Table 1 Clinicopathologic patients’ characteristics at the time of CT starting
|
N (%) of Patients
|
Age median (range)
|
49 (42-56)
|
Menopausal Status
|
Premenopausal
|
143 (57%)
|
Postmenopausal
|
108 (43%)
|
Race
|
Caucasian
|
193 (76.6%)
|
Mediterranean
|
55 (21.8%)
|
African
|
3 (1.2%)
|
Asian
|
1 (0.4%)
|
Hystotype
|
Ductal
|
233 (92%)
|
Lobular
|
9 (4%)
|
Other
|
10 (4%)
|
Phenotype
|
ER and/or PgR pos
|
169 (67%)
|
ER and/or PgR neg
|
83 (33%)
|
HER2 pos.
|
93 (37%)
|
HER2 neg.
|
159 (63%)
|
Ki67 status
|
Ki67 <20%
|
70 (28%)
|
Ki67 ≥ 20%
|
182 (72%)
|
Histological Grade
|
1
|
4 (2%)
|
2
|
97 (38%)
|
3
|
148 (59%)
|
UNK
|
3 (1%)
|
cT/pT
|
1
|
129 (51%)
|
2
|
94 (37%)
|
3
|
20 (8%)
|
4
|
7 (3%)
|
UNK
|
2 (1%)
|
cN/pN
|
0
|
113 (45%)
|
1
|
93 (37%)
|
2
|
23 (9%)
|
3
|
19 (8%)
|
UNK
|
4 (2%)
|
Disease Stage (AJCC)
|
I
|
79 (31%)
|
II
|
117 (46%)
|
III
|
56 (22%)
|
Body Mass Index
|
Healthy Weight
|
132 (52%)
|
Under Weight
|
10 (4%)
|
Over Weight
|
77 (31%)
|
Obesity
|
33 (13%)
|
Smokers
|
Yes
|
26 (10.3%)
|
No
|
198 (78.6%)
|
UNK
|
28 (11.1%)
|
Chemotherapy regimens
|
EC→TXL weekly
|
75 (29.8%)
|
EC
|
72 (28.6%)
|
TXL + TRAST. Weekly
|
46 (18.3%)
|
EC→TXL+TRAST.weekly
|
35 (13.9%)
|
EC→TXL + CBDCA weekly
|
5 (2.0%)
|
EC→TXL+TRAST.+PERT. weekly
|
5 (2.0%)
|
(TXL+CBDCA weekly→EC )+PEMBRO.
|
5 (2.0%)
|
TC
|
5 (2.0%)
|
EC→CMF
|
5 (2.0%)
|
CBDCA+TXL+TRAST. Weekly
|
1 (0.4%)
|
EC→CMF+TRAST
|
1 (0.4%)
|
Anthracycline based regimens
|
Yes
|
197 (78.2%)
|
No
|
55 (21.8%)
|
Chemotherapy setting
|
Adjuvant
|
195 (77.4%)
|
Neoadjuvant
|
57 (22.6%)
|
Abbreviations: AJCC, American Joint Committee on Cancer; ER, estrogen receptor; PgR, progesterone receptor; pos., positive; HER2, human epidermal growth factor receptor 2; TXL, paclitaxel; TRAST, trastuzumab; CBDCA, carboplatin; EC, epirubicin – cyclophosphamide; TC, docetaxel - cyclophosphamide; PEMBRO., pembrolizumab; CMF, cyclophosphamide-methotrexate-5-fluorouracil; PERT., pertuzumab; UNK, unknown.
One hundred and seventeen patients had II stages of the disease according to the AJCC system classification (43%), 31% and 22% of patients had stage I and III respectively. Patients received different chemotherapy regimens, mainly anthracyclines containing regimens (78.2%) and in the adjuvant setting (77.4%)
4.2 Scalp Cooling Efficacy
The scalp cooling device success rate (Dean’s alopecia scale score 0-2 equal to a hair loss ≤ 50%) was 72% for all evaluable patients at the end of chemotherapy treatment (248 out of 252). Differences in response to SC device according to the chemotherapy regimens were listed in Table 2.
Table 2 Features related to chemotherapy regimens used and hair loss degree evaluated by medical staff at the end of all cycles of chemotherapy
Chemotherapy regimens
|
Treatment completed
N. (%) of Patients
|
Hair loss ≤ 50%
N. (%) of Patients
|
Hair loss >50%
N. (%) of Patients
|
Treatment prematurely stopped (%)
|
Reason (n)
|
EC→TXL weekly
|
59 (81.9%)
|
39 (66.1%)
|
20 (33.9%)
|
13 (18.1%)
|
Alopecia (7), coldness (3), others (3)
|
EC
|
64 (88.9%)
|
41 (64.1%)
|
23 (35.9%)
|
8 (11.1%)
|
Alopecia (2), pain (2), others (2)
|
TXL + TRAST. Weekly
|
46 (100%)
|
46 (100%)
|
|
|
|
EC→TXL+TRAST.weekly
|
27 (80%)
|
18 (66.7%)
|
9 (33.3%)
|
8 (22.9%)
|
Alopecia (5), coldness (1), other (1)
|
EC→TXL + CBDCA weekly
|
5 (100%)
|
2 (40%)
|
3 (60%)
|
|
|
EC→TXL+TRAST.+PERT. weekly
|
3 (60%)
|
3 (100%)
|
|
2 (40%)
|
Others (2)
|
(TXL+CBDCA weekly→EC )+PEMBRO.
|
4 (100%)
|
4 (100%)
|
|
|
|
TC
|
5 (80%)
|
3 (75%)
|
1 (25%)
|
1 (20%)
|
Alopecia
|
EC→CMF
|
2 (100%)
|
1 (50%)
|
|
|
|
CBDCA+TXL+TRAST. Weekly
|
1 (100%)
|
1 (100%)
|
|
|
|
EC→CMF+TRAST
|
1 (100%)
|
|
1 (100%)
|
|
|
Abbreviation: TXL, paclitaxel; TRAST, trastuzumab; CBDCA, carboplatin; EC, epirubicin – cyclophosphamide; TC, docetaxel - cyclophosphamide; PEMBRO., pembrolizumab; CMF, cyclophosphamide-methotrexate-5-fluorouracil; PERT., pertuzumab;
Efficacy of 100% in preventing hair loss > 50% was observed in all patients undergoing paclitaxel weekly-based chemotherapy. Moreover, among 46 patients treated with weekly paclitaxel-based chemotherapy (forty-five received paclitaxel-trastuzumab and one paclitaxel-carboplatin-trastuzumab regimen) no one has had a hair loss evaluated by Dean's scale more than G1 and the majority (42 out of 46) did not experienced any hair loss (G0 according to Dean's scale). Even if only four patients received chemotherapy with anthracyclines after weekly paclitaxel, all these patients had hair loss ≤ 50% (Figure 1).
Table 3 The hair loss degree related to anthracycline based versus anthracycline free regimes
Variable
|
Hair loss ≤ 50%
N. (%) of Patients
|
Hair loss >50% (%)
N. (%) of Patients
|
p-value
|
Anthracycline free chemoterapy
|
49 (98.0%)
|
1 (2.0%)
|
<0.001*
(Cramer’s V: 0.310)
|
Anthracycline free chemoterapy
|
108 (65.5%)
|
57 (34.5%)
|
p-value: Fisher’s exact test; *anthracycline based chemotherapy versus anthracycline free chemotherapy
In our series, hair thickness, hair straightening and hair permanent were related to the risk of hair loss; hair structure, previous alopecia, dyed hair and hair density were not significantly related to scalp-cooling failure (Table 4 and Table 5)
Table 4. The hair loss degree according to the hair’s patient characteristics at baseline.
|
N (%) of Patients
|
Hair loss ≤ 50%
N. of Patients (%)
|
Hair loss > 50%
N. of Patients (%)
|
p-Value
|
Hair thickness
|
Fine
|
76 (30.5%)
|
59 (77.6%)
|
17 (22.4%)
|
0,02*
|
Medium
|
94 (37.7%)
|
72 (76.6%)
|
22 (23.4%)
|
Thick
|
79 (31.7%)
|
47 (59.5%)
|
32 (40.5%)
|
Hair structure
|
Straight
|
127 (51%)
|
97 (76.3%)
|
30 (23.6%)
|
0,07*
|
Wavy
|
88 (36%)
|
61 (69.3%)
|
27 (30.6%)
|
Curly
|
32 (13%)
|
18 (56.2%)
|
14 (43.7%)
|
Hair density
|
Low
|
156 (63.1%)
|
109 (69.9%)
|
47 (30.1%)
|
0,64*
|
Medium
|
88 (35.6%)
|
64 (72.7%)
|
24 (27.3%)
|
High
|
3 (1.2%)
|
3 (100%)
|
0 (0.0%)
|
Pevious alopecia
|
Yes
|
10 (4.1%)
|
10 (100%)
|
0 (0.0%)
|
0,67*
|
No
|
237 (95.9%)
|
166 (70.0%)
|
71 (30.0%)
|
Dyed hair
|
Yes
|
171 (69%)
|
121 (70.7%)
|
51 (29.8%)
|
0,64*
|
No
|
76 (31%)
|
56 (73.6%)
|
20 (26.3%)
|
Hair straightening
|
Yes
|
37 (15%)
|
144 (68.5%)
|
66 (31.4%)
|
0,03*
|
No
|
210 (85%)
|
32 (86.4%)
|
5 (13.5%)
|
Hair permanent
|
Yes
|
8 (3%)
|
7 (88%)
|
1 (13%)
|
0,44*
|
No
|
239 (97%)
|
169 (71%)
|
70 (29%)
|
*p-value: Fischer’s exact test.
Table 5. The hair loss degree according to the patient characteristics.
Variables
|
Hair loss ≤ 50%
N. of Patients (%)
|
Hair loss > 50%
N. of Patients (%)
|
p-Value
|
BMI
|
Overweight
|
61 (80.3%)
|
15 (19.7%)
|
0,003*
|
Healthy weight
|
82 (64.1%)
|
46 (35.9%)
|
Underweight
|
5 (50.0%)
|
5 (50.0%)
|
Obese
|
29 (87.9)
|
4 (12.1)
|
Smokers
|
Yes
|
22 (84.6%)
|
4 (15.4%)
|
0,114*
|
No
|
137 (69.2%)
|
61 (30.8%)
|
Age
|
Under 50
|
94 (69.6%)
|
41 (30.4%)
|
0,480*
|
Over 50
|
83 (74.1%)
|
29 (25.9%)
|
Menopausal status
|
|
|
|
Premenopausal
|
91 (67.4%)
|
44 (32.6%)
|
0,117*
|
Postmenopausal
|
83 (76.9%)
|
25 (23.1%)
|
*p-value: Fischer’s exact test.
Univariate and multivariate logistic regression analysis of hair loss degree according to the patient and treatment characteristics were reported in Table 6. The multivariate analysis identified the type of chemotherapy regimen, BMI and high hair thickness as independent factors for hair loss.
Table 6. Univariate and multivariate analysis
Variable
|
Univariate
|
Multivariate
|
OR (95% CI)
|
p-Value
|
OR (95% CI)
|
p-Value
|
Anthracycline based chemotherapy
|
|
|
|
|
No
|
Ref.
|
0.002
|
Ref.
|
|
Yes
|
25.1743 (3.387-187.096)
|
21.447 (2.845-161.676)
|
0.003*
|
Hair thickness
|
|
|
|
|
Medium
|
Ref.
|
|
Ref.
|
|
Fine -Medium
|
0,918 (0.402-2.096)
|
0.838
|
1.031 (0.419-2.535)
|
0.946
|
Thick - Medium
|
2.490 (1.218-5.091)
|
0.012
|
2.577 (1.185-5.601)
|
0.017*
|
Hair straightening
|
|
|
|
|
No
|
Ref.
|
|
Ref.
|
|
Yes
|
0.428 (0.157-1.165)
|
0.097
|
0.391 (0.132-1.162)
|
0.091
|
Hair permanent
|
|
|
|
|
No
|
Ref.
|
|
|
|
Yes
|
0.385 (0.0463-3.202)
|
0.377
|
|
|
BMI
|
|
|
|
|
Healthy Weight
|
Ref.
|
|
Ref.
|
|
Overweight-Healthy weight
|
0.438 (0.2242-0.857)
|
0,016
|
0.549 (0.247-1.219)
|
0.141
|
Under-weight-Healty-weight
|
1.783 (0.4902-6.483)
|
0,380
|
3.519 (0.724-17.099)
|
0.119
|
Obese-Healthy-weight
|
0.246 (0.0814-0.743)
|
0,013
|
0.250 (0.067-0.926)
|
0.038*
|
*p-value < 0.05
4.3 Adverse Events
Sixty-nine percent of all evaluable patients had almost one adverse event, 13% of them early interrupted the use SC system, the majority within the first cycles. The most frequent cause was incoming alopecia in 15 patients (47%). Fear of scalp metastases was the reason of early discontinuation in one patient. Five patients early discontinued due to coldness and two patients due to cervical pain and headache respectively.
Table 7 show the frequencies of the adverse events associated with the SC device reported by patients. Coldness was the most frequently adverse event registered in 152 out of 247 (62%) patients, severe in 43%. The second most common adverse event was headache (129/247), mostly mild in gravity (24%).
Table 7. Frequencies of SC adverse events and their gravity according to the patient’s description.
Adverse Events
|
No (%)
|
Yes (%)
|
Mild (%)
|
Moderate (%)
|
Severe (%)
|
Headache
|
118 (47.8%)
|
129 (52.2%)
|
66 (51.2%)
|
39 (30.2%)
|
24 (18.6%)
|
Scalp Pain
|
138 (55.9%)
|
109 (44.1%)
|
54 (49.1%)
|
37 (33.6%)
|
19 (17.3%)
|
Neck Pain
|
147 (59.5%)
|
100 (40.5%)
|
49 (49.5%)
|
35 (35.4%)
|
15 (15.2%)
|
Coldness
|
94 (38.1%)
|
153 (61.9%)
|
54 (35.8%)
|
36 (23.8%)
|
61 (40.4%)
|
Heavy Head
|
110 (44.5%)
|
137 (55.5%)
|
51 (37.2%)
|
64 (46.7%)
|
22 (16.1%)
|
4.4 Follow-up
Median follow-up was 35 months (range 12-62) for 248 evaluable patients; fifteen (6%) patients had a relapse of breast cancer. The site of relapse was known in ten patients: one local, six in the liver, two in the brain and one in the bone. At the end of the data cut-off, 11 died for breast cancer relapse (4.4%). No scalp metastasis was reported among the study participants.