Matched case-control phase 2 study to evaluate the use of a frozen sock to prevent docetaxel-induced onycholysis and cutaneous toxicity of the foot




Onycholysis occurs in approximately 30% of patients treated with docetaxel. The efficacy and safety of an Elasto-Gel frozen sock (FS) was investigated for the prevention of docetaxel-induced nail and skin toxicity of the feet.


Patients receiving docetaxel at a dose of 70 to 100 mg/m2 every 3 weeks were eligible for this matched case-control study. Each patient wore an FS for 90 minutes on the right foot. The unprotected left foot acted as control. Nail and skin toxicities were assessed using National Cancer Institute Common Toxicity Criteria (version 3) and compared using a 2-sample Wilcoxon matched-pairs rank test adjusted for tied values.


Fifty consecutive patients were included between April 2005 and January 2007. Nail toxicity was significantly lower in the FS-protected foot compared with the control foot (grade 0: 100% versus 79%; and grade 1 and 2: 0% versus 21%, respectively) (P = .002). Skin toxicity was grade 0: 98% versus 94%; and grade 1 and 2: 2% versus 6% in the FS-protected and the control feet, respectively. The median times until toxicity occurrence were not found to differ significantly between the groups. One patient experienced discomfort because of cold intolerance.


Cold therapy using FS significantly reduced the incidence of docetaxel-induced foot nail toxicity, as previously demonstrated using frozen gloves for the hands. Cancer 2008. © 2008 American Cancer Society.

The taxanes (docetaxel and paclitaxel), have emerged as among the most powerful classes of cytotoxic agents, exhibiting a wide spectrum of activity against solid tumors including breast, ovarian, lung, head and neck, stomach, and prostate cancer.1 Taxanes have a predictable and manageable toxicity profile. Specific toxicities include asthenia, alopecia, cutaneous reactions, stomatitis, fluid retention syndrome for docetaxel, and neuropathy, myalgia, and arthralgia for paclitaxel.2–5 Myelosuppression and hypersensitivity reactions are common for both drugs. Cutaneous toxicity reported with docetaxel treatment manifested as erythema and desquamation of the skin of the extremities (hand and foot syndrome) and nail changes.6–8 Although hand and foot syndrome does not appear to be a common adverse event with docetaxel treatment,6 nail changes were more often described with docetaxel than with paclitaxel, with an overall incidence of approximately 30%.7 In a recent study performed in 55 metastatic breast cancer patients treated with docetaxel, the incidence of nail toxicity reached 58% after 4 cycles, with an increase to 88% after 3 additional cycles.9 A large proportion of patients experienced the nail changes as a cosmetic nuisance, and >32% had functional problems.9

Nail changes include hyperpigmentation, splinter hemorrhage, subungual hematoma or/and hyperkeratosis, orange discoloration, Beau-Reil lines (indicating the cessation of nail growth), acute paronychia, and onycholysis (the loosening or separation of a fingernail or toenail from its bed).10 These abnormalities are in most cases not serious, but hemorrhagic onycholysis and subungual abscesses can occur, producing important morbidity. These side effects can cause potentially severe complications because patients may become neutropenic during chemotherapy, enhancing the risk of sepsis. Functional disturbance can also occur, leading to patients withdrawing from treatment. The type of nail changes is related to the number of cycles of chemotherapy, and to our knowledge to date few preventive measures are available.11 Topic antibiotics or antifungal agents may be required to treat nail bed infections, which appear to be a complication after the nail's detachment.

In a previous study, we reported that the use of an Elasto-Gel frozen glove (−25 to −30°C) significantly reduced the incidence of nail and skin toxicity of the hands induced by docetaxel administered at a dose of 75 mg/m2 every 3 weeks, either alone or in combination.12 The application of frozen gloves significantly decreased the overall occurrence of nail toxicity from 51% to 11%, with onycholysis occurring in none of the protected hands compared with 22% in the control hand. Skin toxicity was observed in 24% versus 53% of patients (protected vs unprotected hand), respectively.12 These favorable results warrant additional studies to confirm the efficacy of cold protection in preventing nail and/or cutaneous toxicities related to docetaxel.

To our knowledge, there are no data regarding the incidence of specific docetaxel-induced toxicity to the foot. The current matched, case-control, phase 2 study was designed to assess the efficacy and safety of cold therapy in the prevention of docetaxel-induced onycholysis and skin toxicity of the foot. Each patient wore a protective frozen sock (FS) on the right foot and had no protection on the contralateral foot, which was considered the control side. Using this method, it was possible to obtain a comparative incidence of nail and skin disorders with or without cold protection.



Patients enrolled in this prospective study were undergoing treatment for a variety of tumor types with docetaxel administered at doses of between 70 mg/m2 and 100 mg/m2 as a 1-hour intravenous infusion every 3 weeks, either alone or in combination with other cytotoxic agents. Inclusion criteria were as follows: no prior treatment with taxanes, the absence of skin and nail disorders at the initiation of chemotherapy, a life expectancy of at least 3 months, and an Eastern Cooperative Oncology Group (ECOG) performance status (PS) ≤2. Patients were excluded if they had Raynaud syndrome, distal bone or cutaneous metastases, ungual pathology, arteriopathy, cold intolerance, or peripheral neuropathy (grade ≥2). All patients provided written informed consent before inclusion. The study was approved by an external Ethics Committee and was declared to the French National Health Agency.

Cold Therapy

Patients wore an Elasto-Gel (Akromed, France) flexible FS. This patented sock contains glycerin, which has thermal properties that allow its use in hot or cold therapies (Figure 1). The gel-filled sock covers the foot to the ankle. Before use it was refrigerated for at least 3 hours at −25°C to −30°C. For each docetaxel infusion, the patient wore an FS for a total of 90 minutes on the right foot (15 minutes before the administration of docetaxel, during the 1-hour docetaxel infusion, and 15 minutes after the end of the infusion). The study was not blinded. Two FS were used successively (for 45 minutes each) to maintain a consistently low temperature of the foot. The left foot was not protected by the FS and acted as the control.

Figure 1.

The Elasto-Gel (Akromed, France) flexible frozen sock.

Toxicity and Patient Satisfaction

Onycholysis and skin toxicity were assessed before each cycle of chemotherapy by the investigator (F.S. or E.L.), and double-checked by a second observer (a research nurse), using the National Cancer Institute Common Toxicity Criteria (NCI-CTC, version 3). The NCI-CTC nail and skin toxicities criteria were updated in the most recent version, the third one.13 The nail disorder categories were expanded into 3 grades, similar to that noted in hand and foot syndrome: grade 1, discoloration, ridging (koilonychia), or pitting; grade 2, partial loss of nail(s) (onycholysis) or pain in nail beds not interfering with function; and grade 3, partial loss of nail(s) (onycholysis) or pain in nail beds interfering with function, or complete loss of nail(s). Patient comfort was assessed using a 4-point rating scale that determined whether patients were: dissatisfied (0), not satisfied (1), satisfied (2), or very satisfied (3). The use of the FS was prematurely stopped when patients experienced cold intolerance, had a serious adverse event, or withdrew consent.

Statistical Analysis

The primary endpoint was the efficacy in preventing nail toxicity, and secondary endpoints were the assessment of efficacy in the prevention of skin toxicity, time to occurrence of nail and skin toxicities, and patient comfort. Analyses of toxicities were performed on the per-protocol population defined as the totality of included patients, which used the FS at least 1 time at the first cycle. The 2-sample Wilcoxon matched-pairs rank test adjusted for tied values was used as the main method to determine the statistical significance of difference between the incidence of nail and skin toxicities between FS-protected and unprotected feet. Because time-to-event endpoints were used, Kaplan-Meier and log-rank methods were used to estimate and compare differences in time to toxicity occurrence. Patients were censored if no toxicity occurred at the end of chemotherapy or during follow-up. The impact on the time to nail toxicity occurrence of some confounders such as the ECOG PS, sex, number of cycles, and dose of docetaxel was studied using multivariate Cox regression analysis (forward stepwise selection). Hazard ratios (HRs) were estimated with 95% confidence intervals (95% CIs). Medians with range and frequency percentages were estimated for quantitative and categoric variables, respectively. No adjustment for multiplicity of tests was performed. Subgroup analyses were performed according to the docetaxel dosing schedule: ≤75 mg/m2 and >75 mg/m2. The relational database was created using FileMaker Pro 8.5 (vesion 2; Santa Clara, Calif) software. The statistical analysis was performed using SPSS 15 software (SPSS Inc, Chicago, Ill) and Epi Info (version 3.2.2; Centers for Disease Control and Prevention, Atlanta, Ga).


Fifty consecutive patients were enrolled in a single French center between April 2005 and January 2007. One patient was found to be ineligible at the timeof study entry because of previous taxane treatment. A second patient withdrew informed consent before the start of chemotherapy. Therefore, data for 48 patients were analyzed as a per-protocol population.

Patient Characteristics

The characteristics of the patients at inclusion are summarized in Table 1. There were 36 men and 12 women, with a median age of 62 years, treated with docetaxel for prostate (52%), lung (21%), breast (19%), or other cancers (urinary bladder, head and neck, primary unknown for 8% of cases). The majority of patients (88%) had an ECOG PS of 0 or 1. Docetaxel was administered either as monotherapy (42%) or in combination therapy (58%) as a first-line treatment for 69% of patients. The other agents in combination were estramustine phosphate (56%), carboplatin (16%), cisplatin (12%), and others (erlotinib, trastuzumab, doxorubicin, and cyclophosphamide) in 16% of cases. The median number of cycles administered was 5 (range, 1–9 cycles) and the median cumulative dose was 720 mg (range, 150–1260 mg). There were 4 different dosages at Day 1 of the first cycle of docetaxel: 70 mg/m2, 75 mg/m2, 85 mg/m2, and 100 mg/m2; the percentage of corresponding patients were 38%, 29%, 6%, and 27%, respectively. Twelve of 48 patients (25%) experienced a docetaxel dose reduction, principally because of hematologic toxicity.

Table 1. Patient Characteristics (Per-protocol Population, n = 48)
CharacteristicMedian (Range)No. of patients%95% CI
  1. 95% CI indicates 95% confidence interval; ECOG, Eastern Cooperative Oncology Group.

SexMale 367560–86
Female 122514–40
Age, y 62 (36–80)   
ECOG performance status0 224631–61
1 204228–57
2 6125–25
Type of tumorProstate 255237–68
Lung (nonsmall cell) 102111–35
Breast 9199–33
Other 482–20
Docetaxel treatmentSingle agent 204228–57
Combination therapy 285843–72
No. of cycles5 (1–9)   
Cumulative dose, mg720 (150–1260)   
Docetaxel schedule, mg/m270 183824–53
75 142917–44
85 361.3–17
100 132715–42
Prior chemotherapyYes 153119–46
No 336954–81

Nail and Skin Toxicity

The application of a FS significantly reduced the incidence of nail toxicity, with grade 1–2 toxicity occurring in none of the FS-protected feet compared with 21% in the unprotected feet (P = .002) (Table 2). Onycholysis occurred in only 2% of the control feet (Figures 2 and 3). Overall skin toxicity was registered in 2% of the FS-protected feet versus 6% of the unprotected feet. The difference was not statistically significant (P = .18) (Table 2). The time until nail toxicity occurrence for unprotected foot was significantly associated with the total number of cycles of docetaxel administered (HR of 0.36; 95% CI, 0.17–0.77; [P = .008]); a borderline correlation was found for ECOG PS (HR of 0.31; 95% CI, 0.09–1.10 [P = .07]). Patients with a poor ECOG PS presented with a short time interval until the occurrence of nail toxicity, with the risk found to be correlated with the cumulative number of cycles of chemotherapy. Age, previous chemotherapy, and the weight of the patients were not found to be associated with the risk of nail toxicity. The percentage of patients with no nail toxicity at 1 month, 2 months, and 3 months for the control foot were 98%, 81%, and 67%, respectively. No differences were observed in terms of time to the occurrence of nail and skin toxicity between protected and unprotected feet (105 days vs 87 and 101 days vs 101 days, respectively). An exploratory analysis performed on the data of patients treated with docetaxel in combination with different agents demonstrated no impact of the introduction of a second drug on the endpoint of study (nail toxicity).

Figure 2.

Grade 1 nail toxicity (dyschromia) (using National Cancer Institute Common Toxicity Criteria [version 3]), which was present on the (control) left foot but absent on the protected (right) foot.

Figure 3.

Grade 2 nail toxicity (onycholysis) (using National Cancer Institute Common Toxicity Criteria [version 3]), which was present on the (control) left foot but absent on the protected (right) foot.

Table 2. Toxicity*
Grade of toxicityNail toxicity after docetaxel treatment
Control foot(n = 48)Protected foot(n = 48)P
%95% CI%95% CI
  • 95% CI indicates 95% confidence interval.

  • *

    Toxicity was graded using National Cancer Institute Common Toxicity Criteria (version 3).

 Skin toxicity after docetaxel treatment 
 Control foot (n = 8)Protected foot (n = 48) 
Toxicity grade%95% CI%95% CIP

Patient Comfort

Assessment of patients' global comfort included factors such as sock contact, temperature tolerance, and immobilization constraints. Greater than half of the patients (58%) were satisfied and even very satisfied (19%) with the FS protection. Only 1 patient (2%) was dissatisfied because of FS-related cold intolerance (Table 3).

Table 3. Patient Comfort
Comfort assessmentGlobal comfort(n = 48)Cold tolerance(n = 48)
%95% CI%95% CI
  1. 95% CI indicates 95% confidence interval.

Very satisfied199–33178–30


Taxanes are important drugs in the treatment of breast, head and neck, lung, stomach, and prostate cancer. Although nail and skin toxicities are not life-threatening, clinicians should manage these particular side effects that could lead to an impairment of quality of life and early discontinuation of chemotherapy.14 Nail toxicity is now considered an important issue in taxane therapy. Even weekly schedules of docetaxel demonstrated a potential risk of nail toxicity, as demonstrated by a review performed in breast cancer patients.15 Not only taxanes are associated with important nail changes; antiangiogenic agents such as sunitinib and sorafenib were associated with distal subungual splinter hemorrhages in 30% and 60% of cases, respectively.16

Cold temperature had been found to decrease the toxicity (incidence and severity) of some cytotoxic agents as a direct consequence of the induced vasoconstriction, associated with a temporary decrease of blood flow. This approach was applied to prevent alopecia with the use of cold caps17, 18 and to prevent 5-fluorouracil-related mucositis using oral cryotherapy.19 The same concept was successfully tested for the prevention of nail and skin toxicities induced by docetaxel using an Elasto-Gel flexible frozen glove12 and therefore was extended in the current study to nail changes of the foot using a similar FS.

This matched, case-control, phase 2 study, performed in 45 patients with solid tumors who were treated with docetaxel-based regimens, demonstrated the feasibility and efficacy of cold therapy using the Elasto-Gel flexible FS to prevent docetaxel-induced nail toxicity of the foot. We believe the usefulness of cold therapy in preventing docetaxel-induced nail toxicity was confirmed and validated in the current study. The use of an FS reduced the incidence and intensity of nail toxicity associated with docetaxel, administered every 3 weeks, alone or in combination. No nail toxicity was observed in the FS-protected foot compared with an incidence of 21% in the control foot, including dischromia and onycholysis for 19% and 2% of cases, respectively. The incidence of skin toxicity was also reduced, but this finding was not statistically significant. Few patients presented grade 1 or 2 skin toxicity in the control group. In daily clinical practice, feet are less frequently affected than the hands by docetaxel-induced nail changes. This may be because of sock wear and light exposure prevention. In the previous study assessing frozen gloves, grade 1 and 2 nail toxicity occurred in 29% and 22% of the unprotected hand (compared with 19% and 2% in the unprotected foot of the current study), respectively. Grade 1 and 2 skin toxicity occurred in 44% and 9% of the control hand, compared with 4% and 2% in the control feet, respectively.12 The appearance of nail and/or skin toxicities of the foot was not significantly delayed by cold therapy, as had been observed using the frozen gloves. Overall, it appears that the FS was less effective, but the results are still statistically significant despite the small sample size for this cohort of patients.

The global comfort of patients as well as the cold tolerance assessed using a 4-point rating scale were excellent; only 1 patient was not satisfied with the FS, experiencing cold intolerance. These safety results appeared better than those obtained in the frozen gloves study, in which 14% of patients were dissatisfied, including 11% of those who withdrew from the study because of cold intolerance during glove contact.12 Therefore, the FS appears to be better tolerated than the gloves.

To our knowledge, the physiopathology of nail toxicity is actually unknown. Some studies have suggested that the antiangiogenic properties of taxanes may be involved,20, 21 whereas another study hints at the existence of a neurogenically mediated inflammatory process.22 De Giorgi et al.23 suggested that onycholysis might represent a side effect related to the vehicle cremophor for paclitaxel or Tween 80 for docetaxel, possibly through neurotoxic damage.

To our knowledge to date, no effective preventive measures are available to avoid taxane-induced nail changes. This technique, using Elasto-Gel frozen gloves and/or socks, is easy to apply and is well tolerated by the patients, with no major side effects reported. It may be recommended for routine use with taxane-based therapies and should be considered as a new valuable tool in supportive care. Because the FS did not affect skin toxicity, a modified version of the sock, covering only toes and nails, should be tested. The cold therapy must be tested with other cytotoxic agents known to be associated with important nail and skin toxicities such as liposomal doxorubicin. A similar study is ongoing in our center and the results will be presented in the near future.