1g0375.qxd

Support Care Cancer (2002) 10:529–537DOI 10.1007/s00520-002-0375-y Christel Protière
Efficacy and tolerance of a scalp-cooling system Katrin Evans
Jacques Camerlo
for prevention of hair loss and the experience Marie-Pierre d’Ingrado
of breast cancer patients treated by adjuvant Geneviève Macquart-Moulin
Patrice Viens
Dominique Maraninchi
Dominique Genre
Abstract The applicability and effi-
Published online: 15 August 2002 Springer-Verlag 2002 than in those without (n=28). Differ- by nurses (cycles 1–3: P<0.01; cycle 4: P<0.05). Tolerance was generally K. Evans · J. Camerlo · M.-P. d’Ingrado Medical Oncology Department, Institut Paoli-Calmettes, 232, the distress associated with thissymptom. Nurses’ ratings (n=105) Keywords Alopecia · Scalp-cooling
University of Mediterranean (Aix-Marseille II), Marseille, France evidence of the disease to the outside world, etc.) [3, 8,24, 31]. Moreover, it induces different physical problems Among chemotherapy side-effects, nausea/vomiting and (discomfort caused by the capillary prosthesis, irritability alopecia are the ones most feared by patients. The use of because of hair loss, anxiety) [29]. Lastly, alopecia can setrons has allowed vomiting to be enormously reduced, incite some patients, particularly women, to totally but the problem of alopecia still remains unresolved.
refuse potentially curative treatment.
Nevertheless, hair loss is known to have a strong psy- Consequently, alopecia is a cause of concern to clini- chological influence on patients (alteration of self-image, cians involved in breast cancer care. In the context of self-esteem, feelings of difference in the eyes of others, standard breast cancer for which an adjuvant chemother- apy is likely to be offered, the most commonly used regi- Fig. 1 Study design
men is a combination of cyclophosphamide, methotrex-ate and 5 fluorouracil (CMF) [5]. Another well-estab- in breast cancer patients receiving adjuvant chemothera- lished active agent is doxorubicin, which is also widely py associating mitoxantrone and cyclophosphamide.
used in combination with other agents [1, 6]. However, Women accepting the scalp cooling system were com- these drugs have severe toxic effects (hair loss, nausea pared for alopecia to those who refused it and to patients and vomiting, mucositis, cardiomyopathy and marrow included in the previous above-cited study.
suppression). For this reason, several analogues anddrugs with similar biochemical mechanisms have beendeveloped with less toxic effects but similar efficacy.
These drugs to which mitoxantrone belongs [4] are likelyto be better accepted by patients.
Study design is summarized in Fig. 1.
This has led the clinicians to introduce mitoxantrone into adjuvant chemotherapy for standard breast cancer.
Indeed, mitoxantrone is generally considered as well tol- erated and associated with only moderate alopecia [14, 15, 20, 23, 25, 32]. Moreover, mitoxantrone has been re-ported to cause selective loss of white hair [2]. However, One hundred and five patients suffering from breast cancer with its toxicity, when associated with cyclophosphamide that no metastases and fewer than nine involved axillary nodes re- is known as a severely alopecia-inducing drug, merited ceived an adjuvant treatment associating polychemotherapy andconcomitant or delayed radiotherapy. They were recruited at the to be fully explored. A quality of life study carried out at Institut Paoli-Calmettes (Regional Hospital for Cancer Care) in the Institut Paoli-Calmettes (Marseilles, France) among Marseilles (France) between 10 February 1997 and 4 May 1999.
109 breast cancer patients receiving a regimen of mitox- Chemotherapy doses were based on ideal body weight. The antrone and cyclophosphamide (NC chemotherapy) in protocol (NC chemotherapy) comprised of mitoxantrone (12 mg/m2)4 cycles of 21 days showed that hair loss was reported and cyclophosphamide (600 mg/m2) administered in four cycles of21 days in the outpatient clinic. Mitoxantrone (Novantrone) was by patients in about 74% of cycles [22], what remained administered as a 15-min i.v. infusion and cyclophosphamide (En- relatively high. Following these findings, it was decided doxan) as a 30-min i.v. infusion on 1 day. Antiemetic prophylaxis in the future to offer the use of a scalp cooling system to consisted of granisetron (Kytril) administered in 1-mg tablet form breast cancer patients included in the NC protocol.
1 h before each cycle of chemotherapy and within 12 h after eachfirst administration.
The benefit of preventing hair loss, a problem due to Patients were offered the scalp-cooling system by physicians antimitotics, was revealed about 30 years ago. It is based after being informed of the technique of its use, its drawbacks, the on the theory that reduction of cutaneous scalp circula- potential risks (mainly scalp metastasis) and the anticipated effects tion, produces vasoconstriction of blood vessels, and in- [12]. Verbal, informed consent was required from all patients.
Scalp cooling was started with the first cycle of chemotherapy cidentally reduces the quantity of antimitotics which per- fuse the capillary follicles. The first method describedused compression with a tourniquet system. Then, onlycooling methods were used, such as ice cubes, cryogels, “Reference” group of patients (n=109) propelled air and refrigerated helmets [9].
These patients (n=109) were recruited at the Institut Paoli- The aim of this paper is to report the applicability and Calmettes (Regional Hospital for Cancer Care) in Marseilles be- efficacy of a scalp-cooling system (refrigerated helmet) tween 12 May 1995 and 4 February 1997, with the same inclusion criteria as those required for study patients. The treatment protocol tients accepting the scalp-cooling system (n=77) were compared comprised the same polychemotherapy as that used in patients in- for alopecia with patients who refused (n=28). In addition, a sub- cluded in the current study and also radiotherapy administered sample of study patients (n=27) who accepted the scalp-cooling concomitantly with chemotherapy. The scalp-cooling system was system and self-assessed their alopecia was compared with the ref- not proposed but these patients self-assessed their alopecia in the erence group of 109 patients treated by NC chemotherapy who context of a prospective study of treatment side-effects [22].
had similarly assessed their hair loss. Cohen’s kappa coefficient[10] was used as the measure of agreement between the frequencyof alopecia assessed by patients and that noted simultaneously by nurses (using WHO criteria). Based on general statistical guide-lines [17], the strength of agreement was labeled as follows: The equipment used was the FRIGECAP helmet (PETERS LABO ≤0.40, poor to fair agreement; 0.41–0.60, moderate agreement; PHARMA), which required the hair to be moistened and the head 0.61–0.80, good agreement; 0.81–1, excellent agreement. Finally, protected with a cap. The helmet was worn for 15 min before and tolerance of the scalp-cooling system was analyzed in a descrip- 45 min after treatment. Helmet temperature was checked after tive way. All statistical tests were two sided. Significance was as- 45 min and the helmet was replaced if it had become warm (which happened in 4 of the 77 accepting patients). After use, it was re-moved, disinfected and decontaminated. After drying, the helmetwas frozen (–25°C) for a minimum of 48 h.
Efficacy and tolerance of the scalp-cooling system The efficacy and tolerance of the scalp-cooling system were as- One hundred five patients were entered into the current sessed by nurses using a standard form developed by paramedicalstaff and a physician (D.G.). This form was completed by nurses study. Of 105 patients approached prior to cycle 1 of consulting with the patients (face-to-face interview) at each cycle chemotherapy, 77 accepted the scalp-cooling system and of chemotherapy. Hair loss was evaluated according to the World 28 refused it. Among the 77 accepting patients, 4 defini- Health Organization (WHO) criteria for alopecia (grade 0: no hair tively stopped using the system after cycle 1, 5 after loss; grade 1: minimal hair loss; grade 2: moderate hair loss,patchy alopecia; grade 3: complete hair loss, reversible; grade cycle 2 and 2 after cycle 3 (Fig. 1).
4:complete hair loss, irreversible). Tolerance of the scalp-cooling Clinical characteristics of the study patients are pre- system was assessed by the nurse applying and supervising its use.
sented in Table 1. There were no statistically significant The patient was questioned during anticancer drug perfusion and differences in terms of age, hormonal status, number of the following 7 days. The 7-day assessment was retrospective andconcerned the patient’s experience during the previous cycle. Tol- involved axillary lymph nodes, type of surgery and mo- erance was rated as: very good, good, moderate, or poor. All side- dalities of administration of radiotherapy (before or after effects related to helmet use reported by the patient were systemat- chemotherapy) between the patients who accepted the ically noted by the nurse. To this end, she systematically ques- scalp-cooling system (n=77) and those who refused it tioned the patient on specific items (problems related to helmet during the entire course of radiotherapy (n=28). Further- weight and coldness, headache, neck pain), but also collected pa-tient’s spontaneous comments.
more, there were also no statistically significant differ-ences for any of the initial characteristics between the105 patients entered on the current study and the 109 pa- tients included in our previous study (reference group Hair loss was assessed by patients using a specific self-report patients). However, it must be noted that in the reference questionnaire including three items measuring frequency, severity group radiotherapy was systematically administered con- and distress associated with this symptom. This self-report ques- tionnaire is a part of a 19-item self-report questionnaire used formeasuring symptoms related to NC chemotherapy in the referencegroup of 109 patients [22]. Patients were asked about hair lossduring the previous cycle. Severity of alopecia and the distress Efficacy of the scalp-cooling system as assessed caused by it were scored on four-point scales ranging from 1 by nurses: comparison of the patients with the scalp-cooling(“a little”) to 4 (“totally”) for severity and from 1 (“not at all”) to system (n=77) vs those without (n=28) 4 (“very much”) for distress. The three-item self-report question-naire was completed by patients at the start of each cycle of che-motherapy from cycle 2 onward, and at the visit made at the end Table 2 shows the frequency of alopecia as assessed of chemotherapy (planned for 1 month after the last cycle of che- by nurses at each cycle of chemotherapy using WHO motherapy). This administration schedule was the same as that criteria, both in patients who accepted the scalp-cooling used in the reference group of patients.
system and in patients who refused it. In Table 2, pa-tients are classified as follows: no alopecia (WHO grade 0) and alopecia (WHO grades 1–4). After the firstNC chemotherapy cycle, the frequency of alopecia was Statistical analysis was performed using SPSS software [28].
13.0% among the patients who accepted the scalp-cool- Qualitative data were analyzed using the Pearson Chi-square test ing system, as against 32.1% among those who refused (or Fisher exact test, if necessary). Quantitative data were com-pared by parametric (Student t-test) or nonparametric methods it. Figure 2 shows the frequency of alopecia in the two (Mann-Whitney test). At each cycle of chemotherapy, study pa- groups after the first cycle of NC chemotherapy, this Table 1 Patients’ characteristics
RadiotherapyConcomitant with chemotherapy a Number and percentage, except for age and number of involved b Incomplete numbers correspond to missing data Table 2 Alopecia assessed by nurses in the 105 study patients
motherapy in 73.1% (76/104), while no episodes werenoted in 26.9% (28/104). Compared with patients who were assessed by nurses as having had at least one epi- sode of hair loss, patients without any episode of hair loss during the entire course of chemotherapy were more frequently assigned to receive concomitant radiotherapy [82.1% (23/28) vs 60.8% (45/74) – P=0.041]. When thedata were examined cycle by cycle, it appeared that simi- lar results were found for cycles 3 [76.9% (40/52) vs 57.1% (28/49) – P=0.034] and 4 [81.1% (30/37) vs 58.7% (37/63) – P=0.022], but not for cycles 1 and 2.
Nevertheless, there was a similar percentage of patients who accepted use of the scalp cooling system in the two modalities of radiotherapy administration. Furthermore,there were no statistically significant relationships be- tween alopecia and the four following variables: age, number of involved axillary lymph nodes, hormonal sta- a Number (percentage); incomplete numbers correspond to missingdatab Comparison patients with vs without scalp cooling system: Efficacy of the scalp cooling system as assessed by patients: comparison of patients with scalp-coolingsystem (n=27) vs patients without (n=109) time broken down according to WHO alopecia grades.
Not surprisingly, a progressive increase in frequency was Table 3 shows the frequency of alopecia self-assessed at observed in the two groups of patients over subsequent each cycle of chemotherapy by a subgroup of 27 study cycles. Differences between the two groups of patients patients who accepted the scalp-cooling system. Their were statistically significant for cycles 1 (P=0.024) and 3 assessments were compared with those provided by the (P=0.038). At the end of cycle 4, nurses noted hair loss 109 patients belonging to the reference group (no use of in 56.1% of patients who accepted the system and in the scalp-cooling system). There were statistically signif- icant differences between the two groups of patients for Among the 105 study patients, nurses noted at least each cycle of chemotherapy (P<0.01 for the three first one episode of hair loss during the entire course of che- cycles and P<0.05 for the last cycle).
Fig. 2 Alopecia as assessed by
nurses in the 105 study patients
using WHO criteria (cycle 1)
Table 3 Frequency of alopecia
assessed by a subsample of 27
vs without scalp-cooling system:P<0.05 Alopecia frequencies self-reported by these 27 pa- apy, in patients with scalp cooling system than in pa- tients at each cycle of chemotherapy were compared tients without. However, differences between the two with those noted simultaneously by nurses. According to groups of patients were not statistically significant both Landis and Koch’s criteria, concordance between pa- for severity and distress scores in none of four cycles of tients’ and nurses’ assessments was good for cycles 1, 2 and 3 (kappa coefficients = 0.72, 0.74, and 0.75, respec-tively) and excellent for cycle 4 (kappa coefficient = 0.91).
Table 4 displays alopecia severity and distress scores Tolerance and safety of the scalp cooling system (n=48) (range: 1–4), reported both by the 27 study patientsagreeing to use the scalp cooling system and the 109 ref- Tolerance of the scalp cooling system was evaluated in erence group patients (no use of the scalp-cooling 48 patients of 77 accepting patients. Table 5 displays tol- system). Table 4 shows that the percentage of patients erance scores at each cycle of chemotherapy before and giving severity and distress scores 3–4 (high intensity of after perfusion and shows high tolerance scores both dur- severity/distress) was lower, at each cycle of chemother- Table 4 Severity and distress
caused by alopecia, assessed by
tients compared with referencegroup of 109 patients were calculated when symptomwas present. Incomplete num- ence group patients: P>0.05 for Table 5 Evaluation of toler-
ance of the scalp-cooling
system, during and after perfu-sion, in a subsample of 48 a Number (percentage); incomplete numbers correspond to patients who stopped using the scalp-cool-ing system (4 after cycle 1, 3 after cycle 2 and 2 after cycle 3) and/or to missing data In this sample of 48 patients, 4 patients stopped using accepted the scalp-cooling system (n=77) had reported the scalp-cooling system after the first cycle of chemo- therapy because of neck pain, fear of skin metastasis, si-nusitis, or febrile syndrome; 3 withdrew after the secondcycle because of headaches or cold sensations or for per- sonal reasons; and 2 patients after the third cycle becauseof psychological rejection and intolerance to wearing the The aim of this study was to test the efficacy of a scalp- helmet. During chemotherapy perfusion, patients’ com- cooling system in the prevention of hair loss in women ments essentially concerned the weight of the helmet, suffering from breast cancer treated by a chemotherapy pronounced coldness for 5–10 min or a feeling of burn- regimen associating mitoxantrone and cyclophospha- ing and headaches. After perfusion, the following symp- mide (NC chemotherapy) and to explore patients’ expe- toms were the most frequently described : headaches, neck pain, and psychological difficulties in wearing the Among the methods of alopecia prevention, the hypo- helmet (violent aversion to cold, ice and freezing).
thermic methods, used since 1978, have become themost widely used [9, 12]. Two studies have shown thatthe efficacy of the hypothermic methods is inversely as- sociated with epi- and subcutaneous scalp temperatures[7, 13, 18]. Based on these studies, a pre- and postche- After a median follow-up of 44 months (range: 23–63) motherapy scalp-cooling system has been used to main- from the start of chemotherapy, none of patients who had tain steady, low scalp temperatures. The temperature lev- el obtained at scalp level is also a cause of concern for NC chemotherapy who were not offered the scalp-cool- investigators. Gregory [13] reported that hair preserva- ing system. Statistical analysis demonstrated that pa- tion was only obtained when scalp temperature was re- tients included in the current study presented similar ini- duced below 22°C, while for Cooke et al. [11] hair con- tial clinical characteristics to those belonging to the ref- servation only occurred when the scalp temperature was erence group. Second, tolerance of the helmet was stud- reduced to 24°C or below. Another subject of discussion ied in only 48 of 77 accepting patients, and the self-mea- is the duration of scalp cooling required both before and sures of alopecia were performed on a small subsample after chemotherapy infusion. According to Tollonear et (n=27) of these same patients. The reasons for this were al. [30], cooling times of at least 15 min before and at only logistical. In particular, it was not easy to submit least 30 min after the infusion are indicated when anthra- the study patients to a double evaluation of their hair loss cycline is used as the sole alopecia-inducing drug. In the in the context of the outpatient clinic. It is possible also recent study published by Katsimbri et al. [16], the peri- that uneven numbers in different groups of patients could od of application of their scalp-cooling system added ap- induce bias in the results of comparative analyses. Final- proximately 2.5 h to the duration of each chemotherapy ly, the study would have had to be completed by an eval- session. When refrigerated at –25°C for 24 h, the scalp- uation of the costs of the procedure in terms of time in- cooling system used in our study might maintain a tem- vested by nurses and the time a patient occupied a place perature below 25°C for 60 min. Pre- and postinfusion in the outpatient clinic. This evaluation would show cooling times of 15 min and 45 min, respectively, were whether or not a method which can provide a global de- in accordance with those advocated by Tollonear et al.
crease in alopecia frequency by about 22–31% (depend- [30]. Furthermore, it must be noted that in our study, the ing on whether the alopecia assessment was done by helmet was replaced because of warming only in 5% of nurses or patients) can be justified. The burden for the patients, which is very low. Some of mitoxantrone’s patient should also be considered. However, it must be pharmacokinetic characteristics may explain the relative noted that in our study the majority of women were very efficiency of the scalp-cooling system. Indeed, in previ- satisfied with the procedure. Despite some criticisms, we ous pharmacokinetic studies, the initial drug distribution believe that data obtained in this study are useful and phases were shown to be very short, with T 1/2α and have important implications for patients’ well-being.
T 1/2β of 2.5 and 16 min. [20]. Consequently, in our This study clearly demonstrates the efficacy of the study, the cooling duration exceeded the length of expo- scalp-cooling system in preventing alopecia induced by sure to high plasmatic mitoxantrone concentrations, po- NC chemotherapy. At cycle 4, nurses noted hair loss in tentially contributing to the reduction in hair toxicity.
56.1% of those accepting the system vs 69.2% in pa- Reaching a consensus on the effectiveness of scalp tients who had refused it (NS). For the subsample of 27 cooling has been hampered by the large variety of cyto- patients who self-assessed their alopecia and who were toxic agents and doses used, the small patient cohorts compared with the reference group of 109 patients with- and a failure to randomize patients or to incorporate out the scalp-cooling system, results were even more comparative groups [16, 19, 26, 27, 29, 30]. Another marked. (At cycle 4, hair loss was reported in 59.1% of limitation was the use of nonstandard methods or criteria patients who used the scalp-cooling system, as against of hair loss evaluation (mainly, photos of patient scalp, 83.9% of patients who did not (P=0.017). In contrast to graded scale, use or not of a wig). Furthermore, they also our previous experience with physicians [21], concor- failed to take direct account of the patients’ point of dance between patients’ and nurses’ reports was high.
view in parallel with physicians’ or nurses’ measures.
Our results can be favorably compared with those of Our study has the advantage of comparing the study Ron et al. [27], who found 47% hair loss in patients with patients accepting the scalp-cooling system with those scalp-cooling system vs 81% without (P=0.014). Their refusing it and also with a reference group of 109 women study group of 35 breast cancer patients consisted of 19 previously treated with the same chemotherapy protocol hypothermically treated patients and 16 age-matched but without the use of a scalp-cooling system. The num- control patients treated at ambient temperature. All were ber of patients entered was relatively large (n=105), and treated with the same highly depilatory protocol CMF alopecia was assessed by nurses for all patients and also (cyclophosphamide, methotrexate and 5-fluorouracil).
by the women themselves in a subsample of 27 patients.
Our findings concerning a higher frequency of hair loss However, some limitations of this study merit discus- at cycles 3 and 4 in patients assigned to sequential treat- sion. First, patients were not randomized. Indeed, we de- ment may be explained as the expression of some appre- liberately decided against the randomization procedure, hension at the prospect of a second complementary ther- for ethical considerations. Furthermore, we knew that apy, i.e. the radiotherapy. This situation would not arise patients were aware of the possibility of using the scalp- in patients receiving concurrent radiochemotherapy who cooling system and were entirely free to request the pro- cedure from the medical team. In addition to this, we al- Compared to the “reference” group of 109 patients, ready had access to a large group of patients treated by the severity and distress associated with hair loss were always lower at each chemotherapy cycle in the sub- Despite the doubts about the efficacy of scalp hypother- sample of 27 patients who self-assessed their hair loss mia for some drugs and questions remaining about subse- and who used the scalp cooling system. These results quent scalp metastases, scalp hypothermia appears simple, confirm the effectiveness of the scalp cooling system relatively easy for nursing staff, low cost and low-risk. As and demonstrates the additional benefit in stress reduc- patients’ quality of life becomes a priority, the routine use of a scalp-cooling system as part of adjuvant chemothera- Although there were 4 cases of withdrawal from py, especially in cancers with a low prevalence of scalp scalp-cooling in the subsample of 48 patients evaluated metastasis, should be seriously considered. This technique with regard to scalp-cooling system tolerance, patients’ can help maintain patients’ self-image and thereby reduce acceptability was high. Complaints were in keeping with the devastating psychological effects associated with can- those reported by other investigators. In spite of prob- cer, and, sometimes, still more with its treatment.
lems encountered, women were generally very satisfiedwith their choice and did not regret their decision to ac- Acknowledgements We wish express our gratitude to PETERS
cept the scalp-cooling system. In fact, the majority con- LABO PHARMA for providing the FRIGECAP helmet used inthis study. We also thank the patients and the members of the sidered that the psychological comfort related to its use Medical Oncology Department of the Institut Paoli-Calmettes for outweighed minor physical difficulties.
1. Aisner J, Weinberg V, Perloff M, et al cinoma of the breast. A CALGB study.
scaled disagreement or partial credit.
netics, efficacy in the treatment of solid ing care – a literature study. Eur J Can- 12. Dorr VJ (1998) A practitioner’s guide therapy of metastatic breast carcinoma.
23. Panasci L, Shenouda G, Begin L, et al trial of Adriamycin versus methotrex- induced alopecia can be prevented.
Lancet 346:371–372 breast cancer. Eur J Cancer [A]29:2106–

Source: http://paxman.sk/files/protiere-paper.pdf