Intraoperative High Dose Rate ³²P Brachytherapy for Diffuse Recalcitrant Conjunctival Neoplasms: A Retrospective Case Series and Report of Toxicity

Abstract

Importance

Adjunct treatments for conjunctival malignancies are needed when standard therapy provides limited benefits or fails.

Objective

To describe the results of patients with diffuse conjunctival neoplasms treated with a radioactive phosphorus (³²P) impregnated flexible film.

Design

Retrospective case series between 2010 and 2013.

Setting

Memorial Sloan-Kettering Cancer Center – a tertiary referral center

Participants

The study was conducted on 7 eyes of 6 patients treated for diffuse conjunctival squamous cell carcinoma, sebaceous carcinoma, or lymphoma that had recurrent or residual disease after primary treatment.

Intervention

Patients underwent mapping biopsies and detailed conjunctival drawings to delineate the pathologic extent of the disease. The brachytherapy film used for treatment was the RIC Conformal Source Model 100 (RIC-100, R.I. Consultants, Hudson, NH). The RIC-100 is a flexible, thin (~0.5 mm) film made of a polymer chemically bound to ³²P. The radioactive ³²P film was placed intraoperatively, allowed to stay in place until the prescription dose was reached, and then removed. Median dose at the prescription point (1 mm from the surface of the film) was 15 Gy (range, 5–17 Gy).

Main Outcome Measures

Patients were tested for best-corrected visual acuity (BCVA), recurrence-free survival, and adverse events scored by using the adult comorbidity evaluation-27 (ACE-27) scale.

Results

Between 2010 and 2013, 7 eyes of 6 patients were treated. Median age was 70 years old. All patients had a recurrent or persistent neoplasm. Four patients with squamous cell carcinoma, 1 with sebaceous carcinoma, and 1 with metachronous bilateral lymphomas were treated. Median treatment time was 19 minutes (range, 10–52 minutes). Median follow-up was 24.9 months (range, 3.1–38.2 months). Recurrence-free survival 24 months after brachytherapy was 75% (95% CI 19–89.1%). Two moderate adverse events and one severe adverse event occurred. Vision was stable or improved in five of the seven eyes.

Conclusions and Relevance

Our results show the use of Intraoperative High Dose Rate ³²P Brachytherapy in selected cases of recalcitrant diffuse conjunctival neoplasms. This technique offers a novel adjunct in the treatment of these cancers. Further follow-up and study is warranted.

Trial Registration

N/A

Introduction

Diffuse conjunctival cancers that are non-resectable are difficult to cure. Eradicating the lesion often destroys the normal function of these structures. Loss of the ocular surface can lead to loss of vision or the globe. Current eye-sparing treatments include surgery, cryotherapy, immunotherapy, chemotherapy, and radiation therapy. Surgical options are limited by the extent and location of the disease; larger lesions make complete resection impractical or ineffective because of the extent of specialized tissue that needs to be reconstructed. In squamous cell carcinoma of the conjunctiva, reports have shown that more advanced (T2-T3) tumors, classified using the American Joint Committee on Cancer (AJCC) clinical staging system, have higher recurrence rates after surgery.¹ Topical chemotherapy and immunotherapy using mitomycin C (MMC), 5-fluorouracil (5FU), and alpha interferon 2b (IFN) are widely used, with good results for tumor control (74% MMC; 72–85% IFN; 57% 5FU); however, secondary complications and recurrences still occur.^2–5 For diffuse or advanced disease or for those patients who fail both surgical and topical therapy, radiation therapy has been used to avoid exenteration. Radiation techniques such as external beam radiation, proton beam therapy, and brachytherapy that have been used as adjuvant and definitive treatments for these diseases, however, carry associated complications. Here, we present our experience with brachytherapy using a unique radioactive phosphorus (³²P) impregnated flexible film for patients with recurrent diffuse conjunctival neoplasms. This technique allows a custom-sized, thin flexible film to be placed on the eye, delivering a full dose in one intraoperative treatment, providing a more localized delivery for select cases without the need for deeper orbital treatment.

Methods

This retrospective clinical research study was carried out with the permission of the institutional review board of Memorial Sloan-Kettering Cancer Center (MSKCC); the study was HIPAA compliant. Patients with ophthalmic neoplasms treated with ³²P were identified in the brachytherapy treatment-planning database. In all cases, a pathologist at our center confirmed the histopathologic diagnosis. Comorbidities were classified and scored according to the adult comorbidity evaluation-27 (ACE-27) scale. After treatment, patients were evaluated every 3–6 months to assess for local control and adverse events.

The brachytherapy film used for treatment was the RIC Conformal Source Model 100 (RIC-100, R.I. Consultants, Hudson, NH). The RIC-100 is a flexible, thin (~0.5 mm) film made of a polymer chemically bound to ³²P. ³²P is an exclusive beta emitter (<0.05% gamma emission) with maximum energy of 1.71 MeV, average energy of 0.695 MeV, and half-life of 14.26 days. The maximum range is approximately 7 mm in water or biological tissues. The RIC-100 source has been approved by the US Food and Drug Administration as a single-use device and the cost at our center is just over $ 6000 for the source.

An ophthalmic oncologist and radiation oncologist evaluated patients jointly. Patients underwent mapping biopsies, as described by Putterman,⁶ and detailed conjunctival drawings to delineate the pathologic extent of the disease. Cases selected for therapy had thin non-nodular tumors that were diffuse, covering a significant portion of the ocular and tarsal surface, or had multifocal or scleral recurrences. The gross tumor volume (GTV) consisted of clinically apparent neoplasm, and the clinical target volume (CTV) was the region adjacent to the GTV identified as abnormal by mapping biopsies and a 2-mm margin; this included the region proximal to the GTV in the case of negative mapping biopsies. Under topical anesthetic, a flexible and transparent “dummy” film was cut to a custom size that encompassed the surface area and shape of the CTV, ensuring direct contact of the film on the surface of the conjunctiva (Figure 1. A, B). A print of the “dummy” film was then sent to the supplier (R.I. Consultants, Hudson, NH) for fabrication of the radioactive ³²P film. After fabrication, the ³²P film was shipped to our institution. Prior to use, the radiation dose rate away from the film was measured and used to determine the duration of time necessary to deliver the prescribed total radiation dose at the prescription point (1 mm from the surface of the film). Patients were brought to the operating room and general anesthesia was administered. Under sterile technique, the GTV and CTV were identified by the ophthalmic oncologist and radiation oncologist. The sterilized “dummy” film was positioned to confirm stability of the film in situ. The film was created to fit snuggly in the fornices that secured it in position, similar to a scleral shell or conformer. Patient #4 required a limbal mattress suture to be preplaced to secure the smaller sized film. After confirming acceptable positioning of the “dummy” film, attention was turned to the sterilized radioactive ³²P film. The radioactive source was carefully removed from the shielded carrier with long smooth forceps and handled behind a 5-mm acrylic shield, avoiding abrasion, cutting, or grinding of the film. A wipe test was performed on the ³²P film using a sterile cotton-tipped swab and analyzed in the operating room with a thin-windowed Geiger Mueller (GM) radiation detector. Upon confirming no residual radioactivity on the source, it was placed over the GTV and CTV (Figure 1. C, D). After secure placement was confirmed, the eye was covered with wet towels, and operating room personnel increased their distance from the source to minimize radiation exposure. The film was left in place for the preplanned amount of time to deliver the prescription radiation dose. After completion of brachytherapy (removal of the ³²P film), the patient was surveyed for residual radioactivity in the operating room with a thin-windowed GM radiation detector. Occupational radiation dose measurements of the extremities were performed for the operating ophthalmologist and radiation oncologist using sterilized thermoluminescent dosimeters (TLDs). The source characterization, in-hospital quality assurance procedures, as well as discussion of the clinical implementation and radiation safety precautions are described in detail by Cohen et al.⁷

Figure 1.

A, The flexible “dummy” ³²P film shown after custom sizing prior to ordering the live ³²P film. B, The “dummy” ³²P film in position, confirming proper fit and coverage. C, The active ³²P shown ready for placement; note the soft-tipped long forceps to avoid abrasion or ripping of the film. D, The intraoperative placement of the ³²P film.

Medical records were reviewed for ophthalmic adverse events, as defined by the Common Terminology and Criteria for Adverse Events (version 4.02). Visual acuity and intraocular pressure were assessed at each clinic visit with the ophthalmic oncologist, and were collected for study.

Recurrence-free survival was calculated by the Kaplan-Meier method Graph generation was performed using GraphPad Prism 6.

Results

Between 2010 and 2013, 7 eyes of 6 patients were identified and deemed eligible for study. Table 1 presents patient demographics, disease, treatment history, and comorbidities at the time of brachytherapy. Median age was 70 years (range, 51–80). All patients had a recurrent neoplasm. Most patients had squamous cell carcinoma of the conjunctiva that had recurred after a median of 3 prior therapies (range, 1–5).

Table 1.

Patient demographics, disease, treatment history, and comorbidities at the time of brachytherapy

Subject	Age (years)	Gender	Location	Disease	Stage	Prior treatment(s)	ACE-27 score
1	56	M	Left conjunctiva	SCC	T3N0M0	Excision, Cryo, MMC, IFN	1
2	51	F	Left conjunctiva	T-cell lymphoma	T2N0M0	Excision	2
2	52	F	Right conjunctiva	T-cell lymphoma	T2N0M0	Excision, Cryo	2
3	63	F	Right eyelid and conjunctiva	Sebaceous carcinoma	T3N0M0	Excision, Cryo	2
4	76	M	Left conjunctiva	SCC	T3N0M0	Excision, Cryo	1
5	80	M	Left conjunctiva	SCC	T3N0M0	Excision, Cryo, 5-FU, IFN	1
6	79	M	Left conjunctiva	SCC	T3N0M0	Excision, Cryo, MMC	1

ACE-27, adult comorbidity evaluation-27; M, male; F, female; SCC, squamous cell carcinoma; MMC, topical mitomycin-C; IFN, topical interferon; Cryo, cryotherapy; 5-FU, topical 5-fluorouracil

Table 2 displays parameters of each course of brachytherapy. Median dose at the prescription point (1 mm from the surface of the film) was 15 Gy (range, 5–17). Median dose rate at the prescription point was 29 Gy/hour (range, 20–64). Median treatment time was 19 minutes (range, 10–52). Median film activity was 222 MBq (range, 50–277.5). Median film area was 5 cm² (range, 1–6.7). In some cases, radioactivity levels greater than the background count rate (50 cpm) were detected upon removal of the ³²P film, as detected with the thin-windowed GM radiation detector in close proximity to the treatment site. After copious irrigation, levels returned to background cpm. Occupational radiation dose measurements of the extremities for the radiation oncologist and the ophthalmic oncologist did not exceed background radiation levels in 4 of the 7 cases. In cases in which dose exceeded background radiation levels, the median net was 0.14 mSv (range, 0.014–0.04). No staff received excessive radiation doses.

Table 2.

Brachytherapy details, follow-up, and treatment response

Subject	Dose (Gy)*	Dose rate (Gy/h)*	Treatment duration (minutes)	Activity (mCi)	Film area (cm2)	Follow-up (months)	Local recurrence
1	15	25	36	5.8	5	37.9	No
2	5	29	10	5.7	5.46	31.3	No
2	5	21	14	4.42	4.5	23.4	No
3	17	20	52	5.5	6.7	30.2	Yes
4	15	31	30	1.35	1	7.7	No
5	15	46	19	6	3.2	16.7	No
6	15	64	14	7.5	3.1	3.1	No

^*at prescription point, 1 mm from film surface

Median follow-up was 24.9 months (range, 3.1–38.2). Recurrence-free survival 24 months after brachytherapy was 75%(95% CI 19–89.1%) (Figures 2) and was the same if we excluded patients with less than 24 months of follow up. One patient experienced recurrence at eyelid margins and medial canthus as well as a perforated bacterial corneal ulcer 4.2 months after brachytherapy, which necessitated salvage orbital exenteration. This patient has been well controlled, with no evidence of disease recurrence since exenteration. No patients have developed scleral necrosis or metastases.

Figure 2. Clinical photos before and after high dose rate ³²P brachytherapy.

A, Pre-high dose rate ³²P brachytherapy clinical photograph of Subject #1 showing diffuse squamous carcinoma involvement of the inferior fornix and bulbar conjunctiva with extension onto the inferior cornea. B, Subject #1 more than 3 years after ³²P brachytherapy showing no recurrence. C, Pre-high dose rate ³²P brachytherapy clinical photograph of Subject #4 showing recurrent squamous carcinoma involvement arising from the deep cornea and sclera. D, Subject #4 7 months after ³²P brachytherapy showing resolution of the lesion without residual corneal or scleral disease.

Table 3 presents the ophthalmic adverse events observed. Two moderate and one severe adverse event occurred. The former events were cataract development, which required surgery 29.8 and 7.6 months, respectively, after brachytherapy. The latter event was an infectious corneal ulcer and perforation, which developed concurrently with disease recurrence. This event necessitated orbital exenteration.

Table 3.

Adverse events after brachytherapy

Adverse event	Grade 1*	Grade 2*	Grade 3*	Grade 4*
Watering eyes	1, 0.6
Conjunctivitis	1, 0.6	2, 0.5
Cataracts			1, 29.8; 5, 7.6
Dry eye	2, 1.2
Blurred vision	2, 20.1
Eye pain	2, 10.7; 4, 1.4
Corneal ulcer				4, 0.5

^*subject number, months after brachytherapy

Visual acuity was better than 20/70 in the 5 patients who retained their treated eye. No patients developed glaucoma in their retained, treated eyes.. Four patients noted a decline (not worse than 20/200) in visual acuity soon after brachytherapy, but this improved and stabilized in all retained eyes 2–3 months after treatment.

Discussion

Radiation has been used extensively for tumors in the head and neck region as well as eyelid neoplasms.^{8, 9} Beta radiation for the eye has been used since the 1930s when first reported by Moore using radon seeds for intraocular tumors.¹⁰ Since then, brachytherapy has been widely used for intraocular tumors and to a lesser extent conjunctival tumors.^11–14 In a case series by Jones et al., eyes with conjunctival intraepithelial neoplasia (CIN) were treated with a strontium-90 applicator to ocular surface lesions using two fractions of 1200–1524 cGy, totaling 2400–3050 cGy. There were no recurrences of the CIN in the 4 cases over a 3-year follow–up, and the only reported complication was progression of cataract.¹⁴ Another study, using beta radiation for conjunctival lymphoma using a strontium-90–yttrium-90 applicator, had an 87% control rate using 4000–8000 cGy divided into 500–2000 cGy weekly fractions. Their complications included cataracts in 15%, and conjunctivitis and keratitis in 25%. ¹³ The higher doses were associated with a greater incidence of side effects. Photon, proton, and electron beam radiation therapy have also been used for adjuvant and definitive treatment of conjunctival neoplasms. In a similar patient population to our study, external beam radiation was used to treat conjunctival squamous carcinoma. A control rate of 75% with doses of 50 to 60 Gy in 2.0- to 2.5-Gy daily fractions was obtained.¹⁵ In the one reported case of invasive conjunctival squamous carcinoma treated with proton beam radiation therapy, a dose of 3200 cGy was given in 4 fractions with local control for 9 months.¹⁶

The side effects of radiation treatment of the eye have been well described and include cataract, dry eye syndrome, conjunctivitis, keratitis, glaucoma, scarring of the conjunctiva and cornea, episcleral telangiectasia or avascularity, conjunctival and scleral atrophy, keratinization of the conjunctiva, and poor wound healing. Rates of cataract with beta radiation range from 21% to 45% and were found to be dose and location dependent; and with the strontium-90 applicator, a dose over 5000 cGy was thought to cause cataracts.^{11, 14,17}

The use of intraoperative ³²P high dose rate brachytherapy has been reported for use in brain and spinal tumors.¹⁸ This technique allowed an intraoperative adjunct to surgical resection or treatment of otherwise non-resectable lesions. Because of the film’s flexible nature, it conforms to specific areas during surgery and allows focal superficial treatment without damage to the underlying deep tissues. Local control rates of up to 80% have been obtained with this technique, even in lesions previously treated with radiation.¹⁸ Conjunctival lesions require the dose at the surface and not to the adjacent, deeper, more sensitive tissues as with spinal lesions. The ability of the material to bend and its limited depth of penetration make it an ideal tool for treatment of diffuse conjunctival neoplasms without orbital disease. Another advantage of the technique is that it is completed in a single treatment, averaging 19 minutes, in a controlled operating room environment. The radiation doses chosen for treatment were based on the presumed radiosensitivity of the neoplasm harbored by the patient (lymphoma being most radiosensitive and requiring the lowest dose, and sebaceous carcinoma being the most radioresistant and requiring the highest dose). In addition, the safety profile of other beta-emitting isotopes frequently used on the conjunctiva (i.e., ⁹⁰Sr for pterygia) was considered when selecting the dose. A prospective clinical trial established that 25 Gy delivered at a high dose rate in a single treatment for pterygia using ⁹⁰Sr yields no significant complications.¹⁹ The estimated median dose to the surface of the conjunctiva with the presently reported ³²P brachytherapy technique is estimated to be 255% of the prescription dose 1 mm from the surface of the applicator, or 38.25 Gy for a 15 Gy prescription dose. The rapid radiation dose fall off derived from beta particles is likely responsible for the tolerability of this technique, but further studies are necessary to determine the optimal dose.

All of our cases were referred to us after having a biopsy or prior therapy elsewhere, and had recurrent or residual disease. Only one eye, the second eye of the patient with lymphoma, had ³²P treatment immediately after the initial biopsy because of the favorable results of the opposite eye. The majority (6/7) of the eyes had failed multiple topical or surgical treatments before being treated with the ³²P brachytherapy. All patients remain free of disease at last follow-up. The one case that recurred had advanced pagetoid spread of sebaceous carcinoma with involvement of 100% of the bulbar and tarsal conjunctiva. The nodular component was resected from the upper eyelid at the time of map biopsy, and ³²P plaque was placed to treat the entire conjunctiva. The patient developed a bacterial corneal ulcer with perforation. The patient was treated for the ulcer, and repeat map biopsies showed recurrent disease at the eyelid margins and caruncle. It was decided at that time to proceed with an exenteration, and the patient has been disease free without systemic spread. It is possible that one reason for recurrence was the type of cancer, since all other successful cases were not sebaceous carcinoma. Another reason may be related to the large GTV, in that possible occult eyelid margin disease and caruncle disease received less of a dose due to their location being further from the radiation source.

Our treatment side effects included two grade 1 ocular surface symptoms and two grade 3 cataracts that required treatment following brachytherapy; however, both of these patients had significant cataracts that were not treated prior to the ³²P brachytherapy because of the underlying active neoplasm. The grade 4 side effect was discussed above with the case of recurrence (Table 3). The side effects were comparable to those of the above-referenced series. Despite some of the adverse events, final vision remained equal or better than 20/40 in 5 of the 7 eyes.

Our results show the use of intraoperative high dose rate ³²P brachytherapy in selected cases of diffuse conjunctival neoplasms. The technique was used for diffuse neoplasia in contrast to thick nodular disease or cases with suspected orbital disease and showed its best preliminary results with limited complications in cases of squamous cell carcinoma. This technique offers an additional, novel approach for treatment of select diffuse conjunctival neoplasms. The study is limited by its retrospective nature, limited follow up in some cases and small number of subjects which resulted in wide confidence intervals regarding results presented. Additional cases and continued follow-up of this technique needs to be explored, along with the investigation of other potential applications.