Japanese Journal of Clinical Oncology 31:195-202 (2001)
© 2001 Foundation for Promotion of Cancer Research
Influence of Postsurgical Residual Tumor Volume on Local Control in Radiotherapy for Maxillary Sinus Cancer
Departments of 1Radiology and 2Head and Neck Surgery, National Cancer Center Hospital East, Kashiwa, Chiba and 3Department of Radiology, University of Tsukuba, Tsukuba, Ibaraki, Japan
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Background: The aim was to study the influence of postsurgical gross residual tumor volume on local control of maxillary sinus cancer treated with radiotherapy combined with debulking surgery.
Methods: Forty-three patients who underwent combined surgery and radiotherapy (50–72 Gy, median 60 Gy) for squamous cell carcinoma of the maxillary sinus were reviewed. Gross residual tumor volume (GRTV) after surgery was measured on computed tomograms obtained during the radiotherapy planning. Patients were classified according to GRTV as follows: group AA, GRTV = 0 (microscopic residual, n = 2); group A, GRTV <10 cm3 (n = 24); group B, 10–40 cm3 (n = 9); and group C, 
40 cm3 (n = 8). The relationship between local control and GRTV was analyzed using univariate and multivariate analysis.
Results: The 2-year local control rate for all patients was 62%. The differences in local control rates between groups AA, A and B were not significant (P > 0.05), but the rate was significantly lower in group C than in the other groups (69% at 2 years vs 31% at 1 year, P < 0.001). Multivariate analysis showed that GRTV (P = 0.002) and histological differentiation (poorly differentiated histology was favorable, P = 0.035) were independent prognostic factors and that intra-arterial chemotherapy and administered total dose were not. Local control in groups A and B significantly depended on the total dose of radiotherapy, with 2-year control rates of patients receiving 50 Gy (n = 6) and 60 Gy (n = 27) of 17% vs 79%, respectively (P < 0.001).
Conclusions: Our data suggest that adequate, not complete, debulking associated with a total radiotherapy dose of 60 Gy can provide satisfactory local control for patients with squamous cell carcinoma of the maxillary sinus.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Treating maxillary sinus cancer is challenging because of the difficult problem of the proximity of critical structures, such as the eye and the brain, which preclude wide surgical excision and high-dose radiotherapy. The clinical course is indolent at most and a substantial proportion of patients have advanced disease at the time of diagnosis. Combined-modality therapy consisting of surgery and radiotherapy (RT) with or without intra-arterial chemotherapy (IAC) is generally used to treat this disease and the reported 5-year local control and survival rates are 50–78 and 39–64%, respectively. However, an appropriate treatment strategy in terms of surgical procedure, RT methods and their sequence is still a matter of controversy (1–7).
In 1970, Sato et al. reported that curtailment of the surgical procedures, with the aim of preserving the bony framework of the maxillary antrum, combined with RT and IAC resulted in a significantly better cosmetic and functional outcome as well as an improvement in survival (8) and since then many authors have insisted on the benefits of piecemeal debulking combined with radiotherapy (4,9,10). However, conventional two-dimensional RT caused radiation-induced damage of surrounding normal tissues, such as the brain, bone and soft tissue, that was frequently dose-limiting and a total dose of 50–70 Gy per 5–7 weeks of RT was usually adopted (1–10). On the other hand, both we and others have reported that whether the tumor volume is less or greater than 30–40 cm3 is crucial to predicting tumor radiocurability of various head and neck cancers with a total dose of 60–70 Gy by conventional or hyperfractionated regimens (11–13). In this context, how much postsurgical residual tumor volume can be eradicated by a total dose of around 60 Gy per 6 weeks of RT becomes a clinically relevant question. This retrospective study was therefore undertaken to analyze the interrelationship between gross residual tumor volume (GRTV) after debulking surgery and local control after RT at a total dose of 50–72 Gy per 5–7 weeks by the conventional two-dimensional technique.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Patients’ Characteristics
Of the 48 patients with biopsy-proven squamous cell carcinoma of the maxillary sinus who presented between October 1992 and February 1999, the 43 patients who underwent surgical excision combined with radiotherapy at a total dose of 50 Gy or more were reviewed. No patients had distant metastasis at the time of diagnosis. Five patients were excluded: two because of extensive and inoperable diseases who were treated with palliative radiotherapy alone, two with T3N0 cancer treated with 40 Gy followed by total maxillectomy and one with bilateral disease.
Disease was classified according to the TNM classification (UICC, 1992) and the results are summarized in Table 1. Twenty-six (60%) patients had tumors which invaded sphenoid sinus, orbital contents beyond the floor or base of skull (T4) and seven (16%) had neck metastasis at the time of diagnosis. One patient with T4N0 disease had a history of T3N2 hypopharyngeal cancer treated by total pharyngolaryngectomy and bilateral neck dissection 3.5 years before treatment for the maxillary sinus cancer and died of local recurrence of the maxillary sinus cancer 10 months after the commencement of RT. No other patients had a history of previous malignancy. There were 38 men and five women and their median age was 56 years (range: 36–84 years). The length of follow-up of the 30 survivors was 3–72 months (median: 26 months). Two patients were lost to follow-up at 3 and 8 months and both experienced unresectable local failures. All other patients were followed for more than 1 year.
|
Treatment
Surgery, consisting of 37 partial and six total maxillectomies (three with orbital exenteration), was carried out according to the extent of the disease. The transoral approach was mainly adopted and three (12%) of 26 patients with T4 disease underwent total maxillectomy with orbital exenteration. In the remaining patients, the orbital floor was preserved after surgery, even though the tumor had invaded the orbital floor. Thirty-six patients underwent postoperative radiotherapy (three followed by necrotomy). Six were treated by pre- and postoperative radiotherapy (sandwich treatment) and one was treated by partial maxillectomy followed by sandwich treatment (Table 2). Pathological examination revealed a positive resection margin in every case and all but two had macroscopic residual disease within the irradiated volumes. Two patients who had small disease causing erosion of the hard palate or lateral antral wall (T2) had only microscopic residual tumor. Of the seven patients presenting with nodal diseases at the time of diagnosis, six underwent neck dissection and one received definitive RT for nodal disease. No patients without palpable nodes received elective neck irradiation.
|
Combined chemotherapy with various regimens was applied to 39 patients according to the judgement of the physicians in charge. Thirty-five underwent IAC with continuous infusion of 5-FU via the superficial temporal artery (daily dose 150–250 mg/body) during radiotherapy. Total doses of 5-FU varied between 1500 and 4400 mg (median 3000 mg). Of the remaining eight patients, three underwent systemic bolus infusion of cisplatin at a daily dose of 6 mg/m2 at 30 min prior to RT to a total dose of 84, 132 and 180 mg/body and one underwent a single course of systemic chemotherapy with cisplatin (40 mg/m2 on day one of RT) and 5-FU (400 mg/m2 on days 1–4) concurrent with RT. Four did not receive any chemotherapy.
RT was carried out mainly with a wedged-pair two-field technique using 6 MV X-rays. The total tumor doses ranged from 50 to 72 Gy (median 60 Gy) and they were given at 2 Gy/fraction, excluding one patient with huge T4 disease invading the skull base who received sandwich treatment. This patient received 30 Gy preoperative RT at 1.5 Gy twice daily and postoperative RT with 42 Gy by conventional fractionation. Three patients received a total dose of 70–72 Gy according to the judgement of their physicians in view of their far-advanced disease. Eight patients in the early period received a total dose of 50 Gy. The remaining 32 patients received 60 Gy/30 fractions. The median elapsed RT treatment time was 46 days, with a range of 36–90 days and 41 patients (95%) completed their RT within 2 months. The median duration of treatment as a whole, consisting of surgery and RT, was 55 days, ranging from 46 to 96 days, and in 35 cases (81%) it was completed within 2 months.
When possible, relapse at the primary site was treated by further surgical excision and isolated nodal recurrence was treated by salvage neck dissection.
CT Volumetry of Gross Residual Tumor Volume (GRTV)
Tumor extension into the wall of the maxillary sinus, orbit, pterygopalatine and infratemporal fossa, nasopharynx, ethmoid and sphenoid sinus, frontal sinus and skull base was estimated separately based on preoperative computed tomography (CT) scans and/or magnetic resonance imaging (MRI). Intraoperative findings were also reviewed to confirm the estimates and the residual volume of soft tissue densities, regarded as residual tumors, within each compartment described above was measured on CT scans that were obtained during RT planning after surgical excision. Therefore, in patients who received sandwich treatment, GRTV means residual tumor volume prior to the postoperative segment of the RT. The residual tumor was outlined with a mouse device and the area was measured with a built-in image analysis system. GRTV was calculated by multiplying each cross-sectional area by slice thickness and adding the slices together. The slice thickness was 5 mm in all but two patients, in whom it was 7 mm. CT-based RT planning was performed with a Toshiba X Force/SH CT scanner and CT PORT software.
Data Analysis
Local control was the endpoint of this study. Overall, local relapse-free and disease-specific survival were also analyzed. Survival time was calculated from the commencement of RT. In local recurrence-free survival (LFS), local recurrences and disease-specific death were counted as failures regardless of all other types of recurrence (nodal and distant). In disease-specific survival (DSS), those who survived after salvage treatment for any types of recurrence were considered successes. These endpoints were estimated according to the Kaplan–Meier method (14) and significance was evaluated by the log-rank test. Multivariate analysis was performed by Cox’s proportional hazards analysis (15).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
GRTV
Of the five patients with T2 disease, in two the gross lesions were completely excised by transoral partial maxillectomy; however, pathological examination of surgical specimens revealed that tumor cells were exposed at the cut edges and GRTV was estimated to be 0 cm3 in both of them (group AA). The remaining 41 patients had macroscopic residual disease after excision and the GRTV estimates are shown in Fig. 1. The median GRTV was 7 cm3 and ranged from 5 to 196 cm3. We classified patients into group A, B or C according to their GRTV as follows: group A, GRTV < 10 cm3 (i.e. almost complete resection of the primary tumor); group B, 10 cm3
GRTV < 40 cm3; and group C, GRTV 40 cm3 (11–13). One patient in group B had a GRTV of 33 cm3 (Fig. 2). The patient remained disease-free for 43 months after treatment by transoral debulking followed by 60 Gy of RT. All of the others in group B had a GRTV of 25 cm3. The six patients who had measurable GRTV (5–20 cm3, median 5 cm3) had received only 50 Gy/5 weeks in the early period.
|
|
Local Control
Both 2- and 5-year local control rates for all patients were 62% (95% confidence interval: 46–78%). No statistically significant differences in local control rates were observed between the patients in groups AA, A and B, as shown in Fig. 3. Two patients in group AA, who received a total RT dose of 50 Gy, remained disease-free for 68 and 72 months after RT.
|
Three of the 23 patients in groups A and B who received a total RT dose of
60 Gy combined with IAC experienced local failure (2-year local control rate 86%, range 70–100%). However, in two the failure occurred in the anterior portion of the ethmoid sinus, which was outside the RT-treated volume. Preoperative MRI images in both cases showed gross invasion by the tumor in this region and these failures were therefore attributed to inadequate RT planning. Among the remaining 21 patients, only one experienced local failure and the 5-year local control rate was 95% (range 85–100%). Five of the six patients in groups A and B who received only 50 Gy for their postsurgical residual tumor and five of the eight patients in group C also experienced local failure within the treated volume. Univariate analysis revealed a statistically significant difference in local control rates according to GRTV (group C vs the others, 2-year local control rate: 31 vs 69%, P = 0.001). In addition, histology (poorly vs moderately-well differentiated, 2-year local control rate: 100 vs 48%, P = 0.010) and IAC (+ vs –: 72 vs 18%, P = 0.018) also had a significant influence on local control. T stage (T2/3 vs T4: 76 vs 51%, P = 0.176) and sequence of surgery and RT (postoperative RT vs others: 63 vs 60%, P = 0.541) had no influence on local control (Table 3). The total dose of RT also had no influence on local control in the patients as a whole (60 Gy vs 50 Gy: 68 vs 38%, P = 0.132); however, it did have a significant influence among the patients in groups A and B (79 vs 17%, P < 0.001, Fig. 4). Among the patients in groups A and B who received 60 Gy of RT, IAC remained a significant prognostic factor for local control; however, only four patients in this group underwent RT without IAC. All four had advanced T4 disease requiring total maxillectomy including orbital exenteration in one and two of the four experienced local failure within the treated volume.
|
|
Multivariate analysis revealed that GRTV (P = 0.002) and histology (P = 0.034) were independent prognostic indicators for local control and that IAC was not (P = 0.203). Among the patients with moderate to well differentiated histology, the influence of GRTV on local control remained significant (57% at 2 years vs 17% at 6 months, P < 0.001).
Nodal and Distant Failures
Six (including three of the seven patients presenting with nodal diseases) of 43 patients experienced nodal recurrences as the first events of treatment failure and four were successfully salvaged. Failures at remote sites occurred in the lungs in four patients and in bone in one, and in three they were the first events of treatment failure. The incidence of nodal and distant failure was not correlated with the treatment procedure.
Survival
The cumulative OAS and DSS at 2 years were 72% (range 57–86%) and 77% (range 63–90%), respectively. DSS at 2 years for the patients in groups AA, A and B and in group C were 82% (range 67–97%) and 29% (range 0–63%), respectively (P < 0.001, Fig. 5). All relapses had occurred within 2 years and it was estimated that the 2-year LFS for the patients in groups AA, A and B was 67% (range 50–83%) and that in group C, it was 16% (range 0–44%) (P < 0.001, Fig. 6).
|
|
Complications
One patient was found to have bone necrosis in the zygomatic arch 1 year after RT with a total dose of 72 Gy by sandwich treatment and partial maxillectomy. This patient was treated conservatively, but experienced nodal and distant failure and died 35 months after RT. Soft tissue necrosis and cellulitis without local failure were observed in two patients and one required ligation of external carotid artery 4 months after RT. No blindness or vision impairment requiring aggressive treatment of the ipsilateral eye was observed in any of the patients, excluding those who underwent orbital exenteration. Brain necrosis was not observed.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Surgery combined with RT (bimodality therapy) with or without chemotherapy is generally the treatment of choice for maxillary sinus cancer; however, no standard treatment has yet been established. Total maxillectomy with or without orbital exenteration aimed at en bloc excision of the tumor frequently leads to cosmetic deterioration or impairment of eye movement, while transoral piecemeal debulking aimed at preservation of the bony framework of the antrum, which is a unique strategy for this disease, frequently leaves gross residual tumor at the excision site. The latter procedure is preferable in terms of cosmetic and functional outcome when the residual tumor can be eradicated by RT. Therefore, it has been our policy to perform transoral piecemeal debulking followed by RT to manage this disease whenever possible.
Although many clinicopathological and biological parameters have been reported as indicators of tumor radiocurability, obviously one of the most important prognostic factors is the volume of the tumor (11–13,16). In this study, there were significant differences in local control between the patients in groups AA, A and B and in group C. Uni- and multivariate analysis showed that GRTV was an independent prognostic factor for local control, whereas T classification (T1–3 vs T4) was not. These findings suggest that the treatment procedure should not be selected according to the T stage, but on the basis of GRTV according to precise estimations of tumor invasion and the anticipated limits of debulking procedures at the time of diagnosis.
Measurement of GRTV on CT images is difficult, mainly because of postoperative inflammatory change. However, we think that adequate estimates were obtained with regard to tumor extension into adjacent areas according to the preoperative and intraoperative findings. In addition, it is noteworthy that the local control rates of the patients in groups AA/A and B were not significantly different, as shown in Fig. 3. Furthermore, the local control rate was 95% at both 2 and 5 years among the patients in groups A and B, who not only received a total dose of 60 Gy or more concurrent with IAC but also adequate RT planning. These findings suggest that adequate debulking followed by definitive RT at a total dose of 60 Gy offers the same possibility of local cure as extensive resection. A total dose of 50 Gy is not considered adequate to eradicate postsurgical residual tumor, with few exceptions, such as small T2 lesions classifiable in group AA. All but one patient (GRTV = 33 cm3) in group B had a GRTV of 25 cm3, suggesting that a GRTV of less than around 30 cm3 is a more reliable predictor of local cure after a total RT dose of 60 Gy.
The role of IAC was not conclusive for the following reasons: (1) there were considerable variations of total 5-FU dose among patients who received IAC; (2) the fact that all four patients who did not undergo IAC had far-advanced T4 disease that required total maxillectomy, including orbital exenteration in one; and (3) a 5-year local control rate of more than 74%, which is comparable to the results in our series, has been reported after surgery plus postoperative RT without IAC (6). However, in view of the fact that 95% of the 2-year local control rate was achieved in such patients receiving IAC as mentioned above, further study is needed to determine the benefit of IAC.
Longer follow-up is needed to estimate accurately the incidence of late adverse events. Jiang et al. reported latency periods for the onset of adverse events of 5–74 months (6). The incidence of detrimental late events has been reported to increase at a total dose of >65–70 Gy (7). Dose escalation beyond this dose should be performed with caution, considering the risks and benefits, and it should be restricted to cases of far-advanced disease or cases associated with medical problems that preclude adequate surgical debulking. It was also unclear whether pre- or postoperative RT was more suitable for this disease in this study. Because adequate local control rates were obtained with postoperative RT in groups AA, A and B, we consider that this is a standard procedure for this disease at present.
In conclusion, the results of this study suggest that adequate debulking of squamous cell tumor of the maxillary sinus followed by 60 Gy of RT offers a good possibility of local cure even when a certain amount of macroscopic tumor remains unresected. This strategy does not compromise cosmetic and functional results, in contrast to wide resection of the maxillary sinus. Therefore, this procedure should be considered as the treatment of choice for patients whose tumor is adequately debulkable even if cosmetically important bony framework of the maxillary sinus, such as orbital floor, is preserved. Aggressive wide resection or further escalation of total RT dose should be considered for those who have more advanced disease.
|
FOOTNOTES |
|---|
+ For reprints and all correspondence: Mitsuhiko Kawashima, Department of Radiology, National Cancer Center Hospital East, 6–5–1 Kashiwanoha, Kashiwa, 277-8577 Chiba, Japan. E-mail: mkawashi@east.ncc.go.jp
Abbreviations: RT, radiotherapy; IAC, intra-arterial chemotherapy; GRTV, gross residual tumor volume; CT, computed tomography; MRI, magnetic resonance imaging; LFS, local recurrence-free survival; DSS, disease-specific survival
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
1 Bush SE, Bagshaw MA. Carcinoma of paranasal sinuses. Cancer 1982;50:154–8.[Web of Science][Medline]
2 Hu Y, Tu G, Qi Y, Xu G, Wu X, Cai W, et al. Comparison of pre- and postoperative irradiation in the combined treatment of carcinoma of maxillary sinus. Int J Radiat Oncol Biol Phys 1982;8:1045–9.[Medline]
3 Knegt P, de Jong PC, van Andel JG, de Boer MF, Eykemboom W, van der Schans E. Carcinoma of the paranasal sinuses. Cancer 1985;56:57–62.[Medline]
4 Sakai S, Hohki A, Fuchihata H, Tanaka Y. Multidisciplinary treatment of maxillary sinus carcinoma. Cancer 1983;52:1360–4.[Medline]
5 Tsujii H, Kamada T, Arimoto T, Mizoe J, Shirato H, Matsuoka Y, et al. The role of radiotherapy in the management of maxillary sinus carcinoma. Cancer 1986;57:2261–6.[Medline]
6 Jiang GL, Ang K, Peters LJ, Wendt CD, Oswald MJ, Goepfert H. Maxillary sinus carcinomas: natural history and results of postoperative radiotherapy. Radiother Oncol 1991;21:193–200.[Web of Science][Medline]
7 Paulino AC, Marks JE, Bricker P, Melian E, Reddy SP, Emami B. Results of treatment of patients with maxillary sinus carcinoma. Cancer 1998;83:457–65.[Medline]
8 Sato Y, Morita M, Takahashi H, Watanabe N, Kirikae I. Combined surgery, radiotherapy and regional chemotherapy in carcinoma of the paranasal sinuses. Cancer 1970;25:571–9.[Medline]
9 Sakata K, Aoki Y, Karasawa K, Nakagawa K, Hasezawa K, Muta N. Analysis of the results of combined therapy for maxillary carcinoma. Cancer 1993;71:2715–22.[Medline]
10 Itami J, Uno T, Aruga M, Ode S. Squamous cell carcinoma of the maxillary sinus treated with radiation therapy and conservative surgery. Cancer 1998;82:104–7.[Medline]
11 Bentzen SM, Thames HD. Tumor volume and local control probability: clinical data and radiobiological interpretations. Int J Radiat Oncol Biol Phys 1996;36:247–51.[Medline]
12 Dubben HH, Thames HD, Beck-Bornholdt HP. Tumor volume: a basic and specific response predictor in radiotherapy. Radiother Oncol 1998;47:167–74.[Medline]
13 Kawashima M, Ogino T, Fujii H, Ishikura S, Ito Y, Ikeda H. Local-regional control by conventional radiotherapy according to tumor volume in patients with squamous cell carcinoma of the pharyngolarynx. Jpn J Clin Oncol 1999;29:467–73.
14 Kaplan EL, Meier P. Non-parametric estimation from incomplete observation. J Am Statist Assoc 1958;53:457–81.[Web of Science]
15 Cox DR. Regression models and life-tables. J R Statist Soc 1972;34:187–220.
16 Brenner DJ. Dose, volume and tumor-control predictions in radiotherapy. Int J Radiat Oncol Biol Phys 1993;26:171–9.[Medline]
Received October 12, 2000; accepted January 15, 2001.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||