Treatment outcomes after definitive radio(chemo)therapy for 17 lacrimal sac squamous cell carcinoma

Abstract

Objectives:

Tumors of the lacrimal sac are rare and life-threatening. Because of their rarity, no extensive clinical data on their management and prognosis exist. We investigated the application of definitive radiation therapy and its outcome in patients with lacrimal sac squamous cell carcinoma (LSSCC).

Methods:

We retrospectively studied 17 patients with LSSCC at a single institution between 2003 and 2017. All the patients were treated with definitive radiotherapy, and 11 patients were delivered with cisplatin-based chemotherapy. The patients’ clinical records were reviewed for symptoms, pathological types, the volume and dosimetry of the tumors and their adjacent structures, radiation coverage of lymph node drainage areas, treatment outcomes, and complications from definitive radiotherapy.

Results:

Median follow-up was 38.9 months, and age at diagnosis was 48 years.The 2-year and 5-year overall survival, progression-free survival, locoregional control, and disease metastasis-free survival rates were 94.1 and 84.7%, 88.2 and 73.5%, 93.8%, 94.1, and 78.4%, respectively. A total dose of 6600–7000 cGy was prescribed to the tumor. Levels Ⅰb, Ⅶa, Ⅷ, and Ⅸ were covered with the clinical target volume regardless of lymph involvement. Acute Grade 3 radiation dermatitis occurred in seven patients (17.6%), but no acute Grade 4 or Grade 5 toxicity of any type occurred. Seven (41.2%, 7/17) of the treated eyes had moderated vision impairments; 17.6% (3/17) of patients developed cataracts, and glaucoma and radiation retinopathy were found in 5.9% (1/17) of patients.

Conclusions:

Definitive radiotherapy could be a treatment option for those who refuse surgery or have unresectable LSSCC.

Advances in knowledge:

Radiation alone is a treatment option for LSSCC.

Introduction

The clinical presentation of lacrimal sac squamous cell carcinoma (LSSCC) resembles that of chronic dacryocystitis, which does not cause alarm until the appearance of specific symptoms, such as blood-stained tearsora palpable lump. So, diagnosis is often delayed in LSSCC. Fewer than 15% of cases are diagnosed within 2 months, and 72% of patients initiate treatment within 12 months.¹ LSSCC can involve the lacrimal sac and also grow through the nasolacrimal duct to invade its peripheral organs and structures. Lymph node involvement at diagnosis was found in 36.2% of patients. The most common sites of lymph node metastasis are cervical, retropharyngeal, preauricular, and submandibular.²

Multidisciplinary management of LSSCC is becoming accepted increasingly, and the standard therapy is local resection of the tumor followed by adjuvant radiation therapy, with or without chemotherapy.³ En bloc tumor and lacrimal drainage duct excision are recommended for cases of LSSCC.⁴ Therefore, orbital exenteration is needed for some cases with extensive tumor invasion to the orbital and its adjacent structures; however, some patients refuse such surgery because of concerns about disfigurement. Radiation therapy also plays an essential role in the treatment of the LSSCC, especially for those patients who refuse surgery or have unresectable tumors.² Previous results demonstrated the treatment efficacy of irradiation being used as adjuvant therapy in postoperative patients with lacrimal sac tumors.⁵ Especially for the patients with LSSCC who are unfit for multidisciplinary therapy, or refuse such therapy, definitive radiation therapy could be used as a valid treatment modality. Delivery of radiation doses high enough to achieve local control of most epithelial tumors of the orbit and ocular adnexa might cause unacceptable toxicity to the globe and its nearby structures.⁶ So, how to employ radiation therapy to achieve better outcomes and reduce impairment is extremely important for the treatment of LSSCC.

The purposes of this study are to share our experience of radiation therapy for LSSCC, to assess the impact of radiation therapy on disease control, and present toxicities and complications in patients with LSSCC. Besides, we discuss the possibility of definitive radiation therapy as a radical treatment for LSSCC. We present, herein, our radiation therapy techniques and parameters, along with the contouring and dosage volume distributions.

Methods and materials

Patients

This study was approved by the institutional review board of, and informed consent for research was obtained from patients. Between January 2003 and May 2017, 17 patients were treated with definitive radiation therapy for a histologically proven LSSCC. The ratio between males and females was 11:6. The median age was 48 years (range, 22–80). Patient characteristics are described in Table 1. None of the patients were operated on for reasons such as unresectable lesions or refusing surgery. All patients were free of distant metastasis at the time of diagnosis. We retrospectively reviewed the patients’ medical records, including the general characteristics, pathological reports, symptoms and signs, the volume and dosimetry of the tumor and the adjacent structures, and radiation coverage of lymph node drainage areas. We evaluated the survival, local control, and the types and severity of treatment-related complications. No defined staging system is in the lacrimal tumors, all the cases were classified following Wang’s protocol in this cohort, that the lacrimal squamous cell carcinoma was divided into four clinical stages combing physical examination, imaging features, and tumor prognosis.⁷ Delineation of the cervical lymph refers to the DAHANCA, EORTC, HKNPCSG, NCICCTG, NCRI, RTOG, TROG consensus guidelines.⁸

Table 1.

Patient characteristics (n = 17)

Characteristic	n	%
Sex
Male	11	64.7
Female	6	35.3
Age
>50	8	47.1
≤50	9	52.9
Lymph nodes
Negative	9	52.9
Positive	8	47.1
Symptoms
Epiphora	7	41.2
Mass	7	41.2
Nose bleeds/nasal obstruction	4	29.4
Clinical stage
Ⅱ	4	23.5
Ⅲ	5	29.4
Ⅳ	8	47.1
Chemotherapy
Yes	11	64.7
No	6	35.3
Radiation therapy technique
IMRT	11	64.7
CRT	6	35.3

CRT, conventional radiation therapy.; IMRT, intensity modulated radiation therapy.

All patients underwent definitive radiotherapy, with some (11/17, 64.7%) also receiving chemotherapy, including induction chemotherapy, concurrent chemotherapy, or adjuvant chemotherapy. Induction chemotherapy was either TPF (docetaxel +cisplatin +5-fluorouracil), TP (docetaxel +cisplatin), PF (cisplatin +5-fluorouracil), or GP (gemcitabine +cisplatin) in ten patients. Two cycles of induction chemotherapy were followed by the concurrent chemoradiotherapy after three weeks. The application of chemotherapeutics in every three weeks with gemcitabine 1000 mg/m² on day 1 and day 8, docetaxel 70 mg/m² on day 1, cisplatin 75 mg/m² on day 1 to day 3, and 5-FU 500 mg/m² from day 1 to day 4. Concurrent chemotherapy consisted of cisplatin 75 mg/m² divided evenly across three days in a 3-week interval. Induction chemotherapy was used for the patients who had cervical lymph nodes, and the concurrent chemotherapy is mainly for the extensive primary tumor invasion. 4–6 cycles of chemotherapy were generally administered, and eight patients were delivered with platinum-based concurrent chemotherapy if less than four cycles of before. Four patients received subsequent adjuvant chemotherapy that primarily consisted of PF within three weeks after completing radiation therapy. Among all the 11 patents, five patients received induction chemotherapy (IC) and concurrent chemotherapy (CCRT), four had CCRT and adjuvant chemotherapy, and two had IC and adjuvant chemotherapy.

Radiotherapy details

A head-and-neck mask was made for immobilization at the time of the computed tomography (CT) simulation (GE Hispeed F/X), with a shape adaptive bolus if necessary. The CT scan from superior to the frontal sinus to the upper mediastinum and the axial images with 1 mm or 3 mm slice thickness of the head and neck were obtained, respectively. Treatment planning was performed on a 3-dimensional CT image-based planning system (Philips Pinnacle).³ The eyeballs, lens, optic nerves, and optic chiasma were outlined on each CT slice. The gross tumor volume (GTV) was defined as the gross extent of the tumor by imaging and physical examination, including the primary tumor in the lacrimal sac as gross tumor volume of tumor boundary (GTVtb or GTV1) and lymph nodes in the neck as nodal gross tumor volume (GTVnd or GTV2). GTVtb and GTVnd corresponded to the initial tumor volume before chemotherapy in case of induction chemotherapy. According to Grégoire et al,⁹ we delineated two clincial target volumes (CTVs) for the primary tumor, the so-called CTV1 and CVT2, corresponding to a higher and lower tumor burden, which should be associated with a higher and a lower dose prescription, respectively. CTV1 consisted of a 5 mm–to 10 mm expansion from the GTVtb, into the ipsilateral lacrimal duct, and partial nasal cavity, the anterior ethmoid sinus, and the partial posterior ethmoid sinus/ sphenoid sinus, with subsequent adjustment according to the involvement of the primary tumor. CTV2 consisted of a 5 mm to 10 mm expansion from the CTV1. CTV2 covered the entire ipsilateral nasal cavity, ethmoid sinus, partial sphenoid sinus, nasopharyngeal cavity, and the skin of the infraorbital margin. For some patients with nasal cavity or sinus cavity invasion, the contralateral nasal cavity or the whole involved sinus cavity was also included in CTV2. CTV2 also covered the elective lymphatic drainage areas.

All CTVs were created by at least two radiation oncologists to ensure coverage of areas at risk of tumor spread. The planning target volume (PTV) was derived by expanding the GTV and CTV with a margin of 1–3 mm depending on the anatomical relationship to critical structures. Some vital structures, like the eyeball, retina, optical nerve, and optical chiasm, were expanded to a planning risk volume with a margin of 2 mm. In cases of overlap between the PTV of tumor and planning risk volume, the margin of the PTV might be adjusted later to reflect actual tumor shrinkage.^10,11

Patients were irradiated using intensity-modulated radiation therapy (IMRT) or conventional three-dimensional-conformed radiation therapy (3D-CRT) during the entire radiation therapy treatment. A6 MV linear accelerator or electron-beam irradiation to the primary tumor was performed. A total dose of 6600–7000 cGy was prescribed so that at least 95% of GTVtb and at most 110% of GTVtb received the prescribed dose. The nine irradiation beams were angled in IMRT to avoid the cornea and retina, and protection against radiation damage to the cornea was used in 3D-CRT and electron-beam irradiation. 3D-CRT plans were delivered with a conventional fractionation schedule of GTVtb200 cGy per fraction in five daily fractions per week. IMRT plans were delivered with a schedule of GTVtb 215–225 cGy per fraction, GTVnd 200–215 cGy per fraction, CTV1 180–200 cGy per fraction, and CTV2 160–200 cGy per fraction in five daily fractions per week.

The gantry angles were selected to avoid critical structures and to minimize the projection of the treatment target. In 3D-CRT, the cornea was shielded from the X-ray field, and in IMRT, we can adjust the nine beams that form the radiation fields to avoid high doses to the cornea. A bolus is needed for superficial tumors to improve the radiation dose distribution to the target volume, and the bolus should conform well and minimize air gaps between the skin and the bolus. For patients with cervical lymph node metastasis, the ipsilateral neck should be irradiated; for patients with skin involvement, levels VIII, IX, and I should be considered for radiation coverage using CTV2. A CT scan was performed to evaluate the regression of the primary tumor and lymph nodes when the radiation dose of GTVtb reached 5000cGy. If the radiation dose to optical organs was beyond the tolerance of limit, 2–3 adapted plans were required to protect the eyeball or optical nerves better. We delivered almost 5500–5800 cGy of GTVtb in Phase 1 and 900–1100 cGy of GTVtb in the left phase, depending on the volume of tumor shrinkage.

Follow-Up

All patients were examined weekly during definitive radiotherapy treatment. Acute radiation-related toxicity was evaluated and recorded weekly during the treatment by two radiation oncologists, according to the National Cancer Institute’s Common Terminology Criteria for Adverse Events version 3.0 (CTCAE v3). Post-treatment assessments of patients, including a physical examination, late toxicities, and vision impairment evaluation, were planned for every 6–8 weeks during the first year, every 10–12 weeks during the second year, and then every 4–6 months after that. Magnetic resonance imaging (MRI) or CT scan of the head and neck was performed approximately every 6 months during the follow-up. Patients were also asked to see an ophthalmologist and otolaryngologist periodically.

Statistical analysis

Assessed outcomes were OS, PFS, locoregional control rate (LRC), and distant metastasis-free survival (DMFS). OS time was measured between the date of the initial diagnosis to the date of death; PFS was defined as the time between the date at which the patients first sought a diagnosis to the date of disease progression; LRC was defined as freedom from disease in the lacrimal sac or lymph nodes of the neck; DMFS was defined as the time between the date of the initial diagnosis and distant metastasis.The vision was also recorded from clinical examinations performed at initial assessment and during the follow-up.

Qualitative parameters were described by frequency and percentage, quantitative parameters by the mean and standard deviation. Clinical outcomes such as 5-year OS and PFS were determined by the Kaplan-Meier method. The differences were calculated using the log-rank test, and p < 0.05 was statistically significant. Statistical analyses were performed using SPSS software v.20.0 (IBM Corp., Armonk, NY, USA).

Results

Patient characteristics

A total of 17 patients were included in the analysis. An incisional biopsy was initially performed in all patients, and all were diagnosed with squamous cell carcinomas. Fifteen of the 17 patients (88.2%) were diagnosed with moderate-to-poorly differentiated pathological type, and the others with well-differentiated tumors (Figure 1). Seven patients (41.2%) had a history of epiphora, which is the most common symptom, and five patients (29.4%) complained of nose bleeds or nasal obstruction at their first visits. Seven patients (41.2%) presented with a palpable mass in the lacrimal sac. Four patients in stage Ⅱ that tumor invades the eyeball, or naso-lacrymal duct, or lacrimal canaliculi, or palpebral conjunctiva. Five patients in stage Ⅲ that tumor invades the nasal cavity, or sinus, or the peripheral bone, or the skin. Eight patients in stage Ⅳ that tumor invades the orbital apex, or meninges, or bran, or lymph nodes. CT/MRI scans showed bone erosion of the lacrimal sac in 12 patients (70.6%), and tumor invasion into the adjacent sinuses in seven patients (41.2%). A total of eight patients (47.1%) were found to have lymph node involvement in the neck, and the most common site was levels II-IV (5/17, 47.1%), followed by level Ib (3/17, 17.6%), level VII (3/17, 17.6%), and levels VIII-IX (3/17, 17.6%; Figure 2).

Figure 1.

Histopathology using hematoxylin and eosin (H&E,×20) showing features of a representative squamous cell carcinoma

Figure 2.

A 39-year-old male patient with squamous cell carcinoma of the right lacrimal sac. Axial post-contrast T₁-weighted MRI scans show an enhancement mass in the right lacrimal sac area, lymph node in right retropharyngeal space (B), and lymph nodes in the right neck (C, arrow) before radiation therapy. Axial post-contrast T₁-weighted MRI scan demonstrated that the primary tumor (D), right retropharyngeal lymph node (E), and right cervical lymph nodes (F, arrow) decreased significantly after completion of treatment.

Survival and local control

All patients were treated with definitive radiotherapy, with some (11/17, 64.7%) also receiving chemotherapy as induction chemotherapy or/and concurrent chemotherapy and adjuvant chemotherapy.The median follow-up period for survivors was 38.9 months (range, 7.4–106.4 months). The 2 and 5 year OS rates were 94.1 and 84.7%, respectively. The 2 and 5 year PFS rates were 88.2 and 73.5%, respectively. The 2 and 5 year LRC rates were 93.8%. The 2 and 5 year DMFS rates were 94.1 and 78.4%, respectively. The OS, PFS, LRC, and DMFS curves are shown in Figure 3. At the time of analysis, 14 patients were alive. Two patients had died of distant metastasis, and one patient had died of local recurrence. Among the three patients who died, one progressed with distant metastasis to the abdominal nodes, one had a local recurrence, and another progressed with local control failure and distant metastasis to the lung.The pathological types of these three cases were: two cases of poorly differentiated squamous cell carcinomas and one case of squamous cell canceration. The clinical-stage distribution of the three cases were: two cases of stage IV and one case of stage III.

Figure 3.

Kaplan-Meier estimate of 5-year overall survival rate (A), locoregional control rate (B), progression-free survival rate (C), and distant metastasis-free survival rate for patients with lacrimal sac squamous cell carcinoma treated by definitive intensity-modulated radiation therapy.

Although, the patients received multidisciplinary treatment including surgery, chemotherapy, and radiation therapy, the mortality rate, locoregional recurrence rate, and distant metastasis rate of lacrimal sac tumor is varied in different studies. The investigated results from eight studies inducing our cohort were shown in Table 2.^1,12–18

Table 2.

Eight studies reporting the treatment outcomes of lacrimal sac tumor.

Reference	PE	TM	RT	FT (m)	MR	RR	DMR
Ni et al.12	67	S, R, C	67/67	/	31.3%	13.4%	/
Parmaret al.13	15	S, R, C	9/15	2–204	13.3%	20%	20%
Valenzuela et al.14	11	S, R	4/11	6–84	18.2%	0	27.3%
Kang et al.15	10	S, R	4/10	3–239	20%	10%	20%
Montalbanet al.1	7	S, R, I	5/7	6–204	14%	28.6%	0
Skinner et al.16a	13	S, R, C	10/13	3–460	32.6%	39.1%	39.1%
Alabiadet al.17	14	S, R	9/14	9–149	14%	14%	7%
EI-Sawyet al.18	14	S, R, C	12/14	6–96	21.4%	28.6%	21.4%
Song et al.	17	R, C	17/17	7–106	17.6%	5.9%	11.8%

C, chemotherapy; DMR, distant metastasis rate; FT(m), follow-up time (months); I, immunotherapy; MR, mortality rate; PE, patients evaluated; R, radiotherapy; RR, recurrent rate; RT, radiation therapy; S, surgery; TM, treatment modality.

^athe data in lacrimalsac/duct referring to the 46 cases with tumors of the lacrimal apparatus

The planning results

A summary of the dose-volume histogram for an IMRT plan is presented, and the conformal avoidance of the optic nerves and eyes are readily visible in Figure 4D. Table 3 shows the irradiation doses of GTVs and CTVs to the primary tumor and lymph node are in all patients. The irradiation doses to the optic structures are summarized in Table 4. The contralateral optic structures received a lower dose than the optic structures ipsilateral to the tumor.

Figure 4.

Axial (A) and coronal (B) views of the definitive intensity-modulated radiation therapy plan of a patient with lacrimal sac squamous cell carcinoma showed 6791.8 cGy to be delivered to the primary tumor cavity. Sagittal (C) views of intensity-modulated radiation therapy plan showed 6791.8 cGy to the primary tumor and 6552.8 cGy to the cervical lymph nodes. Dose-volume histogram (D)

Table 3.

Dose (cGy) to primary tumor and lymph node area (n = 17)

	Max	Min	Mean
PGTV1	7069.7 ± 193.3	5733.4 ± 940.3	6761.7 ± 142.4
PGTV2	6729.3 ± 103.5	6214.0 ± 515.2	6469.9 ± 56.4
PCTV1	7102.0 ± 176.9	4165.2 ± 1351.7	6457.3 ± 203.8
PCTV2	7071.0 ± 147.0	3468.7 ± 1440.7	5968.1 ± 113.8

Table 4.

Dose (cGy) to critical structures (n = 17)

	Max	Min	Mean
Ipsilateral optic nerve	6075.3 ± 682.3	4107.8 ± 1239.0	5198.1 ± 919.7
Contralateral optic nerve	4698.6 ± 946.2	2747.0 ± 1298.0	3839.9 ± 898.9
Ipsilateral eye	6742.7 ± 329.8	2418.4 ± 1619.6	4747.3 ± 1108.2
Contralateral eye	4221.7 ± 933.5	1333.6 ± 702.5	2761.3 ± 988.7
Optic chiasm	4409.2 ± 1080.4	2409.6 ± 1074.3	3312.6 ± 1119.6
Affected side-lens	4604.0 ± 1465.8	2870.0 ± 1272.3	3462.5 ± 1425.5
Contralateral lens	2367.2 ± 1057.2	1752.8 ± 844.4	2002.2 ± 941.7

Lymph node drainage areas were covered in terms of the specific lymph node involvement. Although no cervical lymph was found before treatment, the levels Ib, VIIa, VIII, and IX should be covered as CTV2. The elective nodal irradiation for cervical lymph node drainages should be covered to the next station of positive lymph node if there were lymph nodes involvement. Typical countering and dose distributions for an IMRT plan are shown in Figure 4 (Figure A, B, C).

Acute and chronic toxicities

Acute toxicities occurred during RT and within the first 3 months after treatment completion. Acute Grade 3 radiation dermatitis occurred in seven patients (17.6%), but no acute Grade 4 or Grade 5 toxicities of any type were reported. No corneal ulcers occurred. Ten (58.8%) patients underwent acute conjunctivitis with symptoms of watering and discomfort.

All the patients in this study had dry eye syndrome at different degrees. The vision impairment levels were measured according to the criteria from a systematic review by Bourne et al.¹⁹ Four (23.5%) of the treated eyes had moderate vision impairment, and one of the treated eyes had severe vision impairment. Cataracts were the most common complication (3/17), followed by glaucoma (1/17) and radiation retinopathy (1/17). Two patients experienced epiphora due to nasolacrimal duct obstruction and underwent surgical placement of tubes to assist tear drainage.

Discussion

LSSCC is rare, and the etiology is currently unclear. EBV has been reported as a likely cause of undifferentiated lacrimal sac carcinomas,^20,21 which is very similar to the situation in nasopharyngeal carcinoma. Multidisciplinary therapy, including surgery, chemotherapy, and radiotherapy, is the primary treatment modality. Extensive surgical en bloc resection of lacrimal sac tumor with medial maxillectomy or total maxillectomy is favored with good success rates for local disease control. Orbit exenteration, resection of the paranasal sinuses, or lymph node dissection is performed in certain advanced cases.^14,15 Song et al.⁷ reported that the outcomes of comprehensive treatment were quite encouraging, and the 5-year overall survival (OS) rate and 5 year progression-free survival (PFS) rate were 87.6±4.8% and 76.3±6.4%, respectively. Radiation therapy plays a vital role in treatment, especially for patients who are unfit for or refuse multidisciplinary therapy. To the best of our knowledge, this is the first study to present definitive radiation therapy for the treatment of LSSCC.

Local control failure is the leading cause of LSSCC fatalities, and the 5-year local recurrence rate after surgery is approximately 50%; however, the definitive radiotherapy used in our study achieved a 5-year local control rate of 93.8%, with only two patients developing a local recurrence. Importantly, the eyeballs were well-preserved in all patients. The pathological types in our study were moderately to poorly differentiated squamous cell carcinoma in 88.2% patients, the majority of which were poorly differentiated type, which is known to be radiosensitive. In poorly differentiated lacrimal sac cancer, chemoradiotherapy alone may be a valid treatment option. In this study, we only performed definitive radiotherapy in those patients in whom complete resection of the primary tumor was not possible or who refused surgery. The mortality rate, recurrence rate, and distant metastasis rate are 17.6%, 5.9%, and 11.8%, respectively. Compared with the other eight studies in which patients were delivered multidisciplinary treatment, definitive radiotherapy in our cohort showed comparable results. However, the application of radiotherapy in other studies varied between 29.4 and 100%, and none of the cases was reported for radiotherapy alone. So, the role of radiotherapy in the treatment of the lacrimal tumors was not well defined. After the completion of radiotherapy and chemotherapy, we recommended patients to see an ophthalmologist’s assist in deciding whether further surgery or follow-up is warranted.

A characteristic feature of poorly differentiated squamous cell cancer is lymph nodes metastasis in the related locoregional area. There are three major lymph metastasis routes in LSSCC: 1) from the inner can thus to the check (level IX), and then to the submandibular triangle (level Ib); 2) from outside of the eye to the parotid gland (level VIII), and then to the submandibular triangle (level Ib); 3) from retropharyngeal (level VIIa) to the submandibular triangle (level Ib). Krishna et al.⁴ discussed how lacrimal sac tumors spread via the lymphatics to the preauricular, submandibular, or cervical lymph nodes in less than a third of the cases. In our cohort, 52.9% of patients had lymph node involvement in the head and neck, with the most common site being level II-IV (5/17, 47.1%), followed by level Ib (3/17, 17.6%), level VIIa (3/17, 17.6%), and levels VIII, IX(3/17, 17.6%). The difference in lymph node metastasis rate between our study and previous results might be due to more complete imaging, including high-resolution CT and MRI. MRI images are more favorable than CT scans to distinguish lymph nodes, and the imaging examination should include the head and the whole neck. The irradiation coverage of the lymph drainage area is essential to prevent locoregional lymph node recurrence. We decided which areas should be covered according to the typical route of lymph node metastasis, and the elective nodal irradiation range is down to the next stop of the last positive lymph node in clinical work. The average dosage of the lymph drainage area (CTV2) was 5968.1 ± 113.8 cGy in our cohort, and none of the patients developed locoregional lymph node metastasis, so prophylactic radiation therapy for lymph node drainage regions was shown to be effective. However, the dosage of CTV2 is too high for the ipsilateral parotid gland and submandibular gland and would severely damage gland function. The question is how to handle the lymph node area for those patients without lymph nodes metastasis while minimizing the radiation damage to the glands. In our experience, for the patient without nodal involvement, if the primary tumor was locally located in the lacrimal sac without skin involvement, we only covered level VIIa; if the primary tumor invaded the surrounding structures, especially to the skin, we suggested irradiation coverage of levels VIII, IX.

However, the cornea, lens, retina, optic nerve, and optic chiasm are sensitive to radiation. Severe complications will occur if these structures receive radiation dose beyond their tolerance. It is a big challenge to protect the organ at risk while ensuring the treatment efficacy of radiation therapy for patients with lacrimal sac tumors. Partial or total orbital irradiation may cause a broad spectrum of early and late toxicities, ranging from transient irritation side-effects to permanent blindness. One high risk of irradiation is the reduction of vision, and three patients in this cohort had different levels of visual loss. The loss of vision may occur when doses above 50 Gy are delivered to the optic nerve and retina, and at the doses ≥ 60 Gy, there is an increased risk of radiation-induced optic neuropathy^22,23. The retina dosage was not completely calculated in our treatment planning system, but the average irradiation dose of the affected side’s optic nerve was 5198.1 ± 919.7 cGy. Mayo et al.²⁴ reported that the optic nerve was relatively safe with a maximum irradiation dose <55 Gy, which is consistent with the results in our cohort. Seven of the ipsilateral eyes had moderate vision impairment relative to the pretreatment baseline, and no patients suffered from contralateral eye vision impairment.

Acute toxicity of radiotherapy is low and is limited to conjunctival or cutaneous hyperemia, and delayed toxicity leads to xerophthalmia, radiation retinopathy, glaucoma, cataract, or keratitis with corneal ulcerations.⁵ Weekly physical examination during radiation therapy can diagnose acute toxicities and handle them promptly. No patients in this study ceased or delayed radiation therapy because of severe acute toxicity.

Cataracts, glaucoma, and radiation retinopathy were the most common complications for patients treated with definitive radiation therapy, and all eyeballs were successfully preserved. Some scholars insist that exenteration does not improve prognosis and causes substantial facial disfigurement that may lead to severe psychological and psychiatric disorders.^1,25 Compared to exenteration, the above complications are much more acceptable.

Conclusions

In this study, we showed that radiation therapy alone achieved excellent long-term clinical outcomes, including OS, PFS, and LCR, and may be a viable treatment option for patients who refuse surgery or have unresectable tumors. The acute and delayed toxicities of radio(chemo)therapy were well-tolerated. However, more clinical data and prospective studies are warranted.