X-Ray Therapy of Skin Cancer at Critical Facial Sites: Results of a Tertiary Referral Center

Abstract

Background

Therapeutic management of skin cancer (SC) in the “H-area” of the face is challenging due to the demands of function, cosmesis, and efficacy.

Objective

To evaluate the efficacy and safety of superficial radiotherapy (RT) for SC in the “H-area” of the face and to identify predictors of recurrence.

Methods

Retrospective analysis of patients with SC in the “H-area” of the face treated with superficial RT at the University Hospital of Zurich, Switzerland, between 2010 and 2021. The study included cases of basal cell carcinoma (BCC), squamous cell carcinoma (SCC), in situ SCC (isSCC), lentigo maligna melanoma (LMM), lentigo maligna (LM), and primary cutaneous T/B-cell lymphoma.

Results

We identified 546 lesions in 382 patients. Estimated relapse rates at 5 years were lowest for SCC (3.2%), isSCC (1.9%), and BCC (9.7%), while LMM and LM showed estimated relapse rates of 28.9% and 35.0%, respectively. Significant predictive factors for overall recurrence were increased Common Terminology Criteria for Adverse Events (CTCAE) grades with a hazard ratio (HR) of 1.9 (1.1–3.4) and the treatment of secondary lesions with an HR of 2.6 (1.4–5.1).

Conclusion

RT is a valuable treatment option for malignant skin tumors, especially non-melanoma SC, in high-risk areas, including the facial “H-area,” providing favorable results with minimal side effects.

Introduction

Skin cancer (SC) represents a significant and increasing burden of disease both in Switzerland and globally and is classified into melanoma and non-melanoma skin cancer (NMSC) [1]. Melanoma, which has the highest incidence rate in Switzerland among European countries at 25.8 per 100,000 inhabitants, has become a critical problem. NMSC constitutes the majority of epithelial neoplasms in humans, with Switzerland exhibiting one of the highest incidence rates in Europe. These include basal cell carcinoma (BCC) with a prevalence of 75/10,000 inhabitants in Switzerland, accounting for 80% of all NMSC, squamous cell carcinoma (SCC) with an incidence of 29/100,000 per year in Switzerland, and the less common SCs [2, 3]. These statistics underline the importance of taking appropriate preventive and therapeutic measures to reduce the impact of SC, both from a patient and healthcare economic sustainability perspective. Surgical excision, particularly Mohs micrographic surgery, is recognized as the primary treatment option for BCC, SCC, and melanoma for total eradication and minimizing recurrence rates [2, 4–9]. For premalignant lesions such as actinic keratosis (AK) and superficial tumors associated with SCC, BCC, and lentigo maligna (LM), non-invasive treatments such as cryosurgery, photodynamic therapy, and topical treatment can be used [4, 5, 10, 11]. In addition, superficial radiotherapy (RT) has become a preferred option for older patients who are not candidates for surgery, as well as for the treatment of difficult-to-treat areas, including the “H-area” of the face, where a balance between therapeutic efficacy and cosmetic results is crucial [12]. It can be used as a definitive, adjuvant, or palliative solution. The “H-area” of the face as illustrated in Figure 1 comprises orbital, periorbital, nasal, buccal, jaw, auricular, preauricular, and temple regions [6, 10]. It is defined by its heightened exposure to UV light, a critical risk factor for SC, and its increased recurrence risk compared to other areas of the body. Due to distinct entity, histology, and genetics of SCs, as well as variations in RT protocols, the number of clinical trials consolidating SC research in this area remains limited. Consequently, contemporary therapeutic guidelines predominantly rely on the insights derived from small and short timed retrospective cohort studies and expert consensus [8, 9, 11]. Through our research, we endeavor to enrich the body of evidence supporting the efficacy of RT for SC located in the “H-area” by undertaking a retrospective examination of a significant cohort of patients diagnosed with SC who underwent RT at our institution. Specifically, our study focuses on the analysis of health-related patient data routinely collected during daily clinical practice at our institution.

Fig. 1.

Area (dark blue) depicting the “H-area” of the face.

Patients and Methods

In this retrospective study, we analyzed the data of 383 patients aged 60 years and older with 546 SC lesions in the “H-area” of the face. These patients were treated with a completed cycle of RT at our institution between 2010 and 2021, with a minimum follow-up period of 24 months. Data were collected from available medical records. RT was delivered using the Gulmay D3100 Superficial X-Ray Therapy System (Gulmay Ltd., Surrey, UK). Treatment was prescribed for each patient by a senior physician based on our established protocol. For AK and Bowen’s disease (BD), Grenz rays (10–20 kV) were used. Dose prescription and fractionation were 16–48 Gy in 2–8 fractions given every 3–7 days. For soft ray therapy (20–50 kV) for SCC and BCC, total prescribed doses were 40–56 Gy given in fractions of 4–8 Gy daily to every 4 days. Recommended safety margins for NMSCs were 1 cm from the visible tumor. For LMM, the protocol specified total doses of 42–54 Gy or 100–120 Gy for surgically pretreated lesions with in situ remnants, fractionated into 6 Gy or 10 Gy, respectively, using boundary beam, soft beam, or orthovoltage (10–100 kV) therapy. For LM, the dose prescription and fractionation for Grenz rays was 100–120 Gy given in fractions of 10 Gy every 4 days. LM with invasion along follicles was treated with total doses of 42–50 Gy (20 kV). The safety margins of LM and LMM were at least 1.5 cm from the visible tumor border. An external photon beam of 30–100 kV is used in primary cutaneous lymphoma. For primary cutaneous B-cell lymphoma (pcBCL), total doses of 8–40 Gy were proposed, depending on the subtype, in fractions of 2–4 Gy each, administered two to three times a week. For primary cutaneous T-cell lymphoma (pcTCL; mainly mycosis fungoides [MF]), the protocol proposed total doses of 8–24 Gy in single fractions of 2–4 Gy twice a week. The safety margin for lymphoma was 1–2 cm from the visible tumor border. The Common Terminology Criteria for Adverse Events (CTCAE) v5.0 radiation dermatitis scale was used to grade the severity of adverse events. After RT, lesions can remain visible for several weeks and are often misinterpreted as recurrences. Therefore, outcomes were assessed based on treatment response (absence of clinically detectable tumor 3 months after RT) and recurrence (histological confirmation of tumor in the treated area at least 3 months post-RT, including marginal recurrence). Follow-up appointments were scheduled in accordance with our protocol at 3, 6, 12, 18, and 24 months, with the typical course depicted in Figure 2. As a referral center, some follow-up visits were conducted by peripheral physicians who remained in close contact with our hospital and submitted follow-up questionnaires at each scheduled interval. For patients lost to follow-up after 24 months, we contacted their treating physicians to verify any further relapses and reached out to the patients themselves to confirm survival status. The study was authorized by the Local Ethics Committee (Ethikkommission Zürich, Switzerland) under the project ID 2023-02147.

Fig. 2.

Clinical presentation and response to RT before RT treatment, during RT, 4–12 weeks after treatment, and 12 months after treatment of a 70-year-old female patient with LMM of the left lower eyelid (a–d), an 81-year-old male patient with nodular BCC (e–h), and a 68-year-old female patient with moderate SCC of the right upper lip (i–l).

Statistics

Data analysis was performed using R (RStudio Inc., Boston, MA, USA). Numerical data were expressed as mean and standard deviation, and median and range were given when normality could not be assumed. Nominal data were expressed as absolute numbers and percentages, and Pearson’s χ² or Fisher’s test was performed for significant tests. Recurrence analysis was estimated using Kaplan-Meier curves and the log-rank test for significance. We used mixed-effects Cox models to assess the association of age, gender, tumor location, immunosuppression, total dose (Gy), field size, CTCAE grade, and secondary lesions with relapse. Each predictor was initially tested using a univariate Cox proportional hazards model (coxph). For the mixed-effects Cox models (coxme) across all histology subtypes and for NMSC subtypes, CTCAE grade and the presence of secondary lesions – both significant in univariate analysis – were included, along with age and gender as clinically relevant covariates. Patient ID was included as a random effect to account for repeated measures [13]. For the BCC model, only field size was significant and included, alongside age, secondary lesions, and patient ID number as a random effect due to the limited number of events. The STROBE guidelines were followed for reporting purposes [14].

Results

Patient Characteristics

Of 550 patients who underwent RT of “H-area” SC, 383 patients with 546 lesions met the inclusion criteria. The cohort had a median age of 78 years (range 60–100 years), was predominantly male (n = 217, 56.7%) compared to female (n = 166, 43.5%), and had a median follow-up of 54 months (range 24–157 months) (Table 1). The main site of lesion was the nose (46.3%, n = 253). The main skin type according to the Fitzpatrick scale was grade 2 with 91.2%.

Table 1.

Patient and lesion characteristics

Category	Variable	Count	Mean (range)
Age, years	​	​	78 (60–100)
Gender	​	100% (n = 383)	​
​	Female	43.3% (n = 166)	​
	Male	56.7% (n = 217)	​
Skin type grade		100% (n = 546)	​
​	Grade 1	1.8% (n = 10)	​
	Grade 2	91.2% (n = 498)	​
	Grade 3	6.8% (n = 37)	​
	Grade 4	0.2% (n = 1)	​
Location of index lesion		100% (n = 546)	​
​	Auricular	9.9% (n = 54)	​
	Buccal	8.1% (n = 44)	​
	Frontal	2.8% (n = 15)	​
	Nasal	46.3% (n = 253)	​
	Periocular	13.7% (n = 75)	​
	Preauricular	9.5% (n = 52)	​
	Temporal	9.7% (n = 53)	​
Lesion	​	100% (n = 546)	​
​	Secondary lesion	9.2% (n = 50)	​
	Primary lesion	90.8% (n = 496)	​
Secondary lesion		100% (n = 546)	​
​	Other	10% (n = 5)	​
	Physical	18% (n = 9)	​
	Surgery	58% (n = 29)	​
	Topical	14% (n = 7)	​
Other malignancies		100% (n = 546)	​
​	No other malignancies	22.5% (n = 123)	​
	Other malignancies	77.5% (n = 423)	​
Follow-up, months		​	54 (24–157)
Recurrence	​	100% (n = 546)	​
​	No recurrence	90.5% (n = 494)	​
	Recurrence	9.5% (n = 52)	​
Acute toxicity grade		100% (n = 546)	​
​	Grade 1	33.7% (n = 184)	​
	Grade 2	43.8% (n = 239)	​
	Grade 3	22.5% (n = 123)	​
Final cosmetic outcome		100% (n = 546)	​
​	Acceptable	7.1% (n = 39)	​
	Good	55% (n = 300)	​
	Unknown	37.9% (n = 207)	​

Grades of acute toxicity as defined by CTCAE v5.0.

The analysis revealed that the most common lesions were BCC (50.2%, n = 274), in situ SCC (isSCC; 20.5%, n = 112), and SCC (11.4%, n = 62) (Table 2). Less common were LM (6.8%, n = 37), LMM (5.5%, n = 30), and primary cutaneous lymphoma (5.7%, n = 31) (pcTCL, n = 13 and pcBCL, n = 18). Histologically, BCCs were predominantly nodular (81.0%, n = 222), while isSCC included AK (75.0%, n = 84), BD (21.4%, n = 24) and isSCC (3.6%, n = 4). SCC grading showed 25.8% (n = 16) well-differentiated, 56.5% (n = 35) moderately differentiated, and 1.6% (n = 1) poorly differentiated cases. Histologic subgroups of pcTCL included predominantly MF and pcBCL follicular lymphoma. MF staging ranged from IB to IVB1, with 46% (n = 6) predominantly at stage II. Adverse events, as assessed by NCI-CTCAE v5.0, consisted primarily of mild to moderate acute radiodermatitis (grade 1: n = 184, 33.7%; grade 2: n = 239, 43.8%). A minority of patients experienced grade 3 reactions (n = 123, 22.5%), and no grade 4 reactions were observed. Long-term outcomes were favorable, with no serious adverse events reported. Telangiectasia or hypopigmentation occurred in 38 patients, while 300 patients had no long-term adverse events. 10.4% (n = 57) of all lesions had been previously treated. Four patients had to be excluded from the study due to a lack of clarity with regard to the progression of their lesion. Patient and lesion characteristics are summarized in Tables 1 and 2. After 5 years, 48 patients had died, with only one documented direct link to the treated lesion in a patient with MF. Three lesions had missing field size data and were not included in the COX analysis.

Table 2.

Histology and subgroups categorized by counts and percentages

Diagnosis	Subgroup	Percentage	Count
BCC		50.2	274
​	Nodular	81.0	222
	Trichogen	3.6	10
	Scirrhous	3.3	9
	Micronodular	2.9	8
	Superficial	2.6	7
	Metaplastic	2.2	6
	Multifocal	1.8	5
	Unknown	1.8	5
	Adnexal	0.7	2
isSCC		20.5	112
​	AK	75.0	84
	BD	21.4	24
	In situ	3.6	4
SCC		11.4	62
​	Moderate	56.5	35
	Good	25.8	16
	Unknown	16.1	10
	Poor	1.6	1
LM		6.8	37
LMM		5.5	30
T-/B-cell lymphoma		5.7	31
​	B-cell lymphoma	58.1	18
	T-cell lymphoma	41.9	13
Total	​	​	546

Estimated Relapse Rates at Timepoints and Total Relapse Rates

When stratified by histology, the lowest estimated relapse rates at 1, 3, and 5 years were observed for isSCC with rates of 0.0% (CI 0–0%, n = 0), 1.9% (CI 0–4.4%, n = 2), and 1.9% (CI 0–4.4%, n = 2), and SCC with 1.6% (CI 0–4.7%, n = 1), 3.2% (CI 0–7.5%, n = 2), and 3.2% (CI 0–7.5%, n = 2), respectively (Fig. 3). BCC showed slightly higher recurrence rates of 2.6% (CI 0.7–4.4%, n = 7), 6.3% (CI 3.3–9.1%, n = 17), and 9.7% (CI 5.7–13.5%, n = 23). Conversely, higher relapse rates were observed for LM at 5.4% (CI 0–12.4%, n = 2), 16.9% (CI 3.5–31–28.5%, n = 6), and 35.0% (CI 15.2–50.2%, n = 11), and for LMM at 6.7% (CI 0–15.2%, n = 2), 14.5% (CI 0.1–26.8%, n = 5), and 28.9% (CI 7.7–45.2%, n = 8). For pcTCL and pcBCL each, the relapse rates were 0.0% (CI 0–0%, n = 0), 8.3% (CI 0–22.7%, n = 1), and 8.3% (CI 0–22.7%, n = 1). Analysis of recurrence rates by histology showed significant differences (p value <0.05). However, within the isSCC, SCC, and BCC subcategories, there was no significant variance in the recurrence rate for specific subgroups (p values >0.05). Notably, in the SCC group, recurrences were detected only in lesions with moderate grading with estimated recurrence rates of 2.9% (CI 0–8.2%, n = 1), 5.7% (CI 0–13.1%, n = 2), and 5.7% (CI 0–13.1%, n = 2), respectively. The lowest total recurrence rates were observed for isSCC and SCC, with rates of 1.8% (CI 0.2–6.3%, n = 2) and 3.2% (CI 0.4–11.2%, n = 2), respectively. BCC showed slightly higher total relapse rates at 10.2% (CI 6.9–14.4%, n = 28), as did pcTCL at 7.7% (CI 0.2–36%, n = 1) and pcBCL at 5.6% (CI 0.1–27.3%, n = 1). The highest total relapse rates were observed for LM with 29.7% (15.9–47%, n = 11) and for LMM with 23.3% (CI 9.9–42.3%, n = 7). Analysis of overall recurrence rates by histology showed significant differences (p value <0.05).

Fig. 3.

Kaplan-Meier curves for relapse-free survival. Yellow: BCC, pink: T/B-cell lymphoma, light blue: LM, dark blue: LMM, light green: isSCC, dark green: SCC.

Predictors of Relapses

Recurrence correlated significantly (p value <0.05) with a linear increase in CTCAE grade hazard ratio (HR) 1.9 (CI 1.09–3.40, p = 0.02) and secondary lesions HR 2.4 (CI 1.12–5.16, p = 0.02). No significant correlations were found for age, gender, tumor location, immunosuppression, total dose (Gy), and field size (p values >0.05). For NMSC, significant correlations (p values <0.05) for recurrence were found only based on linearly increasing CTCAE grade HR 2.3 (CI 1.13–5.78) p = 0.02 and secondary lesions with HR 3.7 (CI 1.51–9.03) p = 0.01. For BCC, field size showed an HR of 1.1 (CI 1.02–1.11, p = 0.01), and when subdivided into nodular and all other subgroups, significance (p value <0.05) could only be demonstrated for subgroups other than nodular at HR 1.1 (CI 1.02–1.18, p = 0.01) vs. 1.1 (0.99–1.14, p = 0.08). There were no significant predictors (p value >0.05) for LM and LMM. For SCC and T/B-cell lymphoma, the number of events (SCC and isSCC n = 4, T/B-cell lymphoma n = 2) was too small to assess.

Discussion

The study investigated relapse rates following RT of SC within the “H-area” of the face, a high-risk recurrence area for SC, in a large patient cohort (n = 383) with a minimum follow-up of 24 months.

Non-Melanoma Skin Cancer

NMSC, while rarely metastatic and lethal, is highly prevalent and a burden to healthcare systems [4, 11]. Although surgical excision remains the gold standard except for AK, isSCC, and BD, where topical and physical treatments are recommended as well, RT is a valuable option for elderly or polymorbid patients and for tumors in surgically challenging sites [2, 6, 9, 10, 15].

Our observed recurrence rates after RT were for BCC 2.6%, 6.3%, and 9.7% at 1, 3, and 5 years; for isSCC 0.0%, 1.9%, and 1.9%; and for SCC 1.6%, 3.2%, and 3.2%. These values are consistent with the literature [16–35], though many previous studies either lacked long-term follow-up [17, 18, 23–33, 35–37], calculated total relapse rates only [16, 17, 23–25, 28–33], or included lesions outside the “H-area” [16–18, 23, 25, 28, 29, 31–34, 36, 37], likely underestimating recurrence rates. Notably, our study showed ongoing recurrence of BCC for up to 6 years, stressing the need for extensive follow-up. Compared to previous institutional data using older RT equipment, we observed improved outcomes. For example, a prior study by Zagrodnik et al. [37] using a similar RT protocol reported a 15.8% 5-year recurrence rate for BCC [37], nearly double our rate. This improvement can likely be attributed to more sophisticated technical implementation and increased clinical experience. Unlike Zagrodnik et al. [37], we observed no elevated recurrence in high-risk BCC subtypes, possibly due to limited sample sizes and protocol refinements stemming from greater experience. Similarly, for SCC, our recurrence-free survival exceeds previous reports. Barysch et al. [36] reported an 86.2% 5-year recurrence-free survival estimate under similar conditions [36], despite including fewer poorly differentiated tumors (57.2% vs. 26.7%). This may have led to more meticulous treatment planning. However, no significant differences in recurrence were found by SCC subtype or grade, likely due to low event numbers.

LM and LMM

LM, the most common in situ melanoma, primarily affects sun-exposed areas like the head and neck and carries a 50% risk of progression to LMM, the third most common melanoma subtype [38]. In our department, elderly patients with LM or LMM unfit for surgery are treated with RT using Grenz rays or superficial X-rays. We observed recurrence rates for LM of 5.4%, 16.9%, and 35% and for LMM of 6.7%, 14.5%, and 28.9% at 1, 3, and 5 years, respectively. Compared to earlier institutional studies reporting lower total recurrence rates (7% and 15%) [21, 39], our findings may reflect differences in anatomical site and case complexity. Notably, the previous analyses lacked a defined minimum follow-up and calculated only total relapse rates, potentially underestimating recurrence. Looking at other publications, there is a wide range of reported recurrence rates compared to NMSC [38, 40–43], while most of them defined no minimum follow-up period and included non-high-risk areas. Hedblad and Mallbris [40] analyzed 593 patients treated with Grenz rays (10 kV) and total doses of 100–160 Gy, a similar RT protocol to ours, with overall recurrence rates of 12% and 17% for primary lesions of LM and LMM, respectively. In 72% of the study population, the lesions were followed for more than 2 years. Only 30% of the lesions were located in the “H-area” of the face. In contrast, Lee et al. compared RT of LM with a total dose of 50 Gy, which differs from our RT protocol, to surgical excision with no specified minimum follow-up and found total recurrence rates of 29% and 4.2%, respectively [41]. Our cohort focused exclusively on high-risk facial zones (“H-area”), often with convex/concave topographies (e.g., ear, eyelid), where uniform dose delivery and adequate safety margins of >1.5 cm (from the macroscopically visible tumor volume to include all abnormal cells in the periphery of the lesion) are difficult to achieve. This likely contributed to higher recurrence, especially since some lesions were marginal recurrences or had previously relapsed post-surgery. Our data highlight the importance of lesion site, technical challenges in irradiation, and sufficient safety margins in managing LM and LMM with RT.

Primary Cutaneous T- and B-Cell Lymphoma

Cutaneous lymphomas are highly radiosensitive, and RT, or total skin irradiation, is therefore a preferred treatment option as a monotherapy or in combination with systemic therapies [44]. The National Comprehensive Research Center recommends RT for MF and pcBCL for local radiation at different stages [45]. Current RT treatment recommendations are heterogeneous and lack consensus, with low-dose and RT doses above 20 Gy being proposed [44, 46]. Our results show for pcTCL and pcBCL recurrence rates of 0.0%, 8.3%, and 8.3% at 1, 3, and 5 years, which are consistent with the existing literature [44, 47–51].

Adverse Events

Regarding adverse events, our findings of 77.5% acute CTCAE grade 1 and 2 reactions and no grade 4 reactions or serious long-term adverse events are consistent with current evidence showing that mild to moderate CTCAE grade 1 to 3 acute radiation toxicities are common [21, 23–25, 35], but long-term toxicities are rare [16, 23, 25, 35].

Predictors of Recurrence

In our analysis, across all histologies and NMSC relapse rates, they were significantly influenced by linearly increasing CTCAE grades with HR 1.9 and HR 2.3 and for secondary lesions with HR 2.4 and 3.7, respectively. No significant correlations were found for age, sex, high or low total Gy doses, prior treatment, immunosuppression, and field size. Several mechanisms may explain the correlation of higher CTCAE grades with higher relapse rates. Skin toxicity during radiation therapy can impact relapse rates by causing treatment interruptions, giving cancer cells a chance to recover, though this was only the case in seven of our patients, who all still received the required minimal dose treatment. This affected 5 patients with BCC, where four had a break of 1 week and one treatment was stopped earlier due to the patient’s wishes, as well as 1 LM and 1 LMM patient. Additionally, skin toxicity might necessitate dose reductions, and lower doses may be less effective in eradicating cancer cells, potentially affecting local control and increasing relapse likelihood [26, 32]. However, we did not demonstrate a correlation between total Gy doses and relapse rates in our cohort. Furthermore, skin toxicity can trigger an inflammatory response, promoting tumor growth and angiogenesis, thus creating a more favorable environment for cancer cells to survive and spread [52–54]. Lastly, toxicity-induced changes in the skin’s microenvironment can affect immune responses and cell signaling pathways, impacting tumor behavior and recurrence risk. This finding is crucial in practice as radiation dose and side effects are continuously evaluated for therapy continuation or modification. The increased recurrence risk in secondary lesions likely reflects the inherently more aggressive biology of previously relapsed tumors. At our center, recurrent lesions often receive RT when surgery is not feasible, introducing potential selection bias due to their complexity and location. In BCC, field size showed a minor correlation with recurrence (HR 1.1), more pronounced in non-nodular subtypes (HR 1.1, p = 0.014) compared to nodular ones (p = 0.09). However, due to the same HR and only marginal differences in significance, we consider this distinction negligible. LM and LMM showed no significant correlation, probably due to the small number of events. Several studies have reported correlations between RT for NMSC and predictors such as tumor size [18, 26, 32, 36, 55], secondary lesions [26, 32, 55], immunosuppression [55], pathology [55], and age [36]. However, other potential associations are either controversial or have yielded negative results, such as age [26, 35, 55], histology [26, 32], Gy dose [55], gender [32, 35], and location [55]. In summary, while our findings support existing evidence that secondary lesions and treatment-related toxicity are predictive of recurrence, further validation with larger cohorts is needed to strengthen these associations.

Other Treatment Modalities

Mohs surgery is the gold standard for the treatment of both NMSC and melanoma, even in challenging areas such as the “H-area” [56, 57]. Recent technological advances and scientific evidence support the use of RT or brachytherapy for patients who are ineligible for surgery or in cases where preservation of functional and cosmetic integrity, such as the eyelids, nose, or ears, is essential [8–10, 12]. Reviews published in Clinical and Experimental Dermatology have reported lower overall recurrence rates for Mohs surgery than in our study for BCC at 0.2%–6% vs. 10.2%, for SCC at 2.1%–3% vs. 3.2%, and for LMM and LM at 0%–4.3% (with one publication at 33%) vs. 29.7% for LM and 23.3% for LMM [58–60], but no minimum follow-up time or specification of high-risk areas was applied. Tomás-Velázquez et al. [61] reported overall recurrence rates after Mohs surgery in a study of 4,400 patients, with 80% of lesions located in the “H-area” and a mean follow-up of 2.7 years for BCC of 3.5% and for SCC of 11%. A systematic review of 27 articles describing LM and LMM Mohs surgery found a recurrence rate of 1.35% with follow-up ranging from 1 month to 5 years and including lesions from non-high-risk areas, but again without restricting to high-risk areas or longer follow-up durations [62]. Cosmesis and preservation of function are often cited as potential advantages of RT. The only randomized prospective trial comparing surgery and RT for cosmesis favored surgery after 4 years of follow-up [63]. It bears mentioning that the study was published in 1997, and the RT arm included different treatment modalities and RT protocols that have been updated since then. Especially with these caveats, the reported recurrence rates for NMSC are not drastically different between Mohs surgery and RT [59–61]. This underlines the assumption formulated in the current guidelines that RT can be considered a valid alternative to Mohs surgery for inoperable patients or surgery with an unfavorable cosmetic or functional outcome. Existing data suggest that imiquimod is effective for LM and LMM, though evidence is limited due to small study populations and a lack of comparative trials with surgery or RT, making direct comparisons challenging [64]. Nevertheless, several international guidelines recommend imiquimod for LM as an alternative for patients not eligible for surgery or RT. German guidelines propose its use in selected cases of cutaneous metastatic melanoma [5]. Various treatments are available for AK, isSCC, and BD, including cryotherapy, PDT, and topical agents. Due to the limited number of direct comparison studies, robust evaluation remains challenging [11]. Yoo et al. [65] reported for AK recurrence rates of 3.5% for cryotherapy, 6.7% for PDT, and 10.5% for imiquimod. A systematic review reported recurrence rates for isSCC and SCC of 2.0% following electrodessication, 1.6% after cryotherapy, 29.0% for photodynamic therapy, 26.6% for 5-fluorouracil, and 16.1% for imiquimod [66]. These rates are similar to, or in some cases higher than, those observed in our study, highlighting the comparable effectiveness of our treatment approach.

Limitations

The strength of our data set lies in the size of the data set, the long follow-up period, and the inclusion of various tumors whose treatment with surface RT is established. Nevertheless, a retrospective and single-institution setting may lead to a potential selection bias. Despite the large number of patients collectively, the numbers for some subgroup analyses were too small to draw conclusions. Since the majority of lesions analyzed belong to patients with skin type grade 2, the conclusions drawn from this study may not fully represent higher grade skin types.

Conclusion

In summary, this study not only shows an improvement in recurrence rates compared with our previous results but also highlights the importance of longer follow-up. The minimum follow-up period of 2 years after RT is noteworthy as the majority of previous publications show significantly shorter observation periods. The increase in recurrence rate with longer follow-up underlines the importance of long-term aftercare. The study also highlights the efficacy and safety of RT in the treatment of SC, particularly in the high-risk “H-area” of the face, with promising results for NMSC and T/B-cell lymphoma. RT has been validated as a viable alternative to surgery, particularly in cases where surgery is not feasible or when preserving function and cosmesis is a priority. Good skin tolerability and efficacy make it an attractive option in these cases. Higher grade acute toxicity and treatment of recurrence with RT might predict a relapse. The same applies to the field size for BCC subtypes other than nodular BCC. Where possible, safety margins should be more generous.

The optimization of surface RT for different types of melanomas remains a challenge and highlights the need for further research. The results support personalized treatment approaches to improve patient care and outcomes in the treatment of SC.

Acknowledgments

We thank all patients for contributing to the advancement of knowledge in RT. The patients in this manuscript have given written informed consent for publication of their case details. We are very grateful to the radiotherapy staff for their assistance with the data.

Statement of Ethics

The study was approved by the Cantonal Ethics Committee Zurich (Project ID: ID 2023-02147). The patients in this manuscript have given written informed consent to publication of their case details and any accompanying images.

Conflict of Interest Statement

S.T., A.C., L.I., and R.D. have no conflicts of interest related to this work.

Funding Sources

No funding was received for this study.

Author Contributions

S.T. and L.I. contributed to the conception and design of the work; acquisition, analysis, and interpretation of data for the work; and writing of manuscript. A.C. and R.D. contributed to writing the manuscript.

Funding Statement

No funding was received for this study.

Data Availability Statement

Extended data analysis results are presented in the supplementary material (for all online suppl. material, see https://doi.org/10.1159/000548154). The data that support the findings of this study are available from the corresponding author upon reasonable request. The data are not publicly available due to privacy or ethical restrictions.