Skin-directed radiotherapy for primary cutaneous T-cell lymphomas

Article information

Radiat Oncol J. 2024;42(3):228-236

Publication date (electronic) : 2024 September 24

doi : https://doi.org/10.3857/roj.2024.00444

Ha Un Kim¹, Yeon Joo Kim¹, Mi Woo Lee², Woo Jin Lee², Sang-wook Lee¹, Youngju Song¹, Byungchul Cho¹, Si Yeol Song^,1

¹Department of Radiation Oncology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Republic of Korea

²Department of Dermatology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Republic of Korea

Correspondence: Si Yeol Song Department of Radiation Oncology, Asan Medical Center, University of Ulsan College of Medicine, 88 Olympic-ro 43-gil, Songpa-gu, Seoul 05505, Republic of Korea. Tel: +82-2-3010-4431 E-mail: coocoori@amc.seoul.kr

Received 2024 July 4; Revised 2024 August 7; Accepted 2024 August 14.

Abstract

Introduction

Primary cutaneous T-cell lymphoma (CTCL) is a rare neoplasm characterized by a heterogeneous group of T-cell lymphoproliferative neoplasm localized to the skin. The incidence of CTCL is 10 per 1 million persons, according to the Surveillance, Epidemiology, and End Results program [1]. Mycosis fungoides (MF) is the most common type of CTCL, accounting for 54% of CTCL diagnoses from 2001 to 2005 [2]. According to the 2016 revision of the World Health Organization classification, CTCL other than MF include Sezary syndrome primary cutaneous CD30+ lymphoproliferative disorders, subcutaneous panniculitis-like T-cell lymphoma (SPTL), extranodal NK/T-cell lymphoma (ENKTL) nasal type, adult T-cell lymphoma, and primary cutaneous peripheral T-cell lymphoma unspecified [3].

CTCL treatments include topical corticosteroids (TCS) or alkylating agents, phototherapy, and radiotherapy (RT). RT is a highly effective curative and palliative treatment for both localized and systemic lesions in patients with CTCL. For generalized lesions, total skin electron beam therapy (TSEBT) can successfully deliver superficial doses to the skin. The guidelines for field and dose were introduced by the International Lymphoma Radiation Oncology Group (ILROG) [4]. Since MF constitutes most of CTCL incidences, dose guidelines for CTCL other than MF are usually based on the limited published data and clinical experience that report the treatment outcomes due to the rarity of the disease.

In Korea, MF accounts for approximately 23%−43% of all CTCL cases, somewhat lower than the 54% observed in Western countries [5,6], and the incidences of NKTL and SPTL are more common than those reported in Western countries [7]. Based out of a large tertiary medical center in Korea, we aimed to evaluate the efficacy and toxicities of skin-directed RT in patients with various CTCL subtypes.

Materials and Methods

We retrospectively reviewed the medical records of patients who underwent RT at Asan Medical Center from January 2000 to December 2022. Only patients with confirmed histopathologic diagnosis of primary CTCL were included. Patients with a diagnosis other than CTCL or missing data were excluded. This study was approved by the Institutional Review Board of Asan Medical Center (IRB No. 2024-0806). Informed consent was waived based on the retrospective nature of this study. A total of 42 patients (57 lesions) were included for analysis. The 57 lesions were analyzed in terms of histology, stage, treatment technique, and response to RT. Based on histology, 57 lesions were classified into two categories: MF and non-MF. The non-MF group included all CTCL subtypes except MF. Early and advanced stages were defined differently in the MF and non-MF groups. In the non-MF group, the early stage was defined as T1−2, and T3 was considered as an advanced stage. In the MF group, T1 was defined as early stage, while T2 or higher was defined as advanced stage. T staging in the MF and non-MF groups followed the classification recommended by the International Society for Cutaneous Lymphomas (ISCL) and the European Organization for Research and Treatment of Cancer [8,9].

Patients in the early stage group presented only early-stage lesions. When multiple lesions were present, each underwent individual biopsy for pathological confirmation. For the advanced stage group receiving local RT, pathological confirmation was obtained for 70.8% of lesions (17 out of 24). The remaining seven lesions, while not biopsied, underwent comprehensive clinical assessment. This evaluation involved detailed patient interviews and meticulous physical examinations by experienced dermatologists, including precise mapping of skin involvement. This thorough clinical approach ensured accurate diagnosis of CTCL, even for lesions that did not undergo biopsy.

Two groups were formed, based on the treatment modality: TSEBT and local RT. TSEBT was defined as a course of treatment intended to treat the entire skin surface. The Stanford technique was used for all courses of TSEBT. It comprises a six dual field technique with the six positions spaced at 60°: anterior, posterior, right anterior oblique, right posterior oblique, left anterior oblique, and left posterior oblique. A scattering panel was used to achieve even distribution at the skin surface. Devices for shielding nails and eyes were applied. The optical stimulated luminescent dosimeters were taped to patients' extremities and bodies and measured. Two different fractionation schedules for TSEBT were utilized. The majority of patients (9 out of 11, or 81.8%) received 2 Gy per fraction, with a total of 2 fractions administered per week. The remaining patients (2 out of 11, or 18.2%) were treated with 1.2 Gy per fraction, with a total of 4 fractions administered per week. In our institution, boosts to specific regions such as the perineum, vertex, soles, and epaulets were not routinely administered during TSEBT, as patients received individualized maintenance therapy following TSEBT on a case-by-case basis. Based on each patient's specific condition and disease characteristics, boost treatments were administered at the discretion of clinicians. Specifically, 8 out of 11 patients (72.7%) in the TSEBT group received boost doses to areas including the fingers, vertex, anus, inguinal regions, soles, and hips.

Local RT was defined as limited radiation to a single lesion or group of adjacent lesions. RT planning included conventional two-dimensional (2D) and intensity-modulated RT (IMRT). A 2D-RT was administered using the following modalities: megavoltage photons and electrons with conventional blocks or intra-operative radiation therapy (IORT) cones. IORT cones are RT applicators used to irradiate superficial lesions. In the linear accelerator room where the patient was positioned, the appropriate type of IORT cone was selected based on the size and location of lesions, as shown in Fig. 1. Direct visualization of lesions was possible due to the transparent walls of the IORT cones. Then, the IORT cone was docked to the treatment head of the linear accelerator instead of conventional blocks.

Fig. 1.

Simulation with an intra-operative radiotherapy cone.

Additionally, a three dimensional (3D)-printed bolus was used to treat lesions much smaller than the smallest size of IORT cones that we had or to increase the surface dose to certain millimeters (mm). For both electrons with conventional blocks and IORT cones, margins of about 1−3 cm were added beyond lesions to cover any microscopic disease. The choice of treatment modalities was at the discretion of radiation oncologists, considering the size and depth of lesions, anatomic relationships, and geometric situations.

For IMRT treatment planning, a computed tomography (CT) scan with a slice thickness of 2.5 mm was used. During CT simulation, a non-metallic skin marker was employed to delineate the target lesion.

The response to treatment was assessed at every follow-up visit according to the ISCL criteria [10]. Complete response (CR) was defined as 100% clearance of skin lesions. Partial response (PR) was defined as 50%−99% clearance of skin disease from baseline. Stable disease was defined as less than 25% increase to less than 50% clearance in skin disease from baseline. Progressive disease was defined as greater than 25% or equal to 25% from baseline. The interval from RT completion to the best response was calculated. Local progression was specifically defined as recurrence within the radiation field. Acute toxicity was evaluated according to the Common Terminology Criteria for Adverse Events (version 5.0). Local progression probability was calculated from the initiation of RT to local recurrence. Survival curves for local progression probability were depicted using the Kaplan−Meier method, and log-rank tests were performed to compare the local progression probability across different dose groups. All statistical analyses were performed using STATA software package version 18 (StataCorp, College Station, TX, USA).

Results

Fifty-two lesions were classified into three groups: early stage with local RT, advanced stage with local RT, and advanced stage with TSEBT. The baseline characteristics of the lesions are summarized in Table 1. The median age for each group was 47, 56, and 58 years, respectively. In the advanced stage with local RT group, 17 lesions were T3–4, while seven lesions were T1–2. Regarding the previous history of treatments, 20 lesions (91.0%) in the early stage with local RT were treatment-naïve, while most in the TSEBT group (90.9%) already received treatments other than RT. For both the early and advanced stages with local RT, the most common sites of lesions were the head and neck. The next most common sites of lesions were extremities and torso in the early and advanced stages, respectively. Based on pathology, 55% in the early stage with local RT were diagnosed with CD4+ primary cutaneous small/medium T-cell lymphoproliferative disorder (PCSM-TLPD). The next most common pathology was anaplastic large cell lymphoma (ALCL). In the advanced stage with local RT and TSEBT, most cases were MF.

Table 1.

Lesion and treatment characteristics

The median doses in the early stage with local RT, the advanced stage with local RT, and the advanced stage with TSEBT were 30.6 Gy, 27 Gy, and 32 Gy, respectively. Electron beam energies ranged between 4−9 MeV with conventional blocks used in 33 lesions. Electron beam energy of 4 MeV with an IORT cone was used in 12 lesions. A 6 MeV energy was used in all patients who received TSEBT.

The median interval between the completion of RT and the best response achievement was 3.5 months (range, 2.0 to 18.1 months). Twenty-one lesions (95.5%) in the early stage with local RT and 17 lesions (70.8%) in the advanced stage with local RT achieved CR. In the advanced stage of TSEBT, 10 patients (90.9%) had partial remission (Table 2). Fig. 2 demonstrates an example of successfully treated lesions in the early stage with local RT. An 8-year-old girl presented with an erythematous patch on her right cheek and was diagnosed with stage I PCSM-TLPD (Fig. 2A). The lesion was treatment-naïve, and she was referred for RT. The IORT cone with a diameter of 1.9 cm and 45° was chosen to minimize the effect of RT on critical structures such as the eyes and nose. Precisely 8 months after she received 25.2 Gy in 14 fractions, total clearance of the PCSM-TLPD lesion was observed without any acute or chronic complications.

Table 2.

Best overall response and recurrence sites

Fig. 2.

An 8-year-old girl initially presented with an erythematous patch on her right cheek (A). Total clearance of the lesion 8 months after the treatment (B).

Recurrence at irradiated sites was observed in the advanced stage (Table 2). Specifically, seven local recurrences (29.2%) occurred in the local RT group, while seven patients (63.6%) in the TSEBT group experienced local failure. Notably, all recurrences were observed in lesions and patients with MF. In the local RT group, the best overall responses before recurrence were CR in three lesions and PR in four lesions. Conversely, in the TSEBT group, all seven local recurrences showed PR before recurrence. The median follow-up was 24.5 months (range, 10 to 136 months) for the local RT with advanced stage group and 34 months (range, 4 to 120 months) for the TSEBT group. The cumulative incidence of local progression for the local RT group with advanced stage at 6 months was 4.2%, and at 1 year, 20.8%. The cumulative incidence of local progression for the TSEBT group was 9.1% at 6 months, and 62.1% at 1 year (Fig. 3).

Fig. 3.

Local progression probability in (A) local radiotherapy with advanced stage and (B) total skin electron beam therapy group.

We analyzed the local progression rates among the three groups based on the presence or absence of previous treatment history. No statistically significant differences were observed in the local progression rates between patients and lesions with and without prior treatment history within each group (p > 0.05 for all comparisons).

Despite maintenance therapy with TCS, chemotherapy, and phototherapy following RT completion, seven lesions and seven patients experienced recurrence. Among the seven lesions with local recurrences in the advanced stage within the local RT group, five received re-irradiation (re-RT), predominantly TSEBT (60%), while the remaining two received chemotherapy. In the advanced stage with TSEBT, five of seven patients (71.4%) with local recurrences underwent re-RT, including both local RT and TSEBT. The remaining two patients received phototherapy and chemotherapy, respectively. In the advanced stage with local RT, five out of seven recurrent lesions received re-RT, with doses ranging from 20 Gy to 30 Gy delivered in 10–15 fractions. In the TSEBT group, five out of seven patients with recurrences received re-RT to multiple sites, with doses ranging from 16 Gy to 36 Gy delivered in 8–18 fractions. Of the 14 recurrences, 10 lesions were treated with additional RT, resulting in nine lesions achieving PR and one lesion achieving CR.

We analyzed the dose-response relationship in MF as presented in Fig. 4. We found that there was no significant association in local progression rates between <30 Gy and ≥30 Gy in the TSEBT group (Fig. 3A). Similarly, this relationship was observed in the local RT group with MF (Fig. 3B). For other pathologies, the comparison using log-rank test was not feasible owing to a small number of patients. A median dose of 35 Gy was used in nine lesions diagnosed with ALCL, and all achieved CR. Regardless of stages, there was no local progression. Two lesions (3.5%) diagnosed with SPTL were treated with an RT dose of 49 Gy, and they all attained CR. Three lesions (5.3%) diagnosed with ENKTL were treated with a dose range of 30−49.2 Gy. They all showed CRs with no observed progression (Supplementary Table S1).

Fig. 4.

Local progression rate for (A) local RT group in MF and (B) TSEBT group stratified by RT dose. RT, radiotherapy; TSEBT, total skin electron beam therapy; MF, mycosis fungoides.

The acute toxicities of RT are detailed in Table 3. The acute radiation-induced side effects were grade 1 or 2 and included dermatitis (n = 18; 31.6%), dry skin (n = 2; 3.5%), hyperkeratosis (n = 1; 1.8%), blurred vision (n = 1; 1.8%), and fatigue (n = 1; 1.8%). There were no reports of grade 3 or higher acute toxicities and grade 2 or higher chronic toxicities.

Table 3.

Acute adverse events

Discussion and Conclusion

In this study, skin-directed RT was highly effective in both early and advanced stages. We noted very high response rates to RT in the early (100%) and advanced stage (100%) regardless of local RT or TSEBT. Local progression developed in 29.2% of the local RT with advanced stage and in 63.6% of the TSEBT group, while no progression was observed in the local RT with early stage group. Since all recurrences were observed in MF cases, we analyzed the dose-response relationship in MF based on the treatment modality. We found that there was no significant association in local progression between <30 Gy and ≥30 Gy. The treatment-related acute toxicities were observed in 40% of lesions, with 78% of them being grade 1. The rates of dermatitis were higher in the TSEBT group than in the local RT group, potentially because the entire skin surface was treated.

CTCL is a group of several disease entities, and RT for each entity is usually reported in case reports or retrospective studies. Treatments for CTCL are based on the clinical condition and tumor stage of the patient. The most used skin-directed treatments include topical steroids, phototherapy, and RT. TCS are widely prescribed for CTCL. Despite their widespread usage, there remains a paucity of reported outcomes. Zackheim et al. [11] conducted the largest prospective study to date, reporting CR rates of 63% in the T1 stage and 25% in the T2 stage of the disease, with PR rates of 31% in T1 and 57% in T2. Upon discontinuation of TCS therapy, only 37% of T1 and 18% of T2 patients remained in complete remission over a median follow-up period of 9 months. The drawbacks of TCS include systemic absorption, leading to cortisol elevation, and application to large surface areas in the systemic stage. Phototherapy is an effective treatment with a durable response. Narrowband ultraviolet B (UVB) is typically used for the early stage of CTCL, yielding 81% and 71% CR rates, respectively [12]. Due to its deeper dermal penetration compared to UVB, photochemotherapy (PUVA) is recommended for thicker plaques, while UVB is advised for patches or thin plaques. Adjuvant interferon or retinoids following PUVA therapy have been reported to extend relapse-free survival [13,14]. Unfortunately, PUVA is currently unavailable in Korea. Given the limitations of phototherapy in penetrating certain depths and the propensity for CTCL located on the head and neck or the palmoplantar area to exhibit more aggressive behavior, RT can be considered a promising treatment option.

Since the early 1900s, with its initial documentation by Scholtz [15], local RT has consistently demonstrated an objective response rate (ORR) exceeding 90% [16-18]. We found that the CR rates were 95.5% in the local RT group with early stage disease, compared to 70.8% in the local RT group with advanced stage disease. The lower CR rates in the advanced stage group could be attributed to the fact that all seven lesions that did not show CR were MF and carried high disease burden. Furthermore, these lesions were refractory to other skin-directed therapies. Previous studies have reported ORRs of approximately 98% across various TSEBT dose regimens [19]. Our study observed an ORR of 100% in the TSEBT group. The utilization of a higher dose, with a median of 32 Gy, could account for this since the higher response rates are associated with doses exceeding 30 Gy [20]. In this study, among the 21 lesions in the head and neck region, the CR rate was 85%, with only one instance of local progression observed. Hence, RT was deemed effective for these lesions.

For decades, several studies have raised the interest of low-dose RT in MF. Neelis et al. [21] analyzed 65 lesions of MF treated with local RT and demonstrated a CR rate of 92% with 16 Gy in 2 fractions. Similarly, Thomas et al. [22] evaluated 58 patients with MF treated with 7−8 Gy in a single fraction and reported a CR rate of 94.4% in 270 lesions. Furthermore, several studies have shown that a lower total dose in TSEBT was associated with fewer short-term complications and provided opportunities for re-RT. Lower doses of TSEBT with <30 Gy yielded comparable ORR (95%) to that of the standard dose (≥30 Gy) [20]. As depicted in Fig. 4, our study showed that there was no significant difference in local progression rates between <30 Gy and ≥30 Gy in TSEBT and local RT groups. Since CTCL with a high disease burden tends to relapse, lower dose TSEBT has the advantages of shorter treatment time, better ability to re-treat with RT, and equivalent response.

CTCL excluding MF comprised 50.8% of total lesions in our study. Based on the reviews of published data, for ALCL, a dose between 24 Gy and 50 Gy is known to achieve a durable complete remission, according to the ILROG [23,24]. Although there is little data on radiation dose for SPTL, the recommended RT dose for SPTL based on the ILROG is usually ≥40 Gy [4]. Similarly, ENKTL nasal type is usually treated with 50 Gy with a boost of 5–10 Gy to residual disease [4]. Our study showed favorable outcomes with median doses of 35 Gy, 49 Gy, 30.6 Gy, and 40 Gy in ALCL, SPTL, CD4+ PCSM-TLPD, and ENKTL, respectively. Importantly, no recurrences were observed, suggesting that RT is effective in treating responsive CTCL lesions. As MF was the most common CTCL subtype, and most lesions receiving low doses were MF, it was challenging to evaluate the outcomes of low doses in CTCL, irrespective of MF. Another reason supporting the feasibility of lower doses in CTCL treatment is the observed effectiveness and practicality of re-RT. In this study, we observed favorable outcomes in patients treated with RT, particularly noting the efficacy of re-RT in recurrent lesions; approximately 71% of recurrent lesions were successfully managed with re-RT, achieving PR in nine lesions and CR in one lesion. This indicates that re-RT is a viable option for treating resistant lesions, demonstrating feasibility and effectiveness in managing CTCL recurrences. These findings suggest the consideration of initiating treatment with lower doses of RT, particularly for lesions expected to be responsive, while reserving re-RT for resistant lesions. This approach also provided a strategic option for managing recurrences, ultimately enhancing patient outcomes and quality of life. Future studies should focus on evaluating the clinical outcomes of low-dose RT and re-RT in a larger cohort of patients with CTCL, including various subtypes, to validate these findings.

This study had several limitations. First, due to the retrospective nature of this study, incomplete documentation or missing data within medical records might have introduced potential biases, especially the risk of underreported toxicities. Secondly, given the relative rarity of the disease and its various subtypes, the sample size of analyzed lesions remained small despite the 20-year study period. Notably, patients diagnosed with Sezary syndrome, primary cutaneous acral CD8+ T-cell lymphoproliferative disorder, or aggressive epidermotropic cytotoxic T-cell lymphoma were not included in the study cohort. The relative frequencies of these subtypes are reportedly very low, ranging from less than 1% to 2% [25]. Thirdly, during the long enrollment, there were various approaches to decide radiation treatment doses. However, since the treatment was carried out at a single institution over a long period and the guidelines were not significantly revised, we can say that there was negligible heterogeneity in treatments. Lastly, while not all lesions in the advanced stage group were pathologically confirmed, rigorous clinical diagnosis by experienced dermatologists was used to ensure the inclusion of only CTCL lesions in our analysis.

The major strength of this study is that we utilized various RT techniques for skin lesion treatment. While photon beams were sparingly used for complex anatomical cases, electron beam therapy was used as the primary modality, offering distinct advantages, especially in CTCL cases, due to its ability to deliver high surface doses. Field shaping techniques, including conventional blocks and IORT cones, achieve precise RT delivery to lesions while minimizing exposure to surrounding tissues. However, electron beams inherently produce penumbra, necessitating larger lateral margins for treatment efficacy. Our hospital adopted IORT cones to mitigate this issue. Fig. 5 demonstrates the comparison of the dose profile at the surface between a conventional block and an IORT cone. Measurements were taken by placing a film on a solid phantom at a depth of 7 mm, corresponding to the D_max of 4 MeV. When using an IORT cone with a diameter of 38 mm, the distribution on the surface is more homogeneous compared to a conventional block. Additionally, the 3D bolus was used to treat lesions much smaller than the smallest size of IORT cones, and we had to increase the surface dose to a specific millimeter.

Fig. 5.

Comparison of dose profile at the surface between a conventional block and an intra-operative radiotherapy (IORT) cone at D_max of 4 MeV.

In conclusion, in this retrospective study, skin-directed RT in CTCL was effective for local control and well-tolerated with less toxicity. Although there was no significant difference between <30 Gy and ≥30 Gy in lesions with MF, further investigation is needed to evaluate the use of low doses.

Supplementary Materials

Article information Continued