Prospective, observational practice survey of applied skin care and management of cetuximab-related skin reactions: PROSKIN study

Cetuximab is a chimeric monoclonal antibody that binds and inactivates the epidermal growth factor receptor (EGFR). The mechanism of blocking EGFR is an important strategy in the treatment of malignancies of epithelial origin such as colorectal cancer and squamous cell cancer of the head and neck [1, 2]. While this therapeutic approach is usually better tolerated than conventional chemotherapy, it has a unique side-effect profile related to the mechanism of action. EGFR is not solely expressed on tumor cells, but also on cells of the epidermis. Thus, skin toxicities, including rash, pruritus, dry skin, desquamation, hypertrichosis, and nail disorders, are seen in approximately 80% of patients treated with cetuximab [3‐5]. Unless properly managed, these can result in dose reductions and discontinuation of treatment, in about 15–25% of patients [6‐8]. The most common (90% of patients with cutaneous toxicity) and clinically most relevant skin toxicity associated with cetuximab is the papulopustular rash, also called acneiform rash [9]. Dermatologic toxicities are rarely life threatening; however, they impair quality of life (QoL) and treatment compliance. Therefore, an effective management of skin toxicities is crucial to maximize treatment efficacy and maintain QoL.

The cetuximab prescribing information states that skin reactions are very common and may require treatment interruption or discontinuation [10]. It recommends that according to clinical practice guidelines, prophylactic use of oral tetracyclines (6–8 weeks) and topical application of 1% hydrocortisone cream with moisturizer should be considered. Medium- to high-potency topical corticosteroids or oral tetracyclines have been used for the treatment of skin reactions [10, 11].

In daily clinical practice, multiple care and management options to reduce the severity of skin reactions are used, including but not limited to topical or oral antibiotics, or glucocorticosteroids. Antihistamines and local anesthetics can be administered to reduce pruritus. The frequent use of cetuximab in cancer therapy and the lack of specific clinical studies necessitate the need for studies evaluating the measures taken to alleviate skin reactions in patients treated with cetuximab.

The current observational study was initiated to gain information about the perceived effectiveness of the measures taken to alleviate skin reactions in cetuximab-treated patients. Furthermore, we aimed at assessing the impact of skin reactions on the treatment course, QoL and the reason of physicians’ choice for specific therapies.

The optimal approach on how to prevent and manage the cutaneous side effects of cetuximab and other EGFR-targeting antibodies has not been clearly established. The prospective observational PROSKIN study provides detailed insights into current preventive measures and treatment of skin reactions related to cetuximab in Switzerland. The study included patients from 23 sites in the country including university hospitals, secondary- and primary-level hospitals and oncological practices, therefore reflecting the full spectrum of oncological patient care in the country. The study demonstrates a broad variety of preventive and therapeutic skin care that is used in patients undergoing cetuximab therapy. Each patient receives on average three products at the same time. The most frequently used products in this study were moisturizing, lipid-regenerating and urea-containing products (i.e., no typical pharmaceutical products), followed by systemic antibiotics and vitamin K1 cream. Thus, physicians mainly used low-cost topical products.

Moderate to very strong effectiveness was perceived by a majority of physicians for patients who received systemic antibiotics or vitamin K1 cream, while moisturizing, lipid-regenerating and urea-containing products were perceived as less effective. An interesting finding was that a specific pattern in effectiveness was observed for all product classes over the whole time of the study: either the product showed no effect at all or a moderate/strong effect. No product class appeared to be substantially superior to the others on an individual patient level. We therefore conclude that the best treatment option needs to be explored individually.

Only one in five patients received systemic antibiotics as primary prophylaxis from the beginning, so this drug class is often reserved for treatment-associated adverse events or as secondary prophylaxis. As expected, the frequency of acneiform rash grade ≥ 2 increased during the treatment course, from 12.6% at week 2 to 21.7% at week 16. These numbers are reassuring in the sense that the majority of patients did not develop severe skin reactions. Nevertheless, the proportion of patients whose skin reactions did impact on daily life steadily increased during treatment, reaching 76.7% at week 16. In line with this observation, the reactive use of all products tended to increase during the treatment course. These findings underline the importance of assessing not only severe toxicities, as also mild or moderate toxicities might affect QoL and patients’ daily life. The toxicity rates confirm published data from prospective clinical trials investigating cetuximab in patients with mCRC and r/m SCCHN.

The results of our study need to be discussed in the context of prior studies investigating different approaches of skin toxicity management in patients treated with cetuximab. In a US-American monocenter randomized controlled trial on 48 patients, prophylaxis with oral minocycline decreased the severity of acneiform rash during the first month of cetuximab treatment, while topical tazarotene was associated with significant irritation [12]. In the same institution, a prospective randomized trial of topical pimecrolimus for cetuximab-associated acne-like eruption in 24 patients failed to show clinically meaningful benefit [13]. Conversely, in a case series of 20 patients treated with different epidermal growth factor receptor inhibitors in Greece, pimecrolimus cream 1% (substituted by metronidazole 1% cream) was effective in most patients (> 50% reduction of erythema, pustules and pruritus) [14]. Minocycline was used in seven patients in the PROSKIN study, none was treated with topical tazarotene or pimecrolimus. In a study with historical controls as comparator, in 40 patients, cetuximab prophylaxis with topical vitamin K1 cream did not translate into clinically meaningful benefit in terms of reducing acneiform rash [15]. However, an Italian monocenter study of 41 patients with metastatic colorectal cancer treated with cetuximab suggested a possible benefit of topical vitamin K1-based cream as prophylaxis for skin rash [16]. In our PROSKIN study, most physicians perceived a moderate effect of topical vitamin K1 cream.

Ocvirk et al. reported on a group of 31 patients with acne-like rash, who, after the first documented cutaneous toxicity, received topical use of emollients. Patients with grade 2 rash received emollients and topical antibiotics. Patients developing grade 3 rash discontinued therapy with cetuximab until recovery and were treated with emollients, topical and systemic antibiotics. Of 31 patients, six had grade 3 rash, 16 patients grade 2 and nine patients grade 1 acne-like rash [17]. As in our study, no grade 4 skin reactions were observed.

A multicenter observational study in 55 patients reported delayed occurrence and milder course of skin reactions with the use of a pre-defined prophylactic skin care regimen including vitamin K1 ointment and oral doxycycline [17].

Overall, similar to our findings, results from previous studies show conflicting results and could so far not establish an evidence-based prophylaxis or treatment regimen for cetuximab-related skin toxicities. Nevertheless, the topic was discussed at consensus meetings and recommendations have been published [18, 19]. Another relevant factor is the low rate of consensus between oncologists and dermatologists in labeling and grading of skin reactions and the frequent interobserver variability [20].

A number of limitations and strengths of the study design need to be considered when discussing the results of this study. The study was small, had no control arm and was unblinded and thus should be considered as real-world observational study with focus on skin products used and their perceived effectiveness over time. In many subgroups, the number of patients was small and therefore results have to be carefully interpreted. There was a high rate of patients that were lost to follow-up. All patients came from Switzerland which had advantages in terms of the access of patients to products and uniform procedures, but might limit the generalizability of the findings to patients in other health-care settings and countries. It was the clinical decision of the respective treating physicians to assign selected patients to certain therapies and not to other available treatment options that potentially may introduce allocation or channeling bias and confound the association between treatment and outcomes [21]. However, this is an advantage as physicians already have some experience with the regimen they use. With a pre-defined regimen, there would be a learning curve at the beginning and possible skepticism that could influence how physicians perceive the effectiveness. Physicians and patients willing to participate in non-interventional studies such as ours may be particularly motivated or interested in science and therefore also be patient to selection bias. Finally, the follow-up period of 16 weeks was relatively short. However, it has been shown that skin toxicities usually develop within the first few weeks of cetuximab therapy. Therefore, we defined this time window. Moreover, at week 16 a large proportion of patients show tumor progression and these patients have to be discontinued from the study introducing another potential selection bias.

In conclusion, in Switzerland oncologists used a great variety of low-cost general skin care products for the prevention or treatment of cetuximab-related skin toxicity in patients with mCRC or r/m SCCHN. They gave preference to systemic antibiotics and vitamin K1 cream. Patients perceived overall “moderate” efficacy of the various measures. For most patients, skin reactions did not influence their willingness to continue cetuximab therapy.