 1.8-2 Gy per day, 5 days per week. The whole breast dose was prescribed to a point on the central axis, measured 1 cm from the posterior border of the tangential fields. In more recent years, this point was moved to 1.5 cm from the posterior border on the central axis. A boost to the primary tumor bed was given by interstitial implantation in 905 women
Volume 23, Number 5, 1992
(56%) or external beam, the majority with electrons, in 595 women (37%). Thirty-eight patients (2%) had a combination boost treatment, and 86 women (5%) had no boost. The median total dose to the primary tumor site was 64.8 Gy (range 44-84 Gy), with 80% of women receiving between 60-70 Gy. The median dose to the supraclavicular field, usually prescribed at a depth of 3 cm, was 46 Gy (range, 19.850.4 Gy), with 96% of women receiving 44-50 Gy. The median dose to the supraclavicular/axillary field, usually prescribed at a depth of 5 cm, was 46 Gy (range, 6.2-75 Gy), with 95% of women receiving 44-55 Gy. Eighty percent of women were treated on a 4 MV linear accelerator. In the earlier years of this study, only women who were found to have a large separation (defined as 26 cm or greater measured along the medial to lateral plane of the posterior border) were treated on an 8 MV linear accelerator. In more recent years, 6 and 8 MV accelerators have been used almost exclusively. Adjuvant chemotherapy was given at the discretion of the medical oncologist to 379 women. This included cyclophosphamide, methotrexate, and 5-fluorouracil (CMF), with or without vincristine or prednisone, in 276 women (72%); doxorubicin and cyclophosphamide in 15 women (4%); and a combination of doxorubicin and CMF-based regimens in 54 women ( 14%). The remaining 34 patients received other drugs. The sequencing of the chemotherapy with the RT has been previously described (28) and was not found to contribute significantly to the complications in this review. The first follow-up visit with the radiation oncologist was usually 3-4 weeks after the completion of RT, then every 3-4 months for the next l-2 years, and semiannually thereafter. The development of complications was recorded at the follow-up examinations. Detailed analyses of the complications radiation pneumonitis and arm edema have been previously published for this patient population (24, 28) and will be briefly reviewed in the discussion of this paper. Brachial plexopathy was defined as the development of new numbness, paresthesias, pain or weakness in the ipsilateral arm after RT. Women who were subsequently found to have an axillary recurrence were excluded from this group. Brachial plexopathy was scored as transient if the symptoms or signs completely resolved, or progressive if they increased in severity or did not resolve with time. Mild brachial plexopathy describes minimal interruption in daily function and severe represents marked impairment. Neurologic consultation and electromyograms (EMG) were obtained at the discretion of the radiation oncologist. Patients were recorded as having a rib fracture when chest wall pain, anatomically related to the rib cage, was present in the treatment field, with or without x-ray confirmation. Tissue necrosis was scored when it was severe enough to require surgical intervention. Patients were presumed to have pericarditis if they developed chest pain
917
Radiation complications in breast cancer 0 S. M. PIERCEet al.
requiring in-patient investigation after receiving RT to the left breast. All time intervals were measured from the start of RT. Patients were not censored from this analysis at the time of recurrent breast cancer, but remained at risk for complications through the time of their last follow-up or their death. The Kaplan-Meier method was used to compute actuarial rates. The Fisher exact test was used to compare differences between crude proportions. All tests of statistical significance were two-tailed with a p-value of 0.05 or less considered significant. RESULTS
Brachial plexopathy Twenty of 1624 patients (1.2%) developed a brachial plexopathy. The median time to its appearance was 10.5 months (range 1S-77 months). The median age of women with brachial plexopathy was 44 years old (range, 36-65), which was younger than that of the entire group. Eighteen (90%) of these 20 women presented with numbness or paresthesias. This was in the ulnar nerve distribution in seven patients, the median nerve region in two women, the upper arm/axilla in three, and in six patients the distribution was not specified. Two patients presented with pain in the arm. Motor dysfunction was found in 10 women (50%), with a decrease in deep tendon reflexes noted in five patients, and three patients were found to have weakness of their hand grip. In 16 of the 20 patients (80%), brachial plexopathy was transient and mild, and resolved within 1 year. One woman had a transient but severe brachial plexopathy, which resolved over 2 years. Three women had progressive and severe involvement. Thus, severe brachial plexus impairment occurred in 4 of the 1624 women in the entire group (0.25%). Electromyography (EMG) and nerve conduction studies were performed in only six patients, including two patients with transient and mild involvement and in all four patients with severe brachial plexopathy. Slowing in brachial plexus nerve conduction velocity across the thoracic outlet segment was demonstrated in all five cases in which it was measured. The EMG ranged from normal
showing loss of functioning motor units. The loss of functioning motor units occurred either because of neural demyelination without axonal damage (i.e., decreased total number of motor units but normal unit morphology), or demyelination with axonal damage (i.e., evidence of acute denervation, in the form of fibrillations, and/or chronic denervation, in the form of decreased total number of motor units with units of increased size and polyphasia). In the first patient with transient and mild brachial plexopathy, slowing across the thoracic outlet was demonstrated for both the median and ulnar nerves. The EMG was normal. The second patient with transient and mild brachial plexopathy did not have brachial plexus conduction velocity measurements, but had low amplitude median and ulnar sensory evoked potentials, consistent with brachial plexus rather than radicular localization of her neurological lesion. Her EMG showed a mild increase in the average size of the motor units in C8 innervated intrinsic hand muscles, with fibrillations and fasciculations in the abductor pollicis brevis. The EMG was normal for C7 muscles. The patient with the transient and severe dysfunction had slow thoracic outlet segment conduction for both the median and ulnar nerves and a moderate decrease in the total number of motor units in C6,7, and 8 segments, without evidence of denervation or reinnervation. Follow-up nerve conduction studies returned to normal over 2 years, which correlated with her subjective and objective improvement. One woman with gradually progressive and severe clinical involvement demonstrated commensurate electrophysiological brachial plexus slowing across the thoracic outlet segment. The EMG was normal 1 year after her symptoms appeared and was not subsequently repeated. The remaining two patients with progressive and severe involvement had slowing in brachial plexus nerve conduction, but EMG data were not obtained. The technique of treatment (2-field vs 3-field), axillary dose, and use of chemotherapy were significantly associated with the development of brachial plexopathy, as shown in Table 1. Of the 1,117 patients treated with a three-field approach, 20 (1.8%) had a brachial plexopathy, compared to none of the 507 patients treated with tangents only (p = 0.0009). The remainder of the analysis was re-
to
Table 1. Incidence of brachial plexopathy in relation to radiation technique, dose, and chemotherapy No CT Technique Two-field Three-field Three-field technique Axillary dose 5 50Gy >50Gy
CT = chemotherapy.
0% (O/458) 0.6% (S/787) -pPOoool-
CT
Total
0% (O/49) 4.5% (15/330)
0% (O/507) 1.8% (20/l 117)
only: 0.4% (3/724) 3.2% (2163) -p=NS
p = 0.05
p = 0.0002 -
3.7% (10/267) b 7.9% (5/63)
p = NS
1.3% (13/991) 5.6% (7/126) p = 0.004
918
1. J. Radiation Oncology 0 Biology0 Physics
Volume 23, Number 5, 1992
Table 2. Incidence of rib fracture in relation to machine energy and dose to the whole breast Whole breast dose 5 45 Gy > 45-< 50 Gy
4MV
6or8MV
0.4% (l/279)
0% (O/131) 0.8% (l/120)
> 50Gy Total
stricted to the patients treated with a three-field technique. The incidence of brachial plexopathy was significantly higher when the axillary dose of radiation therapy was greater than 50 Gy, compared to an axillary dose of 50 Gy or less (5.6% and 1.3%, respectively, p = 0.004). The use of chemotherapy was also associated with an increased rate ofbrachial plexopathy (4.5% for patients treated with adjuvant chemotherapy and 0.6% for patients treated with radiation alone, p < 0.0001). These factors appeared to be additive. For example, a higher incidence of brachial plexopathy was seen when both chemotherapy and a dose greater than 50 Gy were used (7.9%) than when a dose of greater than 50 Gy was used but no chemotherapy was given (3.2%), although this difference was not statistically significant. The timing of chemotherapy and RT (concurrent vs sequential) was not significantly associated with the development of brachial plexopathy. The axillary doses of the four women with severe impairment were 4800, 5100, 5200, and 5350 cGy, which are all higher than the median dose of 4750 cGy for the patients who had a mild brachial plexopathy. One woman with a severe brachial plexopathy received fractions of 2.5 Gy for part of her treatment. The interval to the development of brachial plexopathy was not different for those who had severe plexopathy than for those with mild symptoms. Two of the four patients with severe brachial plexopathy received chemotherapy. Factors that were not found to be significantly associated with the development of brachial plexopathy included the use of an axillary dissection, the number of lymph nodes recovered in the axillary dissection, fraction size ( 1.8 Gy vs 2 Gy), and treatment of the supraclavicular area alone as compared to the supraclavicular plus axillary region. Rib fracture
Twenty-nine of the 1624 patients (1.8%) developed a rib fracture, with a median time to its occurrence of 12 months (range, l-57 months). Radiographic confirmation of the fracture was obtained in 2 1 of the 29 women (73%), and the remaining patients had signs and symptoms consistent with a rib fracture. In 20 patients, plain x-ray revealed the fracture. Six patients had x-rays taken which did not reveal a rib fracture, and three patients did not have an x-ray. A bone scan was obtained in six women; four showed increased uptake corresponding to the x-ray abnormalities, one patient had negative x-rays and a pos-
0% (O/25) * 0.4% (l/276) ??
Total 0.2% (l/410) 1.3% (1 l/845) 5.3% (17/321) I .8% (29/l 576)
itive bone scan, and in one woman both studies were normal. Of the 2 1 patients with radiographic confirmation of a fracture, 11 women had involvement of the anterior rib cage (the medial chest wall and inframammary areas) and 10 had a fracture in the lateral chest wall along the anterior axillary line. The number of ribs involved were detailed in 16 patients, with the majority (94%) of these having more than one rib fractured. The fourth, fifth or sixth rib was involved in 88% of women. We evaluated rib fracture in relation to machine energy and radiation dose to the breast (Table 2). Its occurrence was greater for patients treated with a 4 MV machine energy (2.2%) than for those treated with a 6 or 8 MV machine (0.4%, p = 0.005). Among patients treated with 4 MV, the rate of rib fracture was correlated with the radiation dose given to the whole breast. Patients who received 45 Gy or less had a 0.4% incidence, patients who received more than 45 Gy but less than 50 Gy had a 1.4% incidence: and for women treated with 50 Gy or more, the rate was 5.7%. The difference between less than 50 Gy and 50 Gy or more whole breast dose was statistically significant (r, = 0.0001). Patients who received a whole breast dose of exactly 50 Gy had a significantly lower rib fracture rate (4.3%) than patients who received more than 50 Gy ( 14.6%, p = 0.02). We also assessed the contribution of chemotherapy to the development of a rib fracture (Table 3). With a whole breast dose of less than 50 Gy, the addition of chemotherapy was associated with an increased rate of rib fracture: 0.5% without chemotherapy and 2.3% with chemotherapy (p = 0.01). When 50 Gy or more was given to the whole breast, 4.7% of patients treated with radiation alone developed a rib fracture compared with 7.4% of patients who received adjuvant chemotherapy @ = NS). In all cases, the rib fracture healed without intervention. One patient had subsequent, multiple fractures in the treated area which all eventually healed. Table 3. Incidence of rib fracture in relation to chemotherapy and the dose to the whole breast Whole breast dose
Inr I Radirrrion Oncology Bwl Ph.v.5 Vol. 23. pp. 915-923 Pnnted in the U.S.A. All rights reserved.
??Clinical Original Contribution
LONG-TERM RADIATION COMPLICATIONS FOLLOWING CONSERVATIVE SURGERY (CS) AND RADIATION THERAPY (RT) IN PATIENTS WITH EARLY STAGE BREAST CANCER SUSAN M. PIERCE,M.D.,’ ABRAM RECHT,M.D.,’ TATIANA I. LINGOS,M.D.,’ ANTHONY ABNER, M.D.,’ FRANK VICINI, M.D.,’ BARBARASILVER,B.A.,’ ANDREW HERZOG, M.D.* AND JAY R. HARRIS,M.D.’ ‘Joint Center for Radiation Therapy and Department of Radiation Oncology, Harvard Medical School, Boston, MA; and ‘Department of Neurology, Beth Israel Hospital, Boston, MA
Thefrequenpy of brachial plexopathy, rib fracture, tissue necrosis, pericarditis, and second non-breast malignancies occurring iI the treatment field among 1624 patients with early stage breast cancer treated with conservative surgery and radiation therapy at the Joint Center for Radiation Therapy between 1968 and 1985 is reported. The median follow-up time for survivors was 79 months (range S-233 months). Brachial plexopathy was related to the use of a third field, the use of chemotherapy and the total dose to the axilla. Brachial plexopathy developed in 20 of 1117 women (1.8%) who received supraclavicular irradiation with or without axilhuy irradiation. The median time to its occurrence was 10.5 months (range 1.5-77 mo), and the majority (80%) of cases completely resolved. Among patients treated with a three-field technique, the incidence of brachial plexopathy was 1.3% (13/991) in patients treated with a dose to the axilla of 5 50 Gy, compared with 5.6% (7/126) in women treated with an axilhuy dose of > 50 Gy. The incidence of brachial plexopathy was 4.5% (15/330) among patients receiving chemotherapy, compared with 0.6% (5/787) when chemotherapy was not used (p < 0.0001). Rib fracture was seen in 29 patients (1.8%), at a median time of 12 months following treatment (range l-57). In all cases, the rib fracture healed without intervention. The incidence of rib fracture was 2.2% (28/X300) among patients treated on a 4 MV linear accelerator, compared with 0.4% (l/276) for patients treated on a 6 or 8 MV machine (p = 0.05). Of patients treated on a 4 MV machine, 0.4% (l/279) developed a rib fracture when a whole breast dose of 45 Gy or less was given, 1.4% (10/725) after receiving between 45 and 50 Gy, and 5.7% (17/296) following 50 Gy or higher. Tissue necrosis requiring surgical correction developed in three patients (0.18%) 22,25, and 114 months after treatment. Presumed pericarditis (requiring hospitalization) was seen in 0.4% of women (3/831) who received radiation therapy to the left breast 2, 2, and 11 months after the start of treatment. Three women (0.18%) developed sarcomas in the treatment field at 72,107, and 110 months, for a lo-year actuarial rate of 0.8%. Two of these sarcomas developed in areas of probable match-line overlap. One patient (0.06%) developed an in-field basal cell carcinoma at 42 months. In conclusion, the risk of significant complications following conservative surgery and radiation therapy for early stage breast cancer is low. Small alterations in treatment, such as using a 6 MV machine and limiting the dose to the whole breast and axilla to 50 Gy or lower, may reduce their occurrence. Breast cancer, Conservative surgery, Radiation therapy, Chemotherapy, Complications, Brachial plexopathy, Rib fracture, Tissue necrosis, Radiation-induced second malignancies.
numbers, several years of follow-up and careful documentation. The reported incidence of severe complications following treatment with CS and RT has been 1% to 3% at institutions that use fractionation schemes and radiation technique consistent with current practice (10, 16, 22, 26, 36, 43, 46). This report details the incidence of brachial plexopathy, rib fracture, tissue necrosis, pericarditis, and second nonbreast infield malignancies occurring in patients with early stage breast cancer treated with CS and RT at this insti-
INTRODUCIION
Conservative surgery (CS) and radiation therapy (RT) have been used with increasing frequency in the treatment of early stage breast cancer over the past two decades. Improvements in radiation technique have evolved from observations of the acute effects of radiation on the breast and the subsequent cosmetic result, as well as analysis of local failure rates. However, the accurate definition of the risk of long-term complications requires large patient Presented in part at the 33rd Annual Meeting of the American Society for Therapeutic DC, November 199 1.
Reprint requests to: Susan M. Pierce, M.D., Joint Center for Radiation Therapy, 50 Binney St., Boston, MA 02 115. Accepted for publication 9 March 1992.
Radiology and Oncology, Washington 91s
916
1. J. Radiation Oncology 0 Biology 0 Physics
tution. We have identified an association between these complications and certain treatment factors and suggest alterations in radiation technique which may reduce their occurrence. METHODS
AND MATERIAL
Between 1968 and 1985, 1624 women with Stage I or II breast cancer were treated with CS + RT at the Joint Center for Radiation Therapy (JCRT). Patients with a previous history of malignancy (except non-invasive carcinoma of the cervix and non-melanoma skin cancer) or synchronous bilateral breast cancer were excluded from this study group. The median age at diagnosis was 5 1 years (range, 25-93). The median follow-up time for survivors was 79 months, with a range of 37-233 months (excluding the nine patients lost to follow-up). Surgery consisted of excisional biopsy, defined as gross removal of the tumor without regard to the microscopic margins of resection in 1593 patients (98%) while 3 1 patients (1.9%) had gross residual disease remaining in the breast after surgery. Axillary dissection was performed in 1193 women (73%) at the discretion of the surgeon, and usually was confined to removal of the level I and II lymph nodes. The remaining 43 1 women did not have an axillary dissection. The median age of the women who did not undergo an axillary dissection was 66 years (range. 2793 years), compared 48 years (range, 25-85) in women having an axillary dissection. External beam radiation therapy was given to the entire breast using either tangential fields alone (n = 507) or a three-field technique (n = 1117) as selected by the individual radiation oncologist. The posterior edges of the tangential fields are made coplanar. When used, the supraclavicular or supraclavicular/axillary field was usually angled slightly off the vertical (approximately 10 degrees) to avoid treating the spinal cord. A standard supraclavicular field extended from the sternal notch medially to the border between the distal third and proximal two thirds of the clavicle laterally. The lateral border of the supraclavicular/axillary field usually extended to include two thirds of the humeral head. Matching of the tangential field and this third field was done using the “hanging block” or “corner block” technique in the majority of cases (39, 45). A posterior or en face axillary boost was used in some women. A “hockey stick” directed at the internal mammary lymph nodes was rarely employed. The dose to the whole breast was 45-46 Gy in the majority of women, with a median dose of 46 Gy (range, 25.2-63 Gy), usually given in fractions of 1.8-2 Gy per day, 5 days per week. The whole breast dose was prescribed to a point on the central axis, measured 1 cm from the posterior border of the tangential fields. In more recent years, this point was moved to 1.5 cm from the posterior border on the central axis. A boost to the primary tumor bed was given by interstitial implantation in 905 women
Volume 23, Number 5, 1992
(56%) or external beam, the majority with electrons, in 595 women (37%). Thirty-eight patients (2%) had a combination boost treatment, and 86 women (5%) had no boost. The median total dose to the primary tumor site was 64.8 Gy (range 44-84 Gy), with 80% of women receiving between 60-70 Gy. The median dose to the supraclavicular field, usually prescribed at a depth of 3 cm, was 46 Gy (range, 19.850.4 Gy), with 96% of women receiving 44-50 Gy. The median dose to the supraclavicular/axillary field, usually prescribed at a depth of 5 cm, was 46 Gy (range, 6.2-75 Gy), with 95% of women receiving 44-55 Gy. Eighty percent of women were treated on a 4 MV linear accelerator. In the earlier years of this study, only women who were found to have a large separation (defined as 26 cm or greater measured along the medial to lateral plane of the posterior border) were treated on an 8 MV linear accelerator. In more recent years, 6 and 8 MV accelerators have been used almost exclusively. Adjuvant chemotherapy was given at the discretion of the medical oncologist to 379 women. This included cyclophosphamide, methotrexate, and 5-fluorouracil (CMF), with or without vincristine or prednisone, in 276 women (72%); doxorubicin and cyclophosphamide in 15 women (4%); and a combination of doxorubicin and CMF-based regimens in 54 women ( 14%). The remaining 34 patients received other drugs. The sequencing of the chemotherapy with the RT has been previously described (28) and was not found to contribute significantly to the complications in this review. The first follow-up visit with the radiation oncologist was usually 3-4 weeks after the completion of RT, then every 3-4 months for the next l-2 years, and semiannually thereafter. The development of complications was recorded at the follow-up examinations. Detailed analyses of the complications radiation pneumonitis and arm edema have been previously published for this patient population (24, 28) and will be briefly reviewed in the discussion of this paper. Brachial plexopathy was defined as the development of new numbness, paresthesias, pain or weakness in the ipsilateral arm after RT. Women who were subsequently found to have an axillary recurrence were excluded from this group. Brachial plexopathy was scored as transient if the symptoms or signs completely resolved, or progressive if they increased in severity or did not resolve with time. Mild brachial plexopathy describes minimal interruption in daily function and severe represents marked impairment. Neurologic consultation and electromyograms (EMG) were obtained at the discretion of the radiation oncologist. Patients were recorded as having a rib fracture when chest wall pain, anatomically related to the rib cage, was present in the treatment field, with or without x-ray confirmation. Tissue necrosis was scored when it was severe enough to require surgical intervention. Patients were presumed to have pericarditis if they developed chest pain
917
Radiation complications in breast cancer 0 S. M. PIERCEet al.
requiring in-patient investigation after receiving RT to the left breast. All time intervals were measured from the start of RT. Patients were not censored from this analysis at the time of recurrent breast cancer, but remained at risk for complications through the time of their last follow-up or their death. The Kaplan-Meier method was used to compute actuarial rates. The Fisher exact test was used to compare differences between crude proportions. All tests of statistical significance were two-tailed with a p-value of 0.05 or less considered significant. RESULTS
Brachial plexopathy Twenty of 1624 patients (1.2%) developed a brachial plexopathy. The median time to its appearance was 10.5 months (range 1S-77 months). The median age of women with brachial plexopathy was 44 years old (range, 36-65), which was younger than that of the entire group. Eighteen (90%) of these 20 women presented with numbness or paresthesias. This was in the ulnar nerve distribution in seven patients, the median nerve region in two women, the upper arm/axilla in three, and in six patients the distribution was not specified. Two patients presented with pain in the arm. Motor dysfunction was found in 10 women (50%), with a decrease in deep tendon reflexes noted in five patients, and three patients were found to have weakness of their hand grip. In 16 of the 20 patients (80%), brachial plexopathy was transient and mild, and resolved within 1 year. One woman had a transient but severe brachial plexopathy, which resolved over 2 years. Three women had progressive and severe involvement. Thus, severe brachial plexus impairment occurred in 4 of the 1624 women in the entire group (0.25%). Electromyography (EMG) and nerve conduction studies were performed in only six patients, including two patients with transient and mild involvement and in all four patients with severe brachial plexopathy. Slowing in brachial plexus nerve conduction velocity across the thoracic outlet segment was demonstrated in all five cases in which it was measured. The EMG ranged from normal
showing loss of functioning motor units. The loss of functioning motor units occurred either because of neural demyelination without axonal damage (i.e., decreased total number of motor units but normal unit morphology), or demyelination with axonal damage (i.e., evidence of acute denervation, in the form of fibrillations, and/or chronic denervation, in the form of decreased total number of motor units with units of increased size and polyphasia). In the first patient with transient and mild brachial plexopathy, slowing across the thoracic outlet was demonstrated for both the median and ulnar nerves. The EMG was normal. The second patient with transient and mild brachial plexopathy did not have brachial plexus conduction velocity measurements, but had low amplitude median and ulnar sensory evoked potentials, consistent with brachial plexus rather than radicular localization of her neurological lesion. Her EMG showed a mild increase in the average size of the motor units in C8 innervated intrinsic hand muscles, with fibrillations and fasciculations in the abductor pollicis brevis. The EMG was normal for C7 muscles. The patient with the transient and severe dysfunction had slow thoracic outlet segment conduction for both the median and ulnar nerves and a moderate decrease in the total number of motor units in C6,7, and 8 segments, without evidence of denervation or reinnervation. Follow-up nerve conduction studies returned to normal over 2 years, which correlated with her subjective and objective improvement. One woman with gradually progressive and severe clinical involvement demonstrated commensurate electrophysiological brachial plexus slowing across the thoracic outlet segment. The EMG was normal 1 year after her symptoms appeared and was not subsequently repeated. The remaining two patients with progressive and severe involvement had slowing in brachial plexus nerve conduction, but EMG data were not obtained. The technique of treatment (2-field vs 3-field), axillary dose, and use of chemotherapy were significantly associated with the development of brachial plexopathy, as shown in Table 1. Of the 1,117 patients treated with a three-field approach, 20 (1.8%) had a brachial plexopathy, compared to none of the 507 patients treated with tangents only (p = 0.0009). The remainder of the analysis was re-
to
Table 1. Incidence of brachial plexopathy in relation to radiation technique, dose, and chemotherapy No CT Technique Two-field Three-field Three-field technique Axillary dose 5 50Gy >50Gy
CT = chemotherapy.
0% (O/458) 0.6% (S/787) -pPOoool-
CT
Total
0% (O/49) 4.5% (15/330)
0% (O/507) 1.8% (20/l 117)
only: 0.4% (3/724) 3.2% (2163) -p=NS
p = 0.05
p = 0.0002 -
3.7% (10/267) b 7.9% (5/63)
p = NS
1.3% (13/991) 5.6% (7/126) p = 0.004
918
1. J. Radiation Oncology 0 Biology0 Physics
Volume 23, Number 5, 1992
Table 2. Incidence of rib fracture in relation to machine energy and dose to the whole breast Whole breast dose 5 45 Gy > 45-< 50 Gy
4MV
6or8MV
0.4% (l/279)
0% (O/131) 0.8% (l/120)
> 50Gy Total
stricted to the patients treated with a three-field technique. The incidence of brachial plexopathy was significantly higher when the axillary dose of radiation therapy was greater than 50 Gy, compared to an axillary dose of 50 Gy or less (5.6% and 1.3%, respectively, p = 0.004). The use of chemotherapy was also associated with an increased rate ofbrachial plexopathy (4.5% for patients treated with adjuvant chemotherapy and 0.6% for patients treated with radiation alone, p < 0.0001). These factors appeared to be additive. For example, a higher incidence of brachial plexopathy was seen when both chemotherapy and a dose greater than 50 Gy were used (7.9%) than when a dose of greater than 50 Gy was used but no chemotherapy was given (3.2%), although this difference was not statistically significant. The timing of chemotherapy and RT (concurrent vs sequential) was not significantly associated with the development of brachial plexopathy. The axillary doses of the four women with severe impairment were 4800, 5100, 5200, and 5350 cGy, which are all higher than the median dose of 4750 cGy for the patients who had a mild brachial plexopathy. One woman with a severe brachial plexopathy received fractions of 2.5 Gy for part of her treatment. The interval to the development of brachial plexopathy was not different for those who had severe plexopathy than for those with mild symptoms. Two of the four patients with severe brachial plexopathy received chemotherapy. Factors that were not found to be significantly associated with the development of brachial plexopathy included the use of an axillary dissection, the number of lymph nodes recovered in the axillary dissection, fraction size ( 1.8 Gy vs 2 Gy), and treatment of the supraclavicular area alone as compared to the supraclavicular plus axillary region. Rib fracture
Twenty-nine of the 1624 patients (1.8%) developed a rib fracture, with a median time to its occurrence of 12 months (range, l-57 months). Radiographic confirmation of the fracture was obtained in 2 1 of the 29 women (73%), and the remaining patients had signs and symptoms consistent with a rib fracture. In 20 patients, plain x-ray revealed the fracture. Six patients had x-rays taken which did not reveal a rib fracture, and three patients did not have an x-ray. A bone scan was obtained in six women; four showed increased uptake corresponding to the x-ray abnormalities, one patient had negative x-rays and a pos-
0% (O/25) * 0.4% (l/276) ??
Total 0.2% (l/410) 1.3% (1 l/845) 5.3% (17/321) I .8% (29/l 576)
itive bone scan, and in one woman both studies were normal. Of the 2 1 patients with radiographic confirmation of a fracture, 11 women had involvement of the anterior rib cage (the medial chest wall and inframammary areas) and 10 had a fracture in the lateral chest wall along the anterior axillary line. The number of ribs involved were detailed in 16 patients, with the majority (94%) of these having more than one rib fractured. The fourth, fifth or sixth rib was involved in 88% of women. We evaluated rib fracture in relation to machine energy and radiation dose to the breast (Table 2). Its occurrence was greater for patients treated with a 4 MV machine energy (2.2%) than for those treated with a 6 or 8 MV machine (0.4%, p = 0.005). Among patients treated with 4 MV, the rate of rib fracture was correlated with the radiation dose given to the whole breast. Patients who received 45 Gy or less had a 0.4% incidence, patients who received more than 45 Gy but less than 50 Gy had a 1.4% incidence: and for women treated with 50 Gy or more, the rate was 5.7%. The difference between less than 50 Gy and 50 Gy or more whole breast dose was statistically significant (r, = 0.0001). Patients who received a whole breast dose of exactly 50 Gy had a significantly lower rib fracture rate (4.3%) than patients who received more than 50 Gy ( 14.6%, p = 0.02). We also assessed the contribution of chemotherapy to the development of a rib fracture (Table 3). With a whole breast dose of less than 50 Gy, the addition of chemotherapy was associated with an increased rate of rib fracture: 0.5% without chemotherapy and 2.3% with chemotherapy (p = 0.01). When 50 Gy or more was given to the whole breast, 4.7% of patients treated with radiation alone developed a rib fracture compared with 7.4% of patients who received adjuvant chemotherapy @ = NS). In all cases, the rib fracture healed without intervention. One patient had subsequent, multiple fractures in the treated area which all eventually healed. Table 3. Incidence of rib fracture in relation to chemotherapy and the dose to the whole breast Whole breast dose
