Practical Guideline for Prevention of Patchy Hair Loss following CyberKnife Stereotactic Radiosurgery for Calvarial or Scalp Tumors: Retrospective Analysis of a Single Institution Experience

Hair is a major aspect of body image, a symbol of life and identity, and it plays an important role in social communication. It can reflect social class, sex, profession, religious belief, and both social and political convictions [1]. Hair follicles cycle through three phases: anagen (regeneration), catagen (degeneration), and telogen (rest), with 90% of their cycle spent in the anagen phase [2]. The high dose of radiation beams of stereotactic radiosurgery (SRS) induces cell death by damaging the DNA and potentially disrupting several proliferative signals for hair follicle stem cells. Hair follicles are most sensitive to radiation during the anagen phase, ultimately resulting in temporary or potentially permanent hair loss [2, 3].

Alopecia, hair loss, is a recognized phenomenon previously reported as the most feared side effect by up to 58% of women preparing for cancer treatment [4]. It has been reported that alopecia is significantly associated with lower quality of life, including physical, psychological, and social well-being [4]. Alopecia can produce anxiety and depression and often cause patients to subsequently refuse treatment [5]. Although it is not a life-threatening complication, alopecia causes negative impacts which leads to the need for investigating methods to prevent it.

Due to the steep radiation dose falloff outside tumors in SRS, radiation-induced patchy alopecia is not commonly reported in the group of patients with intracranial target, treated with SRS. However, the possibility of radiation-induced patchy alopecia becomes significantly higher for those patients undergoing SRS to the calvarium and/or scalp which delivers higher radiation dose to the hair follicles nearby. This is also one of the most frequently asked questions posed by patients in clinics while discussing SRS. However, studies related to the topic of patchy alopecia following SRS are still greatly limited, and still there are no clear and practical guidelines to use in the clinical field. Our objective is to retrospectively review the incidence and outcome of patchy alopecia following SRS for patients with calvarial or scalp lesions and to investigate how we may effectively and practically avoid this problem in the future.

We retrospectively reviewed the IRB-approved brain tumor database from our institution between January 2014 and September 2022. Our analysis focused on a cohort of 3,620 patients who underwent CyberKnife SRS during this timeframe. Written informed consent from participants was not required in accordance with local/national guidelines. The clinical and radiological outcomes of patients with calvarial or scalp lesions who underwent CyberKnife SRS at our institute were extracted and analyzed. Within this group, we identified a total of 20 patients (median age 66 years, range: 17–80) with a total of 30 lesions (range: 1–6 per patient). Detailed patient demographics are described in Table 1. The patients were 9 males and 11 females. Eighteen patients had metastatic tumors, and 2 patients had meningioma (one WHO grade 2 and one WHO grade 3 meningioma) invading the calvarium. Among those 30 lesions, 29 lesions were located in the calvarium, and only a single lesion was located on the scalp. As most of the patients had metastatic disease, chemotherapy agents were reviewed whether the agent had a radiosensitizing effect, assuming that it can have a potential increase of radiation effect on the scalp area adjacent to the target. Chemotherapy was performed either before or during the SRS treatment on 26 lesions, and 11 lesions had radiosensitizing chemotherapy agents, such as temozolomide, cisplatin, carboplatin, docetaxel, etc. (Table 2).

All patients underwent CyberKnife (Accuray, Inc., Sunnyvale, CA, USA) treatment as previously described [6]. The median SRS target volume was 9.85 cc (range, 0.81–110.7 cc). The median prescription dose was 27 Gy (range, 16–40 Gy), with a maximal dose of 34.53 Gy (range, 20–50 Gy), administered in 1–5 fractions (median, 3). The most common dose/fractionation schemes were median marginal dose of 27 Gy (range, 24–30 Gy) in 3 fraction (12 lesions), 30 Gy (range, 25–40 Gy) in 5 fractions (10 lesions), and 18 Gy (range, 16–22 Gy) in a single fraction (8 lesions). The median prescribed isodose line was 76% (range, 70–82%), with a median conformity index of 1.23 (range, 1.0–1.48).

We defined radiation-induced alopecia as any degree of hair loss following SRS completion. We focused only on a patchy hair loss on the area over the SRS treatment. For further analysis, overlying scalp areas on top of the target lesions were contoured retrospectively via Accuray Precision^® CyberKnife treatment planning system (Accuray, Inc., Sunnyvale, CA, USA), and the radiosurgical data of the overlying scalp was collected for statistical analysis (Fig. 1). The median follow-up was 15 months (range: 1–57). For the patients with short follow-up of less than 12 months, alopecia was evaluated at the time of the last follow-up.

The biologically effective dose (BED) delivered to each lesion was calculated using the linear-quadratic (LQ) model and compared [7‒9]. The α/β ratio in the equation is used to model the repair capability of a certain type of cell. Based on previous literature, an α/β ratio of 10 Gy was used for metastatic brain tumors [10, 11], and a ratio of 3 Gy was used for meningiomas [12]. An α/β ratio of 10 Gy was used for the normal scalp overlying the tumor [13]. The BED was converted to single-fraction equivalent dose (SFED), using LQ model with α/β ratio of 2 Gy, for further comparison [14‒17]. The median BED was 51.30 Gy (range, 37.50–146.67 Gy) and the median SFED was 18.19 Gy (range, 15–22 Gy) for the treated lesions. Maximal dose was used instead of mean dose for the overlying scalp as the contour of the overlying scalp was subjective. The median BED was 50.91 Gy (range, 7.97–94.84 Gy) and the median SFED was 18.11 Gy (range, 5.23–26.20) for the overlying scalps.

For statistical analysis, categorical variables were evaluated through frequencies and percentages; means, medians, and ranges were provided for continuous variables. We conducted Student’s t test and Fisher’s exact test for statistical comparisons. Statistical analysis was performed using SPSS Statistics (Version 29, IBM^® SPSS^® Inc., Chicago, IL, USA) and Microsoft Excel (Version 2209 Build 16.0.15629.20208, Microsoft^® Excel^®, Redmond, WA, USA). The tumor and overlying scalp volumes were contoured and measured using the Accuray Precision^® CyberKnife treatment planning system (Accuray, Inc., Sunnyvale, CA, USA).

Out of the 30 lesions observed, 11 had patchy hair loss over the target area, while 19 had no hair loss at all. The patient with scalp mass did not present any hair loss during the 57-month follow-up. All 11 cases of patchy alopecia were located in the calvarium, and patients reported that their hair grew back to normal within a year. Although the recovery period was uncertain for most patients, 63.6% of them recalled that their hair loss recovered within 3–6 months after SRS treatment. In comparison of those two groups which revealed patchy alopecia and no alopecia, there were no statistical significances between age, gender, pathology, and location of the tumor. Additionally, there was no statistically significant difference found between patients who received radiosensitizing chemotherapy agents versus those who did not (p = 0.12).

Radiosurgical data of the target lesions, including the volume of the lesions, marginal dose, maximal dose, fractions, isodose line, conformality index, BED, and SFED, were statistically compared and analyzed between the two groups (patchy alopecia vs. no alopecia) (Table 2). All variables from the treated lesions, such as volume, marginal dose, maximal dose, fractions, isodose line, conformality index, BED, and SFED, did not demonstrate a statistical significance. The radiosurgical data of the manually contoured overlying scalp of the treated lesions were analyzed as well. No statistically significant differences in volume, fraction, mean radiation dose, or maximal radiation dose were found between those two groups. However, the difference of BED and SFED (61.0 vs. 46.4 Gy and 20.0 vs. 16.6 Gy, respectively) was statistically significant (p < 0.05) (Table 2). Both BED and SFED were significantly higher in the group with patchy alopecia. The threshold between those groups for each of the indicators was defined using confidence intervals, and the groups were statistically compared via Fisher’s exact test (Table 3). As described in Table 3, the BED larger than 60 Gy and SFED larger than 20 Gy demonstrated significantly higher probability of temporary patchy alopecia following the SRS treatment of the calvarial lesion compared to those indicators below each threshold. Patients with a BED of the overlying scalp higher than 60 Gy or SFED higher than 20 Gy exhibited 9.3 times higher odds of having temporary alopecia than did the patients with it lower than 60 Gy and 20 Gy, respectively. None of the patients presented other adverse radiation effects, such as scalp irritation, ulceration, or necrosis. The 1-year local tumor control rate was 93.3% for all calvarial and scalp lesions. The recurrent lesions required surgical resection at the time of recurrence.

Radiation-induced alopecia is a well-understood phenomenon contemporary to the discovery of X-rays since 1896, when Daniel [18] reported radiation-induced alopecia in the human scalp, soon after the introduction of X-rays by Roentgen. Hair follicles are one of the most radiosensitive organs of the body. In the 1950s, it was reported that alopecia began 2–3 weeks after radiation, and the new hair began to regrow within 2–3 months after completion of radiotherapy [19].

Compared to whole brain radiotherapy, chances of radiation-induced alopecia are significantly less after SRS due to sharp dose falloff. However, the risk of alopecia related to the exposure of hair follicles to radiation becomes significantly higher when the lesion is located in the calvarium and/or scalp, particularly when single-fraction high-dose SRS is delivered. A prospective study focused on short-term adverse effects following GammaKnife SRS reporting new-onset patchy alopecia by 1 month in nearly 10% of a total of 76 patients with diverse diagnosis [20]. However, this side effect seemed to be more closely related to the pin-site issue from the head frame. The same group reported a retrospective study of 21 patients with calvarial and skull base metastases, and 1 patient out of 7 calvarial lesions developed a patchy alopecia over the treatment field [21]. Detailed information as to this side effect was not provided.

Currently, various protocols refer to pre-existing organ at risk guidelines, such as AAPM TG 101 [22], Timmerman et al. [23], the UK Consensus document [24], and NRG-BR001. However, the dose constraints suggested in these guidelines included skin as a serial tissue and ulceration as the end point but did not include scalp specifically nor alopecia as the end point.

Several other studies reported that the radiation dose caused permanent hair loss following radiation. However, it varies widely. International Commission on Radiological Protection Publication 85 has stated that permanent alopecia occurs at 7 Gy (single fraction) [25]. Shakespeare et al. [26] reported that 5% (median, range 0–80%) risk of permanent alopecia with a dose of 36 Gy and 15% (median, range 5–100%) risk of permanent alopecia with a dose of 45 Gy for conventional radiotherapy (2 Gy per fraction).

Currently, the most widely used guideline for prevention of radiation-induced alopecia in the scalp after SRS is based on the report from Lawenda et al. [27, 28] The authors studied 26 patients with 61 lesions, using the follicular dose to analyze the dose-response relationship. The average depth of human scalp follicles is approximately 4.5 mm, according to various literature sources [29‒31]. As a result, the authors suggested that the follicle dose should be measured and calculated at this depth of 4.5 mm. This calculation involved adding the entrance and exit doses for each radiation field that crossed the scalp area of interest. The authors reported that D₅₀, the follicle dose at which 50% of the patients developed permanent alopecia, was estimated to be 43 Gy. However, this method appears impractical for use in a clinical setting when planning SRS treatment.

Here we propose a more practical method to reduce the risk of radiation-induced patchy alopecia following SRS treatment. We suggest contouring the overlying scalp separately and measuring the maximal radiation dose to the overlying scalp when the target is located in the calvarium or scalp nearby the hair follicles. Using α/β ratio of 10 Gy for the normal scalp overlying the tumor and maximal dose to the overlying scalp instead of marginal dose, BED and SFED can be easily calculated. Based on our study, we propose measuring and calculating these modified BED and SFED for the overlying scalp which demonstrated statistical significance between patient who had radiation-induced alopecia and those who had no alopecia. We suggest limiting the BED under 60 Gy and SFED under 20 Gy for the overlying scalp to prevent even temporary patchy alopecia. Modifying the contour, reducing the marginal dose, or hypofractionation can be considered options to decrease the BED and SFED.

Our study has several significant limitations. The retrospective design and small cohort size prevent us from drawing definitive conclusions. The fragility of the scalp may be affected by different pathologies such as brain metastases and meningioma since the tumor’s growth pattern can vary. Moreover, the follow-up duration varied widely due to different pathologies, which could affect the patient’s clinical status during SRS treatment planning. Due to the retrospective design, we could not conduct a thorough investigation on diverse and personalized chemotherapy regimens, although they could impact the outcome. Instead, we grouped chemotherapy regimens into radiosensitizing agents and non-radiosensitizing agents, but no significant difference was found in the incidence of patchy alopecia in our cohort. Other studies have reported patchy alopecia following SRS treatment, but no clinically practical guideline exists to avoid it specifically for calvarial or scalp tumors. To confirm the effectiveness of our proposed guideline, larger multicenter studies would be helpful.

Additionally, we acknowledge that there is ongoing controversy and debate regarding the applicability of the LQ model in radiosurgery, specifically for single-dose treatments. While we understand that alternative models, such as the incomplete repair model or the multitarget model, have been proposed for estimating the biological effects of single-dose radiosurgery, it is important to note that the LQ model has been widely used in the field of radiation oncology and has provided valuable insights into treatment planning and outcomes. In our study, we chose to utilize the LQ model based on established practices and the available literature at the time of data collection and analysis. While we acknowledge the controversies surrounding its use in single-dose radiosurgery, our objective was to assess the outcomes within the framework of the prevailing standard practices.

SRS is an effective and safe treatment modality for patients with calvarial or scalp mass with respect to local control and prevention of the radiation-induced patchy alopecia. Our data suggest that limiting the BED under 60 Gy and SFED under 20 Gy for the overlying scalp can be helpful in preventing radiation-induced patchy alopecia for SRS treatment of the calvarial or scalp mass nearby the hair follicles. Larger prospective studies are encouraged to validate the guidelines for SRS treatment for calvarial or scalp lesions.

Approval for this study protocol was obtained from the Stanford Institutional Review Board (IRB), with the assigned approval number 3910. Written informed consent from participants was not required in accordance with local/national guidelines. The handling of clinical information adheres to the principles outlined in the Helsinki Declaration.

The authors affirm that the content of this article was developed without any commercial or financial relationships that could be considered potential conflicts of interest.

No external funding was received for this research.

Conception and design: Park and Meola. Acquisition of data: Park, Tayag, Ustrzynski, and Emrish. Analysis and interpretation of data and statistical analysis: Park, Soltys, and Meola. Drafting the article: Park. Critically revising the article: Park, Soltys, Chang, and Meola. Reviewed submitted version of manuscript: Park, Marianayagam, Yener, Tayag, Ustrzynski, Emrish, Pollom, Soltys, Chang, and Meola. Approved the final version of the manuscript on behalf of all authors: Meola. Administrative/technical/material support: Pollom, Soltys, Chang, and Meola. Study supervision: Park, Chang, and Meola.