- Email: [email protected]
Hyperfractionation in radiotherapy using an extended linear quadratic model
243
Proceedings of the 32nd Annual ASTRO Meeting
1037 POSTOPERATIVE RADIATION THERAPY FOR MALIGNANTFIBROUS HISTIOCYTOMA HM Fagundes,’
1
PP Lai,
1 Radiation Oncology Surgery, Washington
LP Dehner,
2
BN Emami,
Center, Mallinckrodt University School of
1
WG Kraybill,
Institute Medicine,
3
and NA Kucikl
2 Department of Radiology; St. Louis, MO
of
Pathology,
3
Department
of
Between 1974 and 1989, 49 patients with histologically confirmed malignant fibrous histiocytoma received postoperative radiotherapy at the Mallinckrodt Institute of Radiology for primary (41) or recurrent (8) disease. The median age of the patients was 63 years, and the median follow-up period was 41 months. Patients were grouped according to the 1988 AJC staging classification: stage IA (one patient), stage IIA (4 patients), stage IIB (9 patients), stage IIIA (15 patients), stage IIIB (18 patients), and stage IVA (2 patients). Distribution of primary lesions included 18(37%) in lower extremities, 11(23X) in upper extremities, 8(16%) in pelvis, 8(16X) in trunk, and 4(8%) in head and neck areas. Primary surgical procedures included incisional biopsy (4), excisional biopsy (19), narrow margin excision (14), wide local excision (9), and removal of the entire compartment (3). Based on pathology reports, the margins of resection were as follows: 23 positive margins (5 gross, 18 microscopic), 5 close margins, 10 unknown, and 11 negative margins. Patients were treated with shrinking field technique, and the median radiation dose was 6000 cGy with more than 95% of patients receiving at least 4500 cGy. In addition, 7 patients received chemotherapy NED survival, The 5-year overall survival, local control, and freedom from distant postoperatively. metastasis rates were 62X, 64X, 68X, and 85%, respectively. Five patients developed distant metastases, with 80% occurring within 2 years. Thirteen patients had local recurrences with > 75% recurring within 3 years. The sites of recurrence were as follows: trunk (3), pelvis (3), lower extremities (4), and head and neck areas (3). A direct correlation of local failure with positive surgical margin was observed: 9 local failures had positive margins; 2 had close margins; 1 was unknown, and 1 had a negative margin. Three of 13 patients with persistent or recurrent disease were salvaged by additional treatment, rendering the ultimate local control of 80% (39149). Thirty-four of 36 patients with local control obtained good to excellent function. Treatment complications consisted of 7 patients with skin and soft tissue fibrosis, 6 patients with delayed edema of the extremities, and 2 patients with necrotic skin ulcers (successfully treated with hyperbaric oxygen). In summary, postoperative radiotherapy for malignant fibrous histiocytoma is indicated to achieve local control as well as preservation of functional limbs with acceptable morbidity.
1038 HYPERFRACTIONATION
IN RADIOTHERAPY
USING
AN
EXTENDED
LINEAR
QUADRATIC
MODEL
J. van de Geijn, T. Goffman, Junwen Chen Radiation Oncology Branch,
Division of Cancer Treatment, NCI, NIH Bethesda MD 20892
Accelerated and hyperfractionated radiotherapy are explored increasingly in efforts to enhance the therapeutic ratio, in particular for tumors suspected of containing sizable fractions of rapidly proliferating cells. Shorter overall delivery time allows less tumor proliferation during the treatment course and thus diminishes the number of cells to be killed before the end of the treatment. The same or even a smaller total dose might then be more effective toward tumor control. However, shortening of overall time also is likely to enhance acute normal tissue reactions. In addition to this proliferation and repopulation related “time factor”, shorter inter-fraction time intervals may give rise to extra lethal damage due to incompletely repaired sublethal damage (the “II?” component). The IR-effect is desirable for tumor control but not for normal tissues. In this presentation the IR- and the time factor are added to the conventional a- and j3components to form an extended LQ-model. This poster then examines the relative importance of the four components by comparing computed control rates to clinical results for hyperfractionation schemes currently in use in four major centers. To this end, reasonable estimates are made of the tumor volume and its clonogenic fraction at the start of treatment, the half-time of repair, the acute doubling time and the delay in its onset from the beginning of treatment. Acceptable sets a$ can be selected based on the local control rate obtained with a conventional treatment regimen: the control rate p is assumed to be given by the Poisson relationship p=expf-N1, N being the number of clonogenic cells left at the end of treatment. N is given by their number No at the beginning, corrected for growth during treatment and multiplied by the survival rate, which is determined by the treatment regimen, that is the conventional a and j3terms, and now including the IR -correction. It appears that for the alternative regimens examined here, the IR factor should be negligible compared to the a- and P-components. The limitations caused by normal tissue tolerance create competition between the conventional “dose factor” (Edose=and+3nd2=(a+j3d)D] and the counteracting time factor (Etme = 0.693. (Ttot - Tdelay)/Tpot]. Both Tpot and T&lay are unknown, however. If shortening of overall time Tt,trequires lowering of the Edose (through lowering of d and/or D), a net loss in effectiveness will result unless Eume is reduced bv more (which can be done only through shortening of Ttot). CHART drops the conventional dose by 10%. but then has to count on a net gain in total effectiveness due to greater drop in the time factor, assuming a very short Tpot and a very short T&lay as well. Slower components of tumors might get underdosed that way, however, for a loss of control. Overall, because of the uncertainties about Tpot and Tklcly, uninterrupted hyperfractionation to higher dose over conventional overall times would seem to be preferable. Tables and graphs will illustrate these observations.
Proceedings of the 32nd Annual ASTRO Meeting
1037 POSTOPERATIVE RADIATION THERAPY FOR MALIGNANTFIBROUS HISTIOCYTOMA HM Fagundes,’
1
PP Lai,
1 Radiation Oncology Surgery, Washington
LP Dehner,
2
BN Emami,
Center, Mallinckrodt University School of
1
WG Kraybill,
Institute Medicine,
3
and NA Kucikl
2 Department of Radiology; St. Louis, MO
of
Pathology,
3
Department
of
Between 1974 and 1989, 49 patients with histologically confirmed malignant fibrous histiocytoma received postoperative radiotherapy at the Mallinckrodt Institute of Radiology for primary (41) or recurrent (8) disease. The median age of the patients was 63 years, and the median follow-up period was 41 months. Patients were grouped according to the 1988 AJC staging classification: stage IA (one patient), stage IIA (4 patients), stage IIB (9 patients), stage IIIA (15 patients), stage IIIB (18 patients), and stage IVA (2 patients). Distribution of primary lesions included 18(37%) in lower extremities, 11(23X) in upper extremities, 8(16%) in pelvis, 8(16X) in trunk, and 4(8%) in head and neck areas. Primary surgical procedures included incisional biopsy (4), excisional biopsy (19), narrow margin excision (14), wide local excision (9), and removal of the entire compartment (3). Based on pathology reports, the margins of resection were as follows: 23 positive margins (5 gross, 18 microscopic), 5 close margins, 10 unknown, and 11 negative margins. Patients were treated with shrinking field technique, and the median radiation dose was 6000 cGy with more than 95% of patients receiving at least 4500 cGy. In addition, 7 patients received chemotherapy NED survival, The 5-year overall survival, local control, and freedom from distant postoperatively. metastasis rates were 62X, 64X, 68X, and 85%, respectively. Five patients developed distant metastases, with 80% occurring within 2 years. Thirteen patients had local recurrences with > 75% recurring within 3 years. The sites of recurrence were as follows: trunk (3), pelvis (3), lower extremities (4), and head and neck areas (3). A direct correlation of local failure with positive surgical margin was observed: 9 local failures had positive margins; 2 had close margins; 1 was unknown, and 1 had a negative margin. Three of 13 patients with persistent or recurrent disease were salvaged by additional treatment, rendering the ultimate local control of 80% (39149). Thirty-four of 36 patients with local control obtained good to excellent function. Treatment complications consisted of 7 patients with skin and soft tissue fibrosis, 6 patients with delayed edema of the extremities, and 2 patients with necrotic skin ulcers (successfully treated with hyperbaric oxygen). In summary, postoperative radiotherapy for malignant fibrous histiocytoma is indicated to achieve local control as well as preservation of functional limbs with acceptable morbidity.
1038 HYPERFRACTIONATION
IN RADIOTHERAPY
USING
AN
EXTENDED
LINEAR
QUADRATIC
MODEL
J. van de Geijn, T. Goffman, Junwen Chen Radiation Oncology Branch,
Division of Cancer Treatment, NCI, NIH Bethesda MD 20892
Accelerated and hyperfractionated radiotherapy are explored increasingly in efforts to enhance the therapeutic ratio, in particular for tumors suspected of containing sizable fractions of rapidly proliferating cells. Shorter overall delivery time allows less tumor proliferation during the treatment course and thus diminishes the number of cells to be killed before the end of the treatment. The same or even a smaller total dose might then be more effective toward tumor control. However, shortening of overall time also is likely to enhance acute normal tissue reactions. In addition to this proliferation and repopulation related “time factor”, shorter inter-fraction time intervals may give rise to extra lethal damage due to incompletely repaired sublethal damage (the “II?” component). The IR-effect is desirable for tumor control but not for normal tissues. In this presentation the IR- and the time factor are added to the conventional a- and j3components to form an extended LQ-model. This poster then examines the relative importance of the four components by comparing computed control rates to clinical results for hyperfractionation schemes currently in use in four major centers. To this end, reasonable estimates are made of the tumor volume and its clonogenic fraction at the start of treatment, the half-time of repair, the acute doubling time and the delay in its onset from the beginning of treatment. Acceptable sets a$ can be selected based on the local control rate obtained with a conventional treatment regimen: the control rate p is assumed to be given by the Poisson relationship p=expf-N1, N being the number of clonogenic cells left at the end of treatment. N is given by their number No at the beginning, corrected for growth during treatment and multiplied by the survival rate, which is determined by the treatment regimen, that is the conventional a and j3terms, and now including the IR -correction. It appears that for the alternative regimens examined here, the IR factor should be negligible compared to the a- and P-components. The limitations caused by normal tissue tolerance create competition between the conventional “dose factor” (Edose=and+3nd2=(a+j3d)D] and the counteracting time factor (Etme = 0.693. (Ttot - Tdelay)/Tpot]. Both Tpot and T&lay are unknown, however. If shortening of overall time Tt,trequires lowering of the Edose (through lowering of d and/or D), a net loss in effectiveness will result unless Eume is reduced bv more (which can be done only through shortening of Ttot). CHART drops the conventional dose by 10%. but then has to count on a net gain in total effectiveness due to greater drop in the time factor, assuming a very short Tpot and a very short T&lay as well. Slower components of tumors might get underdosed that way, however, for a loss of control. Overall, because of the uncertainties about Tpot and Tklcly, uninterrupted hyperfractionation to higher dose over conventional overall times would seem to be preferable. Tables and graphs will illustrate these observations.