Updated results of 3,050 non-melanoma skin cancer (NMSC) lesions in 1725 patients treated with high resolution dermal ultrasound-guided superficial radiotherapy, a multi-institutional study

Abstract

Background

Image-guided superficial radiation therapy (IGSRT) using a high resolution dermal ultrasound, is becoming a non-surgical highly effective treatment option for non-melanoma skin cancer (NMSC). In a previous study, we reported results from a multi-institutional study of 1616 patients with 2917 NMSC lesions treated with IGSRT showing a 99.3% rate of local control (LC) with mean follow-up of 16.06 months.

Methods

In this study, we analyze 133 additional lesions from 93 patients, as well as update previous findings with a longer follow-up duration, and perform subgroup analysis and Kaplan-Meier statistics. A retrospective analysis of 1709 patients with 3,050 Stage 0, I, and II NMSC lesions treated from 2017 to 2020 was performed.

Results

With image guidance, lesions received a median of 20 fractions of 50, 70, or 100 kilovoltage(kV) IGSRT. Average follow-up was 25.1 months with a maximum follow up of 65.6 months for the entire cohort. Sixty-eight patients expired, with deaths due to unrelated causes, who had no-evidence of disease (NED) at last follow-up prior to death, leading to Disease-Specific-Survival of 100% (Overall survival was 96%). Absolute LC of 99.2% was achieved in 3,027 of 3,050 lesions with overall absolute LC for BCC, SCC, and SCC-is being 99.0%, 99.2%, and 99.8%, respectively. As of January 2022, no other late complications were found.

Discussion

These updated results demonstrates that IGSRT should be considered a first-line option for the non-surgical treatment of NMSC as it continues to achieve low complication rates while maintaining a high level of LC.

Background

Non-melanoma skin cancer (NMSC) is the most common cancer in the United States (US) [1]. The overwhelming majority of NMSC are comprised of basal cell carcinoma (BCC) and squamous cell carcinoma (SCC), which account for 99% of NMSC. Since NMSC are not reported to national cancer registries, the most current estimate of NMSC occurrences are from 2012, where it was estimated that there was 5.43 million NMSC lesions in the US population and 3.32 million patients treated for NMSC [2]. The incidence of NMSC is expected to be increasing by two to three percent yearly [3]. In 2023, this translates to 6.75 to 7.52 million lesions in 4.13 to 4.60 million individuals. Additionally, individuals often have more than one NMSC lesion and the incidence of subsequent lesions increases after initial lesion diagnosis [3].

NMSC is considered nonfatal and curable due to its slow growth, low recurrence, and rare metastasis [4]. It is typically treated via surgical modalities, including Mohs Micrographic Surgery (MMS). However, NMSC occurs on the head and neck in 70–80% of cases, and surgery can leave scars and cosmetic defects [5]. There are numerous non-surgical treatment modalities with moderate control rates. Image-Guided Superficial Radiotherapy (IGSRT) which incorporates a high resolution dermal ultrasound has emerged recently as a modality that has excellent control rates comparable to MMS and is an attractive non-surgical treatment modality for the treatment of NMSC [5, 6].

From March 2016 to January 2020 a retrospective chart review of 1616 patients from seven out-patient dermatology practices with 2917 early-stage non-melanoma skin cancer (NMSC) lesions treated with image-guided superficial radiotherapy (IGSRT) was conducted. The study’s objective was to assess the efficacy and safety of IGSRT in treating a large number of patients with NMSC lesions. In June 2021, the results were published, which showed a local control (LC) of 99.3% with mean follow-up of 16 months [5]. These results have continued to allow physicians to offer IGSRT to their patients as a treatment for their early-stage NMSC lesions.

This purpose of this analysis is to add an additional 93 patients with 133 lesions from another out-patient dermatology practice, update the previous results with a longer follow-up duration, and provide a subgroup analysis.

Methods

The details regarding patient selection, treatment guidelines, study endpoints, and statistical analyses were discussed in the previous publication, and remain consistent for this updated analysis [5]. Statistical methods were augmented with Kaplan-Meier analysis.

Patients

A retrospective chart review of 1709 patients, including 93 additional patients treated with IGSRT for NMSC between March 2016 and January 2020 at 8 outpatient dermatology practices across the US were analyzed. Lesion characteristics and treatment data at the time of treatment were collected retrospectively during the chart review. Prior to treatment with IGSRT, lesion diagnosis and staging were confirmed via biopsy performed by a dermatologist at each practice.

The study cohort comprised of 756 female and 953 male patients (mean age 74 years [SD ± 10.6]) with 3,050 lesions treated from years 2016 to 2020. All lesions included in this study were stage Tis, T1, or T2 and each lesion was regarded as an individual cancer lesion. Based on the American Joint Committee on Cancer (AJCC) 8th Ed. Cancer Staging Manual, patients exhibited no clinical evidence of distant disease or regional lymph node involvement (N0 and M0) at presentation [7]. The AJCC 8th Ed. staging is specific to cutaneous SCC of the head and neck, but for consistency these same criteria were applied to all BCC, SCC, and squamous cell carcinoma in-situ (SCC-IS) lesions in this study.

Initially, data and follow-up intervals were extracted manually from written and electronic medical records. Subsequent updates to patient data and follow-up were accessed via an electronic algorithmic analysis supplied by a healthcare data company (Sympto Health, Inc.).

The study protocol was reviewed and determined to be exempt from IRB approval by an IRB committee (Western Institutional Review Board (WIRB)-Copernicus Group) under 45 CFR 46.104 (d) [4]. The information obtained was recorded by the investigator in such a manner that the identity of the human subject could not be readily ascertained directly or through identifers linked to the subjects, the investigator does not contact the subjects, and the investigator will not re-identify subjects. Any health information used in this study has been de-identifed. This study was performed in compliance with the pertinent sections of the Helsinki Declaration and its amendments, as well as the “Common Rule” established in the Federal Policy for the Protection of Human Services. All methods were carried out in accordance with relevant guidelines and regulations. Informed consent was received from all patients prior to treatment.

Treatment

Treatment characteristics are summarized descriptively in Table 1. All lesions received a median of 20 fractions (range 13–30 fractions) of 50, 70, 100, or mixed (i.e. 50/70, 70/100, 50/70/100) kilovoltage (kV) IGSRT. Ultrasound imaging allowed for energy selection before and during treatment. The median dose per fraction was 259 cGy (mean 259.2 cGy [SD +/- 10.4]). The median total dose received was 5184 cGy (mean 5216.0 cGy [SD +/- 222.3]). Median treatment duration was 7 weeks (mean 6.9 weeks [SD +/- 1.5]). Median time dose fractionation (TDF), a numerical value used to represent biologic dose effect, was 90 (mean 90.8 [SD +/- 4.8]). During IGSRT treatment, patients were evaluated at each session clinically and with ultrasound imaging and treatment dosing adjustments were made if necessary.

Table 1 Treatment characteristics for tumor treatment with Image-Guided superficial radiation therapy (IGSRT)

Follow-up

Once treatment ended patients were evaluated 2–6 weeks after treatment completion and every 1–6 months thereafter.

Study endpoints

This study’s main outcome measure was lesion recurrence assessed by raw local control (LC) rate and Kaplan-Meier (KM) LC analysis with follow-up through January 2022. LC was assessed at follow-up visits clinically by dermatology providers. IGSRT’s safety was assessed by Radiation Treatment Oncology Group (RTOG) toxicity, which was prospectively documented in the charts routinely after every 5-fractions. RTOG data was extracted and documented as the highest RTOG grade from the entire treatment course. Certain study locations did not maintain a procedure to record RTOG toxicity grades for IGSRT patients in the initial study year(s), accounting for the missing values on this measure of safety/toxicity. Overall survival was calculated as the percent of patients alive at the time of analysis, including death from all causes. Disease specific survival was calculated as percent of patients who did not die from NMSC during the study period.

Statistical methods

Treatment information, patient demographics, tumor characteristics, and RTOG toxicity were summarized descriptively. Treatment information included total dose received (cGy), number of fractions, dose per fraction (cGy), energy (kV), treatment duration, and time dose fractionation (TDF)^A. Duration of follow-up was defined as the last date of follow-up minus treatment completion date plus one day and was then converted to weeks and months.^B Patient demographics included total patients, age at first treatment, gender, follow-up interval, and deaths. Tumor characteristics included number of lesions, lesion histopathology, lesion size, lesion recurrences, and lesion location. Lesion histopathology separated by tumor stage and lesion histopathology separated by energy treatment were also provided. Acute toxicities were graded with RTOG (Radiation Treatment Oncology Group) toxicity scoring [8].

Categorical outcomes were reported with percentages and frequency, while the number of observations (n), mean, standard deviation, median, interquartile range, and range were reported as summary statistics for continuous outcomes. Missing data were not included. SAS^® Studio was used for all analyses and verification with R Studio.

All patients were included in the intent-to-treat analysis. Overall LC and KM LC were reported. KM LC was reported based on lesion histology was compared using log-rank tests.

Subgroup analyses of patients with ≥ 12-month follow-up, ≥ 24-month follow-up, and breakdown by histology (BCC, SCC, SCC-IS) were performed. Number of lesions, lesions recurred, absolute local control, overall 5-year KM LC, and 5-year KM LC by histology subtype were reported.

For this updated report, the results are based on all the information received by January 14th, 2022.

Results

Patient demographics and tumor characteristics

Table 2 details patient demographics. Median age at first treatment was 74 years (mean 74.0 years [SD +/- 10.6 years]). There was slight predominance of male patients at 55.8% (953 patients) with 756 female patients (44.2%). Median follow-up was 25 months (mean 25.1 months [SD +/- 17.1 months]).

Table 2 Patient demographics at time of lesion treatment with Image-Guided Superficial Radiation Therapy (IGSRT)

Tumor characteristics are described in Table 3. All lesions were stage Tis, T1 or T2 with BCC representing 47.9%, followed by SCC (30.3%) and SCC-IS (21.2%). Median lesion size was 1.0 cm (mean 1.2 cm [SD +/- 0.7 cm]). Lesion histopathology separated by lesion stage is depicted in Table 4. Lesion histopathology separated by treatment energy is depicted in Table 5. Lesion histopathology separated by lesion location is depicted in Table 6.

Table 3 Patient demographics at time of lesion treatment with image-guided superficial radiation therapy (IGSRT)

Table 4 Lesions separated by histopathology and tumor stage

Table 5 Lesions separated by histopathology and energy (kV) treatment

Table 6 Lesions separated by histopathology and lesion location

Table 7 Acute toxicities RTOG (Radiation Treatment Oncology Group) grades for lesions treated with IGSRT (Image-Guided Superficial Radiation Therapy) [8]

As of January 2022, overall survival was 96%, with 1,641 of 1,709 patients alive and 68 patients expired and disease specific survival was 100%.

Outcomes

Absolute local control (LC)

At a mean follow-up of 25.1 months, 3,027 of 3,050 lesions achieved local control with 23 recurrences (15 BCC, 7 SCC, 1 SCC-IS), resulting in an overall absolute LC of 99.2%. Absolute local control for BCC, SCC and SCC-IS were 99.0%, 99.2%, and 99.8%, respectively.

Kaplan Meier local control (KM LC) at ≥ 5 years

Overall 5-year KM LC was 98.8% and unchanged at maximum follow-up of 65 months (Fig. 1). KM LC for BCC was 98.2% at 5-years. KM LC for SCC was 99.0% at 5 years (Fig. 2). KM LC for SCC-IS was 99.7% at 5-years. Comparison of KM LC between histologic subtypes was not statistically significant using log-rank tests (p = 0.0630, alpha = 0.05).

Fig. 1

Kaplan-Meier Local control (KM LC) for all 3,050 lesions treated with IGSRT (Image-Guided Superficial Radiation Therapy). Dots represent censored events. (Note: Y-axis starts at 60%). NAR = number at risk

Fig. 2

Kaplan-Meier Local control (KM LC) by histologic subtype of 3,032 lesions (Basal Cell Carcinoma (BCC) = 1,460 lesions, Squamous Cell Carcinoma (SCC) = 924 lesions, SCC In-Situ (SCC-IS) = 648 lesions) treated with IGSRT (Image-Guided Superficial Radiation Therapy). Dots represent censored events. (Note: Y-axis starts at 60%). NAR = number at risk

Treatment tolerance

All lesions had minimal or mild toxicity (RTOG 0, 1, 2) with only 20 lesions having severe or significant toxicity (RTOG 3, 4) based on the RTOG toxicity scoring [8] (Table 7). No additional late complications were found to date as of January 2022.

Subgroup analysis

Subgroup analysis of lesions with follow-up greater than or equal to 12 months and 24 months is summarized in Table 8 and Table 9. Lesions with follow up less than 12 months and less than 24 months were excluded respectively from subgroup analysis.

Table 8 Subgroup analysis comparing absolute local control (LC) and overall 5-year kaplan Meier local control (KM LC) for all lesions, lesions with follow-up $\:\ge\:$ 12-months, and lesions with follow-up $\:\ge\:$ 24-months

Table 9 Subgroup analysis comparing 5-year Kaplan Meier local control (KM LC) separated by histopathologic subtype (BCC, SCC, SCC-IS only) for all lesions, lesions with follow-up ≥ 12-months, and lesions with follow-up ≥ 24-months

Subgroup with ≥ 12-month follow-up

A total of 2174 lesions had a follow-up time of greater than or equal to 12 months, representing 71.3% of all lesions. A total of 9 lesions recurred. Resulting in an overall absolute LC of 99.6% and overall KM LC was 99.3% at 5-years. KM LC for BCC lesions was 98.9%, for SCC lesions was 99.5%, and for SCC-IS lesions 99.7% at 5-years (Fig. 3). Comparison of KM LC between histologic subtypes was not statistically significant (p = 0.5776, alpha = 0.05).

Fig. 3

Kaplan-Meier Local control (KM LC) by histologic subtype of 2,174 lesions treated with IGSRT (Image-Guided Superficial Radiation Therapy) with a follow-up of greater than or equal to 12 months. Dots represent censored events. (Note: Y-axis starts at 60%). NAR = number at risk

Subgroup with ≥ 24-Month Follow-up

A total of 1615 lesions had a follow-up time of greater than or equal to 24 months, representing 53.0% of all lesions. A total of 4 lesions recurred. Resulting in an overall absolute LC of 99.8% and overall KM LC was 99.6% at 5-years. KM LC for BCC lesions was 99.3%, for SCC lesions was 99.8%, and for SCC-IS lesions 99.7% at 5-years (Fig. 4). Comparison of KM LC between histologic subtypes was not statistically significant (p = 0.9301, alpha = 0.05).

Fig. 4

Kaplan-Meier local control (KM LC) by histologic subtype of 1,615 lesions treated with IGSRT (Image-Guided Superficial Radiation Therapy) with a follow-up of greater than or equal to 24 months. Dots represent censored events. (Note: Y-axis starts at 60%). NAR = number at risk

Discussion

This analysis updates the results of a large multi-institutional analysis of the utilization of image-guided superficial radiation therapy (IGSRT) for the curative treatment of early-stage NSMC. Our updated results support the initial findings and this analysis with longer follow-up has a similar absolute LC rate to previously published results. Absolute LC rate of 99.2% (mean follow-up 25.1 months) versus the previously reported local control rate of 99.3% (mean follow-up 16.1 months), show that the absolute LC remains substantively unchanged with additional patients and longer follow-up [5].

Currently there are few modern studies on the use of IGSRT or even non-image guided superficial radiation therapy (SRT) for the treatment of early-stage NMSC. An extensive literature search conducted by the American Society for Radiation Oncology (ASTRO) from 1988 to 2018 showed that there were limited randomized controlled trials (RCT) on radiotherapy [9]. Additionally, our own PubMed search revealed that there are no published studies looking at the use of IGSRT for the curative treatment of NMSC, besides the previous analysis that this manuscript updates and an additional two manuscripts published by the authors [4,5,6]. One of the more recently published analyses by the authors that utilized IGSRT for the curative treatment of NMSC lesions demonstrated an overall absolute LC of 99.7% and a 5-year KM LC of 99.4% for 1899 NMSC lesions in 1243 patients [6]. The study by Tran et al. utilized a different cohort of patients having similar histology and stages of lesions showing the results were comparable to our current analysis which updated the 2021 study demonstrating that IGSRT yields consistently excellent results.

Our updated results have a mean follow-up of greater than 2 years with approximately ¾ of the lesions analyzed having a follow-up of greater than or equal to 1 year (71.3%) and approximately ½ of the lesions analyzed having a follow-up of greater than or equal to 2 years (53.0%). In the previously published manuscript, we compared our results to a study by Cognetta et al. from 2012 that utilized non-image guided SRT and is one of the more modern studies using SRT to treat NMSC with curative intent. However, at that time only 55% of our patients had a follow-up greater than or equal to 1 year. The study by Cognetta et al. reported a 2-year recurrence rate for BCC and SCC lesions at 2.0% and 1.8%, respectively [10]. Thus, the 2-year absolute LC would be 98.0% and 98.2% for BCC and SCC, respectively. Our results continue to be consistent and exceed those reported by Cognetta et al.

The likelihood of local recurrence for early-stage NMSC is small after two years. A majority (70–80%) of SCC are thought to recur within the first 2 years after primary tumor treatment with 95% of SCC recurrences occurring within 5 years [11]. Whereas for BCC, approximately 50% of recurrences occur within the first 2 years after treatment of the primary tumor with 80% of BCC recurrences occurring with 5 years [12]. Although some reports suggests that BCC recurrences primarily occur within the first 4–12 months [11]. Our results corroborate with the latter.

Additionally, we previously compared our preliminary results to another SRT study by Hernández-Machin et al. [13]. Hernández-Machin et al. reported 5-year absolute LC rates of 94.4% for BCC lesions and 92.7% for SCC lesions. With these results, the authors recommended SRT as a first line treatment option for NMSC. Our estimated 5-year KM LC remains higher at 98.8%. This is consistent with our belief that image guidance is most likely responsible for the improved outcomes.

Our improved absolute LC results over those of other studies that utilized SRT alone or external beam radiation therapy (XRT) without image guidance, is statistically significant. Two recent studies, a metanalysis and a logistic regression analysis, compared IGSRT studies to large well run SRT/XRT studies without image guidance. Both analyses found that the LC rates for the IGSRT studies were statistically superior to the non-image-guided SRT/XRT studies [14, 15]. The image-guided component of SRT utilizes a 22 megahertz (MHz) dermal ultrasound with color doppler to visualize the superficial depth of the skin. This image-guided component allows the physician to visualize the lesion prior to, during, and after treatment. During treatment, the physician can make any necessary adjustments to the treatment regimen.

As IGSRT gains more popularity, the role of this modality in the adjuvant setting may be considered in the future [16]. The definitive treatment of NMSC at an early stage using this and other modalities can decrease delay in seeking treatment that may eventually lead to less need to manage difficult to treat NMSC [17].

IGSRT treatment is well suited for lesions that are not deeper than 6 mm due to the low penetrance of the superficial 50–100 kV energy. It also has limitations in covering surfaces with substantial (> 1 cm) concavity or convexity where a surface flap or mold may have better coverage [18]. Further, IGSRT is not recommended for post-radiotherapy recurrences. There have been reports of salvage brachytherapy for post-radiation recurrences, however these show low success rates [19], and no lesions fitting this category were included in the study.

Limitations of this study include the retrospective study design and moderately short median follow up of slightly over two years. As this was a retrospective study, follow up was maintained as part of routine clinical care and was not prospectively maintained. Less patient data was available up to 5 years, which is commonly reported in oncology, however, since most NMSC recurrences occur within the first 2–3 years, as noted above, we feel the data is unlikely to change substantially with additional follow up.

Finally, although there are currently no randomized, controlled trials directly comparing IGSRT to MMS, a recent study shows that IGSRT may be comparable or possibly superior in terms of local control at 2 years [14, 15, 20]. Further research with randomized controlled trials to directly compare MMS to IGSRT can be considered, although may be difficult to achieve as patients are unlikely to accept randomization between surgical and nonsurgical modalities.

Conclusions

Our large patient sample from multiple institutions support that this office-based technology, IGSRT, is feasible, effective, safe and easily tolerable in an out-patient setting. Almost all lesions had minimal or mild toxicity scores (RTOG 0, 1, 2). Our updated results support that IGSRT continues to offer excellent local control rate. This study shows that IGSRT continues to be an effective non-invasive, non-surgical treatment modality for NMSC with its major advantages of avoiding the potential pitfalls of surgery clinically, cosmetically, physiologically and psychologically. These updated results validate our prior statement that IGSRT should be considered among primary treatment options presented to patients for treatment of their early-stage NMSC lesions.