- Email: [email protected]
Intraoperative electron beam radiotherapy for previously irradiated advanced head and neck malignancies
Int. J. Radiation Oncology Biol. Phys., Vol. 42, No. 5, pp. 1085–1089, 1998 Copyright © 1998 Elsevier Science Inc. Printed in the USA. All rights reserved 0360-3016/98/$–see front matter
PII S0360-3016(98)00289-2
●
Clinical Investigation INTRAOPERATIVE ELECTRON BEAM RADIOTHERAPY FOR PREVIOUSLY IRRADIATED ADVANCED HEAD AND NECK MALIGNANCIES SUBIR NAG, M.D.,* DAVID E. SCHULLER, M.D.,† RAFAEL MARTINEZ-MONGE, M.D.,* SILVIA RODRIGUEZ-VILLALBA, M.D.,* JOHN GRECULA, M.D.,* AND CONSTANCE BAUER, M.D.* *Divisions of Radiation Oncology and †Otolaryngology, Arthur G. James Cancer Hospital and Research Institute, The Ohio State University, Columbus, OH Purpose: This is a retrospective review to evaluate the role of surgery and intraoperative electron beam radiotherapy (IOERT) in the treatment of patients with previously irradiated advanced head and neck cancers. Methods and Materials: Between January 1992 and March 1997, 38 patients (31 males, 7 females; median age of 62 years) with recurrent head and neck cancer were treated with maximal resection and IOERT at the Ohio State University (OSU). All had been previously treated with full-course radiotherapy (median 65.1 Gy, range 50 –74.4 Gy). Twenty-nine patients (76%) had previously undergone one or more surgical procedures. After maximal surgery the tumor bed was treated with IOERT (single field in 36 patients and 2 fields in 2 patients), most commonly with 6 MeV electrons (87%). The dose administered (at 90% isodose line) was 15 Gy for close or microscopically positive margins in 34 patients and 20 Gy for gross disease in 1 patient. Further external beam radiation therapy (EBRT) was not given. Results: After a median follow-up of 30 months (range 8 –39 months), 24 of the 38 patients (66%) recurred within the IOERT field. Median time to IOERT failure was 6 months (95% CI: 4.3–7.7). The 6-month, 1-, and 2-year control rates within the IOERT volume were 41%, 19%, and 13 %, respectively. Thirty of the 38 patients (79%) recurred in locoregional areas. Median time to locoregional failure was 4 months (95% CI: 3.3– 4.7). The 6-month, 1-, and 2-year locoregional control rates were 33%, 11%, and 4%, respectively. Distant metastases ocurred in 7 patients, 5 in association with IOERT failure and 2 with locoregional failure. Median overall survival was 7 months (95% CI: 4.7–9.3). The 6-month, 1-, 2-, and 3-year actuarial survival rates were 51%, 21%, 21%, and 8%, respectively. Major treatment-related complications occurred in 6 patients (16%). Conclusion: IOERT alone, at the dose used, is not sufficient for control of recurrent, previously irradiated head and neck cancers. Since higher IOERT doses are associated with high morbidity, we are currently evaluating the addition of limited EBRT dose and/or brachytherapy to improve the local control of these poor prognostic recurrent tumors, with acceptable morbidity. © 1998 Elsevier Science Inc. Head and neck tumors, Recurrences, Intraoperative, Radiotherapy; Electron beam.
INTRODUCTION Previously irradiated, advanced head and neck tumors have a dismal prognosis. Their therapeutic options are limited due to prior treatments received. Physicians are usually reluctant to prescribe meaningful doses of external beam radiation therapy (EBRT) to patients who have previously received full doses of EBRT due to the tolerance of the normal tissues. As a result, palliative surgery, brachytherapy, hyperthermia, chemotherapy, or therapeutic abstention are the usual options. However, salvage radiation therapy might be justified if the expected complications could be minimized.
Intraoperative Electron Beam Radiation Therapy (IOERT) has been used in the management of locally advanced and recurrent head and neck cancer (1– 8). IOERT allows the delivery of radiation at the time of surgery, with direct visualization of the tumor bed, minimizing the risk of a geographical miss. Normal tissues can be retracted or shielded and therefore, higher doses can be given to the tumor. The rapid dose fall-off from IOERT minimizes the volume of tissue irradiated, reducing the risk of morbidity. We therefore evaluated the role of palliative surgery and IOERT in 38 patients with previously irradiated recurrent head and neck cancer.
Reprint requests to: Subir Nag, M.D., Professor and Chief of Brachytherapy, The Arthur G. James Cancer Hospital, Ohio State University, 300 West Tenth Avenue, Columbus, OH 43210. E-mail: [email protected] Acknowledgments—Supported in part by grant no. 5P30CS16058 from National Cancer Institute, Bethesda, MD. Dr. Rafael Mar-
tinez-Monge was supported by a Brachytherapy Fellowship from the Government of Navarre (Spain). The authors wish to express their gratitude to Mrs. Rita Planitzer for her help in the preparation of this manuscript and Mr. David Carpenter for editorial assistance. Accepted for publication 23 July 1998. 1085
1086
I. J. Radiation Oncology
●
Biology
●
Physics
Table 1. Previous radiotherapy No. of patients EBRT (once)
33
EBRT (twice) EBRT 1 Implant HDR EBRT 1 I-125 seed
1 1 2
EBRT 1 IOHDR*
1
Volume 42, Number 5, 1998
Table 2. Tumor characteristics n
%
2 1 35
5 3 92
9 11 10
30 37 33
12 6 5 1 10 6 2 1 1 3 1 1 1 3 2 1 3 2 1 2 1 1 1 1 1
32
16 18
47 53
Dose median 53 Gy range 46–74.4 Gy 50.4 Gy/18.3 Gy 46.8 Gy 1 20 Gy (5 Gy 3 4) 54 Gy 1 50 Gy 66 Gy 1 50 Gy 45 Gy 1 12.5 Gy
*IOHDR 5 Intraoperative high dose rate brachytherapy.
METHODS AND MATERIALS Patients Between January 1992 and March 1997, 38 patients (31 male, 7 female; median age 62; range 28 – 81 years) with local or regional recurrent head and neck cancer were treated with IOERT at the James Cancer Center, The Ohio State University (OSU). Eighteen patients received IOERT during the management of the first locoregional recurrence, 13 (34%) during the treatment of the second, and 7 (18%) during subsequent recurrences. All had been previously treated with full-course radiotherapy (with EBRT and/or brachytherapy; median 65.1 Gy, range 50 –74.4 Gy) (Table 1). Twenty-nine patients had undergone prior surgery; 1 procedure in 17 patients (45%), 2 procedures in 9 patients (24%), and 3 procedures in 3 patients (8%). Seven (18%) patients had prior chemotherapy. Tumor characteristics are depicted in Table 2. Sites of recurrence included the primary site only in 11 patients (29%), neck only in 14 (37%), tracheal stoma only in 8 (21%), both primary and neck in 4 (11%), and stoma and neck in 1 (3%). IOERT technique OSU has a dedicated linear accelerator (linac) for IOERT in the operative suite, with electron beams in the 6 –15 MeV energy range. The gantry of the linac can be rotated 25° from the vertical axis. The appropriate applicator, selected to encompass the area to be irradiated, was manually positioned over the target area and attached to the surgical table with a Bookwalter clamp. Gauze packing, retractors and/or pliable 1–2-mm-thick lead shields were used to displace or protect normal tissues near or inside the treatment field. The applicator was aligned to the linac by moving the table under the guidance of a laser docking system. There was no physical contact between the linac and the applicator (soft docking system). The treatment field was suctioned to prevent any accumulated fluids from acting as a bolus. The staff then left the operating room, and the patient was observed with remote monitors while IOERT was delivered. Maximal resection was attempted prior to IOERT in all patients. However, gross residual disease was left in 3 patients, and microscopic or close margins remained in 35 (92%) patients. The target volume included the tumor bed
Pathology: Adenocarcinoma Large cell carcinoma Squamous Grade: 1 2 3 Initial location: Larynx Supraglottis Glottis Both Oral cavity Lateral tongue Floor of mouth Buccal mucosa Gum Oropharynx Tonsillar fossa Soft palate Retrom. trigone Hypopharynx Pyriform sinus Postcricoid region Paranasal sinuses Maxillary antrum Ethmoid Parotid gland Unknown primary (neck node) Nasopharynx Orbit Lower lip Skin (nasolabial fold) Initial stage: Stage I/II Stage III/IV
26
8
8 8 5 3 3 3 3 3
with a 1–2-cm margin. In 36 patients the target volume was encompassed with a single IOERT field. In one patient, two separate sites were treated; and, in the other patient with gross residual tumor, a “field-within-a-field” technique was used. The applicator size varied from 5 to 12 cm in diameter; the 6-cm applicator being the most frequently used (29%). Shielding and bolus were used where required (Table 3). The IOERT doses given were according to the following protocol: 15 Gy for close or microscopically positive margins in 34 patients and 20 Gy for gross disease in 1 patient. Two patients with close margins were treated off-protocol with 10 Gy because of high (. 70 Gy) EBRT doses previously received. One additional patient was treated by a field within a field technique, receiving 25 Gy to the area of gross residual and 15 Gy to the surrounding areas of microscopic residual. The IOERT dose was prescribed to the 90% isodose line which encompassed the area at risk. Median treatment volume was 39 cc (range 9 –137 cc). Since these patients had previously received full dose EBRT, postoperative EBRT was not recommended. Control at the IOERT site, locoregional control and over-
IOERT in head and neck cancer
Bolus: Yes No % Target lead shielded: None 1–5% 6–10% 11–15% 16–20% 30% Cone diameter (cm): 5–6 7–8 9–11 Beam Energy (MeV): 6 9 12
S. NAG et al.
1087
Table 4. Patterns of failure
Table 3. IOERT characteristics Parameter
●
n
%
18 20
47 53
24 1 4 1 3 5
63 3 11 3 8 13
19 12 7
50 32 18
33 4 1
87 11 3
all survival rates were calculated using the Kaplan-Meier product limit method and compared using the log-rank test (9). Patient characteristics (age, sex, histology, histologic grade, initial location, number of previous recurrences, number of previous surgeries, previous chemotherapy) and treatment factors (IOERT volume, applicator diameter, IOERT dose, IOERT energy, residual disease, IOERT site, and postoperative chemotherapy) were analyzed using IOERT control, locoregional control, and overall survival as endpoints. Median follow-up was 30 months (range 8 –39 months). Three patients were lost to follow-up.
Failure within IOERT volume Loco-regional failure Distant failure Combined, local, and distant failure
n
%
24 30 7 7
63 79 18 18
Median time to IOERT failure was 6 months (95% CI: 4.3–7.7 months). Twenty-four of the 38 patients (66%) recurred within the IOERT field. Table 4 summarizes the patterns of failure. Locoregional control The 6-month, 1-, and 2-year locoregional control rates were 33%, 11%, and 4%, respectively (Fig. 2). Median local failure was 4 months (95% CI: 3.3– 4.7). Thirty of 38 patients (79%) recurred in locoregional areas. All patients who recurred in the IOERT field and those who failed above the clavicles were considered locoregional failures. None of the factors analyzed significantly influenced locoregional control. Distant failure Distant metastases occurred in seven patients, five with concomitant IOERT failure and two with locoregional failure. Involved organs were the lung in one, liver in two, bone in two, brain in one, and lung and liver synchronously in one.
Control within IOERT volume The 6-month, 1-, and 2-year control rates within IOERT volume were 41%, 19%, and 13%, respectively (Fig. 1).
Survival The 6-month, 1-, 2-, and 3-year actuarial survival rates were 51%, 21%, 21%, and 8%, respectively (Fig. 3). Median overall survival was 7 months (95% CI: 4.7–9.3). Those patients treated with 5– 8-cm diameter applicators survived longer (7 of 31 patients alive) than patients treated
Fig. 1. Actuarial control within the IOERT volume.
Fig. 2. Actuarial locoregional control.
RESULTS
1088
I. J. Radiation Oncology
●
Biology
●
Physics
Fig. 3. Actuarial overall survival.
with larger applicators (0 of 7 patients alive) (p , 0.009). This difference reflects the better prognosis of smaller volume recurrent disease, which could be encompassed within smaller applicators. Patients treated to neck sites had a better overall survival (4 of 14 patients alive) than those with primary (2 of 11), stomal (1 of 8) or multiple recurrences (0 of 6). This difference was of borderline significance (p 5 0.054). At the time of the analysis, 31 patients have expired; 26 of tumor recurrence and 4 of non-cancer-related causes without evidence of recurrence. One patient died due to treatment related complications (tracheovascular fistulae 1 month after IOERT). Five (13%) are alive with tumor 2, 10, 25, 32, and 33 months after IOERT. Complications Severe treatment-related complications occurred in six patients (16%) and included orocutaneous fistulae in two, tracheal dehiscence in one, wound dehiscence in one, carotid occlusion in one, and fatal tracheovascular fistula in one. DISCUSSION Patients with advanced, previously irradiated head and neck cancer have very limited treatment options. Even in patients with potentially resectable lesions, surgical resection is often unsuccessful due to the microscopic extension of the disease and the technical difficulties of surgical resection after radiation therapy. Radiation-induced fibrosis makes recognition of tumor boundaries and dissection through normal tissue planes very difficult. Recurrent tumors may be more radioresistant than primary head and neck cancers. In vitro cultured tumor cells from head and neck squamous cell carcinoma recurrences after a complete course of EBRT have been shown to be radioresistant (10). Previously irradiated tissues are more hypoxic and therefore
Volume 42, Number 5, 1998
more radioresistant. Furthermore, the inability to deliver meaningful doses of adjuvant EBRT in previously irradiated tumors makes local control unlikely (4, 5, 11). While IOERT combined with postoperative EBRT, achieves local control in 40 –70% of the patients with primary head neck tumors (1, 2, 8, 11–13), decreased local control and survival is seen in patients who do not receive or complete the supplementary postoperative EBRT (5, 14). Martinez-Monge and Calvo reported on a series of 31 patients with locally advanced or recurrent head and neck cancer treated with maximal surgical resection and 10 –15 Gy IORT at the University of Navarre, Spain (4). The 7-year control rate for 17 previously unirradiated patients was 46%, versus 19% for 14 previously irradiated patients. Six previously irradiated patients received low-dose EBRT (median 30 Gy). The inability to deliver EBRT because of prior EBRT was statistically significant for local control (p 5 0.004). In our previously published series, local control decreased from 79% to 50% in those patients who did not receive the postoperative EBRT (5). We integrated IOERT into a surgical salvage program in an attempt to improve local control and survival results in these previously irradiated patients with dismal prognosis. IOERT allows combining maximal debulking surgery and “in situ” radiotherapy, minimizing the possibility of geographical miss and avoiding irradiation of uninvolved structures. In the head and neck area, EBRT reirradiation of certain normal tissues might be a problem. Significant areas of bone and uninvolved aerodigestive mucosa would be reirradiated, even with the most sophisticated treatment planning. Some representative EBRT reirradiation series have reported a severe late sequelae rate of 9 –28% (15–18) with reirradiation doses in the 30 – 60 Gy range. The probability of tumor control is a function of the dose administered. Using the linear-quadratic bio-effective model, 15 Gy and 20 Gy IOERT are radiobiologically equivalent (using a/b 5 10 Gy) to 31.3 Gy and 50 Gy given as fractionated EBRT in terms of tumor effect (19). Radiobiologically, using single large doses of IOERT is disadvantageous because it does not permit the repair of sublethal damage of the surrounding normal tissues or reoxygenation or redistribution of tumor cells to more radiosensitive mitotic phases. This disadvantage is compensated for by the reduction of the dose to the normal tissues, with dose reductions of 20 –30% required to achieve the same isoeffect dose (20), depending on the assumed half-time repair of tumor cells. This disadvantage can also be minimized by giving IOERT as a boost to the tumor bed to supplement a moderate dose of EBRT to a greater volume, thus achieving higher doses to the tumor while minimizing morbidity. Reirradiation with IOERT and EBRT (with or without chemotherapy) has also been used with good results in recurrent previously irradiated colorectal cancers (21). Reirradiation with EBRT has been explored in several studies (15– 18). Some of these studies have used hyperfractionated regimens between 1.2–1.6 Gy B.I.D. with an interfraction gap of at least 6 h to minimize the incidence of late tissue
IOERT in head and neck cancer
damage (15, 18). Treated patients include a broad category of recurrent tumors or second primaries arising in a previously irradiated area. The treatments given included EBRT reirradiation (30 – 60 Gy), surgery and intracavitary or interstitial brachytherapy boost whenever possible. Overall, these reirradiation series have reported long-term local control in 27–36% of the patients (15–18). When areas at risk are technically inaccessible to the electron beam applicator (e.g., steeply sloped surfaces or narrow cavities), we employ intraoperative high-dose-rate brachytherapy (IOHDR) (14). The flexible applicators used in IOHDR are able to access narrow cavities and can homogeneously irradiate curved and sloping surfaces. The single dose employed has the same disadvantages previously described for IOERT; however, it is possible to deliver postoperative fractionated high-dose-rate brachytherapy several days after the surgery (perioperative brachytherapy), using the implanted catheters in situations in which further external beam radiation cannot be given. IOERT alone does not provide good control of recurrent, previously irradiated head and neck cancers. Our experience
●
S. NAG et al.
1089
is confirmed by other authors (7) who have reported high rates of locoregional failure (88.8%) in a series of 95 patients (94% of whom had previous EBRT). The poor control could be due to the limited IOERT doses that can be given to previously irradiated tumors. Higher IORT doses cause significant morbidity (22). Hence, alternative avenues being explored are the addition of postoperative EBRT (which is now routinely given) or brachytherapy (LDR or fractionated HDR) to maximal surgery and IOERT in an attempt to improve the dismal prognosis of these patients.
CONCLUSION IOERT alone, at the dose used, is not sufficient for control of recurrent, previously irradiated head and neck cancers. Since higher IOERT doses are associated with high morbidity, we are currently evaluating the addition of limited EBRT dose and/or brachytherapy to improve the local control of these poor prognostic recurrent tumors, with acceptable morbidity.
REFERENCES 1. Coleman C, Roach M III, Ling S, et al. Adjuvant electronbeam IORT in high-risk head and neck cancer patients. Front Radiat Ther Oncol 1997; 31:105–111. 2. Freeman SB, Hamaker RC, Singer MI, et al. Intraoperative radiotherapy of head and neck cancer. Arch Otolaryngol Head Neck Surg 1990; 116:165–168. 3. Garrett P, Pugh N, Ross D, et al. Intraoperative radiation for advanced or recurrent head and neck cancer. Int J Radiat Oncol Biol Phys 1987; 13:785–788. 4. Martı´nez-Monge R, Azinovic I, Alcalde J, et al. IORT in the management of locally advanced or recurrent head and neck cancer. Front Radiat Oncol 1997; 31:122–125. 5. Nag S, Schuller D, Pak V, et al. Intraoperative radiation therapy using electron beam or HDR brachytherapy for previously unirradiated head and neck cancers. Front Radiat Ther Oncol 1997; 31:112–116. 6. Rate W, Garrett P, Hamaker R, et al. Intraoperative radiation therapy for recurrent head and neck cancer. Cancer 1991; 67:2738 –2740. 7. Spaeth J, Andropulus D, Unger T, et al. Intra-operative radiotherapy 5 years of experience in the palliative treatment of recurrent and advanced head and neck cancers. Oncology 1997; 54:208 –213. 8. Toita T, Nakano M, Takizawa Y, et al. Intraoperative radiation therapy (IOERT) for head and neck cancer. Int J Radiat Oncol Biol Phys 1994; 30:1219 –1224. 9. Kaplan EL, Meier P. Nonparametric estimation for incomplete observations. J Am Stat Assoc 1958; 4:57– 81. 10. Weichselbaum R, Beckett M, Scwartz J, et al. Radioresistant tumor cells are present in head and neck carcinomas that recur after radiotherapy. Int J Radiat Oncol Biol Phys 1988; 15: 575–579. 11. Pearlman N. Treatment outcome in recurrent head and neck cancer. Arch Surg 1979; 114:39 – 42. 12. Kokal W, Neifield J, Eisert D, et al. Management of locore-
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
gional recurrent oropharyngeal carcinoma. Am J Surg 1983; 146:436 – 438. Regine W, Valentino J, Patel P, et al. Efficacy of postoperative radiation therapy for recurrent squamous cell carcinoma of the head and neck. Int J Radiat Oncol Biol Phys 1997; 39:297– 302. Nag S, Schuller D, Pak V, et al. Pilot study of intraoperative high dose rate brachytherapy for head and neck cancer. Radiother Oncol 1996; 41:125–130. Benchalal M, Bachaud JM, Francois P, et al. Hyperfractionation in the reirradiation of head and neckcancers. Result of a pilot study. Radiother Oncol 1995; 36:203–210. Levendag PC, Meeuwis CA, Visser AG. Reirradiation of recurrent head and neck cancers: External and/or interstitial radiation therapy. Radiother Oncol 1992; 23:6 –15. Stevens KR Jr, Britsch A, Moss WT. High-dose reirradiation of head and neck cancer with curative intent. Int J Radiat Oncol Biol Phys 1994; 29:687– 698. Tercilla OF, Schmidt-Ullrich R, Wazer DE. Reirradiation of head and neck neoplasms using twice-a-day scheduling. Strahlenther Onkol 1993; 169:285–290. Nag S, Martinez- Monge R, Gupta N. Intraoperative radiation therapy using electron beam and high dose rate brachytherapy. Cancer J 1997; 10:94 –101. Nag S, Orton C. Development of intraoperative high dose rate for treatment of resected tumor bed in anesthetized patients. Endocuriether/Hypertherm Oncol 1993; 9:187–193. Gunderson L, Haddock M, Nelson H, et al. Locally recurrent colorectal cancer: IOERT and EBRT 6 5-FU and maximal resection. Front Radiat Oncol 1997; 31:224 –228. Shaw E, Gunderson L, Martin K, et al. Peripheral nerve and ureteral tolerance to intraoperative radiation therapy: Clinical and dose-response analysis. Radiother Oncol 1990; 18:247– 255.
PII S0360-3016(98)00289-2
●
Clinical Investigation INTRAOPERATIVE ELECTRON BEAM RADIOTHERAPY FOR PREVIOUSLY IRRADIATED ADVANCED HEAD AND NECK MALIGNANCIES SUBIR NAG, M.D.,* DAVID E. SCHULLER, M.D.,† RAFAEL MARTINEZ-MONGE, M.D.,* SILVIA RODRIGUEZ-VILLALBA, M.D.,* JOHN GRECULA, M.D.,* AND CONSTANCE BAUER, M.D.* *Divisions of Radiation Oncology and †Otolaryngology, Arthur G. James Cancer Hospital and Research Institute, The Ohio State University, Columbus, OH Purpose: This is a retrospective review to evaluate the role of surgery and intraoperative electron beam radiotherapy (IOERT) in the treatment of patients with previously irradiated advanced head and neck cancers. Methods and Materials: Between January 1992 and March 1997, 38 patients (31 males, 7 females; median age of 62 years) with recurrent head and neck cancer were treated with maximal resection and IOERT at the Ohio State University (OSU). All had been previously treated with full-course radiotherapy (median 65.1 Gy, range 50 –74.4 Gy). Twenty-nine patients (76%) had previously undergone one or more surgical procedures. After maximal surgery the tumor bed was treated with IOERT (single field in 36 patients and 2 fields in 2 patients), most commonly with 6 MeV electrons (87%). The dose administered (at 90% isodose line) was 15 Gy for close or microscopically positive margins in 34 patients and 20 Gy for gross disease in 1 patient. Further external beam radiation therapy (EBRT) was not given. Results: After a median follow-up of 30 months (range 8 –39 months), 24 of the 38 patients (66%) recurred within the IOERT field. Median time to IOERT failure was 6 months (95% CI: 4.3–7.7). The 6-month, 1-, and 2-year control rates within the IOERT volume were 41%, 19%, and 13 %, respectively. Thirty of the 38 patients (79%) recurred in locoregional areas. Median time to locoregional failure was 4 months (95% CI: 3.3– 4.7). The 6-month, 1-, and 2-year locoregional control rates were 33%, 11%, and 4%, respectively. Distant metastases ocurred in 7 patients, 5 in association with IOERT failure and 2 with locoregional failure. Median overall survival was 7 months (95% CI: 4.7–9.3). The 6-month, 1-, 2-, and 3-year actuarial survival rates were 51%, 21%, 21%, and 8%, respectively. Major treatment-related complications occurred in 6 patients (16%). Conclusion: IOERT alone, at the dose used, is not sufficient for control of recurrent, previously irradiated head and neck cancers. Since higher IOERT doses are associated with high morbidity, we are currently evaluating the addition of limited EBRT dose and/or brachytherapy to improve the local control of these poor prognostic recurrent tumors, with acceptable morbidity. © 1998 Elsevier Science Inc. Head and neck tumors, Recurrences, Intraoperative, Radiotherapy; Electron beam.
INTRODUCTION Previously irradiated, advanced head and neck tumors have a dismal prognosis. Their therapeutic options are limited due to prior treatments received. Physicians are usually reluctant to prescribe meaningful doses of external beam radiation therapy (EBRT) to patients who have previously received full doses of EBRT due to the tolerance of the normal tissues. As a result, palliative surgery, brachytherapy, hyperthermia, chemotherapy, or therapeutic abstention are the usual options. However, salvage radiation therapy might be justified if the expected complications could be minimized.
Intraoperative Electron Beam Radiation Therapy (IOERT) has been used in the management of locally advanced and recurrent head and neck cancer (1– 8). IOERT allows the delivery of radiation at the time of surgery, with direct visualization of the tumor bed, minimizing the risk of a geographical miss. Normal tissues can be retracted or shielded and therefore, higher doses can be given to the tumor. The rapid dose fall-off from IOERT minimizes the volume of tissue irradiated, reducing the risk of morbidity. We therefore evaluated the role of palliative surgery and IOERT in 38 patients with previously irradiated recurrent head and neck cancer.
Reprint requests to: Subir Nag, M.D., Professor and Chief of Brachytherapy, The Arthur G. James Cancer Hospital, Ohio State University, 300 West Tenth Avenue, Columbus, OH 43210. E-mail: [email protected] Acknowledgments—Supported in part by grant no. 5P30CS16058 from National Cancer Institute, Bethesda, MD. Dr. Rafael Mar-
tinez-Monge was supported by a Brachytherapy Fellowship from the Government of Navarre (Spain). The authors wish to express their gratitude to Mrs. Rita Planitzer for her help in the preparation of this manuscript and Mr. David Carpenter for editorial assistance. Accepted for publication 23 July 1998. 1085
1086
I. J. Radiation Oncology
●
Biology
●
Physics
Table 1. Previous radiotherapy No. of patients EBRT (once)
33
EBRT (twice) EBRT 1 Implant HDR EBRT 1 I-125 seed
1 1 2
EBRT 1 IOHDR*
1
Volume 42, Number 5, 1998
Table 2. Tumor characteristics n
%
2 1 35
5 3 92
9 11 10
30 37 33
12 6 5 1 10 6 2 1 1 3 1 1 1 3 2 1 3 2 1 2 1 1 1 1 1
32
16 18
47 53
Dose median 53 Gy range 46–74.4 Gy 50.4 Gy/18.3 Gy 46.8 Gy 1 20 Gy (5 Gy 3 4) 54 Gy 1 50 Gy 66 Gy 1 50 Gy 45 Gy 1 12.5 Gy
*IOHDR 5 Intraoperative high dose rate brachytherapy.
METHODS AND MATERIALS Patients Between January 1992 and March 1997, 38 patients (31 male, 7 female; median age 62; range 28 – 81 years) with local or regional recurrent head and neck cancer were treated with IOERT at the James Cancer Center, The Ohio State University (OSU). Eighteen patients received IOERT during the management of the first locoregional recurrence, 13 (34%) during the treatment of the second, and 7 (18%) during subsequent recurrences. All had been previously treated with full-course radiotherapy (with EBRT and/or brachytherapy; median 65.1 Gy, range 50 –74.4 Gy) (Table 1). Twenty-nine patients had undergone prior surgery; 1 procedure in 17 patients (45%), 2 procedures in 9 patients (24%), and 3 procedures in 3 patients (8%). Seven (18%) patients had prior chemotherapy. Tumor characteristics are depicted in Table 2. Sites of recurrence included the primary site only in 11 patients (29%), neck only in 14 (37%), tracheal stoma only in 8 (21%), both primary and neck in 4 (11%), and stoma and neck in 1 (3%). IOERT technique OSU has a dedicated linear accelerator (linac) for IOERT in the operative suite, with electron beams in the 6 –15 MeV energy range. The gantry of the linac can be rotated 25° from the vertical axis. The appropriate applicator, selected to encompass the area to be irradiated, was manually positioned over the target area and attached to the surgical table with a Bookwalter clamp. Gauze packing, retractors and/or pliable 1–2-mm-thick lead shields were used to displace or protect normal tissues near or inside the treatment field. The applicator was aligned to the linac by moving the table under the guidance of a laser docking system. There was no physical contact between the linac and the applicator (soft docking system). The treatment field was suctioned to prevent any accumulated fluids from acting as a bolus. The staff then left the operating room, and the patient was observed with remote monitors while IOERT was delivered. Maximal resection was attempted prior to IOERT in all patients. However, gross residual disease was left in 3 patients, and microscopic or close margins remained in 35 (92%) patients. The target volume included the tumor bed
Pathology: Adenocarcinoma Large cell carcinoma Squamous Grade: 1 2 3 Initial location: Larynx Supraglottis Glottis Both Oral cavity Lateral tongue Floor of mouth Buccal mucosa Gum Oropharynx Tonsillar fossa Soft palate Retrom. trigone Hypopharynx Pyriform sinus Postcricoid region Paranasal sinuses Maxillary antrum Ethmoid Parotid gland Unknown primary (neck node) Nasopharynx Orbit Lower lip Skin (nasolabial fold) Initial stage: Stage I/II Stage III/IV
26
8
8 8 5 3 3 3 3 3
with a 1–2-cm margin. In 36 patients the target volume was encompassed with a single IOERT field. In one patient, two separate sites were treated; and, in the other patient with gross residual tumor, a “field-within-a-field” technique was used. The applicator size varied from 5 to 12 cm in diameter; the 6-cm applicator being the most frequently used (29%). Shielding and bolus were used where required (Table 3). The IOERT doses given were according to the following protocol: 15 Gy for close or microscopically positive margins in 34 patients and 20 Gy for gross disease in 1 patient. Two patients with close margins were treated off-protocol with 10 Gy because of high (. 70 Gy) EBRT doses previously received. One additional patient was treated by a field within a field technique, receiving 25 Gy to the area of gross residual and 15 Gy to the surrounding areas of microscopic residual. The IOERT dose was prescribed to the 90% isodose line which encompassed the area at risk. Median treatment volume was 39 cc (range 9 –137 cc). Since these patients had previously received full dose EBRT, postoperative EBRT was not recommended. Control at the IOERT site, locoregional control and over-
IOERT in head and neck cancer
Bolus: Yes No % Target lead shielded: None 1–5% 6–10% 11–15% 16–20% 30% Cone diameter (cm): 5–6 7–8 9–11 Beam Energy (MeV): 6 9 12
S. NAG et al.
1087
Table 4. Patterns of failure
Table 3. IOERT characteristics Parameter
●
n
%
18 20
47 53
24 1 4 1 3 5
63 3 11 3 8 13
19 12 7
50 32 18
33 4 1
87 11 3
all survival rates were calculated using the Kaplan-Meier product limit method and compared using the log-rank test (9). Patient characteristics (age, sex, histology, histologic grade, initial location, number of previous recurrences, number of previous surgeries, previous chemotherapy) and treatment factors (IOERT volume, applicator diameter, IOERT dose, IOERT energy, residual disease, IOERT site, and postoperative chemotherapy) were analyzed using IOERT control, locoregional control, and overall survival as endpoints. Median follow-up was 30 months (range 8 –39 months). Three patients were lost to follow-up.
Failure within IOERT volume Loco-regional failure Distant failure Combined, local, and distant failure
n
%
24 30 7 7
63 79 18 18
Median time to IOERT failure was 6 months (95% CI: 4.3–7.7 months). Twenty-four of the 38 patients (66%) recurred within the IOERT field. Table 4 summarizes the patterns of failure. Locoregional control The 6-month, 1-, and 2-year locoregional control rates were 33%, 11%, and 4%, respectively (Fig. 2). Median local failure was 4 months (95% CI: 3.3– 4.7). Thirty of 38 patients (79%) recurred in locoregional areas. All patients who recurred in the IOERT field and those who failed above the clavicles were considered locoregional failures. None of the factors analyzed significantly influenced locoregional control. Distant failure Distant metastases occurred in seven patients, five with concomitant IOERT failure and two with locoregional failure. Involved organs were the lung in one, liver in two, bone in two, brain in one, and lung and liver synchronously in one.
Control within IOERT volume The 6-month, 1-, and 2-year control rates within IOERT volume were 41%, 19%, and 13%, respectively (Fig. 1).
Survival The 6-month, 1-, 2-, and 3-year actuarial survival rates were 51%, 21%, 21%, and 8%, respectively (Fig. 3). Median overall survival was 7 months (95% CI: 4.7–9.3). Those patients treated with 5– 8-cm diameter applicators survived longer (7 of 31 patients alive) than patients treated
Fig. 1. Actuarial control within the IOERT volume.
Fig. 2. Actuarial locoregional control.
RESULTS
1088
I. J. Radiation Oncology
●
Biology
●
Physics
Fig. 3. Actuarial overall survival.
with larger applicators (0 of 7 patients alive) (p , 0.009). This difference reflects the better prognosis of smaller volume recurrent disease, which could be encompassed within smaller applicators. Patients treated to neck sites had a better overall survival (4 of 14 patients alive) than those with primary (2 of 11), stomal (1 of 8) or multiple recurrences (0 of 6). This difference was of borderline significance (p 5 0.054). At the time of the analysis, 31 patients have expired; 26 of tumor recurrence and 4 of non-cancer-related causes without evidence of recurrence. One patient died due to treatment related complications (tracheovascular fistulae 1 month after IOERT). Five (13%) are alive with tumor 2, 10, 25, 32, and 33 months after IOERT. Complications Severe treatment-related complications occurred in six patients (16%) and included orocutaneous fistulae in two, tracheal dehiscence in one, wound dehiscence in one, carotid occlusion in one, and fatal tracheovascular fistula in one. DISCUSSION Patients with advanced, previously irradiated head and neck cancer have very limited treatment options. Even in patients with potentially resectable lesions, surgical resection is often unsuccessful due to the microscopic extension of the disease and the technical difficulties of surgical resection after radiation therapy. Radiation-induced fibrosis makes recognition of tumor boundaries and dissection through normal tissue planes very difficult. Recurrent tumors may be more radioresistant than primary head and neck cancers. In vitro cultured tumor cells from head and neck squamous cell carcinoma recurrences after a complete course of EBRT have been shown to be radioresistant (10). Previously irradiated tissues are more hypoxic and therefore
Volume 42, Number 5, 1998
more radioresistant. Furthermore, the inability to deliver meaningful doses of adjuvant EBRT in previously irradiated tumors makes local control unlikely (4, 5, 11). While IOERT combined with postoperative EBRT, achieves local control in 40 –70% of the patients with primary head neck tumors (1, 2, 8, 11–13), decreased local control and survival is seen in patients who do not receive or complete the supplementary postoperative EBRT (5, 14). Martinez-Monge and Calvo reported on a series of 31 patients with locally advanced or recurrent head and neck cancer treated with maximal surgical resection and 10 –15 Gy IORT at the University of Navarre, Spain (4). The 7-year control rate for 17 previously unirradiated patients was 46%, versus 19% for 14 previously irradiated patients. Six previously irradiated patients received low-dose EBRT (median 30 Gy). The inability to deliver EBRT because of prior EBRT was statistically significant for local control (p 5 0.004). In our previously published series, local control decreased from 79% to 50% in those patients who did not receive the postoperative EBRT (5). We integrated IOERT into a surgical salvage program in an attempt to improve local control and survival results in these previously irradiated patients with dismal prognosis. IOERT allows combining maximal debulking surgery and “in situ” radiotherapy, minimizing the possibility of geographical miss and avoiding irradiation of uninvolved structures. In the head and neck area, EBRT reirradiation of certain normal tissues might be a problem. Significant areas of bone and uninvolved aerodigestive mucosa would be reirradiated, even with the most sophisticated treatment planning. Some representative EBRT reirradiation series have reported a severe late sequelae rate of 9 –28% (15–18) with reirradiation doses in the 30 – 60 Gy range. The probability of tumor control is a function of the dose administered. Using the linear-quadratic bio-effective model, 15 Gy and 20 Gy IOERT are radiobiologically equivalent (using a/b 5 10 Gy) to 31.3 Gy and 50 Gy given as fractionated EBRT in terms of tumor effect (19). Radiobiologically, using single large doses of IOERT is disadvantageous because it does not permit the repair of sublethal damage of the surrounding normal tissues or reoxygenation or redistribution of tumor cells to more radiosensitive mitotic phases. This disadvantage is compensated for by the reduction of the dose to the normal tissues, with dose reductions of 20 –30% required to achieve the same isoeffect dose (20), depending on the assumed half-time repair of tumor cells. This disadvantage can also be minimized by giving IOERT as a boost to the tumor bed to supplement a moderate dose of EBRT to a greater volume, thus achieving higher doses to the tumor while minimizing morbidity. Reirradiation with IOERT and EBRT (with or without chemotherapy) has also been used with good results in recurrent previously irradiated colorectal cancers (21). Reirradiation with EBRT has been explored in several studies (15– 18). Some of these studies have used hyperfractionated regimens between 1.2–1.6 Gy B.I.D. with an interfraction gap of at least 6 h to minimize the incidence of late tissue
IOERT in head and neck cancer
damage (15, 18). Treated patients include a broad category of recurrent tumors or second primaries arising in a previously irradiated area. The treatments given included EBRT reirradiation (30 – 60 Gy), surgery and intracavitary or interstitial brachytherapy boost whenever possible. Overall, these reirradiation series have reported long-term local control in 27–36% of the patients (15–18). When areas at risk are technically inaccessible to the electron beam applicator (e.g., steeply sloped surfaces or narrow cavities), we employ intraoperative high-dose-rate brachytherapy (IOHDR) (14). The flexible applicators used in IOHDR are able to access narrow cavities and can homogeneously irradiate curved and sloping surfaces. The single dose employed has the same disadvantages previously described for IOERT; however, it is possible to deliver postoperative fractionated high-dose-rate brachytherapy several days after the surgery (perioperative brachytherapy), using the implanted catheters in situations in which further external beam radiation cannot be given. IOERT alone does not provide good control of recurrent, previously irradiated head and neck cancers. Our experience
●
S. NAG et al.
1089
is confirmed by other authors (7) who have reported high rates of locoregional failure (88.8%) in a series of 95 patients (94% of whom had previous EBRT). The poor control could be due to the limited IOERT doses that can be given to previously irradiated tumors. Higher IORT doses cause significant morbidity (22). Hence, alternative avenues being explored are the addition of postoperative EBRT (which is now routinely given) or brachytherapy (LDR or fractionated HDR) to maximal surgery and IOERT in an attempt to improve the dismal prognosis of these patients.
CONCLUSION IOERT alone, at the dose used, is not sufficient for control of recurrent, previously irradiated head and neck cancers. Since higher IOERT doses are associated with high morbidity, we are currently evaluating the addition of limited EBRT dose and/or brachytherapy to improve the local control of these poor prognostic recurrent tumors, with acceptable morbidity.
REFERENCES 1. Coleman C, Roach M III, Ling S, et al. Adjuvant electronbeam IORT in high-risk head and neck cancer patients. Front Radiat Ther Oncol 1997; 31:105–111. 2. Freeman SB, Hamaker RC, Singer MI, et al. Intraoperative radiotherapy of head and neck cancer. Arch Otolaryngol Head Neck Surg 1990; 116:165–168. 3. Garrett P, Pugh N, Ross D, et al. Intraoperative radiation for advanced or recurrent head and neck cancer. Int J Radiat Oncol Biol Phys 1987; 13:785–788. 4. Martı´nez-Monge R, Azinovic I, Alcalde J, et al. IORT in the management of locally advanced or recurrent head and neck cancer. Front Radiat Oncol 1997; 31:122–125. 5. Nag S, Schuller D, Pak V, et al. Intraoperative radiation therapy using electron beam or HDR brachytherapy for previously unirradiated head and neck cancers. Front Radiat Ther Oncol 1997; 31:112–116. 6. Rate W, Garrett P, Hamaker R, et al. Intraoperative radiation therapy for recurrent head and neck cancer. Cancer 1991; 67:2738 –2740. 7. Spaeth J, Andropulus D, Unger T, et al. Intra-operative radiotherapy 5 years of experience in the palliative treatment of recurrent and advanced head and neck cancers. Oncology 1997; 54:208 –213. 8. Toita T, Nakano M, Takizawa Y, et al. Intraoperative radiation therapy (IOERT) for head and neck cancer. Int J Radiat Oncol Biol Phys 1994; 30:1219 –1224. 9. Kaplan EL, Meier P. Nonparametric estimation for incomplete observations. J Am Stat Assoc 1958; 4:57– 81. 10. Weichselbaum R, Beckett M, Scwartz J, et al. Radioresistant tumor cells are present in head and neck carcinomas that recur after radiotherapy. Int J Radiat Oncol Biol Phys 1988; 15: 575–579. 11. Pearlman N. Treatment outcome in recurrent head and neck cancer. Arch Surg 1979; 114:39 – 42. 12. Kokal W, Neifield J, Eisert D, et al. Management of locore-
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
gional recurrent oropharyngeal carcinoma. Am J Surg 1983; 146:436 – 438. Regine W, Valentino J, Patel P, et al. Efficacy of postoperative radiation therapy for recurrent squamous cell carcinoma of the head and neck. Int J Radiat Oncol Biol Phys 1997; 39:297– 302. Nag S, Schuller D, Pak V, et al. Pilot study of intraoperative high dose rate brachytherapy for head and neck cancer. Radiother Oncol 1996; 41:125–130. Benchalal M, Bachaud JM, Francois P, et al. Hyperfractionation in the reirradiation of head and neckcancers. Result of a pilot study. Radiother Oncol 1995; 36:203–210. Levendag PC, Meeuwis CA, Visser AG. Reirradiation of recurrent head and neck cancers: External and/or interstitial radiation therapy. Radiother Oncol 1992; 23:6 –15. Stevens KR Jr, Britsch A, Moss WT. High-dose reirradiation of head and neck cancer with curative intent. Int J Radiat Oncol Biol Phys 1994; 29:687– 698. Tercilla OF, Schmidt-Ullrich R, Wazer DE. Reirradiation of head and neck neoplasms using twice-a-day scheduling. Strahlenther Onkol 1993; 169:285–290. Nag S, Martinez- Monge R, Gupta N. Intraoperative radiation therapy using electron beam and high dose rate brachytherapy. Cancer J 1997; 10:94 –101. Nag S, Orton C. Development of intraoperative high dose rate for treatment of resected tumor bed in anesthetized patients. Endocuriether/Hypertherm Oncol 1993; 9:187–193. Gunderson L, Haddock M, Nelson H, et al. Locally recurrent colorectal cancer: IOERT and EBRT 6 5-FU and maximal resection. Front Radiat Oncol 1997; 31:224 –228. Shaw E, Gunderson L, Martin K, et al. Peripheral nerve and ureteral tolerance to intraoperative radiation therapy: Clinical and dose-response analysis. Radiother Oncol 1990; 18:247– 255.