Boron neutron capture therapy for vulvar melanoma and genital extramammary Paget’s disease with curative responses

Although vulvar cancer is generally considered rare, it is the fourth most common gynecologic malignancy in the United States [1]. Vulvar melanoma (VM) is the second-most common type of cancer involving the vulva after squamous cell carcinoma and usually occurs in the fifth to seventh decades of life. VM accounts for 5%–10% of vulvar cancers and has an incidence of 0.2 per 100,000 women in the United States [2]. Mert et al. [3] have reported differences in clinicopathologic features and survival patterns between patients with vulvar/vaginal melanomas and those with cutaneous melanomas from the data of Surveillance Epidemiology and End Results (SEER). Included in the study were 762 patients with vulvar/vaginal melanomas and 55,485 patients with cutaneous melanoma. Twenty-eight patients of the vulvar/vaginal group and 334 patients of the cutaneous group were black (3.6% vs 0.6%, respectively). The median age at the time of diagnosis was 68 years in the vulvar/vaginal group and 52 years in the cutaneous group. Three hundred fifty patients (45.9%) in the vulvar/vaginal and 46,499 patients (83.8%) in the cutaneous group presented with localized disease. The median survival of the black patients was 16 months in the vulvar/vaginal group and 124 months in the cutaneous melanoma group. The median survival in the nonblack population was 39 months in the vulvar/vaginal group compared to 319 months in the cutaneous melanoma group. This study indicated significant differences in median age at diagnosis, racial distribution, and survival of women with vulvar/vaginal melanomas compared to those with cutaneous tumors.

Extramammary Paget’s disease (EMPD) is a rare, slow-growing cutaneous adenocarcinoma of apocrine gland-bearing skin. It presents as an erythematous, eczematous, hyperkeratotic plaque with occasional areas of hypopigmentation and superficial erosions. Histologic diagnosis of both mammary Paget’s disease and EMPD is based on the presence of large infiltrating, round intraepithelial malignant cells with glandular differentiation, which are distributed as individual cells or in clusters [4]. The most common sites of involvement are the vulva in women and scrotal and penile skin in men, and perineal and perianal areas in both men and women. EMPD of the genital region constitutes 1%–5% of all vulvar malignancies in woman, with a peak incidence at age 65 [4]. The incidence of these malignancies in populous Asian countries, such as China, India, and Japan, is unknown to our knowledge, but it is seen mostly in males in Western countries [5].

The most commonly recommended treatment for melanoma and EMPD of the genital region, including the vulva, penis, scrotum, and perianal area, is wide surgical excision of the lesion, with or without lymph node dissection, and reconstruction with a skin graft or a skin flap [6–8]. More recently neoadjuvant chemotherapy has been administered [9]. Although wide surgical excision has been the standard procedure, it is highly invasive, especially in older patients, and it can lead to a variety of functional and sexual problems that result in a poor quality of life [10, 11]. Alternative treatment modalities, such as topical chemotherapy [12], immunotherapy, carbon ion radiotherapy [13], and photodynamic therapy, have been administered for local control when wide local excision was not feasible. Fukuda and Funakoshi [14] recently reviewed current therapies for EMPD and concluded that current systemic chemotherapeutic regimens are not very effective. However, recent genomic analysis indicate increased frequency of mismatch repair mutations in patients with EMPD suggesting that these patients might be candidates for immunotherapy with anti-PD1 antibody [14]. The treatment of vulvar melanomas has been similarly unsatisfactory [15], making a strong case for new therapeutic approaches.

Boron neutron capture therapy (BNCT) is based on a nuclear reaction between the non-radioactive isotope boron-10 (¹⁰B) and either low energy thermal neutrons or higher energy epithermal neutrons (Fig. 1). These are captured by ¹⁰B atoms, resulting in the production of alpha (a) particles (⁴He) and lithium atoms (⁷Li) (Fig. 1a). The α particles have high linear energy transfer (LET) and very short path lengths (≤ 10 µm), which are approximately the diameter of a single tumor cell (Fig. 1b). If sufficient amounts of ¹⁰B atoms are selectively localized in tumor cells, the resulting ¹⁰B(n,α)⁷Li capture reaction can kill them and spare surrounding normal cells. Theoretically, BNCT is an ideal type of radiation therapy because it is both biologically and physically targeted and the structure and function of normal tissues are spared.

In 1972, experimental studies on BNCT for cutaneous melanoma were initiated by Yutaka Mishima at Kobe University in Japan, and his colleagues, who included physicists, chemists, radiation biologists, and physicians [16, 17]. In 1987, after 15 years of basic research, his team initiated the first clinical use of BNCT to treat a patient with a metastatic melanoma to the skin of the left occipital region of the scalp, using para-boronophenylalanine (BPA) as the boron delivery agent [18]. This was administered by perilesional injections of BPA-HCl, followed by thermal neutron irradiation, which resulted in complete regression of the tumor. This prompted Mishima et al. [19] to treat a patient with an acral melanoma on the sole of the right foot. BPA-fructose (BPA-F), which increased the water solubility of BPA [20], was injected perilesionally, followed by neutron irradiation. Again, there was complete regression of the tumor. Based on the Japanese clinical results, other reports followed that described the treatment of cutaneous melanomas using BNCT [21, 22].

Although BPA was developed as a boron delivery agent for BNCT of melanoma, it selectively accumulates in other types of malignant tumors. Coderre et al. [23] first reported that BPA was taken up by non-pigmented tumors, including a rat brain tumor, the 9 L gliosarcoma. This suggested that other uptake mechanisms existed that are independent of melanin synthesis, which might explain its tumor-localizing properties. The selective uptake of BPA most likely is due to increased l-type amino acid transport activity in tumor cells [24]. Shortly after the report of Coderre et al. BPA started to be used as a boron delivery agent for patients with brain tumors [25] and head and neck cancers [26]. It also has been evaluated for potential use in treating patients with mesotheliomas [27] and colon cancer metastatic to the liver [28]. Here, we report the results obtained using BNCT to treat one patient with VM and three with EMPD.

To our knowledge, the present report is the first describing the treatment of patients with VM and EMPD using BNCT.

Because VM and EMPD are very rare, no prospective, randomized, clinical trials have been performed to determine the effectiveness of various treatment options for these malignancies. At present, the most commonly recommended treatment is wide local excision of the lesion. However, this approach can be highly mutilating and can significantly diminish the quality of life. Effective alternative treatments for the primary tumor are therefore needed. High LET carbon ion radiotherapy is reportedly efficacious for cutaneous melanomas, with actual local control rates at 1 and 3 years of 85.7% and 42.9%, respectively [13]. However, no patients with EMPD and VM were included in this study [13]. Karasawa et al. [34] reported the outcomes of 23 patients with gynecologic melanomas treated with carbon ion radiotherapy. Fourteen of the melanomas were located in the vagina, 6 in the vulva, and 3 in the cervix. Total doses of 57.6 Gy-Eq in 16 fractions each were administered to 22 patients, and 64 Gy-Eq to one patient. Six patients showed complete responses, and partial responses were noted in 17 patients. Finally, the 3-year local control and overall survival rates were 49.9% and 53%, respectively; the authors concluded that carbon ion radiotherapy could be an acceptable alternative to surgery.

Surgery for EMPD of the genital region is limited from the prognostic point of view because of its multifocal nature and its frequent association with severe morbidity and functional impairment. Photon radiotherapy has been used in certain conditions, such as elderly patients who are medically unfit for surgery, or patients who refuse surgery, or as an alternative therapy for those with recurrence after repeated operations. Although optimal radiation doses have not been definitively determined, several authors [35, 36] have recommended 40–60 Gy. The morbidity associated with radiotherapy is minimal.

BNCT has three significant advantages over carbon ion therapy or conventional photon radiotherapy. First, although melanomas are generally considered to be resistant to conventional photon irradiation, they can be eradicated by BNCT, and both oxic and anoxic tumor cells are equally susceptible. Second, BNCT can be administered to relatively large areas, thereby allowing a wide margin, because the BPA selectively accumulates in both melanotic and non-melanotic tumor cells. These cells then are killed by the ¹⁰B(n,α)⁷Li capture reaction without significant damage to the surrounding normal tissue. In contrast with carbon ion radiotherapy, which has a Bragg peak, the dose in BNCT is uniformly delivered within the target volume. This advantage of BNCT is particularly useful for treating EMPD because histological involvement characteristically extends beyond the grossly visible lesion [37]. EMPD is usually multifocal with sub-clinical extension, which sometimes hinders the establishment of precise limits with normal skin. It has been reported that the surgical margin is positive despite wide excision with, grossly, a 2 cm margin. Biopsy examination to judge the margin cannot guarantee surgical margin negativity because the extent of histologic involvement is greater than that of the gross lesion [37]. Therefore, avoiding local recurrence following both surgery and radiotherapy requires adding a wide safety margin to the visible area [36, 37]. Third, BNCT is administered in a single fraction and high radiation doses can be delivered selectively to cancer cells. Hypothetically, doses as high as 60–80 Gy-Eq can be delivered to the malignant cells after the uptake and retention of BPA-F within approximately 1 h following its intravenous administration. In comparison, such radiation doses are usually administered over 6–7 weeks when conventionally fractionated photon radiation is used. This advantage enables patients to quickly resume their normal activities.

Our results are preliminary, as the patient cohort was very small. Furthermore, we based our dose calculations on data obtained from previous studies on cutaneous melanomas [30]. That is, we did not measure the boron concentration in each tumor because no easy way to do so directly was available, other than taking biopsies, followed by neutron irradiation to determine it by means of prompt gamma emission [31]. Nevertheless, we believe that the responses observed in tumor, skin, and mucosa following BNCT indicate that our calculations were accurate. Early and late responses, such as ulceration, necrosis, or residual tumor, did not occur during the 1.1–6.9 years of follow-up of the four patients. The most important goal of cancer treatment is to achieve tumor control while sparing surrounding normal tissue and preserve its function, and this was achieved in our patients with genital cancers.

Recent advances in immunotherapeutic approaches [38, 39] to treat metastatic melanoma combined with BNCT of the primary tumor might represent a breakthrough in treating this malignancy, which has a high propensity to metastasize. A Phase III clinical trial has shown that high-dose interferon-alpha (IFNα) can significantly increase overall survival [38]. A recent Phase III clinical trial also showed that cytotoxic T lymphocyte antigen-4 (CTLA-4) blockade with ipilimumab significantly improved overall survival [36]. Finally, targeting the programmed cell-death-1(PD-1) ligand with multiple anti-PD-1 monoclonal antibodies has been evaluated in Phase III trials and these also had impressive results [39]. Since BNCT spares normal cells, and more specifically immune effector cells, it may fit in ideally with immunotherapeutic approaches to treat VM. VM has an unfavorable prognosis and a relatively unpredictable biologic behavior and it tends to recur locally, and metastasize by hematogenous dissemination [3]. Thus, local BNCT and systemic immunotherapy are mutually complementary and potentially synergistic, because BNCT spares immune effector cells at the site of the tumor.

This is the first clinical report of treatment of patients with VM and EMPD by BNCT, which resulted in complete local tumor control. Our results suggest that BNCT is a promising treatment modality for VM and EMPD, which were heretofore considered to be radio- and chemotherapeutically resistant.