Abstract
This study evaluated survival in 154 patients with stages II–IV oral squamous cell carcinoma (OSCC) treated with preoperative concurrent chemoradiotherapy and assessed the future use of this treatment strategy. 14 patients exhibited advanced stage II, 73 exhibited stage III and 67 exhibited stage IV. All patients received 40 Gy irradiation and concurrent cisplatin-based chemotherapy in two courses. Radical surgery was undertaken after 2–6 weeks. The clinical tumour response, histopathologic regression grade, residual tumour grade (RGrade) in the primary tumour and the level of residual pN+ were associated with prognosis. 90% of patients with complete response and 73% of patients with good partial response in the primary tumour were RGrade 0 (no residual tumour cells) or RGrade 1 (viable tumour cells remained within central superficial portion). In patients with complete response in the neck, residual pN+ was only seen in levels IB (8%) and IIA (8%); the higher the level of residual pN+, the lower the survival rate ( p < 0.0001). This treatment strategy was excellent for stages II–IV OSCC. It may be possible to perform minimally invasive surgery in which the extent of resection in primary tumour and neck is reduced in patients who achieve good response following preoperative chemoradiotherapy.
Loco-regionally advanced squamous cell carcinoma of the oral cavity (OSCC) continues to be a major clinical problem. The treatment of advanced OSCC requires multimodal therapy to control the loco-regional disease better than either surgery or radiation therapy alone. Traditional therapy has consisted of surgery with postoperative radiotherapy. A high percentage of patients develop recurrent disease at the primary or regional sites within the first 2 years of treatment so 3- and 5-year survival rates are generally low (<40%) .
To improve these treatment outcomes, additional chemotherapy is applied concurrently with radiotherapy (RT) or alone as induction chemotherapy or adjuvant chemotherapy. For concurrent chemoradiotherapy (CRT), the beneficial effects of RT and chemotherapy were shown. As RT is likely to be more effective in otherwise unaffected and well-oxygenated tissue, the authors have adopted preoperative concurrent CRT in the management of advanced OSCC.
The authors reported previously that preoperative concurrent CRT in resectable advanced OSCC was tolerated and resulted in high rates of significant tumour responses and good prognosis. Several reports in the literature also demonstrated that this approach is feasible with tolerable adverse effects and results in promising overall and disease-free survival rates .
In the current study, the authors report data on long-term overall and disease-free survival in advanced OSCC with preoperative concurrent CRT and radical surgery. They discuss the future possibility of organ preservation with minimally invasive surgery for patients who achieve a good response with preoperative concurrent CRT.
Materials and methods
In this study, 154 patients with advanced but potentially resectable OSCC treated at the Department of Oral and Maxillofacial Surgery, Nara Medical University, from 1988 to 2008 were enrolled. The median age of the patients was 62.5 years (range 29–74). 92 patients were male and 62 were female. A minimum tumour size of 30 mm (advanced T2) was required. The presence of regional lymph nodes did not interfere with inclusion (T2-4, N0-3, M0). 14 patients exhibited advanced stage II, 73 exhibited stage III and 67 exhibited stage IV.
All patients had an Eastern Cooperative Oncology Group (ECOG) criteria performance status of <2, provided that there was acceptable risk for head and neck surgery, that they had adequate hematologic, renal and hepatic functions to receive cisplatin- or carboplatin-based chemotherapy, and that they had given informed consent for all aspects of the treatment regimen. Pretreatment evaluation was completed in all patients and included a complete medical history, physical examination, a computed tomography (CT) scan or magnetic resonance imaging (MRI) of the head and neck, chest X-ray, and other staging procedures for distant metastases if clinically indicated. Complete biochemical studies of liver, kidney (creatinine clearance), hematologic status, urinalysis and electrocardiogram were performed on all patients. Patients also underwent pretreatment dental evaluation with appropriate care and received instruction regarding diet and the use of tube feeding. Patient and tumour characteristics are shown in Table 1 .
| No. of patients | 154 |
| Male/female ratio | 92:62 |
| Median age (year) | 62.5 (29–74) |
| Median performance status | 1 (range 0–2) |
| Primary sites | |
| Tongue | 74 (48.1%) |
| Lower gingiva | 34 (22.1%) |
| Upper gingiva | 24 (15.5%) |
| Floor of mouth | 10 (6.5%) |
| Buccal mucosa | 7 (4.5%) |
| Hard palate | 5 (3.3%) |
| Stage | |
| Advanced II | 14 (9.1%) |
| III | 73 (47.4%) |
| IV | 67 (43.5%) |
| Radiotherapy 40 Gy | 154 (100.0%) |
| Chemotherapy regimen | |
| CBDCA + 5FU | 87 (56.5%) |
| CDDP + PEP | 40 (26.0%) |
| CDDP + 5FU | 23 (14.9%) |
| CDGP + 5FU | 4 (2.6%) |
Treatment regimen and follow-up
The rational for this treatment design has already been described , as have the details of concurrent chemotherapy and radiation therapy. All patients received 40 Gy at 2 Gy per day in 20 fractions of external beam irradiation to the primary tumour and lymphatics for 4 weeks. Concurrently, during the course of radiation therapy, the patients received cisplatin (15 mg/m 2 ) or carboplatin (70–100 mg/m 2 ) on days 1–3 and peplomycin (5 mg/day) or 5FU (500–750 mg/day) on days 4–7, which was run continuously for more than 4 h with intra-arterial infusion. A second cycle of this chemotherapy regimen was repeated in the fourth week of radiation therapy.
Patients were seen every day by a physician and fully evaluated (except for radiologic examination) at least once a week during treatment.
Radical surgery was undertaken 2–6 weeks after the completion of concurrent CRT and not later than 6 weeks after. The surgical decision was based on the pretreatment staging, and the surgery consisted of a composite resection of any residual primary tumour with immediate reconstruction using bone grafts, free flaps or myocutaneous flaps. Radical or modified radical neck dissections were carried out for all patients with clinically positive neck nodes. In cases of midline transgression neck dissection was performed bilaterally. An accurate assessment to ensure satisfactory tumour resection was achieved by combining visual examination with careful palpation and taking frozen sections during the operation. Resected tumour tissue and the cervical lymph node specimen were routinely histologically analysed for postoperative TNM stage.
Clinical response to therapy was determined at 2–3 weeks after completion of preoperative concurrent CRT. To evaluate response, the same diagnostic procedures (e.g. CT, MRI and clinical examination) as initially applied were performed. Standard criteria were used to assess response. Complete response (CR) was defined as complete clinical regression of tumour. Fair partial response (fair PR) was defined as >50% to <85% reduction and good partial response (good PR) was defined as >85% to <100% reduction in the cross-sectional area of a measurable tumour using the product of the two largest perpendicular diameters. Stable disease (SD) was defined as <50% reduction or <25% increase in the size of a measurable tumour . Histopathologic responses were defined on the basis of the regression grading system of S himosato et al. using semiserial sections of whole surgical specimens. The histopathologic regression grade was classified according to a 4-stage scale with some subclassifications: grade I, characteristic changes are noted in tumour cells, but tumour structures have not been destroyed (there is no defect in tumour nests resulting from the lysis of individual tumour cells); grade IIa, destruction of tumour structures is of a mild degree, ‘viable tumour cells’ are frequently observed; grade IIb, destruction of tumour structures is of a severe degree; grade III, markedly altered, presumably non-viable tumour cells are present singly or in small clusters and ‘viable cells’ are hardly seen; grade IV, no tumour cells remain in any sections. Those with grades IIb, III and IV were defined as good pathologic responders and those with grade I and grade IIa were poor responders. In particular, grades III and IV were considered as a pathological complete response (pCR) because viable tumour cells were hardly seen in grade III and no viable tumour cells were seen in grade IV.
The degeneration of tumour cells and the distribution of the remaining viable tumour cells were examined microscopically and defined as residual tumour grades (RGrades) originally . The distribution of the residual viable tumour cells in the primary tumour was classified using the following four grades: RGrade 0, no residual tumour cells; RGrade 1, viable tumour cells remained within the central submucosa or shallow in the muscularis propria (central superficial portion); RGrade 2, viable tumour cells remained widely within the superficial portion; RGrade 3, viable tumour cells remained in the central region and deep in the muscularis propria; and RGrade 4, viable tumour cells remained both widely and deeply distributed.
After the completion of therapy, patients were followed at regular and frequent intervals by members of the treatment team. Minimum intervals of recall were 1 month during the first year, 2–3 months from the second to third year, 4–5 months from the fourth to fifth year and twice a year until the tenth year. Careful clinical examinations were performed at each visit and suspected recurrences were biopsied. All patients were also checked for second primary carcinoma by panendoscopy, CT scan, MRI, chest X-ray, scintigrams and positron emission tomography (PET).
Survival curves were constructed using the Kaplan–Meier method and compared using the log-rank test. Correlations amongst the clinical response, histopathologic regression grade, and RGrades were compared using the χ 2 test. p values less than or equal to 0.05 were considered to indicate statistical significance.
Treatment regimen and follow-up
The rational for this treatment design has already been described , as have the details of concurrent chemotherapy and radiation therapy. All patients received 40 Gy at 2 Gy per day in 20 fractions of external beam irradiation to the primary tumour and lymphatics for 4 weeks. Concurrently, during the course of radiation therapy, the patients received cisplatin (15 mg/m 2 ) or carboplatin (70–100 mg/m 2 ) on days 1–3 and peplomycin (5 mg/day) or 5FU (500–750 mg/day) on days 4–7, which was run continuously for more than 4 h with intra-arterial infusion. A second cycle of this chemotherapy regimen was repeated in the fourth week of radiation therapy.
Patients were seen every day by a physician and fully evaluated (except for radiologic examination) at least once a week during treatment.
Radical surgery was undertaken 2–6 weeks after the completion of concurrent CRT and not later than 6 weeks after. The surgical decision was based on the pretreatment staging, and the surgery consisted of a composite resection of any residual primary tumour with immediate reconstruction using bone grafts, free flaps or myocutaneous flaps. Radical or modified radical neck dissections were carried out for all patients with clinically positive neck nodes. In cases of midline transgression neck dissection was performed bilaterally. An accurate assessment to ensure satisfactory tumour resection was achieved by combining visual examination with careful palpation and taking frozen sections during the operation. Resected tumour tissue and the cervical lymph node specimen were routinely histologically analysed for postoperative TNM stage.
Clinical response to therapy was determined at 2–3 weeks after completion of preoperative concurrent CRT. To evaluate response, the same diagnostic procedures (e.g. CT, MRI and clinical examination) as initially applied were performed. Standard criteria were used to assess response. Complete response (CR) was defined as complete clinical regression of tumour. Fair partial response (fair PR) was defined as >50% to <85% reduction and good partial response (good PR) was defined as >85% to <100% reduction in the cross-sectional area of a measurable tumour using the product of the two largest perpendicular diameters. Stable disease (SD) was defined as <50% reduction or <25% increase in the size of a measurable tumour . Histopathologic responses were defined on the basis of the regression grading system of S himosato et al. using semiserial sections of whole surgical specimens. The histopathologic regression grade was classified according to a 4-stage scale with some subclassifications: grade I, characteristic changes are noted in tumour cells, but tumour structures have not been destroyed (there is no defect in tumour nests resulting from the lysis of individual tumour cells); grade IIa, destruction of tumour structures is of a mild degree, ‘viable tumour cells’ are frequently observed; grade IIb, destruction of tumour structures is of a severe degree; grade III, markedly altered, presumably non-viable tumour cells are present singly or in small clusters and ‘viable cells’ are hardly seen; grade IV, no tumour cells remain in any sections. Those with grades IIb, III and IV were defined as good pathologic responders and those with grade I and grade IIa were poor responders. In particular, grades III and IV were considered as a pathological complete response (pCR) because viable tumour cells were hardly seen in grade III and no viable tumour cells were seen in grade IV.
The degeneration of tumour cells and the distribution of the remaining viable tumour cells were examined microscopically and defined as residual tumour grades (RGrades) originally . The distribution of the residual viable tumour cells in the primary tumour was classified using the following four grades: RGrade 0, no residual tumour cells; RGrade 1, viable tumour cells remained within the central submucosa or shallow in the muscularis propria (central superficial portion); RGrade 2, viable tumour cells remained widely within the superficial portion; RGrade 3, viable tumour cells remained in the central region and deep in the muscularis propria; and RGrade 4, viable tumour cells remained both widely and deeply distributed.
After the completion of therapy, patients were followed at regular and frequent intervals by members of the treatment team. Minimum intervals of recall were 1 month during the first year, 2–3 months from the second to third year, 4–5 months from the fourth to fifth year and twice a year until the tenth year. Careful clinical examinations were performed at each visit and suspected recurrences were biopsied. All patients were also checked for second primary carcinoma by panendoscopy, CT scan, MRI, chest X-ray, scintigrams and positron emission tomography (PET).
Survival curves were constructed using the Kaplan–Meier method and compared using the log-rank test. Correlations amongst the clinical response, histopathologic regression grade, and RGrades were compared using the χ 2 test. p values less than or equal to 0.05 were considered to indicate statistical significance.
Results
All patients but one were evaluable for clinical and histopathologic responses and RGrade. The overall clinical response rate for the primary tumours was 93%. The CR rate was 60%, the good PR rate was 17% and the fair PR rate was 16%. No progression of disease was observed in this study.
A histopathologic review of the resected specimen was carried out for all patients but one and revealed good pathologic responses (grades IIb, III and IV) in 85% and the pCR (grades III + IV) rate was 71%. 82 (89%) of 92 patients with CR and 19 (73%) of 26 patients with good PR had pCR, whilst 8 (33%) of 24 patients with fair PR and none of 11 patients with SD had pCR. The pCR rate was very low in fair PR and SD patients. The tumour regression rate was correlated with histopathologic regression grade. This showed that the better the clinical response, the better the histopathological response. In particular, more than 70% of patients with good PR and CR had pCR ( Table 2 ).
Stay updated, free dental videos. Join our Telegram channel
VIDEdental - Online dental courses
