TORn in Two Over Breast Cancer Drug Resistance

In recent years it has become increasingly evident that the phosphatidylinositol 3-kinase (PI3K)/Akt/mammalian target of rapamycin (mTOR) signaling pathway is deregulated in breast cancer and plays a significant role in mediating drug resistance.[1-3] Indeed, breast cancer genome sequencing studies have revealed recurrent somatic mutations in several of the PI3K pathway components, including the PI3K catalytic subunit, p110α, the downstream protein kinase, Akt, and the tumor suppressor, PTEN.[4] Not surprisingly, these discoveries have driven investigators to develop drugs for many of the PI3K/mTOR pathway proteins. Much of the effort has focused on new drug derivatives of rapamycin (termed rapalogs), which selectively inhibit the downstream protein kinase complex, mTORC1, but not mTORC2. However, more recent work has expanded to include development of inhibitors that target multiple components within the PI3K/mTOR pathway.

In this issue of ONCOLOGY, Drs. Vinayak and Carlson provide an informative review of the relevant studies aimed at evaluating the effectiveness of PI3K/mTOR pathway inhibition for breast cancer. The authors highlight several important considerations for future studies aimed at targeting the PI3K/mTOR pathway within the different breast cancer subtypes-namely, the importance of accurately identifying patients most likely to benefit from these specific agents, the advantage of dual-acting PI3K/mTOR inhibitors over mTORC1 inhibitors, and the relevant clinical considerations concerning the notable toxicity associated with current rapalog therapies. To date, the majority of the clinical studies evaluating rapalogs for breast cancer have focused on estrogen receptor (ER)-positive refractory disease, as most patients within this subgroup develop resistance to their respective therapy, and preclinical evidence has suggested a high incidence of aberrant PI3K/mTOR signaling activity, presumably mediated through somatic mutations in constituent genes. However, emerging work and early-phase clinical results suggest that targeted PI3K inhibitors may be highly effective for additional subtypes of breast cancers.

At least for combating drug resistance with rapalogs in ER-positive patients, a combinatorial approach will largely be favored, as heavily pretreated hormone receptor–positive patients with metastatic disease have displayed little benefit from rapalog monotherapies.[5,6] In fact, Vinayak and Carlson suggest that for patients with ER-positive disease, first-line hormone therapy might be necessary to sensitize tumors to possible synergistic effects of rapalogs combined with alternative endocrine therapies. It is possible that hormone therapy pretreatment may lead to the selection of cells dependent upon PI3K/mTOR pathway activity, such as cells expressing low levels of PTEN or harboring point mutations in PIK3CA or AKT1. Nonetheless, these observations and lessons learned from past trials suggest that combinatorial approaches will likely predominate in future endeavors.

In spite of the numerous preclinical studies implicating PI3K/mTOR signaling in breast cancer drug resistance, recent trial data evaluating rapalogs have demonstrated relatively modest clinical responses. Moreover, some studies have displayed conflicting correlations between PI3K pathway aberrations and response to mTORC1 inhibitors. These findings beget the question: given that PI3K/mTOR signaling contributes to drug resistance, what improvements can we make to achieve better clinical responses in future studies? First, it seems that patient selection will be an important factor. Many of the early trials aimed at combating resistance mediated by PI3K signaling did not investigate the presence of specific lesions that might predict for response to rapalogs. In addition, for the trials that have evaluated PIK3CA mutation status or PTEN expression levels, most utilized archival tissue samples. As a consequence, tumor DNA sequencing and immunohistochemical data obtained from these specimens may have been outdated. That is, information obtained from the primary lesion may not represent the current profile of the malignancy. Two recent studies support this hypothesis where a discordance in PIK3CA status was observed between the primary and metastatic lesions.[7,8]

Another problem with current approaches resides with the limited interrogation and understanding of potential components mediating the drug resistance. As several recent breast cancer genome sequencing studies have demonstrated, other oncogenic lesions may contribute to increased PI3K signaling activity, such as somatic mutations in AKT1.[9] Moreover, recent research has uncovered important negative feedback loops within the PI3K pathway, where activation of mTORC1 leads to attenuation of upstream PI3K signaling. Relieving the negative feedback, through mTORC1 inhibition with the current panel of rapalogs (everolimus [Afinitor], temsirolimus [Torisel], and ridaforolimus [AP23573, MK-8669]), may lead to prolonged PI3K signaling activity. Overall, it is becoming increasingly clear that the PI3K signaling pathway is incredibly complex, containing several different key nodes that have the potential for crosstalk through other signaling pathways, such as the mitogen-activated protein kinase (MAPK) pathway.[10]

These analyses underscore the need for both better predictive biomarkers and a more personalized approach as we move forward with new clinical trials. One improvement may come through the deep DNA sequencing of patient tumors. Whole-genome sequencing of patient tumors should help us more accurately elucidate predictive markers, and consequently, help us make more informed treatment decisions. In addition, knowing the genetic makeup of patient tumors permits the development of new DNA-based diagnostics, such as BEAMing and PARE. The two complementary technologies, BEAMing (Beads, Emulsion, Amplification, Magnetics) and PARE (Personalized Analysis of Rearranged Ends), are highly sensitive, noninvasive approaches that enable clinicians to follow circulating tumor DNA found in the blood.[11,12] By incorporating these new tools into future trials, we should be able to concurrently learn about drug resistance and significantly improve patient responses.

Financial Disclosure: The authors have no significant financial interest or other relationship with the manufacturers of any products or providers of any service mentioned in this article.

References:

REFERENCES

1. Berns K, Horlings HM, Hennessy BT, et al. A functional genetic approach identifies the PI3K pathway as a major determinant of trastuzumab resistance in breast cancer. Cancer Cell. 2007;12:395-402.

2. Miller T. Hyperactivation of phosphatidylinositol-3 kinase promotes escape from hormone dependence in estrogen receptor-positive human breast cancer. J Clin Oncol. 2010;120:1-8.

3. Knuefermann C, Lu Y, Liu B, et al. HER2/PI-3K/Akt activation leads to a multidrug resistance in human breast adenocarcinoma cells. Oncogene. 2003;22:3205-12.

4. Cancer Genome Atlas Network. Comprehensive molecular portraits of human breast tumours. Nature. 2012;490:61-70.

5. Ellard SL, Clemons M, Gelmon KA, et al. Randomized phase II study comparing two schedules of everolimus in patients with recurrent/metastatic breast cancer: NCIC Clinical Trials Group IND.163. J Clin Oncol. 2009;27:4536-41.

6. Chan S, Scheulen ME, Johnston S, et al. Phase II study of temsirolimus (CCI-779), a novel inhibitor of mTOR, in heavily pretreated patients with locally advanced or metastatic breast cancer. J Clin Oncol. 2005;23:5314-22.

7. Dupont Jensen J, Laenkholm A-V, Knoop A, et al. PIK3CA mutations may be discordant between primary and corresponding metastatic disease in breast cancer. Clin Cancer Res. 2011;17:667-77.

8. Gonzalez-Angulo AM, Ferrer-Lozano J, Stemke-Hale K, et al. PI3K pathway mutations and PTEN levels in primary and metastatic breast cancer. Mol Cancer Ther. 2011;10:1093-101.

9. Carpten JD, Faber AL, Horn C, et al. A transforming mutation in the pleckstrin homology domain of AKT1 in cancer. Nature. 2007;448:439-44.

10. Carracedo A, Pandolfi PP. The PTEN-PI3K pathway: of feedbacks and cross-talks. Oncogene. 2008;27:5527-41.

11. Diehl F, Schmidt K, Choti MA, et al. Circulating mutant DNA to assess tumor dynamics. Nature Med. 2008;14:985-90.

12. Leary RJ, Kinde I, Diehl F, et al. Development of personalized tumor biomarkers using massively parallel sequencing. Sci Transl Med. 2010;2:20ra14.