“Rash” was reported in the early clinical trials for both vemurafenib and dabrafenib. While this can take many forms including the classical maculopapular drug-related exanthema,25 in our experience the most common rash induced by BRAFi's is Grover's disease. This occurs in up to 45% of patients on dabrafenib,1 and 39% on vemurafenib.3 It commonly presents on the trunk with the limbs infrequently involved. Treatment varies depending on its severity (Table 1). One group has reported the development of Darier's-like disease26 that on histology looked similar to Grover's disease.

Plantar keratoderma usually occurs at sites of friction and also on the hands (Fig. 1b). As these lesions are quite tender and interfere with patient's quality of life, early treatment is essential.1

Grover's disease and plantar keratoderma can be treated with moisturisers and topical keratolytics (urea or salicylic acid). Oral acitretin has also been reported to be useful in minimising the severity of Grover's disease and plantar keratoderma.19,20,22

Changes in melanocytic naevi have been reported, including with the new BRAFi, LGX818.27–29 These changes include new naevi, regression of existing naevi and hyperpigmentation. There is a debate about the increased frequency of melanoma development in these patients30,31 with figures from 2.5%3 to 21%32 and up to 58%28 of patients treated with BRAFi. The new melanomas have been shown to be wild-type for the BRAF mutation, suggesting that paradoxical activation of the MAPK pathway may also be contributing to their development.27 New or changing naevi are best monitored with serial dermoscopy examinations, and any atypical lesions require excision (Table 1).28

Pruritus is also another commonly reported cutaneous AE. Approximately 13% and 30% of patients experience pruritus of any grade receiving dabrafenib and vemurafenib respectively.22,33 In patients receiving vemurafenib, grade 2 and 3 pruritus appears in 6–7%.8,34 Pruritus could develop secondary to drug-induced xerosis, Darier's or Grover's diseases.6,22,26

General measures such as the use of soap-free body wash and regular application of emollients together with a first generation oral anti-histamines such as diphenhydramine HCL or hydroxyzine HCL control symptoms in most cases.35 It seldom requires dose interruption or discontinuation.5 Topical antipruritic medications (camphor 0.5%, menthol 0.5%, pramoxine hydrochloride 1%; doxepin 5%) can also be used to provide symptomatic reliefs.36 Medium strength potency topical corticosteroids can be applied twice daily. Urea-containing creams can also be used for symptom control. Oral doxepin, gabapentin or a short course of low dose oral corticosteroids at 0.5–1mg/kg for less than a week may be beneficial.6 Additionally, patients should be encouraged to wear loose fitting clothing in a cool ambient environment.6

BRAFi also induces changes of hair follicle. This includes alopecia, changes in the hair structure causing curly/grey hair, folliculitis, keratosis pilaris and cysts.22

Photosensitivity is more commonly seen with vemurafenib. Approximately 52% of patients experience significant reaction to light,22 and 12% have grade 2 or 3 reactions, although with proper photo-testing it seems 92% reacts to light.37 This is induced by ultraviolet A exposure37–39 and requires sun-avoidance and the frequent use of broad spectrum sunscreens (UVA and UVB). There have been some cases of dabrafenib-induced photosensitivity in 0.8–3% of patients, mostly grade 1 or 2 reactions.9,40,41

Although accurate frequency is unknown, both vemurafenib and dabrafenib induced panniculitis have been described in literature,22 with frequency between 2.5% for dabrafenib and 11% for vemurafenib in one paper.3

As we stated before, there are multiple reports of “rash” in patients on vemurafenib and dabrafenib that could be classified in many of the clinical manifestations described above. In our patients, classical maculopapular drug-related exanthemas were only seen in a few patients on dabrafenib (0.8%) and, more frequently, with vemurafenib (11.1%).25

Infrequent but serious cutaneous toxicities including drug reaction with eosinophilia and systemic signs (DRESS),42,43 sweet syndrome44 and toxic epidermal necrolysis (TEN) have also been described.45–48 In one phase 3 clinical trial, grade 3 rash and pruritus were observed in 8% and 1% respectively.8 Two other phase 3 clinical trials reported a few other grade 3 cutaneous AEs besides cuSCC. These include hyperkeratosis, hand-foot syndrome, new primary melanoma, and rash, and the frequencies were less than 1%.49,50

The most frequently observed cutaneous AE of trametinib is acneiform eruption (Fig. 1c). Falchook et al.53 reported 82%; Kim et al.54 reported 75% and Flaherty et al.55 reported 57% of rash/acneiform dermatitis development with MEKi use. In another study, although only a small number of patients (ten) were included,56 they represented 77% of cases of purely acneiform eruptions occurring in patients in trametinib.

Acneiform eruptions usually appear on the face and trunk, predominately where there are more sebaceous glands rather than in sun-exposed areas. Clinical presentations are more inflammatory (erythematous, papular and pustular) than cystic in nature.57

There are a few hypotheses suggesting the mechanisms of acneiform eruptions in patients receiving trametinib. Development of acne is associated with Insulin-like growing factor-1 (IGF-1) inducing sebaceous glands lipogenesis via activation of Phosphoinositide 3-kinase (PI3K)–AKT pathway.58 On the other hand, a similar clinical presentation associated to epidermal growth factor receptor inhibitors (eGFRi), is related to an abrupt blockage of mitogen-activated protein kinase (MAPK) pathway.

Treatments vary according to the presentation and the severity of each particular case. Topical treatments such as clindamycin, together with topical corticosteroids and oral doxycycline, are enough to control these lesions,3 but oral isotretinoin should be considered in more severe cases (Table 1).

Pruritus has been reported in up to 27% of patients either as a separate entity or in combination with drug-induced xerosis.53,59 Paronychia has been reported infrequently.60 MEKi is associated with fewer and milder cutaneous AEs compared with other treatments of advanced melanoma.55

Folliculitis is the most common AE present in patients receiving combined BRAFi and MEKi (40%) (Fig. 1d).3 Managements are based on clinical presentations. As most of them are mild, antiseptic wash (triclosan and chlorhexidine) are usually sufficient. Oral antibiotics can be used in moderately severe cases.22

There is a noticeable reduction in the development of cuSCC in patients receiving combined BRAFi and MEKi compared with BRAFi alone (7% vs. 19% respectively)63 due to the inhibition of the excessive signalling produced by the paradoxical activation of MAPK pathway. Similarly, the reduced frequencies of other cutaneous AEs such as verrucal keratosis, Grover's disease, hyperkeratosis, palmo-plantar keratoderma, alopecia and changes in the hair follicles (grey or curly hair) can be explained by this inhibition.3 Contrastingly, the number of new primary melanomas was similar in both groups (2% vs. 1%, respectively).50 Interestingly, photosensitivity was reported to be more frequent in vemurafenib and cobimetinib group compared to vemurafenib alone group (28% vs. 15%, respectively),61 although the incidence appeared low in the vemurafenib group in this study compared to others.22,37

Grade 3 cutaneous AEs besides cuSCCs described in phase 3 clinical trials were limited to poorly described rash, photosensitivity reaction and one case of alopecia in the vemurafenib and cobimetinib combination,61 and hand-foot syndrome, new primary melanoma in the dabrafenib and trametinib combination.49,50 Management advice is outlined in Table 1. Overall, severe cutaneous AEs are less frequent in a combination group compared to a single agent group.64

Patients receiving ipilimumab commonly develop maculopapular exanthema (Fig. 3a). According to three major studies, 47–68% of patients receiving ipilimumab are expected to develop maculopapular ‘rash’ after 2–4 weeks.5,6,36,71–74 These are usually of mild to moderate severity appearing on the trunk and extremities, which may be pruritic.6,75 In rare cases, this may develop into generalised erythema. A pustular acneiform eruption and lichenoid dermatitis with violaceous eczematous papules and plaques have also been observed.6 Pathologically, epidermal spongiosis, and perivascular lymphocytic infiltrate with predominant eosinophils and CD4+ T cells have been described.67

Figure 3.

Anti-CTLA4 induced (a) maculopapular exanthema, (b) vitiligo, (c) anti-PD1 induced eczema and (d) bullous pemphigoid.

(0.35MB).

Often maculopapular exanthema is managed symptomatically without discontinuation or dose reduction of ipilimumab.36 To control the inflammation, medium potency topical corticosteroid or topical calcineurin inhibitor can be used.6 However, in more severe cases, discontinuation of ipilimumab and administrating a tapering course of oral prednisone 1–2mg/kg daily over a month may be indicated.36 Clinicians should be aware that a rapid tapering of steroid can lead to recurrence and exacerbation of symptoms.35 Interestingly, it is reported that the use of an immunosuppressant such as oral corticosteroid maintains anti-tumour response and selectively down-regulate severe irAEs, suggesting that autoimmune reactions may be a separate entity to anti-tumour activity.35,69 In our practice, low-dose acitretin with topical steroid and emulsifying ointment wet dressing (Table 1) showed marked skin improvement.

Ipilimumab-induced pruritus typically occurs regardless of the presence of maculopapular exanthema or concurrent xerosis. Approximately 31% of the patients receiving ipilimumab have reported pruritus.76 Management has been discussed above.

Approximately 4–11% of patients receiving ipilimumab have reported loss of skin pigmentation (Fig. 3b).36,71,77 Additionally, autoimmune reactions involving follicular melanocytes may also lead to loss of hair colour and localised or generalised alopecia.77 Patients should be educated to be compliant with photo-protection measures as they are more susceptible to sunburn (Table 1).6,35

There have been reported cases of neutrophilic dermatosis such as pyoderma gangrenosum-like ulcerations78 and Sweet's syndrome.75 Aggressive wound care and high-dose oral corticosteroids may be helpful.78 A case of a psoriasiform eruption has also been previously reported with ipilimumab, treated with both topical and oral corticosteroids.6 Additionally, a case of cutaneous radiation sensitivity with blistering and a photosensitivity reaction have been reported.78

The reported cases of TEN and Stevens–Johnson syndrome (SJS) are less than 1%.35 DRESS has also been described.78 In these grade 4 cutaneous AEs, immediate and permanent discontinuation of the treatment, together with initiation of aggressive therapy for severe drug reaction and prompt referrals to specialised units are recommended. Detailed skin care including mucosal surfaces should be performed on regular basis.6,35,36

Patients receiving anti-PD1 antibodies often develop violaceous pruritic papules and plaques, sometimes resembling lichen planus, a few months into the treatment.2,83 We reported that 17% of our patients developed biopsy proven lichenoid reactions, and an estimated one quarter of this population developed this reaction within 8.3 months.2 Other trials have reported approximately 20–29% of patients experience ‘rash’ or a maculopapular eruption during the treatment.81,84,85 Due to the non-specific term of “rash” being used in trials, the true incidence of lichenoid reaction is difficult to ascertain, and some cases of these may have been misclassified.

Lichenoid reactions are predominantly distributed on the body and typically, mucosal surfaces are spared.2 Some studies described acneiform eruptions separate to lichenoid reaction.81 Interestingly, we have observed a few patients with histology proven lichenoid reaction clinically presenting as acneiform eruptions.

These mild to moderately severe lichenoid reactions are best managed with medium potency topical corticosteroid. Oral anti-histamines along with emollients may also be beneficial.81,83,86 In rare cases, we have used systemic prednisone or oral acitretin (Table 1).

Vitiligo is another frequent cutaneous irAEs. An 8–24% of patients have been reported to develop vitiligo during anti-PD1 therapy use.2,82,85 There are two recent studies describing the possible positive association between the development of vitiligo and survival benefit.81,82 However, proper statistical analysis should be performed as these irAE are time-dependent.81,87

Pruritus is one of commonly reported AEs for anti-PD1 therapy use. In one study, 12% of patients developed pruritus.85 Pruritus is usually managed with supportive treatment and good skin care. Management has been discussed above.

There has been a number of reports of vesiculo-bullous reactions, mainly bullous pemphigoid (Fig. 3d),88–90 but also bullous lichenoid reactions91 and reactions described as Steven–Johnson-like with mucosal involvement.92 The presence of severe interface dermatitis is consistent with the lichenoid changes observed in these patients,2 but more interesting is the development of immunoglobulin mediated diseases such as bullous pemphigoid88 suggesting that PD1 could be involved in B-cell biology.

Various cutaneous manifestations have been described with anti-PD1 therapy use including exacerbation of psoriasis,93–95 psoriasiform reactions,96 Sweet's syndrome and alopecia.97 It is expected that new but less frequent side effects will be described in the future with increasing use of anti-PD1 therapy.

Two cases resembling DRESS associated with the use of anti-PD1 therapy prior to vemurafenib treatment have been described with a hypothesis that anti-PD1 therapy may be priming the immune system, predisposing these patients to develop DRESS.98 TEN like reaction with satellite cell necrosis has also been described.7

The authors declare that no experiments were performed on humans or animals for this study.

The authors declare that no patient data without signed consents appear in this article.

The authors declare and guarantee that they are in possession of a document signed by the patients whose personal data is included in the article.