ORIGINAL ARTICLE
Orbital carcinomas treated with adjuvant intensity-modulated radiation therapy Randa Tao, MD,1 Dominic Ma, BS,1 Vinita Takiar, MD, PhD,1 Steven J. Frank, MD,1 Clifton D. Fuller, MD, PhD,1 G. Brandon Gunn, MD,1 Beth M. Beadle, MD, PhD,1 William H. Morrison, MD,1 David I. Rosenthal, MD,1 Mark A. Edson, MD, PhD,1 Bita Esmaeli, MD,2 Michael E. Kupferman, MD,3 Ehab Y. Hanna, MD,3 Adam S. Garden, MD,1 Jack Phan, MD, PhD1* 1
Department of Radiation Oncology, The University of Texas MD Anderson Cancer Center, Houston, Texas, 2Section of Ophthalmology, The University of Texas MD Anderson Cancer Center, Houston, Texas, 3Department of Head and Neck Surgery, The University of Texas MD Anderson Cancer Center, Houston, Texas.
Accepted 6 March 2015 Published online 6 July 2015 in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/hed.24044
ABSTRACT: Background. The purpose of this study was to assess outcomes of patients with orbital carcinomas treated with orbital exenteration and intensity-modulated radiation therapy (IMRT). Methods. Twenty-nine patients were treated with orbital exenteration and postoperative IMRT between 2002 through 2011; their medical records were retrospectively reviewed. Results. Adenoid cystic carcinoma represented the most common histology (41%) followed by squamous cell carcinoma (21%). Perineural invasion (PNI) was identified in 22 patients (76%). The median radiation dose was 60 Gy (range, 60–70). Seven patients (24%) received neck radiation. The median follow-up was 43 months (range, 5–102 months). Five-year local control, overall survival (OS), and disease-free survival
rates were 83%, 60%, and 55%, respectively. PNI (p 5 .01) and especially involvement of a named nerve (p 5 .001) significantly correlated with worse OS. Conclusion. Favorable disease control rates for orbital carcinomas are achievable with IMRT after orbital exenteration even for patients with advanced disease. Toxicity for the contralateral eye was minimal. C 2015 Wiley Periodicals, Inc. Head Neck 38: E580–E587, 2016 V
INTRODUCTION
mary treatment of orbital carcinomas is surgical resection. Adjuvant radiation is indicated for patients with high-risk features, including lymph node involvement, perineural invasion (PNI), locally advanced stage of disease, recurrent tumors, and close or positive surgical margins.3,4 Radiation techniques for treating orbital carcinomas can vary depending on the anatomy and the era of treatment.5–7 Conventional 2D external beam radiotherapy with a single en face photon or electron beam, or via wedged anterior and lateral fields, have been historically utilized after orbital exenteration; albeit with the risk of contralateral eye toxicity, hypothalamic and pituitary dysfunction, and radiation-associated secondary cancers.8–10 Electrons have attractive depth dose characteristics with rapid dose fall off and can be delivered safely to orbital targets with excellent coverage, but are limited by their depth of coverage, particularly for PNI requiring coverage of nerve tracks. Threedimensional conformal external beam radiotherapy allows for some improved tumor dose conformality but produces unacceptable plans for treatment of irregular targets close to critical structures. Compared to these traditional techniques, intensity-modulated radiation therapy (IMRT) allows for conformal shaping of radiation doses to irregular tumor volumes and thus more effective doses delivered to the tumor with fewer adverse effects.11–14 Since 2002, patients undergoing head and neck cancer radiation therapy at The University of Texas MD Anderson Cancer Center (MD Anderson) were routinely treated
The orbit is an anatomic structure that describes the cavity and contents containing the eye, the optic nerve, extraocular muscles, and other orbital soft tissue. Malignancies of the eye and orbit are rare, with an incidence of 0.8 per 100,000.1 In an analysis of data from the Florida Cancer Registry, lymphomas were found to be the most common malignancy of the orbit, accounting for over half (55%) of the histologic types represented.2 Orbital carcinomas then followed in incidence and represented 25% of orbital malignancies as a group. As a whole, orbital carcinomas include primary epithelial malignancies arising from orbital contents and surrounding structures, such as the skin, eyelids, conjunctiva, and the lacrimal apparatus. Treatment of orbital carcinomas poses unique challenges. The goal of achieving local control must be weighed against the toxicity and long-term effects of treatment, which can result in significant morbidity, such as loss of vision and/or poor cosmetic impact. The pri-
*Corresponding author: J. Phan, Department of Radiation Oncology, The University of Texas MD Anderson Cancer Center, 1515 Holcombe Blvd., Unit 97, Houston, TX 77030. E-mail: [email protected] This work was presented in part at the 55th Annual Meeting of the American Society for Therapeutic Radiology and Oncology, September 22–25, 2013, Atlanta, Georgia. Ehab Y. Hanna, MD, Editor, was recused from consideration of this manuscript.
E580
HEAD & NECK—DOI 10.1002/HED
APRIL 2016
KEY WORDS: orbital carcinomas, exenteration, intensity-modulated radiation therapy (IMRT), patterns of recurrence, preserved eye function
ORBITAL
with IMRT. Prior studies on orbital carcinomas have included a heterogeneous group of patients treated with surgery, definitive radiation, and postoperative radiation therapy with different radiation and surgery techniques used.6,15–17 There are limited data on patients treated with combined surgery and radiation with even less data on those treated with modern radiation techniques. The purpose of this study was to assess the outcomes of patients with malignant orbital carcinomas exclusively treated with orbital exenteration followed by adjuvant radiation with IMRT.
MATERIALS AND METHODS This study was approved by the Institutional Review Board. Patients treated in the Department of Radiation Oncology at MD Anderson between 2002 through 2011 with carcinomas of the orbital tissues formed the parent cohort. Patients with lymphoma, melanoma, or sarcoma histology were excluded from further analysis because these histologies have very different radiosensitivity and are therefore treated with a wide range of radiation doses. As this article attempts to study a homogenous population of patients treated with IMRT after orbital exenteration, those patients treated with other surgical or radiation approaches were also excluded. The medical records and institutional databases were retrospectively reviewed for clinical and pathologic characteristics, surgery, radiation treatment details, and patient outcomes. Staging was based on the American Joint Committee on Cancer seventh edition for the respective sites of involvement. All patients were evaluated by the multidisciplinary team and presented at the weekly head and neck multidisciplinary conference attended by head and neck surgeons, ophthalmologists, medical oncologists, and radiation oncologists. Patients were seen in follow-up approximately 6 to 8 weeks after completion of radiation and then every 3 to 4 months with repeat imaging, including a head and neck CT and/or MRI orbit for the first year. This was generally followed by repeat imaging with a CT and/or MRI every 6 months until 3 to 4 years after treatment; afterward, patients underwent yearly follow-up imaging. All patients underwent baseline evaluation by the ophthalmology service and were seen in follow-up 1.5 to 6 months after completion of radiation with an ophthalmologic examination. This was followed by yearly ophthalmologic examinations. The primary outcomes assessed were overall survival (OS), disease-specific survival (DSS), and patterns of failure. Local control was defined as freedom from disease in the periorbital tissues and postoperative bed. Regional control was defined as freedom from disease in the draining regional lymphatics. A local in-field recurrence was defined as recurrence within the high-dose clinical target volume (CTV1). Survival outcomes were calculated using the Kaplan–Meier method with prognostic factors determined by the log-rank test. Statistical analysis was performed using SPSS version 21 (IBM, Chicago, IL) and JMP version 10 (SAS Institute, Cary, NC). Side effects from treatment were graded using the Common Toxicity Criteria for Adverse Events version 4.03 guidelines. We performed nonparametric comparisons of means using the Wilcoxon rank sum test to determine whether there were
CARCINOMAS TREATED WITH
IMRT
TABLE 1. Patient and tumor characteristics. Variables
No. of patients
%
7 22
24 76
11 18
38 62
6 12 1 4 1 2 1 1 1
21 41 3 14 3 7 3 3 3
0 3 2 11 1 10 2
0 10 7 38 3 34 7
27 1 1
93 3 3
6 22 1
21 76 3
25 4
86 14
5 7 17
17 24 59
19 2 7 1
66 7 24 3
Sex Female Male Laterality Left Right Histology Squamous cell carcinoma Adenoid cystic carcinoma Adenocarcinoma Basal cell carcinoma Neuroendocrine carcinoma Sebaceous carcinoma Adnexal carcinoma Salivary duct carcinoma Other (spindle-cell carcinoma) T classification T1 T2 T3 T4a T4b Recurrent Tx/no staging available N classification N0 N1 N2 PNI No Yes Unknown/not evaluable Named nerve invasion No Yes Lymphovascular invasion No Yes Unknown Margin status Negative Close (12 wk Chemotherapy administered No Yes Chemotherapy agent Cisplatin (single-agent) Carboplatin (single-agent) Carboplatin and docetaxel Carboplatin, paclitaxel, and herceptin Intra-arterial cisplatin and adriamycin Erlotinib TPF Chemotherapy sequencing Surgery->chemo->radiation Chemo->surgery->radiation Chemo->surgery->concurrent chemoRT Surgery->radiation->chemo
No. of patients
%
21 2 3 1 2
72 7 10 3 7
22 7
76 24
27 2
93 7
22 7
76 24
20 9
69 31
3 1 1 1 1 1 1
10 3 3 3 3 3 3
2 5 1 1
22 56 11 11
Abbreviations: fx, fraction; TPF, docetaxel, cisplatin, and 5-fluorouracil; chemo, chemotherapy; chemoRT, chemoradiation.
Patterns of failure The median follow-up was 43 months (range, 5–102 months). Six patients (21%) had local recurrences, including 1 patient experiencing 2 distinct local failure events (therefore there were a total of 7 local recurrence events) and 1 patient also having a regional recurrence. The 3year and 5-year local recurrence-free survival rates were 91% and 83%, respectively (see Figure 3). Four of the 7 local recurrence events were within the treated radiation field and were salvaged with additional surgery. Of the 3 recurrence events outside the radiation field, 1 was in an unirradiated cavernous sinus and the other 2 were dermal recurrences in the scalp. Two patients with recurrent disease in the unirradiated scalp were salvaged with surgery; 1 patient eventually developed distant metastatic disease and the other remained disease-free. One patient with the cavernous sinus recurrence was locally salvaged with stereotactic radiosurgery but eventually developed distant metastatic disease. The median time to local failure was 25 months (range, 1–79 months) after completion of radiation treatment. Three of the 6 patients with local recurrences had adenoid cystic carcinomas, 2 patients had squamous cell carcinomas, and 1 patient had an adenocarcinoma. One patient who experienced a regional recurrence in the unirradiated ipsilateral neck developed the recurrence
CARCINOMAS TREATED WITH
IMRT
6 months after completion of radiation. The patient was initially treated with an orbital exenteration for a recurrent squamous cell carcinoma of the eyelid with surgical pathology demonstrating extensive PNI and a close (1 mm) margin. He was treated with adjuvant radiation and received a total dose of 60 Gy in 30 fractions targeting the resection cavity and a 54 Gy volume targeting the trigeminal nerve extending into Meckel’s cave. He was found to have ipsilateral right neck nodal recurrence, which was treated with neck dissection followed by radiation to the neck. He then experienced a local recurrence within the original radiation field with locally advanced disease. This was treated with chemotherapy. There were no other patients who experienced regional failures in the lymph nodes. Eight patients (28%) experienced distant metastatic disease at a median time of 29 months (range 5 6–56 months) after initial diagnosis. Five of the 8 patients who developed metastatic disease had adenoid cystic carcinomas, 2 patients had sebaceous carcinoma histology, and 1 patient had a poorly differentiated carcinoma with neuroendocrine features. Six of the 8 patients also had a T4 classification or recurrent disease at the primary site. The most common site of involvement was the skeleton with metastases observed in 4 patients (50%), followed by multiple sites of metastases in 3 patients (38%). The remaining 1 patient had metastatic disease to the lungs (12%). The 3-year and 5-year distant metastasis-free survival rate was 80% and 73%, respectively (see Figure 3).
Survival outcomes The 3-year and 5-year OS rates were 70% and 60%, respectively. The 3-year and 5-year DSS rates for the entire group were 72% and 66%, respectively (see Figure 3). The 3-year and 5-year disease-free survival rates were 70% and 55%, respectively. On univariate analysis, PNI significantly correlated with worse OS (p 5 .01) and DSS (p 5 .03); involvement of a named nerve was also associated with a worse OS (p 5 .001). There were no local or regional failures in the 6 patients who did not have PNI. However, 2 of these patients developed distant metastatic disease. The 5-year local control rate for patients with PNI was 75% as compared to 100% in patients without PNI; this difference was not statistically significant (p 5 .24). Surgical margin status did not have a significant impact on OS (p 5 .30) or DSS, although there was a trend toward better DSS with negative margins with 5year DSS rates of 83% versus 31% (negative vs positive margins; p 5 .06). Patients who received chemotherapy or elective nodal radiation had a trend toward worse OS outcomes, although the differences were not statistically significant. Tumor histology did not significantly impact OS, but those with basal cell histologies had improved DSS and local control rates compared with those with squamous cell and adenoid cystic histologies (see Figure 4). The 5year DSS rate for patients with basal cell carcinoma histology was 100% compared to 57% for adenoid cystic carcinoma and 56% for squamous cell carcinoma (p 5 .006). Similarly, the 5-year local control rate was 100% for basal cell carcinomas, 80% for adenoid cystic carcinomas, and 56% for squamous cell carcinomas HEAD & NECK—DOI 10.1002/HED
APRIL 2016
E583
TAO ET AL.
FIGURE 3. Survival curves for (A) local-regional control, (B) disease-free survival, (C) distant metastasis–free survival, (D) overall survival, and (E) disease-specific survival.
(p 5 .04; Figure 4). Multivariate analysis was not performed given the small number of patients.
Toxicity All patients maintained function of the contralateral eye after treatment. Based on follow-up evaluation by the ophthalmology service, overall contralateral eye function was graded as “with impairment,” “without impairment,” or “nonfunctional.” Twenty-four patients (83%) had complete function of the contralateral eye without any impairment, whereas 5 patients (17%) experienced symptoms that caused impairment. These results are summarized in Table 3. Two patients developed a grade 3 contralateral eye cataract requiring cataract surgery; both patients subsequently regained 20/20 vision. There were 4 patients who were found to have a grade 1 or 2 nuclear sclerotic change/cataract on detailed ophthalmologic examinations but did not experience detectable visual impairment. A total of 2 patients experienced a detectable change in visual acuity from baseline after radiation; all patients had no more than a 5-point Visual Acuity Scale change (eg, from baseline of 20/20 to 20/25). Other documented side effects included dry eye (2 patients), eye pain (1 patient), and eyelid abnormalities, including blepharitis (1 patient). There were no other late toxicities observed for the contralateral eye, including decreased visual fields, conjunctivitis, glaucoma, or keratitis. There were no statistically significant differences in the mean or maximum doses to the contralateral optic structures for patients who experienced side effects. E584
HEAD & NECK—DOI 10.1002/HED
APRIL 2016
As for nonophthalmologic toxicities, 1 patient developed grade 2 hearing loss in the ipsilateral ear (on the same side as the radiated orbit) that did not require further intervention. There was 1 patient who experienced a grade 1 ipsilateral temporal lobe necrosis who was treated to a dose of 66 Gy in 33 fractions. Other late toxicities included grade 1 skin changes in 3 patients (10%), including hyperpigmentation, telangiectasia, and sinocutaneous fistula formation. One patient who received elective nodal radiation developed grade 1 neck pain. Aside from the patients who developed a contralateral eye cataract, there were no other reported grade 3 or 4 late toxicities.
DISCUSSION IMRT is now widely used for definitive and adjuvant treatment of head and neck cancers18 and prospective studies have shown the possibility to achieve decreased toxicity, such as xerostomia, without compromising treatment efficacy.19,20 Some of the best data reporting the benefits of IMRT treatment have focused on patients with head and neck cancers; however, these have largely concentrated on the more common head and neck sites, including oropharyngeal, laryngeal, and hypopharyngeal cancers. With the rarity of orbital carcinomas, there is limited information on the outcomes with IMRT treatment. In addition, the potential benefits of IMRT to contralateral eye function are important given the possible long-term damage to ocular structures and particularly in this cohort with mono-ocular vision. Therefore, we performed this retrospective analysis to determine whether
ORBITAL
CARCINOMAS TREATED WITH
IMRT
FIGURE 4. Survival curves for (A) locoregional control (LRC) by histology and (B) disease-specific survival (DSS) by histology. There were 6 patients with squamous cell carcinomas (SCCs), 12 patients with adenoid cystic carcinomas (ACCs), 4 patients with basal cell carcinomas, and the remaining 5 patients had other histologies.
IMRT offered lower toxicity for the treatment of orbital carcinomas. Our results demonstrate a high rate of disease control in patients who require both orbital exenteration and adjuvant radiation. This represents a homogeneous group of patients who are selected for aggressive treatment and the local control rates observed compare favorably to other series that included patients with locally advanced disease. It is important to note that the current body of literature on orbital carcinomas includes a majority of patients with early stage disease, with few studies including patients treated with orbital exenteration followed by adjuvant radiation therapy. One of the largest series of patients treated with radiation for carcinoma of the eyelid was reported from the Institut Curie,6 which included 850 cases but only 15 patients (
Orbital carcinomas treated with adjuvant intensity-modulated radiation therapy Randa Tao, MD,1 Dominic Ma, BS,1 Vinita Takiar, MD, PhD,1 Steven J. Frank, MD,1 Clifton D. Fuller, MD, PhD,1 G. Brandon Gunn, MD,1 Beth M. Beadle, MD, PhD,1 William H. Morrison, MD,1 David I. Rosenthal, MD,1 Mark A. Edson, MD, PhD,1 Bita Esmaeli, MD,2 Michael E. Kupferman, MD,3 Ehab Y. Hanna, MD,3 Adam S. Garden, MD,1 Jack Phan, MD, PhD1* 1
Department of Radiation Oncology, The University of Texas MD Anderson Cancer Center, Houston, Texas, 2Section of Ophthalmology, The University of Texas MD Anderson Cancer Center, Houston, Texas, 3Department of Head and Neck Surgery, The University of Texas MD Anderson Cancer Center, Houston, Texas.
Accepted 6 March 2015 Published online 6 July 2015 in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/hed.24044
ABSTRACT: Background. The purpose of this study was to assess outcomes of patients with orbital carcinomas treated with orbital exenteration and intensity-modulated radiation therapy (IMRT). Methods. Twenty-nine patients were treated with orbital exenteration and postoperative IMRT between 2002 through 2011; their medical records were retrospectively reviewed. Results. Adenoid cystic carcinoma represented the most common histology (41%) followed by squamous cell carcinoma (21%). Perineural invasion (PNI) was identified in 22 patients (76%). The median radiation dose was 60 Gy (range, 60–70). Seven patients (24%) received neck radiation. The median follow-up was 43 months (range, 5–102 months). Five-year local control, overall survival (OS), and disease-free survival
rates were 83%, 60%, and 55%, respectively. PNI (p 5 .01) and especially involvement of a named nerve (p 5 .001) significantly correlated with worse OS. Conclusion. Favorable disease control rates for orbital carcinomas are achievable with IMRT after orbital exenteration even for patients with advanced disease. Toxicity for the contralateral eye was minimal. C 2015 Wiley Periodicals, Inc. Head Neck 38: E580–E587, 2016 V
INTRODUCTION
mary treatment of orbital carcinomas is surgical resection. Adjuvant radiation is indicated for patients with high-risk features, including lymph node involvement, perineural invasion (PNI), locally advanced stage of disease, recurrent tumors, and close or positive surgical margins.3,4 Radiation techniques for treating orbital carcinomas can vary depending on the anatomy and the era of treatment.5–7 Conventional 2D external beam radiotherapy with a single en face photon or electron beam, or via wedged anterior and lateral fields, have been historically utilized after orbital exenteration; albeit with the risk of contralateral eye toxicity, hypothalamic and pituitary dysfunction, and radiation-associated secondary cancers.8–10 Electrons have attractive depth dose characteristics with rapid dose fall off and can be delivered safely to orbital targets with excellent coverage, but are limited by their depth of coverage, particularly for PNI requiring coverage of nerve tracks. Threedimensional conformal external beam radiotherapy allows for some improved tumor dose conformality but produces unacceptable plans for treatment of irregular targets close to critical structures. Compared to these traditional techniques, intensity-modulated radiation therapy (IMRT) allows for conformal shaping of radiation doses to irregular tumor volumes and thus more effective doses delivered to the tumor with fewer adverse effects.11–14 Since 2002, patients undergoing head and neck cancer radiation therapy at The University of Texas MD Anderson Cancer Center (MD Anderson) were routinely treated
The orbit is an anatomic structure that describes the cavity and contents containing the eye, the optic nerve, extraocular muscles, and other orbital soft tissue. Malignancies of the eye and orbit are rare, with an incidence of 0.8 per 100,000.1 In an analysis of data from the Florida Cancer Registry, lymphomas were found to be the most common malignancy of the orbit, accounting for over half (55%) of the histologic types represented.2 Orbital carcinomas then followed in incidence and represented 25% of orbital malignancies as a group. As a whole, orbital carcinomas include primary epithelial malignancies arising from orbital contents and surrounding structures, such as the skin, eyelids, conjunctiva, and the lacrimal apparatus. Treatment of orbital carcinomas poses unique challenges. The goal of achieving local control must be weighed against the toxicity and long-term effects of treatment, which can result in significant morbidity, such as loss of vision and/or poor cosmetic impact. The pri-
*Corresponding author: J. Phan, Department of Radiation Oncology, The University of Texas MD Anderson Cancer Center, 1515 Holcombe Blvd., Unit 97, Houston, TX 77030. E-mail: [email protected] This work was presented in part at the 55th Annual Meeting of the American Society for Therapeutic Radiology and Oncology, September 22–25, 2013, Atlanta, Georgia. Ehab Y. Hanna, MD, Editor, was recused from consideration of this manuscript.
E580
HEAD & NECK—DOI 10.1002/HED
APRIL 2016
KEY WORDS: orbital carcinomas, exenteration, intensity-modulated radiation therapy (IMRT), patterns of recurrence, preserved eye function
ORBITAL
with IMRT. Prior studies on orbital carcinomas have included a heterogeneous group of patients treated with surgery, definitive radiation, and postoperative radiation therapy with different radiation and surgery techniques used.6,15–17 There are limited data on patients treated with combined surgery and radiation with even less data on those treated with modern radiation techniques. The purpose of this study was to assess the outcomes of patients with malignant orbital carcinomas exclusively treated with orbital exenteration followed by adjuvant radiation with IMRT.
MATERIALS AND METHODS This study was approved by the Institutional Review Board. Patients treated in the Department of Radiation Oncology at MD Anderson between 2002 through 2011 with carcinomas of the orbital tissues formed the parent cohort. Patients with lymphoma, melanoma, or sarcoma histology were excluded from further analysis because these histologies have very different radiosensitivity and are therefore treated with a wide range of radiation doses. As this article attempts to study a homogenous population of patients treated with IMRT after orbital exenteration, those patients treated with other surgical or radiation approaches were also excluded. The medical records and institutional databases were retrospectively reviewed for clinical and pathologic characteristics, surgery, radiation treatment details, and patient outcomes. Staging was based on the American Joint Committee on Cancer seventh edition for the respective sites of involvement. All patients were evaluated by the multidisciplinary team and presented at the weekly head and neck multidisciplinary conference attended by head and neck surgeons, ophthalmologists, medical oncologists, and radiation oncologists. Patients were seen in follow-up approximately 6 to 8 weeks after completion of radiation and then every 3 to 4 months with repeat imaging, including a head and neck CT and/or MRI orbit for the first year. This was generally followed by repeat imaging with a CT and/or MRI every 6 months until 3 to 4 years after treatment; afterward, patients underwent yearly follow-up imaging. All patients underwent baseline evaluation by the ophthalmology service and were seen in follow-up 1.5 to 6 months after completion of radiation with an ophthalmologic examination. This was followed by yearly ophthalmologic examinations. The primary outcomes assessed were overall survival (OS), disease-specific survival (DSS), and patterns of failure. Local control was defined as freedom from disease in the periorbital tissues and postoperative bed. Regional control was defined as freedom from disease in the draining regional lymphatics. A local in-field recurrence was defined as recurrence within the high-dose clinical target volume (CTV1). Survival outcomes were calculated using the Kaplan–Meier method with prognostic factors determined by the log-rank test. Statistical analysis was performed using SPSS version 21 (IBM, Chicago, IL) and JMP version 10 (SAS Institute, Cary, NC). Side effects from treatment were graded using the Common Toxicity Criteria for Adverse Events version 4.03 guidelines. We performed nonparametric comparisons of means using the Wilcoxon rank sum test to determine whether there were
CARCINOMAS TREATED WITH
IMRT
TABLE 1. Patient and tumor characteristics. Variables
No. of patients
%
7 22
24 76
11 18
38 62
6 12 1 4 1 2 1 1 1
21 41 3 14 3 7 3 3 3
0 3 2 11 1 10 2
0 10 7 38 3 34 7
27 1 1
93 3 3
6 22 1
21 76 3
25 4
86 14
5 7 17
17 24 59
19 2 7 1
66 7 24 3
Sex Female Male Laterality Left Right Histology Squamous cell carcinoma Adenoid cystic carcinoma Adenocarcinoma Basal cell carcinoma Neuroendocrine carcinoma Sebaceous carcinoma Adnexal carcinoma Salivary duct carcinoma Other (spindle-cell carcinoma) T classification T1 T2 T3 T4a T4b Recurrent Tx/no staging available N classification N0 N1 N2 PNI No Yes Unknown/not evaluable Named nerve invasion No Yes Lymphovascular invasion No Yes Unknown Margin status Negative Close (12 wk Chemotherapy administered No Yes Chemotherapy agent Cisplatin (single-agent) Carboplatin (single-agent) Carboplatin and docetaxel Carboplatin, paclitaxel, and herceptin Intra-arterial cisplatin and adriamycin Erlotinib TPF Chemotherapy sequencing Surgery->chemo->radiation Chemo->surgery->radiation Chemo->surgery->concurrent chemoRT Surgery->radiation->chemo
No. of patients
%
21 2 3 1 2
72 7 10 3 7
22 7
76 24
27 2
93 7
22 7
76 24
20 9
69 31
3 1 1 1 1 1 1
10 3 3 3 3 3 3
2 5 1 1
22 56 11 11
Abbreviations: fx, fraction; TPF, docetaxel, cisplatin, and 5-fluorouracil; chemo, chemotherapy; chemoRT, chemoradiation.
Patterns of failure The median follow-up was 43 months (range, 5–102 months). Six patients (21%) had local recurrences, including 1 patient experiencing 2 distinct local failure events (therefore there were a total of 7 local recurrence events) and 1 patient also having a regional recurrence. The 3year and 5-year local recurrence-free survival rates were 91% and 83%, respectively (see Figure 3). Four of the 7 local recurrence events were within the treated radiation field and were salvaged with additional surgery. Of the 3 recurrence events outside the radiation field, 1 was in an unirradiated cavernous sinus and the other 2 were dermal recurrences in the scalp. Two patients with recurrent disease in the unirradiated scalp were salvaged with surgery; 1 patient eventually developed distant metastatic disease and the other remained disease-free. One patient with the cavernous sinus recurrence was locally salvaged with stereotactic radiosurgery but eventually developed distant metastatic disease. The median time to local failure was 25 months (range, 1–79 months) after completion of radiation treatment. Three of the 6 patients with local recurrences had adenoid cystic carcinomas, 2 patients had squamous cell carcinomas, and 1 patient had an adenocarcinoma. One patient who experienced a regional recurrence in the unirradiated ipsilateral neck developed the recurrence
CARCINOMAS TREATED WITH
IMRT
6 months after completion of radiation. The patient was initially treated with an orbital exenteration for a recurrent squamous cell carcinoma of the eyelid with surgical pathology demonstrating extensive PNI and a close (1 mm) margin. He was treated with adjuvant radiation and received a total dose of 60 Gy in 30 fractions targeting the resection cavity and a 54 Gy volume targeting the trigeminal nerve extending into Meckel’s cave. He was found to have ipsilateral right neck nodal recurrence, which was treated with neck dissection followed by radiation to the neck. He then experienced a local recurrence within the original radiation field with locally advanced disease. This was treated with chemotherapy. There were no other patients who experienced regional failures in the lymph nodes. Eight patients (28%) experienced distant metastatic disease at a median time of 29 months (range 5 6–56 months) after initial diagnosis. Five of the 8 patients who developed metastatic disease had adenoid cystic carcinomas, 2 patients had sebaceous carcinoma histology, and 1 patient had a poorly differentiated carcinoma with neuroendocrine features. Six of the 8 patients also had a T4 classification or recurrent disease at the primary site. The most common site of involvement was the skeleton with metastases observed in 4 patients (50%), followed by multiple sites of metastases in 3 patients (38%). The remaining 1 patient had metastatic disease to the lungs (12%). The 3-year and 5-year distant metastasis-free survival rate was 80% and 73%, respectively (see Figure 3).
Survival outcomes The 3-year and 5-year OS rates were 70% and 60%, respectively. The 3-year and 5-year DSS rates for the entire group were 72% and 66%, respectively (see Figure 3). The 3-year and 5-year disease-free survival rates were 70% and 55%, respectively. On univariate analysis, PNI significantly correlated with worse OS (p 5 .01) and DSS (p 5 .03); involvement of a named nerve was also associated with a worse OS (p 5 .001). There were no local or regional failures in the 6 patients who did not have PNI. However, 2 of these patients developed distant metastatic disease. The 5-year local control rate for patients with PNI was 75% as compared to 100% in patients without PNI; this difference was not statistically significant (p 5 .24). Surgical margin status did not have a significant impact on OS (p 5 .30) or DSS, although there was a trend toward better DSS with negative margins with 5year DSS rates of 83% versus 31% (negative vs positive margins; p 5 .06). Patients who received chemotherapy or elective nodal radiation had a trend toward worse OS outcomes, although the differences were not statistically significant. Tumor histology did not significantly impact OS, but those with basal cell histologies had improved DSS and local control rates compared with those with squamous cell and adenoid cystic histologies (see Figure 4). The 5year DSS rate for patients with basal cell carcinoma histology was 100% compared to 57% for adenoid cystic carcinoma and 56% for squamous cell carcinoma (p 5 .006). Similarly, the 5-year local control rate was 100% for basal cell carcinomas, 80% for adenoid cystic carcinomas, and 56% for squamous cell carcinomas HEAD & NECK—DOI 10.1002/HED
APRIL 2016
E583
TAO ET AL.
FIGURE 3. Survival curves for (A) local-regional control, (B) disease-free survival, (C) distant metastasis–free survival, (D) overall survival, and (E) disease-specific survival.
(p 5 .04; Figure 4). Multivariate analysis was not performed given the small number of patients.
Toxicity All patients maintained function of the contralateral eye after treatment. Based on follow-up evaluation by the ophthalmology service, overall contralateral eye function was graded as “with impairment,” “without impairment,” or “nonfunctional.” Twenty-four patients (83%) had complete function of the contralateral eye without any impairment, whereas 5 patients (17%) experienced symptoms that caused impairment. These results are summarized in Table 3. Two patients developed a grade 3 contralateral eye cataract requiring cataract surgery; both patients subsequently regained 20/20 vision. There were 4 patients who were found to have a grade 1 or 2 nuclear sclerotic change/cataract on detailed ophthalmologic examinations but did not experience detectable visual impairment. A total of 2 patients experienced a detectable change in visual acuity from baseline after radiation; all patients had no more than a 5-point Visual Acuity Scale change (eg, from baseline of 20/20 to 20/25). Other documented side effects included dry eye (2 patients), eye pain (1 patient), and eyelid abnormalities, including blepharitis (1 patient). There were no other late toxicities observed for the contralateral eye, including decreased visual fields, conjunctivitis, glaucoma, or keratitis. There were no statistically significant differences in the mean or maximum doses to the contralateral optic structures for patients who experienced side effects. E584
HEAD & NECK—DOI 10.1002/HED
APRIL 2016
As for nonophthalmologic toxicities, 1 patient developed grade 2 hearing loss in the ipsilateral ear (on the same side as the radiated orbit) that did not require further intervention. There was 1 patient who experienced a grade 1 ipsilateral temporal lobe necrosis who was treated to a dose of 66 Gy in 33 fractions. Other late toxicities included grade 1 skin changes in 3 patients (10%), including hyperpigmentation, telangiectasia, and sinocutaneous fistula formation. One patient who received elective nodal radiation developed grade 1 neck pain. Aside from the patients who developed a contralateral eye cataract, there were no other reported grade 3 or 4 late toxicities.
DISCUSSION IMRT is now widely used for definitive and adjuvant treatment of head and neck cancers18 and prospective studies have shown the possibility to achieve decreased toxicity, such as xerostomia, without compromising treatment efficacy.19,20 Some of the best data reporting the benefits of IMRT treatment have focused on patients with head and neck cancers; however, these have largely concentrated on the more common head and neck sites, including oropharyngeal, laryngeal, and hypopharyngeal cancers. With the rarity of orbital carcinomas, there is limited information on the outcomes with IMRT treatment. In addition, the potential benefits of IMRT to contralateral eye function are important given the possible long-term damage to ocular structures and particularly in this cohort with mono-ocular vision. Therefore, we performed this retrospective analysis to determine whether
ORBITAL
CARCINOMAS TREATED WITH
IMRT
FIGURE 4. Survival curves for (A) locoregional control (LRC) by histology and (B) disease-specific survival (DSS) by histology. There were 6 patients with squamous cell carcinomas (SCCs), 12 patients with adenoid cystic carcinomas (ACCs), 4 patients with basal cell carcinomas, and the remaining 5 patients had other histologies.
IMRT offered lower toxicity for the treatment of orbital carcinomas. Our results demonstrate a high rate of disease control in patients who require both orbital exenteration and adjuvant radiation. This represents a homogeneous group of patients who are selected for aggressive treatment and the local control rates observed compare favorably to other series that included patients with locally advanced disease. It is important to note that the current body of literature on orbital carcinomas includes a majority of patients with early stage disease, with few studies including patients treated with orbital exenteration followed by adjuvant radiation therapy. One of the largest series of patients treated with radiation for carcinoma of the eyelid was reported from the Institut Curie,6 which included 850 cases but only 15 patients (
