Abstract
Background. Chemotherapy-induced alopecia is a frequently occurring side effect of cancer treatment with a high psychological impact which can be prevented by scalp cooling. With this multi-centre patient series we estimated the results of scalp cooling for currently used chemotherapies to provide patient information and we identified characteristics associated with the results. Material and methods. The Dutch Scalp Cooling Registry collected data on scalp-cooled patients in 28 Dutch hospitals. Nurses and patients completed questionnaires on patients, chemotherapy and scalp cooling characteristics. Logistic regression analysis was used to examine associated characteristics of the scalp cooling result. Results. Overall, 50% of the 1411 scalp-cooled patients did not wear a head cover during their last chemotherapy session. Patients were satisfied with the results in 8% of cases after TAC chemotherapy and up to 95% after paclitaxel treatment. Besides type of chemotherapy, higher dose and shorter infusion time, older age, female gender and non-West-European type of hair significantly increased the proportion head cover use. Hair length, quantity, chemical manipulation (dyeing, waving, colouring), wetting hair before scalp cooling, and treatment with chemotherapy ever before did not influence the degree of head covering among patients. Conclusions. Scalp cooling results as recorded in this open patient registry were positive for most regimens, justifying it's use by all eligible patients, except for those needing TAC. Lengthening infusion time may improve the results.
Chemotherapy-induced alopecia (CIA) is a common side effect of cancer treatment and one of the most distressing side effects for many patients [Citation1,Citation2]. CIA may be prevented by scalp cooling prior to, during and some time after chemotherapy infusion. Scalp cooling has been practised since the 1970s but the physiological and biochemical aspects have scarcely been modelled [Citation3]. Until 2011 only 48 publications focussed on scalp cooling results [Citation4–11], largely from monocentric studies with small patient numbers. Generally, efficacy is exhibited, although it is not defined uniformly and is inconclusive for most of the currently used chemotherapies. In fact, determinants of scalp cooling results vary tremendously, i.e. especially type, dose and frequency of chemotherapy infusion, the type of scalp cooling equipment and duration [Citation6].
Clinical impressions of oncologists and nurses are often insufficient for predicting the scalp cooling result for a patient needing a particular chemotherapy regimen [Citation12]. Systematic registration of all patients receiving chemotherapy and scalp cooling aims to standardise and clarify the various issues of hair preservation in currently used chemotherapies.
Material and methods
The Dutch Scalp Cooling Registry started in January 2006 with eight community hospitals and one academic hospital. In December 2009, the registry comprised 28 hospitals, two of which were academic hospitals. Nurses asked all patients who started scalp cooling in that time period to participate, i.e. chemonaive as well as patients with chemotherapy ever before. Nurses reported the year of birth, gender, cancer type, chemotherapy type, dose and infusion time, the adjuvant or palliative nature and reasons for stopping scalp cooling. Infusion times depended on the protocol of a particular hospital. Some of the examined chemotherapies were sequential schemes.
During the first scalp cooling session patients reported hair characteristics: the length and quantity, the type of hair determined by ethnical background, and whether they had dyed, waved or coloured it within two months of the start of chemotherapy. Patients also stated whether they had received alopecia-inducing chemotherapy ever before. During each session nurses recorded the scalp cooling times and whether they had dampened the patient's hair or used hair conditioner before the start of scalp cooling, thus trying to achieve a lower scalp skin temperature. Tolerance was evaluated by recording reasons for stopping scalp cooling.
Scalp cooling was performed using the Paxman PSC1 or PSC2 system. The result was evaluated by asking patients whether or not they wore a head cover (including wig) during their last scalp cooling session. Head cover use is the most important outcome measure, because it represents perceived satisfaction with hair preservation by scalp cooling; for some patients with minimal CIA, this is a severe burden so they choose to wear a head cover, but the opposite is also true. The result of scalp cooling was also evaluated by patients indicating the severity of hair loss on the WHO scale, with score 0 for none, 1 for minimal, 2 for severe and 3 for total alopecia [Citation13]. Final analyses included only patients who had completed at least two scalp cooling sessions or if they discontinued scalp cooling because of severe CIA after the first session.
Statistics
Differences between patients with and without head covering were compared by the χ2 test or Mann Whitney U test. Whether the patient, chemotherapy and scalp cooling characteristics were associated with not wearing head covering after scalp cooling was evaluated by logistic regression analyses and expressed as Odds Ratios (OR). In these overall analyses chemotherapy dosages were not taken into account, because of too much variance within the regimens. For the largest subgroup, i.e. women treated with 5-fluorouracil, epirubicine and cyclophosphamide (FEC) chemotherapy (with epirubicine dosages from 50–100 mg/m2) separate regression analyses were performed, now including dosage. The various patient, chemotherapy and scalp cooling characteristics, included for adjustment in the multivariate analyses, were determined a priori [Citation14]. Statistical differences were indicated if p < 0.05 and reported p-values were two-sided. Statistical analyses were performed using SAS (version 9.1 for Windows, SAS institute Inc., Cary NC, USA).
Results
Between 2006 and 2009 1411 patients were registered in the Dutch Scalp Cooling Registry (). The majority of patients were women (96%) with breast cancer (86%), who were treated in the adjuvant setting (69%). The mean age was 53 years (range 18–81 years). The median pre-infusion cooling time was 38 minutes (Standard Deviation (SD) 12, range 3–115) and the median post-infusion cooling time was 90 minutes (SD 17, range 15–210) (data not shown). The median number of cooling sessions was 4 (SD 3, range 1–27).
Table I. Result of scalp cooling according to patient and scalp cooling characteristics (n = 1411).
Half of the patients (n = 709, 50%) wore no head cover during the last scalp cooling session (). Use of head covering varied according to type and dose of chemotherapy from 8% to 94% of patients. Results were best for monotherapy with low dose taxanes: 94% of patients on docetaxel (D75) wore no head cover, as well as 81% of patients with paclitaxel (T70 - 90) chemotherapy. Results were worst for TAC chemotherapy (8%), despite the relatively low dose of taxane (D75) and anthracycline (A50). In the multivariate analysis, these chemotherapy regimens (D, T and TAC), together with irinotecan, independently influenced the scalp cooling result (). WHO scores for alopecia (WHO 0, 1, 2, 3) for patients wearing head covering, were 2%, 2%, 9% and 87%, respectively, and for patients not wearing head covering 26%, 49%, 25% and 0%.
Table II. Head cover use during the last scalp cooling session according to type of chemotherapy.
Table III. Univariate and multivariate logistic regression analysis of all patients (n = 1411); odds of wearing no head cover during the last scalp cooling session (hc) and odds of WHO score 0 - 1 (who).
Multivariate analysis, including all patients, showed that elderly patients (OR 0.6, p = 0.04) and those with Asian type of hair (OR 0.4, p = 0.008) exhibited the worst scalp cooling results (). Similar results were found for the more homogeneous group of women receiving FEC chemotherapy (). Male gender was positively associated (OR 6.3, p = 0.0004) (). When using WHO scores 0 and 1 (‘no/ minimal’) versus WHO 2 and 3 (‘severe/ total’ alopecia) multivariate analyses showed similar results.
Table IV. Univariate and multivariate logistic regression analysis of females with breast cancer, receiving FE50C-FE100C chemotherapy (n = 751); odds of wearing no head cover during the last scalp cooling session.
The subgroup analysis of patients receiving FEC showed that increase in dosage led to inferior scalp cooling results (). The indication of a dose-effect relationship was also observed when comparing D75 to D100 (94% vs. 61% no head cover) and T70 - 100Carbo to T175Carbo (75% vs. 38% no head cover) (). Longer FEC infusion times reduced the use of head covering (). This effect was also observed marginally for the total patient group ().
Patients with chemically manipulated hair because of dying, waving or colouring did not exhibit higher head cover use after scalp cooling, neither did those with longer length and larger quantities of hair or longer post-infusion cooling times ( and ). Wetting hair also did not contribute to the result. Corrected for the type of chemotherapy in the multivariate analysis, the results were better for patients who had had no chemotherapy before.
Scalp cooling was stopped because of intolerance in only 3% of patients. No scalp skin metastases were reported following the last chemotherapy session up to August 2011.
Discussion
To our knowledge this is the largest prospective multicentre patient series on scalp cooling among patients receiving chemotherapy reported in the literature and the first one to study multivariate characteristics associated with head cover use after scalp cooling. The Dutch Scalp Cooling Registry data showed that no head cover was used by 50% of patients who received chemotherapy regimens that normally cause severe CIA. This outcome is not very optimal; however when a patient has a 50% chance to keep their hair during chemotherapy, it will be an incentive for many patients to choose scalp cooling. Our results are in accordance with the literature, when comparing corresponding dosages [Citation6]. However, in our registry dosages were generally higher than in several recent studies [Citation15–19] and several new chemotherapy regimens were evaluated. Since all scalp-cooled patients in the hospitals could participate, significant patient groups with a specific chemotherapy regimen emerged, which properly reflected daily practice at day care units.
The outcome parameters – proportion head cover use and WHO score – are only an indication of scalp cooling efficacy. For example, patients receiving docetaxel were twice as likely to wear no head cover as patients on FEC (OR 2.0 vs. 1.0). However, scalp cooling improved the chance of no head covering in FEC from about 5% [Citation12] to 49% (factor 10) and in docetaxel from about 30% [Citation20] to 75% (factor 2.5). The net scalp cooling efficacy remains unknown for most chemotherapy regimens, while severity of CIA without scalp cooling was not evaluated and varies tremendously in phase II and III trials.
Less head cover use after longer infusion times has not been reported previously. It seems that a lower peak plasma concentration causes more damage to hair root cells than a longer chemotherapeutic exposure time. Lengthened infusion times are regularly used to prevent hypersensitivity reactions and to lower the risk of cardiotoxicity [Citation21,Citation22]. However, variation in infusion time may alter treatment outcomes and toxicity differently in each regimen [Citation23,Citation24].
The finding that after scalp cooling patients older than 65 years were more likely to wear a head cover is new. Only Macduff et al. reported in a small study that patients aged 50 or over lost more hair than those younger than 50 years [Citation25]. It is possibly due to higher chemotherapeutic concentrations in hair root cells during scalp cooling. Aged skin has a diminished cold-induced vasoconstriction [Citation26] and an age-related decline in organ function may increase toxicity [Citation27]. Chemotherapeutics mainly affect anagen hairs, i.e. hair in the growth phase, and cause a sharp constriction of the hair shaft, where hairs may break [Citation28]. Therefore, the reduced hair diameter at older age may also increase the risk of breakage.
Reduced scalp cooling results for Asian patients may be caused by a lower maximum tolerable chemotherapeutic dose and higher toxicity rates compared to their Western counterparts [Citation29]. The satisfactory scalp cooling results reported in a recent Japanese study may be attributed to lower chemotherapy dosages (T60, C400, E40) than currently used in the Netherlands [Citation17]. Biochemical characteristics of hair from people of different ethnical backgrounds have proved to be very similar, whereas geometry, mechanical properties and phase and rate of hair growth are particularly different [Citation30–33]. These factors might contribute to hair breakage rates.
Unsatisfied scalp-cooled patients exhibited overall worse quality of life and body image compared to patients who did not undergo scalp cooling [Citation1]. Therefore, scalp cooling should not be advised if poor results are expected, i.e. up to now with TAC.
The unspecified patient group with very diverse chemotherapy regimens had good scalp cooling results. The regimens and dosages for this group might have caused somewhat less often severe CIA, also without scalp cooling.
The good scalp cooling results for males may be caused by a difference in chemotherapy. Males most frequently received docetaxel 75 mg/m2 (52%), for which scalp cooling results are particularly good [Citation20]. Moreover, the result in males is likely to be somewhat overestimated, since they are in general less inclined to wear a wig or head cover. The low number of male patients is not because they do not value their hair [Citation34], but probably because scalp cooling is rarely offered to them. In one of our studies all eligible patients were actively recruited and one third of the 168 scalp-cooled patients were males [Citation20].
Continuous multicentric registration ensures rapid availability of the scalp cooling results patients may expect in new or uncommon chemotherapy regimens. Furthermore, registration may be used to detect easily differences in results between hospitals; actions may be undertaken in case of extra good outcomes or if outcomes are below average. As found in the past in the Netherlands, several hospitals quickly stopped scalp cooling after disappointing results in a minimal number of patients.
In conclusion, for daily practice this study implies that scalp cooling results are better with certain chemotherapy types (taxanes), decrease at higher doses and might be improved by longer infusion times. Medical doctors and nurses have to be aware that males are also eligible for scalp cooling and those patients of older age or non-West-European hair type may have somewhat less chance of satisfactory results. Scalp cooling results need to be improved, mainly by a more patient-tailored approach. Progress can be achieved by a better patient selection and when optimal scalp skin temperature and cooling times for each type of chemotherapy are known. Scalp cooling intensity should be adapted to the optimal scalp skin temperature for each patient. Post-infusion cooling times (PICT) are currently only used arbitrarily, with the exception of docetaxel [Citation20]. Optimal PICTs may vary in particular chemotherapy regimens, while half life times of chemotherapeutics vary tremendously. We continued multicentre registration, because it improved patient information while analysis of characteristics provided a better understanding of the determinants that improve scalp cooling results. This registry should be expanded internationally to obtain a consistent picture.
Acknowledgements
Special thanks go to all patients who participated in the registration, to the nurses and MDs who performed the registration (www.scalpcooling.nl), to A. Willemse who assisted in data collection and Eindhoven Cancer Registry/ Comprehensive Cancer Centre South for hosting the scalp cooling research centre. The authors have declared no conflicts of interest.
Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.
References
- Hurk van den CJ, Mols F, Vingerhoets AJ, Breed WP. Impact of alopecia and scalp cooling on the well-being of breast cancer patients. Psycho-oncology 2010;19:701–9.
- Mulders M, Vingerhoets A, Breed W. The impact of cancer and chemotherapy: Perceptual similarities and differences between cancer patients, nurses and physicians. Eur J Oncol Nurs 2008;12:97–102.
- Janssen FE. Modelling physiological and biochemical aspects of scalp cooling. Eindhoven: Technical University Eindhoven; 2007. Thesis.
- Hurk van den CJG, Gerrits P, Graat J, Kolen B, Laar van de-Müskens J, Breed WPM. Positieve ervaringen met hoofdhuidkoeling in drie ziekenhuizen in Nederland. Tijd voor een standaard aanbod van hoofdhuidkoeling? Oncologica 2005;22:162–67 [Dutch].
- Christodoulou C, Tsakalos G, Galani E, Skarlos DV. Scalp metastases and scalp cooling for chemotherapy-induced alopecia prevention. Ann Oncol 2006;17:350.
- Breed WPM, Hurk van den CJG, Peerbooms M. Presentation, impact and prevention of chemotherapy-induced hair loss; scalp cooling potentials and limitations. Expert Review Dermatol 2011;6:109–25.
- Grevelman EG, Breed WP. Prevention of chemotherapy-induced hair loss by scalp cooling. Ann Oncol 2005;16:352–8.
- Konishi Y, Kuroki T. Prevention of adrimycin-induced alopecia by scalp hypothermia with a deep-frozen Duncool-cap. Gan To Kagaku Ryoho 1988;15:3081–5.
- Nakazawa M, Sakamoto T, Honda M, Sawa Y, Sato K, Oguro K, . Scalp hypothermia: A prevention of hair-loss in clinical oncology. Dokkyo J Med Sci 1985;12:17–24.
- Keizer-Heldens P. Hoofdhuidkoeling bij haarverlies door chemotherapie. Oncologica. 2009;29–34 [Dutch].
- Kargar M, Sarvestani RS, Khojasteh HN, Heidari MT. Efficacy of penguin cap as scalp cooling system for prevention of alopecia in patients undergoing chemotherapy. J Adv Nurs 2011;67:2473–7.
- Geerts P, Hendriks M, van den Hurk C, Dercksen M, Breed W. Hair preservation in FEC-chemotherapy; not an exception. 2011 (submitted).
- World Health Organisation. Handbook for Reporting Results of Cancer Treatment. Geneva: WHO Offset Publ.; 1979.
- Babyak MA. What you see may not be what you get: A brief, nontechnical introduction to overfitting in regression-type models. Psychosom Med 2004;66:411–21.
- Auvinen PK, Mahonen UA, Soininen KM, Paananen PK, Ranta-Koponen PH, Saavalainen IE, . The effectiveness of a scalp cooling cap in preventing chemotherapy-induced alopecia. Tumori 2010;96:271–5.
- Massey CS. A multicentre study to determine the efficacy and patient acceptability of the Paxman Scalp Cooler to prevent hair loss in patients receiving chemotherapy. Eur J Oncol Nurs 2004;8:121–30.
- Kato M, Sakuyama A, Imai R, Kobayashi T. Evaluation of dignicap system for the prevention of chemotherapy-induced hair loss in breast cancer patient. Breast 2011;20(Suppl 1):S80, abstr 811.
- Lenaerts E, Meyen M, Maes T, El Mousati A, Lemmens J, . Scalp cooling in the prevention of anthracycline-induced alopecia. Eur J Cancer 2001;37(Suppl 6):360.
- Peck HJ, Mitchell H, Stewart AL. Evaluating the efficacy of scalp cooling using the Penguin cold cap system to reduce alopecia in patients undergoing chemotherapy for breast cancer. Eur J Oncol Nurs 2000;4:246–8.
- Hurk van den CJG, Breed WPM, Nortier JWR. Short post-infusion scalp cooling times in the prevention of docetaxel-induced alopecia. 2011 (submitted).
- Kingsley CD. Hypersensitivity reactions. In: Perry MC, editor. The chemotherapy source book. Philadelphia: Wolters Kluwer Health; 2008.
- Rasheed ZA, Rubin EH. Topoisomerase-interacting agents. In: DeVita VT, Lawrence TS, Rosenberg SA, editors. Cancer principles & practice of oncology. 8th ed. Philadelphia: Wolters Kluwer Health; 2008. pp. 437–44.
- Williams C, Bryant A. Short versus long duration infusions of paclitaxel for any advanced adenocarcinoma. Cochrane database of systematic reviews (Online). 2011;5:CD003911.
- Mikkelsen TS, Sparreboom A, Cheng C, Zhou Y, Boyett JM, Raimondi SC, . Shortening infusion time for high-dose methotrexate alters antileukemic effects: A randomized prospective clinical trial. J Clin Oncol 2011;29:1771–8.
- Macduff C, Mackenzie T, Hutcheon A, Melville L, Archibald H. The effectiveness of scalp cooling in preventing alopecia for patients receiving epirubicin and docetaxel. Eur J Cancer Care (Engl) 2003;12:154–61.
- Holowatz LA, Kenney WL. Peripheral mechanisms of thermoregulatory control of skin blood flow in aged humans. J Appl Physiol 2010;109:1538–44.
- Hurria A, Lichtman SM. Pharmacokinetics of chemotherapy in the older patient. Cancer Control 2007;14:32–43.
- Hurk van den CJG, Breed WPM, Mols F. Chemotherapy-induced hair loss. In: Preedy VE, editor. Handbook of Hair in Health and Disease. Wageningen, The Netherlands: Wageningen Academic Publishers; 2012; p. 403–16.
- O'Donnell PH, Dolan ME. Cancer pharmacoethnicity: Ethnic differences in susceptibility to the effects of chemotherapy. Clin Cancer Res 2009;15:4806–14.
- Blume-Peytavi U, Tosti A, Whiting DA, Trueb RM. Hair growth and disorders. Berlin: Springer Verlag; 2008.
- Franbourg A, Hallegot P, Baltenneck F, Toutain C, Leroy F. Current research on ethnic hair. J Am Acad Dermatol 2003;48:S115–9.
- Takahashi T, Hayashi R, Okamoto M, Inoue S. Morphology and properties of Asian and Caucasian hair. J Cosmet Sci 2006;57:327–38.
- Loussouarn G, El Rawadi C, Genain G. Diversity of hair growth profiles. Int J Dermatol 2005;44(Suppl 1):6–9.
- Hilton S, Hunt K, Emslie C, Salinas M, Ziebland S. Have men been overlooked? A comparison of young men and women's experiences of chemotherapy-induced alopecia. Psycho-oncology 2008;17:577–83.