Radiosensitising Cancer Using Phosphatidylinositol-3-Kinase (PI3K), Protein Kinase B (AKT) or Mammalian Target of Rapamycin (mTOR) Inhibitors

Radiotherapy is routinely used as a neoadjuvant, adjuvant or palliative treatment in various cancers. There is significant variation in clinical response to radiotherapy with or without traditional chemotherapy. Patients with a good response to radiotherapy demonstrate better clinical outcomes universally across different cancers. The PI3K/AKT/mTOR pathway upregulation has been linked to radiotherapy resistance. We reviewed the current literature exploring the role of inhibiting targets along this pathway, in enhancing radiotherapy response. We identified several studies using in vitro cancer cell lines, in vivo tumour xenografts and a few Phase I/II clinical trials. Most of the current evidence in this area comes from glioblastoma multiforme, non-small cell lung cancer, head and neck cancer, colorectal cancer, and prostate cancer. The biological basis for radiosensitivity following pathway inhibition was through inhibited DNA double strand break repair, inhibited cell proliferation, enhanced apoptosis and autophagy as well as tumour microenvironment changes. Dual PI3K/mTOR inhibition consistently demonstrated radiosensitisation of all types of cancer cells. Single pathway component inhibitors and other inhibitor combinations yielded variable outcomes especially within early clinical trials. There is ample evidence from preclinical studies to suggest that direct pharmacological inhibition of the PI3K/AKT/mTOR pathway components can radiosensitise different types of cancer cells. We recommend that future in vitro and in vivo research in this field should focus on dual PI3K/mTOR inhibitors. Early clinical trials are needed to assess the feasibility and efficacy of these dual inhibitors in combination with radiotherapy in brain, lung, head and neck, breast, prostate and rectal cancer patients.


Introduction
Radiotherapy is widely used in the treatment of cancer. Radiotherapy in isolation or in combination with chemotherapy may be administered to cancer patients with curative intent [1,2]. Approximately 40% of all patients cured of their cancer will have received radiotherapy as part of their treatment [3]. Radiotherapy can also be used as a neoadjuvant, adjuvant or palliative treatment. Patient response to radiotherapy is widely variable [4,5]. Patients with a good response to radiotherapy generally benefit from a survival advantage over their poorly responding counterparts [4,6]. Those with a good response can also be managed conservatively using "watch and wait" strategies, potentially avoiding the need for further invasive treatment (e.g., surgery) [7,8].
The quest to better understand radiotherapy response and resistance has seen decades of research focus on the intracellular mechanisms activated in response to radiotherapy. Ionising radiation administered as radiotherapy leads to DNA double strand break (DSB) within the cellular genome [9]. This DNA damage activates intracellular pathways geared at either repairing damaged DNA or triggering cell death [10,11]. The two main types of DNA repair processes following DNA DSB include non-homologous end joining (NHEJ) repair and homology directed recombinant (HDR) repair [12,13]. DNA repair through HDR uses template donor DNA from a sister chromatid resulting in a more accurate repair. NHEJ repair joins the damaged ends of the DNA strand together. This can lead to the deletion of DNA segments resulting in changes in the genetic code. Where repair is impossible the cell will undergo apoptosis followed by autophagy. Radiotherapy also promotes significant changes within the tumour microenvironment [14]. The immune system also plays an important role in eliminating damaged or dying cells following radiotherapy through the dispersion of immune system stimulating tumour antigens [15]. Traditionally, 5-fluorauracil (5-FU) based chemotherapy has been used to sensitise tumours to radiotherapy. However, the efficacy of these drugs in patients remains poor [16]. This is arguably due to the clonal selection of chemoradioresistant subpopulations of cells within the treated tumour; harbouring pro-survival and anti-apoptotic mutations [17].
Significant stresses are exerted upon the cell and its microenvironment by radiotherapy. This leads to the activation of pro-survival pathways such as the phosphatidylinositol-3-kinase (PI3K), protein kinase B (AKT) and mammalian target of rapamycin (mTOR) pathway [18]. This complex intracellular pathway contains multiple downstream effectors involved in cell proliferation, growth, survival, cell migration and membrane trafficking (Figure 1 / Figure S1). It is tightly regulated with several activators and negative feedback loops. The PI3K/AKT/mTOR pathway is activated by substrates binding to G-protein coupled receptor (GPCR) or receptor tyrosine kinase (RTK) leading to the activation of membrane bound class-I PI3K proteins [19,20]. Class-I PI3K may also be activated directly by RAS [21]. Activated class-I PI3K phosphorylates transmembrane protein phosphatidylinositol-4,5-bisphosphate (PIP2) to phosphatidylinositol-3,4,5-bisphosphate (PIP3) [22]. This process is inhibited by phosphatase and tensin homolog (PTEN) [23]. PIP3 activates AKT directly by phosphorylation or indirectly via the recruitment of phosphoinositide-dependent protein kinase (PDK1) [24]. mTORC2 is also capable of direct AKT activation [25].
There are numerous pharmacological inhibitors of the PI3K/AKT/mTOR pathway which are widely commercially available ( Figure 2). In vitro research has also shown that the inhibition of RAS by EGFR inhibitors leads to enhanced radiation response by indirectly inhibiting the PI3K/AKT/mTOR pathway [42]. Therefore, recent attention has shifted towards direct PI3K/AKT/mTOR pathway component inhibition to enhance radiotherapy response (Tables 1-4). This literature review identified several studies utilising PI3K, AKT and/or mTOR inhibitors to enhance the radiosensitivity of different types of tumour. Most research investigating the radiosensitising properties of pharmacological inhibitors of the PI3K/AKT/mTOR pathway remains at in vitro or in vivo xenograft stages. A few drugs have undergone early clinical trials ( Table 5). The research was broadly categorised based on the type of cancer (adenocarcinoma: Table 1, squamous cell carcinoma (SCC) and non-small cell lung cancer (NSCLC): Table 2, glioblastoma multiforme  (GBM): Table 3, and other tumour types: Table 4), and further subcategorised by type of inhibitor (e.g., single or dual).

Sensitising Adenocarcinoma to Radiotherapy
Neoadjuvant long course chemoradiotherapy (LChRT) followed by surgical resection with total mesorectal excision (TME) is the gold standard treatment for patients with locally advanced rectal cancer (LARC) [43]. Neoadjuvant treatment modalities involving radiotherapy have been shown to significantly reduce the risk of local recurrence and increased rates of clear resection margins (R0 resections) in LARC patients. Complete pathological response (pCR) rates of up to 10-30% and 2-6% have been reported in LARC patients following LChRT or short course radiotherapy (SCRT) respectively [44,45]. Patients with pCR demonstrate a decreased risk of local, distant recurrence and improved survival [45,46]. It is also important to note that there is a significant proportion of rectal cancer patients with a poor response to neoadjuvant therapy [47]. There is an urgent clinical need to improve response to treatment especially in this cohort, given that resistance to chemoradiotherapy poses a significant treatment challenge.
Research focusing on the use of PI3K/AKT/mTOR pathway inhibitors to improve radiotherapy response in rectal cancer is still in its infancy. In 2015, Chen et al. (1) described the radiosensitising role of dual class-I PI3K and mTORC1/C2 inhibitor (NVP-BEZ235) in vitro using KRAS mutant (HCT 116, SW 620) and wild type (HT 29) colorectal (CRC) cell lines [18]. The authors demonstrated activation of the PI3K/AKT/mTOR pathway following radiotherapy in the treated cell lines. There was a gradual increase in the levels of pAKT and phosphorylated mTOR observed up to three hours following exposure to 5 Gy radiotherapy. Dual inhibition with NVP-BEZ235 led to an increase in radiosensitivity of all three CRC lines. The authors observed a reduction in cell proliferation as well as loss of expression of phosphorylated DNA-PKcs and phosphorylated Ribosomal Protein S6 (rpS6) (protein involved in cell growth and proliferation). A significant reduction in HCT 116 CRC mouse xenografts tumour size following NVP-BEZ235 and radiotherapy treatment was also observed (p < 0.001). Chen et al. (2) also demonstrated the radiosensitising potential of combined radiotherapy and NVP-BEZ-235 treatment, followed by continued treatment with the drug; in vitro using CRC cell lines HCT 116, HT 29 and SW480, and in vivo using HCT 116 mouse xenografts [48]. This study demonstrated that prolonged treatment after radiotherapy with a dual PI3K/mTOR inhibitor NVP-BEZ235 led to enhanced apoptosis, inhibition of mTOR signalling, prolonged inhibition of cellular viability and disruption of DSB repair pathways.
Prevo et al. and Djuzenova et al. evaluated the role of PI3K/AKT/DNA-PK inhibitor PI-103 on radiosensitising HCT116, DLD-1 and SW480, SW48 CRC cell lines respectively [23,49]. Djuzenova et al. demonstrated that PI-103 effectively radiosensitised CRC cell lines in the presence of the Heat Shock Protein 90 (HSP-90) inhibitor NVP-AUY922, when both drugs were added three hours before irradiation and continued for 24 h after [49]. A reduction in cell proliferation as well as a rise in level of γ-H2AX was observed. The latter indicating impaired and prolonged DNA DSB repair. The authors also found that any earlier treatment with PI-103 up to 24 h before radiotherapy led to a decline of the radiosensitising effect. The authors concluded that the activation of the MEK/ERK pathway following prolonged inhibition circumvented the PI3K/AKT/mTOR pathway resulting in radioresistance. Prevo et al. demonstrated radiosensitisation independent of EGFR overexpression, RAS mutation status and PI3K overexpression in CRC cell lines treated with PI-103 [23]. Reduced pAKT levels, increased G2-M phase delay, delayed DNA DSB repair and cell proliferation contributed to radiosensitisation. The role of isolated mTOR inhibition with everolimus as a radiosensitiser was evaluated by Manegold et al. using CT-26 murine CRC cell lines and an in vivo xenograft mouse model [50]. The authors found that the decreased expression of vascular endothelial growth factor (VEGF) following mTOR inhibition, and the consequent antiangiogenic effects on the tumour microvasculature, was the underlying mechanism of radiosensitivity. This is one of the limited number of studies demonstrating the cancer radiosensitisation following isolated mTOR inhibition.
Two early Phase I/II clinical trials evaluated the role of mTOR inhibitors in conjunction with SCRT in rectal cancer patients (Table 5). Buijsen et al. treated patients with rapamycin one week before and during radiotherapy [51]. The study evaluated the changes in tumour perfusion status and treatment feasibility. This study included thirteen patients in phase I and identified a maximum tolerated dose (MTD) of 6 mg. In phase II, 31 patients were treated with 6 mg of rapamycin and radiotherapy. A higher rate of post-operative complications were observed in phase-I. Consequently, in phase II, surgery was performed after 6 weeks as opposed to 3 days following treatment in phase I. The authors concluded that the combination of SCRT and rapamycin was feasible and safe when using the phase II treatment regimen. There was a significant reduction in tumour metabolic activity on positron emission tomography (PET) scans following treatment (p < 0.05). However, no significant increase in pCR was observed in this small sample sized study. Gelsomino et al. also treated 12 patients with 5-FU, radiotherapy and mTOR inhibitor everolimus in another Phase I/II clinical trial [52]. Everolimus was administered 14 days prior to the start of chemoradiotherapy and continued throughout the four-week course of 5FU and radiotherapy. The authors identified a MTD of 10 mg. No significant increase in pCR rates was observed. However, the authors demonstrated that the combination of long course chemoradiotherapy and everolimus was feasible. Further clinical trials are warranted to evaluate the role of mTOR inhibitors and dual pathway inhibitors in improving pCR rates in rectal cancer patients.

Adenocarcinoma of the Prostate
Radiotherapy alone or in combination with chemotherapy and surgery is used in the treatment of localised as well as locally advanced PC respectively [8]. However, PC may recur in approximately 39% of patients who receive radiotherapy for localised disease after a median of 3 years, with an overall mortality rate of 23% at 5 years [53]. In patients with locally advanced PC treated with radiotherapy and docetaxel chemotherapy, 22.7% had a recurrence by 10 years, and the overall 5year mortality was 41.9% [54]. Recurrence following radiotherapy in local or metastatic PC affects a significant proportion of patients and can be challenging to treat [55]. This is due to the clonal selection and propagation of radioresistant PC cells within these recurrent tumours. There is scope for improving response to radiotherapy in recurrent PC as well as improving response to index radiotherapy treatment in PC patients with targeted therapy.
Chang et al. evaluated the role of class-I PI3K inhibitor (BKM120), mTORC1 inhibitor (rapamycin), dual class-I PI3K and mTORC1/C2 inhibitors (NVP-BEZ235, PI-103 [also a DNA-PK] inhibitor) in enhancing response to radiotherapy using three radioresistant PC cell lines [56]. The results demonstrated that dual pathway component inhibitors were superior to single component inhibitors in radiosensitising PC cell lines. There was increased apoptosis, autophagy and DNA DSB following the use of dual inhibitors. The authors also observed decreased expression of proteins involved in cell cycle checkpoint inactivation and DNA repair (NHEJ as well as HDR). Potiron et al. and Zhu et al. also demonstrated that NVP-BEZ235 can radiosensitise the PC-3 cell line [57,58].
Potiron et al. also tested the drug on DU145 PC cell line and evaluated the effects of the dual pathway inhibitor on both cell lines under hypoxic and normoxic conditions. In their study, dual inhibition of the pathway successfully radiosensitised both cell lines independent of oxygen concentration or PTEN mutation status [57]. All three studies demonstrated G2-Metaphase (G2-M) arrest within the cell cycle when treated with dual inhibitors.
Diaz et al. studied the effects of isolated AKT inhibitor Palomid 529 on NCI-60 cell line panel in vitro and in vivo using PC-3 xenograft mice [59]. The cell lines were more sensitive to radiotherapy following treatment with the AKT inhibitor. There were reduced levels of pAKT, VEGF and matrix metalloproteinase (MMP)-2, 9. Decreased levels of inhibitor of differentiation/DNA binding protein 1 (Id-1) and BCL-2/BAX protein ratio was also noted. The above proteins are involved in DNA DSB repair, cell survival and proliferation. Further mouse model work demonstrated greater tumour shrinkage (77.4%) after treatment with Palomid 529 and 6 Gy radiotherapy compared to single treatments-(Palomid 529 only-42.9% tumour shrinkage, radiotherapy only 53% tumour shrinkage). There was inhibited cell proliferation and increased apoptosis within the tumour xenograft. Dumont et al. treated mouse xenografts containing PC-3 cells with radioisotope (77Lu) tagged gastrin releasing peptide receptor (GRPr) antagonist, RM2 and mTORC1 inhibitor rapamycin [60]. Combined treatment with 177Lu-RM2 and rapamycin yielded the highest xenograft median survival (82 days), compared to 26 days for untreated mice, and 62 days for radioisotope therapy with GRPr antagonist treatment only group.
A phase-I clinical trial of 15 patients with locally advanced PC conducted by Azria et al. demonstrated that combined treatment with mTORC1 inhibitor everolimus, hormonal therapy and radiotherapy was feasible [61]. The MTD of everolimus was 7.5 mg/day. However, given the frequency of minor side effects, the authors recommend a treatment dose of 5 mg/day for future studies. In a separate phase-I clinical trial, Narayan et al. used radiotherapy and everolimus to treat patients with a biochemical recurrence of PC following surgery [62]. The authors concluded that this combination was feasible in patients with recurrent PC and that everolimus was well tolerated at doses below 10 mg/day.  [64]. Research has shown that there are significant variations in the oxygen delivered to the different regions of a solid tumour [67]. Subpopulations of cells within a tumour exposed to different oxygen levels exhibit variable chemoradiotherapy sensitivity [64,68]. The drug radiosensitised hypoxic, normoxic and re-oxygenated cells to the same extent. There was induction of autophagy, delayed DNA DSB repair, inhibited PI3K/mTOR signalling after treatment with the dual inhibitor and radiotherapy, independent of oxygenation status. These findings suggest that pathway inhibition could potentially radiosensitise the heterogeneous population of cells within a tumour.
Miyasaka et al. demonstrated that TP53 mutation status and PI3K pathway activation contributed to radiotherapy resistance [65]. Nevertheless, the authors successfully radiosensitised endometrial cancer cells containing TP53 and PI3K mutations using the dual PI3K/mTOR inhibitor NVP-BEZ235. The drug in combination with radiotherapy successfully inhibited PI3K/AKT/mTOR signalling which led to reduced cell proliferation and reduced expression of the hypoxia Inducing Factor 1-α (HIF1-α)/VEGF pathway proteins. HIF1-α silencing with small interfering RNA (siRNA) enhanced radiosensitivity further by reducing the sub-G1 cell population. The authors demonstrated the significant relationship between the PI3K/AKT/mTOR pathway and the HIF-α/VEGF pathway in radioresistance, and identified VEGF, HIF-α as potential targets for radiosensitising therapies. Approximately half a million patients are diagnosed annually with head and neck (H&N) cancer worldwide [69]. Over 90% of H&N cancer is SCC, originating anywhere along the nasal, para-nasal sinuses, oral cavity or pharyngeal mucosa [70]. Radiotherapy is used in the treatment of H&N SCC in combination with surgery and chemotherapy in the neoadjuvant or adjuvant setting [71]. Prognosis following treatment depends on tumour site and stage [72].
We identified several studies evaluating the role of PI3K/AKT/mTOR pathway inhibition radiosensitising H&N SCC ( Table 2). Seven of these studies demonstrated the efficacy of dual PI3K and mTOR inhibitors in radiosensitising H&N SCC cell lines [23,37,[73][74][75][76][77]. Five of these seven studies Yu et al. compared the dual inhibitor NVP-BEZ235 against isolated mTOR inhibitors (everolimus and AZD2014) in vitro using H&N SCC cell lines and in vivo using a xenograft model [37]. In this study, dual inhibition was superior to isolated mTOR inhibition, at radiosensitising H&N cell lines. Furthermore, a statistically significant reduction in xenograft tumour size was observed following treatment with a dual inhibitor and radiotherapy (p = 0.000017). Impaired AKT/mTOR signalling and G1 phase arrest contributed to radiosensitivity in vitro and in vivo in the dual inhibitor group. Bozec et al. also explored the role of isolated mTOR inhibition in radiosensitising H&N SCC CAL33 mouse xenografts [78]. Treatment with mTOR inhibitor and radiotherapy failed to demonstrate a significant reduction in xenograft tumour size over the control group. However, when combined with traditional chemotherapy, a VEGFA inhibitor (bevacizumab) and radiotherapy, there was significant anti-proliferative effects and a reduction in xenograft tumour size. Isolated mTOR inhibition is therefore unlikely to radiosensitise H&N SCC cells.
Two early clinical trials evaluating pathway inhibitors as radiosensitisers are currently in progress. A Phase Ib study (ClinicalTrials.gov identifier, NCT02113878) is evaluating the role of BKM120 (PI3K Inhibitor) in combination with cisplatin and radiotherapy in high risk locally advanced H&N SCC. Another Phase I trial is evaluating the feasibility of radiotherapy with everolimus and cisplatin for patients with head and neck cancer (ClinicalTrials.gov identifier, NCT00858663). The results from these trials are pending. The combination of Z-DEVD and RAD001 more potently radiosensitised H460 cells than individual treatment alone. Increased radiosensitisation predominantly through enhanced autophagy

Non-Small Cell Lung Cancer
NSCLC which accounts for over 85% of lung cancer is the leading cause of cancer related death worldwide and may comprise of SCC, adenocarcinoma or other rarer histological subtypes [84]. Radiotherapy is used in the treatment of NSCLC with curative intent, palliative intent or as adjuvant treatment. Outcomes following radiotherapy for NSCLC remains poor even in stage I disease [85]. Numerous studies have explored the role of PI3K/AKT/mTOR pathway inhibitors in radiosensitising NSCLC (Table 2).
Kim et al. (1) and Konstantinidou et al. tested dual inhibitor NVP-BEZ235 in vitro using NSCLC cell lines, and in vivo using xenograft mice [36,80]. Kim et al. (1) reports a statistically significant reduction in the in vitro NSCLC cell viability following treatment with the dual inhibitor (p = 0.01) [80]. The two studies demonstrated radiosensitising effects through increased autophagy, cell proliferation and cell cycle arrest at G1 phase. There was a significant delay in tumour growth observed in the treated xenografts. Additionally, Konstantinidou et al. tried LY294002 with rapamycin, as a dual pathway inhibition treatment [36]. LY294002 is a potent PI3K inhibitor which also weakly inhibits other PI3K-like kinases such as mTOR, ATM and DNA-PK [86]. Using this combination, the authors observed radiosensitising effects which were comparable to NVP-BEZ235 treatment. Toulany et al. also used PI3K, mTOR and DNA-PKcs inhibitor PI-103, and MEK inhibitor PD98059 on an in vitro cell line and in vivo xenograft experiment [38]. The authors demonstrated radiosensitising effects through inhibited DNA DSB repair through suppressed DNA-PKcs activity [38]. In this study, short term PI-103 only treatment led to radiotherapy sensitivity, but long-term treatment for 24 h led to MEK/ERK dependent activation of the PI3K/AKT/mTOR pathway, resulting in radioresistance.  3) found that isolated mTOR inhibition did not radiosensitise NSCLC [39,81,83]. However, Holler et al. demonstrated that when a mTOR inhibitor is combined with an AKT inhibitor (MK2206), significant radiosensitising effects may be observed in NSCLC cell lines [39]. Deutsch et al. showed feasibility of everolimus with radiotherapy in NSCLC patients in an early Phase I clinical trial [87]. Two further clinical trials are currently in progress evaluating everolimus (ClinicalTrials.gov identifier, NCT00374140) and BKM120 (ClinicalTrials.gov identifier, NCT02128724) as radiosensitisers in NSCLC patients.

Sensitising Glioblastoma Multiforme to Radiotherapy
GBM is the most common malignant central nervous system tumour in adults and is routinely treated with chemoradiotherapy with or without surgery [88]. Survival remains poor with studies reporting a median survival of just 14.6 months following chemoradiotherapy [89]. The majority of research exploring the radiosensitising properties of PI3K/AKT/mTOR pathway inhibitors comes from research using GBM cell lines and from a few early clinical trials in GBM patients (Tables 3 and  5 [90,93]. The most frequently used dual PI3K/mTOR inhibitor was NVP-BEZ235. The drug consistently demonstrated enhanced radiosensitisation of GBM cell lines and in two in vivo mouse xenograft models (Table 3). Two studies using PI3K/mTOR/DNA-PK inhibitor PI-103 also demonstrated enhanced radiosensitisation of GBM cell lines in vitro [49,94]. Dual inhibitors were found to lead to autophagy, apoptosis, G2/M cell cycle arrest and impaired DNA DSB repair. Li et al., Kao et al. and Nakamura et al. tested LY294002 (PI3K and PI3K-like kinase inhibitor) in GBM cell lines [95][96][97]. These studies showed enhanced radiosensitisation of GBM cell lines following treatment with LY294002. Rapamycin had no effect on cell lines However, the drug significantly enhanced the response to radiotherapy on xenograft and spheroids following treatment Choi et al., Nakamura et al. and Eshleman et al. tested mTOR inhibitor rapamycin in vitro [94,97,100]. All three studies showed no significant radiosensitising effect of this drug on GBM cell lines. However, Eshelman et al. demonstrated rapamycin enhances radiotherapy sensitivity of U87 spheroids and U87 xenografts [100]. Shinohara et al. studied the effects of mTOR inhibitor everolimus on glioma cells and endothelial cells using in vitro cell line and in vivo mouse model experiments [99]. This study demonstrated the role of mTOR inhibition on impaired tumour blood supply thus leading to inhibition of tumour growth.
A phase-I trial by Sarkaraia et al. demonstrated that everolimus can be safely administered to GBM patients in conjunction with a chemoradiotherapy treatment regimen comprising of temozolomide and radiotherapy [101]. The authors recommend an everolimus dose of 70 mg/week (10 mg/day) for future studies. However, a phase-II trial by Ma et al. demonstrated that the addition of 70 mg/week of everolimus to standard therapy led to a moderate increase in toxicity with no appreciable benefit in overall survival (p = 0.28) or progression free survival (p = 0.15) [102]. Another phase-II trial by Chinnaiyan et al. which included 171 patients with GBM found that the addition of everolimus (10 mg/day) to standard therapy led to a significant reduction in overall survival (p = 0.008) [103]. Wen et al. conducted a Phase I clinical trial using dual PI3K/mTOR inhibitor XL 765 [104]. The study demonstrated feasibility of the drug in combination with current standard treatment. There was moderate inhibition of the pathway components noted on skin biopsies from patients. The MTD reported by the authors was 90 mg/day.

Sensitising Other Tumour Cells and Endothelial Cells to Radiotherapy
The radiosensitising properties of PI3K/AKT/mTOR inhibitors have also been demonstrated in soft tissue sarcoma, bladder transitional cell carcinoma, hepatocellular carcinoma, metastatic renal cell carcinoma Burkitt's lymphoma (Table 4), SCC of the cervix (Table 2) and pancreatic cancer (Table  1). Fokas et al. (2) radiosensitised soft tissue sarcoma cell lines and xenograft mice using NVP-BEZ235 [76]. This dual PI3K/mTOR inhibitor led to significant changes within the tumour microenvironment by prolonging the time taken for the normalisation of tumour vascularity. The authors also compared the efficacy of isolated PI3K inhibition using BKM120 against dual inhibition with NVP-BEZ235. The dual inhibitor was a more potent radiosensitiser and led to impaired remodelling of the tumour vasculature [76]. also demonstrated that mTOR inhibition can effectively radiosensitise cervical SCC cell lines in vitro, at radiotherapy doses as low as 2 Gy [79]. Other types of cancer that may be radiosensitised by pathway inhibition include renal cell carcinoma, hepatocellular carcinoma and Burkitt's lymphoma [106][107][108]. Given that radiotherapy is less frequently used in the treatment of the aforementioned cancers, the clinical significance of this research is somewhat limited.
Several studies have also evaluated the role of pathway inhibition in radiosensitising endothelial cells [50,77,99,105]. Resistance to radiotherapy may be induced by tumour revascularisation and inhibiting cellular hypoxia [14]. The pathway inhibitors which sensitise tumour as well as endothelial cells to radiotherapy, suppress the revascularisation of the tumour resulting in an enhanced response to radiotherapy.

Overcoming Radiotherapy Resistance by Pathway Blockade
Research has identified several biological processes which contribute to the radiosensitisation of cancer, following PI3K/AKT/mTOR pathway inhibition. Most studies demonstrated a reduction in cell proliferation following effective pathway inhibition and irradiation (Tables 1-4). The majority also reported impaired DNA DSB repair following treatment with pathway inhibitors, characterised by the persistence of excess λ-H2AX foci [18,23,39,49,64,[73][74][75]. Several studies demonstrated reduced expression of DNA-PKcs, the protein catalytic subunit of DNA-PK [18,38,66,90,93]. DNA-PK plays an important role in DNA DSB repair [93]. There was enhanced apoptosis and autophagy [56,80,81,83]. Cell cycle progression was inhibited with numerous studies demonstrating cell cycle arrest at G2-M phase following treatment with pathway inhibitors and radiotherapy [56][57][58]66,73,77,78,91,93]. Several others showed cell cycle arrest at G1 phase [23,36,37,65]. The pathway inhibitors successfully led to cell cycle checkpoint activation by inhibiting the inhibitors of cell cycle checkpoint proteins such as TP53. Pathway inhibition also led to significant changes within the tumour microenvironment. There was impaired tumour revascularisation likely through the inhibition of HIF-α and VEGF [65,78]. There was inhibited cell migration [48,73,75]. Pathway inhibitors demonstrated the ability to radiosensitise cells growing in the most hypoxic regions of a tumour, which harbour extreme pro-survival and anti-apoptotic mutations [57,64].
Single pathway inhibition can be ineffective given that the PI3K/AKT/mTOR pathway has numerous triggers ( Figure 1) and direct activators of its downstream components, circumventing the effects of this class of inhibitor. Isolated inhibition of mTOR or PI3K likely leads to activation of AKT (the main effector of the pathway) by phosphorylation via the EGFR/RAS/RAF/MEK/ERK pathway [38,49]. AKT inhibition in isolation is also unlikely to be effective because downstream components of the pathway (e.g., mTOR) may be activated via alternative biological pathways [39]. This indirect activation of the PI3K/AKT/mTOR pathway is a crucial mechanism of radiotherapy resistance [38,39,109]. A minimum two-point inhibition of key pathway components is therefore essential to ensure effective downstream inhibition. This may explain the higher efficacy of dual inhibitors such as NVP-BEZ235. The latter was the most frequently used dual PI3K and mTOR inhibitor identified within this type of research. The drug consistently demonstrated radiosensitising effects on in vitro cell line and in vivo xenograft experiments; in PC, CRC, GBM, H&N SCC, NSCLC, breast and gynaecological cancer. Other potential dual or pan pathway inhibitor combinations (e.g., AKT/mTOR or PI3K/mTOR and DNA-PK inhibition) have also been tried [38,39]. However, the current scientific evidence available is limited for making discernible conclusions regarding the radiosensitising potential of these alternative pathway inhibitor combinations.

Class of inhibitor Drug No. of studies utilising drug as a radiosensitiser
In-vitro/In-vivo Clinical trials

Limitations, Challenges and Potential for Clinical Translation
There is significant variation in the methodology used in this research. There were multiple different cell lines being used even within a given tumour type. The majority of this research in rectal cancer utilised cell lines or xenograft models derived from colon cancer [18,23,[48][49][50]. However, radiotherapy is only used in the treatment of rectal cancer and not in colon cancer. This remains a significant limitation for rectal cancer research. The advent of novel in vitro experimental models such as three-dimensional primary cell cultures (organoids) using primary rectal tumour tissue, will undoubtedly contribute to novel insights [110]. Variations in radiotherapy regimes used across the different studies was also a significant limitation. Standardisation of radiotherapy regimes preferably at a clinically relevant dose for that tumour type will ensure greater comparability across future research. The duration of treatment with pathway inhibitors varied considerably across the different studies. Even though most authors emphasised the need for pathway inhibitors to be commenced prior to traditional chemoradiotherapy, continued during and preferably for a period thereafter; there remains a lack of consensus on the exact duration of treatment. There was also significant variations in the outcome measures used within this in vitro and in vivo research. These factors render comparability across different studies extremely difficult.
The challenges of toxicity, the complexity of pathway inhibitor combinations used, as well as the sheer volume of different inhibitors that are available might be hindering translational research in this field (Table 5, Figure 2). The dilemma facing the clinical researcher reviewing the scientific evidence is which inhibitor they must choose for an early clinical trial. For example, we identified at least four different dual PI3K and mTOR inhibitors of this pathway that have been used across the different tumour types. This is a significant limitation in research in this area. Therefore, scientific research should focus on a select few pathway inhibitors and pursue them to clinical trials to ensure effective clinical translation.
A limited number of pharmacological inhibitors of this pathway have undergone early clinical trials in combination with radiotherapy ( Table 5). Several of these early clinical trials used mTOR inhibitors. Frequent side effects of the most commonly used mTOR inhibitor everolimus included mucositis, cutaneous rash and diarrhoea [61,62]. mTOR inhibitors were found to be safe in combination with chemotherapy and/or radiotherapy in cervical cancer, PC and rectal cancer. However, a Phase-II clinical trial from 2018 in GBM patients found overall survival in the treatment group with everolimus to be significantly poorer compared to the standard treatment arm [103]. Most clinical trials using mTOR inhibitors and radiotherapy evaluating pCR rates as an outcome have not shown a clinically significant benefit [51,52]. However, this result should be interpreted with caution given the smaller sample sizes observed in these early clinical trials. Therefore, more phase-II/III multi-arm randomised control trials are needed to make clinically relevant conclusions to this regard.
Only one clinical trial involved a dual PI3K and mTOR inhibitor (Xl-765). The common adverse effects observed with this drug when combined with radiotherapy included nausea, fatigue, thrombocytopenia and diarrhoea [104]. NVP-BEZ235 was the most commonly used dual inhibitor identified in this review. The drug has already been safely used in cancer patients in early clinical trials, as the sole treatment or in combination with other chemotherapy [111]. However, it has yet to undergo an early clinical trial in combination with radiotherapy. The growing evidence from scientific research on the efficacy of dual PI3K and mTOR inhibitors as a radiosensitiser should be the primary focus of current and future scientific research in this area. Clinical trials involving these dual pathway inhibitors and radiotherapy remain scarce. There remains a need for early phase I/II clinical trials using dual PI3K/inhibitors (e.g., NVP-BEZ235) in conjunction with radiotherapy (with or without chemotherapy) in rectal cancer, PC, breast cancer, GBM, H&N SCC and NSCL. Neoadjuvant radiotherapy and rapamycin feasible with no significant increase in toxicity (MTD = 6mg) No increase in post-operative complications rates after delaying duration to surgery to 6 weeks Significant reduction in tumour metabolic activity but pCR rate was still 10% within small sample sized study

Conclusions
The PI3K/AKT/mTOR pathway has been identified as a potential radiotherapy resistance inducing pathway in cancer. There is abundant evidence in the current scientific literature which supports the notion that pharmacological inhibition of the PI3K/AKT/mTOR pathway components can radiosensitise cancer. Changing the tumour microenvironment, disrupting DNA DSB repair, inhibiting cell growth, cell proliferation and promoting cell death are a few of the mechanisms by which radiosensitisation may occur. Many pharmacological inhibitors of this pathway have undergone safety and feasibility studies in humans. However, there is limited clinical research available to make reliable conclusions regarding the potency and safety of these inhibitors in radiosensitising tumours in patients.
Research has shown that isolated mTOR or PI3K inhibition can radiosensitise tumours in vitro and in vivo. However, these drugs failed to demonstrate significant clinical efficacy during preliminary trials. Consequently, the attention of researchers has shifted towards the radiosensitising properties of dual PI3K and mTOR inhibitors. Recent scientific research has consistently demonstrated that dual pathway component inhibition to be far superior at radiosensitising different tumour cell types. Therefore, future laboratory as well as clinical research in this area should focus on the translational potential of dual PI3K and mTOR inhibitors as radiosensitisers in cancer patients.
Author Contributions: The corresponding author was responsible for the article concept, literature review writing the main body of the text, designing tables and figures. All remaining authors equally contributed in providing assistance, supervision and guidance in these tasks. All authors have read and agreed to the published version of the manuscript.

Conflicts of Interest:
The authors declare no conflict of interest.