Any opinions, findings, conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of UK Essays.
To preserve genome integrity and stability, cells have developed elaborate kinetic pathways to cope with the spectrum of DNA lesions induced by endogenous and exogenous sources of damage. Most natural DNA damage is caused by chemical damage (e.g. oxidative stress). Exposure to ionizing radiation stimulates free radical formation that oxidize proteins and lipids generating abasic sites and single-strand breaks which collectively induce mitotic failure and subsequent inflammatory responses in tissues.
Depending upon the level of acquired DNA damage, DNA repair machineries manipulate specific cell-cycle checkpoints (either G1/S or G2/M) to prevent continuation of the cell-cycle to allow repair of various DNA lesions. DNA double-strand breaks are the most dangerous type of DNA damage, but can be repaired by multiple repair machineries – homologous recombination repair, base excision repair, non-homologous end joining repair, and nucleotide excision repair. Base excision repair and mismatch repair are triggered in response to base damages and single-strand breakage for efficient restoration of the fidelity of the DNA molecule (Mladenov and Iliakis, 2011). However, blockage of these pathways results in increased radiosensitivity. If tumour cells can proficiently repair the acquired radiation-induced damage, radioresistance develops enabling cell-cycle continuation. However, if persistence of residual un-rejoined double-strand breaks remains or mismatched, these repair pathways induce apoptosis to prevent accumulation of mutations; thus, restricting mutagenesis as such lesions can lead to chromosome aberration and loss of genetic material.
Radiotherapy dosage is measured according to the energy that body tissue absorbed in units gray (Gy), which equals joule/kilogram (J/kg). Slowly proliferating and dormant cells are less radiosensitive in comparison to those with high proliferation rates. The goal of curative radiotherapy is to inactivate all tumour cells with minimal damage to surrounding normal tissues for avoidance of treatment-related dysfunctions. However, the dose-effect relationship is biased towards the principle that the higher the dose, the higher the tumour-control probability. Nevertheless, to target the tumour effectively, radiotherapy always must be executed with margins. Yet, the dose-limiting factor is the risk of toxicity in surrounding normal tissues. Although most patients tolerate treatment, some cancer patients achieve survivorship at the cost of treatment complications occurring in these surrounding tissues.
Toxicity development within patients whom undergone curative radiotherapy can be acute and chronic. Although acute toxicity occurs within proliferating tissues during treatment or shortly post-completion, it resolves within a few months post-treatment. Whereas, chronic toxicity symptoms persist several months post-treatment completion, which negatively affects patient’s quality-of-life. Fibrosis, an example of chronic toxicity, can result in the incidence of hardening in the breast post-radiation of tumours in the breast, or obstruction of bowel and/or urethra post-radiation of tumours in the pelvis. The implementation of different endpoints from a variety of scoring systems hinders the identification of clinical biomarkers associated with radiotoxicity making it cumbersome for comparative study analyses. Therefore, the Common Terminology Criteria for Adverse Events (CTCAE) version 4 scoring system to assess toxicity during and post-radiotherapy for breast and prostate cancer is used in this thesis to identify relevant phenotypes. This thesis focused on gastrointestinal and genitourinary complication manifestations within PC patients, as urinary bladder, rectum and urethra surrounds the prostate gland. Breast appearance alterations amongst BC patients were assessed (i.e., erythema (acute inflammation/reddening of the skin), telangiectasia (chronic radiation dermatitis), ulceration, atrophy, and induction (a proxy for fibrosis)).
The ability to stratify patients for radiation-induced toxicity in normal tissues could influence clinical treatment decisions. Indeed, modification of radiotherapy for radiosensitive patients could potentially avoid unnecessary worsening of quality-of-life incurred when the likelihood of radiotoxicity-related morbidities undermines the risk of tumour recurrence. Therefore, development of user-friendly predictive tests employing minimally-invasive sampling methods for the prediction of complications pre-therapy is critical.
The integration of phenotype profiles (clinical data and treatment parameters) within well-calibrated predictive models is one clinical approach. Indeed, the assessment of the relationship between toxicity end-points and radiation-induced damage have been investigated via several predictive tests for identification of individual cancer patient radiosensitivity. Radiation-induced apoptosis in peripheral blood T-lymphocyte subsets post in vitro irradiation gives the most promising results in prospective clinical trials for correlation with chronic toxicity post-radiotherapy. In addition, scoring of radiation-induced DNA damage and repair kinetics by using the Comet assay may be useful for prediction of reactions to radiotherapy.
Apoptosis occurs through two main pathways: intrinsic and extrinsic. DNA damage and oxidative stress initiates the intrinsic pathway. Binding of death ligands to death receptors triggers the extrinsic pathway. Radiation-induced lymphocyte apoptosis assay is based on the initiation of CD8+ T-lymphocyte apoptosis, and this assay has been shown to predictive toxicity in cancer patients. The RILA is a fluorescence-activated cell sorting-based assay in which whole blood cell culture is irradiated in tissue culture media then it undergoes a 48-hour incubation period that allows the apoptotic process to occur. This FACS assay involves an antibody for CD8+ (a cell surface marker on cytotoxic T-cells). As the last stage of apoptosis involves nuclear membrane breakdown exposing the DNA, propidium iodine (DNA marker) is also used to measure the amount of DNA. The proportion of cells in the apoptotic phase (fluorescence decrease of nuclear DNA) is quantified to give information on how radiosensitive the cells are.
Studies have assessed whether radiation-induced T-lymphocyte apoptosis was correlated with hypersensitive patients (acute and/or chronic toxicities) to determine whether RILA could identify patients at risk of suffering from radiotoxicity. Henriquez-Hernández et al. (2011) reported that BC patients with higher levels of radiation-induced lymphocyte apoptosis were at a minimal risk of suffering late toxicity. Findings by Azria et al. (2015) later confirmed that hypersensitive patients displayed a compromised apoptotic response. Azria et al. (2015) reported an inverse relationship between the percentage of apoptosis in irradiated lymphocytes and chronic toxicity occurrence in BC patients (i.e. patients with low radiation-induced lymphocyte apoptosis scores had high toxicity grade). In Dr Talbot’s group, RILA showed to predict one-year bowel toxicity in a cohort of 50 PC patients (data unpublished but illustrated in Figure 1.5.1). A one-way between-group analysis of variance was conducted to explore the impact of radiation-induced apoptotic scores, as measured by RILA, on the development of bowel toxicity in PC patients. Patients were divided into three groups according to their RILA scores based on previous publications (low score<14.8; medium core=14.8-24.6; high score>24.6). There was a statistically significant difference at the p<.05 level in RILA scores for the three groups: F(2,47)=4.47, p=.017. This result revealed there was an inverse correlation between RILA score and bowl toxicity. That is, PC patients who had bowel toxicity at one-year had low RILA scores. The mechanism behind this is not yet clear. However, RILA seem to be a promoising assay to identify hypersensitive patients as its negative predictive value is high. Therefore, this assay was included in this study for assay comparative analysis (with Comet assay). Interestingly, Pinar et al. (2010) reported no relation was found between DNA damage measured by Comet assay and radiation-induced lymphocyte apoptosis at 1, 2 and 8Gy, albeit in only 26 patients.
The single-cell gel electrophoresis, or Comet assay, is a valuable method for the assessment of DNA strand breakage in a single cell, based on the alkaline lysis of DNA at sites of damage. Cells are embedded in agarose, lysed, electrophoresed, and stained before DNA damage is visualized using fluorescence microscopy. Breakage in the DNA molecule catalyses unwinding of its complex supercoil. The damaged, liberated DNA which leaves the nucleus during electrophoresis, leads to the formation of what appears a “comet-like” structure with a bright fluorescent head, in which the head consists of intact nucleoid body containing undamaged DNA and the tail the damaged DNA strand breakage streaming away from it (as illustrated in Figure 1.5.2). The amount of DNA in the tail is thus proportional to the level of damaged DNA in the respective cell. Therefore, it was hypothesized that there will be less DNA in the comet tail as time interval increased (i.e. DNA repair occurred).
Several studies have controversially reported on the implementation of the Comet assay in prospective settings for the identification of radiosensitivity. Sterpone et al. (2010) analysed the response to ionizing radiation exposure in sporadic BC patients and healthy controls by measuring DNA damage through alkaline Comet assay. This group analysis identified BC patients with reduced efficiency in DNA repair had an increased risk of toxicity. This was in agreement with results reported by Alapetite and colleagues whom observed that BC patients with most severe complications showed impaired rejoining as analysed through alkaline Comet assay (Alapetite et al., 1999). That is, according to these studies, reduced DNA repair correlated with risk of toxicity. Yet, there are also studies challenging the predictive potential of the Comet assay. The Popanda et al. (2003) and Padjas et al. (2012) studies found no correlation between risk of toxicity (acute and chronic side-effects) and DNA degradation using the Comet assay.
The reason for the discrepancy may be that previous studies employing the alkaline Comet assay were relatively small. Hence, this study conducted evaluation of the alkaline Comet assay amongst breast and PC patients whom underwent radiotherapy, to identify phenotypic characteristics for predisposition for clinical radiosensitivity. For this, DNA damage and oxidative stress was induced in cancer patients with solid tumours (breast and prostate) by exposing their peripheral blood to X-ray irradiation for the investigation of DNA repair.
While radiation dose is the main factor influencing toxicity, additional factors, including pre-existing conditions (viral infection), may modify the dose–toxicity relationship. If the ends of viral genomes are exposed and recognized as double-strand breaks, these viruses can activate the DNA damage checkpoint pathway. For instance, ataxia-telangiectasia mutated checkpoint pathway, which responds to double-strand breaks, is activated, blocked and mislocalized (along with checkpoint kinase 2) to a cytoplasmic virus assembly zone in response to human cytomegalovirus (CMV) DNA replication (Smolarz et al., 2015). In contrast to other members of the Herpesviridae family, CMV is not an oncogenic virus. CMV infects 70-90% of the general population, causing acute, persistent or lifelong latent infection. Furthermore, CMV can reactivate under some conditions triggering the DNA damage checkpoint and blockage of cell-cycle progression if the host is immunosuppressed (Gaspar and Shenk, 2006). There is evidence that CMV can induce alterations of T-cells function in response to chronic infection stimulation and its antigen-presenting capacity. CMV infection may result in atypical lymphocytes (non-functional or limited functionality – known as T-cell exhaustion), induce chromosomal breaks and aberrations that can lead to genomic instability (Noriega et al., 2012).
The weakening of the DNA response is necessary for effective establishment of CMV viral latency. CMV-specific immunity is characterized by an increase of CMV-specific cytotoxic T-cells in seropositive individuals. Due to T-cell exhaustion, the accumulated damage renders cells susceptible to apoptosis. CMV seropositive patients have less naïve T-cells as these have switched to effectors in response to the recurrent viral infection activation due to circulating antibody levels to CMV infection (i.e. decreased helper function and increased suppressor-cell activities). Radiation-induced lymphocyte apoptosis have been shown to be inversely correlated with toxicity, and CMV infection can induce damage that renders cells susceptible to apoptosis. Thus, cancer patients with CMV infection could possibly by radioresistant. That is, CMV protects against toxicity due to an increase of CMV-specific cytotoxic T-cells in seropositive individuals which increases apoptosis and might reduce risk of radiotoxicity. Patients without CMV infection will experience high toxicity.
While serological evidence for CMV in cancer patients is limited, Neote et al. (1993) found elevated levels of CMV IgG preceded the development of breast cancer in a Norwegian cohort. Furthermore, Kuo et al. (2008) case study, involving 15 cancer patients, showed that CMV reactivation were elevated during chemotherapy. The reason for this viral infection analysis was to assess whether an anti-CMV IgG response could be associated with increased risk of radiotoxicity in cancer patients undergoing radiation treatment.
How to elevate tumour radiation response while reducing radiosensitivity in adjacent normal tissues has become a core issue in understanding radiation-induced complications in cancer patients. Indeed, avenues of predicting patients’ likelihood of radiotoxicity-related morbidities post-radiotherapy and innovative approaches for their consequent improvement in the therapeutic ratio is crucial. Recruitment of patients with different solid tumours can also strengthen pre-clinical models for predictive and prognostic value as radiosensitivity maybe cancer specific. Accordingly, the purpose of this study was to investigate normal tissue response in cancer patients undergoing curative radiotherapy.
Repair of radiation-induced DNA damage plays a critical role for both subsequent cancer risk and susceptibility to adverse reactions post-radiotherapy. This study aimed to investigate different candidate cellular radiosensitivity assays for prediction of normal tissue complications, in a prospective case-study of breast and prostate cancer patients. The first objective was to evaluate whether DNA repair data determined ex vivo correlated with radiotoxicity. It was hypothesized in this part of the study that patients with radiotoxicity have reduced efficiency in DNA repair capacity. This led to the determination of whether Comet assay ex vivo responses to radiation might lead to a useful clinical test by using groups larger than those previously investigated (e.g. Padjas et al., 2012; Sterpone et al. 2010). The second objective was to conduct a viral infection analysis to assess whether an anti-CMV IgG response could be associated with increased risk of radiotoxicity in cancer patients undergoing radiation treatment. Patients were analysed for CMV infection as this virus might influence radiosensitivity risk. It was thus hypothesized that CMV seropositive patients would have a reduced risk of radiotoxicity.
Assays were carried out on peripheral blood drawn pre-radiotherapy from cancer patients (breast and prostate) recruited from the REQUITE cohort (n=600) (West et al., 2014). X-ray induced DNA damage and relative repair evaluations in a randomly selected subset of the same cohort (n=229) was performed with the alkaline Comet assay following ex vivo irradiation. Associations of individual measures and risk scores with toxicity phenotypes were analysed using regression analysis. In addition, IgG-class CMV antibody detection using a very sensitive and specific ELISA was undertaken in a subset of the same cohort of patients (n=366) irrespective of toxicity
occurrence. Regarding the Comet assay, generally a low mean % of DNA in the tail was observed in unirradiated cells, which, however, is increased considerably post-irradiation with 8Gy. To quantify the DNA damaged induced (i.e. DNA strand breakage), % of DNA tail was measured after alkaline electrophoresis. The % of DNA in comet tail of the healthy donor peripheral blood was constant (data illustrated in Figure 4.1.1); an indication that preparation and subsequent processing of patients’ peripheral blood did not introduce considerable damage to their DNA. This healthy donor was excluded from the analysis.
Figure 4.1.2 shows the baseline level, X-ray induced DNA damage, and DNA damage repair in peripheral blood of 81 BC patients and 58 PC patients measured as % of DNA in the tail. The average and variation shown in Table 1. Interestingly, PC patients did not exhibit a significant (p=.993) mean level of background DNA damage when compared with BC patients (15.56 vs 14.91). Immediately post-irradiation, the acquired mean level of X-ray induced DNA damage did not differ between BC patients and PC patients (45.53 vs 50.9, p=.115). The DNA repair also did not differ between BC patients and PC patients (32.47 vs 36.77, p=.0.276). Concerning DNA damage repair, in both cohort of patients, DNA damage slowly decreased 30min post-irradiation, never reaching the background level.
As sample size (n=139; 81 BC and 58 PC) is small, Comet assay results done by other experimenters were included. Therefore, an experimenter comparative analysis was conducted. Blood samples were collected from a total of 600 patients (350 BC patients and 250 PC patients). A total of 230 (130 BC and 100 PC) randomly selected patients were analysed by the alkaline Comet assay of which Kerstie conducted the Comet assay on 45 patients, Max analysed 31 patients, and 15 patients were analysed by Lilas. A one-way analysis of variance (ANOVA) revealed there is evidence of a difference in the overall mean of the % of tail DNA measured by the alkaline Comet assay amongst the four different experimenters. Kerstie, Max and Taniqua’s mean were not significantly different, with Kerstie having the lowest mean (as illustrated in Table 2). However, Lilas’ mean differed significantly from the other experimenters, having the highest mean amount. The data points were evenly spread within Kerstie, Max and Taniqua’s Comet results (data not shown) indicating equal variances. Yet, Lilas’ data points were not fairly evenly spread (data not shown). Lilas’ Comet assay results, therefore, were deemed unreliable and excluded from further Comet assay analysis. Figure 4.1.3 illustrates the difference between experimenter group means, and the mean difference is shown in Table 2. Thus, a total of 215 patients (125 BC and 90 PC) % of DNA in the tail as measured by the Comet assay underwent further analysis towards predicting radiotoxicity.
The induction and repair of DNA lesions such as DNA strand breaks are key factors that modulate individual radiation sensitivity, and may thus be altered in individuals developing clinical adverse reactions of the co-irradiated normal tissues post-radiotherapy. The induction and relative repair of X-ray induced DNA damage was evaluated in 215 patient-derived peripheral blood samples using the Comet assay to investigate its predictive value. In BC cohort (n=125), a significantly higher value of DNA damage post 30min was observed when it was compared with background level damage (30.35 vs 13.92, p<0.001). In PC cohort (n=90), a similar trend was observed (34.02 vs 14.4, p<0.001). A somewhat higher level of induced damage was observed in PC patients in comparison to BC patients. That is, the BC patients have non-significantly lower X-ray induced DNA damage. This tendency was evident in unirradiated cells (i.e., PC patients had higher background level damage). X-ray induced DNA damage was computed by deduction of 8Gy(0mins)-0Gy. Rate of repair was computed as 8Gy(0mins)-8Gy(30min repair). Relative repair was computed as the percentage of DNA repair divided by induced DNA damage. Comet assay experimental parameters summarized in Table 3, and X-ray induced damage and rate of repair graphically illustrated in Figure 4.2.2.
It is noteworthy that assessment of late toxicity is normally conducted at three years post-radiotherapy. In this study, due to time constraints, late toxicity at one-year was analysed. It is expected that over time late toxicity will develop progressively worse (i.e., patients with grade 1-2 toxicity at one-year will have grade 3-4 at three years). Patients in whom late toxicity is not observed at one-year, will develop grade 1-2 later. Furthermore, blood samples were taken at pre-radiotherapy planning clinics, and acute and one-year toxicity data have not been completely collected for several patients. Acute toxicity and one-year toxicity scores were computed by deduction of baseline symptom score from score at the end of radiation treatment session and one-year post completion, respectively. A composite score of overall toxicity (STAT score) was calculated and adjusted for known predictors by regression (rSTAT score).
Radiation-induced toxicity occurrences in breast cancer patients
Data on outcome is reported by number of patients instead of number of episodes. Full clinical toxicity outcome data were available (acute and one-year toxicity) in 228 patients of 350 BC patients recruited. Of the 125 BC patients assessed via Comet assay, at the time of analysis, 104 patients had acute toxicity data, and 89 patients had one-year toxicity data available. Table 4 summaries the cohort population characteristics and Table 5 shows distribution of adverse reactions within the cohort as assessed using the CTCAE scale within the patient cohort.
Studies have shown that several patient factors (i.e., body mass index, breast volume, age, diabetes, smoking status, and cardiovascular disease) influence risk of radiosensitivity. Therefore, these patient factors were analysed to establish whether correlation points exist in relation to toxicity using the rSTAT score derived within the cohort population. Nonparametric Spearman’s rho correlation was used to assess this relationship, as data not normally distributed. Interestingly, age, diabetes nor breast volume did not correlate with neither acute rSTAT nor one-year rSTAT within the patient cohort. In addition, cardiovascular status did not correlate with either acute rSTAT or one-year rSTAT. Unexpectedly, body mass index and smoking status negatively correlated with acute rSTAT (distribution depicted in Figure 4.2.3 and Figure 4.2.4, respectively). Whereas, neither correlated with one-year rSTAT. Full correlation analysis is depicted in Table 6.
The relation between acute toxicity and one-year toxicity in the radiosensitive patients amongst this cohort population was also investigated. It was of interest to assess whether patients who developed acute toxicity also developed toxicity at one-year. The relation between acute rSTAT and one-year rSTAT was investigated using Pearson product-moment correlation coefficient. There was a weak, positive correlation between acute rSTAT and one-year rSTAT, r=.19, n=228, p=.004, with high one-year toxicity scores associated with high acute toxicity scores (distribution illustrated in Figure 4.2.5).
When considering the Comet assay, X-ray induced damage and relative repair were analysed to establish whether correlation points exist in relation to toxicity using the rSTAT score. Work by Sterpone et al. (2010) concluded induced damage and repair capacity correlated with radiosensitivity in breast cancer patients. It was hypothesized that impaired DNA repair capacity may increase the risk of radiation adverse reactions as mutations in the non-homologous end joining proteins leads to mismatch repair. Nonparametric Spearman’s rho correlation was used to assess this relationship, as data not normally distributed. X-ray induced DNA damage as measured by Comet assay showed no correlation in relation to acute rSTAT (ρ=-.14, p=.172) and one-year rSTAT (ρ=-.135, p=.231). When relative repair was considered, the risk of acute and one-year toxicity did not correlate with relative repair as measured by the Comet assay (ρ=-.066, p=.509 and ρ=.197, p=.064, respectively).
Radiation-induced toxicity occurrences in prostate cancer patients
Data on outcome is reported by number of patients instead of number of episodes. Full acute clinical toxicity outcome data was available for all 250 PC patients. However, bowel, urinary and erectile dysfunction toxicity data were available for 190, 163 and 168 PC patients, respectively. Of the 90 PC patients assessed via Comet assay, bowel, urinary and erectile dysfunction toxicity data were available for 63, 54 and 53 patients, respectively. Table 7 summaries the cohort population characteristics.
Patient factors (i.e., weight, alcohol status, age, diabetes, and smoking status) can influence risk of radiosensitivity. Therefore, these patient factors were analysed to establish whether correlation points exist in relation to toxicity using the STAT scores derived within the cohort population. Nonparametric Spearman’s rho correlation was used to assess the relationship, as data not normally distributed. Unexpectedly, all the patient factors showed no correlation with acute STAT, bowel STAT, urinary STAT, and erectile STAT. Full correlation analysis is depicted in Table 8.
Concerning X-ray induced DNA damage as measured by Comet assay, Pearson product-moment correlation coefficient revealed no correlation between DNA damage with acute STAT (p=.186), bowel STAT (p=.786), urinary STAT (p=.118), and erectile STAT (p=.929). When relative repair was considered, the risk of neither acute STAT, bowel STAT, nor urinary STAT correlated with relative repair within this cohort population (p=.469, p=.65, and p=.655, respectively). There was a trend towards correlation between erectile STAT and relative repair (p=.054).
Radiation-induced lymphocyte apoptosis assay has the best evidence to predict radiosensitivity response in cancer patients, and this assay could work in combination with the Comet assay because it tests a different but related pathway. One unresolved issue is whether the signal for apoptosis is triggered by radiation-induced lesions accumulated in the cell or by mismatched DNA that has been repaired. Derived experimental measures from the Comet assay (X-ray induced DNA damage and relative repair) were used to establish whether correlation points exist between these radiobiology assays.
Breast cancer patients
RILA correlates with relative repair (Comet assay)
Interestingly, Pearson product-moment correlation coefficient revealed although X-ray induced DNA damage did not correlate (r=-.13, p=.198), there was a weak, positive association between RILA and relative repair (r=.36, p<0.001, R2=13.2) (graphically illustrated in Figure 4.3). Patients with a faster relative rate of repair were found to have an increase rate of apoptosis (i.e. high radiation-induced lymphocyte apoptosis scores). The repaired DNA double-strand breakage may have been mismatched or there might have been mutations in the non-homologous end-joining pathway resulting in an increase level of apoptosis as the DNA was being repaired (i.e. elevated levels of repair associated with elevated levels of apoptosis). It is postulated that 13.2% of radiation-induced lymphocyte apoptosis observed is explained by mismatch DNA repair level.
Prostate cancer patients
RILA did not correlate with Comet assay
In this patient cohort population, Pearson product-moment correlation coefficient revealed X-ray induced DNA damage did not correlate (r=-.02, p=.88) with RILA nor did relative repair (r=-.12, p=.26).
To search for features which could better distinguish radiosensitive cancer patients, CMV serologic status was detected in 366 samples (183 BC and 183 PC) from the entire cohort of 600 patients to assess the effect of anti-CMV immunoglobulin G (IgG) positivity. Anti-CMV ELISA assay was conducted on thawed plasma samples, and CMV serology was assessed by IgG ELISA as described in methods. This assay was performed on at least two occasions, and patients were defined as CMV seropositive or seronegative according to the absorbance validity range. A quantitative measurement of CMV IgG was calculated as described. Figure 4.4 shows an illustration of an anti-CMV ELISA plate, and Table 9 reports the results interpretation obtained from this Anti-CMV IgG plate.
Based on the cut-off value computed from this anti-CMV IgG ELISA plate and using the results interpretation described in methods, this Anti-CMV IgG plate revealed 32 patients were CMV seronegative (21 PC and 11 BC) and 42 patients were CMV seropositive (24 PC and 18 BC).
Cases of CMV infection in this study should be understood in the context of the persistent presence of CMV antigen (IgG) rather than primary disease (IgM) or reactivation of CMV. The objective of CMV serologic testing was to assess whether an anti-CMV response could be associated with increased risk of radiosensitivity in patients undergoing radiation treatment. It is thus hypothesized that patients with CMV infection have less naïve T-cells as these have switched to effectors in response to circulating antibody levels to CMV infection (i.e. increased suppressor-cell activities) based on previous research (Chidrawar et al., 2009). Therefore, CMV seropositive patients would have a reduced risk of radiotoxicity. Of the 366 samples, 187 patients were CMV seropositive and 179 patients tested CMV seronegative.
There were 95 (51.9%) CMV seropositive patients, and 88 (48.1%) CMV seronegative patients within this cohort population (n=183). The median age of the CMV seronegative group was 58 years old, and of the CMV seropositive group the median age was 65 years old. Considering acute clinical radiotoxicity as defined by CTCAE score, CMV serostatus had no significant effect on risk of acute radiosensitivity (p=.299, by Mann-Whitney). However, when the one-year rSTAT was analysed, there was a significant relation between CMV serostatus and radiotoxicity risk at one-year (p=.043, by Mann-Whitney), as shown in Figure 4.5. This result indicates that the majority of patients who were CMV seropositive were also patients who had worse toxicity; a complete opposite to the proposed hypothesis that CMV infection would protect against toxicity.
CMV status and radiotoxicity occurrence
There were 92 (50.3%) CMV seropositive patients, and 91 (49.7%) CMV seronegative patients within this cohort population (n=183). The median age of the CMV seronegative group was 69 years old, and of the CMV seropositive group the median age was 70 years old. Considering clinical radiotoxicity, in a univariate analysis, Mann-Whitney revealed no significant difference in risk of toxicity in relation to CMV serostatus (illustrated in Table 10). The results imply CMV serostatus neither increase nor decrease patients’ risk of suffering radiotoxicity.
A cross assay analysis was conducted to investigate whether the derived measures from the Comet assay (X-ray induced DNA damage and relative repair) correlated to CMV serostatus in this cohort population (n=366; 183 BC and 183 PC).
Breast cancer patients
Pearson product-moment correlation coefficient revealed no correlation between both CMV serostatus and X-ray induced DNA damage (p=.801) nor CMV serostatus and relative repair (p=.836) as measured by Comet assay.
Prostate cancer patients
Pearson product-moment correlation coefficient similarly revealed there was no correlation between neither CMV serostatus and X-ray induced DNA damage (p=.409) nor CMV serostatus and relative repair (p=.263) as measured by Comet assay.
The result from both patient cohort population reveal CMV infection did not affect the acquired level of damage nor influence DNA repair capacity.
It has been demonstrated that CMV infection has profound effects on the immune system of chemotherapy-treated patients (Kuo et al., 2008). CMV infection increases CMV-specific cytotoxic T-cells in CMV seropositive individuals. There were 187(51.01%) CMV seropositive patients (95 BC and 92 PC), and 179(48.9%) CMV seronegative patients (88 BC and 91 PC) within the 366-sample population screened for anti-CMV IgG. Radiation-induced lymphocyte apoptosis was measured by gating around the lymphocyte population to determine the subpopulation of CD8 T-cells. There is a clear association between CMV serostatus and radiation-induced lymphocyte apoptosis within this cohort population. In relation to BC patients, RILA scores were significantly higher in CMV seropositive patients; the mean value in CMV seronegative patients was 56.53 compared to 85.26 in CMV seropositive patients (p<0.001, by Mann-Whitney) (Figure 4.6.1). This was a similar trend in the PC cohort; the mean value in CMV seronegative patients was 63.95 in comparison to 85.06 in CMV seropositive patients (p=.003, by Mann-Whitney) (Figure 4.6.1).
The results indicated that CMV seropositive patients had high radiation-induced apoptotic scores. It was then investigated to what extent CMV serostatus influenced RILA predictive power in relation to radiotoxicity. Thus, the dataset was then split in cases based on CMV serostatus and a correlation assessed whether RILA predicted toxicity within the various cohort population (BC and PC).
Breast cancer patients
Amongst BC patient cohort population, Pearson product-moment correlation coefficient indicated there was no correlation with relation to RILA and acute rSTAT and one-year rSTAT (as shown in Table 11) irrespective of CMV serostatus.
Prostate cancer patients
In PC patient cohort, there was a different trend. There was a non-significant trend amongst many of the toxicity endpoints and RILA irrespective of CMV serostatus (as shown in Table 12). However, there was a significant trend in relation to urinary STAT and RILA predictive power. It is noteworthy that the RILA assay shown to predict urinary STAT within this cohort population. Pearson product-moment correlation coefficient indicated there was a weak, negative correlation between RILA scores and urinary toxicity (r=-.183, n=130, p=.037) (Figure 4.6.2). This result implies that high radiation-induced lymphocyte apoptosis scores were associated with low urinary toxicity scores. Interestingly, when the cohort was split based on CMV serostatus, RILA predicted urinary STAT only in CMV seronegative patients.
Pearson product-moment correlation coefficient revealed a weak, negative correlation between urinary toxicity and RILA scores amongst CMV seronegative patients, r=-.308, n=51, p=.028 suggesting that CMV infected prostate cancer patients had developed a resistance against toxicity (graphically shown in Figure 4.6.3). There was no correlation between the other toxicity endpoints irrespective of CMV serostatus, as shown in Table 11.
Figure 4.3 shows a weak, positive correlation between RILA and relative repair as measured by Comet assay amongst the BC cohort. When the dataset was split based on CMV serostatus, this relationship was investigated once more. Interestingly, there was a positive correlation between RILA and relative repair in CMV seropositive patients (r=.49, p=.001) (Figure 4.7); whereas, there was no correlation between RILA and relative repair in CMV seronegative (r=.24, p=.14). This result implies CMV serostatus influence DNA repair capacity which subsequently affects the efficacy in DNA repair and induce apoptosis. This positive correlation might be due to a reactivation of the CMV infection within these patients.
X-ray induced DNA damage and repair in relation to adverse reactions
Although radiotherapy is primarily a local treatment, patients are exposed to a risk of toxicities in the treatment field and surrounding tissues, which may develop acutely and chronically. Because an intact DNA template is a fundamental prerequisite for normal cellular function, this part of the study evaluated the impact of X-ray induced DNA damage and relative repair in relation to the development of radiotoxicity in normal tissues post-radiation treatment in breast cancer and prostate cancer patients. The hypothesis behind this part of the study was that deficiency in DNA repair in leucocytes correlates with increased risk of radiosensitivity post-radiotherapy.
Alkaline Comet assay is a common test to investigate induced DNA damage and repair kinetics of DNA damage in human peripheral blood. In this study, the presence of X-ray induced DNA strand breakage and relative repair was evaluated. The data showed that BC patients have a non-significantly lower mean value of induced DNA damage in comparison to PC patients (Figure 4.2.1) which is not in agreement with Sterpone et al. (2010) work that showed BC patients have a high mean value of induced damage immediately post-irradiation. However, X-ray induced DNA damage did not correlate with radiosensitivity in either cohort population. Concerning DNA repair capacity, in both cohort population (BC and PC) the relative repair did not correlate with radiosensitivity in either cohort population. The results, although not accounting for BRCA1 and BRCA2 mutations, but on a sizable sample (n=215) of randomly selected breast cancer and prostate cancer patients, are in line with Padjas et al. (2012) that the Comet assay has not correlated with radiosensitivity and ex vivo irradiated peripheral blood amongst both cohort population.
This negative result is not unexpected in view of the conflicting results published by others in which radiation-induced damage was analysed in human peripheral blood. The Comet assay derived measurements as biomarkers for individual radiosensitivity are controversially discussed. Sterpone et al. (2010) work successfully identified double-strand breakage repair defective breast cancer patients had an increased risk for high grade toxicities during radiotherapy. That is, patients who were radiosensitive had delayed repair of DNA damage or deficiency within the DNA repair capacity in peripheral blood. However, there are studies also challenging the predictive potential of the Comet assay. Along with Padjas et al. (2012), Popanda et al. (2003) work using Comet assay did not establish correlation between DNA damage and acute toxicity in breast cancer patients. Greve et al. (2012) work found no correlation between late toxicity and Comet assay in breast cancer patients.
One reason for the discrepancy might reside in the tumour type amongst the patient cohort recruited as well as toxicity grade for inter-laboratory comparison or patient factors which were not accounted for in the analyses. It is speculated that the Comet assay might better predict associations with toxicity in patients with extreme cases. Alapetite et al. (1999) noted that the association between DNA repair and toxicity was found amongst patients with extreme toxicity scores as measured by RTOG scale. These patients had high % of DNA in the tail in the Comet assay (indication of high residual damage percentages). Therefore, to better predict normal tissue reactions in cancer patients using the Comet assay model merits further investigation. The Comet assay can evaluate the intensity of DNA damage and repair, but it is unable to visualize the fidelity of how the damaged DNA is repaired. Genetic variants in DNA repair genes can affect the amount of DNA damage repaired, and its repair integrity post-irradiation exposure. In addition, the incubation period of 30min post-irradiation is enough time for single-strand breakage but not double-strand breakage. It could be that assessing double-strand break repair specifically would still predict, however, more comprehensive analyses is needed to optimize repair kinetics for different radiation-induced DNA lesions.
It would be of interest to examine ataxia telangiectasia (ATM) nucleo-shuttling as this is a cell-based assay of DNA damage response pathway. Evidence suggest that the delays in the nucleo-shuttling of ATM proteins may be reliable parameters for predicting any level of individual radiosensitivity (Ferlazzo et al., 2014). Mutations in repair genes which influence damaged DNA repair was found to be enhanced in ataxia telangiectasia patients. Phosphorylated ATM forms move from the cytoplasm where they are stored to the nucleus to phosphorylate H2AX, an upstream actor required for repairing double-strand breakage. Any impairment or delay in the nucleo-shuttling of ATM may result in radiosensitivity and/or genomic instability (Ferlazzo et al., 2014). Also, ATM is blocked and mislocalized to a cytoplasmic virus assembly zone in response to human cytomegalovirus DNA replication (Smolarz et al., 2015).
Breast cancer: patient factors correlate with adverse reactions
Patient factors can influence risk of radiosensitivity (Tureson et al., 1996). Although age, diabetes, breast volume and cardiovascular status did not correlate with neither acute rSTAT nor one-year rSTAT within the patient cohort, unexpectedly, body mass index and smoking status negatively correlated with acute rSTAT (Figure 4.2.3 and Figure 4.2.4 respectively). This result indicated high body mass index was associated with low acute toxicity score.
In addition, high acute toxicity score was found in the majority of BC patients who were non-smokers. Juneja et al. (2016) found no association between breast volume and toxicity. This agreed with the result reported here. However, as toxicity scores were collected immediately post-radiation completion and one-year later, the follow-up time frame may not accurately report toxicity in this patient cohort. Since patients progressively worsen over time, follow-up assessment is normally done at three years and beyond as this allows an accurate recovery time assessment of normal tissue complication rates.
It was of interest to also investigate whether radiosensitivity patients showing acute toxicity also showed toxicity symptoms at one-year. The result showed a weak, positive correlation (p=.004) between acute toxicity and one-year toxicity in this cohort population (Figure 4.2.5). This result agrees with Tureson et al. (1996) work that showed an association between acute and late adverse reactions in 402 BC patients in which erythema and telangiectasia were primary skin toxicity endpoints. In addition, Greve et al. (2012) work revealed BC patients who had developed acute toxicity had an increased risk of developing chronic adverse reasons. The mechanism behind this might be due to mismatched DNA in cells that have went undetected. Although prolong DNA damage can initiate the apoptotic process, and as RILA was shown to inversely correlate with late toxicity in BC patients, this might not have been stimulated possibly due to BRCA1 mutations. However, like this present study, none of these studies accounted for BRCA1 and BRCA2 mutations.
Prostate cancer: patient factors did not correlate with adverse reactions
None of the patient factors (weight, age, diabetes, smoking status, alcohol intake) showed a significant association to the toxicity endpoints (acute, bowel, urinary and erectile) (Table 8).
Radiation-induced lymphocyte apoptosis assay has the best evidence to predict radiosensitivity response in cancer patients, and this assay could work in combination with the Comet assay because it tests a different but related pathway. Derived experimental measures from the Comet assay (X-ray induced DNA damage and relative repair) were used to establish whether correlation points exist between these radiobiology assays. Amongst the PC patient cohort, none of the derived measures from the alkaline Comet assay correlated with RILA. However, in the BC patient cohort, although induced damage did not correlate with RILA, a positive correlation was found between relative repair and RILA scores. That is, high relative repair percentages were associated with high radiation-induced lymphocyte apoptosis scores in this cohort. This indicated that although DNA damage was repaired, it was not being repaired efficiently thus triggering apoptosis. When damage occurs to the DNA of a cell, several responses are possible, including cell-cycle checkpoint activation, DNA repair and apoptosis. Apoptosis is a response though to protect the integrity of the DNA genome. Thus, apoptosis is the ultimate mechanism in avoiding chromosomal aberrations and abnormal cell division due to unrepaired or mis-repaired DNA damage. The result obtained in this study suggest apoptosis occurs only in apoptosis-prone cells (i.e., deficiency in how the DNA was being repaired in this cohort population triggered apoptotic process).
On the contrary, normal function of the p53 gene is apparently required for radiation-induced lymphocyte apoptosis. However, p53 gene might be mutated in cancer patients causing cells to lack the propensity for apoptosis as the cells are not expressing the required genes for apoptosis. This might explain the observation in the PC patient cohort as radiation-induce apoptosis is a signalled event instead of a passive consequence of lethally damaged cell. Furthermore, cells that undergo DNA damage and apoptotic pathway activation but survive may have altered function that could contribute to tumour resistance in cancer patients. It is thus of interest to genotype the patient cohort recruited in this study for further analysis.
CMV infection is a well-known complication in immunocompromised patients as this virus can establish latent and persistent infections in the host. However, little evidence is available regarding CMV infection influence in radiation-treated cancer patients. Cases of CMV infection in the present study should be understood in the context of the presence of the CMV capsid antigen (IgG) rather than primary disease (IgM) or reactivation of CMV. Due to time constraints, patient anti-CMV IgG ELISA analysis could only be carried out once rather in duplication. Given the impact of CMV infection on immune system (i.e. increase CMV-specific cytotoxic T-cells in CMV seropositive individuals), the next step was to investigate whether CMV infection brought about changes associated risk of with radiotoxicity. The objective of CMV serologic testing was to assess whether an anti-CMV response could be associated with increased risk of radiosensitivity.
CMV infection correlation with adverse reactions in cancer patients
Of the 366 samples, 187 patients (95 BC and 92 PC) were CMV seropositive and 179 patients (88 BC and 91 PC) tested CMV seronegative. When considering radiotoxicity as defined by CTCAE scale system, amongst the PC cohort, CMV serostatus shown to not correlate with risk of radiosensitivity. The result implies CMV serostatus did not influence the risk level of radiosensitivity. However, there was a significant different trend amongst the BC cohort. Although CMV serostatus did not correlate with acute toxicity in this cohort population, CMV infection did correlate with one-year toxicity (p=.036). The majority of CMV seropositive BC patients were also amongst those patients who had developed one-year toxicity (Figure 4.4). This indicated that CMV seropositive patients had worse toxicity. CMV infection can induced DNA damage and inhibit DNA repair genes or manipulate these mechanisms in such a way that is permissive for its viral replication. Oseguera et al. (2017) findings indicated an association between CMV and breast cancer. While this study have been conducted in vitro using cell lines, Oseguera et al. (2017) observed that CMV was associated with promoting malignant spread of breast cancer cells. Furthermore, there is evidence to suggest that CMV contributes to loss of cellular function (Pawelec et al., 2009). This might explain the result observed amongst CMV seropositive BC patient cohort in this study. CMV infection might be stimulating cellular changes that facilitate worse prognosis in BC patients; hence the majority of CMV seropositive BC patients in this cohort were also amongst those who developed one-year toxicity.
It has been demonstrated that CMV infection has profound effects on the immune system of chemotherapy-treated patients (Kuo et al., 2008). Suppression of a host’s cellular immune response is the major underlying predisposing factor for CMV infection. The CMV serostatus testing was performed to investigate the mechanism by which RILA predicts toxicity. There was an association between radiation-induced lymphocyte apoptosis and CMV in both patient cohort population, an indication that RILA scores increases in CMV seropositive patients (Figure 4.6.1). The magnitude of naïve CD8+ T-cells is significantly decreased in CMV seropositive individuals. Furthermore, increased apoptotic signalling in the cells might influence the rate and extent in which tissue damage heals post-irradiation. This might explain the correlation observed between CMV and one-year toxicity in BC patients. Nielson et al. (1982) concluded that collagen synthesis and fibroblast mitosis were function roles of T-cells. Therefore, if many T-cells have switched to effectors along with an increased apoptosis, interference of healing process can increase risk of chronic toxicity.
Chidrawar et al. (2009) noted that there was a profound increase in the terminally differentiated CD8+ T-cell compartment. This leads to speculation that CMV infection is a confounding factor which might be explained by reactivation of CMV infection in cancer patients (i.e. chronic CMV infection can stimulate an increase CMV-specific immune response and subsequent immunological dysfunction and decline). The majority of CMV seropositive patients in both cohorts (BC and PC) were also amongst patients who had high RILA scores; an indication of many terminal cells underwent apoptosis. CMV seropositive patients have an increased risk of CMV infection reactivation when undergoing intensified immunosuppressive therapies. It is conceivable that these immunosuppressive effects, such as decreased proliferation of T-cells to CMV (Chidrawar et al., 2009), may cause subclinical CMV reactivation in infected patients. Kuo et al. (2008) noted that in patients with head and neck cancers, lung cancers and rectal cancer, the risk of CMV reactivation in these patients receiving conventional chemotherapy is high. Of the 15 patients, Kuo et al. (2008) observed 14 patients experienced CMV reactivation.
Breast cancer patients
It was then investigated to what extent CMV serostatus influenced RILA predictive power in relation to radiotoxicity. Thus, the dataset was then split in cases based on CMV serostatus and correlation assessed whether RILA predicted toxicity within the various cohort population (BC and PC). When the dataset was split based on CMV serostatus, Pearson product-moment correlation coefficient indicated there was a non-significant trend with relation to RILA and acute toxicity and one-year toxicity (Table 11) irrespective of CMV serostatus. This result implies that radiation-induced lymphocyte apoptosis was triggered via other means. In this cohort, a positive correlation was observed between RILA and relative repair. This relationship was investigated once more. Interestingly, there was a positive correlation between RILA and relative repair in CMV seropositive patients (Figure 4.7). This indicated that CMV serostatus influence DNA repair capacity which subsequently affects the efficacy in DNA repair and induce apoptosis. This positive correlation might be due to a reactivation of the CMV infection within these patients. However, blood samples in this present study was taken prior both chemotherapy and radiotherapy. Therefore, CMV analysis would need to be conducted in blood samples drawn post-treatment.
Prostate cancer patients
In PC patient cohort, there was a different trend. There was no significant trend amongst many of the toxicity endpoints and RILA irrespective of CMV serostatus (as shown in Table 12). However, there was a significant trend in relation to urinary STAT and RILA predictive power. It is noteworthy that the RILA assay shown to predict urinary STAT within this cohort population (Figure 4.6.2). Interestingly, when the dataset was split based on CMV serostatus, however, urinary toxicity was not observed in CMV seropositive patient despite an increase number of cells undergoing apoptosis as assessed by RILA (Figure 4.6.3). This implies CMV infection in this cohort of prostate cancer patients protects against toxicity (i.e. these patients were hyposensitive to toxicity – CMV seropositive patients were at a minimal risk of suffering toxicity). This trend is in opposition to what was observed amongst CMV seropositive BC patients. Neote et al. (1993) found elevated levels of CMV IgG preceded the development of breast cancer in a Norwegian cohort. This might be the reason why CMV seropositive BC patients had worse toxicity. CMV infection in BC patients might be involved in tumour progression. Oseguera et al. (2017) study found elevated CMV levels in the serum of some BC patients correlated with poor prognosis. CMV infection was associated with promoting malignant spread of BC cell.
This study was subject to a series of limitations. First, BRCA1 and BRCA2 mutations were not accounted for in this cohort population as these mutations can influence inter-individual responses. BRCA1 functions to maintain genomic stability, manage the activation of DNA damage-induced cell-cycle checkpoint, and this protein is involved in repairing damage DNA along with initiation of apoptosis. Lastly, anti-CMV ELISA testing results were not duplicated due to time constraints. In addition, the results observed in this study should be validated in a different cohort. It would be of interest to investigate other viruses such as Hepatitis C, Herpes simplex virus and Epstein-Barr virus. Epstein-Barr virus, in particular, not only can establish persistent infection thus can alter T-cell function in patients, but this is an oncovirus. Consequently, this virus can severely hamper cell functions and might increase risk of normal tissue radiotoxicity in seropositive patients.
The objective of normal tissue sensitivity testing is to develop biological tests for individualizing radiotherapy dose prescriptions. That is, to permit dose escalations without increased normal tissue complication rates in patients with more resistant normal tissues or reduction of toxicity in radiosensitive patients. It is evident that toxicity from radiotherapy is a grave problem for cancer patients. The prospective study design allowed for a standardized data collection on adverse reactions of radiotherapy at defined points during radiotherapy, and the use of a predefined classification system comparable to other studies. In this cohort of unselected breast cancer patients receiving radiotherapy of the breasts, body mass index and smoking status were associated with risk of acute toxicity. In this study of randomly selected breast cancer and prostate cancer patients, there was no correlation of radiation response in vivo as assessed by the CTCAE scale with either an increased induced DNA damage level or relative repair in alkaline Comet assay. However, there was a positive association between relative repair and radiation-induced lymphocyte apoptosis in the breast cancer cohort. In addition, this study showed CMV serostatus correlated with radiation-induced lymphocyte apoptosis (i.e. CMV seropositivity increased radiation-induced lymphocyte apoptosis). Furthermore, CMV serostatus correlated with one-year toxicity in breast cancer patients.
Cite This Work
To export a reference to this article please select a referencing stye below:
Related ServicesView all
DMCA / Removal Request
If you are the original writer of this essay and no longer wish to have the essay published on the UK Essays website then please: