Intensity modulated radiation treatment for head and neck squamous cell

Intensity-modulated radiation treatment for head-and-neck squamous cell carcinoma—the University of Iowa experience Presented in part at the 90th Scientific Assembly and Annual Meeting of the Radiological Society of North America, Chicago, IL, November 28–December 3, 2004. Min Yao M.D., Ph.D., Kenneth J. Dornfeld M.D., Ph.D., John M. Buatti M.D, Mark Skwarchuk Ph.D., Huaming Tan M.S., Thanh Nguyen M.D., Judith Wacha C.M.D., John E. Bayouth Ph.D., Gerry F. Funk M.D., Russell B. Smith M.D., Scott M. Graham M.D., Kristi Chang M.D. and Henry T. Hoffman M.D. Department of Radiation Oncology, University of Iowa Health Care, Iowa City, Iowa, USA Department of Otolaryngology, University of Iowa Health Care, Iowa City, IA Department of Biostatistics, College of Public Health, University of Iowa, Iowa City, IA Received 17 December 2004; revised 8 February 2005; accepted 14 February 2005. Available online 16 May 2005. Purpose: To review the University of Iowa experience with intensity-modulated radiotherapy (IMRT) in the treatment of head-and-neck squamous cell carcinoma. Methods and Materials: From October 1999 to April 2004, 151 patients with head-and-neck squamous cell carcinoma were treated with IMRT for curative intent. One patient was lost to follow-up 2 months after treatment and therefore excluded from analysis. Of the remaining 150 patients, 99 were treated with definitive IMRT, and 51 received postoperative IMRT. Sites included were nasopharynx, 5; oropharynx, 56; larynx, 33; oral cavity, 29; hypopharynx, 8; nasal cavity/paranasal sinus, 8; and unknown primary, 11. None of the patients treated with postoperative IMRT received chemotherapy. Of 99 patients who had definitive IMRT, 68 patients received concurrent cisplatin-based chemotherapy. One patient received induction cisplatin-based chemotherapy, but no concurrent chemotherapy was given. Three clinical target volumes (CTV1, CTV2, and CTV3) were defined. The prescribed doses to CTV1, CTV2, and CTV3 in the definitive cohort were 70–74 Gy, 60 Gy, and 54 Gy, respectively. For high-risk postoperative IMRT, the prescribed doses to CTV1, CTV2, and CTV3 were 64–66 Gy, 60 Gy, and 54 Gy, respectively. For intermediate-risk postoperative IMRT, the prescribed doses to CTV1, CTV2, and CTV3 were 60 Gy, 60 Gy, and 54 Gy. Results: The median follow-up was 18 months (range, 2–60 months). All living patients were followed for at least 6 months. There were 11 local–regional failures: 7 local failures, 3 regional failures, and 1 failure both in the primary tumor and regional lymph node. There were 16 patients who failed distantly, either with distant metastasis or new lung primaries. The 2-year overall survival, local progression-free survival, locoregional progression-free survival, and distant disease-free survival rates were 85%, 94%, 92%, and 87%, respectively. The median time from treatment completion to local–regional recurrence was 4.7 months (range, 1.8 to 15.6 months). Only one marginal failure was noted in a patient who had extensive tonsil cancer with tumor extension into the orbit and cavernous sinus. Patients with oropharyngeal cancer did significantly better than patients with oral cavity and laryngeal cancer, with a 2-year local–regional control rate of 98%, compared with 78% for oral cavity cancer and 85% for laryngeal cancer (p = 0.005). There was no significant difference in local–regional control for patients who received postoperative radiation or definitive radiation (p = 0.339) and for patients who had chemotherapy or not (p = 0.402). Neither T stage nor N stage had a significant effect on local–regional control (p = 0.722 and 0.712, respectively). Conclusions: Our results have confirmed the effectiveness of IMRT in head-and-neck cancer. It offers excellent outcomes in local–regional control and overall survival. More studies are necessary to further improve the outcomes of laryngeal cancer as well as oral cavity cancer. Keywords: Dose prescription; Head-and-neck cancer IMRT; Patterns of failure; Target volume delineation Article Outline Introduction Methods and materials Patient characteristics Chemotherapy Surgery Simulation and treatment delivery Determination of gross tumor volume Determination of clinical target volumes Determination of normal structures Dose prescription Follow-up Analysis of failure Statistical analysis Results Treatment outcomes Characteristics of local–regional failures Prognostic factors Discussion Acknowledgements References Introduction Radiation treatment is a major modality in head-and-neck cancer management. It can be given postoperatively as an adjuvant treatment or as definitive treatment with or without chemotherapy. Because a high dose of radiation is required for local–regional control and because many critical structures, including parotid glands, spinal cord, orbits, and brainstem, are immediately adjacent to the treatment targets, there are significant short-term and long- term side effects. Intensity-modulated radiation treatment (IMRT) is a highly conformal radiation technique that allows delivery of high radiation doses to the gross tumor, bulky lymph nodes, and high-risk areas while sparing the adjacent organs. Intensity-modulated radiation treatment is ideal for head-and-neck cancer treatment. Recent studies on IMRT in headand-neck cancer have shown that IMRT can potentially improve local–regional control (1 and 2), reduce side effects (especially xerostomia) (3 and 4), and improve quality of life (5 and 6). However, it is noted that long-term outcome data of IMRT in head-and-neck cancer are limited and that there is great variation in the IMRT delivery technique, including target delineation and dose prescription. It is important to confirm that the long-term favorable outcomes can be reliably reproduced. To develop a more consistent treatment technique upon which further clinical trials can be developed, it is also important to compare the long-term outcomes results, including tumor control, quality of life, and IMRT technique used by each institute. Since October 1999, we have been using IMRT to treat head-and-neck cancer in our institution. Initially, the primary application was for reirradiation in patients who had failures from previous conventional radiation. Since 2002, we have been using IMRT as our standard approach to head-and-neck cancer patients. The IMRT techniques have evolved during this period. The objectives of this study were to review our experience and report the treatment outcomes in using IMRT in head-and-neck squamous cell carcinoma. This was a retrospective study approved by the Institutional Review Board of the University of Iowa. Methods and materials Patient characteristics From October 1999 to April 2004, 151 patients with head-and-neck squamous cell carcinoma (excluding skin and parotid squamous cell carcinoma) completed IMRT in the Department of Radiation Oncology, University of Iowa. All patients were treated with curative intent. One patient lost to follow-up 2 months after treatment was excluded from analysis. The remaining 150 patients were analyzed in this study. Of the 150 patients, 118 were men and 32 were women, with a median age of 56 years (range, 20–90 years). ninetynine patients were treated with definitive IMRT, and 51 patients received postoperative IMRT. Sites included were nasopharynx, 5; oropharynx, 56; larynx, 33; oral cavity, 29; hypopharynx, 8; nasal cavity/paranasal sinus, 8; and unknown primary, 11. The clinical characteristics of these patients are summarized in Table 1. Stage distribution is summarized in Table 1 and Table 2. The American Joint Committee on Cancer staging system was used, according to the primary tumor location. The 2 patients with stage rT0N2c were patients with nasal cavity cancer who developed bilateral cervical lymph node metastasis after the primary tumor was controlled. Table 1 Patient characteristics Sex Male Female Age Radiation therapy (IMRT) Definitive With chemotherapy Without chemotherapy Postoperative Initial primary tumor site Hypopharynx Larynx Nasopharynx Oral cavity Oropharynx Sinus/nasal cavity Unknown primary AJCC stage Stage I Stage II Stage III Stage IV Unknown 1 (0.6) 10 (6.7) 25 (16.7) 103 (68.7) 11 (7.3) 8 (5.3) 33 (22) 5 (3.3) 29 (19.3) 56 (37.3) 8 (5.3) 11 (7.3) 99 (66) 69* 30 51 (34) 118 (79) 32 (21) 56 (20–90) Abbreviations: AJCC = American Joint Committee on Cancer IMRT = intensity-modulated radiotherapy. Data are presented as n (%), except for age, presented as median (range). One with induction chemotherapy only. Table 2 AJCC stage distribution of 150 patients N0 T0 Tx T1 T2 T3 T4 0 0 1 10 9 16 N1 N2 N3 Total 0 0 1 7 6 4 18 2 9 14 20 19 20 84 0 2 2 2 1 5 12 2 11 18 39 35 45 150 Total 36 Abbreviation: AJCC = American Joint Committee on Cancer. These patients with nasal cavity cancer presented with cervical lymph node metastasis after the primary tumors were controlled. Chemotherapy Of 99 patients who had definitive IMRT, 68 patients received concurrent cisplatin-based chemotherapy. Most of them received cisplatin 100 mg/m2 every 3 weeks. One of them received concurrent intra-arterial cisplatin chemotherapy. One patient received induction cisplatin-based chemotherapy, but no concurrent chemotherapy was given. Thirty of 99 patients with definitive IMRT did not received chemotherapy. None of 51 patients treated with postoperative IMRT received chemotherapy. Surgery Fifty-one patients had definitive surgery followed by postoperative IMRT. Of 99 patients who had definitive IMRT, 6 had neck dissection or excision of the cervical lymph nodes before IMRT, and 9 had neck dissection after IMRT because of persistent cervical lymphadenopathy. For patients who have complete response after treatment, planned neck dissection is not routinely performed at our institution. Simulation and treatment delivery Patients were placed in a supine position on a head support and immobilized with a thermoplastic facemask. Computed tomographic (CT) imaging from the vertex to 2 cm below the clavicle was obtained with the patient immobilized in the treatment position. Slice thickness and spacing were 3 mm throughout imaging. The CT images were transferred to the inverse treatment planning system (Corvus Treatment Planning System, NOMOS, version 3.0, Cranberry Township, PA). The treatment target volumes and critical structures were outlined in each CT slice as described below. All patients reported in this study were treated with the Peacock system (Nomos Corporation, Cranberry Township, PA) with multivane intensity-modulating collimator. Determination of gross tumor volume For definitive IMRT, gross tumor volume (GTV) included the primary tumor and involved cervical lymph nodes. The extent of the primary tumor was primarily determined by diagnostic CT images and endoscopic findings. Magnetic resonance imaging (MRI) was also obtained for nasopharyngeal cancer, and for some cases of base of tongue and tonsil primary tumors. [18F] fluorodeoxyglucose positron emission tomography (FDG-PET) was obtained for most patients as part of the staging workup, and abnormal FDG foci were correlated with CT images. Since August 2003, we have been using a CTI Biograph PET/CT system (Siemens Medical Systems, Hoffman Estates, IL) that obtains PET and CT images concurrently and produces coregistered images for head-and-neck cancer patients. This system provides valuable information for accurate delineation of GTV. An involved lymph node was defined as any lymph node with focal necrosis, or >1.0 cm in diameter on CT or MR imaging, or with abnormal focal FDG uptake in the staging PET scan. For postoperative IMRT, the GTV is the tumor bed that includes the preoperative volume of the primary tumor and involved lymph nodes. This is generally determined from the preoperative diagnostic imaging, surgical and pathologic findings, and surgical defects and postoperative changes on the postoperative CT scan. All cases were presented in the multidisciplinary head-and-neck tumor board. All radiologic images were reviewed with radiologists and surgical and pathologic findings with surgeons and pathologists before target delineation. Determination of clinical target volumes Three clinical target volumes, CTV1, CTV2, and CTV3, are defined and outlined on each CT slice. Clinical target volume 1 is defined as GTV (including primary tumor and involved lymph nodes) with 5–10-mm margins based on clinical and radiologic information. Clinical target volume 2 is the high-risk areas harboring microscopic disease. This includes normal structures immediately surrounding the CTV1 with high risk of local tumor invasion (primary tumor CTV2) and the high-risk lymphatic regions determined by the primary site, T stage, and involved lymph nodes (lymphatic CTV2). Clinical target volume 3 is defined as the intermediate-risk lymphatic areas. The primary tumor CTV2 volume depends on the location of the primary tumor and possible microscopic extension of the tumor. For example, in nasopharyngeal cancer, the CTV2 includes the entire nasopharynx, base of skull, superior parapharyngeal space, posterior third of nasal cavity, and posterior third of maxillary sinus. For tonsil and base of tongue cancer, the CTV2 generally includes 2 cm around the primary tumor, tailored by potential extension of the tumor. This often includes the entire base of tongue and contralateral tonsil. For hypopharyngeal cancer, the CTV2 is defined as the entire hypopharynx because there is no barrier inside the hypopharynx to prevent cancer spread. This is also true for laryngeal cancer; the entire larynx is defined as CTV2. If a tracheostomy was performed for patients with hypopharyngeal or laryngeal cancer and the stoma was included in the IMRT field, the stoma was also outlined as CTV2 in both definitive IMRT and postoperative IMRT cases. In some cases, the stoma was treated in the low anterior neck field that matched the IMRT field. It is treated to 60 Gy, regardless of the technique used to prevent recurrence at the stoma. The lymphatic CTV2 includes the involved lymphatic regions with 1 to 2 cm cranial and caudal margins from the involved lymph nodes and high-risk uninvolved lymphatic regions. The nodal classification is evolving, and there are several guidelines developed (7, 8, 9, 10, 11, 12 and 13). We based our approach on the one presented by Som et al. (7). The risk of lymph node involvement is based on the surgical and pathologic findings in the literature (14, 15 and 16) and has been recently reviewed by several investigators (10, 13 and 17). We generally include uninvolved lymph nodal regions with a probability of microscopic involvement higher than 15–20% in lymphatic CTV2. The retropharyngeal lymph nodes are also included in CTV2 for primary tumors in nasopharynx, oropharynx, and hypopharynx, or any tumor with posterior pharyngeal wall involvement. Clinical target volume 3 includes lymph nodal regions with lower risk of microscopic involvement. In general, this includes the nodal levels outside lymphatic CTV2 that would be included in a conventional radiation field. Determination of normal structures The parotid glands, spinal cord, optic nerves, orbits, brainstem, and posterior neck skin are outlined. The deep lobes of the parotids might be omitted in cases in which sparing the deep lobe would compromise tumor target coverage. The mandible is identified and spared to the extent that target dose is not compromised, especially for tumors in the oral cavity. In cases with no risk of larynx involvement, such as cancer in the nasopharynx, oral cavity, or oropharynx, the larynx is also outlined as a dose-limited organ. In some plans, unwanted hot spots appeared in the posterior cervical musculature. A reference structure (pseudovolume) was therefore added to limit the dose, and a new plan was generated. Planning organs-at-risk volumes were created during treatment planning by expanding the spinal cord and brainstem 5 mm and optic nerves 1 mm circumferentially. Dose prescription The CTV1, CTV2, and CTV3 are expanded 5–8 mm in each direction to create planning target volumes (PTVs). For definitive IMRT, the dose objectives for PTV1, PTV2, and PTV3 were 70–74 Gy (most patients received 70 Gy), 60 Gy, and 50–54 Gy, respectively. For high-risk postoperative IMRT (those with extracapsular extension, positive or close margins, or soft tissue involvement of tumor), the dose objectives for PTV1, PTV2, and PTV3 were 64–66 Gy, 60 Gy, and 50–54 Gy, respectively. For intermediate-risk postoperative IMRT (those without extracapsular extension, positive or close margins, or soft tissue involvement of tumor), the dose objectives for PTV1, PTV2, and PTV3 were 60 Gy, 60 Gy, and 50–54 Gy, respectively. Dose prescription and delivery has evolved in our institute during these years. Initially, the dose was delivered through several field reduction steps with all PTVs receiving the same dose per fraction (2 Gy) during the treatment. At first, all PTVs received 50 Gy, and then 10 Gy was delivered to PTV1 and PTV2. A subsequent boost to PTV1 only was used to deliver the final total dose. More recently, it was reported by Mohan et al. (18) that delivery of all doses to all targets in a single IMRT plan throughout the treatment course offered better dose conformality. Therefore, several of our patients were treated with a single IMRT plan. However, because of the uncertainties of the modified fraction dose, either with a higher fraction dose to the PTV1, as in simultaneous modulated accelerated radiation therapy (SMART) (19) and in simultaneous integrated boost (SIB) (20), or with lower fraction dose to the prophylactic lymphatic regions (13), and our preference to use standard fraction dose (1.8 Gy or 2.0 Gy), beginning in October 2002 we developed a unique dose prescription scheme. At first, a simultaneous integrated boost plan is given, with PTV1 and PTV2 treated to 60 Gy at 2 Gy per fraction and PTV3 treated to 54 Gy at 1.8 Gy per fraction in one plan. For definitive IMRT, the PTV1 is then treated as a boost for an additional 10–14 Gy at 2 Gy per fraction. For high-risk postoperative IMRT (those with extracapsular extension, positive or close margins, or soft tissue involvement of tumor), the PTV1 is then treated as a boost with an additional 4–6 Gy at 2 Gy per fraction. For intermediate-risk postoperative IMRT (those without extracapsular extension, positive or close margins, or soft tissue involvement of tumor), no further boost treatment is given. With this dose prescription scheme, the majority of the dose and the treatment are delivered by the SIB technique, followed by only a short course of sequential boost (SEB). Standard fraction doses are used, and the treatment course is finished in the period established by conventional radiation. The dose prescription is summarized in Table 3. The planning dose objectives are based on the PTVs; however, the final dose prescriptions are based on coverage of the CTVs by evaluation of dose–volume histograms (DVHs) of the CTVs. In general, we try to make sure that at least 95% of the CTVs received the prescribed dose. Recently, we use the guidelines in Radiation Therapy Oncology Group (RTOG) protocol RTOG H-0022 to evaluate our plans. In summary, no more than 20% of CTVs (rather than PTVs, as required in the protocol) receive more than 110% of their prescribed doses, and no more than 1% of any CTVs receive less than 93% of their prescribed doses. Table 3 Dose specification to target volume in SIB-SEB IMRT for head-and-neck cancer Definitive High risk postoperative SIB CTV1/CTV2 CTV3 SEB CTV1 60 (2) 54 (1.8) 10–14 (2) 60 (2) 54 (1.8) 4–6 (2) Intermediate risk postoperative 60 (2) 54 (1.8) No Abbreviations: CTV = clinical target volume; SEB = sequential boost; SIB = simultaneous integrated boost. Data are presented as total (daily) dose in Gy. The dose constraints to the normal tissues are summarized in Table 4. Maximal planning dose constraints represent the maximal doses that are put in the prescription during the initial planning process. If the resulting plan is not optimal in dose coverage, these parameters can be modified accordingly, depending on where the plan needs to be improved. A new plan is then generated and evaluated. The actual doses from the SIB plan and from the SEB plan are added and compared with the maximal final allowed doses. The maximal final allowed doses are summarized from the literature; most are well established from conventional radiation. These represent the maximal doses allowed to any given portion of the organ, except for parotids and larynx. For parotids and larynx (particularly for parotids), the DVH is also analyzed during evaluation of the plan. For critical structures, such as spinal cord, brainstem, and optic nerves, PRVs were used for dose constrains. Because the maximal planning doses we input in the prescription are much lower than the maximal final allowed doses, the sum of actual doses from the SIB plan and from the SEB plan is generally within the limits of the maximal allowed doses. Table 4 Dose constraints to normal tissues Maximal planning dose constraints (Gy) SIB Spinal cord Brain stem Optic nerves/chiasm Retina Temporal lobes Glottic larynx* Mandible Parotid 30 35 15 15 35 35 35 20 2 2 2 2 4 3 6 2 SEB 45 54 54 50 60 2/3 below 50 Gy 70 Mean dose of <30 Gy or 50% of either parotid <30 Gy Tissue Maximal final allowed dose (Gy) Abbreviations: SEB = sequential boost; SIB = simultaneous integrated boost. For tumors not involving the larynx and hypopharynx. Examples of target delineation and dose prescription are illustrated in Fig. 1 and Fig. 2. Figure 1 is a patient with stage T3N1 supraglottic cancer. The gross tumor and the involved lymph nodes were first outlined as GTV, and CTV1 (red) was then created by adding 5 mm margins to the GTV. Then, the entire larynx, the lymphatic region around the involved lymph nodes, and ipsilateral Levels II and III were defined as CTV2 (purple). For more advanced tumors, or if the primary tumor crosses the midline, the contralateral Level II lymph nodes would also be included as CTV2. Finally, the remaining cervical lymphatic regions and the supraclavicular areas were defined as CTV3 (yellow). Figure 1a represents the initial treatment plan, and Fig. 1b is the sequential boost for the CTV1. Figure 2 is a postoperative IMRT plan for a patient with Stage T3N2c laryngeal cancer status after total laryngectomy and bilateral neck dissection. There was extracapsular extension of the left cervical lymph nodes, which was outlined as CTV1 (red). Because the surgical margins were negative and there was no extracapsular extension in the right cervical lymph nodes, the primary tumor bed and the right neck were defined as CTV2 (purple). Note that the stoma was included in CTV2. The lower neck and the supraclavicular areas were included as CTV3 (yellow). Figure 2 represents the initial treatment plan, which was followed by a boost of 6 Gy to the CTV1. Fig. 1. An example of target volume delineation and dose prescription for a definitive case, a patient with T3N1 supraglottic carcinoma. CTV1 is shown in red, CTV2 in purple, and CTV3 in yellow. (A) Initial SIB plan. (B) Sequential boost to the CTV1 (with 5-mm expansion). The final total doses to the CTV1, CTV2, and CTV3 were 70 Gy, 60 Gy, and 54 Gy, respectively. CTV = clinical target volume. Fig. 2. An example of target delineation and dose prescription for a postoperative case, a patient with Stage T3N2c laryngeal cancer status after total laryngectomy and bilateral neck dissection, with extracapsular extension in the left neck. CTV1 is shown in red, CTV2 in purple, and CTV3 in yellow. The stoma is included in CTV2. This represents the initial SIB plan. An additional 6 Gy was given to the CTV1 in a sequential boost because of extracapsular extension. The final total doses to the CTV1, CTV2, and CTV3 were 66 Gy, 60 Gy, and 54 Gy, respectively. CTV = clinical target volume. Follow-up Patients were seen in follow-up 1 month after their completion of radiation and every 6–8 weeks thereafter in the first year and every 2–3 months in the second year. Physical examination, including fiberoptic laryngoscopy, was performed at each follow-up. When possible, FDG-PET and CT images were obtained 3–4 months after treatment to assess treatment response and 1 year after treatment for surveillance. For patients with suspicious findings on physical examination, CT imaging, and/or abnormal FDG-PET results at the primary site, panendoscopy and directed biopsies were performed under general anesthesia. If biopsies were positive, salvage surgery was performed in suitable surgical candidates. If biopsies were negative, close follow-up with physical examination and repeated imaging studies was recommended. Analysis of failure Failure patterns were determined by comparison of imaging studies, including CT and FDG-PET at the time of detection of persistent/recurrent disease to the treatment plan. In some cases, the CT at the time of detection of recurrence/persistence was coregistered with the planning CT, and the relationship of the recurrent/persistent tumor with the isodose distribution could then be directly visualized. Figure 3 illustrates one such example. This is a patient with Stage T3N0 laryngeal cancer who received definitive IMRT to 70 Gy without chemotherapy. Figure 3a represents the treatment plan (the SIB phase). An FDG-PET image was obtained 4.5 months after completion of treatment and showed an increased FDG uptake focus in the larynx (Fig. 3b). This corresponded to the soft tissue thickening in the larynx in the CT obtained on the same day (Fig. 3c). Initial biopsy of this area revealed carcinoma in situ. A repeat biopsy 1.5 months later showed invasive squamous cell carcinoma. Figure 3d is obtained by coregistration of Fig. 3c to Fig. 3a within the Nomos planning system and reveals that the recurrent tumor is inside the CTV1. Fig. 3. Analysis of failure pattern. A patient with Stage T3N0 laryngeal cancer received definitive intensity-modulated radiotherapy volume to 70 Gy with no concurrent chemotherapy. (a) Treatment plan (the SIB phase). (b) [18F] fluorodeoxyglucose position emission tomography (FDG-PET) was obtained 4.5 months after treatment completion, showing a focus of increased FDG uptake. (c) A computed tomography obtained on the same day showed soft tissue thickening in the larynx. (d) Coregistration of a and c, showing that the recurrent tumor is within the CTV1. CTV = clinical tartet radiotherapy Statistical analysis The overall survival, local recurrence-free survival, local–regional recurrence-free survival, and distant disease-free survival were calculated by the Kaplan-Meier product-limit method. Salvage of recurrences was not taken into account in the evaluation of local recurrence-free survival and local–regional recurrence-free survival. The differences in local– regional recurrence-free survival among groups were assessed by the log–rank statistic. Values of p < 0.05 were considered as significant. All analyses were performed with S-PLUS 2000 software (Mathsoft, Seattle, WA). Results Treatment outcomes For all patients, the median follow-up was 18 months (range, 2.0–60 months). All living patients were followed for at least 6 months. There were 11 local–regional failures: 7 local failures, 3 regional failures, and 1 failure at both the primary tumor and regional lymph node. There were 16 patients who failed distantly, either with distant metastasis or new lung primary tumors (3 patients). The 2-year overall survival, local recurrence-free survival, local–regional recurrence-free survival, and distant disease-free survival rates were 85%, 94%, 92%, and 87%, respectively. The 3year overall survival, local recurrence-free survival, local–regional recurrence-free survival, and distant disease-free survival rates were 82%, 94%, 92%, and 81%, respectively (Table 5, Fig. 4, Fig. 5, Fig. 6 and Fig. 7). Nine patients who had definitive IMRT had neck dissection after radiation because of persistent lymph nodes. Four of them had residual cancer in the lymph nodes. Because neck dissection was part of their initial treatment, these 4 patients are not tallied as treatment failures. Table 5 Overall performance of head-and-neck patients treated with IMRT 2-year survival (%) Overall survival Local recurrence-free survival Local–regional recurrence-free survival Distant disease-free survival 85 94 92 87 3-year survival (%) 82 94 92 81 Abbreviations: IMRT = intensity-modulated radiotherapy. Fig. 4. Kaplan-Meier estimates of overall survival of all patients. Fig. 5. Kaplan-Meier estimates of local recurrence-free survival of all patients. Fig. 6. Kaplan-Meier estimates of local–regional recurrence-free survival of all patients. Fig. 7. Kaplan-Meier estimates of distant disease-free survival of all patients. Characteristics of local–regional failures The median time from treatment completion to local–regional recurrence was 4.7 months (range, 1.8–15.6 months). The characteristics of local and regional failures are summarized in Table 6. Five patients failed after postoperative IMRT, and 6 failed after definitive IMRT. Table 6 Clinical characteristics and patterns of locoregional failures Time to recurrence (mo) 5.13 15.63 5.23 4.73 5.33 6.37 4.1 Patien t No. 1 2 3 4 5 6 7 Tumor site R Llp Supraglottis R oral tongue L oral tongue Larynx Larynx L oral tongue T T2 T4 T3 T3 T2 T3 T4 N N1 N2C N2B N2B N0 N0 N2C Aim Postoperative Definitive Postoperative Postoperative Definitive Definitive Postoperative Chemotherap y No Yes No No No No No Failure site R Level II Supraglottis L Level IB Oral tongue Larynx Larynx Tongue BOT Location CTV1 CTV1 CTV2 CTV1 CTV1 CTV1 CTV1 CTV2 Out of 8 L oral tongue T1 N1 Postoperative No R Level III 3.0 Patien t No. Tumor site T N Aim Chemotherap y Failure site Location field Time to recurrence (mo) 9 R buccal mucosa T4 N2B Definitive Induction R buccal mucosa L Level IB CTV1 CTV2 CTV1 CTV1 3.13 10 11 L tonsil Larynx T4 T4 N0 N2B Definitive Definitive Yes Yes L tonsil Larynx 1.83 4.0 Abbreviations: CTV = clinical target volume; L = left; R = right. Patient 4 finished his radiation course over 106 days, and the interval between his surgery and completion of radiation was 180 days, owing to his initial refusal to have radiation and long break in the radiation course for personal reasons. Patient 8 had a Stage T1N1 oral tongue cancer after surgery. Because of the small lateral primary tumor and small metastatic lymph node, per treating physician, only the oral tongue, bilateral Levels I and II, and the ipsilateral neck were included in the radiation. She failed in the contralateral lower neck that was not irradiated. Patient 9, with a Stage T4N2B buccal mucosal cancer, was recommended to have surgery followed by postoperative radiation; however, she refused surgery. Patient 11 with laryngeal cancer with bulky tumor and thyroid cartilage erosion also refused the recommended surgery with postoperative radiation. Both were treated with definitive IMRT and failed in the primary tumor as well as regional lymph nodes. Patient 10 had an extensive tonsil cancer extending into posterior orbit and cavernous sinus. He failed both in the primary tumor and in the base of skull, with recurrent tumor extending into the brain and orbit. This is the only marginal failure in our experience. For Patient 7, the majority of recurrent tumor was within CTV1, which received 66 Gy, but it extended into CTV2, which received 60 Gy. Interestingly, 3 patients had failures in the contralateral lymph nodes: Patients 3 and 9 failed in contralateral Level IB, which received 60 Gy, and Patient 8 failed in contralateral Level III, which was not irradiated. Three patients with local recurrences could not be salvaged and died from local disease. Two refused salvage surgery; 1 of them died from local disease, and 1 was alive with local disease. Six patients underwent salvage surgeries. Two of them were alive and well at last follow-up. One developed local recurrence and died 1 year after the salvage surgery. Three died from distant diseases (1 distant metastasis, 1 new lung primary tumor, and 1 indeterminate). Prognostic factors Univariate analysis, taking gender, site, T stage, N stage, type of IMRT (postoperative vs. definitive), and chemotherapy use as prognostic factors, was performed for local–regional control. Patients with oropharyngeal cancer did significantly better than patients with oral cavity and laryngeal cancer, with a 2-year local–regional control rate of 98%, compared with 78% for oral cavity cancer and 85% for laryngeal cancer (p = 0.005; Fig. 8). Most oropharyngeal cancer patients were treated with definitive IMRT with concurrent chemotherapy, whereas most oral cavity cancer patients were treated with postoperative IMRT without chemotherapy. There was no significant difference in local– regional control for patients who received postoperative radiation or definitive radiation (Fig. 9, p = 0.339) and for patients who had chemotherapy or not (p = 0.402). Neither T stage nor N stage had a significant effect on local– regional control (p = 0.722 and 0.712, respectively) in this series. Fig. 8. Kaplan-Meier estimates of local–regional recurrence-free survival according to primary tumor site. Fig. 9. Kaplan-Meier estimates of local–regional recurrence-free survival according to the type of treatment (definitive vs. postoperative). Discussion Intensity-modulated radiation therapy is a major advance in radiation treatment. It is a highly conformal radiation technique that permits delivery of high-dose radiation to target tissues while sparing adjacent critical structures. Many published studies have shown the advantages of IMRT in dosimetry compared with conventional techniques. However, only a few studies reported treatment outcome of IMRT in head-and-neck cancer. Lee et al. (21) reported the University of California, San Francisco, experience with IMRT in nasopharyngeal cancer, showing that the 4-year estimates of local progression-free, local–regional progression-free, and distant metastasis-free rates were 97%, 98%, and 66%, respectively, and the 4-year estimate of overall survival was 88%. Chao et al. (1) reported the Washington University experience in 126 head-and-neck cancer patients treated with IMRT, showing a 2-year actuarial locoregional control rate of 85%. Recently, Chao et al. (22) reported outcomes of IMRT in oropharyngeal cancer, showing a 4-year estimate of locoregional control of 87%. Eisbruch et al. (23) also reported excellent treatment outcomes with IMRT in head-and-neck cancer, showing 3-year actuarial locoregional recurrence-free survival rates for definitive IMRT and postoperative IMRT of 81% and 84%, respectively. We show here that, for 150 patients treated with IMRT at our institution, the 2-year overall survival, local recurrence-free survival, local–regional recurrence-free survival, and distant disease-free survival rates were 85%, 94%, 92%, and 87%, respectively. Our results further confirm the effectiveness of IMRT in head-and-neck cancer. It is noted that there is a significant variation in target delineation, dose prescription, and dose delivery in IMRT for head-and-neck cancer. There has been great effort in developing consensus and guidelines in target delineation (11). In terms of dose prescription and dose delivery, there are two general approaches. A course of IMRT can be given in multiple steps with consecutive plans using conventional fractionation. Alternatively, one single plan throughout the treatment course can be used with different dose rates to different target volumes. Mohan et al. (18), using a phantom, illustrated that IMRT given by single plan (SIB for simultaneous integrated boost) offered higher dose conformality than IMRT given by multiple steps. They found that, although both techniques provided the requisite coverage for the gross disease and equivalent sparing of the spinal cord and parotid glands, the dose to normal tissue outside the tumor volume was lower for the SIB plan than for the sequential plan. In their study, the sequential IMRT consisted of an initial 50 Gy in 25 fractions to all the targets, followed by 20 Gy in 10 fractions to the GTV for a boost. To include all targets into a single treatment plan, fraction doses to different targets need to be adjusted, either giving conventional fraction dose to the gross tumor and lower fraction dose to the microscopic areas or giving conventional fraction dose to the microscopic areas and higher fraction dose to the gross tumor. The former technique might prolong the treatment course to the microscopic areas, and the latter might potentially result in more significant late side effects due to high dose per fraction, especially when given with concurrent chemotherapy. Because of these uncertainties, we developed a modified dose delivery scheme in which the majority of the treatment is given in an SIB plan followed by only a short course of sequential boost. We use a conventional fraction dose (i.e., 1.8 Gy and 2 Gy) throughout the treatment course, and the treatment course is finished in the time period established by conventional radiation. A recent abstract by Crimaldi et al. (24) proposed a modified SIB planning, SIB-II, which is similar to our approach. Using this approach, they replanned 10 patients treated with SIB-IMRT. They found that SIB-II dose distributions for GTV and CTV were comparable to those of the SIB plan, but the total doses and the biologically equivalent doses to the parotids were lower in SIB-II than those in SIB, which translated into reductions of normal tissue complication probability of 26.1% and 40.0% for total and distant parotid glands, respectively. Further studies in long-term outcome, including tumor control and quality of life, are needed to determine the optimal dose prescription and delivery in head-and-neck IMRT. There were 11 local regional recurrences found in our cohort of patients. The median time from treatment completion to recurrences was 4.7 months (range, 1.8–15.6 months). This is shorter than that reported by the University of Michigan group (8–9 months [23 and 25]). This might be because of our follow-up schedule, in that an FDG-PET is routinely obtained 3–5 months after completion of treatment. Most of these local–regional recurrences were detected by the posttherapy FDG-PET (26). Consistent with previous studies (1 and 25), most of our local–regional failures were in-field failures. Only one marginal failure occurred in a patient with tumor extending into orbit and cavernous sinus. Our results indicate that our approach in target delineation, and coverage is adequate for head-and-neck cancer patients receiving IMRT. It should be emphasized that the multidisciplinary approach in target delineation is critical. All our patients were presented in the multidisciplinary head-and-neck tumor board. All radiologic images were reviewed with radiologists and surgical and pathologic findings with surgeons and pathologists before target delineation. Most of our patients had PET and recently PET/CT for staging workup. This provides valuable information in target delineation. We are planning to use PET/CT for simulation, target delineation, and treatment planning for all head-and-neck IMRT. We show that patients with oropharyngeal cancer did significantly better than patients with oral cavity and laryngeal cancer, with a 2-year local–regional control rate of 98%, compared with 78% for oral cavity cancer and 85% for laryngeal cancer. This is consistent with the report by Eisbruch et al. (23). They showed that the 3-year local control rate for oropharyngeal cancer was 94%, compared with 59% and 60% for oral cavity cancer and laryngeal cancer, respectively. An excellent outcome for oropharyngeal cancer treated with IMRT was also reported by Chao et al. (22) and Huang et al. (27). Eisbruch et al. (23) showed no difference in postoperative IMRT vs. definitive IMRT for oropharyngeal cancer. However, Chao et al. (22) showed that postoperative IMRT was superior to definitive IMRT, with 4-year local–regional control rates of 77% in definitive IMRT and 95% in postoperative IMRT. Most of our patients with oropharyngeal cancer were treated by definitive IMRT with concurrent chemotherapy, as were the patients reported by Huang et al. (27). Both showed excellent local–regional control. Longer follow-up is necessary to determine the optimal treatment modality for oropharyngeal cancer, and quality-of-life issues should also be taken into consideration. Interestingly, we noticed a high incidence of contralateral lymph node recurrences in oral cavity cancer. Three patients had failures in the contralateral lymph nodes: 2 failed in contralateral Level IB, which received 60 Gy, and 1 failed in contralateral Level III, which was not irradiated. Chow et al. (28) reported 8 contralateral lymph node recurrences in 72 oral cavity cancer patients after treatment and found that positive ipsilateral lymph node was a significant predictive factor. Kurita et al. (29) reported 19 of 129 oral cavity cancer patients with contralateral lymph node metastasis. They found that, although T stage, number of ipsilateral lymph node involvement, and histopathologic grading were independent and significant predictors, contralateral lymph node metastasis never occurred in patients without ipsilateral lymph node metastasis. For oral cavity cancer with ipsilateral lymph node involvement, it is therefore critical to include bilateral neck in the radiation field. For locally advanced oral cavity cancer, it might be necessary to include bilateral Level I nodes in the CTV1. As mentioned above, oral cavity cancer patients did much poorer in local–regional control. Most of these patients were treated by surgery and postoperative IMRT without chemotherapy. Even though in this heterogeneous group of patients, we could not find a benefit of chemotherapy in local–regional control, in light of the recent studies (30 and 31) showing benefit of postoperative chemoradiation over postoperative radiation alone in local–regionally advanced head-and-neck cancer, it might be reasonable to consider postoperative IMRT with concurrent chemotherapy for local–regionally advanced oral cavity cancer. In conclusion, our results have confirmed the effectiveness of IMRT in head-and-neck cancer. It offers excellent outcomes in local–regional control and overall survival. Further studies are necessary to improve the outcomes for laryngeal cancer and oral cavity cancer. Acknowledgements We thank Ms. Kellie Bodeker for her editorial assistance and manuscript preparation. References 1 K.S. Chao, G. Ozyigit and B.N. Tran et al., Patterns of failure in patients receiving definitive and postoperative IMRT for head-and-neck cancer, Int J Radiat Oncol Biol Phys 55 (2003), pp. 312–321. | 2 N. Lee, P. Xia and N.J. Fischbein et al., Intensity-modulated radiation therapy for head-and-neck cancer The UCSF experience focusing on target volume delineation, Int J Radiat Oncol Biol Phys 57 (2003), pp. 49–60. 3 A. Eisbruch, H.M. Kim and J.E. Terrell et al., Xerostomia and its predictors following parotid-sparing irradiation of head-and-neck cancer, Int J Radiat Oncol Biol Phys 50 (2001), pp. 695–704. | 4 K.S. Chao, J.O. Deasy and J. Markman et al., A prospective study of salivary function sparing in patients with headand-neck cancers receiving intensity-modulated or three-dimensional radiation therapy Initial results, Int J Radiat Oncol Biol Phys 49 (2001), pp. 907–916. 5 A. Lin, H.M. Kim and J.E. Terrell et al., Quality of life after parotid-sparing IMRT for head-and-neck cancer A prospective longitudinal study, Int J Radiat Oncol Biol Phys 57 (2003), pp. 61–70. 6 M.B. Parliament, R.A. Scrimger and S.G. Anderson et al., Preservation of oral health-related quality of life and salivary flow rates after inverse-planned intensity–modulated radiotherapy (IMRT) for head-and-neck cancer, Int J Radiat Oncol Biol Phys 58 (2004), pp. 663–673. 7 P.M. Som, H.D. Curtin and A.A. Mancuso, An imaging-based classification for the cervical nodes designed as an adjunct to recent clinically based nodal classifications, Arch Otolaryngol Head Neck Surg 125 (1999), pp. 388–396. 8 P.J. Nowak, O.B. Wijers and F.J. Lagerwaard et al., A three-dimensional CT-based target definition for elective irradiation of the neck, Int J Radiat Oncol Biol Phys 45 (1999), pp. 33–39. 9 O.B. Wijers, P.C. Levendag and T. Tan et al., A simplified CT-based definition of the lymph node levels in the node negative neck, Radiother Oncol 52 (1999), pp. 35–42. 10 V. Gregoire, E. Coche and G. Cosnard et al., Selection and delineation of lymph node target volumes in head and neck conformal radiotherapy. Proposal for standardizing terminology and procedure based on the surgical experience, Radiother Oncol 56 (2000), pp. 135–150. 11 V. Gregoire, P. Levendag and K.K. Ang et al., CT-based delineation of lymph node levels and related CTVs in the node-negative neck DAHANCA, EORTC, GORTEC, NCIC, RTOG consensus guidelines, Radiother Oncol 69 (2003), pp. 227–236. 12 A. Eisbruch, R.L. Foote and B. O’Sullivan et al., Intensity-modulated radiation therapy for head and neck cancer Emphasis on the selection and delineation of the targets, Semin Radiat Oncol 12 (2002), pp. 238–249. 13 K.S. Chao, F.J. Wippold and G. Ozyigit et al., Determination and delineation of nodal target volumes for head-andneck cancer based on patterns of failure in patients receiving definitive and postoperative IMRT, Int J Radiat Oncol Biol Phys 53 (2002), pp. 1174–1184. 14 R. Lindberg, Distribution of cervical lymph node metastases from squamous cell carcinoma of the upper respiratory and digestive tracts, Cancer 29 (1972), pp. 1446–1449. 15 R.M. Byers, P.F. Wolf and A.J. Ballantyne, Rationale for elective modified neck dissection, Head Neck Surg 10 (1988), pp. 160–167. 16 J.P. Shah, Patterns of cervical lymph node metastasis from squamous carcinomas of the upper aerodigestive tract, Am J Surg 160 (1990), pp. 405–409. 17 S.K. Mukherji, D. Armao and V.M. Joshi, Cervical nodal metastases in squamous cell carcinoma of the head and neck What to expect, Head Neck 23 (2001), pp. 995–1005. 18 R. Mohan, Q. Wu and M. Manning et al., Radiobiological considerations in the design of fractionation strategies for intensity-modulated radiation therapy of head and neck cancers, Int J Radiat Oncol Biol Phys 46 (2000), pp. 619–630. 19 E.B. Butler, B.S. Teh and W.H. Grant 3rd et al., SMART (simultaneous modulated accelerated radiation therapy) boost A new accelerated fractionation schedule for the treatment of head- and neck-cancer with intensity-modulated radiotherapy, Int J Radiat Oncol Biol Phys 45 (1999), pp. 21–32. 20 Q. Wu, R. Mohan and M. Morris et al., Simultaneous integrated boost intensity-modulated radiotherapy for locally advanced head-and-neck squamous cell carcinomas. I: Dosimetric results, Int J Radiat Oncol Biol Phys 56 (2003), pp. 573–585. 21 N. Lee, P. Xia and J.M. Quivey et al., Intensity-modulated radiotherapy in the treatment of nasopharyngeal carcinoma An update of the UCSF experience, Int J Radiat Oncol Biol Phys 53 (2002), pp. 12–22. | 22 K.S. Chao, G. Ozyigit and A.I. Blanco et al., Intensity-modulated radiation therapy for oropharyngeal carcinoma Impact of the tumor volume, Int J Radiat Oncol Biol Phys 59 (2004), pp. 43–50. 23 A. Eisbruch, L.H. Marsh and L.A. Dawson et al., Recurrences near base of skull after IMRT for head-and-neck cancer Implications for target delineation in high neck and for parotid gland sparing, Int J Radiat Oncol Biol Phys 59 (2004), pp. 28–42. 24 A.J. Crimaldi, Y. Wu and O. Abayomi et al., SIB-II Improved parotid gland sparing using a two-phase planning strategy for head and neck intensity-modulated radiotherapy (IMRT), Int J Radiat Oncol Biol Phys 57 (2003), p. S432. 25 L.A. Dawson, Y. Anzai and L. Marsh et al., Patterns of local-regional recurrence following parotid-sparing conformal and segmental intensity-modulated radiotherapy for head-and-neck cancer, Int J Radiat Oncol Biol Phys 46 (2000), pp. 1117–1126. 26 M. Yao, M.M. Graham and R.B. Smith et al., Value of FDG PET in assessment of treatment response and surveillance in head- and-neck cancer patients after intensity-modulated radiation treatment—A preliminary report, Int J Radiat Oncol Biol Phys 60 (2004), pp. 1410–1418. 27 K. Huang, N. Lee and P. Xia et al., Intensity-modulated radiotherapy in the treatment of oropharyngeal carcinoma A single institutional experience, Int J Radiat Oncol Biol Phys 57 (2003), p. S302. 28 T.L. Chow, T.K. Chow and T. Chan et al., Contralateral neck recurrence of squamous cell carcinoma of oral cavity and oropharynx, J Oral Maxillofac Surg 62 (2004), pp. 1225–1228. 29 H. Kurita, T. Koike and J. Narikawa et al., Clinical predictors for contralateral neck lymph node metastasis from unilateral squamous cell carcinoma in the oral cavity, Oral Oncol 40 (2004), pp. 898–903. 30 J.S. Cooper, T.F. Pajak and A.A. Forastiere et al., Postoperative concurrent radiotherapy and chemotherapy for high-risk squamous-cell carcinoma of the head and neck, N Engl J Med 350 (2004), pp. 1937–1944. 31 J. Bernier, C. Domenge and M. Ozsahin et al., Postoperative irradiation with or without concomitant chemotherapy for locally advanced head and neck cancer, N Engl J Med 350 (2004), pp. 1945–1952. Reprint requests to: Min Yao, M.D., Ph.D., University of Iowa Health Care, Department of Radiation Oncology, W189Z GH, 200 Hawkins Dr., Iowa City, IA 52242. Tel: (319) 356-7603; Fax: (319) 356-1530