PI3K in CLL: are 2 isoforms better than 1?

Andrew J. Steele

In this issue of Blood, Dong et al present a series of experiments with the novel B-cell receptor (BCR) kinase inhibitor IPI-145 (phosphatidylinositol 3-kinase γ/δ [PI3Kγ/δ] inhibitor) and show that chronic lymphocytic leukemia (CLL) samples resistant to ibrutinib remain sensitive to killing by this agent.1

Binding of antigen/autoantigen to the BCR results in formation of the signalosome that includes SYK, BTK, and PI3K, all of which have been targeted by kinase inhibitors that have shown promise for the treatment of CLL. However, the PI3K complex also regulates signaling from IL-4, CD40L, CXCL12, BAFF, and the Toll-like receptor (TLR). Several signaling pathways are activated downstream of these kinases, which results in the elevation of NFAT, MYC, JUN, nuclear factor-κB (NF-κB), S6, Bim, Noxa, and β-catenin, which are involved in survival, apoptosis, proliferation, and migration. Mammalian target of rapamycin (mTOR) immediately downstream of PI3K/AKT signaling is the catalytic subunit of the mTORC1 and mTORC2 complexes. mTORC2 is capable of inducing a positive feedback loop that leads to the phosphorylation of serine 473 (P S473) and reactivation of AKT downstream signaling.

In normal healthy B cells, the BCR detects and responds to antigens displayed on pathogens. Upon engagement, it activates 2 key processes; first, it triggers the formation of the “signalosome,” a complex of kinases and proteins leading to downstream signal transduction consequent to apoptosis, proliferation, and/or migration (see figure); second, it internalizes the captured antigen for processing and subsequent presentation to T cells.

Strong evidence now exists to suggest that BCR signaling has a major role in regulating the behavior of CLL cells, and roles in the development, progression, and clinical response to treatment have all been proposed.2 These observations provide a strong rationale for targeting the BCR in CLL. BCR kinase inhibitors for the treatment of CLL have shown impressive clinical responses, which has led to significant therapeutic advances in the treatment of difficult-to-treat and otherwise refractory patients. Two inhibitors of the BCR pathway have recently been approved by the Food and Drug Administration for various cohorts of CLL: the Bruton tyrosine kinase (BTK) inhibitor ibrutinib and the PI3Kδ inhibitor idelalisib. Both drugs were hypothesized to exert their therapeutic effects by inhibiting CLL ingress into and/or promoting efflux out of the lymph nodes into the blood, where they no longer receive proliferative and prosurvival signaling. However, there are data to suggest BCR kinase inhibitors can also exert direct cytotoxicity,3,4 and more so in lymph nodes.4 Unfortunately, resistance to these drugs is already emerging in patients with high-risk cytogenetic features—those patients carrying del(11q22.3), del(17p13.1), or a complex karyotype.5 Therefore, it is important to fully understand how these inhibitors work and to demonstrate whether, once resistance emerges, alternative BCR pathway inhibitors can be utilized. Moreover, it is still too early to tell whether these agents with long-term use will be curative or will just keep the disease suppressed. Therefore, there is clearly a need to develop alternative BCR signaling inhibitors, including PI3K inhibitors, with greater efficacy.

Class 1 PI3K signaling is pivotal for cell survival in numerous cancers and is known to be overactive in CLL, in patients with unmutated immunoglobulin heavy chain variable (IGHV) having increased PI3K expression/activity compared to patients with mutated IGHV.6 The class 1 PI3K comprises class 1A (PI3Kα, PI3Kβ, and PI3Kδ) and class 1B (PI3Kγ) isoforms. Although PI3Kδ and PI3Kγ expression is largely restricted to leukocytes, PI3Kα and PI3Kβ are generally ubiquitous. PI3Kδ and PI3Kγ are pivotal for a number of leukocyte functions, including proliferation, antibody secretion, survival, and migration, while PI3Kβ has recently been described to be involved in neutrophil activation and generation of reactive oxygen species in response to immune complexes.7 There is also evidence of functional redundancy because multiple PI3K isoforms require inhibition to fully reverse the neutrophil survival induced by granulocyte-macrophage colony-stimulating factor.8 Generally, the more PI3K isoforms that are inhibited, the greater the toxicity to CLL cells. However, the caveat is that there is more off-target toxicity to other nonhematological tissues if PI3Kα and PI3Kβ are also inhibited.

In this issue, Dong et al1 investigate in vitro an alternative PI3K inhibitor, IPI-145, which is currently in a phase 3 clinical trial as a monotherapy for CLL (; #NCT02004522). Unlike idelalisib, IPI-145 targets both δ and γ isoforms of PI3K that are expressed in leukocytes. In vitro, IPI-145 induced up to ∼30% apoptosis of primary CLL cells in a dose- and time-dependent manner, which is similar to published data with idelalisib.3 Importantly, samples that were resistant to ibrutinib were still susceptible to killing from IPI-145 even though BTK exerts its effect both upstream and downstream of PI3K, suggesting PI3K may compensate after BTK/phospholipase C-γ2 mutations/disruption. IPI-145 significantly inhibited signaling with anti-immunoglobulin M at much lower concentrations than idelalisib1,9; however, while pAKTT308 and ERK 1/2T202/Y204 were completely abrogated and confirm PI3K inhibition, pAKTS473 was only partially inhibited. This may suggest that AKTS473 is regulated by another pathway or that the known positive feedback mechanism induced by mammalian target of rapamycin complex 2 (mTORC2) and observed in other hematological malignancies10 is active. Furthermore, while IPI-145 was not toxic to normal control T and B cells, CLL T, B, and natural killer cells were more sensitive to this agent. Further work is required to understand this mechanism and what it will mean to patients.

Although these data confirm BCR signaling was inhibited by IPI-145, there are still a significant number of unanswered questions. Will IPI-145 inhibit CXCR4/CXCL12 signaling and subsequent migration and T-cell (CD40L and interleukin-4 [IL-4]), stromal (BAFF, IL-6), and TLR–mediated signaling as previously shown with idelalisib? Will IPI-145 be superior to idelalisib in patients? Will IPI-145 need to be combined with another agent such as BH3 mimetics or monoclonal antibodies (mAbs) to obtain greater efficacy? In theory, both should synergize well with IPI-145 through the previously mentioned mechanism of retaining CLL cells in the blood, both reducing the anti-apoptotic signaling and providing easy access for mAbs. Finally, is inhibiting 2 isoforms better than inhibiting 1? Might the inhibition of the γ isoform of PI3K, induced by IPI-145 but not idelalisib, in fact lead to more tumor toxicity in vivo? Similarly, might it reduce the effector functions of various immune effectors when used in combination with mAbs? For these answers, we eagerly await the results of the upcoming trials and further in vitro experimentation.

Despite these unresolved questions, this article provides exciting new data and insight into the biology of the PI3K signaling pathway in CLL cells, which will enable the development of more effective drugs for the treatment of this currently incurable disease. Moreover, the more inhibitors we have for the BCR signaling pathway, the more tools we will have to further dissect its critical signaling functions in malignant cells and the more opportunities we will have to explore rational combinations for improved therapeutic efficacy in the future.


  • Conflict-of-interest disclosure: The author declares no competing financial interests.