- Email: [email protected]
Second non-breast malignancies after conservative surgery (CS) and radiation therapy (RT) for early-stage breast cancer
Proceedings
of the 42nd Annual ASTRO Meeting
145
had node-negative disease. There was no significant difference in the distribution of pathological and treatment variables between patients treated for left- and right-sided cancers. There were 70 MI events among 56 patients following breast irradiation. According to MONICA criteria 53. 6 and I I were characterized as definite, possible or definitely not MI. Twenty-six patients treated for left-sided and twenty-three patients treated for right-sided breast cancer experienced at least one definite or possible MI (Log-rank, p = 0.66). The incidence of MI in the study cohort was comparable to that in the age-matched general population of women in Ontario.
Conclusion:
67
Radiotherapy
following
breast conserving
Second cancers after conservative women at increased risk
B. Fowble. A. Hanlon,
G. Freedman,
surgery and radiation
N. Nicolaou,
Fox Chase Cancer Cerrtrc Philndelphitr,
surgery is not associated
with an increased risk of fatal or non-fatal MI.
for stage I-II breast cancer: identifying
a subset of
P. Anderson
PA
Purpose: Second malignancies (SM) following radiation for breast cancer are of concern especially when they can be attributed to treatment. Information regarding these second cancers has often been obtained from population-based studies in which independent risk factors are not assessed and older treatment techniques have been employed. Randomized trials comparing conservative surgery and radiation to mastectomy have provided little information on second cancers. This study assesses the risk and patterns of a SM in a group of women treated in a relatively contemporary manner and identifies a subgroup of these women at increased risk for a second cancer. Materials and Methods: From 1978-1994, 1253 women with unilateral stage I-II breast cancer underwent wide excision. axillary dissection and radiation. Median followup was 8.9 years with 446 pts. followed for 2 IO years. Median age was 55 yrs. (range 22-88). 68% had TI tumors and 73% were axillary node negative. Radiation was directed to the breast only in 78%. Ad.juvant therapy consisted of chemotherapy in 19%, tamoxifen in 19%. and both in 7%. Factors analyzed for their association with the cumulative incidence (CI) of second malignancies (SM), contralateral breast cancer (CBC) and non-breast cancer malignancy (NBCM) were age. menopausal status, race, family history, obesity. smoking; tumor size. location. histology, pathologic nodal status: region(s) treated with radiation and the ube and type of adjuvant therapy. Results: 182women developed a SM (93 CBC at median interval 5.8 yrs., range .3-15.2 and 98 NBCM at median interval 7.2 yrs., range .3-l 8.4). Nine women had both a CBC and NBCM. The most common NBCM were skin cancers (29 pts.), gynecologic (19 pts.), and gastrointestinal (18 pts,) malignancies. There were 2 breast sarcomas. Table I presents the cumulative incidence (CI) of all SM, CBC and NBCM 2 skin cancers at 5 and IO yrs. The 5 and 10 yr. CI of a SM were 5% and 16%. Patient age was a signiticant factor for all SM, CBC, and NBCM. Young age was associated with an increased risk of CBC while older age increased the risk of a NBCM, 26% of which were skin cancers. Family history of breast cancer increased the risk of all SM and CBC but not NBCM. The risk of a CBC increased as the number of affected relatives increased but NBCM was not aigniticantly affected. LCIS was not a significant factor for CBC. The 5 and IO yr. CI for leukemia and lung cancer were .0X% and .2% and 0.8% and 1%. There was no significant effect of chemotherapy or region(s) treated with radiation on SM. CBC, or NBCM. Tamoxifen resulted in a nonsignificant decrease in CBC and an increase in NBCM. Summary: The IO yr. CI of a second cancer in this study is 16%. Only patient related factors were significantly associated with the risk of a SM. Young age and family history predicted for an increased risk of CBC and older age predicted for an increased risk of NBCM. We could not identify any treatment related factors which significantly contributed to this risk. 5 and (IO) Year Cumulative Incidence 13”Second Malignancy
68
secon d non-breast breast cancer
S. Galper,‘.i
R. Gelman,‘.’
malignancies
A. Recht,‘,’
All SM
CBC
NBCM
NBCM excluding skin
5 (16) .00035 5 (1%
3 (71 .03 3(13)
3 (8) <.OOOl
<.OOOl
5 (12) J(l3) 7 (27) .00X s (1.1) 601)
4 (8)
1(5) 0 (0)
I (5)
I(I) I (4) 3 (8)
3 (8) <.OOOl 2 (4) 4(13)
5 (181 .83 3 (8) 2 (8)
3 (II) .98 2 (5) I (5)
5 (20)
3 (IO)
7(1X) I3 (39) 7 (22)
6 (14) I I (37)
2 (‘)I I (6)
after conservative
surgery (CS) and radiation
4(l7)
B. Silver,’ J. S. Wang.“,’
S. J. Schnitt,’
2(5)
2(6)
3 (3)
0 (3) 0 (0)
I (8)
I (5)
therapy (RT) for early-stage
J. L. Connolly,’
J. R. Harris.‘,5
‘Brighnm md Womw‘s Hmpitul. Boston, MA, “Daru-Ferber Cancer Institute, Boston, MA, .‘Beth I.wnel Deaconess Medico1 Center, Boston. MA, ‘Mrtro West Medico1 Center, Boston. MA, SJoint Center for Radiation Therapy, Boston, MA Purpose: To determine the incidence (SNBM-RT) among longterm survivors as first cancers in the SEER population.
of all second non-breast malignancies (SNBM) and SNBM in the radiation field and to compare the risk of these events to the age-specific incidence of malignancies
146
I. .I. Radiation
Oncology
l Biology
l Physics
Volume 48, Number 3, Supplement,
2000
Materials and Methods: We analyzed the likelihood of SNBM for 1,884 patients with clinical stage I-II breast cancer treated with gross excision and 260 Gy (median 63) to the breast between 1970-1987.58% received supracIavicular/axillary radiation (median dose 45 Gy, range 20-60 Gy) and 28% received systemic therapy. The median age at diagnosis was 52. Median clinical tumor size was 2 cm. Patients were considered at risk for a SNBM until the first of distant metastases, contralateral breast cancer, or death or, if alive and disease-free, until last follow up. SNBM-RT were defined as any invasive tumors of the lung, bronchus, pleura, mediastinum, trachea, esophagus, larynx and thyroid. Expected numbers of cancers were obtained from the SEER databases for 1982-1986, using age-specific incidence for white females within 5-year age groups (p-values from exact Poisson). The median time at risk for a SNBM was 10.9 years (range 0.2-27.9). Results: 151 SNBM sites were observed in 147 patients: 33 lung, 23 colorectal, 1 esophagus, 9 other GI, 12 ovary, 16 other GYN, 14 melanoma, 8 lymphoma, 7 sarcoma, 6 meningioma, 4 urinary tract, 3 leukemia, 2 neuroendocrine, 2 multiple myeloma, 2 mesothelioma, 2 head and neck, 2 GBM, 1 astrocytoma, 1 neuroectodermal tumor, 1 polycythemia Vera, 1 thyroid and 1 unknown primary. Within the first 5 years, observed and expected SNBM were identical (47 vs 46.9). After 5 years, 24% more SNBM were observed than expected (100 vs 80.8, p = 0.02). Among patients < 50 years old at diagnosis, there were 43% more observed SNBM than expected (40 vs 28, p = 0.02). For patients 2 50, there were 7% more SNBM observed than expected (107 vs 99.7, p = 0.25). The observed incidence of ovarian cancer was significantly greater than expected among patients < 50 (7 vs 196, p = 0.004) but was not different than expected for patients 2 50 (5 vs 5.3, p = 0.61). SNBM-RT have been observed in 35, 33% more than predicted by SEER (p = 0.06). Among the 7 sarcomas, 1 developed in the ipsilateral shoulder, 3 in the ipsilateral breast, and 1 in the ipsilateral pleura. Four of these received three-field radiation using an old matchline technique and were noted to have matchline fibrosis with subsequent development of the sarcoma along the matchline. Conclusions: Patients with breast cancer treated with CS and RT are at increased risk of second non-breast malignancies compared to an age-matched population. This risk increases with time and is higher in younger patients. Some of the increased risk appears related to RT, and some may be related to underlying genetic factors seen in patients with breast cancer. In particular, there was an increased incidence of in-field SNBM-RT. However, the current technique of matching the axillarysupraclavicular field to the breast tangents may reduce the incidence of radiation-induced sarcomas.
69
The role of endorectal
D. H. Clarke,’ D. N. Blair’
S. J. Banks,’
‘Inova Alexandria
coil MRI in patient selection and treatment
A. R. Wiederhorn,’
Cancer Center, Alenandriu,
planning for prostate seed implants
J. W. Klousia,* .I. M. Lissy,’ A. A. Able,’ C. Artilles,’ VA, ‘Inova Alexandria
Hospital, Alexandria,
W. V. Hindle,’
VA
Based on data by D’Amico regarding the staging efficacy of endorectal coil MRI (erMRI), we initiated a permanent prostate seed implant program utilizing erMR1 as a routine staging procedure. This presentation will summarize our staging results, the impact of erMR1 on patient selection and the value of erMR1 in pre-planning an optimal implant. Between lo/94 and 12/98, more than 1,000 erMR1 studies were performed on prostate cancer patients evaluated in radiation oncology; and 390 prostate seed implants (99% Pd-103, 1% I-125) were performed. 327/390 seed patients (84%) had an erMR1. 239/327 patients (73%) were upstaged and 84/327 patients (26%) were upstaged to T3 with 8/327 patients (2.5%) being T3c. 51/327 patients (16%) had perineural invasion and this strongly correlated with T3 stage that was seen in 31/51 patients (61%). Of the 84% without perineural invasion, only 24% had T3 disease by erMR1. erMR1 findings had a major impact on treatment. 65/327 patients (20%) had a change in the overall treatment recommendation; i.e. combined external beam and seeds versus monotherapy. 185/327 patients (57%) had a change in seed distribution. Most of these changes related to pre-planned extracapsular coverage or seminal vesicle implantation. Between lo/94 and 12/96, 205 patients were implanted with a mean follow-up of 37 months. 72% of these patients had pre-treatment erMR1. 57% received combined therapy including external RT and seeds and 37% received hormonal downsizing. Only patients with documented PSA failure received hormonal therapy post-implant. Despite the fact that 39% had a Gleason Score (GS) of 7 or higher, there have been only 17 PSA failures. 15/17 patients had PSA doubling times of less than 6 months; the 1 patient with a doubling time greater than 24 months had a positive prostascint scan in paraaortic lymph nodes. There have been no documented local recurrences. D’Amico previously defined an intermediate risk group of patients (PSA 4 - 10 and GS 5 - 7, PSA 0 - 4 and GS 7, and PSA 10 _ 20 and GS 2 ~ 7) who also had extracapsular extension (ECE) or seminal vesicle invasion @VI) on erMRI. Following radical prostatectomy the PSA progression rate was 79% at 36 months. In our series, we identified 26 patients who fell into the intermediate risk group and had T3 disease on erMR1. Following external radiation and seeds, our PSA progression rate at 35 months was zero. We believe that our extensive use of erMR1 as a routine staging procedure has significantly improved our ability to select appropriate candidates for seeding and also has been instrumental in identifying patients who would be better treated with combined therapy. erMR1 has had a dramatic influence on seed distribution allowing for accurate implantation of periprostatic disease, including tumor involvement of the neurovascular bundle and the proximal seminal vesicles. Furthermore, erMR1 has also enabled us to plan for a ‘boost’ within the implanted volume by precisely delineating the location of bulky tumor. We believe our excellent results are largely attributable to our use of erMR1.
70
Ten year biochemical
outcomes
following
1254odine
monotherapy
for early stage prostate cancer
P. D. Grimm, J. C. Blasko, J. E. Sylvester, W. A. Cavanagh, R. Meier Seattle Prostate Institute, Seattle, WA One-hundred twenty-six (126) patients who presented with well to moderately differentiated carcinoma of the prostate between January 1988 and December 1990 were treated by the ‘Seattle‘ method of prostate brachytherapy. One patient was lost to follow-up, leaving 125 for PSA-based analysis. Patients were staged clinically by DRE as follows: Tl 30 (24%); T2a - 77 (62%); T2b - 17 (14%); T2c - 1 patient. All patients presented with a combined Gleason score less than 7. Median serum PSA at presentation was 5.1 ng/mI; mean 8.1; 96 (77%) presented with serum PSA less than 10 ng/ml. Median PSA-based follow-up time is 8 14 months, 94.4 months when considering only living, non-failed patients. PSA
of the 42nd Annual ASTRO Meeting
145
had node-negative disease. There was no significant difference in the distribution of pathological and treatment variables between patients treated for left- and right-sided cancers. There were 70 MI events among 56 patients following breast irradiation. According to MONICA criteria 53. 6 and I I were characterized as definite, possible or definitely not MI. Twenty-six patients treated for left-sided and twenty-three patients treated for right-sided breast cancer experienced at least one definite or possible MI (Log-rank, p = 0.66). The incidence of MI in the study cohort was comparable to that in the age-matched general population of women in Ontario.
Conclusion:
67
Radiotherapy
following
breast conserving
Second cancers after conservative women at increased risk
B. Fowble. A. Hanlon,
G. Freedman,
surgery and radiation
N. Nicolaou,
Fox Chase Cancer Cerrtrc Philndelphitr,
surgery is not associated
with an increased risk of fatal or non-fatal MI.
for stage I-II breast cancer: identifying
a subset of
P. Anderson
PA
Purpose: Second malignancies (SM) following radiation for breast cancer are of concern especially when they can be attributed to treatment. Information regarding these second cancers has often been obtained from population-based studies in which independent risk factors are not assessed and older treatment techniques have been employed. Randomized trials comparing conservative surgery and radiation to mastectomy have provided little information on second cancers. This study assesses the risk and patterns of a SM in a group of women treated in a relatively contemporary manner and identifies a subgroup of these women at increased risk for a second cancer. Materials and Methods: From 1978-1994, 1253 women with unilateral stage I-II breast cancer underwent wide excision. axillary dissection and radiation. Median followup was 8.9 years with 446 pts. followed for 2 IO years. Median age was 55 yrs. (range 22-88). 68% had TI tumors and 73% were axillary node negative. Radiation was directed to the breast only in 78%. Ad.juvant therapy consisted of chemotherapy in 19%, tamoxifen in 19%. and both in 7%. Factors analyzed for their association with the cumulative incidence (CI) of second malignancies (SM), contralateral breast cancer (CBC) and non-breast cancer malignancy (NBCM) were age. menopausal status, race, family history, obesity. smoking; tumor size. location. histology, pathologic nodal status: region(s) treated with radiation and the ube and type of adjuvant therapy. Results: 182women developed a SM (93 CBC at median interval 5.8 yrs., range .3-15.2 and 98 NBCM at median interval 7.2 yrs., range .3-l 8.4). Nine women had both a CBC and NBCM. The most common NBCM were skin cancers (29 pts.), gynecologic (19 pts.), and gastrointestinal (18 pts,) malignancies. There were 2 breast sarcomas. Table I presents the cumulative incidence (CI) of all SM, CBC and NBCM 2 skin cancers at 5 and IO yrs. The 5 and 10 yr. CI of a SM were 5% and 16%. Patient age was a signiticant factor for all SM, CBC, and NBCM. Young age was associated with an increased risk of CBC while older age increased the risk of a NBCM, 26% of which were skin cancers. Family history of breast cancer increased the risk of all SM and CBC but not NBCM. The risk of a CBC increased as the number of affected relatives increased but NBCM was not aigniticantly affected. LCIS was not a significant factor for CBC. The 5 and IO yr. CI for leukemia and lung cancer were .0X% and .2% and 0.8% and 1%. There was no significant effect of chemotherapy or region(s) treated with radiation on SM. CBC, or NBCM. Tamoxifen resulted in a nonsignificant decrease in CBC and an increase in NBCM. Summary: The IO yr. CI of a second cancer in this study is 16%. Only patient related factors were significantly associated with the risk of a SM. Young age and family history predicted for an increased risk of CBC and older age predicted for an increased risk of NBCM. We could not identify any treatment related factors which significantly contributed to this risk. 5 and (IO) Year Cumulative Incidence 13”Second Malignancy
68
secon d non-breast breast cancer
S. Galper,‘.i
R. Gelman,‘.’
malignancies
A. Recht,‘,’
All SM
CBC
NBCM
NBCM excluding skin
5 (16) .00035 5 (1%
3 (71 .03 3(13)
3 (8) <.OOOl
<.OOOl
5 (12) J(l3) 7 (27) .00X s (1.1) 601)
4 (8)
1(5) 0 (0)
I (5)
I(I) I (4) 3 (8)
3 (8) <.OOOl 2 (4) 4(13)
5 (181 .83 3 (8) 2 (8)
3 (II) .98 2 (5) I (5)
5 (20)
3 (IO)
7(1X) I3 (39) 7 (22)
6 (14) I I (37)
2 (‘)I I (6)
after conservative
surgery (CS) and radiation
4(l7)
B. Silver,’ J. S. Wang.“,’
S. J. Schnitt,’
2(5)
2(6)
3 (3)
0 (3) 0 (0)
I (8)
I (5)
therapy (RT) for early-stage
J. L. Connolly,’
J. R. Harris.‘,5
‘Brighnm md Womw‘s Hmpitul. Boston, MA, “Daru-Ferber Cancer Institute, Boston, MA, .‘Beth I.wnel Deaconess Medico1 Center, Boston. MA, ‘Mrtro West Medico1 Center, Boston. MA, SJoint Center for Radiation Therapy, Boston, MA Purpose: To determine the incidence (SNBM-RT) among longterm survivors as first cancers in the SEER population.
of all second non-breast malignancies (SNBM) and SNBM in the radiation field and to compare the risk of these events to the age-specific incidence of malignancies
146
I. .I. Radiation
Oncology
l Biology
l Physics
Volume 48, Number 3, Supplement,
2000
Materials and Methods: We analyzed the likelihood of SNBM for 1,884 patients with clinical stage I-II breast cancer treated with gross excision and 260 Gy (median 63) to the breast between 1970-1987.58% received supracIavicular/axillary radiation (median dose 45 Gy, range 20-60 Gy) and 28% received systemic therapy. The median age at diagnosis was 52. Median clinical tumor size was 2 cm. Patients were considered at risk for a SNBM until the first of distant metastases, contralateral breast cancer, or death or, if alive and disease-free, until last follow up. SNBM-RT were defined as any invasive tumors of the lung, bronchus, pleura, mediastinum, trachea, esophagus, larynx and thyroid. Expected numbers of cancers were obtained from the SEER databases for 1982-1986, using age-specific incidence for white females within 5-year age groups (p-values from exact Poisson). The median time at risk for a SNBM was 10.9 years (range 0.2-27.9). Results: 151 SNBM sites were observed in 147 patients: 33 lung, 23 colorectal, 1 esophagus, 9 other GI, 12 ovary, 16 other GYN, 14 melanoma, 8 lymphoma, 7 sarcoma, 6 meningioma, 4 urinary tract, 3 leukemia, 2 neuroendocrine, 2 multiple myeloma, 2 mesothelioma, 2 head and neck, 2 GBM, 1 astrocytoma, 1 neuroectodermal tumor, 1 polycythemia Vera, 1 thyroid and 1 unknown primary. Within the first 5 years, observed and expected SNBM were identical (47 vs 46.9). After 5 years, 24% more SNBM were observed than expected (100 vs 80.8, p = 0.02). Among patients < 50 years old at diagnosis, there were 43% more observed SNBM than expected (40 vs 28, p = 0.02). For patients 2 50, there were 7% more SNBM observed than expected (107 vs 99.7, p = 0.25). The observed incidence of ovarian cancer was significantly greater than expected among patients < 50 (7 vs 196, p = 0.004) but was not different than expected for patients 2 50 (5 vs 5.3, p = 0.61). SNBM-RT have been observed in 35, 33% more than predicted by SEER (p = 0.06). Among the 7 sarcomas, 1 developed in the ipsilateral shoulder, 3 in the ipsilateral breast, and 1 in the ipsilateral pleura. Four of these received three-field radiation using an old matchline technique and were noted to have matchline fibrosis with subsequent development of the sarcoma along the matchline. Conclusions: Patients with breast cancer treated with CS and RT are at increased risk of second non-breast malignancies compared to an age-matched population. This risk increases with time and is higher in younger patients. Some of the increased risk appears related to RT, and some may be related to underlying genetic factors seen in patients with breast cancer. In particular, there was an increased incidence of in-field SNBM-RT. However, the current technique of matching the axillarysupraclavicular field to the breast tangents may reduce the incidence of radiation-induced sarcomas.
69
The role of endorectal
D. H. Clarke,’ D. N. Blair’
S. J. Banks,’
‘Inova Alexandria
coil MRI in patient selection and treatment
A. R. Wiederhorn,’
Cancer Center, Alenandriu,
planning for prostate seed implants
J. W. Klousia,* .I. M. Lissy,’ A. A. Able,’ C. Artilles,’ VA, ‘Inova Alexandria
Hospital, Alexandria,
W. V. Hindle,’
VA
Based on data by D’Amico regarding the staging efficacy of endorectal coil MRI (erMRI), we initiated a permanent prostate seed implant program utilizing erMR1 as a routine staging procedure. This presentation will summarize our staging results, the impact of erMR1 on patient selection and the value of erMR1 in pre-planning an optimal implant. Between lo/94 and 12/98, more than 1,000 erMR1 studies were performed on prostate cancer patients evaluated in radiation oncology; and 390 prostate seed implants (99% Pd-103, 1% I-125) were performed. 327/390 seed patients (84%) had an erMR1. 239/327 patients (73%) were upstaged and 84/327 patients (26%) were upstaged to T3 with 8/327 patients (2.5%) being T3c. 51/327 patients (16%) had perineural invasion and this strongly correlated with T3 stage that was seen in 31/51 patients (61%). Of the 84% without perineural invasion, only 24% had T3 disease by erMR1. erMR1 findings had a major impact on treatment. 65/327 patients (20%) had a change in the overall treatment recommendation; i.e. combined external beam and seeds versus monotherapy. 185/327 patients (57%) had a change in seed distribution. Most of these changes related to pre-planned extracapsular coverage or seminal vesicle implantation. Between lo/94 and 12/96, 205 patients were implanted with a mean follow-up of 37 months. 72% of these patients had pre-treatment erMR1. 57% received combined therapy including external RT and seeds and 37% received hormonal downsizing. Only patients with documented PSA failure received hormonal therapy post-implant. Despite the fact that 39% had a Gleason Score (GS) of 7 or higher, there have been only 17 PSA failures. 15/17 patients had PSA doubling times of less than 6 months; the 1 patient with a doubling time greater than 24 months had a positive prostascint scan in paraaortic lymph nodes. There have been no documented local recurrences. D’Amico previously defined an intermediate risk group of patients (PSA 4 - 10 and GS 5 - 7, PSA 0 - 4 and GS 7, and PSA 10 _ 20 and GS 2 ~ 7) who also had extracapsular extension (ECE) or seminal vesicle invasion @VI) on erMRI. Following radical prostatectomy the PSA progression rate was 79% at 36 months. In our series, we identified 26 patients who fell into the intermediate risk group and had T3 disease on erMR1. Following external radiation and seeds, our PSA progression rate at 35 months was zero. We believe that our extensive use of erMR1 as a routine staging procedure has significantly improved our ability to select appropriate candidates for seeding and also has been instrumental in identifying patients who would be better treated with combined therapy. erMR1 has had a dramatic influence on seed distribution allowing for accurate implantation of periprostatic disease, including tumor involvement of the neurovascular bundle and the proximal seminal vesicles. Furthermore, erMR1 has also enabled us to plan for a ‘boost’ within the implanted volume by precisely delineating the location of bulky tumor. We believe our excellent results are largely attributable to our use of erMR1.
70
Ten year biochemical
outcomes
following
1254odine
monotherapy
for early stage prostate cancer
P. D. Grimm, J. C. Blasko, J. E. Sylvester, W. A. Cavanagh, R. Meier Seattle Prostate Institute, Seattle, WA One-hundred twenty-six (126) patients who presented with well to moderately differentiated carcinoma of the prostate between January 1988 and December 1990 were treated by the ‘Seattle‘ method of prostate brachytherapy. One patient was lost to follow-up, leaving 125 for PSA-based analysis. Patients were staged clinically by DRE as follows: Tl 30 (24%); T2a - 77 (62%); T2b - 17 (14%); T2c - 1 patient. All patients presented with a combined Gleason score less than 7. Median serum PSA at presentation was 5.1 ng/mI; mean 8.1; 96 (77%) presented with serum PSA less than 10 ng/ml. Median PSA-based follow-up time is 8 14 months, 94.4 months when considering only living, non-failed patients. PSA