Background: Hyperactivation of the PI3K/AKT pathway correlates with impaired anti-tumor responses, including reduced T cell infiltration into tumor, and reduced efficacy of immune checkpoint inhibitors. Blockade of this pathway synergizes with PD-L1/PD-1 axis blockade preclinically.

Methods: This Phase I clinical trial (NCT03673787) assessed the safety, pharmacodynamic, and preliminary clinical activity of Ipa (200mg or 400mg OD) given in combination with A 1200mg q3 wk in refractory pts. Serial paired blood and tumor samples were analysed to interrogate the effect of Ipa on the tumor micro-environment and host immune system prior to the addition of the immune check point inhibitor, A.

Results: 18 adult pts were treated in dose escalation. Median age 49 yrs. All pts had ECOG PS 0-1 and median 7 prior therapies. Most common TRAEs (>15%) were mild Gr1-2 diarrhea (56%), rash (50%), fatigue (33%), nausea (33%), raised ALT/AST (33%), headache (28%) and arthralgia (22%). 1 pt had G2 systemic immune activation; 2 pts had G3 rash, both rapidly reversible. 1 DLT of G3 raised ALT seen at 200mg (1 DLT/9 evaluable pts) but none at 400mg (0 DLT/6). Of 14 RECIST evaluable patients, there were 2 confirmed PRs, and 5 SD (clinical benefit rate 50%). Reductions of CD4+FOXP3+ Tregs in tumor microenvironment were seen after 2wks of single agent Ipa, regardless of PIK3/AKT somatic mutation status (Table 1). Responding pts had a >400% median increase in intra-tumoral CD8+ Teff cell infiltration, effectively switching from a desert phenotype to an inflamed phenotype. Paired changes in FACS, transcriptome and cytokine will also be presented.Conclusions: The RP2D of Ipa 400mg OD combination with A was well tolerated with early efficacy signals. Further biomarker work is ongoing and will be evaluated in expansion cohorts.

Table 1:

Changes in immune cell populations as assessed by multicolour Immunofluorescence in paired biopsies of breast/gynae patients, % change in cell number/mm2 from baseline (median [min,max$])&

Post 2 weeks single agent Ipatasertib(n=9)Post 1 cycle of combination Ipatasertib and Atezolizumab(n=7)
CD4+FOXP3+Tregs cellsCD 8+ Teff cellsCD4+FOXP3+Tregs cellsCD 8+ Teff cells
Intra-tumourstromaIntra-tumourstromaIntra-tumourstromaIntra-tumourstroma
All patients -23.9*[-89.7, BL0-30.0*[-91.6, BL0-37.7*[-84.4, -24.5] -28.4[-92.4, 259.8] 335.9[-44.0,BL045.4[-51.0, BL059.6[-60.6,493.3] 64.7[-51.7,293.3] 
Stratified by somatic PI3K/AKT/PTEN mutational status 
Pathogenic mutations (mt) 11.1[-82.2, BL0]# -10.7[-91.6, BL0]Φ ns ns ns ns -30.5[-60.6,-0.5] 11.3[-51.7,50.0] 
Wildtype (wt) -63.1[-89.7,19.0]# -47.5[-77.0,11.1]Φ ns ns ns ns 426.5[59.6,493.3] 126.7[79.4,293.3] 
Stratified by response 
Responders (PR + SD>4 cycles). 1 ER+ HER2+ breast cancer (wt), 1 ER+ HER2- breast cancer (wt) 459.9[426.5,493.3]@ 103.1[79.4,126.7] 
Non-responders (PD at 4 cycles) 1 cervical cancer, 4 ER+ breast cancer -0.5[-60.6, 59.6]@ 30.6[-51.7,293.3] 
Post 2 weeks single agent Ipatasertib(n=9)Post 1 cycle of combination Ipatasertib and Atezolizumab(n=7)
CD4+FOXP3+Tregs cellsCD 8+ Teff cellsCD4+FOXP3+Tregs cellsCD 8+ Teff cells
Intra-tumourstromaIntra-tumourstromaIntra-tumourstromaIntra-tumourstroma
All patients -23.9*[-89.7, BL0-30.0*[-91.6, BL0-37.7*[-84.4, -24.5] -28.4[-92.4, 259.8] 335.9[-44.0,BL045.4[-51.0, BL059.6[-60.6,493.3] 64.7[-51.7,293.3] 
Stratified by somatic PI3K/AKT/PTEN mutational status 
Pathogenic mutations (mt) 11.1[-82.2, BL0]# -10.7[-91.6, BL0]Φ ns ns ns ns -30.5[-60.6,-0.5] 11.3[-51.7,50.0] 
Wildtype (wt) -63.1[-89.7,19.0]# -47.5[-77.0,11.1]Φ ns ns ns ns 426.5[59.6,493.3] 126.7[79.4,293.3] 
Stratified by response 
Responders (PR + SD>4 cycles). 1 ER+ HER2+ breast cancer (wt), 1 ER+ HER2- breast cancer (wt) 459.9[426.5,493.3]@ 103.1[79.4,126.7] 
Non-responders (PD at 4 cycles) 1 cervical cancer, 4 ER+ breast cancer -0.5[-60.6, 59.6]@ 30.6[-51.7,293.3] 

*significant change (p≤0.05; Wilcoxon sign-rank test) from baseline, $maximum values denoted by BL0indicate that the baseline value was zero, and so percentage change from baseline is not defined. For the analysis, the baseline value has been replaced by a nominal value of 0.1 so that a large percentage increase is associated with these cases. Note that these large percentage increases do not affect the non-parametric statistical tests used.#no significant difference in distribution of reduction in intra-tumoural CD4+ FOXP3+Tregsbetween pts with pathogenic mutations in PI3K/AKT and those without (p=0.30; Wilcoxon rank-sum test)Φno significant difference in distribution of reduction in stromal CD4+FOXP3+Tregsbetween pts with pathogenic mutations in PI3K/AKT and those without (p=0.44; Wilcoxon rank-sum test) @ difference between responders and non-responders p=0.083; Wilcoxon rank-sum test)

mt pathogenic mutations in PI3K/AKT and PTEN as per COSMIC database present in tumour or PTEN loss by IHC. wt no pathogenic mutations in PI3K/AKT and PTEN as per COSMIC database detected in tumour and intact PTEN expression by IHC. &exploratory analyses with no adjustment for multiple testing

Citation Format: Juanita S. Lopez, Andrea Biondo, Crescens Tiu, Mariana Scaranti, Malaka Ameratunga, Anna Zachariou, Alison Turner, Nina Tunariu, Toby Prout, Mona Parmar, Hannah Badham, Karen Swales, Wei Yuan, Ricardo Morilla, Mateus Crespo, Rob Daly, Ines Figueiredo, Bora Gurel, Rita Pereira, Ruth Riisnaes, Igor Vivanco, Anna Minchom, Ben Jenkins, Christina Yap, Udai Banerji, Johann De Bono. Proof-of-concept evidence of immune modulation by blockade of the phosphatidylinositol 3-kinase (PI3K)-AKT signaling pathway in the phase I dose escalation study of Ipatasertib (Ipa) in combination with atezolizumab (A) in patients (pts) with advanced solid tumors (Ice-CAP) [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr CT140.