EVA-Scalp: Evaluation of Patient Satisfaction with a Scalp Cooling Device to Prevent Chemotherapy-Induced Alopecia in Breast Cancer Patients

Abstract

Purpose

Scalp cooling (SC) offers a chance to reduce hair loss (HL), but patient satisfaction, the effect on well-being, as well as patient selection criteria have not been sufficiently assessed yet.

Methods

In the EVAScalp trial, SC was offered to 70 breast cancer patients who received chemotherapy between November 2015 and September 2018. For SC, the Paxman-Orbis-II System was used. Satisfaction was measured by a questionnaire evaluating the level of depression with the WHO-5 well-being index (WHO-5) plus questions addressing the patient's experiences and side effects using the SC device. To evaluate efficacy, documentation by photo, by a physician, and by an HL-diary was conducted.

Results

Regarding efficacy, a significant difference between chemotherapy regimens is seen. Anthracycline-based therapies led to a stop of SC in 71% of the patients, whereas taxane-based therapies without anthracyclines were associated with a high acceptance of SC, and 88% of patients with paclitaxel-based therapies continued SC throughout their chemotherapy. Overall, only 7.69% of the patients stopped because of side effects. As an indicator for quality of life, WHO-5 was higher (65.8%) in patients with successful SC compared to in patients who stopped SC because of HL or side effects (only 53.0%). The majority of patients (82.22%) with successful SC would recommend SC to other patients.

Conclusions

Patients tolerated SC as long as HL was successfully prevented. The well-being of patients with successful SC was significantly higher than that of patients who stopped SC prematurely. In general, SC is a promising approach and improves patient well-being, but there are still limitations to its utility depending on the chemotherapy regimen used.

Introduction

Chemotherapy-induced alopecia (CIA) is one of the most common side effects after chemotherapy treatment in breast cancer (BC) patients and was first described by Otto Braun-Falco in 1961 [1]. CIA affects quality of life (QoL) significantly [2, 3]. Almost half of all patients (47%) experience CIA as the most traumatic side effect of their treatment. Up to 8% would even reject chemotherapy and thus risk an inferior outcome because of their fear of hair loss (HL) [4]. Patients with CIA suffer from depression and fear more often than patients without CIA [5]. Taken together, CIA is one of the biggest stigmata in patients undergoing chemotherapy treatment [6].

Cooling reduces cytostatic effects on keratinocytes in vivo [7]. The rationale for the development of scalp cooling (SC) was that cooling leads to vasoconstriction and therefore smaller amounts of chemotherapeutic agents reach the hair follicles. This hypothesis led to the first studies with SC in the early 1990s, in which the cooling was carried out using cool pads during chemotherapy. The pads warmed up fast and were not able to provide continuous cooling. The results were inconsistent, and the investigators argued that SC had no place in the prevention of CIA [8]. Modern SC devices use caps which are placed on the patient's head and have a continuous flow of cooling medium. They are controlled by a sensor which measures temperature continuously. With these techniques, modern devices significantly reduced CIA and consequently the need for wigs [9]. Best effects were seen in patients who underwent taxane-based chemotherapies without anthracyclines. Up to 95% of paclitaxel-treated patients kept their hair and did not use a wig in the analysis of the Dutch Scalp Cooling Registry. These findings regarding the efficacy of modern SC were confirmed by other studies [10, 11, 12, 13]. SC does not lead to an increased risk of scalp metastases in BC patients as this risk amounts to about 1% with or without SC [14, 15]. Furthermore, SC has been shown to be cost effective by reducing the use of wigs by up to 40% [16]. To date, there are only few evaluations regarding the effects of SC on QoL in BC patients [17].

Materials and Methods

The EVAScalp trial was offered to 70 BC patients who received chemotherapy between November 2015 and February 2018 at the Breast Center, Department of Gynecology and Obstetrics and CCC Munich and the Gynecological Office Dr. Wolfgarten, Beta­klinik. Inclusion criteria were confirmed BC and an indication for neoadjuvant, adjuvant, or palliative chemotherapy. Exclusion criteria were alopecia as assessed by the treating oncologist or current scalp metastasis. For SC, the Paxman-Orbis-II-System was used (Fig. 1) according to the manufacturer's instructions. The system is based on a continuous flow of cooling medium through a cool cap. A sensor regulates the temperature to ensure that the scalp does not get warmer than 22°C. Therefore, a pre-cooling time (30 min) before the application of the chemotherapy cycle as well as a post-infusion cooling time (60–120 min) is needed. Different post-infusion cooling times were used according to the manufacturer's instructions for different chemotherapy regimens. The hair situation and SC efficacy were assessed by the treating physician using a modified scale for alopecia, as described by Friedrichs and Carstensen [11], adapted to a visual scale (Fig. 2) from score 0 (no HL) to 4 (total HL). The hair situation was assessed before treatment and during every cycle using a hair diary, where the patients were supposed to enter their daily hair score (Fig. 2) analogous to the adapted visual scale. Furthermore, it was assessed by a physician at the beginning and at the end of the treatment using the same scale. Patients were allowed to quit SC because of side effects or HL. Success of SC was defined as CIA of grade 2 (HL between 25 and 49%) or lower at the end of the treatment and as the use of SC during the complete treatment. Quitting SC because of side effects, HL, or CIA grade 3 or 4 was defined as unsuccessful treatment. An overall success rate of SC during the study was compiled comparing patients with successful SC (group A) with patients with unsuccessful SC (group B). The different chemotherapy regimens were evaluated by assessing the associated SC success rate.

Fig. 1.

Examples of successful and unsuccessful SC. A patient before (A1) and after (A2) 12 cycles of paclitaxel, and a patient before (B1) and after (B2) 2 cycles of EC.

Fig. 2.

Different graduation of alopecia according to WHO, Friedrichs and Carstensen [11], and the adapted visual scale as used in EVAScalp.

Each patient was handed a questionnaire at each treatment cycle. The questionnaire evaluated the level of depression by using the WHO-5 well-being index (WHO-5) [17]. WHO-5 is among the most widely used tools to assess subjective psychological well-being (Table 1). The evaluation of well-being can be expressed as a percentage by adding the 5 values of WHO-5 and multiplying the result by 4. The group with successful SC (group A) was compared with the group with unsuccessful SC (group B) regarding WHO-5.

Table 1.

WHO-5 in the questionnaire

Over the last 2 weeks …	All of the time	Most of the time	More than half of the time	Less than half of the time	Some of the time	None of the time
I felt cheerful and in good spirits	□	□	□	□	□	□
I felt calm and relaxed	□	□	□	□	□	□
I felt active and vigorous	□	□	□	□	□	□
I woke up feeling fresh and rested	□	□	□	□	□	□
My daily life has been filled with things that interested me	□	□	□	□	□	□

Assessment is made by awarding points from 0 (none of the time) to 5 (all the time). The sum of the results is multiplied by 4 to be presented as a percentage.

In addition, the questionnaire included questions regarding patients' experiences and side effects while using the SC device. Patients were asked how they tolerated SC using a visual scale from 1 (smiling face) to 3 (neutral face) to 5 (crying face).

When evaluating the side effects, the maximum value of each category per patient was assessed regardless of the cycle of treatment. The maximum values were summarized and the arithmetic mean over all patients was calculated. Furthermore, the patients were asked if they were afraid of losing their hair, if they expected to lose their hair, and afterwards if they were happy with the result and would therefore recommend SC to other patients.

The primary endpoint of the study was evaluation of the QoL by the WHO-5 in BC patients using an SC device, with additional questions regarding patient experiences, side effects, and recommendation rate. The secondary endpoint of the study was the efficacy of SC defined as successful (group A) or unsuccessful SC (group B). Significance was tested using the Student t test

Results

Seventy patients with BC undergoing chemotherapy with an average age of 54.8 ± 13.8 years (range 31–81) participated in the EVAScalp trial. Overall, 40.62% received chemotherapy for metastatic BC and 59.38% for primary BC. As 6 patients needed to be excluded because of incomplete data sets, a total cohort of 64 patients was finally analyzed. One patient received eribulin and was not considered in the efficacy analysis, but was included in the QoL analysis because of the less commonly observed alopecia rates under eribulin.

Efficacy of SC

The overall success of SC was 61.90% (n = 39) with HL lower than grade 2 and completion of the treatment. In metastatic patients with ongoing treatment the assessment of the success was performed at cycle 12. Unsuccessful SC or quitting because of side effects was observed in 38.10% (n = 24). Of the patients, 7.69% (n = 2) stopped SC treatment because of headache as a side effect, while 91.67% (n = 22) of all patients who quit SC did so because of HL after two cycles of chemotherapy on average. Taxane-based therapies such as paclitaxel (with or without targeted therapies) showed the best success rate of up to 88% (n = 29) with paclitaxel weekly (Table 2). Anthracycline-based regimens as often used in early BC therapies had worse SC outcomes than anthracycline-free therapies. For example, epirubicin/cyclophosphamide (EC) followed by paclitaxel had a success rate of only 29% (n = 2). For dose-dense EC with bi-weekly infusions, HL was observed in every patient (n = 2). Docetaxel/cyclophosphamide (TC) with SC also resulted in no prevention of HL (n = 5). Examples of successful and unsuccessful SC are shown in Figure 1.

Table 2.

Efficacy rates of SC depending on type of chemotherapy for anthracycline-based and taxane-based regimes

Regime	Total, n	Successful, n	%
Paclitaxel q1w (± targeted therapy)	33	29	88
EC q3w×4 → paclitaxel q1w×12	7	2	29
TCbH (+P)	4	1	25
Nab-paclitaxel q1w	3	2	67
Nab-paclitaxel/carboplatin q3w×6	4	2	50
TC q3w×6	5	0	0
Paclitaxel/carboplatin q1w×12	3	1	33
Docetaxel/carboplatin q3w×6	1	1	100
EC q3w×4 → docetaxel +H+P q3w×4	1	1	100
Eribulin	1	1	100
EC q2w → paclitaxel q1w	2	0	0

WHO-5 Well-Being Index

WHO-5 in patients with successful SC was significantly higher (67.83 ± 23.46 vs. 51.40 ± 18.91%) at the end of the treatment (p < 0.01 at a 5% significance level) and therefore can be considered as an indicator for increased well-being of patients with successful prevention of HL. After the first cycle of treatment, WHO-5 was also higher in the group with successful SC (64.00 ± 26.59 vs. 52.28 ± 31.57%, p = 0.17) but the difference between the groups was smaller. Interestingly, the WHO-5 of group A numerically increased slightly over the time of treatment, whereas that of group B decreased (Fig. 3; results not significant).

Fig. 3.

WHO-5 well-being index (%) in groups at the beginning (pale red) and end of SC (dark red).

How Did the Patients Experience SC?

In patients with successful SC, the median of the visual scale was numerically slightly better at 2.09 ± 0.96 compared to that in the group with unsuccessful SC (median 2.50 ± 1.34). Overall, SC was tolerated well.

Side Effects

Side effects are rare and reversible. Most frequently, we observed headache (6.8%) or a feeling of coldness (31.8%). Two patients decided to end SC because of a headache. Side effects were graded in the questionnaire on a scale from 1 (no side effect) to 5 (severe side effects), and the highest value of each patient over all the cycles was taken to calculate the median. The scores for the most frequent side effects over all SC treatment cycles were 1.34 ± 0.74 for headache and 2.06 ± 1.00 for a feeling of coldness (Table 3).

Table 3.

Mean of the severity of side effects (1 = no side effect, 5 = severe side effect) and total of patients who suffered the maximum score for each side effect

Side effect	Mean score of side effect during SC	Mean score of side effect after SC	Overall maximum score (n)
Headache	1.34	1.25	4 (2)
Dizziness	1.31	1.41	3 (3)
Nausea	1.20	1.36	3 (2)
Vomiting	1	1.05	3 (1)
Problems with circulation	1.31	1.44	3 (3)
Tiredness	1.82	2.10	4 (3)
Numbness	1.11	1.18	3 (1)
Painful eyes	1.16	1.16	3 (2)
Coldness	2.07	−	4 (4)
Coldness on head	1.89	−	4 (5)
Contact pain	1.48	−	4 (1)
Hot flush	−	1.59	4 (2)
Sinusitis	−	1.18	2 (11)
Common cold	−	1.16	4 (1)

Patient Recommendations

Participants were asked before starting their treatment if they feared HL and if they thought that SC could prevent HL. In response, 85.93% of all patients did fear HL and 88.52% of all patients hoped to prevent HL with SC.

After their treatment, participants were asked if they would recommend SC to other patients. Overall, 60.94% of all patients would recommend SC to other patients. Regarding the different groups, 82.22% (n = 37) of patients with successful SC would recommend SC compared to only 11.11% (n = 2) of the group with unsuccessful SC. As expected, successful SC leads to a high degree of SC recommendation (Table 4).

Table 4.

Recommendation of SC to other patients by the successful and the unsuccessful groups

Group	Recommendation of SC to others
Successful SC	82.22%
Unsuccessful SC	11.11%

Discussion

In EVAscalp, the efficacy of SC was demonstrated by preventing HL in patients with BC who received adjuvant, neoadjuvant, or palliative chemotherapy in 61.90% (n = 39). These results are concordant with the findings of the SCALP trial by Nangia et al. [18].

Different chemotherapy regimens showed different efficacy in HL prevention. We observed higher success rates in patients who received taxane-based therapies (e.g., 88% with paclitaxel weekly ± targeted therapy) than in those who received anthracycline-based therapies (e.g. 29% in EC-containing therapies; Table 2). In most of the unsuccessful patients, HL took place during the first 2 cycles of treatment. Consequently, we suggest that considering the rare and mostly mild side effects of SC, every patient who receives chemotherapy as part of a BC treatment should be offered the possibility of using SC. If it is not working, SC can be stopped after about two cycles and hardly causes additional expense as the procedure itself is cost effective as long as HL is prevented [16]. Efficacy can be increased by fitting of the SC caps perfectly to the individual head with straps. We learned during treatment how important this aspect is and improved our team's skills. Furthermore, individual hair thickness and texture lead to different results. Thicker hair may have an isolation effect on the scalp, reducing the effectiveness of SC. Further studies are needed to evaluate these observations in more detail.

As Nangia [19] stated in 2018, “quality of life matters…” This was an important issue for our study participants too, and 85.93% of all our patients did fear HL. This shows how important hair is in self-perception and how QoL is affected by CIA. WHO-5 was therefore significantly higher in the group with successful SC (65.80 ± 23.58 vs. 53.00 ± 26.74%). Thus, we were able to show for the first time a significant QoL difference in patients with successful and unsuccessful SC. Interestingly, WHO-5 before treatment was also higher in the group with successful SC. So maybe a more optimistic mindset may also contribute to SC efficacy. As expected, the rate of recommendation was high in the successful group (82.22%, n = 37), and quite low in the unsuccessful group (11.11%, n = 2).

Patients tolerated the SC system well as long as HL was successfully prevented. Side effects were rare, mild, and in every case reversible. Most frequently, we observed headache (6.8%) or a feeling of coldness (31.8%). Headache can be treated with painkillers and disappears after a short time (approx. 15 min) in 66.7% of affected patients. Coldness can be conquered with hot drinks or blankets. A visual scale using different face icons was used to determine which patients tolerated the SC treatment. The opinion was slightly better in the successful group. In both groups, it became better with acclimatization to the SC system. The average was “a slightly smiling face” (2 out of 5). We therefore conclude that SC is associated with only a low treatment burden and, from a patient's perspective, is worth trying.

One limitation of our study is that we could not compare the QoL data with a control group of patients who did not undergo SC. At the point we started SC in our clinic, we observed that nearly every patient did want to use SC and that patients who did not were not willing to fill out the questionnaire at each cycle. Therefore, we decided to concentrate on patients with SC in our study. Another limitation is that we observed a learning curve in our team regarding cap fitting and optimal treatment of the most common side effects. This is why the overall effectiveness may be improved by additional optimization of these parameters in the future. There are also first signs that prolongation of the post-infusion cooling may improve outcome in anthracycline-based therapies [20].

Most of the provided modern SC systems, such as Paxman or Dignicap, work with a continuous sensor-based cooling of the scalp. Older systems such as gel caps are cooled in the refrigerator and changed several times during the application to keep the scalp cool. This does not provide a continuous cooling and seems to be less effective as older publications using theses technique showed no satisfying results [21].

Taken together, SC is the only known effective treatment to prevent CIA and should be offered to BC patients undergoing chemotherapy. The well-being of patients with successful SC is significantly higher than that of patients with unsuccessful SC. It is a promising approach and increases patient well-being, although there are still limitations to its utility in several chemotherapy regimens. Our findings are concordant with the results from previous studies [18, 22] and add information on QoL.

Statement of Ethics

All participants gave their written informed consent. The study protocol was approved by the research institute's committee on human research (approval No. 161-15 from June 26, 2015).

Disclosure Statement

The study was funded by Paxman. There are no further conflicts of interest to declare.

Funding Sources

The study was funded by Paxman.

Author Contributions

The study was designed and conceived by F.-F.B., N.H., and R.W. Acquisition, analysis, or interpretation of data was performed by A.K., T.P.-B., T.V., J.G.K., and M.W. The manuscript was drafted by F.-F.B., R.W., and N.H. Statistical analyses were carried out by T.S. and F.-F.B. The manuscript was critically revised by all authors.