Survival of recurrent/metastatic NSCLC with palliative chemotherapy fails to exceed 1 y and there is an unmet need for new drugs and drug combinations that work through novel mechanisms 46
. In the present study, we propose ritonavir as a candidate drug for metastatic lung adenocarcinoma clinical trials, based on its inhibition of adenocarcinoma lines at concentrations in the 35–45 μM range. Such concentrations are clinically attainable, albeit with significant gastrointestinal toxicity 47
, with ritonavir monotherapy at 600 mg twice daily 48
Study of signaling pathways affected by ritonavir by siRNA profiling revealed that survivin is an important target, whereas c-Src and STAT3 appear to be of lesser importance. Furthermore, forced over-expression of survivin confers relative resistance to ritonavir, confirming importance of survivin as a ritonavir target. Ritonavir reduces survivin, in part, by reducing its mRNA levels. Because survivin is regulated in cancer primarily through its mRNA expression 49
these results suggest that ritonavir is likely attacking a basic mechanism of survivin transcriptional regulation.
Ritonavir inhibits lung adenocarcinoma growth and anchorage-dependent clonogenicity, in part, by inducing G0/G1 cell cycle arrest and, in part, by inducing apoptosis. Survivin is implicated in regulation of the G1/S 8, 9, 11
as well as G2/M checkpoint and therefore its reduction by ritonavir is expected to cause inhibition of the cell cycle. Ritonavir may also inhibit the G1/S checkpoint through down-regulation of CDKs, cyclin D1
and associated Rb phosphorylation, as well as by induction of p27Kip
and wild type p53. While reduction of survivin by ritonavir is expected to promote apoptosis in lung cancer presumably related to loss of the antiapoptotic effect of survivin 50
, the remaining mechanisms by which ritonavir induces apoptosis remain to be determined. These mechanisms could include induction of DNA damage, in part, through increased cleavage of PARP1.
Importantly, drug combination studies revealed that ritonavir is active at lower concentrations when combined with gemcitabine, cisplatin and the gemcitabine/cisplatin combination. In combination with gemcitabine or cisplatin, ritonavir exhibits IC50
values in the range of 15–20 μM. With the gemcitabine/cisplatin combination, the ritonavir IC50
is in the 8 μM range, which should be attainable with 100 mg twice daily dosing 48
. While the gemcitabine/cisplatin combination in advanced NSCLC resulted in the longest median time to progression compared to three other chemotherapy doublets, this was only 4.2 months 45
. We hypothesize, based on in vitro
synergy, that addition of low dose ritonavir to the gemcitabine/cisplatin combination may improve time to progression, with acceptable toxicity. Furthermore, because ritonavir is not myelosuppressive and potentially could be continued through the period of gemcitabine/cisplatin treatment, ritonavir could potentially inhibit re-growth of lung adenocarcinoma between cycles of chemotherapy. Therefore, a phase I study of daily ritonavir in combination with the established gemcitabine and cisplatin schedule is an important next step. While K-ras mutation status did not affect sensitivity to ritonavir, for the H838 K-ras wild type line there was lack of synergy with gemcitabine and antagonism with cisplatin. These results suggest that K-ras mutant lung adenocarcinoma is the best candidate histology for future clinical trials.
Although the mechanisms behind cooperation between ritonavir and gemcitabine and/or cisplatin are not known, it is likely these mechanisms involve survivin effects on DNA repair pathways. Gemcitabine is a DNA strand-terminator that stalls replication forks, causes S phase arrest 51
and double strand breaks (DSB) while inhibiting homologous recombination repair (HRR), which is required for repairing DSB 52, 53
. Survivin has been reported to enhance DSB repair and we hypothesize that reduction of survivin by ritonavir may increase sensitivity to gemcitabine through this mechanism 54
. Survivin reduction may also explain sensitivity of lung adenocarcinoma to ritonavir in combination with cisplatin due to increased PARP1 cleavage.
PARP1 may be involved in repair of cisplatin-induced DNA damage. PARP1 is known to recruit XRCC1 to sites of DNA damage 55
. XRCC1 is a scaffolding factor required for base excision repair (BER) 56
and recently, nucleotide excision repair (NER) 57
. Of interest, interference with NER interferes with repair of cisplatin-induced DNA damage 53
. Although PARP1 has not been implicated as a key regulator of NER, it has been recently been located at sites of cisplatin-induced DNA damage, by two photoaffinity labeling studies 58, 59
. This finding potentially implicates PARP1 in repair of cisplatin-mediated DNA interstrand crosslinks by NER. In addition, PARP1 reduction has also recently been demonstrated to play a critical role in chemosensitivity to the gemcitabine/cisplatin combination in triple negative breast cancer 60
. Future studies will determine the mechanisms by which ritonavir may enhance DNA damage by cisplatin and gemcitabine.
Based on the importance of survivin as a ritonavir target in lung adenocarcinoma, we propose that survivin may be a useful biomarker for ritonavir sensitivity. We hypothesize that among tumors expressing survivin, those exhibiting lower survivin levels will be more likely to respond to ritonavir. Our results from forced survivin over-expression are artificial and may not reflect survivin levels in tumors occurring in patients and therefore we would not recommend excluding patients with high survivin levels from clinical trials of ritonavir. Only the analysis of data from such trials would reveal whether there is a relationship between survivin levels and ritonavir sensitivity.