Dermatologic Adverse Events Related to the PI3Kα Inhibitor Alpelisib (BYL719) in Patients with Breast Cancer

Abstract

Purpose:

Rash develops in approximately 50% of patients receiving alpelisib for breast cancer, often requiring dose modifications. Here, we describe the clinicopathologic, laboratory, and management characteristics of alpelisib-related dermatologic adverse events (dAEs).

Methods:

A single center retrospective analysis was conducted. Data were abstracted from electronic medical records.

Results:

A total of 102 patients (mean age 56 years, range 27-83) receiving alpelisib most frequently in combination with endocrine therapy (79, 77.5%) were included. We identified 41 (40.2%) patients with all-grade rash distributed primarily along the trunk (78%) and extremities (70%) that developed approximately within two weeks of treatment initiation (mean 12.8 ± 1.5 days) and lasted one-week (mean duration 7.1 ± 0.8 days). Of 29 patients with documented morphology of alpelisib-related dAEs, 26 (89.7%) had maculopapular rash. Histology showed perivascular and interface lymphocytic dermatitis. All-grade rash correlated with an increase in serum eosinophils from 2.7% to 4.4%, p<0.05, and prophylaxis with non-sedating antihistamines (n=43) was correlated with a reduction of grade 1/2 rash (OR 0.39, p=0.09). Sixteen (84.2%) of 19 patients with grade 3 dAEs resulted in interruption of alpelisib, which were managed with antihistamines, topical and systemic corticosteroids. We did not observe rash recurrence in 12 (75%) patients who were re-challenged.

Conclusions:

A maculopapular rash associated with increased blood eosinophils occurs frequently with alpelisib. While grade 3 rash leads to alpelisib therapy interruption, dermatologic improvement is evident with systemic corticosteroids; and most patients can continue oncologic treatment at a maintained or reduced dose upon re-challenge with alpelisib.

INTRODUCTION

Elevated phosphatidylinositol-3-kinase (PI3K) signaling, which mediates many cellular processes (i.e. angiogenesis, glucose homeostasis), is considered one of the hallmarks of cancer. In breast cancer, PIK3CA mutations encoding the p110α catalytic subunit of PI3Kα lead to overactivation of the PI3K pathway and occur in approximately 40% of patients [1-3]. As such, efforts to develop cytotoxic therapies in breast cancer have focused on agents targeting the PI3K pathway, including the pan-PI3K and PI3Kα-specific inhibitors.

In 2019, based on results of the SOLAR-1 trial [4], alpelisib (BYL719, Novartis) became the first inhibitor of PI3Kα approved by the Food and Drug Administration (FDA) to treat estrogen receptor positive (ER+) breast cancer harboring a PIK3CA mutation. In this randomized phase III trial, it was demonstrated that alpelisib with fulvestrant improves progression-free survival in metastatic hormone receptor positive / human epidermal growth factor receptor 2 negative (HR+/HER2-) breast cancer patients carrying mutant PIK3CA. All-grade rash was reported in 53.9% of patients; and grade 3 rash in 20.1% [4]. Of note, earlier evidence from phase I trials of alpelisib reported dermatologic adverse events (dAEs) in 42.5% of patients in a dose-dependent manner [5,6]; and in another phase I trial which combined alpelisib with letrozole, 45% were reported to have maculopapular rash [7]. In Japan, a phase I trial of alpelisib caused all-grade rash in 48.5% of patients, with grade 3/4 rash in 24.2% [8].

Second to hyperglycemia, rash has been reported as a common cause of discontinuation of alpelisib (3.2%) [4], signifying an obstacle to achieving therapeutic benefit. However, a comprehensive characterization of alpelisib-related dAEs, which may help guide clinicians in the management and counseling of breast cancer patients receiving this therapy, has not yet been reported. To that end, we present the clinicopathologic and laboratory features of alpelisib-related dAEs, as well as outcomes data on dermatologic management and safety of rechallenge with alpelisib following dose interruption.

METHODS

Utilizing an institutional information systems service (Dataline) and the electronic medical record (EMR), we searched for patients that were treated with alpelisib at Memorial Sloan Kettering Cancer Center from June 1, 2013 to July 31, 2019. Four randomized clinical trial protocols studying alpelisib for metastatic breast cancer were identified (NCT01870505, NCT02167854, NCT02734615, NCT03056755). Patients who were treated with alpelisib for solid tumors subsequent to FDA approval were also included in our study (Table 1, Supplemental Fig. 1).

Table 1.

Characteristics of patients and alpelisib-related dAEs by treatment protocol

	NCT01870505	NCT02167854	NCT02734615	NCT03056755	Post-Approval	All
N (%)	51 (50.0)	23 (22.5)	8 (7.8)	11 (10.8)	9 (8.9)	102 (100)
Sex (M/F)	0/51	0/23	0/8	0/11	2/7	2/100
Race, white (%)	38 (74.5)	16 (69.6)	8 (100.0)	8 (72.7)	5 (55.6)	75 (73.5)
Avg Age (SEM)	55.5 (12.1)	53.5 (8.6)	60.0 (13.0)	59.0 (7.7)	65 (4.5)	56.2 (1.1)
Primary Malignancy	HR+ Breast	HER2+ Breast	HR+ Breast	HR+ Breast	HR+ Breasta Prostateb Lungc EMPDd
Drugs	Alpelisib Letrozole/Exemestane	Alpelisib anti-HER3 Trastuzumab	Alpelisib SERDe	Alpelisib Fulvestrant/Letrozole	Alpelisib Fulvestrant
Dosef (range)	300 mg (200-350 mg)	300 mg (250-350 mg)	225 mg (200-300 mg)	300 mg (300-300 mg)	300 mg (200-300 mg)
Any grade rash (%g)	23 (45.1)	8 (34.8)	4 (50.0)	3 (27.3)	3 (33.3)	41 (40.2)
Grade 1/2 rash (%g)	8 (15.7)	8 (34.8)	4 (50.0)	1 (9.1)	1 (11.1)	22 (21.6)
Grade 3 rash (%g)	15 (29.4)	0 (0.0)	0 (0.0)	2 (18.2)	2 (22.2)	19 (18.6)

^aHR+ Breast, n=6

^bProstate, n=1

^cSquamous Cell Carcinoma of Lung, n=1

^dExtramammary Paget’s Disease (EMPD), n=1

^eselective estrogen receptor degrader

^fmedian dose of alpelisib at drug initiation

^g% of patients within each protocol/column

Patient demographics, clinical characteristics, skin biopsy results, and management of dAEs were abstracted from medical oncology and dermatology assessments notes in the EMR. Clinical features included time to onset and duration of dAEs, prophylactic use of antihistamines, rash morphology, associated symptoms, severity, and affected body surface area. Clinical severity, assessed by either the medical oncologist or oncodermatologist, was reported using the Common Terminology Criteria for Adverse Events version 4.0 (CTCAE) toxicity grading scale, wherein grade 1/2 dAEs = mild/moderate and grade 3/4 dAEs = severe. Telephone calls documented in the EMR were used to supplement clinicopathologic characteristics.

Laboratory data of interest included a complete blood count (CBC) with differential (i.e. absolute eosinophil count) and metabolic panel parameters (i.e. glucose) measured at pre-initiation, 1 week and 2 weeks post-alpelisib. To evaluate if dAEs were associated with adverse effects on kidney or liver function, creatinine and alanine aminotransferase (ALT) were analyzed. In addition, blood insulin C-peptide and glycohemoglobin (HbA1C) values obtained approximately two months prior to alpelisib were examined. Management of alpelisib-related rash, as well as the frequency of dose adjustments and/or discontinuation of alpelisib, based on clinical impression and consideration for protocol management guidelines for adverse events, was also noted.

Statistical data analyses were performed using Excel (Redmond, WA) and Graphpad Prism 8 (San Diego, CA); significance was calculated utilizing two-sided tests. Fisher’s exact test was used to assess the odds ratio (OR) of rash occurrence in relation to prophylactic antihistamine use. Paired t-tests were used to track changes in laboratory values over time. All statistics are depicted as mean ± standard error mean (SEM); figure error bars depict SEM. In calculating the number of days to rash onset and duration, outliers were excluded.

RESULTS

Patient Demographics

We identified a total of 102 patients (mean age: 56 years, range 27-83) that were treated under clinical trial protocols (NCT01870505, NCT02167854, NCT02734615, NCT03056755) and post-FDA approval of alpelisib for PIK3CA mutant (HR+) and HER2 amplified metastatic breast cancers (Table 1). Patients received alpelisib in combination with either endocrine therapy (79, 77.5%) or anti-HER2/3 inhibitors (23, 22.5%). The trials that employed endocrine therapy included letrozole, exemestane, fulvestrant, or selective estrogen receptor degraders [SERDs]) for HR+ subtypes. Fifty-one (50%) patients were enrolled in the NCT01870505 phase I trial, which combined alpelisib with letrozole or exemestane. Daily doses of alpelisib ranged between 200 and 350 mg. Baseline characteristics of individuals in each study are shown in Table 1.

Clinical features of alpelisib-related dermatologic adverse events

Forty-one (40.2%) patients developed all-grade rash (n=102), of which nearly half (19, 46.3%, [18.6% of all patients]) had a severe (grade 3) presentation (Table 1). No grade 4 rashes were observed, and no more than one dAE was recorded for each patient. Forty-three (42%) patients received prophylactic treatment with antihistamines. The majority (37, 86%) were prescribed daily cetirizine 10 mg; others received loratadine 10 mg (3, 7%), cetirizine 5 mg (1, 2%), and hydroxyzine 25 mg (2, 5%). With prophylactic antihistamines, there was a trend towards reduction of grade 1/2 rash (OR 0.39, p=0.09); and although not statistically significant, there was also a protective effect on the incidence of grade 3 dAEs with the use of antihistamines prior to alpelisib (OR 0.60, p=0.43) (Table 2).

Table 2.

Alpelisib-related rash in patients receiving prophylactic antihistamines

	No prophylaxis	Prophylaxis	Odds Ratio	p-value
No rash	31	30
Grade 1/2 rash	16	6	0.39	0.09
Grade 3 rash	12	7	0.60	0.43

Of 29 patients with alpelisib-related dAEs and documented morphology, the majority (26, 89.7%) had maculopapular rash. The remaining 10.3% (n=3) exhibited an acneiform phenotype (Table 3). No patients developed hand-foot syndrome, bullous pemphigoid eruptions, alopecia, lichenoid rash, or psoriasiform dAEs. Notably, all patients with acneiform rash were part of the NCT02167854 trial for HER2 amplified breast cancer, which combined alpelisib with trastuzumab [9] and a novel anti-HER3 monoclonal antibody (mAb). In contrast, all maculopapular rashes were documented in patients who received alpelisib combined with hormone therapy were (Fig. 1; Table 3).

Table 3.

Clinical and management characteristics of alpelisib-related dAEs

	Any grade rash (%a)	Grade 1/2 (%a ,%b)	Grade 3 (%a ,%c)
Morphology (n=29)
Maculopapular	26 (90)	15 (52, 83)	11 (38, 100)
Acneiform	3 (10)	3 (10, 17)	0 (0, 0)
Symptoms (n=19)
Pruritic	7 (36)	2 (11, 25)	5 (26, 46)
Burning	6 (32)	2 (11, 25)	4 (21, 36)
Asymptomatic	6 (32)	4 (21, 50)	2 (11, 18)
Management (n=41)
No treatment	4 (10)	4 (10, 18)	0 (0, 0)
Topical corticosteroids	2 (5)	1 (2, 5)	1 (2, 5)
Oral antihistamine	4 (10)	4 (10, 18)	0 (0, 0)
Oral antihistamine	13 (31)	10 (24, 45)	3 (8, 16)
Topical corticosteroids
Oral antihistamine	4 (10)	1 (2, 5)	3 (8, 16)
Oral corticosteroids
Oral antihistamine	14 (34)	2 (5, 9)	12 (29, 63)
Topical corticosteroids
Oral steroids

^aPercentages derived from fraction reported subset of any grade rash (i.e. n=29 for rash morphology).

^bPercentages derived from fraction of reported subset within same grade rash (i.e. n=18 for grade 1/2 morphology).

^cPercentages derived from fraction of reported subset within same grade rash (i.e. n=11 for grade 3 morphology).

Fig. 1.

Example of grade 2 maculopapular rash on the chest (a) and back (b) of a patient 12 days after initiating 300 mg of alpelisib for advanced breast cancer. Patient was prescribed 10 mg prednisone daily with taper, plus topical clobetasol twice daily. There was no interruption in alpelisib, and rash resolved over 6 days.

In 23 patients whose rash was localized to a specific body area, the trunk (i.e. chest, abdomen, back) (18, 78%) and extremities (16, 70%) were most frequently cited (Fig. 2A). In contrast, only two (9%) patients had rash involving the face, and one (4%) on the scalp. Three (13%) patients developed erythroderma (rash covering > 80% body surface area).

Fig. 2.

(a) Distribution and frequency of alpelisib-related rash (n=23). (b) Time to rash onset for all-grade rashes was 12.8 ± 1.5 days post-initiation of alpelisib (n=38). Grade 1/2 rashes occurred on average within 14.8 ± 2.9 days (n=19), the mean number of days to grade 3 rash was 10.8 ± 0.8 (n=19). (c). Rash duration of all-grade rashes was 7.1 ± 0.8 days (n=20). Grade 1/2 and grade 3 dAEs were present for an average of 7.0 ± 1.0 (n=8) and 7.2 ± 1.3 (n=12) days, respectively. Error bars represent SEM; differences were not statistically significant.

Thirteen (68%) of 19 patients with any-grade rash and documentation of any associated symptoms had burning pain (6, 32%) or pruritus (7, 36%); the rest (6, 32%) were asymptomatic. A greater proportion of symptomatic patients were found in the group with more severe (grade 3) rash (9, 81.8%) than among those with mild to moderately severe (grade 1/2) dAEs (4, 50%) (Table 3).

On average, time to rash onset was 12.8 ± 1.5 days post-initiation of alpelisib (n=38). Whereas grade 1/2 rashes (n=19) occurred on average within 14.8 ± 2.9 days, the mean number of days to grade 3 rash (n=19) was 10.8 ± 0.8 (Fig. 2B). Examination of rash duration shows that grade 1/2 and grade 3 dAEs were present for an average of 7.0 ± 1.0 (n=8) and 7.2 ± 1.3 (n=12) days, respectively (Fig. 2C).

Histopathologic and laboratory findings

Skin biopsies were obtained from two patients treated with alpelisib 300 mg (NCT01870505) (Fig. 1a,b). Hematoxylin and eosin staining showed mild interface and superficial dermal perivascular lymphocytic dermatitis; folliculocentric spongiotic, perivascular, and focal interface dermatitis. Dermatopathological interpretation was consistent with a hypersensitivity reaction.

Available laboratory data was analyzed for patients with both baseline and subsequent blood draws at 1 week and 2 weeks after starting alpelisib. Those who did not have a complete set of labs at these three time points were excluded. CBC analysis showed no changes in lymphocytes, neutrophils or monocytes between patients who developed rash (n=9) and those who did not (n=18) (Fig. 3a-c). Patients who developed rash had an increase in blood eosinophils after receiving alpelisib for two weeks, from 2.7% to 4.4% (p<0.05) (Fig. 3d). Absolute eosinophil count was not statistically significantly elevated for patients, regardless of whether they developed rash (Fig. Supplemental Fig. 2a). A significantly acute elevation in creatinine was observed in patients receiving alpelisib, regardless of rash development (p<0.05). No statistically significant differences were seen in ALT levels, although patients developing rash demonstrated a trend towards elevated ALT after receiving alpelisib. (Supplemental Fig. 2b,c).

Fig. 3.

Complete blood count (CBC) from patients at baseline and after 1 week or 2 weeks of receiving alpelisib, with comparison between patients developing rash (n=9) and patients without rash (n=18). Analysis of changes in lymphocytes (a), neutrophils (b), monocytes (c) and eosinophils (d). Error bars represent SEM. Significance was calculated using paired t-test between baseline and 1-week or 2-week lab values. * p<0.05

Patients’ serum glucose was assessed for hyperglycemia due to the concern for developing alpelisib-related insulin resistance. Serum glucose was higher in patients without dAEs, although this was not statistically significant (Supplemental Fig. 2d). No differences were observed when comparing baseline HbA1C and C-peptide levels in patients with or without dAEs (Supplemental Fig. 2e,f).

Management of alpelisib-related rash

Alpelisib-related dAEs (n=41) were managed according to their grade of severity (Table 3). Among 22 patients with grade 1/2 rash, nearly half (10, 45.5%) were managed with oral antihistamines combined with topical corticosteroids; four (18.2%) received oral antihistamines as monotherapy, and another four (18.2%) did not receive any dermatologic treatment.

Grade 3 dAEs (n=19) were most often managed with a combination of oral antihistamines, topical and systemic steroids (12, 63.2%). Other treatment modalities in this group included a combination of oral antihistamines with topical corticosteroids (3, 15.8%) and administration of systemic steroids as monotherapy (3, 15.8%). Non-sedating oral antihistamines such as cetirizine and loratadine were prescribed; and less frequently, first-generation hydroxyzine or diphenhydramine were recommended when patients’ itch disrupted their sleep. Of the topical corticosteroids, clobetasol was the most frequently recommended, with some instances of triamcinolone, fluocinonide, and desoximetasone. Systemic steroids, such as prednisone, were also administered, recommended at a low dose of 10 mg, three times a day. Exceptions occurred during more aggressive management, such as after rash recurrence in one patient; this patient took 20 mg of prednisone three times a day.

Alpelisib dose modification patterns and re-challenge

Following the onset of rash due to alpelisib, patients may have had the following dose modifications (Fig. 4a): 1) no modifications; maintain alpelisib dose, 2) permanent discontinuation of alpelisib, 3) dose interruption followed by re-initiation of alpelisib at the same or lower dose, 4) dose interruption followed by permanent discontinuation after recurrence of rash with alpelisib.

Fig. 4.

Alpelisib dose modifications related to dAEs (a). Algorithm for management of alpelisib-related dAEs (b). qd: daily; bid: twice a day; tid: three times a day; qhs: at bedtime

Of 22 patients developing grade 1/2 rash from alpelisib, 17 (77.3%) were managed effectively without dose modifications. Four (18.2%) had interruption of therapy followed by rechallenge and no rash recurrence. Among these four patients, one maintained the initial alpelisib dose (300 mg) and three received a reduced dose. One patient (4.6%) had grade 1 rash recurrence and grade 3 mucositis upon re-initiation of alpelisib, leading to permanent discontinuation of alpelisib therapy, which had been administered in combination with trastuzumab and anti-HER3 agents.

Sixteen (84.2%) patients with grade 3 dAEs (n=19) had interruption of alpelisib, followed by re-challenge once the severity of the rash was reduced to grade ≤ 1. We did not observe rash recurrence in 12 (75%) of these 16 patients who re-initiated therapy; the majority (9, 56.3%) were re-challenged without a dose reduction. One (5.3%) of the 19 patients with grade 3 rash was able to continue alpelisib at the initial protocol dose without interruption, and two (10.5%) patients discontinued the drug.

Four (25%) of the 16 patients with anticancer therapy interruption required permanent discontinuation due to rash recurrence within 24 hours of re-initiating alpelisib. One patient developed a painful rash despite dose reduction of alpelisib from 300 mg to 250 mg. While taking oral antihistamines, topical and systemic corticosteroids for management of rash, a second re-initiation trial in this patient at 250 mg failed due to hyperglycemia. Another patient that was re-challenged with alpelisib at a reduced 300 mg dose (from the initial 350 mg), developed rash with burning pain within a few hours of receiving the drug, prompting permanent discontinuation. The third patient re-initiated alpelisib at the same 300 mg dose specified by protocol and developed a burning rash with concomitant fever to 101.7 F, nausea, and headache— also resulting in permanent discontinuation.

The fourth patient was evaluated at the urgent care clinic for recurrence of rash shortly after restarting alpelisib at a reduced 250 mg (decreased from 300 mg). She presented with several new laboratory abnormalities, including eosinophilia at 6.6%, an absolute eosinophil count of 0.3 ×10³/uL and thrombocytopenia (platelets 146 ×10³/mcL), suggesting a drug reaction with eosinophilia and systemic symptoms (DRESS) syndrome (Supplemental Fig. 3) [10]. None of the other three patients developed eosinophilia, defined by over 6% of leukocytes or absolute eosinophil count over 0.7 ×10³/uL. There was no evidence of Stevens Johnson syndrome / toxic epidermal necrolysis (SJS/TEN), or signs of desquamation or skin necrosis in the four patients with rash recurrence.

DISCUSSION

Our study reveals a high frequency (40.2%) of dAEs in patients receiving alpelisib for breast cancer, which remains consistent with results of previous clinical trials wherein dAEs occur in 42.5-53.9% of patients [4-7]. We also found a uniformly maculopapular morphology in the combined alpelisib plus endocrine therapy rash phenotype. This is in contrast to the acneiform type of rash observed with other targeted therapies and immunotherapies, such as in patients receiving EGFR, HER2, and MEK inhibitors [9,11,12].

Overall, alpelisib-related dAEs spare the head and are asymptomatic in 50% of mild-to-moderate presentations, which may be less likely to impact activities of daily living negatively. Nonetheless, patients should be counseled as far as the extent of body surface area involved (i.e. torso, arms, legs) and the typical time to onset of rash (i.e. within two weeks of treatment). Of note, in cases of grade 3 toxicity, there is a trend towards earlier presentation, within 10 days of drug exposure, which correlates with increased incidence of burning pain and pruritic symptoms. Patients should have a lower threshold for reaching out to their provider if they experience rash extending beyond the trunk that is pruritic and/or painful.

In our analysis, there was only one patient with suspected DRESS syndrome, and none that developed SJS/TEN. Therefore, we find that the development of life-threatening dAEs related to alpelisib is rare. Importantly, nearly all (20, 90.9%) patients with grade 1/2 rash were able to continue alpelisib at the standard dose, and the majority (9, 56.3%) of patients who had interruption of oncologic therapy for grade 3 events (n=16) were re-challenged without a dose reduction.

Importance of PI3K signaling in homeostasis of skin repair and inflammation

Maculopapular rashes related to alpelisib are likely an on-target effect of PI3K inhibition. While alpelisib has high selectivity to inhibit the p110α subunit of PI3Kα, small molecule inhibitors with selectivity against various combinations of PI3K isoforms, such as buparlisib, have also demonstrated frequent dAEs with similar maculopapular morphology [13]. Our most extensive in-human experience is with the pan-PI3K inhibitor buparlisib (BKM120, Novartis). Despite the neurologic adverse events that preempted FDA-approval, trials with buparlisib illustrated many safety issues and was a preview for adverse events of other candidate inhibitors of the PI3K pathway. In a phase I study for buparlisib, a primarily pruritic maculopapular rash concentrated on the torso was observed. For grade 3/4 rash occurring in two or more patients, the incidence was 11%; and for all-grade rash, the incidence was 43% [14]. In the BELLE-2 phase III study, there was a 32% incidence of rash, with 8% grade 3/4 rash [15]. Experience with idelalisib, copanlisib, and duvelisib continues to report significant dAEs. Idelalisib, a PI3Kδ specific inhibitor, reported a 13% incidence of any grade rash, with 2-3.6% incidence of grade 3 rash [16,17]. With copanlisib, which selectively inhibits PI3Kα/δ, maculopapular rash of any grade was noted in 13- 16% of all patients with grade 3 rashes occurring in 7-13% of patients [18,19]. In patients receiving duvelisib, which targets p110γ/δ, maculopapular rash was observed in 16.2-18.6% or patients, and grade 3 rash was seen in < 5% of patients [20-23]. It is unlikely that the diverse chemical structures represented in these PI3K inhibitors have similar non-specific activities that cause maculopapular rash.

It is unclear how inhibition of PI3K signaling causes maculopapular rash. As the PI3Kα isoform is expressed at low levels in immune cells [24], it is unlikely that direct inhibition of PI3Kα signaling in this environment is the primary event. In addition, it is unlikely that off-target inhibition of PI3Kδ by alpelisib is the primary event, as there are comparably fewer maculopapular rashes associated with use of δ-targeting molecules—idelalisib, copanlisib, and duvelisib.

Alpelisib may shift recruitment and maturation of different immune cell types as a direct result of disrupted PI3Kα signaling in the skin or an indirect result of disrupted PI3Kα signaling in other tissues, causing systemic exposure to compensatory molecules (i.e. hyperinsulinemia). Upon inhibition of PI3Kα, metabolic tissues have impaired uptake of serum glucose, leading to hyperglycemia and a compensatory increase in insulin [4,25]. However, we found that hyperglycemia is not correlated with alpelisib-related dAEs. Insulin resistance at baseline, as reflected by HbA1C and C-peptide levels, was also not correlated with alpelisib-related dAEs. It is possible that patients that are euglycemic have appropriate hyperinsulinemia which activates certain immune cells. As more patients receive alpelisib, it will be important to trend C-peptide and glucose levels.

Inhibition of PI3Kα signaling may cause intrinsic structural changes to the epidermis and dermis. Activation of PI3K in epithelium does not simply cause tumorigenesis [26]. Rather, the PI3K pathway coordinates keratinocyte fate. PI3Kα activation has been shown to suppress epidermal progenitor cell self-renewal [27], with activation of Akt, the immediate downstream effector, resulting in increased keratinocyte differentiation, while inhibition of PI3Kα induces apoptosis [28]. Also, since estrogen has been demonstrated to have anti-aging and anti-inflammatory effects in skin, there is consideration for the effects of estrogen signaling and crosstalk with the PI3K pathway [29], which may explain the emergence of the maculopapular rash phenotype observed in patients with alpelisib-related dAEs.

Our study additionally suggests that histamine-producing cells and eosinophils have some role in alpelisib-related rash. Based on our data, all patients receiving alpelisib had a trend towards increased eosinophils, with a more pronounced effect in patients with rash. Interestingly, pre-treatment absolute eosinophil count may be lower in patients who develop rash (Supplemental Fig. 2A). We also found that patients developing rash had a statistically significant increase in eosinophils after 2 weeks of receiving alpelisib (p<0.05). Future experiments to monitor serum cytokines and chemokines will help shed light on how alpelisib affects inflammation in the skin.

As additional PI3Kα inhibitors are validated and tested in patients [30,31], we will monitor for dAEs. Antihistamines block H₁- receptors, which are ubiquitously expressed and have a multitude of functions [32]. Histamine stimulation of eosinophils through H₁- receptors increases cell recruitment [33]. We plan to continue investigating alpelisib-related changes in eosinophils to ascertain if there is a clinically significant increase in eosinophils.

Recommendations for management of alpelisib-related rash

Our data showed a trend towards decreased incidence of grade 1/2 (OR 0.39) and grade 3 rash (OR 0.60) with prophylactic antihistamines exposure. While the difference was not statistically significant, our cohort size may have been underpowered. Notably, the manufacturer’s note describes a trend towards reduction of all-grade rash with prophylactic anti-inflammatories during the SOLAR-1 trial, from 54% to 27%. It is difficult to capture how frequently patients take over-the-counter antihistamines for seasonal allergies, independent of their oncologist’s recommendation. Thus, to prevent alpelisib-related rash, we recommend prophylactic treatment with non-sedating antihistamines during the first eight weeks of therapy.

For management of grade 1/2 rashes, we recommend maintaining alpelisib dose and treating the rash with a combination of non-sedating (i.e. cetirizine in the morning) and sedating (i.e. hydroxyzine in the evening) antihistamines with application of high-potency topical corticosteroids twice a day. For management of grade 3 rashes, we additionally recommend interrupting alpelisib and treating with systemic steroids (0.5mg/kg/day of prednisone or equivalent) (Fig. 4b). When the rash is grade ≤ 1, a graded rechallenge with alpelisib while maintaining patient on antihistamines and tapering systemic steroids (0.5mg/kg/day of prednisone or equivalent), plus close monitoring of blood glucose is advised. A graded rechallenge, with a starting with dose of 50mg a day on 1 week, 100mg a day on week 2, 150mg a day on week 3, 200mg a day on week 4, and 250mg a day on week 5, is recommended based on anecdotal evidence. Patients should be instructed to call with symptoms of rash recurrence, especially during the initial 24-hour window after re-challenge with alpelisib. In the event of an acneiform rash, it is important to consider other causative agents and their respective management strategies (i.e. steroid acne, trastuzumab [9] [as observed in one patient analyzed here]).

In summary, maculopapular rash associated with increased blood eosinophils is frequent with alpelisib. Grade 3 rash improves with alpelisib dose interruption and systemic steroids; and the majority of patients are able to continue to derive benefit from this agent at a maintained or reduced dose. However, additional research to identify the mechanisms via which inhibition of PI3Kα alters immune cell signaling and results in clinical manifestations of dAEs is warranted.

ABBREVIATIONS

ALT

alanine aminotransferase

BYL719

alpelisib

CBC

complete blood count

CTCAE

Common Terminology Criteria for Adverse Events

dAEs

dermatologic adverse events

DRESS

drug reaction with eosinophilia and systemic symptoms

EGFR

epidermal growth factor receptor

EMR

electronic medical record

HbA1C

glycohemoglobin

HER2

human epidermal growth factor receptor-2

HER3

human epidermal growth factor receptor-3

HR

hormone receptor

mAb

monoclonal antibody

MEK

mitogen activated protein kinase enzyme

OR

odds ratio

PI3K

phosphatidylinositol-3-kinase

PIK3CA

phosphatidylinositol-4,5-phosphate-3-kinase alpha catalytic subunit

PTEN

phosphatase and tensin homolog

RTK

receptor tyrosine kinase

SEM

standard error mean

SERD

selective estrogen receptor degrader

SERM

selective estrogen receptor modulator

SJS/TEN

Stevens-Johnson syndrome/toxic epidermal necrolysis