Abstract
Objective
Head and neck cancers (HNC) are known for their repopulation ability driven by cancer stem cells (CSCs). While a small fraction of CSCs proliferates, there are quiescent CSCs that are long‐lived and reside outside the cell cycle. Recruitment of quiescent CSCs into the cycle occurs as a response to cell loss and their proliferation may lead to treatment failure. Therefore, CSCs require a more targeted approach to be destroyed. An agent that sensitizes CSC response to treatment is all‐trans‐retinoic acid (ATRA). The aim of this work is to assess the impact of ATRA combined with radiotherapy on HNC and to analyse the interplay between these agents and cell recruitment.
Methods
An in silico model is employed to grow a HNC consisting of all cancer cell lineages, with biologically valid kinetic and dynamic parameters. The fate of both cycling and quiescent cancer stem cells is assessed. The Linear Quadratic model is used to simulate radiotherapy, while cellular recruitment and the effects of ATRA on cancer stem cells are modelled based on literature data.
Results
A Dose Enhancement Factor (DEF) was determined in order to undertake a quantitative assessment of the effect of ATRA on tumour control. Without recruitment, DEF for the tumour population is 1.06, indicating a slight radiosensitizing effect. Yet, when CSCs are being recruited, the dose enhancement factor is significantly greater (DEF = 1.89). Radiation‐induced cell arrest and CSC sensitization by ATRA significantly decreases the dose required for CSC eradication in the cycling population. However, the tumour as a whole is not notably affected as the quiescent cells appear to dictate the shape of the survival curve.
Conclusions
The model shows that ATRA exhibits a powerful effect on CSCs when combined with radiotherapy. However, the more radioresistant quiescent cell population should not be ignored, as it can be a potential threat to treatment outcome when cells are recruited into the cell cycle.
1. INTRODUCTION
1.1. Treatment‐related challenges in head and neck cancer
In head and neck cancer (HNC) accelerated tumour repopulation is a major cause of treatment failure, reason why prolonged radiotherapy schedules should be avoided. This is valid particularly for tumours comprising of large cancer stem cell (CSC) compartment that can fast regrow the tumour as a response to treatment‐induced cell kill.
To overcome repopulation and tumour hypoxia, which are the two main culprits for poor treatment outcome, the conventional radiotherapy schedule (2 Gy/fraction, 5 days a week, over 7 weeks) has been altered. Altered fractionation schedules that were successfully trialled are accelerated radiotherapy and hyperfractionated radiotherapy. Accelerated radiotherapy shortens the overall treatment time, while hyperfractionated radiotherapy is designed to overcome repopulation and promote reoxygenation in between fractions. Based on a meta‐analysis of head and neck cancer trials, hyperfractionated radiotherapy proved to be superior in tumour control to accelerated radiotherapy.1 Within hyperfractionated radiotherapy 2 or more small doses (<2 Gy) of radiation are administered week‐daily, which results in a larger overall delivered dose compared to conventional radiotherapy.
Tumour repopulation is caused by cancer stem cells which have the ability to proliferate indefinitely,2 reason why they require selective targeting. In HNC, cancer stem cells have been first identified and reported by Prince et al,3 who has isolated a cellular subpopulation that exhibited stem‐like properties. To date, there are few quantitative studies reporting on the percentage of CSC in head and neck cancer, with significant differences among the studied head and neck cell lines. Thus, in the experiment reported by Harper et al4 the proportion of CSC in CaLH3 cell line was found to be 12.3%. In another experiment, Tang et al5 has indicated that the CSC proportion in various HNC cell lines ranged between 1.7% and 13.5%. These variations demand more quantitative and qualitative studies to explain the differences.
1.2. Properties of cancer stem cells and their importance for the current study
Cancer stem cells have several specific properties that make them immortal and resistant to therapy. Table 1 presents the most common properties, based on experimental evidence. Resistance to treatment is multifactorial and includes (i) the ability of CSC to efficiently repair damaged DNA, (ii) the capacity to divide symmetrically (ie, symmetric self‐renewal of CSC in mitosis resulting in two CSCs), which contributes to tumour repopulation, (iii) the preference of CSCs to reside in specific microenvironmental niches in order to conserve their status and (iv) the ability of the quiescent CSCs to be recruited into the cell cycle.
Table 1.
Properties and behaviour of cancer stem cells based on experimental evidence
| Cancer stem cell properties based on experimental findings | References |
|---|---|
| Are able to generate all heterogeneous lineages of the original tumour | Al Hajj et al 8 |
| Can recreate themselves by symmetrical division | Morrison et al9 |
| Can be recruited from their niche into the proliferating pool | Vlashi et al10 |
| Are long‐lived and have the capacity to proliferate indeterminately | Moore et al2 |
| Exhibit higher resistance to treatment than non‐stem cancer cells due to enhanced DNA repair | Moore et al2 |
| Are able to repopulate a tumour as a response to treatment‐induced cell kill | Koukourakis et al11 |
| Can be active and invasive (migratory CSCs) or quiescent and non‐invasive (stationary CSCs) | Geissler et al12 |
| Preferentially reside in specific microenvironmental niches within the tumour | Peitzsch et al13 |
| Are highly dynamic and exhibit potential for cellular plasticity that enables shifting from CSC to non‐CSC state and vice versa | Cabrera et al14 |
It was shown that activation of Notch signalling that regulates cell fate, can recruit quiescent stem cells into the cell cycle.6 Fractionated radiotherapy was demonstrated to promote cell recruitment, increasing the cycling CSC population.7
All the above properties of CSCs have a critical role in tumour response to therapy and therefore should be investigated and addressed accordingly. The present study tackles several of these properties in order to establish means of treatment optimization in HNC.
1.3. Targeted therapies for cancer stem cells
Although altered fractionation radiotherapy is a powerful tool in the management of head and neck cancer, CSCs require targeted therapy in order to be destroyed. Several CSC‐targeting pathways have been investigated so far and several others are under investigation.
The focus of this work is on all‐trans‐retinoic acid (ATRA) which is an agent that exhibits cell cycle effects by sensitizing CSC response to treatment. ATRA is a member of the retinoid family and has a potent effect on cell growth, differentiation and apoptosis.15, 16, 17 Retinoids have the ability to influence multiple signalling pathways that are involved in stem cell preservation.15 As demonstrated by in vitro and in vivo pre‐clinical studies, ATRA exhibits powerful effects on CSCs by inducing cell cycle arrest due to the complexity of DNA damage and also apoptosis.16, 17 The same studies have proven the ATRA‐induced differentiation, associated with downregulation of the Wnt pathway, which is a central mechanism for controlling malignant transformation. According to Bertrand et al, ATRA combined with radiotherapy resulted in a decrease in the surviving fraction among the head and neck cancer cell population.17 Furthermore, Zhang et al have shown that ATRA can efficiently inhibit telomerase activity in oral squamous cell carcinoma cell lines that leads to the induction of growth arrest in these tumours.18
1.4. The aim and design of the current work
The current work is based on the above described CSC properties that have been proven to influence response to treatment, particularly in head and neck carcinoma patients. The aim of this study is multifold:
to investigate using modelling tools the effect of cell recruitment on treatment outcome, as a function of the recruited cell type (ie, CSC or differentiated);
to implement the mechanisms of ATRA combined with radiotherapy and to evaluate treatment response for the tumour as a whole;
to assess the interplay between the combined effect of ATRA/radiotherapy and cell recruitment.
To fulfil this goal, a Monte Carlo technique has been adopted. The choice of a stochastic approach is justified by the probabilistic nature of all main phenomena occurring during tumour growth and in response to treatment. The initiation of a malignant transformation, the cellular phenotype, and the hit‐and‐kill effect induced by radiation, are all governed by stochastic processes.
The properties of cancer stem cells have been incorporated in a head and neck tumour model that was grown using biologically realistic parameters. The virtual tumour has been then treated with radiotherapy and ATRA, mimicking the effects of these agents on each individual cell.
2. METHODS
An in silico model based on Monte Carlo techniques was developed to grow a HNC consisting of all lineages of cancer cells. Details of the tumour growth model together with a comprehensive flow chart are presented in a previous work19 and summarized here. The parameters characterizing tumour growth and its behaviour during treatment are listed in Table 2.
Table 2.
Parameters of tumour growth and behaviour during therapy
| Parameters | Modelled values | Literature data/reference |
|---|---|---|
| Tumour growth‐related parameters | ||
| Mean cell cycle time (range) | 33 h | 33 h (20‐60 h)20 |
| Length (proportions) of cell cycle phases before radiotherapy | M:7%; G1:40%; S:30%; G2:23% | S = 1/3 of cell cycle, M = 1‐2 h21 |
| Cell loss factor | 85% | 85%20 |
| Volume doubling time | 52 d | 45 d (33‐150 d)22 |
| Labelling index | 4.7% | 8% (1.2‐30)23 |
| Cell division rate (24 h) | 1.3% | Model‐derived parameters based on the above tumour growth parameters |
| Pre‐treatment probability of CSC symmetrical division | 1.9% | |
| Pre‐treatment percentage of CSCs | 5.9% | |
| Cell phenotype ratio CSC:D:Q | 5.9:7.9:86.2 | |
| Radiotherapy‐related parameters (linear quadratic model parameters) | ||
| Average surviving fraction after 2 Gy (SF2) | 54% | 54%21 |
| Quantification of tumour sensitivity to fractionated radiation (α/β) | 10 | 1024 |
| α value | 0.35 | 0.3524 |
The computational model includes the following modules, as shown in Figure 1:
Tumour development: The virtual tumour is grown from a single stem‐like cell. Cancer stem cells (CSC) have the ability to divide both symmetrically and asymmetrically. Through symmetrical division, cancer stem cells that undergo mitosis create two CSCs, while in asymmetrical division cancer stem cells will self‐replicate and also create a non‐stem cancer cell. Cancer stem cells and differentiated cells (D) that pass through the cell cycle are proliferative cells. Differentiated cells have limited proliferative ability (2 generations of cells) whereas cancer stem cells exhibit unlimited proliferative potential. A large percentage of cells within the tumour are characterized by quiescence, meaning that they are resting in the G0 phase outside the cell cycle. Cells that are quiescent have the ability to proliferate if recruited into the cell cycle as a response to treatment‐induced cell loss. The pre‐treatment percentages of the three cell types have been determined through multiple iterations in order to allow a biologically realistic tumour growth and composition, and they are the following: 5.9% cancer stem cells, 7.9% differentiated cells and 86.2% quiescent cells. This is in agreement with the scientific literature that shows that over 80% of cells reside outside the cell cycle, in the quiescent phase.21 The virtual tumour is modelled to be moderately hypoxic, with an average pO2 value of 6 mm Hg. This consideration is due to the fact that head and neck cancers are known to be generally hypoxic, a property that increases radioresistance.
ATRA: The effects of ATRA (differentiation, cell arrest and apoptosis) are modelled based on scientific literature data. In a previous work, we have illustrated the individual effects of ATRA on CSC only25 while in the current study we are focusing on the overall effect of ATRA on the tumour population as a whole (CSCs and non‐stem cancer cells). Differentiation is simulated by decreasing the initial probability of symmetrical division from 1.9% to 0.1%. The interplay between ATRA and radiotherapy is modelled based on the report of Bertrand et al17 whereby the pre‐radiation administration of ATRA enhanced apoptosis among the CSC cells arrested by radiation in the radiosensitive G2/M phase for up to 48 hour, with a peak of 40% cells being arrested after 24 hour following radiotherapy. Cell arrest in the G2 phase allows some of the cells with radiation‐caused DNA lesion to repair their sublethal damage. Cells that are not able to repair their DNA breaks undergo apoptosis. To simulate the experimental results caused by ATRA, we have modelled the effect rather than the drug dose administered to the cell population. In order to obtain the above‐mentioned effect, in the experiment ATRA was given to CSCs 7 days prior to radiotherapy, at a final concentration of 10 μmol L−1.
Radiotherapy: The Linear Quadratic model is employed to determine the surviving fractions after the administration of altered fractionated radiotherapy. As justified in the Introduction, altered fractionation schedules are superior to conventional radiotherapy. Thus, in this simulation we are considering the effect of hyperfractionated radiotherapy, where 84 Gy overall dose is administered in 1.2 Gy per fraction, twice a day, in 35 fractions. This dose is larger than the conventional radiotherapy dose (70 Gy) administered in daily fractions of 2 Gy each.
Recruitment: Recruitment is modelled using some of the experimental observation reported by Phillips et al on breast cancer‐initiating cells.7 They have shown that after 1 week of fractionated radiotherapy the percentage of cancer stem cells increased from 3.52% to 7.5%, which is more than double the initial percentage. Head and neck cancers are likely to exhibit similar properties given the fact that accelerated repopulation, also demonstrated in breast cancer, was first evidenced in squamous cell carcinomas of the head and neck.26 To extrapolate the findings for the current model, recruitment is simulated by the re‐entry of quiescent cells into the cell cycle which doubles the percentage of CSC over 1 week of fractionated radiotherapy. Two recruitment scenarios are modelled: one that considers recruitment among quiescent non‐CSC cells, and the second scenario, which is clinically more relevant, whereby recruitment occurs among quiescent CSC cells. Based on relatively recent experimental evidence mainly CSCs are getting recruited into the cell cycle.27 Our work modelled the ‘least possible’ scenario where recruitment still occurs but differentiated cells are triggered back into the cell cycle, as opposed to the ‘maximum possible’ scenario, with CSCs being recruited.
Figure 1.
Flow chart of computational modules
The balance reached by the different cell types during tumour development is severely impacted by treatment. As every action determines an opposite reaction, cell kill due to radiation triggers various mechanisms of repopulation. During treatment, the composition of the tumour changes in order to bring more proliferative cells into action to compensate cell loss. ATRA shows to have a powerful impact on tumour dynamics counteracting the self defence mechanisms displayed by the tumour.
3. RESULTS
3.1. Conventional vs hyperfractionated radiotherapy in a moderately hypoxic tumour
As discussed in the Introduction, the results of phased III randomized clinical trials have shown that hyperfractionated radiotherapy is superior to the conventionally fractionated radiotherapy regimen in head and neck cancers. The simulation on the virtual head and neck tumour of the two treatment regimens using the same clinical dosage and fractionation schedules as employed in the trials is represented in Figure 2 and the results are in accordance with the clinical findings. The trial1 showed a superior outcome in terms of locoregional control in patients undergoing hyperfractionated as compared to conventional radiotherapy. This outcome is translated in the model by the more efficient and timely cell kill offered by hyperfractionation compared to the much longer treatment time and dose required from conventional therapy to achieve the same effect. Given that the mentioned randomized trials did not stratify their patients based on hypoxia, the current model considered the ‘average’ patient with a moderately hypoxic tumour.
Figure 2.
Cell survival after conventional vs hyperfractionated radiotherapy in a moderately hypoxic head and neck tumour
In the paragraphs below we are presenting the effects of (i) recruitment as a single mechanism, (ii) ATRA combined with hyperfractionated radiotherapy and (iii) ATRA combined with hyperfractionated radiotherapy when considering the recruitment mechanism.
3.2. The model of cell recruitment and radiotherapy outcome
The behaviour and survival of three cell categories is studied: cancer stem cells (CSC), differentiated cells (D) and quiescent cells (Q). The modelled cell recruitment process considers either CSC or D cell recruitment.
The highly potent effect of CSCs on treatment outcome is well illustrated in Figure 3 and is in accordance with the properties described in Table 1, particularly repopulation and radioresistance.
Figure 3.
Cell surviving curves illustrating the effect of radiotherapy on the tumour population with and without cell recruitment. The cell recruitment process considers either CSC recruitment or D cell recruitment
The effect of proliferation and symmetrical division displayed by CSCs is intensified by the surplus CSC population brought into the cell cycle by the recruitment mechanism, leading to a significant increase in the required dose for tumour kill.
This radiation‐triggered cellular recruitment is a natural response of the tumour to outlive treatment through activation of its DNA repair capacity, symmetrical division and accelerated proliferation.
As an interesting result, recruitment of D cells leads to a lower overall dose required for tumour kill as compared to no recruitment. Once recruited, cells leave the safety net of the G0 phase, so subsequent doses of radiation find them in a more radiosensitive part of the cell cycle. This sensitization to radiation represents the silver lining of the recruitment mechanism.
3.3. The effect of ATRA/radiotherapy on CSC population and on the tumour as a whole without recruitment
We have shown in a previous study that ATRA exhibits powerful effects on cancer stem cells.25 In the mentioned study, the in silico model indicated that the cellular effects displayed by ATRA combined with radiotherapy lead to a more powerful outcome than radiotherapy alone. Namely, cell arrest induced by ATRA resulted in 14.4 Gy lower dose for the same tumour response as compared to radiotherapy‐alone, and the effect of apoptosis decreased the total dose needed for CSC eradication with a further 10.8 Gy. The mentioned study employed hyperfractionated radiotherapy delivering 1.2 Gy radiation dose twice daily, 5 days a week, over 7 weeks. The same treatment schedule is simulated in the current work.
When simulating ATRA combined with radiotherapy on the whole tumour population, it can be noted that the cellular effects induced by ATRA significantly decrease the total dose needed for CSC eradication in the cycling population, as compared to the results illustrated in Figure 3. However, the tumour population as a whole, is not notably affected, concluding that quiescent cells are the ones dictating the shape of the survival curve (Figure 4). The cell surviving curves for the total tumour population and the quiescent cells look nearly identical because of the semi‐logarithmical scale employed on the graph.
Figure 4.
The effect of ATRA combined with radiotherapy on all cell types: Q = quiescent, CSC = cancer stem cell; D = differentiated
Figure 4 illustrates the effect of ATRA combined with radiotherapy on the three cell categories studied, as well as on the tumour as a whole. Being a CSC‐targeting agent, ATRA has the most powerful effect on cancer stem cells. The response of differentiated cells shows a controllable population, however, this response is due to the radioresponsiveness of D cells and not due to the effect of ATRA. Quiescent cells that are more resistant to radiation than cycling cells, dictate the final shape of cell survival curve and are not significantly affected by the combined modality treatment.
3.4. The effect of ATRA/radiotherapy on CSC population and on the tumour as a whole with recruitment
As shown above, if there would be no recruitment, ATRA would have an insignificant effect on the tumour population as a whole, because of the high percentage of quiescent cells residing in the G0 phase outside the cell cycle. The question that arises is whether/to what extent can recruitment of quiescent cells influence tumour response when ATRA is employed. Two scenarios are considered here: (i) when recruitment occurs among quiescent non‐CSC cells, and (ii) recruitment occurs among quiescent CSC cells.
Figure 5 shows the trivial effect of ATRA on the tumour when D cells are triggered into the cycle by radiation. Given the fact that ATRA is a CSC‐targeting agent, this result is somewhat expected.
Figure 5.
Cell surviving curves illustrating the effect of ATRA combined with radiotherapy on the tumour population with and without cell recruitment
Nevertheless, ATRA shows a noteworthy effect on the CSC population, both the existing and newly recruited. ATRA combined with RT resulted in tumour eradication achieved after 36 days of treatment (the equivalent of 52.8 Gy) (Figure 5), whereas radiotherapy as a single agent was not able to overcome repopulation due to recruitment, thus the tumour could not be killed within the clinically set schedule (Figure 3).
A comparative analysis of Figures 3, 4 and 5 shows that recruitment on its own forces the CSC population into untreatable territory (Figure 3). Without recruitment, ATRA has a “magic bullet” effect on the CSCs (Figure 4), while Figure 5 shows the combined effect of the two opposing forces (recruitment and ATRA), illustrating that ATRA provides a valuable service bringing the CSC population under the treatment threshold of altered fractionated radiotherapy.
3.5. Quantitative evaluation of ATRA effects: dose enhancement factor
In order to undertake a quantitative assessment of ATRA effects on tumour control when combined with radiotherapy, the Dose Enhancement Factor (DEF) is determined as the ratio between the dose with radiotherapy and the dose with radiotherapy + ATRA, for the same biological effect (tumour control). Table 3 shows the calculated DEF values for the tumour population as a whole with and without recruitment, based on our model. The powerful effect of ATRA is illustrated by the DEF of 1.89 when CSCs are recruited into the cell cycle, are sensitized by ATRA and targeted by radiation. If no recruitment is considered, DEF is still larger than 1, showing a slight radiosensitizing effect of ATRA on the tumour population. The value of 1.06 is a result of the small fraction of cycling CSCs affected by ATRA. The low value of DEF obtained when D cells are recruited is expected, as ATRA is not affecting non‐stem cancer cells.
Table 3.
Dose enhancement factor of ATRA for the tumour population as a whole
| Treatment agent | DEF |
|---|---|
| Radiotherapy | 1.06 |
| Radiotherapy + ATRA | |
| Radiotherapy with D cell recruitment | 1.09 |
| Radiotherapy with D cell recruitment + ATRA | |
| Radiotherapy with CSC recruitment | 1.89 |
| Radiotherapy with CSC recruitment + ATRA |
4. DISCUSSION AND CONCLUSIONS
Computational models allow treatment simulations and evaluation of tumour behaviour without the expense and time needed for a clinical trial or extensive laboratory work. While by no means can they replace clinical studies, well‐designed in silico models have the advantage of parameter flexibility, reproducibility of experimental outcomes and timely delivery of results. Models can show specific trends, confirm pre‐clinical results, fill in the gaps where quantitative data is lacking. Furthermore, models can identify new research avenues that have not been tackled by experimental studies.
The aim of the current in silico model is to assess the impact of a CSC‐targeting agent (ATRA) combined with radiotherapy in the treatment of HNC, and to evaluate the effect of the combined modality treatment on the tumour population as a whole. We show that while ATRA has a powerful effect on CSCs, the overall behaviour of the tumour is dictated by the large percentage of quiescent cells. The focus therefore is on the quiescent population that can be triggered into the cell cycle by treatment.
Experimental studies provide evidence for cell recruitment, which is the process of re‐entry of quiescent cells into the mitotic cycle triggered by cell loss. Nevertheless, the literature lacks quantitative data concerning the recruited population as well as the quality of recruited cells (ie, cell type). In view of the above, it is critical for future experimental work to quantify recruitment in head and neck cancer in order to design treatments that account for such an effect.
Based on the model results, the following challenging questions need practical attention:
Should quiescent cells be targeted for eradication in the G0 phase?
Should quiescent cells be triggered into the cell cycle and targeted there?
Should quiescent cells be left in their current state?
As shown by our modelling work, the answers to the above questions very much depend on the recruited cell type. If differentiated cells are recruited, the overall outcome can be improved due to the higher radiosensitivity of cycling cells as compared to quiescent cells. While differentiated cells can still proliferate and contribute towards the overall population of cancer cells, they have limited proliferative ability as they undergo abortive division.28 Thus, differentiated cells will not be responsible for tumour regrowth. However, if through the recruitment process the cells re‐entering the mitotic cycle have stem‐like properties, the outcome is worsened due to the ability of cancer stem cells to proliferate indefinitely and divide symmetrically, thus increasing the population of cancer stem cells. Being more resistant to radiation than non‐CSCs, these cells can regrow the tumour and lead to treatment failure. Cancer stem cell targeting therapies are needed to assist in tumour control. In the current work, ATRA as a CSC‐targeting agent is modelled and its effect is quantified. The dose enhancement factor of ATRA for the whole tumour population is 1.06, meaning that the addition of the agent acts as a radiosensitizer, although the value close to unity indicates the limited action of ATRA among non‐CSCs. Yet, when CSCs are being recruited, the dose enhancement factor is significant (DEF = 1.89) due to the increased population of cycling cells that are affected by this agent. This outcome is illustrated by the cell survival curve of the HNC, whereby tumour eradication was achieved after 36 days of hyperfractionated radiotherapy when combined with ATRA.
CSC‐targeting agents, such as ATRA, are innovative approaches in oncology and several agents are presently being investigated. The current work offers a novel perspective on the effect of ATRA and radiotherapy on cancer cells, and at the same time identifies gaps in pre‐clinical research that need to be filled to increase the potential of novel targeted therapies in head and neck cancer management.
In head and neck cancers accelerated tumour repopulation is a major cause of treatment failure. Consequently, protracted radiotherapy schedules must be avoided, particularly in tumours comprising of large CSC compartment that fast regrow the tumour as a response to treatment‐induced cell kill. This shows the necessity to quantify the CSC population in order to design an efficient treatment for the individual patient.
While striving to simulate processes that are close to the biological world, models have their limitations. The paragraphs below illustrate the confines of the current study.
The current work is an in silico model of head and neck cancer treated with CSC‐targeted therapy combined with fractionated radiotherapy. The data used for simulation was supplied by experimental studies reported in the scientific literature in order to achieve a biologically realistic tumour and response to therapy. Where the literature lacked of quantitative data required by the model, the authors have extrapolated parameter values from other solid tumours.
The recruitment model is partially based on experimental findings reported on breast cancer cell lines. However, there are no indications in the literature concerning the pattern of recruitment, thus some assumptions had to be made for the current simulation. Given the fact that re‐entry into the cell cycle is triggered by cell loss, recruitment is modelled after the first week of treatment when cell kill is high enough to onset repopulation via cell recruitment. Afterwards, cellular re‐entry into the cycle occurs after each dose of radiation, leading to a continuous supply of cancer stem cells.
Considering that head and neck cancers are typically poorly oxygenated, the simulated tumour is a moderately hypoxic HNC with an average pO2 (partial oxygen pressure) value of 6 mm Hg. This fact influences overall radioresistance, as hypoxic cells are known to have a more efficient DNA repair capability. Thus, if a tumour is less hypoxic than the one modelled above (pO2 > 6 mm Hg), the overall response to radiotherapy is likely to be superior, whereas in a more hypoxic tumour (pO2 < 6 mm Hg) the outcome is probably worse.
Given the fact that the literature focuses on the effect of ATRA on cancer stem cells only, no interaction between this agent and non‐stem cancer cells has been modelled in this work.
As a final conclusion, ATRA exhibits a powerful effect on CSCs when combined with radiotherapy. However, as shown by the model, the more radioresistant quiescent cell population should not be ignored, as it can be a potential threat to treatment outcome when cells are recruited into the cell cycle.
CONFLICTS OF INTEREST
The authors declare no conflict of interest.
ORCID
Loredana G. Marcu http://orcid.org/0000-0002-6703-979X
ACKNOWLEDGEMENTS
This work was supported by a grant of the Ministry of National Education, CNCS‐UEFISCDI, Project no. PN‐II‐ ID‐PCE‐2012‐4‐0067.
Marcu LG, Marcu D. The effect of targeted therapy on recruited cancer stem cells in a head and neck carcinoma model. Cell Prolif. 2017;50:e12380 10.1111/cpr.12380
REFERENCES
- 1. Bourhis J, Overgaard J, Audry H, et al. Hyperfractionated or accelerated radiotherapy in head and neck cancer: a meta‐analysis. Lancet. 2006;368:843‐854. [DOI] [PubMed] [Google Scholar]
- 2. Moore N, Lyle S. Quiescent, slow‐cycling stem cell populations in cancer: a review of the evidence and discussion of significance. J Oncol. 2011. Pii:396076. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Prince ME, Sivanandan R, Kaczorowski A, et al. Identification of a subpopulation of cells with cancer stem cell properties in head and neck squamous cell carcinoma. Proc Natl Acad Sci U S A. 2007;104:973‐978. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Harper L, Piper K, Common J, Fortune F, Mackenzie IC. Stem cell patterns in cell lines derived from head and neck squamous cell carcinoma. J Oral Pathol Med. 2007;36:594‐603. [DOI] [PubMed] [Google Scholar]
- 5. Tang AL, Owen JH, Hauff SJ, et al. Head and neck cancer stem cells: the effect of HPV–an in vitro and mouse study. Otolaryngol Head Neck Surg. 2013;149:252‐260. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Campa VM, Gutiérrez‐Lanza R, Cerignoli F, et al. Notch activates cell cycle reentry and progression in quiescent cardiomyocytes. J Cell Biol. 2008;183:129‐141. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Phillips TM, McBride WH. Pajonk F The response of CD24(‐/low)/CD44+ breast cancer‐initiating cells to radiation. J Natl Cancer Inst. 2006;98:1777‐1785. [DOI] [PubMed] [Google Scholar]
- 8. Al Hajj M, Wicha MS, Benito‐Hernandez A, Morrison SJ, Clarke MF. Prospective identification of tumorigenic breast cancer cells. Proc Natl Acad Sci U S A. 2003;100:3983‐3988. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Morrison SJ, Kimble J. Asymmetric and symmetric stem‐cell division in development and cancer. Nature. 2006;441:1068‐1074. [DOI] [PubMed] [Google Scholar]
- 10. Vlashi E, McBride WH, Pajonk F. Radiation responses of cancer stem cells. J Cell Biochem. 2009;108:339‐342. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Koukourakis MI, Giatromanolaki A, Tsakmaki V, Danielidis V, Sivridis E. Cancer stem cell phenotype relates to radio‐chemotherapy outcome in locally advanced squamous cell head–neck cancer. Br J Cancer. 2012;106:846‐853. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Geissler C, Hambek M, Leinung M, et al. The challenge of tumor heterogeneity–different phenotypes of cancer stem cells in a head and neck squamous cell carcinoma xenograft mouse model. In Vivo. 2012;26:593‐598. [PubMed] [Google Scholar]
- 13. Peitzsch C, Perrin R, Hill RP, Dubrovska A, Kurth I. Hypoxia as a biomarker for radioresistant cancer stem cells. Int J Radiat Biol. 2014;90:636‐652. [DOI] [PubMed] [Google Scholar]
- 14. Cabrera MC, Hollingsworth RE, Hurt EM. Cancer stem cell plasticity and tumor hierarchy. World J Stem Cells. 2015;7:27‐36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Campos B, Wan F, Farhadi M, et al. Differentiation therapy exerts antitumor effects on stem‐like glioma cells. Clin Cancer Res. 2010;16:2715‐2728. [DOI] [PubMed] [Google Scholar]
- 16. Lim YC, Kang HJ, Kim YS, Choi EC. All‐trans‐retinoic acid inhibits growth of head and neck cancer stem cells by suppression of Wnt/β‐catenin pathway. Eur J Cancer. 2012;48:3310‐3318. [DOI] [PubMed] [Google Scholar]
- 17. Bertrand G, Maalouf M, Boivin A, et al. Targeting head and neck cancer stem cells to overcome resistance to photon and carbon ion radiation. Stem Cell Rev. 2014;10:114‐126. [DOI] [PubMed] [Google Scholar]
- 18. Zhang P, Xu Q, Chen WT, Duan LQ, Zhang ZY, Zhou XJ. Synergistic down‐regulation of telomerase by all‐trans retinoic acid and antisense oligonucleotide in oral squamous cell carcinoma cell line (Tca8113). Oral Oncol. 2005;41:909‐915. [DOI] [PubMed] [Google Scholar]
- 19. Marcu LG, Marcu D, Filip S. In silico study of the impact of cancer stem cell dynamics and radiobiological hypoxia on tumour response to hyperfractionated radiotherapy. Cell Prolif. 2016;49:304‐314. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Hall EJ. Radiobiology for the radiologist. 5th ed, Philadelphia: Lippincott Williams & Wilkins; 2000. [Google Scholar]
- 21. Tannock IF, Hill RP. The basic science of oncology, 3rd ed., New York: McGraw‐Hill; 1998. p. 135‐165. [Google Scholar]
- 22. Terz JJ, Lawrence W, Cox B. Analysis of the cycling and noncycling cell population of human solid tumors. Cancer. 1977;40:1462‐1470. [DOI] [PubMed] [Google Scholar]
- 23. Forster G, Cooke TG, Cooke LD, Stanton PD, Bowie G, Stell PM. Tumour growth rates in squamous carcinoma of the head and neck measured by in vivo bromodeoxyuridine incorporation and flow cytometry. Br J Cancer. 1992;65:698‐702. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Steel GG. Basic clinical radiobiology, 3rd ed., London: Oxford University Press; 2002. [Google Scholar]
- 25. Marcu LG, Marcu D. In silico modelling of a cancer stem cell‐targeting agent and its effects on tumour control during radiotherapy. Sci Rep. 2016;6:32332. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Withers HR, Maciejewski B, Taylor JM, Hliniak A. Accelerated repopulation in head and neck cancer. Front Radiat Ther Oncol. 1988;22:105‐110. [DOI] [PubMed] [Google Scholar]
- 27. Cheung TH, Rando TA. Molecular regulation of stem cell quiescence. Nat Rev Mol Cell Biol. 2013;14:329‐340. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Marcu LG. Tumour repopulation and the role of abortive division in squamous cell carcinomas during chemotherapy. Cell Prolif. 2014;47:318‐325. [DOI] [PMC free article] [PubMed] [Google Scholar]