Anuradha Thiagarajan, MD, 1 James Mechalakos, PhD, 2 Nancy Lee, MD 1

Transcription

1 CASE REPORT Eben L. Rosenthal, MD, Section Editor FEASIBILITY OF REIRRADIATION OF RECURRENT SINONASAL CARCINOMA IN THE PERIORBITAL REGION USING HYPOFRACTIONATED IMAGE-GUIDED INTENSITY-MODULATED RADIATION THERAPY Anuradha Thiagarajan, MD, 1 James Mechalakos, PhD, 2 Nancy Lee, MD 1 1 Department of Radiation Oncology, Memorial Sloan-Kettering Cancer Center, New York, New York. thiagara@mskcc.org 2 Department of Medical Physics, Memorial Sloan-Kettering Cancer Center, New York, New York Accepted 31 December 2009 Published online 19 May 2010 in Wiley Online Library (wileyonlinelibrary.com). DOI: /hed Abstract: Background. Recurrent periorbital tumors pose management challenges because they are often unresectable and chemorefractory. Proximity to critical structures renders reirradiation difficult. With image-guided intensity-modulated radiation therapy (IG-IMRT), real-time corrections to patient setup are possible, enabling significant shrinkage of planning target volume margins and safe dose escalation with hypofractionation. Here, we present our experience with hypofractionated IG-IMRT in a patient with multiply recurrent sinonasal carcinomas in the periorbital region. Methods. The patient is a 67-year-old woman with sinonasal carcinoma, treated with surgery and 2 prior courses of high-dose radiotherapy. She developed a second tumor recurrence in the right lateral orbit, which was reirradiated using hypofractionated IG-IMRT (24 Gy in 3 fractions). Results. Near-complete resolution of her ocular symptoms was observed at 4 months. Progress scans demonstrated a decrease in size of the orbital lesion. Conclusion. Preliminary results of reirradiation using hypofractionated IG-IMRT suggest that it is safe and effective. VC 2010 Wiley Periodicals, Inc. Head Neck 33: , 2011 Keywords: image-guided; head and neck cancers; hypofractionation; reirradiation Correspondence to: A. Thiagarajan VC 2010 Wiley Periodicals, Inc. Recurrent periorbital tumors, both primary and metastatic, pose unique challenges to the treating oncologist. These lesions are usually symptomatic, producing pain, proptosis, and progressive visual loss. Despite advances in multimodal therapy, therapeutic options in this setting are typically limited. Surgical salvage, if at all feasible, would almost certainly entail an orbital exenteration with significant functional, cosmetic, and psychological sequelae. Systemic therapy often results in less than satisfactory outcomes with response rates less than 20% to 30%. 1,2 In addition, any response to chemotherapy, even with multiagent regimens, is usually short-lived, with the majority of patients eventually dying from uncontrolled local disease. Radiation therapy is an option, but traditionally radiation oncologists have been reluctant to consider reirradiation of recurrent tumors particularly when critical structures are in close proximity and have already been treated to tolerance. This is even more of a concern if the time to recurrence interval is short, as is often the case. With the advent of image-guided intensity-modulated radiation therapy (IG-IMRT), it has now become possible to deliver very precise radiotherapy to the designated target volume. By obtaining cone-beam CT and/or 2-dimensional kilovoltage (2D kv) images at the time of treatment, appropriate real-time corrections to patient positioning and setup can be made. Elimination of these types of random and systematic treatment errors permits significant reduction in the normal tissue margin around the target volume and consequently, decreased toxicity to surrounding critical structures. 3 6 At present, there is much focus on the use of IG- IMRT in the management of spinal and paraspinal tumors. However, the scope for application of this new technology is clearly enormous. In this case report, we present our experience with IG-IMRT in the management of a patient with multiply recurrent sinonasal carcinoma in the periorbital region. MATERIALS AND METHODS This study was approved by the Institutional Review Board at Memorial Sloan-Kettering Cancer Center (MSKCC) and granted a waiver from informed consent. The patient is a 67-year-old woman who was initially diagnosed with locally advanced high-grade epithelial carcinoma of the left nasal septum in January 1372 Reirradiation of Recurrent Sinonasal Carcinoma Using IG-IMRT HEAD & NECK DOI /hed September 2011

2 2007 when she presented to her local otolaryngologist for evaluation of progressive nasal congestion of 12 months duration. CT and MRI scans of the head and neck demonstrated a 2.6 cm calcified mass arising from the nasal septum and filling the entire left nasal cavity. There was no evidence of orbital or intracranial involvement and no regional lymphadenopathy was noted. She was subsequently referred to MSKCC where she underwent radical resection of the tumor with a left medial maxillectomy and resection of the nasal septum utilizing a mid-line degloving approach on February 2, Histopathologic examination confirmed a high-grade malignant epithelial neoplasm favoring basaloid squamous cell carcinoma of the nasal septum extending into the left medial maxilla. Lymphovascular invasion was present with a vascular tumor embolus noted at the posterior bone margin. All other margins were free of tumor. Surgery was followed by adjuvant radiation therapy, which was performed at an outside institution between March 26, 2007 and June 14, Radiation was delivered using IMRT to a total dose of 66.6 Gy in 37 fractions. Treatment was complicated by multiple treatment breaks as a result of severe radiation dermatitis and significant weight loss requiring hospital admission for intravenous hydration and temporary gastrostomy placement. In January 2008, the patient returned for review at MSKCC complaining of a several month history of right-sided epiphora. Repeat CT scan of the paranasal sinuses showed new mixed lytic/sclerotic lesions in the right frontal bone and the lateral aspect of the right maxillary sinus where there was also soft tissue extension into the inferolateral orbit. These lesions demonstrated corresponding hypermetabolic activity on a positron emission tomography (PET) scan performed shortly thereafter. A subsequent biopsy of the maxillary sinus lesion showed malignant cells consistent with a poorly differentiated epithelial carcinoma, providing histologic confirmation of recurrence. This case was discussed at the multidisciplinary tumor board and the relevant images were reviewed. Due to the extensive and noncontiguous nature of the recurrences, the lesions were deemed to be unresectable, and hence the patient proceeded to receive IMRT to a total dose of 60 Gy in 30 fractions from April 15 to May 27, This was given with concurrent carboplatin and taxol. In accordance to the practice at our institution for head and neck reirradiation, 2D kv imaging was used daily before treatment to ensure accurate patient setup. The patient tolerated treatment well with minimal acute side effects (G1 xerostomia, dysphagia, and dermatitis). In January 2009, she was found to have a palpable osseous mass on the superior aspect of the manubrium during a routine clinic visit. This lesion demonstrated increased fluorodeoxyglucose (FDG) uptake on PET scan and a subsequent CT-guided biopsy confirmed metastatic involvement by poorly differentiated sinonasal carcinoma. She was then commenced on single-agent methotrexate which she tolerated poorly secondary to significant mucositis. In April 2009, she re-presented with severe right ocular pain, proptosis, and decreased vision. She also reported constant diplopia more pronounced on right lateral gaze. This was on a background of recurrent episodes of periorbital cellulitis that was slow to resolve despite multiple courses of oral antibiotics. Progress CT scan of the orbits and paranasal sinuses showed irregular thickening in the right lateral and superior orbital walls with soft tissue extension into the right lacrimal fossa suspicious for tumor recurrence. Treatment options (orbital exenteration vs hypofractionated image-guided radiation therapy) were discussed at great length with the patient. The patient adamantly refused surgical intervention and elected to proceed with reirradiation despite being advised of the significant risk of blindness from radiation-induced damage to optic structures. She subsequently received hypofractionated IG-IMRT to a total dose of 24 Gy in 3 fractions. Treatment Planning. For the purposes of this report, we will solely describe the treatment planning process of her final course of radiation with hypofractionated IG-IMRT. The patient was simulated with the head held in a thermoplastic immobilization device in a supine, neutral position. Two-millimeter CT slices were obtained through the region of interest. The gross tumor volume (GTV) was defined as the gross extent of tumor demonstrated by CT and MRI imaging studies. Critical structures included the patient s eyes, optic nerves, optic chiasm, and brainstem. In view of her significant previous radiation history and the need to minimize field overlap, the GTV was not expanded further to obtain a clinical target volume (CTV). Instead, a direct 3-dimensional expansion of 3 mm was made from GTV to planning target volume (PTV). The patient was planned using the MSKCC treatment planning system with 6 MV photons. The goal of treatment planning was to get as good a coverage of PTV as possible with relative sparing of the optic chiasm and contralateral optic structures. Of note, a 2-mm margin was applied to these critical structures to aid dose optimization. In addition, target volumes and isodose curves from prior courses of radiation were transferred to the planning CT scan with CT registration to permit accurate calculation of cumulative radiation doses to the sensitive structures in question. Given that this was the patient s second re-treatment and in view of the fact that hypofractionated IG- IMRT was being used, no chemotherapy was given concurrently and in the 2 weeks preceding and following radiation therapy. Reirradiation of Recurrent Sinonasal Carcinoma Using IG-IMRT HEAD & NECK DOI /hed September

3 and the images used to reposition the patient in the x, y, and z directions. On day 1 of treatment, significant rotational errors were noted on the pretreatment cone-beam CT that could not be satisfactorily corrected. Hence, treatment was aborted and the patient was re-simulated. The replanned treatment subsequently proceeded without incident. The patient tolerated treatment well with no acute side effects aside from mild radiation dermatitis. Treatment Outcome. Both clinical and radiographic parameters were used to assess response. At 4 months post completion of radiation, the patient reported near-complete resolution of her ocular symptoms with only minor residual diplopia. Oral antibiotics had been ceased and on physical examination, there was no evidence of proptosis or cellulitis. Progress CT scan confirmed a significant decrease in the size and extent of the previously seen permeative bone changes involving the superolateral orbital wall as well as the extraosseous soft tissue component (Figure 3). FIGURE 1. Isodose curves in the axial, sagittal, and coronal planes taken at the level of the isocenter. Dosimetry. Seven fields were used to treat the recurrent right periorbital disease, with the majority of the beams entering anteriorly. A 5-mm thickness bolus was utilized to adequately treat the superficial portion of the tumor, which extended to within millimeters of the skin surface. Isodose curves in the axial, sagittal, and coronal planes at the level of the isocenter are shown in Figure 1. Dose-volume histograms are shown in Figure 2. The composite doses to the various critical structures were as follows: optic chiasm 6018 cgy, left optic nerve 6045 cgy, brainstem 5600 cgy, and left retina/eye 4485 cgy. To achieve adequate PTV coverage, no dose constraints were placed on the right eye and optic nerve which received cumulative doses of 8000 cgy and 8200 cgy, respectively. However, dosimetrists were instructed to avoid hot spots within these structures. Treatment Course. Orthogonal 2D kv and conebeam CT images were obtained daily before treatment DISCUSSION Radiation oncologists have traditionally avoided reirradiation to the periorbital region due to legitimate concerns about radiation-induced damage to optic structures, a potentially devastating outcome resulting in permanent loss of vision. There are 2 principal technologic advances that have made hypofractionated high-dose reirradiation to these sensitive regions feasible. The first is the advent of IMRT which has the ability to create steep dose gradients between the tumor and neighboring critical structures with a dose fall-off of approximately 10% per millimeter. 3,7 This precise dosimetry is desirable in the management of periorbital tumors, in which at best, only a few millimeters separate disease from the optic apparatus. Retrospective series data from M. D. Anderson Cancer Center and MSKCC have demonstrated encouraging local disease control and survival outcomes from the substitution of conventional radiotherapy techniques with IMRT for re-treatment of recurrent head and neck cancer. 8,9 For patients treated with curative intent, median survival was in excess of 3 years and 4-year overall survival rates approached 50%. The second key element has been the development of sophisticated image-guided systems, which serve to minimize treatment errors associated with patient positioning. With real-time patient setup verification and correction, image guidance improves the accuracy of IMRT delivery, enabling significant reduction of PTV margins. It is important to recognize that even a 2-mm difference in PTV margins results in a 3-dimensional PTV that is considerably smaller. With conventional radiotherapy techniques, the PTV margin added to the GTV and CTV typically ranges from 1 to 1374 Reirradiation of Recurrent Sinonasal Carcinoma Using IG-IMRT HEAD & NECK DOI /hed September 2011

4 FIGURE 2. Dose volume histogram of third course of radiation with hypofractionated image-guided intensity-modulated radiation therapy (24 Gy in 3 fractions). 2 cm. In contrast, a margin of only 2 to 3 mm is used with IG-IMRT. 3 The benefits of this are 2-fold: Firstly, there is reduced normal tissue toxicity. Although reirradiation-related morbidity of IMRT is thought to be less severe than conventional techniques, published reports still quote significant rates of grade 3 and 4 toxicity, ranging from 20% to 30%. 8,9 By comparison, early results of hypofractionated stereotactic radiotherapy using CyberKnife image-guided technology in the re-treatment of recurrent head and neck cancer suggest fewer severe late toxicities at least in part due to shrinkage of PTV margins. 10 Secondly, IG- IMRT allows safe dose escalation with the use of relatively large fraction sizes, thereby overcoming radioresistant clones and improving the probability of tumor control. This is of particular importance in head and neck cancer where studies have shown that the bulk of recurrences occur in-field in high-dose regions. 11 Observations of radioresistance in cultured cells taken from head and neck cancers that recurred after curative radiotherapy lend further credence to the existence of this phenomenon. 12 Without image guidance, it would have been considered unsafe practice to give this patient a second and third course of radiation and administering a meaningful dose would have been impossible. Traditionally, image guidance has relied on megavoltage treatment beams to provide an image of the FIGURE 3. (A) Pretreatment CT scan. (B) CT scan performed approximately 3 months posttreatment. Reirradiation of Recurrent Sinonasal Carcinoma Using IG-IMRT HEAD & NECK DOI /hed September

5 FIGURE 4. (A) Initial cone-beam CT showing rotational error greater than 1 degree. (B) Cone-beam CT showing perfect alignment after resimulation and replanning. patient s bony anatomy. One of the major drawbacks of this approach is the suboptimal definition of bony structures, particularly in patients with a larger body habitus. Recent advances in imaging technology, primarily with the use of kilovoltage imaging, have vastly improved the visualization of both bony and soft tissue structures within the patient. 13 In this regard, 3-dimensional or volumetric images provided by cone-beam CTs are considered to be superior to orthogonal 2D kv images. They have the advantage of not requiring surrogate structures such as bony landmarks to acquire positional information of soft tissue targets. In addition, rotational setup errors are more readily detected when comparing 3-dimensional datasets with each other (ie, cone-beam CT with simulation CT) as was demonstrated in this case. On day Reirradiation of Recurrent Sinonasal Carcinoma Using IG-IMRT HEAD & NECK DOI /hed September 2011

6 FIGURE 5. (A) Lateral 2-dimensional kilovoltage (2D kv) image. (B) Anteroposterior 2-dimensional kilovoltage (2D kv) image. of this patient s radiation course, the rotational error of her setup that was identified on her pretreatment cone-beam CT (Figure 4) was not apparent on the orthogonal 2D kv images that were simultaneously obtained (Figure 5). Hence, in clinical situations involving soft tissue targets and where rotational movements are a concern, periodic cone-beam CTs should be obtained. This gives the treating oncologist greater confidence that radiation is being delivered as intended, mitigating the risk of marginal misses of the target volume or overdosage of surrounding critical structures. Although the aim in this case was to minimize the dose to the optic chiasm and contralateral optic structures, we accepted higher cumulative radiation doses to the optic apparatus than the traditionally estimated TD 5/5, which ranges between 50.4 to 54 Gy. However, it is important to note that the optic structures were treated to near tolerance during her initial course of radiation and contributions from the latter 2 courses were minimal. In fact, despite the hypofractionated radiation schedule, the dose per fraction to the optic chiasm, left optic nerve, and left retina was kept below 2 Gy with these structures only receiving doses of 4.39 Gy, 5.44 Gy, and 5.68 Gy, respectively, from the final course of radiation. On the other hand, due to the proximity of disease, no constraints were placed on the ipsilateral eye and optic nerve, which received cumulative doses in excess of 80 Gy. The treating radiation oncologist accepted potential sacrifice of vision in the right eye to obtain satisfactory PTV coverage and achieve a meaningful interval of local control. Therapeutic options in this scenario are limited and the consequences of uncontrolled local disease are devastating. Hence, despite the heightened risks of late toxicity, it is not unreasonable to consider reirradiation with the aim of improving or at least maintaining the patient s quality of life. In conclusion, the preliminary results of our experience with reirradiation using hypofractionated IG- IMRT are encouraging, with the patient reporting amelioration in visual acuity 4 months after her most recent course of radiation therapy. However, the follow-up duration in this report is too short for complete assessment of delayed morbidity given that the mean latency to radiation-associated optic neuropathy is typically 16 to 18 months with cases occurring as late as 5 years. Despite our favorable early results, it should be stressed that definitive conclusions should not be Reirradiation of Recurrent Sinonasal Carcinoma Using IG-IMRT HEAD & NECK DOI /hed September

7 drawn from a single case report. Further studies with larger patient numbers and longer follow-up are needed to fully evaluate rates of local control and toxicity as well as determine predictors of failure. REFERENCES 1. Forastiere AA, Metch B, Schuller DE, et al. Randomized comparison of cisplatin plus fluorouracil and carboplatin plus fluorouracil versus methotrexate in advanced squamous-cell carcinoma of the head and neck: a Southwest Oncology Group study. J Clin Oncol 1992;10: Forastiere AA, Shank D, Neuberg D, Taylor SG IV, DeConti RC, Adams G. Final report of a phase II evaluation of paclitaxel in patients with advanced squamous cell carcinoma of the head and neck: an Eastern Cooperative Oncology Group trial (PA390). Cancer 1998;82: Yamada Y, Lovelock DM, Bilsky MH. A review of image-guided intensity-modulated radiotherapy for spinal tumors. Neurosurgery 2007;61: ; discussion Dawson LA, Jaffray DA. Advances in image-guided radiation therapy. J Clin Oncol 2007;25: Dawson LA, Sharpe MB. Image-guided radiotherapy: rationale, benefits, and limitations. Lancet Oncol 2006;7: Verellen D, Ridder MD, Linthout N, Tournel K, Soete G, Storme G. Innovations in image-guided radiotherapy. Nat Rev Cancer 2007;7: Lee NY, Terezakis SA. Intensity-modulated radiation therapy. J Surg Oncol 2008;97: Sulman EP, Schwartz DL, Le TT, et al. IMRT reirradiation of head and neck cancer-disease control and morbidity outcomes. Int J Radiat Oncol Biol Phys 2009;73: Lee N, Chan K, Bekelman JE, et al. Salvage reirradiation for recurrent head and neck cancer. Int J Radiat Oncol Biol Phys 2007;68: Voynov G, Heron DE, Burton S, et al. Frameless stereotactic radiosurgery for recurrent head and neck carcinoma. Technol Cancer Res Treat 2006;5: Dawson LA, Anzai Y, Marsh L, 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 2000;46: Weichselbaum RR, Beckett MA, Schwartz JL, Dritschilo A. Radioresistant tumor cells are present in head and neck carcinomas that recur after radiotherapy. Int J Radiat Oncol Biol Phys 1988;15: Lee NY. Intensity-modulated radiation therapy in the treatment of head and neck cancer involving the base of the skull. Int J Radiat Oncol Biol Phys 2007;69(2 Suppl):S Reirradiation of Recurrent Sinonasal Carcinoma Using IG-IMRT HEAD & NECK DOI /hed September 2011