A Phase II Study of MK-2206 in Patients with Incurable Adenoid Cystic Carcinoma (Alliance A091104)

Abstract

Background:

Recurrent/metastatic adenoid cystic carcinoma (ACC) is a rare, incurable disease. MYB is a putative oncogenic driver in ACC that is often overexpressed via a MYB-NFIB rearrangement. We hypothesized that AKT inhibition with the allosteric inhibitor MK-2206 (MSD) can decrease MYB and induce tumor regressions in patients with incurable ACC.

Methods:

Patients with progressive, incurable ACC were enrolled. MK-2206 150 mg weekly was given; escalation to 200 mg was allowed. The primary endpoint was confirmed response. Secondary endpoints were progression-free survival (PFS), overall survival (OS), and safety. An exploratory analysis evaluating MK-2206 impact on MYB expression was conducted in a subset of patients.

Results:

Sixteen patients were enrolled; 14 patients were evaluable for efficacy. No confirmed responses were observed. Thirteen patients had stable disease and one had disease progression as best response. Median PFS was 9.7 months (95% CI: 3.8–11.8) and median OS 18.0 months (95% CI: 11.8–29.9). Nine of 16 (56%) patients had at least one grade 3 treatment-related adverse event with the most common being rash (38%), fatigue (19%), decreased lymphocyte count (13%), and hyperglycemia (13%). Twelve of 14 tumors (86%) tumors had detectable MYB by immunohistochemistry; 7/14 (50%) had a MYB-NFIB gene rearrangement. Serial biopsies revealed decreased MYB levels with MK-2206 in 4 of 5 cases.

Conclusions:

MK-2206 failed to induce clinical responses in patients with incurable ACC. AKT inhibition may diminish MYB protein levels, though the effect was highly variable among patients. Novel approaches to target MYB in ACC are needed.

Précis:

This phase II clinical trial demonstrated that the AKT inhibitor MK-2206 failed to induce clinical responses in patients with recurrent/metastatic adenoid cystic carcinoma. Tumor tissue analyses suggest that MK-2206 treatment may have diminished intratumoral MYB protein levels.

INTRODUCTION

Adenoid cystic carcinoma (ACC) is a rare cancer typically arising from salivary glands. Nearly 40% of patients develop distant metastasis with disease-free intervals ranging from 1 month to 19 years^{1, 2}. Recurrent/metastatic (R/M) ACC is an incurable disease for which there are no standard treatments. Despite recent phase II trials demonstrating activity with multi-targeted kinase inhibitors^3–5, there still is an urgent need to develop effective, biologically rational therapies.

In 2009, Persson et al. was the first to report the discovery of a t(6;9) translocation in ACC tumors⁶. The translocation results in a fusion between the N-terminal domains of the oncogenic transcription factor gene MYB (6q23.3) with the C-terminal end of the transcription factor NFIB (9p22–23). Drier et al. discovered that this translocation event repositions super-enhancers to interact with the MYB promotor, which can be bound by MYB itself to drive a feed-forward mechanism by which MYB is overexpressed⁷. Studies have also shown that MYB can be expressed in fusion-negative ACCs^8–10, suggesting the existence of alternative mechanisms of MYB overexpression. MYB is a proto-oncogene that was first discovered as the cellular homologue of the v-myb transforming oncogene co-opted by the Avian myeloblastosis virus and E26 leukemia virus¹¹. MYB has been implicated in several leukemia subtypes and solid tumors, including breast and gastrointestinal cancers¹¹. A subset of t(6;9) translocation-negative ACCs possess an t(8;9) rearrangement which results in the overexpression of a closely related MYB gene family member, MYBL1^{12, 13}. MYB and MYBL1 share similar DNA binding domains and target some of the same promoters, producing comparable gene expression profiles in ACC tumors¹².

The incidence of MYB rearrangement and overexpression in ACC varies in multiple reports¹⁴. A recent multi-institutional study of approximately 400 ACC tumors observed MYB rearrangements in 39.1% of tumors, MYBL1 rearrangements in 7.1%, and MYB expression by immunohistochemistry (IHC) in 93.2%¹⁵. Hence, MYB overexpression with or without a chromosomal rearrangement represents the most common, recurrent oncogenic event in ACC, nominating it as a therapeutic target of interest for this disease¹⁶. As a transcription factor without targetable enzymatic activity, abrogating MYB requires a strategy that decreases MYB protein levels, disrupts MYB transactivation, and/or disrupts the activation of its downstream gene targets. In preclinical models, depletion of MYB via RNA interference¹⁷ or drugs that suppress MYB protein levels and/or activity^{7, 17–20} can block ACC tumor growth, suggesting that MYB targeted strategies could have therapeutic utility against ACC.

In hematopoietic cells, phosphatidylinositol 3-kinase/protein kinase B (PI3K/AKT) pathway activation has been shown to be required for maintaining MYB stability²¹. PI3K or AKT inhibition can downregulate full-length MYB or MYB-NFIB protein expression in preclinical models^{18, 20–22}. Receptor tyrosine kinases (RTKs) can activate AKT signaling in ACC, and preclinical experiments have shown that the insulin-like growth factor receptor-1 (IGFR-1) specifically can regulate MYB-NFIB expression in an AKT-dependent manner¹⁸. MYB-NFIB positive ACC cells exposed to an AKT or PI3K drug inhibitor resulted in decreased MYB levels^{18, 20}. These data suggest the hypothesis that PI3K or AKT inhibition can be applied to clinically target MYB in ACC. MK-2206 (MSD) was the first allosteric AKT inhibitor to enter clinical development with equal potency against the AKT1 and AKT2, and about 5-fold less potency against AKT3²³. We conducted a phase II trial to evaluate the efficacy of MK-2206 in patients with progressive, incurable ACC, and to analyze the impact of MK-2206 on MYB protein levels in ACC tumor biopsies obtained from trial patients. The first-in-human phase I trial of MK-2206 established 60 mg every other day (QOD) orally as the maximum tolerated dose (MTD)²⁴. The long MK-2206 half-life (60–80 hours) suggested the possibility that weekly dosing could permit administration of higher doses to elicit more potent target inhibition, while preserving manageable drug toxicity. A subsequent phase I trial established MK-2206 200 mg weekly orally as an MTD that produced a 2-fold higher Cmax and comparable systemic exposure, tolerability relative to the 60 mg QOD dosing²⁵. The lower trough concentration with the weekly schedule compared to the QOD schedule was considered an advantageous relief from drug exposure that would make MK-2206 more tolerable, particularly in combination with other drugs²⁵. The most common MK-2206 related adverse events observed with the weekly dosing were fatigue (45.5%), rash (42.4%), diarrhea (27.3%), nausea (27.3%), vomiting (24.2%), decreased appetite (21.2%), stomatitis (15.2%), and pruritis/increased alanine aminotransferase/headache (12.1% each)²⁵. One patient had grade 1 hyperglycemia (3.0%)²⁵. Given this data, we evaluated the MK-2206 weekly dosing in this phase II ACC trial.

MATERIALS & METHODS

Study Patients

Alliance A091104 was a single-arm phase II trial of MK-2206 in patients with ACC conducted through the Alliance for Clinical Trials in Oncology and the National Clinical Trials Network (NCTN). Patients were required to have pathologically confirmed ACC (salivary or non-salivary primaries) not amenable to curative surgery or radiotherapy. Evidence of new or progressive lesion(s) on imaging performed within 6 months prior to study enrollment and/or worsening disease-related symptoms were required. Patients were required to have Response Evaluation Criteria in Solid Tumors version 1.1 (RECIST v1.1) measurable disease to be eligible. Other key inclusion criteria included age ≥18 years, Eastern Cooperative Oncology Group (ECOG) performance status ≤2 or Karnofsky performance status of ≥50%, and the ability to swallow whole tablets. Key exclusion criteria included prior treatment with PI3K/AKT/mTOR inhibitors for R/M ACC, brain metastases, and hemoglobin A1c of ≥8% and/or uncontrolled hyperglycemia in diabetic patients. See Supplementary Figure 1 for the complete inclusion/exclusion criteria. The study was approved by the institutional review boards of participating sites and the study was conducted in accordance with the Declaration of Helsinki and Good Clinical Practice guidelines. Written informed consent was obtained from all patients. Patients were offered consent to an optional correlative tissue sub-study.

Study Procedures

Patients were started on MK-2206 150 mg orally once weekly (1 cycle = 4 weeks). For patients who did not require dose delay or reduction, the MK-2206 was increased to 200 mg weekly beginning with the Cycle 1, Week 3 dose. RECIST v1.1 tumor assessments were performed at baseline and then every 8 weeks for the first 6 cycles; subsequently, assessments were performed every 12 weeks. Adverse events (AEs) were recorded throughout based on Common Terminology Criteria for Adverse Events version 4.0 (CTCAE v4.0). Patients remained on study until disease progression, unacceptable adverse event(s) (AE), or withdrawal of consent. After discontinuing study treatment, patients were followed for survival until death or for a maximum of 3 years after registration. Among the 5 patients who provided consent, core research biopsies were performed on an individual tumor before the start of study therapy (baseline) and after completion of Cycle 1 and prior to initiation of Cycle 2 dosing. Tumor biopsies were immediately flash frozen in liquid nitrogen at time of collection.

MYB Immunohistochemistry (IHC) and MYB-NFIB Fluorescence In Situ Hybridization (FISH)

Paraffin sections were analyzed at Memorial Sloan Kettering Cancer Center (C.R.A., N.K.) with the rabbit anti-MYB phospho (p)-S11 antibody from Abcam (EP769, cat# ab45150 RRID: AB_778878) that binds to the N-terminus of the protein; MYB is constitutively phosphorylated by casein kinase II at this residue^{26, 27}. MYB quantification was assessed as previously published¹⁰: 2+ for strong staining in >50% of cancer cells, 1+ for weak or strong staining in <50% of the cells, and 0 for <5% staining. FISH was performed on paraffin sections utilizing custom probes developed from bacterial artificial chromosomes (BACs) covering and flanking the MYB and NFIB genes. Two-hundred successive nuclei were examined. Detection of a sufficient break-apart signal was interpreted as a positive score.

Western Blot of Frozen Research Biopsies

Frozen tumors were mechanically sheared and lysed in a RIPA buffer containing protease and phosphatase inhibitors. A range of lysate volumes from each time point were loaded for gel electrophoresis. Western blot was subsequently performed (A.L.H, G.S.K, S.D.V) with antibodies directed at N-terminal MYB pS11 (Abcam EP769Y) and cytokeratin-7 (Cell Signaling), according to standard techniques²⁸. Cytokeratin 7 is an epithelial marker of ACC, which was used in these assays as a loading control to allow comparison of MYB protein levels between baseline and on-treatment samples which contained comparable amounts of tumor.

Statistical Analysis

The primary objective was to estimate the proportion of patients with a confirmed tumor response. All patients who met eligibility and began treatment were included in the primary analysis. A two-stage phase II Simon optimal design was used to test whether there was sufficient evidence to determine if the confirmed RR was at least 20% versus at most 5%. A stage 1 analysis was to be performed after the first 12 evaluable patients were enrolled, where at least 1 responder would lead to full accrual of 37 evaluable patients. If at least 4 responses were observed among the first 37 evaluable patients, the treatment would be considered worthy of further investigation. This study design provided 90% power to detect a confirmed response rate of at least 20%, assuming a significance level of 10% if the true confirmed response rate was 5%.

Secondary endpoints included progression-free survival (PFS), overall survival (OS), and safety and tolerability. Time-to-event analyses were conducted using the Kaplan-Meier method. PFS was defined as the time of study entry to the first of either disease progression or death. Overall survival was defined as the time from study entry to death due to any cause. Descriptive statistics (frequency (%), mean, median, etc.) for the baseline data, AE data, and treatment data have been included. All tests were two-sided and analyses were performed using SAS 9.4 (Cary, NC).

Data collection and statistical analyses were conducted by the Alliance Statistics and Data Management Center. Data quality was ensured by review of data by the Alliance Statistics and Data Management Center and by the study chairperson following Alliance policies. All analyses were based on the study database frozen on February 23, 2017.

RESULTS

Patient Characteristics and Study Treatment:

Between August 2012 through February 2013, 16 patients were enrolled. Table 1 summarizes the clinical characteristics of these 16 patients. The median age was 63.5 years (range: 31–76) with the majority being female (62.5%). Of the 15 patients who had sites of metastatic disease recorded, all had metastases to the lung (100%), and a smaller percentage possessed disease in the liver (46.7%) and bone (13.3%). Fourteen of 16 patients (87.5%) had previously received systemic therapy; all had undergone surgery and radiation therapy in the past.

Table 1:

Baseline Characteristics

Patient characteristics	Total (n=16)
Age (years)
Median (Range)	63.5 (31, 76)
Gender, n (%)
Female	10 (62.5%)
Male	6 (37.5%)
Performance Score, n (%)
0	4 (25.0%)
1	12 (75.0%)
Primary Tumor Site, n (%) [n=16]
Major salivary gland	6 (37.5%)
Base of tongue	3 (18.8%)
Sinuses	3 (18.8%)
Maxillary sinus/orbit	1 (6.3%)
Trachea	2 (12.5%)
Hard palate	1 (6.3%)
Sites of Metastases, n (%) [n=15]
Lung	15 (100.0%)
Liver	7 (46.7%)
Abdomen	3 (20.0%)
Bone	2 (13.3%)
Nodal	6 (40.0%)
Other	4 (26.7%)
Prior Surgery, n (%)	16 (100.0%)
Prior Radiation Therapy, n (%)	16 (100.0%)
Prior Systemic Therapy, n (%)	14 (87.5%)

Efficacy

All patients were started with MK2206 at 150 mg orally weekly. Twelve of the 16 patients were escalated to the 200 mg weekly dose. Of the 16 enrolled patients, two were discovered to be ineligible (one was treated prior to completion of registration, and the other had brain metastases at the time of enrollment), leaving 14 evaluable for the efficacy endpoints. Efficacy outcomes are summarized in Table 2. Of the 14 evaluable patients, none had a partial response or complete response, which led to trial closure at the stage 1 analysis per the study design. Thirteen of the 14 patients did have a best response of stable disease (92.9%), while one patient had a best response of disease progression (Table 2). The median PFS was 9.7 months (95% CI: 3.8–11.8) and the 12-month PFS rate was 21.4% (95% CI: 5.2–44.8%) (Figure 1A). With a median follow-up of 41.4 months, the median OS was 18.0 months (95% CI: 11.8–29.9), with a 1-year OS estimate of 78.6% (95% CI: 47.2–92.5%) (Figure 1B).

Table 2:

Efficacy Outcomes

	No. of patients (%)
Best Overall Response (n=14)
Complete Response	0
Partial Response	0
Stable Disease	13 (92.9%)
Progression of Disease	1 (7.1%)
Median PFS (months, 95% CI) (n=14)	9.7 (3.8–11.8)
Median OS (months, 95%CI) (n=14)	18.0 (11.8–29.9)
Reasons for Discontinuation (n=16)
Progression of disease	10 (62.5%)
Toxicity	2 (12.5%)
Refusal of further treatment	4 (25%)

PFS, Progression-free survival

OS, Overall survival

Figure 1:

Kaplan-Meier curves for progression-free and overall survival.

MK-2206 Safety and Reasons for Discontinuation

Sixteen patients were evaluable for treatment-emergent AEs (TEAEs). A median of 5 cycles (range:1–29) were administered. Twelve (75%) patients experienced at least one ≥ grade 3 TEAE, regardless of attribution. Two patients experienced a grade 4 AE (13%; an unrelated grade 4 eye-disorder and unlikely related grade 4 hyperuricemia). No patients experienced a grade 5 AE. Nine (56%) patients experienced a grade 3 AE that was attributed as being at least possibly related to MK-2206, all of which are summarized in Table 3. The most common related grade 3 AEs (>10% of patients) were rash (38%), fatigue (19%), decreased lymphocyte count (13%), and hyperglycemia (13%). The most common reason for study discontinuation was disease progression (10 [62.5%]). Four patients (25%) refused further treatment and 2 (12.5%) were removed for toxicity (Table 2).

Table 3:

Adverse Events across all patients at least possibly related to treatment (maximum grade per patient)

	Grade (N %)
Adverse Event	1		2		3		4
Rash maculo-papular	3	18.8	3	18.8	6	37.5
Fatigue	6	37.5	1	6.3	3	18.8
Platelet count decreased	7	43.8
Hyperglycemia	3	18.8	1	6.3	2	12.5
Lymphocyte count decreased	2	12.5	2	12.5	2	12.5
Blood and lymphat sys disord - Oth spec	6	37.5
Pruritus	3	18.8	2	12.5	1	6.3
Anorexia	1	6.3	3	18.8
Weight loss	3	18.8	1	6.3
Dry skin	3	18.8	1	6.3
Hypophosphatemia	2	12.5	2	12.5
Nausea	4	25.0
Diarrhea	2	12.5			1	6.3
Alkaline phosphatase increased	3	18.8
Mucositis oral	3	18.8
Hypocalcemia	1	6.3	1	6.3
Alanine aminotransferase increased	1	6.3	1	6.3
Dry mouth	1	6.3	1	6.3
Dysgeusia	1	6.3	1	6.3
Pain in extremity	1	6.3	1	6.3
Anemia	1	6.3
Cough	1	6.3
Constipation	1	6.3
Creatinine increased	1	6.3
Dehydration			1	6.3
Dyspnea					1	6.3
ECG QT corrected interval prolonged	1	6.3
Headache			1	6.3
Skin and subcut tissue disord - Oth spec	1	6.3
Vomiting	1	6.3
Allergic rhinitis	1	6.3
Alopecia	1	6.3
Aspartate aminotransferase increased	1	6.3
Cheilitis			1	6.3
Confusion	1	6.3
Dry eye			1	6.3
Dyspepsia	1	6.3
Edema limbs	1	6.3
Hypomagnesemia	1	6.3
Oral pain					1	6.3
Palmar-plantar erythrodysesthesia syndrome			1	6.3
Productive cough			1	6.3
Purpura	1	6.3
Sinus bradycardia	1	6.3

Analysis of Baseline MYB and NFIB Status

MYB IHC and FISH for MYB/NFIB rearrangements were performed on tumors from 14 patients (Table 4). Twelve (86%) had detectable MYB expression by IHC: four (29%) with 1+ staining and 8 (57%) with 2+. For FISH, 7 (50%) harbored both MYB and NFIB break-apart signals, while 4 (29%) were negative for rearrangements of either gene. One had only the MYB break-apart signal, and two had only the NFIB break-apart signal. All 7 of the MYB+/NFIB+ rearranged tumors had detectable MYB protein by IHC (6 cases with 2+, 1 case with 1+); the one MYB+/NFIB- case did not have detectable MYB by IHC. Among the 4 MYB-/NFIB- tumors, three had detectable MYB IHC expression (2 cases with 2+, 1 case with 1+). In the 2 MYB-/NFIB+ cases, both had positive MYB IHC results (both 1+).

Table 4:

MYB Immunohistochemistry (IHC) and MYB/NFIB Fluorescent in situ Hybridization (FISH)

MYB IHC	N=14
Score 0	2
Score +1	4
Score +2	8
MYB/NFIB FISH	N=14
MYB (−)/ NFIB (−)	4
MYB (+)/ NFIB (−)	1
MYB (−)/NFIB (+)	2
MYB (+)/NFIB (+)	7

Impact of MK-2206 Therapy upon Tumor MYB Protein Level

Five patients had the same tumor sampled at baseline (“Pre MK2206”) and on-treatment (“On MK-2206”) between administration of the last Cycle 1 dose and before Cycle 2 (Figure 2). At the time of the on-treatment biopsy, 4 patients were on 200 mg of MK-2206, one was on 100 mg (Patient #4 in Figure 2). All patients had SD as BOR. The biopsies were analyzed by Western blot for full-length MYB expression; 4/5 patients had MYB expression detectable by Western (Figure 2A). Given variable amounts of tumor cells and stroma in the baseline and on-treatment tumor biopsies, a range of lysate volumes for each sample was loaded for Western blot analysis to detect cytokeratin 7, an epithelial marker representative of tumor cell content that was used to normalize comparisons between the serially acquired specimens. For Patients #1 and #2, the initial analysis of a range of lysates was difficult to interpret due to an artifact limited to the cytokeratin 7 bands for the on-treatment samples (Supplementary Figure 2). Repeat Western blots comparing 10-fold more on-treatment lysate than pre-treatment revealed substantially lower MYB protein with MK-2206 in these patients (Figure 2A). For Patients #3 and #4, MYB protein levels were compared between pre- (baseline) and on-treatment lanes that possessed comparable amounts of tumor cells (i.e., similar amounts of cytokeratin 7), revealing only modestly decreased MYB with MK-2206 (Figure 2A). The change in MYB with MK-2206 treatment for Patient #1 was also confirmed via IHC (2+→1+) (Figure 2B).

Figure 2:

Serial tumor analysis in four MK-2206 treated patients. “Pre-” (or “Baseline”) and “On-” MK-2206 (Week 4) tumor biopsies were analyzed. A, Western blots for MYB. Increasing amounts of each lysate were analyzed to normalize tumor content using cytokeratin 7 as a surrogate epithelial marker. For Patients #1 and #2, this analysis (Supplementary Figure 2) led to the experiment shown here comparing 10-fold more on-treatment lysate to pre-treatment lysate. Shorter and longer exposures of each blot are shown. B, Immunohistochemistry for MYB was performed on serial tumor biopsies collected from Patient #1. Another patient with a tumor that was MYB-negative/NFIB-positive by FISH was analyzed by Western blot, but no MYB protein was detectable (data not shown). Crosses denote lanes with comparable tumor content by cytokeratin 7 for patients 3 and 4; “Neg MYB FISH”, indicates no MYB break-apart signal was detected by FISH in the tumor; “Pos MYB FISH”, indicates a MYB break-apart signal was detected by FISH in the tumor.

DISCUSSION

Based on the preclinical finding that MYB expression is at least in part reliant upon native PI3K/AKT pathway signaling, we conducted this clinical trial to test the hypothesis that treatment with the AKT inhibitor MK-2206 could elicit clinical responses in patients with incurable R/M ACC. However, the lack of partial responses observed in the first stage represented a failure to meet the primary endpoint, leading to study closure.

Though the biologic rationale for the trial was based on the putative efficacy of inhibiting MYB in ACC tumors, the trial was designed to evaluate MK-2206 in all-comers since 1) it was presumed that almost all ACCs are likely to express MYB (a recent multi-institutional study reported >90% of ACCs express MYB¹⁵), and 2) a validated MYB expression assay was not available to enrich for MYB expressing tumors. Retrospective testing revealed that 12/14 (86%) patients had evidence of MYB expression by IHC with an antibody directed at the phosphorylated (p)S11 residue in the N-terminus of the protein, suggesting that the biomarker of interest was present in almost all patients. For the two IHC negative patients, MYB expression cannot be ruled out given recent data that a subset of ACC tumors may utilize an alternative promoter to express an N-terminally truncated MYB protein, which is missing the pS11 residue that the antibody used here targets²⁹. It is also possible that these two tumors are truly negative for MYB expression, and are instead MYBL1-driven^{12, 13}, though this was not formally tested. Notably, all 7 of the MYB-NFIB rearranged tumors possessed evidence of MYB expression by IHC.

Serial biopsies from four patients on the trial revealed that MK-2206 treatment was associated with decreased MYB protein levels in patient tumors (Figure 2), consistent with observations from preclinical studies that inhibitors of the PI3K/AKT pathway can downregulate MYB in the salivary cancer context^{18, 20, 22}. However, MYB downregulation in these tumors was incomplete (Figure 2), and the degree to which MYB protein levels decreased was variable. Several other caveats regarding the Western blot analysis should also be taken into consideration. Western blot is semi-quantitative, making it difficult to precisely quantify how much MYB downregulation was achieved in each case. Additionally, the MYB protein band analyzed here is the one that corresponds to the molecular weight for full-length MYB. Two tumors in this analysis (Patients #2 and #3) had MYB break-apart signals by FISH, suggesting truncation of MYB and possibly fusion to a gene partner. Given the high variability in breakpoints observed with MYB rearrangements in ACC³⁰, we cannot determine if the MYB band analyzed represents expression of the MYB alteration or just the full-length protein alone. Another caveat is that both the Western blots and the IHC were performed with the anti-MYB pS11 antibody; hence, we cannot definitively rule out that the changes observed represent loss of phosphorylation at this serine residue rather than decreased levels of total MYB protein. Prior studies suggest that casein kinase II (CK2) phosphorylation of MYB at S11 is constitutive^{26, 27}, and there is no data to suggest that AKT regulates S11 phosphorylation³¹. It is also possible that AKT inhibition switches ACC tumors to preferentially utilize the alternative promoter that generates an N-terminally truncated MYB protein not detected by the antibody utilized in these studies²⁹, though there has been no evidence to date that AKT regulates promoter utilization in ACC tumors. Hence, our current understanding of MYB biology in ACC still favors the interpretation that these observations represent MK-2206 treatment variably reducing MYB protein levels in ACC tumors.

Since only one time point was assessed, the durability of MYB suppression achieved was not evaluated. Prior pharmacodynamic analysis of changes in AKT pathway signaling achieved with weekly MK-2206 dosing demonstrated significant target inhibition for at least 96 hours, but nearly complete recovery of AKT activation by the time for the next weekly drug dose (as reflected by AKT phosphorylation at serine 473)²⁵. It is possible that more sustained AKT inhibition with frequent drug dosing could have more durably suppressed MYB protein levels and translated to greater clinical efficacy. Perhaps the next-generation PI3K and AKT inhibitors now in clinical development could provide both more potent and sustained inhibition of these signals to more robustly deplete MYB. It has also been well established that inhibiting AKT can paradoxically activate receptor tyrosine kinases (RTKs), such as insulin-like growth factor-1 receptor (IGF-1R)³², via release of negative feedback signaling, thereby compromising pathway inhibition. The IGF-1R pathway has also been shown to be critical for promoting MYB-NFIB fusion expression through AKT in ACC tumor cells^{18, 33}. A cancer cell line screen revealed that high MYB expression (at the transcript level) predicted sensitivity to the IGF-1R targeting antibody figitumumab (CP-751,871; Pfizer Oncology)³⁴, supporting the premise that the IGF-1R pathway is a critical mediator of MYB driven tumor cell proliferation. Consistent with these data, figitumumab produced anti-tumor activity in ACC PDX models^{35, 36}. Combined IGF-1R and AKT inhibition would be a rational strategy to potentially optimize pathway inhibition and more potently downregulate MYB levels in ACC.

Beyond AKT signaling, other drug classes have been implicated as potential suppressors of MYB expression, including the BET (bromodomain and extraterminal) inhibitors^{7, 37} and retinoids, specifically all-trans retinoic acid [ATRA]¹⁹. The latter was discovered in a zebrafish chemical screen to lower MYB expression¹⁹, though a subsequent clinical trial did not demonstrate major responses³⁸. A recent publication reported that in a PIK3CA mutant/MYB-NFIB positive ACC PDX model combining ATRA with the PI3Kα inhibitor alpelisib more potently decreased MYB protein levels than either drug alone, correlating with superior inhibition of ACC proliferation²⁰. Further clinical investigation of such drug combinations that more dramatically diminish MYB protein levels in ACC is warranted.

Novel agents that directly disrupt MYB transcriptional activity may possess a superior therapeutic window to induce tumor regression. One example is the development of MYBMIM³⁹, a peptidomimetic inhibitor that can disrupt the assembly of the MYB transcriptional complex, displacing it from MYB-related enhancers to downregulate MYB-target gene expression. MYBMIM and another MYB targeted agent celastrol both have been shown to inhibit the proliferation of MYB-expressing ACC cell line models in vitro¹⁷. Trials testing this class of agents as monotherapies or in combination with other MYB-targeted approaches in ACC patients would be of interest.

With MYB activation presumed to be a truncal oncogenic event in ACC tumorigenesis, the genomic landscape of the disease is otherwise highly heterogeneous¹⁶, nominating other molecules and pathways that could be targeted in combination with MYB inhibition⁴⁰. Approximately 18–25% of R/M ACCs possess activating alterations in NOTCH1, which is associated with a significantly worse prognosis^{16, 40–42}. Even in tumors without NOTCH gene alterations, NOTCH pathway activation has been observed in the luminal subcellular compartment of ACCs⁷, which may be activated via paracrine signals from the myoepithelial cell compartment in tumors that possess this biphasic cellular phenotype⁴³. Preclinical experiments also suggest that high grade, solid type ACCs, which are composed primarily of luminal cells and often possess NOTCH1 mutations, may be more resistant to MYB targeting alone⁷. These lines of evidence suggest that combined MYB and NOTCH pathway targeting may be an effective strategy for ACCs with NOTCH1 mutations, and perhaps even in NOTCH wild-type tumors with luminal cell activation of the pathway. Beyond NOTCH, a recent proteogenomic analysis of ACC tumors identified biologic subsets of the disease that propose for consideration other rational molecular targets for combination with MYB inhibition (e.g. RTKs, BCL-2, PRMT5, etc…)⁴⁰.

CONCLUSION

In summary, MK-2206 alone elicited insufficient clinical activity in patients with incurable ACC, though the correlative tissue analysis suggests that at least partial downregulation of MYB may have been achieved. These results highlight the need for more studies to delineate the utility of MYB as a therapeutic target for patients with ACC.