The hand-foot-syndrome (HFS, palmoplantar erythrodysesthesia, chemotherapy-associated acral erythema) is characterized by painful predominantly palmo-plantar lesions. The association with different chemotherapeutic agents has been known for over 20 years. More recently, HFS has been reported in association with regimens using targeted agents, in particular the multikinase inhibitors (MKI) sorafenib and sunitinib. The HFS associated with MKI has a different distribution and clinical appearance than the traditional disorder. In this review, similarities and differences between chemotherapy- and MKI-associated HFS are discussed and current recommendations for their prophylaxis and management are summarized.
Hand-foot syndrome (HFS) was first reported in 1974 . Because it occurred in patients after beginning cytotoxic chemotherapy, with rapid improvement upon discontinuing therapy, it was believed to be a side effect associated with various cytostatic agents . Nevertheless, there were isolated reports of clinical appearances similar to chemotherapy-associated HFS in patients with neoplasia [3, 4] or graft-versus-host disease . HFS is characterized by painful erythematous lesions which mainly affect palmoplantar surfaces. Thus the syndrome is also referred to as palmoplantar erythrodysesthesia and chemotherapy-induced acral erythema .
It is still unclear whether the disorder is a single disease entity or whether it comprises a mixed collection of various diseases with various underlying mechanisms .
In recent years, HFS has also been often reported as a side effect associated with targeted cancer therapies. It is one of the most commonly reported adverse effects in patients taking the multikinase-inhibitors (MKI) sorafenib and sunitinib. The clinical presentation of HFS during MKI therapy differs from that caused by chemotherapy, and thus some authors distinguish classic from MKI-associated HFS [7, 8].
HFS is rarely life-threatening, but it can significantly impair the quality of life of the patient, possibly necessitating a dose reduction or interruption of therapy thus limiting the use of a potentially effective therapy. This is especially regrettable given with certain substances, such as capecitabine (manufacturer's information on Xeloda®), 5fFluorouracil (5-FU) , and sorafenib , HFS is associated with a clinical response of the tumor to therapy.
We review the current knowledge on the clinical appearances, pathogenesis, prevention, and therapy of HFS from chemotherapy and MKI.
HFS is a common side effect of anthracyclines (doxorubicin, pegylated liposome-encapsulated doxorubicin [PLD]), taxanes (docetaxel), and pyrimidine analogues (cytarabine = cytosine arabinoside, 5-FU and derivatives such as capecitabine). Table 1 summarizes the frequency of HFS in patients taking these substances. HFS has also been reported, though much less frequently, with the use of various other substances such as paclitaxel, hydroxyurea, methotrexate, 6-mercaptopurin, cyclophosphamide, cisplatin, daunorubicin, etoposide, vinorelbine, irinotecan, and epirubicin [2, 11].
Table 1. Reported incidences of hand-foot-sydrome associated with various medical tumor therapies.
Among new “targeted” agents now being used in cancer therapy, HFS is one of the most common side effects of the MKIs sorafenib and sunitinib (Table 1).
Clinical appearance, severity, and histology
HFS from chemotherapy
The clinical presentation of the disorder does not vary depending on the chemotherapy drug used, and thus usually does not yield any clues as to the responsible drug .
After starting therapy, patients first experience palmoplantar dysesthesia. Symptoms worsen within a few days with burning pain and sharply demarcated erythema with edematous swelling which may develop into blistering, desquamation of the horny layer, and subsequent ulceration or erosions (Figure 1). There are also isolated reports of thickening of the horny layer resembling keratoderma [12, 13]. Lesions are mainly found on palmoplantar regions (although involvement of the soles of the feet is often more severe); in severe cases, the dorsal aspects of the hands and feet as well as intertriginous areas and regions underlying tight-fitting clothing may be affected (Figure 1). Some authors have referred to these lesions as “intertrigo-like dermatitis”, and there is overlap with a diagnosis of chemotherapy-associated eccrine squamous syringometaplasia, which is commonly seen with palmoplantar involvement such as in HFS .
Depending on the severity of skin lesions and symptoms, three (NCI classification) or four (WHO classification) degrees of severity are distinguished (Table 2).
Table 2. Grades of severity for hand-foot-syndromes .
WHO correlation histology
National Cancer Institute (NCI-CTCAE* Version 3.0)
National Cancer Institute (NCI-CTCAE* Version 4.02)
* National Cancer Institute Common Terminology Criteria for Adverse Events.
(Decent) erythema, dysesthesia, normal activity possible
Minimal skin changes (e.g. erythema), no pain
Minimal skin changes (e.g. erythema, swelling, or hyperkeratosis), no pain
Erythematous swelling, palmoplantar pain under mechanical stress
Isolated keratinocyte necrosis in higher epidermal layers
Skin lesions (e.g., blistering, bleeding, swelling) or pain, no impairment of functioning
Skin lesions (e.g., blistering, bleeding, swelling, or hyperkeratosis) with pain and restricted daily activities
Ulcerating dermatitis or skin lesions with pain and restricted function
Severe skin lesions (e.g., blistering, bleeding, swelling, or hyperkeratosis) with pain and limited autonomy
Highly painful erythematous swelling, possible blistering and ulcers, impossible to perform usually daily activities
Necrosis throughout the epidermis
In general HFS occurs in the initial weeks of therapy with the triggering agent. For some medications, such as capecitabine and cytarabine, HFS can also occur a few months after beginning therapy [16, 11].
After stopping or interrupting treatment, there is significant improvement of symptoms and healing of skin lesions within weeks which is often accompanied by exfoliative scaling . Permanent damage rarely results after discontinuing the responsible drug. Persistent dysesthesia or cutaneous lesions such as erythema and scales may remain .
Histology of chemotherapy-associated HFS is consistent with the changes due to toxic effects and is not typical for a specific trigger [2, 11]. Histological correlates vary depending on the degree of clinical severity (Table 2) [18, 11].
In grade I disease there is dilation of the capillaries as well as abnormities affecting the basal keratinocytes. In grade II disease there is dilation of the capillaries, papilledema, interface dermatitis, and occasional keratinocyte necrosis. Grade III is characterized by confluent necrotic keratinocytes, especially in the basal layer with separation of the epidermis, edema, and interface dermatitis.
In grade IV disease, necrosis affects the entire epidermis. Hyper- or parakeratosis is common; there is usually only a mild perivascular lymphohistiocytic infiltrate in the dermis, and possible eccrine squamous syringometaplasia [11, 19] (Figure 2).
When histology reveals eccrine squamous syringometaplasia and there is also involvement of intertriginous areas along with the hands and feet, some authors prefer the term “chemotherapy-induced eccrine squamous syringometaplasia” (see below) .
HFS from MKI
During sorafenib therapy HFS usually begins within the first weeks of therapy while in patients taking sunitinib, the onset has been reported even a few months after beginning treatment . One recent publication reported the median onset of HFS caused by sorafenib as occurring after 18.4 days (range of 3–56 days) after beginning therapy; for sunitinib the median onset was after 32.4 days (5–82 days) .
Lesions are especially pronounced on pressure points on the palms and soles of the feet, but can also affect the margins of the feet and the skin between the fingers and toes if these are exposed to mechanical stress (Figure 3) [18, 21]. There have also been reports of lesions occurring at other sites that are under mechanical stress such as around the stump of an amputated leg  or about the elbow  or on the fingertips after intense use of a mobile communication device . Unlike HFS from chemotherapy, there have been as yet no reports of lesions on the dorsal aspects of the hands, feet, or in intertriginous areas. Clinical presentation includes initial dysesthesia and erythema, especially with mechanical or thermal stress. This is followed by a worsening of sharply demarcated erythema which is increasingly painful and callus-like thickening of the horny layer in the erythematous areas. Hence the term “palmoplantar epidermal hyperplasia” is also used . In the thickened areas of the horny layer there may also be large, tense blisters (Figure 3). The clinical appearance of HFS caused by MKI can worsen significantly within days (Figure 4).
As seen in chemotherapy-associated HFS, in MKI-associated HFS there are also histological signs of keratinocyte damage in the epidermis and dermis with dilated blood vessels, a mild to moderate perivascular lymphohistiocytic infiltrate, and occasionally syringometa-plasia of the eccrine sweat glands. There are also signs of accelerated epidermal cell proliferation with acanthosis, papillomatosis, hyperkeratosis, and parakeratosis [25, 18, 21] (Figure 2).
Table 3 compares hand-foot syndrome caused by chemotherapy and by MKI therapy.
Table 3. Comparison of hand-foot-syndrome associated with chemotherapy with that from multikinase inhibitor (MKI) therapy.
Sorafenib: 2nd–4th week of therapy Sunitinib, often after 1–3 months of therapy
With higher-dose pulse therapies: after 1–21 days, with continuous low-dose therapies: up to several months
1–2 weeks after stopping
1–2 weeks after stopping
Tends to be at local pressure points palmoplantar
Patches, can also affect other regions of the body (e.g., intertriginous areas, dorsal aspects of hands/feet)
Painful, callus-like hyperkeratosis and blisters, sometimes surrounded by an erythematous halo
Painful erythema, edema, blistering, desquamation
Possible vascular mechanism
Possible sweat-associated toxicity
The clinical and histological changes seen in HFS are compatible with toxic damage to the skin barrier [2, 11]. There are also additional signs of toxic damage: (1) dose-dependence, which has been reported for chemotherapy drugs such as 5-fluorouracil, capecitabine, PLD [26, 27], cytarabine , and also for MKIs such as sorafenib [28, 10] (Table 1); (2) the more common occurrence after prolonged exposure to the responsible substance (such as with longer-term or continuous low-dose infusion therapies, liposomal preparations with a long half-life, or continuous oral use); (3) an increase in the frequency and severity of HFS with the combined use of two substances which both can cause HFS (Table 1). For instance, the use of docetaxel is an independent risk factor for HFS in patients on capecitabine therapy .
Several hypotheses have been put forth concerning the underlying pathogenetic mechanisms of harm:
1Excretion from the sweat glands leads to a high concentration in the areas of the skin around the sweat glands . This is supported by the observation that certain substances such as taxane and liposome-encapsulated doxorubicin can also affect areas besides the hands and feet that have an abundance of sweat glands such as inguinal areas and the axillae (Figure 1). For certain substances such as PLD elevated concentrations have also been identified in sweat . Yet only in some patients with HFS does histology show changes to the sweat glands, especially in the form of syringometaplasia of the eccrine sweat glands. If, along with palmoplantar involvement, there are also changes affecting other areas of the body with an abundance of sweat glands such as seen with intertriginous lesions and if histology shows syringometaplasia of the eccrine sweat glands, this is referred to by some authors as “chemotherapy-induced syringometaplasia of the eccrine sweat glands”. The boundaries are not always clear, especially considering that the majority of patients with “chemotherapy-induced eccrine squamous syringometaplasia” also have palmoplantar involvement .
2Microtrauma to the capillaries at sites that are under mechanical stress causes the harmful agent to leak into the surrounding tissue. Especially in HFS from MKI there are signs of a connection with the anti-angiogenetic effect of MKI. HFS from MKI therapy occurs predominantly at sites under mechanical stress (Figure 3) and has not been reported in intertriginous areas . In these areas mechanical stress could cause microtrauma to the capillaries, which, given the impaired repair potential (VEGFR- and PDGF-blockade due to MKI), leads to elevated levels of the agent in the tissue. This is also supported by the finding that there is an increased rate of HFS when MKIs are combined with other anti-angiogenetic substances (which alone do not cause HFS). For example, there is a higher incidence of HFS when sunitinib is combined with interferon-α (16 % grade III) than from sunitinib therapy alone (<10 % grade III)  and it is also much higher for sorafenib in combination with the VEGF-antagonists beva-cizumab (total incidence of 79 %, grades II + III 57 %) than for sorafenib alone (total incidence of 31 %, grades II + III 30 %) . A recent publication reported that cediranib, which selectively blocks the VEGF receptor, also triggers HFS, further supporting the hypothesis that an anti-angiogenetic mechanism is at work in HFS in patients receiving MKI therapy.
3Given the high activity of relevant enzymes in the keratinocytes, breakdown products from the harmful agent can accumulate in certain areas of the skin. For example, for thymidine-phosphorylase (TP; the enzyme which activates capecitabine) and for dihydropyrimidine-dehydrogenase (DPD; the enzyme which breaks down 5-FU and capecitabine), greater activity has been reported in the palms than in the skin on the back [32, 19]. A much lower frequency of HFS has been reported in patients on combination therapy with 5-FU and DPD inhibitors (such as uracil, a natural competitor for DPD binding, CDHP, a reversible DPD inhibitor, or eniluracil, an irreversible DPD inhibitor) and in those with an hereditary DPD deficiency . This supports the relevance of the relevant enzymes and breakdown products for the pathogenesis of HFS in patients taking 5-FU or capecitabine. Yet, severe HFS has also been reported in patients hereditary DPD deficiency who are taking 5-FU therapy, suggesting that possibly other pathogenetic mechanisms may also play a role in certain cases . One possibility is elimination of the metabolites, in patients on 5-FU and capecitabine therapy which is primarily via renal excretion. One study reported that capecitabine use in patients with impaired renal function is correlated with HFS; a reduction in creatinine clearance around 10 ml/minute reportedly led to an increased risk of HFS of around 7 %.
These findings apply to chemotherapy agents as well as MKIs. Depending on the specific substance in question, other mechanisms may also play a variously important part.
Most treatment recommendations for the management of HFS in patients undergoing drug therapy for caner are based on case reports, case series, and personal experience rather than data from controlled clinical studies .
A major aspect in the treatment of HFS is modification of the tumor therapy. Reducing or stopping the drug often leads to rapid improvement (Figure 4). Both for certain chemotherapy agents such as capecitabine as well as for the MKIs such as sorafenib, the manufacturer's information contains recommendations concerning dose adjustment. If treatment is discontinued, once HFS diminishes to a level of grade 0–I, therapy may be attempted again, possibly with a dose reduction.
Various topical approaches have also been described which may help alleviate symptoms depending on their form [35, 11].
For hyperkeratosis, topical therapies containing 5–10 % salicylic acid or 10–20 % urea are used. For inflammation, topical steroids (e.g., mometasone, clobetasol) may be given. Oozing lesions may be treated with baths containing dye (such as potassium permanganate) or tanning agents. Cooling the hands and feet 3–4 daily to avoid the development of excessive warmth can also be helpful. A prospective randomized study is currently underway to test four different topical therapies for use in HFS from sorafenib. These include a topical therapy with 40 % urea, one with 0.05 % fluocinonide, one with 0.1 % tazarotene cream, and pure ointments (http://www.clinicaltrials.gov, NCT 00667589).
For therapy of HFS in patients taking doxorubicin, topical DMSO has been recommended. DMSO is a radical interceptor which has been successfully used in the treatment of doxorubicin extravasation . In HFS from PLD there is a case report with two patients on the successful use of 99 % topical DMSO 4 × daily . Its use should be restricted to clinical studies .
For treatment of HFS in patients taking purine antagonists such as 5-FU and capecitabine, topical use of the competitor uridine has been recommended . Formulations for uridine hand-foot cream and an adhesive uridine paste may be found in the New German Formulary (NRF) under “Uridin”. There are as yet no studies available on the efficacy of this therapy.
Thera are also various recommendations concerning systemic therapy of HFS. Vitamin E (300 mg orally per day) has been used in patients with HFS who are undergoing docetaxel or capecitabine therapy . Numerous case series have reported that pyridoxine (vitamin B6) 50–300 mg/day has a preventive effect in patients undergoing chemotherapy (see below) and has also been used for treatment . Systemic steroids appear to be promising based on the results of retrospective analyses in HFS patients who are taking PLD. Yet, due to insufficient data, and especially given that the risk-to-benefit ratio cannot yet be determined, their use is not advised at this time .
For HFS in patients undergoing capecitabine therapy, a positive influence of inhibitors of cyclooxygenase 2 such as celecoxib has been reported .
Preventive measures are central to the management of HFS. Before beginning therapy, patients should be informed of the potential of HFS occurring and should be instructed as to what to do if symptoms appear. Patients should be advised to avoid mechanical stress on the skin such as pressure (heaving lifting/carrying, long walks), friction (tightly-fitting shoes, gloves, or clothing), and heat. The regular use of a moisturizing lotion is recommended and sweat should be rinsed off with lukewarm water. Many drug manufacturers also provide additional patient information. Studies are currently underway to examine the use of urea-based and lactic acid-based topical therapies (http://www.clinicaltrials.gov, NCT 00296036), topical steroids, and dexpanthenol (http://www.clinicaltrials.gov, NCT 00661102) for prevention.
Before beginning therapy, a clinical examination of the hands and feet should be performed including an assessment of predisposing factors such as the presence of hyperkeratosis, eczema, fungal disease, and any malalignment resulting in uneven distribution of pressure. Any apparent risk factors should be treated if necessary (e.g., foot care, dermatological referral, orthopedic treatment of malalignment).
In patients undergoing chemotherapy with drugs such as docetaxel or PLD, which are given as short-term infusions, the frequency and severity of HFS can be significantly reduced by cooling during treatment. Two prospective studies, comparing sides, have been conducted with docetaxel [42, 43]. There is also a prospective study available on PLD with a control group . Numerous retrospective studies  have also been published. The responsible mechanism is believed to be cold-induced vasocon-striction leading to impaired circulation and diminished exposure of the hands and feet due to the chemotherapy agent. A similar mechanism has postulated for the nicotine patch which has been shown in retrospective studies to reduce the frequency and severity of HFS from capecitabine . A prospective study is currently underway in which a nicotine patch is applied daily either at the beginning of capecitabine therapy or after the onset of symptoms of HFS (http://www.clinicaltrials.gov, NCT 00751101). Several retrospective case series with various chemotherapy agents have yielded promising results in patients who were given pyridoxine (vitamin B6, 150–300 mg/daily), who had significantly milder forms HFS compared with controls who were not. Promising results have been reported for pyridoxine in patients taking 5-FU [45, 46], docetaxel  and in a dog model on the use of PLD . These results were not found to apply to capecitabine . A prospective, randomized, placebo-controlled study reported no benefit of 200 mg pyridoxine daily on the development of HFS from capecitabine therapy . Studies are currently examining possible preventive effects of pyridoxine on the development of HFS in prospective, randomized, placebo-controlled studies with capecitabine (http://www.clinicaltrials.gov, NCT 00486213, 00767689 and 00296036, in the latter also in combination with a urea- and lactic acid-based topical therapy) and PLD (http://www.clinicaltrials.gov, NCT 00245050).
The mechanism via which pyridoxine acts is unclear. It is also uncertain whether the generally good results seen with pyridoxine in chemotherapy-associated HFS are applicable to HFS in patients from MKI therapy.
Topical eniluracil is currently being tested in a placebo-controlled, randomized, prospective study in patients receiving capecitabine for the prevention of HFS (http://www.clinicaltrials.gov, NCT 00827580). The drug is thought to reduce the breakdown of capecitabine and thus accumulation of relevant metabolites in the hands and feet.
Retrospective studies on capecitabine have shown that administration of celecoxib can reduce the severity and frequency of HFS . A prospective, randomized study was initiated, but stopped due to lacking recruitment (http://www.clinicaltrials.gov, NCT 0305643). Another option for prevention, which addresses excretion of the chemotherapy agent in sweat, is currently being tested for PLD. This trial is investigating the use of the topical antiperspirant F511 in a prospective, randomized, placebo-controlled study comparing sides (http://www.clinicaltrials.gov, NCT 00992706).
HFS describes a spectrum of diseases with varying clinical appearances and pathogeneses. These disorders all share a common underlying etiology of a toxic effect due to drug treatment of cancer. It appears that two different types of HFS are found at either end of this spectrum, representing two extremes: chemotherapy-associated HFS, which involves severe palmoplantar erythrodysesthesia, and has some overlap with eccrine squamous syringometaplasia, and which appears to involve pathogenetically relevant sweat gland harm. On the other end is MKI-associated HFS, which is characterized by callus-like hyperkeratosis on palmoplantar areas exposed to pressure and in which there are signs of underlying vascular damage.
For any cancer therapy which is associated with HFS, the patient should be thoroughly informed prior to beginning the drug as to what to do in the event that symptoms of HFS appear. In patients with skin disease affecting the hands and feet, dermatological treatment is imperative. Preventive measures such as cooling of the hands and feet during infusion therapy have proven useful for drugs such as docetaxel and pegylated liposomal doxorubicin with a significant reduction in the frequency and severity of HFS. If therapy is continuous, such as in capecitabine or sorafenib, a prompt dose reduction or even briefly interrupting therapy should be considered. Treatment of HFS is based on clinical appearance. There are only small case series available on other therapeutic and preventive measures such as pyridoxine therapy, DMSO- or uridine-based topical therapies.
Conflict of interest
Prof. Dr. med. Ralf Gutzmer has received lecturing fees from Roche, Sanofi Aventis, and Essex Pharma.
Correspondence to Prof. Dr. med. Ralf Gutzmer Klinik und Poliklinik für Dermatologie, Allergologie und Venerologie Hauttumorzentrum Hannover (HTZH) Medizinische Hochschule Hannover Ricklinger Straße 5 D-30449 Hannover Tel.: +49-511-9246-0 Fax: +49-511-9246-284 E-mail: firstname.lastname@example.org