BJR
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
First published online August 2, 2006
British Journal of Radiology (2006) 79, 905-911
© 2006 British Institute of Radiology
doi: 10.1259/bjr/18700614
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Wang, J Z
Right arrow Articles by Mayr, N A
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Wang, J Z
Right arrow Articles by Mayr, N A

Full paper

Dose escalation to combat hypoxia in prostate cancer: a radiobiological study on clinical data

J Z Wang, PhD 1 X A Li, PhD 2 and N A Mayr, MD 1

1 Department of Radiation Medicine, The Ohio State University, Columbus, OH 43210, 2 Department of Radiation Oncology, Medical College of Wisconsin, Milwaukee, WI 53226, USA

Correspondence: Jian Z Wang, Department of Radiation Medicine, James Cancer Hospital and Solove Research Institute, The Ohio State University, 300 W 10th Ave, Rm 094, Columbus, OH 43210, USA. E-mail: wang.993{at}osu.edu


   Abstract
Top
 Abstract
Introduction
 Materials and methods
 Results
 Conclusion and discussion
 References
 
Earlier studies have demonstrated that hypoxic regions exist in human prostate cancer and the degree of hypoxia correlates with the treatment outcome of radiotherapy. Using the concept of the clinical oxygen enhancement ratio (COER), the linear-quadratic (LQ) model was extended to account for the effect of tumour hypoxia. The clinical data collected at the Fox Chase Cancer Center for prostate cancer were analysed based on the LQ model as well as the tumour control probability (TCP) model. The LQ and TCP parameters ({alpha} = 0.15 Gy –1, {alpha}/beta = 3.1 Gy and the number of clonogens K = 106~107 cells) determined in earlier studies were used to derive the COER for prostate cancer: COER = 1.4 with a standard confidence interval (CI) of (1.2, 1.8). The result is consistent with the in vitro OER measurements of human tumour cell lines under chronic hypoxia conditions. This implies that a higher dose is needed to overcome tumour hypoxia. For prostate tumours, the prescription dose required to overcome tumour hypoxia is 165 Gy (CI: 153~186 Gy) for permanent 125I implants and 88 Gy (CI: 74~118 Gy) in 2 Gy fractions for external-beam radiotherapy. The impact of LQ parameters on the calculations of COER and dose escalation was discussed. This study provides a preliminary estimate of the dose escalation needed to overcome tumour hypoxia based on clinical data. More clinical data with better statistics and longer follow-up time are required to further tune the radiobiological modelling of hypoxia for prostate cancer.


   Introduction
Top
 Abstract
 Introduction
Materials and methods
 Results
 Conclusion and discussion
 References
 
Hypoxia has been identified in many human cancers, including head and neck, lung, breast, cervix, bladder and prostate cancer, and the adverse impact of hypoxia on clinical tumour control has been well recognized [1–10].

Recently, Movsas et al [1–3] and Parker et al [4] have demonstrated that hypoxic regions do exist in human prostate carcinoma and the treatment outcome of radiation therapy correlates with the degree of hypoxia. Oxygenation status (oxygen pressure, PO2) may thus be another underlying biological parameter beyond the classic prognostic factors (clinical stage, Gleason score and prostate specific antigen (PSA)) predicting a critical component of treatment failure in prostate cancer. Movsas et al [1] reported a significant difference in the biochemical control rate at 2 years (31% versus 92%) between two patient groups stratified by the prostate/muscle (P/M) PO2 ratio (<0.05 versus ≥0.05). Many other investigators have also found similar correlations between hypoxia and tumour progression [5–10]. It is generally believed that high-risk tumours tend to exhibit more severe hypoxia [3, 10]. Although it is well established that hypoxic cells/regions are less sensitive to radiation and that tumour hypoxia can result in clinical radiation therapy failure, it remains unclear how to overcome hypoxia in the clinical setting. Specific quantitative dose guidelines are needed that can be applied to the clinical setting, particularly to dose escalation trials aimed to combat hypoxia in prostate cancer.

The linear-quadratic (LQ) and tumour control probability (TCP) models extended to address tumour hypoxia were used to analyse the clinical data. Recently, the {alpha}/beta ratio of the LQ model for prostate cancer has become a highly debated topic in the radiation therapy community [11–22]. By taking into account the effect of tumour repopulation, Wang et al [14, 15] and Kal and Van Gellekom [19] took two different approaches, and obtained similar results of {alpha}/beta ratio (around 3.1 Gy). Based on the clinical data of external-beam radiotherapy (EBRT) and low- and high-dose rate (LDR/HDR) brachytherapy [12, 13, 23], Wang et al [14, 15] have derived a new set of LQ parameters ({alpha} = 0.15 Gy–1, {alpha}/beta = 3.1 Gy and repair half-time Tr = 16 min) with estimated numbers of clonogens around 106~107, depending on the patient risk level. The new results provide reasonable estimates of radiosensitivity and the number of clonogens for human prostate tumours [14, 15, 24–26]. The radiobiological modelling with these parameters provides a consistent interpretation for most currently available clinical data, including EBRT, permanent brachytherapy, HDR brachytherapy and their combinations [14, 15, 26].

Fowler et al [16, 21] argued that for slowly growing tumours (e.g. prostate cancer), the onset of tumour repopulation may be delayed significantly to over 200 days. Wang et al [17, 22] pointed out that there is no strong biological evidence for such arguments. The clinical results reported so far have been controversial. In a recently reported study, Perez et al [27] did observe an impact of elapsed treatment time on outcome of EBRT for T2 prostate tumours, implying the onset time of repopulation should be relatively short. Therefore in our study, the original assumption of the onset time of tumour repopulation will be followed for various sets of LQ parameters. Recently, the effect of hypoxia has been taken into account by Nahum et al [18] in their modelling study. However, their model parameters ({alpha}, beta and oxygen enhancement ratio (OER)) were completely based on in vitro measurements, which may not be fully applicable to the clinical micro-environment of prostate cancer in vivo [28]. Tumour heterogeneity of radiosensitivity has been studied [20] and due to the large uncertainty existing in the current clinical data and the large number of parameters used, it is not possible to give a reasonable estimate of {alpha}/beta ratio for any statistical significance using the heterogeneous model. For this reason, we chose the homogeneous model in this study.

The purpose of our study is to quantify the dose escalation to address tumour hypoxia in prostate cancer. The LQ and TCP models are used to interpret the clinical data recently reported from Fox Chase [1] based on the concept of OER. The specific aim of this analysis is to derive the clinical OER (COER) and to determine the dose escalation necessary to offset the hypoxia effect using either permanent brachytherapy or EBRT.


   Materials and methods
Top
 Abstract
 Introduction
 Materials and methods
Results
 Conclusion and discussion
 References
 
LQ and TCP models for prostate cancer
The general LQ model [29] was used in this study. In this model, the surviving fraction S of cells irradiated to a total dose D within an overall treatment time T is given by:


Formula 001

where {alpha} and beta characterize intrinsic radiosensitivity, G is the dose-rate factor accounting for sub-lethal damage repair, {gamma} is the effective tumour repopulation rate [{gamma} = ln(2)/Td, and Td is the effective clonogen doubling time]. A median potential tumour doubling time of 42 days measured in vivo from 7 prostate cancer patients was used [30]. This repopulation rate corresponds to a biological effective dose (BED) of 0.11 Gy per day (for {alpha} = 0.15 Gy –1), compatible with the results of slow-growing tumours presented in a repopulation-dose-equivalent study by Jones et al [31]. The repopulation time T can be replaced by (T–Tk), where Tk is the onset time of tumour repopulation. Because the actual value of this parameter is under debate [16, 17, 21, 22, 27] and no clinical data are available, we followed the original assumptions used by different investigators when they derived the LQ parameters from clinical data, i.e. Tk = 0 for the LQ parameters by Wang et al [14, 15] and by Kal and Van Gellekom [19], and Tk>200 days (i.e. ignoring the repopulation effect) for the LQ parameters by Brenner et al [11, 12] and by Fowler et al [13].

For EBRT, the total dose D equals nd and the dose-rate factor G equals 1/n, where n is the number of dose fractions and d is the dose per fraction. The treatment time T of EBRT can be approximated as the number of treatment fractions multiplied by 1.4 (5 fractions per week). For permanent brachytherapy, D, G and T have been derived in many papers (e.g. [14, 29]) as follows:


Formula 002

where µ is the repair rate of tumour cells [µ = ln(2)/Tr], R0 is the initial dose rate (R0 = D0•{lambda}, D0 is the prescribed dose) and {lambda} is the decay constant for the implanted isotopes [{lambda} = ln(2)/Ts, Ts is the half-life time of the radioisotope]. Ts = 59.4 days for 125I implants.

The Poisson TCP model is often used to link cell killing of a radiotherapy scheme to treatment outcome [32],


Formula 003

where K is the initial number of tumour clonogens. We consider it as an effective number of clonogens associated with the patient risk level.

Two representative sets of LQ parameters derived from clinical data [11–15, 19] and one set of generic LQ parameters for tumour [10] were used in this study (Table 1Go). The numbers shown in the parentheses of Table 1Go indicate the standard confidence intervals (CI) [14] and they were used to estimate the uncertainties of the COER and dose escalations. The clonogen number K of 1.6x106 (CI: 5.6x104~8.8x107) derived from clinical data [14] was used in the calculations with the first parameter set (see the following subsection "Fox Chase clinical data"). Due to limited data for the other two parameter sets, the clonogen numbers were arbitrarily set to match the clinical data and the CIs were not considered. Our calculation was mainly based on the first parameter set. The later two parameter sets were used to test the model dependency of the final results presented in this study.


View this table:
[in this window]
[in a new window]
  		Table 1. Radiobiological parameters used in this study

 
LQ model extended for tumour hypoxia
OER is a dose-modifying parameter, which is defined as the ratio of doses needed to produce the same biological effect for cells without and with oxygen. Traditionally, OER was commonly used to describe the cell survival curves obtained in in vitro measurements. The OER of X-ray irradiation for most mammalian cells in vitro is about 3 at high doses and is possibly lower (about 2) at doses below approximately 2 Gy [10]. In this work, we apply the OER concept to clinical data (i.e. in vivo) to characterize tumours with or without hypoxia [5, 33–35]. Compared with in vitro experiments, the hypoxia in in vivo situations may be quite different. The population for tumours in vivo always reflects a spectrum of hypoxia. Even for well-identified hypoxic tumours, the hypoxic cells may exist only in sub-volumes of tumours. Therefore, the OER obtained retrospectively from clinical data represents only a statistical average and may be different from the corresponding in vitro values for the same type of tumours. Zolzer and Streffer [36] argued that tumours in vivo should be dominated by chronic hypoxia and the radiosensitivity for cells under chronic hypoxia conditions would be higher than those under acute hypoxia conditions. They reported that the OER measurements of two human tumour cell lines under continued/chronic hypoxia conditions (1.3 and 1.5 for MeWo and squamous carcinoma 4451, respectively) were quite different from the OER values measured under acute hypoxia conditions. To distinguish this feature, we term herewith the OER obtained from clinical data as clinical OER (COER). For simplicity, the COER is represented by H throughout this paper.

Given a certain amount of clonogens, to achieve the same TCP for two patient groups, the cell killing efficiency should be the same. That is:


Formula 004

where subscripts h and o label the parameters for tumours with and without hypoxia, respectively. Assuming Dh and Do represent the doses to achieve the same TCP for the two patient groups, respectively, we have Dh = H•Do. To be consistent with Equation (4), we assume that {alpha}h = {alpha}o/H, betah = betao/H2, and they are valid to apply to the LDR 125I brachytherapy and the EBRT with low dose fractions (d = 2 Gy). It appears that the hypoxia impacts more significantly on the quadratic term than on the linear term. Such relationships of the LQ parameters between aerated and hypoxic cells have been observed in in vitro experiments of low dose irradiation (0.5–3 Gy) for Chinese hamster V79-171 cells [37, 38]. Therefore, the LQ model extended to the hypoxia effect can be expressed as (Sh for hypoxic cells):


Formula 005

Similar formulae to describe the hypoxia effect can be found in the literature [5, 33–35].

Because of the limited clinical data, we will focus on the effect of hypoxia on radiosensitivity only, and neglect the influence of hypoxia on repair time or repopulation rate. Hill et al presented data of animal and human carcinoma of the cervix showing no evidence for a difference in repopulation kinetics between aerated or hypoxic tumours [39, 40]. It is generally assumed that the transient hypoxia can be overcome by tumour reoxygenation by protracting radiotherapy (i.e. fractionated EBRT or LDR brachytherapy). The reoxygenation effect has been modelled in previous studies (e.g. [41]). Therefore, it is believed that the oxygen effect on the clinical outcome observed in the 125I data (see next sub-section) should be mainly due to the chronic hypoxia [36]. In this work, we use the above formulae to study the effect of chronic hypoxia.

For patients with tumours that are only partially hypoxic, i.e. containing oxygenated and hypoxic subvolumes, the two-compartment model was used to calculate the overall surviving fraction [18, 33, 35],


Formula 006

where fh is the hypoxic fraction of cells in the tumour and So is for aerated cells. Animal tumour data and human tumour data indicate that the hypoxic fraction ranges from 0 to 50%, with an average value of 15% [5, 10, 42]. This average value along with the range from 5% to 50% was used in this study.

The uncertainty in estimating the COER originates from the uncertainties of the LQ/TCP parameters (as shown in Table 1Go), the hypoxia fraction and the clinical data. The LQ/TCP parameters are strongly correlated to each other and constrained by clinical data; therefore, their uncertainties are not independent from each other [14] and have been considered in this study. The code autoEUD described in Wang and Li [26] was used to propagate the different uncertainties to determine the CIs of COER and dose escalation. For independent parameters, their contributions to the final CIs were calculated based on the root of the quadratic-summation.

For a given patient group of prostate cancer, based on Equations (1) to (6), we could calculate the TCP for a given radiation treatment or determine the required radiation dose D to achieve a given TCP. The software AutoEUD and the typical dose–volume histogram (DVH) of EBRT and brachytherapy described in a previous paper [26] were used to automate these calculations.

Fox Chase clinical data
Since 1999, Movsas et al have published a series of papers to address the hypoxia of prostate cancer [1–3]. For the first time, they demonstrated the existence of hypoxia in human prostate cancer based on the in vivo electrode measurements of oxygen levels, and that the degree of hypoxia in prostate cancer correlated to the treatment outcome of radiotherapy. 57 patients with localized prostate cancer were included in their study [1]. Before radiotherapy treatments, custom-made Eppendorf microelectrodes were used to measure the PO2 in both pathologically involved regions of the prostate and normal muscle (as an internal control). Real-time ultrasound imaging was used to guide the microelectrode measurement. For each patient, approximately 100 PO2 readings were obtained along 3 to 5 tracks in both prostate and muscle [3]. Following the Eppendorf PO2 measurement, 48 patients received permanent 125I brachytherapy. The dose prescribed to the prostate was 145 Gy. Nine patients were treated on a dose escalation protocol involving a 46 Gy (in 23 daily fractions) of EBRT to the pelvic region plus two HDR boost implants (8.75 Gy or 9.75 Gy per fraction). Hormonal therapy prior to radiation therapy was given to nine patients. In their study, the biological failure was defined as two consecutive rises in PSA without a return to baseline. With a median follow-up time of 19 months (ranging from 4 months to 31 months), the overall 2 year bNED (biologically no evidence of disease) was estimated as 81% for the entire patient group (as shown in Figure 1Go of this paper and Figure 2Go of [1]). In order to study the impact of hypoxia, the P/M ratio of PO2, which eliminates the potential interpatient and technical variations, was used to analyse the treatment outcome.


Figure 1
View larger version (18K):
[in this window]
[in a new window]
  		Figure 1. 2 year biologically no evidence of disease(bNED) versus prescription dose of 125I brachytherapy. The symbols represent the Fox Chase clinical data [1] and the curves represent the model calculations. The three points (circle, square and triangle), as well as the three curves (solid, dashed and dash-dotted), are for patient groups with hypoxic, mixed and aerated prostate tumours, respectively. The prescription dose of the clinical data was 145 Gy of 125I implant. The vertical dotted line indicates the prescription dose for the hypoxic group to achieve a bNED of 81%. The vertical error bar shows the standard confidence interval (CI) of the clinical data for the hypoxic group (the CIs for the other two groups are not shown) and the horizontal error bar shows the CI of dose escalation for the hypoxic group to achieve a target bNED of 81%.

 

Figure 2
View larger version (17K):
[in this window]
[in a new window]
  		Figure 2. (a) Tumour control probability (TCP) as a function of the fraction of tumour hypoxia (fh) for individual patients of prostate cancer. The prescription dose is 145 Gy for 125I brachytherapy. (b) Dose escalation of 125I brachytherapy required overcoming tumour hypoxia (to achieve a TCP of 81%) as a function of hypoxia fraction fh.

 
The Fox Chase experience demonstrated that the P/M ratio strongly correlated to the bNED outcome. A threshold analysis showed that the 2 year bNED differed significantly at a P/M ratio of <0.05 (hypoxic group) versus ≥0.05 (aerated group): 31% versus 92% (p<0.0001) (see Figure 1Go). There were 12 and 45 patients in the two groups, respectively. The classic prognosis factors, including clinical stage, Gleason score, pre-treatment PSA and age at treatment were similar in the two patient groups. Based on the numbers of all patients and the patients at risk for each group, using the bi-nominal distribution, we estimated the standard CI of the 2 year bNED for the hypoxic tumour group to be (15%, 47%). Because of the large numbers of patients in the aerated tumour group, the uncertainty of their 2 year bNED is negligible.

According to the patient characteristics of the three prognosis factors, patients of the Fox Chase data were stratified into the low- or intermediate-risk groups. Because most patients presented low-risk prognosis factors, the clonogen number (1.6x106) of low-risk patients was used for this study [14, 15]. Two radiotherapy modalities (125I brachytherapy vs combined EBRT and HDR) were used in the Fox Chase Study. Since the patient number of the combined modality (EBRT + HDR) is relatively small (9 patients) and the treatment outcome for low-risk patients is close for the two regimens, we ignored the difference of the two modalities and combined the data in the following study. For similar reasons, we also ignored the influence of hormone therapy applied to nine patients out of the total 57 patients. A representative DVH obtained from permanent 125I implant (Figure 1Go of Ref. [26]) was used to account for the dose inhomogeneity of the Fox Chase data.


   Results
Top
 Abstract
 Introduction
 Materials and methods
 Results
Conclusion and discussion
 References
 
Clinical OER for hypoxic prostate cancer
The clinical data and the model calculation are summarized in Figure 1Go. It has been reported that, based on LQ and TCP models, the 145 Gy 125I implant has a TCP of 82% for low-risk prostate patients [14, 26]. This calculated TCP value is consistent with the clinically observed 2 year bNED (81%) for the mixed patient group in the Fox Chase study (see "Fox Chase clinical data" in the Materials and Methods section). Using the bNED data for the aerated and hypoxic groups, we derived the COER and the CI for the hypoxic tumours: H = 1.4 (CI: 1.2~1.8) (Table 2Go). This result is very consistent with the in vitro OER measurements of two human tumour cell lines (1.3 and 1.5 for MeWo and squamous carcinoma 4451, respectively) under chronic/continued hypoxia conditions [36]. The large uncertainty is mainly related to the limited information for the hypoxic fraction in the hypoxic tumours. If the uncertainty of hypoxic fraction is not considered, the uncertainty of COER is greatly reduced to±0.13, i.e., in the range of (1.27, 1.53).


View this table:
[in this window]
[in a new window]
  		Table 2. Impact of radiobiological parameters on the calculation of COERs and dose escalations

 
Similar calculations, except for the uncertainty estimation, were performed for other sets of LQ parameters. In the COER derivation with the parameters of Brenner et al (i.e. {alpha} = 0.04 Gy –1, {alpha}/beta = 1.5 Gy, Tr = 1.9 h with repopulation ignored), only the hypoxic clonogens were considered [43]. The COER obtained with this parameter set is H = 1.5. Using the LQ parameters that are considered generic for tumours (i.e. {alpha} = 0.3 Gy –1, {alpha}/beta = 10 Gy, Tr = 1 h and Td = 42 days), we obtained the COER: H = 1.2. Interestingly, the result derived with the parameters of Wang et al lies between the values obtained with these two sets of LQ parameters, and the differences among the COER values obtained with the three parameter sets are smaller than the CIs determined by various uncertainties. More clinical data with better statistics and better knowledge of the hypoxic fraction are required to further narrow down the CIs. To summarize the results, the COERs, as well as the required dose escalations (see the following subsection), derived with different LQ/TCP parameters are listed in Table 2Go.

Dose escalation for hypoxic tumours
Based on the derived COER, the dose required to compensate for tumour hypoxia was calculated using the code autoEUD [26]. To bring the TCP for hypoxic tumours to the same level as that for the general population of prostate patients (TCP = 81%), the prescription dose required for permanent 125I implant should be 169 Gy. This dose may be used to target the hypoxia region if it can be identified. Otherwise, a reduced dose of 165 Gy (CI: 153~186 Gy) should be prescribed to the entire prostate, including the non-hypoxic regions, to achieve the same treatment outcome (a TCP of 81%) as shown in Figure 1Go. This dose escalation is practically feasible, as demonstrated by a recent study [44].

The calculations performed so far provide a guide of dose escalation for a cohort of patients with hypoxic prostate tumours. If the hypoxic subvolume of individual tumours can be detected, for example by biological/functional imaging, the data presented may also be used to target these subvolumes and design the individualized radiation plan. Figure 2Go shows the TCP and dose prescription as a function of the hypoxic fraction fh for individual prostate tumours. The TCP presented in Figure 2aGo is for the dose prescription (145 Gy of 125I implant) used in the current clinical practice. Figure 2bGo shows the dose escalation required to overcome the effect of hypoxia and to bring the TCP to 81% as that of the general population of prostate cancer patients. For individual patients with hypoxic fraction changing from 5% to 50%, the prescription dose is increased from 157 Gy to 181 Gy.

The above results obtained from the Fox Chase data (brachytherapy) data set may be used to derive the dose escalation of EBRT. Various clinical studies have shown that a biological equivalence exists between EBRT and permanent brachytherapy for prostate cancer [13, 26]. For patients with similar risk-levels, the 145 Gy brachytherapy yielded a similar treatment outcome as 71 Gy EBRT in 2 Gy fractions [13]. Based on the LQ and TCP models and the COER derived in this study, we calculated the EBRT dose required to compensate for the effect of hypoxia. This EBRT dose is found to be 88 Gy (CI: 74~118 Gy) in 2 Gy daily fractions.

Similarly, based on the COERs derived with the parameters of Brenner et al or with the generic LQ parameters for tumours (see Table 1Go), we obtained the dose escalation necessary to compensate for prostate hypoxia. With Brenner et al's parameters, the prescription dose would be 193 Gy for 125I brachytherapy and 109 Gy for EBRT. With the generic tumour parameters, these doses would be 155 Gy for the 125I brachytherapy and 102 Gy for EBRT. Because the clinical data used to derive the COERs are from 125I brachytherapy, the dose escalations calculated for the 125I brachytherapy have smaller CIs and they match the sequence of the {alpha}/beta ratio in the three parameter-sets. However, the extrapolated EBRT doses show large uncertainty and the results based on the parameters of Brenner et al and the generic tumour parameters are unusually high; therefore, the prediction of EBRT dose escalation has limited clinical value.


   Conclusion and discussion
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Conclusion and discussion
References
 
In this study, we have analysed the reported clinical data for hypoxic prostate cancer. Based on the LQ and TCP models extended to account for tumour hypoxia, the clinical OER for prostate tumour was obtained: COER = 1.4 (CI: 1.2~1.8). The result agrees with the in vitro OER measurements under chronic/continued hypoxia conditions [36]. To overcome tumour hypoxia, a dose escalation to 165 Gy (CI: 153~186 Gy) (instead of 145 Gy) is required for 125I permanent implants. A similar calculation was performed for EBRT. The EBRT dose needs to be escalated from 71 Gy to 88 Gy in 2 Gy fractions in order to achieve a TCP of 81% for the hypoxic tumour group.

There is in vivo evidence showing that increasing levels of hypoxia in prostate cancer correlate significantly with the increasing clinical stage and patient age [2]. The P/M ratio of PO2 may be used as a prognostic factor for high-risk patients. Dose escalation in both EBRT and brachytherapy has been proven effective to treat high-risk patients of prostate cancer [12, 23, 26, 44, 45]. This study used similar approaches to address the unsatisfactory treatment outcome of hypoxic tumours. The required dose escalations (165 Gy for 125I implant and 88 Gy for EBRT) are practically deliverable [23, 44], and clinical trials with similar dose levels or even higher are being conducted in several institutions.

The results obtained in this work are based on the Fox Chase clinical data [1] that, for the first time, shows the existence of hypoxia in human prostate tumours and the correlation of the degree of hypoxia with treatment outcome. There are several limitations in the Fox Chase study that may compromise the accuracy in the current estimation of the COER and the corresponding target doses. First, the PO2 measurement with Eppendorf microeletrodes is a sampling method which measures only a sampled portion of the prostate. It may not reflect the overall PO2 distribution in the entire tumour. The method also fails to discriminate cell type and viability, resulting in PO2 readings from less significant tissues. Second, this method cannot distinguish the hypoxia types: chronic and transient. The transient hypoxia may be reduced or overcome by providing enough time for reoxygenation during the treatment course. Third, the study contained only a small number of patient samples with limited follow-up time, and furthermore mixed treatment modalities (two radiation methods and hormone therapy) were involved in the small data sample. Because of the limited statistical power, only two patient groups could be identified with an artificially determined threshold of the P/M ratio. More clinical data with larger numbers of patients, longer follow-up time and consistent treatment are certainly desired to further refine our conclusions.

Because of the limited clinical data, it was not possible to derive independent relationships between {alpha}, beta and COER in this study. Instead, the relationships, {alpha}h = {alpha}o/H, betah = betao/H2, from the literature [5, 33–35] were used. Although similar relationships have been observed in in vitro measurements [37, 38], the results obtained in this analysis would be affected if such relationships were not applicable to prostate cancer. The COER calculation was based on an average hypoxic fraction of 15% [5, 10, 42]. A change of this value would change the results of the COER as well as the dose escalation. There are also concerns about the OER changing with the irradiation dose fraction. In this study, we assume that the COER keeps constant for the LDR brachytherapy and the low-dose-fraction (2 Gy) EBRT. The use of alternative models, which include temporal variations in the hypoxic fraction, might also lead to quite different conclusions. Therefore, the results obtained in this study may be limited by the approximations used in the models and the uncertainties shown in the clinical data. Caution needs to be exercised in using the presented data for clinical decision-making purposes.

Received for publication January 4, 2006. Revision received June 2, 2006. Accepted for publication June 5, 2006.


   References
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Conclusion and discussion
 References
 

  1. Movsas B, Chapman JD, Hanlon AL, et al. Hypoxic prostate/muscle Po2 ratio predicts for biochemical failure in patients with prostate cancer: preliminary findings. Urology 2002;60:634–9.[CrossRef][Medline]
  2. Movsas B, Chapman JD, Greenberg RE, et al. Increasing levels of hypoxia in prostate carcinoma correlate significantly with increasing clinical stage and patient age. Cancer 2000;89:2018–24.[CrossRef][Medline]
  3. Movsas B, Chapman JD, Horwitz EM, et al. Hypoxic regions exist in human prostate carcinoma. Urology 1999;53:11–8.[Medline]
  4. Parker CP, Milosevic M, Toi A, et al. Polarographic electrode study of tumor oxygenation in clinically localized prostate cancer. Int J Radiat Oncol Biol Phys 2004;58:750–7.[CrossRef][Medline]
  5. Okunieff P, Hoeckel M, Dunphy EP, et al. Oxygen tension distributions are sufficient to explain the local response of human breast tumors treated with radiation alone. Int J Radiat Oncol Biol Phys 1993;26:631–6.[Medline]
  6. Höckel M, Knoop C, Schlenger K, et al. Intratumoral pO2 predicts survival in advanced cancer of the uterine cervix. Radiother Oncol 1993;26:45–50.[CrossRef][Medline]
  7. Höckel M, Schlenger K, Mitze M, et al. Hypoxia and radiation response in human tumors. Semin Radiat Oncol 1996;6:3–9.[CrossRef][Medline]
  8. Nordsmark M, Overgaard M, Overgaard J. Oxygenation status predicts radiation response in advanced squamous cell carcinoma of head and neck. Radiother Oncol 1996;41:31–9.[Medline]
  9. Brizel DM, Sibley GS, Prosnitz LR, et al. Tumor hypoxia adversely affects the prognosis of carcinoma of the head and neck. Int J Radiat Oncol Biol Phys 1997;38:285–9.[CrossRef][Medline]
  10. Hall EJ. Radiobiology for the radiologist. 5th edn. Philadelphia, PA: Lippincott Williams and Wilkins, 2000:91–111
  11. Brenner DJ, Hall EJ. Fractionation and protraction for radiotherapy of prostate carcinoma. Int J Radiat Oncol Biol Phys 1999;43:1095–101.[CrossRef][Medline]
  12. Brenner DJ, Martinez AA, Edmundson GK, et al. Direct evidence that prostate tumors show high sensitivity to fractionation (low {alpha}/beta ratio), similar to late-responding normal tissue. Int J Radiat Oncol Biol Phys 2002;52:6–13.[CrossRef][Medline]
  13. Fowler J, Chappell R, Ritter M. Is {alpha}/beta for prostate cancer really low? Int J Radiat Oncol Biol Phys 2001;50:1021–31.[CrossRef][Medline]
  14. Wang JZ, Guerrero M, Li XA. How low is the {alpha}/beta ratio for prostate cancer? Int J Radiat Oncol Biol Phys 2003;55:194–203.[CrossRef][Medline]
  15. Wang JZ, Li XA, Yu CX, et al. The low {alpha}/beta ratio for prostate cancer: what does the clinical outcome of HDR brachytherapy tell us? Int J Radiat Oncol Biol Phys 2003;57:1101–8.[CrossRef][Medline]
  16. Fowler JF, Ritter MA, Fenwick JD, et al. How low is the {alpha}/beta ratio for prostate cancer? In regard to Wang et al. [letter]. Int J Radiat Oncol Biol Phys 2003;57:593–5.[Medline]
  17. Wang JZ, Guerrero M, Li XA. Low {alpha}/beta ratio for prostate cancer: in response to Dr. Fowler et al. [letter]. Int J Radiat Oncol Biol Phys 2003;57:595–6.[CrossRef]
  18. Nahum AE, Movsas B, Horwitz EM, et al. Incorporating clinical measurements of hypoxia into tumor local control modeling of prostate cancer: Implications for the {alpha}/beta ratio. Int J Radiat Oncol Biol Phys 2003;57:391–401.[CrossRef][Medline]
  19. Kal HB, Van Gellekom MPR. How low is the {alpha}/beta ratio for prostate cancer? Int J Radiat Oncol Biol Phys 2003;57:1116–21.[CrossRef][Medline]
  20. Carlone M, Wilkins D, Nyiri B, et al. Comparison of {alpha}/beta estimates from homogeneous (individual) and heterogeneous (population) tumor control models for early stage prostate cancer. Med Phys 2003;30:2832–48.[CrossRef][Medline]
  21. Dasu A, Fowler JF. Comments on 'Comparison of in vitro and in vivo {alpha}/beta ratios for prostate cancer'. Phys Med Biol 2005;50:L1–4.[CrossRef][Medline]
  22. Wang JZ, Stewart RD, Carlson DJ, et al. Reply to "comments on 'comparison of in vitro and in vivo {alpha}/beta ratios for prostate cancer'". Phys Med Biol 2005;50:L5–8.[CrossRef]
  23. Levegrün S, Jackson A, Zelefsky MJ, et al. Fitting tumor control probability models to biopsy outcome after three-dimensional conformal radiation therapy of prostate cancer: pitfalls in deducing radiobiologic parameters for tumors from clinical data. Int J Radiat Oncol Biol Phys 2001;51:1064–80.[CrossRef][Medline]
  24. King CR, Mayo CS. Is the prostate {alpha}/beta ratio of 1.5 from Brenner and Hall a modeling artifact? [letter]. Int J Radiat Oncol Biol Phys 2000;47:536–8.[CrossRef][Medline]
  25. Wyatt RM, Beddoe AH, Dale RG. The effects of delays in radiotherapy treatment on tumor control. Phys Med Biol 2003;48:139–55.[CrossRef][Medline]
  26. Wang JZ, Li XA. Evaluation of external beam radiotherapy and brachytherapy for localized prostate cancer using equivalent uniform dose. Med Phys 2003;30:34–40.[CrossRef][Medline]
  27. Perez CA, Michalski J, Mansur D, et al. Impact of elapsed treatment time on outcome of external-beam radiation therapy for localized carcinoma of the prostate. Cancer J 2004;10:349–56.
  28. Wang JZ, Mayr NA, Li XA, et al. Modeling prostate cancer: in regard to Nahum et al. [letter]. Int J Radiat Oncol Biol Phys 2005;61:309–10.[CrossRef][Medline]
  29. Dale RG. The application of the linear-quadratic dose-effect equation to fractionated and protracted radiotherapy. Br J Radiol 1985;58:515–28.[Abstract/Free Full Text]
  30. Haustermans K, Fowler JF. A comment on proliferation rates in human prostate cancer [letter]. Int J Radiat Oncol Biol Phys 2000;48:303[Medline]
  31. Jones B, Cominos M, Dale RG. Application of biological effective dose (BED) to estimate the duration of symptomatic relief and repopulation dose equivalent in palliative radiotherapy and chemotherapy. Int J Radiat Oncol Biol Phys 2003;55:736–42.[CrossRef][Medline]
  32. Porter EH. The statistics of dose/cure relationships for irradiated tumors. Br J Radiol 1980;53:210–27.[Abstract/Free Full Text]
  33. Buffa FM, West C, Byrne K, et al. Radiation response and cure rate of human colon adenocarcinoma spheroids of different size: the significance of hypoxia on tumor control modeling. Int J Radiat Oncol Biol Phys 2001;49:1109–18.[CrossRef][Medline]
  34. Hendry JH, Thames HD. Fractionation sensitivity and the oxygen effect. Br J Radiol 1990;63:79–80.[Abstract/Free Full Text]
  35. Popple RA, Ove R, Shen S. Tumor control probability for selective boosting of hypoxic subvolumes, including the effect of reoxygenation. Int J Radiat Oncol Biol Phys 2002;54:921–7.[CrossRef][Medline]
  36. Zolzer F, Streffer C. Increased radiosensitivity with chronic hypoxia in four human tumor cell lines. Int J Radiat Oncol Biol Phys 2002;54:910–20.[CrossRef][Medline]
  37. Watts ME, Hodgkiss RJ, Jones NR, et al. Radiosensitization of Chinese hamster cells by oxygen and misonidazole at low X-ray doses. Int J Radiat Biol 1986;50:1009–21.[CrossRef]
  38. Skarsgard LD, Harrison I, Durand RE. The radiation response of asynchronous cells at low dose: evidence of substructure. Radiat Res 1991;127:248–56.[Medline]
  39. Speke AK, Hill RP. The effects of claiming and reoxygenation on repopulation during fractionated irradiation. Int J Radiat Oncol Biol Phys 1995;31:857–63.[CrossRef][Medline]
  40. Hill RP, Fyles W, Milosevic M, et al. Is there a relationship between repopulation and hypoxia/reoxygenation? Results from human carcinoma of the cervix. Int J Radiat Biol 2003;79:487–94.[CrossRef][Medline]
  41. Brenner DJ, Hlatky LR, Hahnfeldt PJ, et al. A convenient extension of the linear-quadratic model to include redistribution and reoxygenation. Int J Radiat Oncol Biol Phys 1995;32:379–90.[CrossRef][Medline]
  42. Jones B, Dale G. Estimation of tumor hypoxia fraction from clinical data sets compatible with accelerated repopulation. Acta Oncologica 1998;37:263–8.
  43. Brenner DJ, Hall EJ. In response to Drs. King and Mayo: Low {alpha}/beta values for prostate appear to be independent of modeling details [letter]. Int J Radiat Oncol Biol Phys 2000;47:2753–65.
  44. Li XA, Wang JZ, Stewart RD, et al. Dose escalation in permanent brachytherapy for prostate cancer: dosimetric and biological considerations. Phys Med Biol 2003;48:2753–65.[CrossRef][Medline]
  45. Stock RG, Stone NN, Tabert A, et al. A dose-response study for I-125 prostate implants. Int J Radiat Oncol Biol Phys 1999;41:101–8.



This article has been cited by other articles:


Home page
 Br. J. Radiol.Home page
BJR review of the year - 2006
Br. J. Radiol., March 1, 2007; 80(951): 147 - 151.
[Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Wang, J Z
Right arrow Articles by Mayr, N A
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Wang, J Z
Right arrow Articles by Mayr, N A

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
BJR DMFR IMAGING  ALL BIR JOURNALS