Purpose
Skin cancer treatment options
With four million non-melanoma skin cancer (NMSC) lesions per year in the US, skin cancer is the most common malignancy affecting 2-3 million people each year in the country, reaching an epidemic level [1, 2]. Rates are expected to rise by 3-6% per year, with the largest increase occurring for basal cell carcinoma (BCC), which accounts for 70% of all NMSC cases [3]. It is predicted that 40-50% of Americans who will live to age of 65 years will have either BCC or squamous cell carcinoma (SCC) at least once in their lifetime [4]. According to a study in 2009, nearly 50% of people routinely treated for BCC, developed multiple primary BCCs during 10 years of observation [5]. The impending increase in the incidence of skin cancer warrants a close examination of additional effective treatment options.
Current treatment options for NMSC include topical therapy, photodynamic therapy, cryotherapy, electrodessication and curettage, surgical excision, Mohs micrographic surgery (MMS), and radiation therapy. Factors, such as lesion type, size, and location as well as patient’s age, health, comorbidities, medication use, treatment cost, treatment time, wound care, and patient’s preference are considered in selecting the best therapeutic intervention to achieve optimal results [3, 6]. While MMS is the gold standard treatment for NMSC offering unique margin control, with a reported five-year cure rate of 98% in a prospective randomized study [7], it may not be the most appropriate treatment modality for all NMSCs, nor may be the best option for some patients.
Prior clinical evaluations
The advantage of precise and localized treatment using high-dose-rate (HDR) remote afterloaders is the option to reduce the number of treatment fractions and increase radiation dose per hypofractionation. High-dose-rate brachytherapy treatments are typically administered in 6-10 Gy fractions, with a total dose delivered over 3-4 weeks [8, 9]. Tormo et al. studied HDR brachytherapy with iridium-192 as the radioactive source used in NMSC, and reported 98% local tumor control (single persistence) at 47 months follow-up [8]. Other HDR brachytherapy studies have reported five-year cure rates of 98% [9-11].
X-rays have also been successfully used over the past century for NMSC treatment, with the first reported cases in 1902 and 1903 by Pugh and Sequiera, respectively [12]. Moreover, Grenz rays, contact, superficial (“soft” X-rays), and orthovoltage radiation therapy are viable treatment options despite their varying energy levels. Superficial “soft” X-rays have successfully treated skin cancers, with five-year cure rates of 90-96%, similar to those seen in HDR brachytherapy [13-16]. In fact, both HDR, electronic brachytherapy and superficial X-rays, have recurrence rates comparable to surgical excision, which ranges from 1.3% to 10.1% [17].
An electron beam radiotherapy system using photon radiotherapy doses similar to those previously reported should, theoretically, result in comparable local control rates in other superficial radiotherapy methods and simple excision (90-96% local control) in NMSC. The electron beam system used in this study, Esteya® (Elekta AB, Stockholm, Sweden), generates an X-ray spectrum within the “superficial” range (a peak energy of 69.5 kVp), delivering the collimated X-ray beam from a source positioned approximately 6 cm from the skin surface. A small, prospective study evaluating 20 BCC lesions (prescription dose, 42 Gy) treated with Esteya, demonstrated a 95% local tumor control after one year [18]. A retrospective study on the clinical efficacy and acute toxicities of Esteya [19], at one-year follow-up (75 lesions treated at a total dose of 45-50 Gy, in 8-10 fractions) reported only one persistence, resulting in 98% local tumor control. Acute toxicities included erythema, desquamation, and hypopigmentation, with erythema resolving within three months after treatment.
In other studies, different low-energy electronic X-ray brachytherapy devices (~50 kVp peak energy) were used for treatment of NMSC. Single-center retrospective analysis of patients reported using the Accent® electronic brachytherapy system (Xoft Accent electronic brachytherapy system, Elekta AB, Stockholm, Sweden), showing 100% local tumor control after 10 months [20]. No grade 3 or higher adverse events were observed at any time point. Cosmesis at one year was excellent for 92.9% and good for 7.1% in 42 evaluable lesions. The same treatment system was employed to treat 524 BCC, SCC, and SCC in situ lesions at a single-center [21]. After 12 months of follow-up, there were four recurrences (0.76%), with excellent/good cosmesis reported in all patients.
Material and methods
Study objectives
The primary endpoint of this study was to evaluate the effectiveness of local control in skin surface brachytherapy using the Esteya system for BCC/SCC at two and five years after brachytherapy.
The secondary endpoints were to assess toxicity rates of EBx, cosmetic outcomes, and patient quality of life measures. Adverse events (AEs), cosmetics, and quality of life outcomes were patient-reported and clinician-assessed post-EBx using validated skin toxicity and quality of life (QoL) surveys [see Suppl. Table 1 for RISRAS and skin cancer index (SCI)].
Design summary
This was a multi-center, prospective, observational study investigating the local control, toxicity profile, cosmetic outcome, and QoL of electronic brachytherapy used in patients with early-stage BCC/SCC. A third-party Institutional Review Board (Sterling IRB, Atlanta, GA) reviewed and approved the study protocol and all investigative sites. The study involved 205 patients with 236 lesions (136 BCC, 96 SCC, and 4 other type histology). All lesions were treated with a fractionation scheme that delivered ~70 Gy biologically equivalent dose (BED) at 3-4 mm depth and 20 mm width (lesion diameter). The details of dose rationale are found in the dose rationale and risk/benefits sub-section at the end of the methods section. Table 1 provides the allowable dose-fractionation regimens.
Table 1
Prescription dose schema
| No. of fractions | Dose (Gy) per fraction | Total dose (Gy) | BED (Gy) @ Rx depth (3 mm) |
|---|---|---|---|
| 6 | 7.0 | 42.0 | 71.40 |
| 8 | 5.6 | 44.8 | 69.89 |
| 10 | 4.7 | 47.0 | 69.09 |
| 12 | 4.1 | 49.2 | 69.37 |
A 5-mm margin was added to gross tumor volume for treatment. Based on clinical judgment, this standard fractionation scheme was selected for majority of lesions; however, some alternate schemes were used to avoid significant acute- and late-term toxicities. Treatment locations included head, neck, torso, and upper extremities. Patients were scheduled for follow ups at 1, 3, and 6 months, and then annually until 5 years post-treatment. At each follow-up visit, local control, toxicity (type and grade), and cosmesis were assessed by site’s healthcare provider or physician. Patients were withdrawn at 5 years after last fraction, when experiencing a recurrence within a treated field, withdrew their consent, or were lost to follow-up. The study was terminated in May 2024, and the remaining active patients were also exited at that time.
Study population
Every site’s principal investigator (PI) pre-screened each potential subject’s medical records/data to determine whether the patient met eligibility criteria. Following pre-screening, patients were asked to consider participating in the study, and were provided a copy of informed consent. Patient consent was obtained prior to study enrollment.
The standard of care for SCC/BCC cases included medical history and physical examination, clinical staging and risk group assessment, and review of BCC/SCC biopsy report.
Inclusion and exclusion criteria
There were separate inclusion and exclusion criteria for both subjects and lesions.
Inclusion criteria for subjects
1. Men or women ≥ 55 years old, with estimated life expectancy of at least 5 years; 2. Must be able to understand, read, and speak English; 3. Able and willing to complete RISRAS, ECOG, and SCI standard instruments (see Appendices); 4. Able to provide written informed consent.
Inclusion criteria for each lesion (up to 4 lesions per subject)
1. Primary BCC (superficial or nodular) or primary SCC, as determined by histopathology report; 2. Lesion with greatest diameter ≤ 20 mm; 3. Lesion depth of less than or equal to 4 mm; 4. Clinical stage Tis, T1, or T2, with two or fewer high-risk clinical or pathologic features, as defined by the AJCC Cancer Staging Manual, 7th Ed. [22] (see Table 2).
Table 2
AJCC tumor staging (T–) and “high-risk” features
Exclusion criteria for subjects
1. Evidence of poorly controlled diabetes; 2. Pregnant patients; 3. Receipt of drug that can affect biologic response to radiation (radiosensitizer or radioprotector); 4. High likelihood of protocol non-compliance (in the opinion of investigator); 5. Life expectancy of less than 5 years; 6. Immunosuppressed patients; 7. Known other malignancy within 3 years (except for adequately treated basal cell or squamous cell carcinoma of the skin); 8. Concurrent non-protocol-specified anti-tumor therapy (e.g., chemotherapy, other targeted therapy, or topical therapy, such as 5-Fluorouracil or imiquimod, radiation therapy, surgery, or photodynamic therapy; 9. Recent (within 4 weeks of consent) or planned participation in another experimental drug study; 10. Widespread superficial multifocal BCC considered unresectable due to extent of involvement; 11. Patients likely to be lost to follow-up within 6 months (e.g., plans to move to areas far away, other countries, other states, etc.).
Exclusion criteria for each lesion (up to 4 lesions per subject)
1. Previously treated BCC/SCC lesion (i.e., recurrent BCC/SCC); 2. BCC/SCC in region adjacent to or overlapping with region of prior radiotherapy (within 5 mm of proposed treatment field); 3. BCC/SCC on any surface that cannot be flattened by treatment device (i.e., target area not flat); 4. BCC/SCC adjacent to or overlapping with a burn or scar (within 5 mm of proposed treatment field margins); 5. BCC/SCC in area with compromised lymphatic drainage or vascular supply (e.g., lymphedema); 6. BCC/SCC within 5 mm of another treated or untreated BCC/SCC; 7. BCC/SCC lesion depth > 4 mm (in the clinical opinion of investigator); 8. Prior surgery to the same site.
Study procedures
The radiation delivery system and accessories used in this protocol are available for commercial sale in the United States. Details of treatment procedures implementation for electronic brachytherapy in skin lesions followed that of Pons-Llanas et al. study [23]. Gross tumor volume (GTV) outer boundary was outlined on the perilesional skin. A plastic template (varying circular sizes of 10-30 mm in diameter) was centered over GTV, providing at least 5 mm margin between GTV and edge of the circular template’s “useful treatment beam” circumference. The outer edge of the plastic template was circular, several millimeters concentrically placed outside the useful treatment beam circle. This outer edge can be traced onto the skin indicating precise location to place the Esteya treatment beam applicator. Table 2 provides one such template’s usage of Pons-Llanas et al.
Skin examination
Skin examinations were conducted using standard dermatological methods and practice for suspected/treated NMSCs. A pre-treatment skin biopsy (punch or shave) was performed prior to subject treatment.
Electronic brachytherapy device and radiotherapy delivery
The Esteya device (Figure 1) consists of a treatment unit, laptop with planning software, and treatment control panel. The unit can be positioned over most any site on the body. X-rays produced by the Esteya device have a peak energy of 69.5 kVp. The device’s high-dose-rate radiation delivery (2.7 Gy per min) at 3 mm depth yields treatment times of approximately three minutes per fraction. The output field of X-ray source can be modified to match the tumor size using a surface applicator that functions as a collimator. There are different sizes of surface applicators available (range, 10-30 mm) to allow treatment of tumors up to 20 mm diameter, with an appropriate margin.
Fig. 1
Elekta Esteya® device
Courtesy of Jonathan Cheng, MD, PhD, Genesis Cancer Center, Baytown, TX, USA.
Subjects were placed in supine or recumbent position on a couch dedicated for radiation therapy treatment. GTV was outlined free-hand, and a margin was added to account for sub-clinical disease outside of the GTV, also known as clinical tumor volume (CTV). A 4 mm margin has been shown to be sufficient for low-risk NMSC and 6 mm margin for high-risk lesions [24]. Ruler measurements were used to obtain the CTV. The choice of skin surface applicator (diameter) was the smallest diameter applicator that covered the whole CTV. The unit and treatment applicator were brought towards the subject’s lesion site, and the applicator was used orthogonally to the skin surface. Immobilization was employed as needed. Treatment dose was chosen from possible options depicted in Table 1, with each radiotherapy fraction separated by at least 48 hours.
Study outcome measurements
Radiation-induced skin reaction assessment scale (RISRAS) provides a grading mechanism specific to skin reactions associated with radiation therapy delivery. It has two components: a patient-graded symptom score (4 questions classified on a scale from 1 to 4), and a healthcare provider scale (HCP) that covers the usual acute skin reactions of erythema, dry and moist desquamation, and necrosis, providing specific grading criteria for each reaction.
Subject-reported health-related quality of life outcomes were assessed using skin cancer index (SCI). The SCI is a 1-page instrument consisting of 15 items, with responses provided on a 5-point Likert’s scale based on experience over prior month. Responses are categorized on 3 sub-scales: emotional, social, and appearance, with combined sub-scale scores yielding an overall score ranging from 100 (best) to 0 (worst). Of the 15 items in the instrument, 7 contribute to emotional, 5 to social, and 3 contribute to appearance sub-scales.
Cosmesis was assessed by both subjects and clinicians in this study. Since subject’s assessment of cosmesis is subjective, it was performed using a qualitative descriptor (i.e., excellent, good, fair, or poor). Although patient self-reported description of a “good” result is completely subjective, without criteria or definition for this outcome, it generally means little to no residual change in the skin after treatment. Clinician assessment of cosmesis was both objective and subjective.
Assessing and recording adverse events
The study used two methods of assessing grade of every observed adverse event (AE). Baseline assessment was made prior to initiating radiation therapy using the healthcare provider portion of RISRAS scale. During brachytherapy (at 50% and 100% completion of electronic brachytherapy fractions), and at 4-week and 3-month follow-up visits, AEs were assessed using the healthcare provider portion of RISRAS scale. For adverse events first noted at any time after 3 months, common terminology criteria for adverse events (CTCAE) version 4.0 was applied to evaluate and record any AE. Table 1 illustrates the matrix of study procedures, assessments, and follow ups for all subjects of the study.
Dose rationale and risk/benefits
Historical data on fractionation schema and local control for electronic brachytherapy provided guidance on the limited range of prescription doses allowed in this study. The choice of prescription was based on that used in a 4-year study examining isotope-based brachytherapy delivered with Valencia applicator, which used 42 Gy and had a 98% tumor clearance [8] as well as a pilot study [18] with Esteya using 42 Gy in 20 patients, showing 95% tumor control at 1-year follow-up. These studies used a fractionation scheme based on constant biologically equivalent dose (BED) of approximately 70 at prescription depth. Lower BED values (< 70 BED) appear to have higher recurrence rates, and those above 70 have excellent local control, but higher local toxicity rates.
There is a general sense in the clinical community, which supports the recent published experience, indicating that about 70 BED for body sites without superficial bony structures (i.e., skin of dorsal hand and anterior tibial skin) is a good balance between excellent local control and toxicity. Therefore, in the current study, a narrow range of 69-72 Gy BED was chosen to provide a range of fractionation schemes. The fractionation scheme used in this study is shown in Table 1. Note that the prescription dose of 3 mm depth has a variable number of fractions (with inversely varying dose per fraction), all of which yield similar BED (range, 69.09-71.40 Gy BED). Moreover, the 7 Gy × 6 fraction regimen is a commonly used fractionation scheme for most skin sites, especially for smaller applicator diameters and less sensitive areas.
Here, lesion depth was determined by either a dermatologist and communicated to a radiation oncologist, or by a radiation oncologist directly. Dermatologists used ultrasound or their clinical judgment following review of biopsy samples to determine the depth of lesion as well as literature, which reported depth guidelines based on tumor size and location [25, 26]. Radiation oncologists used either pathology findings directly or a dermatologist’s lesion depth, combined with concerns of anatomic location (e.g., the depth of underlying sensitive/confounding structures, such as tibia, eyelid, etc.) to determine the actual prescription depth.
Results
In total, 205 subjects (236 lesions) were included in this study. Most subjects (71%) were prospectively consented, and 29% were accrued and followed retrospectively. The median age was 74 (range, 56-96) years, with 62% of males and 38% of females (all subjects were of White race). ECOG performance status was 0/1 in all but 2 subjects. Electronic brachytherapy was the patient preference in nearly all the subjects (94%), with patient age (47%) and lesion location (74%) as the other main reasons for non-surgical approach. The most common comorbidities were history of bleeding disorder/anticoagulants (41%), smoking history (27%), and diabetes (13%). The median follow-up was 24.2 months, with a maximum of 73.5 months post-completion of electronic brachytherapy. The patient characteristics are summarized in Table 3.
Table 3
Patient characteristics
Basal cell carcinoma lesion histology was the most common (n = 136, 57%), and SCC was less common (n = 96, 41%). The most common AJCC classification was T1N0M0 (n = 176, 74.6%), followed by TisN0M0 (n = 39, 16.5%). Lesions were primarily located on the head (n = 191, 81%), followed by nose (33%), forehead (10%), and cheek (17%). Upper extremities accounted for 11% of lesions. A more detailed summary of the lesion locations is shown in Table 4A. The median lesion diameter was 8 (range, 1.0-20) mm, with the median depth of 2 (range, 1-3) mm (Table 4B). Radiation therapy data are shown in Table 5. The median number of fractions was 8 (range, 6-13), with the median treatment depth of 3 (range, 2.0-3.5) mm. The median dose per fraction was 5.6 (range, 4.0-7.0) Gy. The median total treatment time (“beam on”) was 17.6 (range, 15.4-31.2) minutes. The most common diameter applicators used were 20 mm and 15 mm diameters (41% and 36%, respectively).
Table 4A
Lesion location
Table 4B
Lesion characteristics
Table 5
Radiation therapy treatment data
The RISRAS assessments (patient symptom score, healthcare provider score, and overall score) are presented in Table 6. A score of zero indicated no concern by either the patient or healthcare provider. The patient (median) scores were flat at a value of 4, except for the time at EBx completion when rising to 5. Scores for healthcare providers were zero (0) at baseline and rising thru EBx (1 at 50% completion, 3 at 100% completion), then fell back to baseline of zero by 3 months.
Table 6
RISRAS over time
The skin cancer index (SCI) was obtained at baseline, halfway through EBx, at EBx completion, and then at various times post-EBx completion. It is important to emphasize that while 71% of subjects were prospective (consented prior to evaluation and treatment), about one-half of the 29% classified as retrospective subjects were actually consented after biopsy, but prior to starting brachytherapy. In terms of lesions treated, 236 lesions were included in database, and 199 of these were located on patients consented prior to starting brachytherapy. Therefore, only 15.7% of lesions were associated with subjects that provided consent after brachytherapy was started or completed.
The sub-scale values and SCI index values over time are given in Table 7 (n, mean/SD, median, and range). The SCI index median was the lowest at baseline (median score, 88.9), and increased modestly throughout radiotherapy (90.1 and 92.0) and follow-up from 1 month (92.5) to 24 months (97.4). A score of 100 was the maximum for the SCI index.
Table 7
SCI over time
Following EBx completion, healthcare professional (HCP) and patient-rated cosmesis was rated excellent/good (E/G) by both the groups at 90-100% rates, except for the HCP rating at 1 month post-EBx, where 85.6% of patients reported E/G (13.2% fair, 0.0% poor). At 36 months post-EBx, cosmesis was rated E/G in a slightly lower percentage of patients than all other times (HCP E/G rating: 83.6%, fair 16.4%, poor 0%; patient E/G rating: 86.6%, fair 10.4%, poor 3.0%). Erythema was the most common acute adverse event (34.1% at 1 month), rebounding back to zero by 6 months. Hypopigmentation rate increased from 18.0% at 1 month rapidly to 43-73% from 3 to 36 months, then dropped to 67% and 27% at 4 and 5 years, respectively (sample size was small at these late times). The detailed data showing cosmesis ratings and adverse events throughout follow-up (1 month to 5 years) are demonstrated in Table 8.
Table 8
Cosmesis and adverse event observations during follow-up
Out of 236 lesions, there was a single lesion recurrence (recurrence rate, 0.42%). When expressed as “patient experiencing a recurrence”, the recurrence rate was 0.49% (1 of 205). For this recurrence, the patient was 82 years old, and the original lesion was located on the left eyelid, measuring 10 mm × 10 mm (BCC with T1N0M0 histology). Brachytherapy was delivered with a prescribed dose of 10 fractions with 4.7 Gy, using a 20 mm applicator with 3 mm prescribed depth. The recurrence was clinically diagnosed at 4.5 months post-brachytherapy completion, located in-field (inside margin) as BCC (biopsy-assessed as T1N0M0).
There were 8 deaths, none of which were deemed related to BCC/SCC or treatments delivered in this study. Table 9 shows the recurrence and subject disposition at study closure.
Table 9
Recurrence and subject disposition
Discussion
The specific method of high-dose-rate electronic brachytherapy device used in this study was the Esteya® system (Elekta AB, Stockholm, Sweden). It falls under the category of superficial therapy, where X-rays are produced at potentials ranging from 50 kV to 200 kV and source-to-surface distance (SSD) usually ranges between 15 and 20 cm. The Esteya system generates an X-ray spectrum with peak energy of 69.5 kVp and skin-to-surface distance of 6 cm. The quality of a superficial beam is suitable for treating NMSC lesions, as their depths rarely exceed 5 mm. Beyond this depth, dose drop-off fails to deliver adequate depth dose.
Smaller and shorter follow-up studies reported local control of nearly 98% [18, 19]. In 2019, a consensus guideline was published [25], providing a review of the local control, rationale for the use of superficial X-ray therapy, and patients who strongly could benefit from this method over surgical treatments. Another large retrospective study reported an overall 5% recurrence rate at 5 years in 1,715 patients [26]. Other definitive sources for skin cancer brachytherapy include the GEC-ESTRO recommendations [27], the American Brachytherapy consensus statement [28], and chapter 29 (skin cancer) of the GEC-ESTRO handbook [29].
The current study comprised 205 patients (236 lesions), with a median follow-up of 24.2 years. While shorter than large retrospective study, most patients were prospectively recruited or consented within 2 years of completing EBx. The evaluation of adverse events, cosmesis, and quality of life appears more detailed and systematic compared with a large retrospective study. Moreover, this study provides a longitudinal perspective on the evolution of AEs, their severity, QoL, and cosmetic evaluations (both by healthcare providers and patients’ self-assessments). Therefore, each set of studies has relative strengths, which are complementary with this study providing shorter-term granularity on secondary endpoints.
The results of this research are consistent with a 2-year tumor control rates of 98% or better in BCC and SCC lesions measuring less than 2 cm in diameter. There was a single lesion recurrence (recurrence rate of 0.42%). For this recurrence, the patient was 82 years old with a lesion on the left eyelid, observed for 4.5 months after brachytherapy completion, located in-field. The eyelid, with skin structure not amenable to flattening by the EBx applicator, may not be appropriate for the use of the Esteya system. Concavity of the target area can change the skin-to-surface distance within the target area in a meaningful manner, which can result in underdosing portions of the target.
The vast majority of toxicities measured in this study were acceptable for both, the patient and the treating physician. No cartilage degradations were reported, especially on nasal lesions.
This study provides further evidence of the similarity of results with Mohs microsurgery for 24 months.
The primary treatment goal for BCC and SCC of the skin is to provide long-term control while optimizing cosmesis in an acceptable manner to the patient. EBx is a viable option for patients based on characteristics of the tumor and patient’s preference. The main limitation of the current study was the duration of follow-up with a median of 24.2 months, although many patients were followed out for 4-5 years. The ideal follow-up time for a study of this nature is approximately 48-60 months.
Conclusions
This is the first report of a phase IV study on NMSC treated with electronic brachytherapy in 205 subjects (236 lesions). The median follow-up was 24.2 years, and the lesion recurrence rate was 0.42% (1 recurrence out of 236 lesions), or 0.49% on a patient basis (1 recurrence out of 205 patients). Patients tolerated the treatments very well as evidenced by strong, longitudinal scores on the skin cancer index, RISRAS assessments, and cosmetic evaluations (performed by both patients and healthcare providers). Adverse event rates were very low, except for the expected acute erythema, chronic hypopigmentation, and telangiectasia. The study provides additional support for the delivery of electronic brachytherapy for low-risk NMSC lesions in patents who prefer a non-surgical treatment, especially in those at risk for surgical problems, surgical cosmesis issues, keloid formation, wound care issues, and use of anticoagulant therapy.
