The use of artificial-intelligence neural networks in the evaluation of treatment plans for external-beam radiotherapy

The purpose of our study was to evaluate whether a neural network is capable of evaluating different treatment plans for external radiotherapy produced by a 3D treatment planning system and presented as dose-volume histograms (DVHs). Three radiotherapists evaluated 27 treatment plans for the external radiotherapy of prostatic adenocarcinoma. DVHs for the dose delivered to the rectum and the bladder were presented. A commercially available neural network with 5x10 input nodes and two output nodes was modified to categorize the plans according to the score of the radiotherapists. The DVHs of the treatment plans were used as the inputs and accepted or not accepted were presented as the outputs. A comparison was made with different models for assessing complication probabilities. The neural network was able to accept or not accept the treatment plans according to the scoring made by the radiotherapists. If the radiotherapists disagreed, the network also expressed the span of opinions. Neural networks can be adapted to evaluate 3D dose-planning treatment plans presented as DVHs. It should be noted that the relation between the amount of data and the size of the neural network in this study was not optimal.     

Use of benchmark dose-volume histograms for selection of the optimal technique between three-dimensional conformal radiation therapy and intensity-modulated radiation therapy in prostate cancer

PURPOSE: The aim of this study was to develop and validate our own benchmark dose-volume histograms (DVHs) of bladder and rectum for both conventional three-dimensional conformal radiation therapy (3D-CRT) and intensity-modulated radiation therapy (IMRT), and to evaluate quantitatively the benefits of using IMRT vs. 3D-CRT in treating localized prostate cancer. METHODS AND MATERIALS: During the implementation of IMRT for prostate cancer, our policy was to plan each patient with both 3D-CRT and IMRT. This study included 31 patients with T1b to T2c localized prostate cancer, for whom we completed double-planning using both 3D-CRT and IMRT techniques. The target volumes included prostate, either with or without proximal seminal vesicles. Bladder and rectum DVH data were summarized to obtain an average DVH for each technique and then compared using two-tailed paired t test analysis. RESULTS: For 3D-CRT our bladder doses were as follows: mean 28.8 Gy, v60 16.4%, v70 10.9%; rectal doses were: mean 39.3 Gy, v60 21.8%, v70 13.6%. IMRT plans resulted in similar mean dose values: bladder 26.4 Gy, rectum 34.9 Gy, but lower values of v70 for the bladder (7.8%) and rectum (9.3%). These benchmark DVHs have resulted in a critical evaluation of our 3D-CRT techniques over time. CONCLUSION: Our institution has developed benchmark DVHs for bladder and rectum based on our clinical experience with 3D-CRT and IMRT. We use these standards as well as differences in individual cases to make decisions on whether patients may benefit from IMRT treatment rather than 3D-CRT.     

Dose-Volume Histograms for bladder and rectum

: A careful examination of the foundation upon which the concept of the Dose-Volume Histogram (DVH) is built, and the implications of this set of parameters on the clinical application and interpretation of the DVH concept has not been conducted since the introduction of DVHs as a tool for the quantitative evaluation of treatment plans. The purpose of the work presented herein is to illustrate problems with current methods of implementing and interpreting DVHs when applied to hollow anatomic structures such as the bladder and rectum.

: A typical treatment plan for external beam irradiation of a patient with prostate cancer was chosen to provide a data set from which DVH curves for both the bladder and rectum were calculated. The two organs share the property of being shells with contents that are of no clinical importance. DVHs for both organs were computed using a solid model and using a shell model. Typical treatment plans for prostate cancer were used to generate DVH curves for both models. The Normal Tissue Complication Probability (NTCP) for these organs is discussed in this context.

: For an eight-field conformal treatment plan of the prostate, a bladder DVH curve generated using the shell model is higher than the corresponding curve generated using the solid model. The shell model also has a higher NTCP. A six-field conformal treatment plan slo results in a higher DVH curve for the shell model. A treatment plan consisting of bilateral 120-degree arcs, results in a higher DVH curve for the shell model, as well as a higher NTCP.

: The DVH concept currently used in evaluation of treatment plans is problematic because current practices of defining exactly what constitutes “bladder” and “rectum.” Commonly used methods of tracing the bladder and rectum imply use of a solid structure model for DVHs. In reality, these organs are shells and the critical structure associated with NTCP is obviously and indisputably the shell, as opposed to its contents. Treatment planning algorithms for DVH computation should thus be modified to utilize the shell model for these organs.     

Emptying the rectum before treatment delivery limits the variations of rectal dose–volume parameters during 3DCRT of prostate cancer

PURPOSE: To investigate the impact of rectum motion on dose - volume histograms of the rectum including filling and of the wall (DVH and DWH, respectively), during 3D-conformal radiotherapy (3DCRT) for localized prostate cancer. MATERIALS AND METHODS: Ten patients received a planning CT scan (CT(0)) and 11-14 CT during 3DCRT for prostate cancer (total CT scans=126). CT images were 3D matched using bony anatomy. A single observer drew the external contours of rectum and rectum wall and the CTV (prostate + seminal vesicles) on CT(0). Patients were asked to empty their rectum before every CT, as generally performed at the Institute for Cancer Research and Treatment (IRCC) before treatment delivery. Bladder was kept full by drinking 500 cm(3) of water 60 min before the scan, according to our protocol. A 4-field box 3DCRT technique was planned and dose statistics/dose - volume histograms of the rectum were calculated for each contour referred to CT(0),CT(1),...,CT(n) for each patient. Average DVHs during treatment were calculated along with their standard deviation (SD(rand)) and compared to the planned DVH. The analyses on the patient population included the assessment of systematic deviation (average difference and SD, named SD(sys)) as well as the average SD(rand) value expressing the random component of organ motion. Rectum shifts were also assessed by anterior and lateral BEV projections. RESULTS: As to the rectum, 8/10 patients showed a "better" average DVH than DVH on CT(0). Wilcoxon test showed a statistically significant reduction when correlating the difference Delta between the average DVH during therapy and planning DVH at CT(0): for instance V(70)Delta = -3.6% and p = 0.022, V(50)Delta = -5.5% and p = 0.022, D(med)Delta = -3.2 Gy and p = 0.007. Average values of DVH systematic difference (average difference between planning scan and treatment), standard deviations (SD(sys)) and average standard deviations of the random fluctuation (SD(random)) were -4.0%, 4.7% and 6.6%, respectively. Whilst the fluctuation results were slightly smaller for DWH. Volume analysis showed a slight systematic variation of the rectal volume between planning and treatment BEV. The average rectal volume during therapy was larger than at the planning CT in 8/10 patients. The systematic shifts of the rectal wall between the planning phase and the treatment were rather small, both below and above the flexure. The larger random fluctuation of the rectum shape was found to be in the cranial half (1 SD=4.4 mm). CONCLUSIONS: The practice of carefully emptying the rectum during simulation and therapy for prostate cancer, which is a safe and simple procedure, reduces the impact of organ motion on dose - volume parameters of the rectum.     

Normal tissue complication probabilities: dependence on choice of biological model and dose-volume histogram reduction scheme

PURPOSE: To evaluate the impact of dose-volume histogram (DVH) reduction schemes and models of normal tissue complication probability (NTCP) on ranking of radiation treatment plans. METHODS AND MATERIALS: Data for liver complications in humans and for spinal cord in rats were used to derive input parameters of four different NTCP models. DVH reduction was performed using two schemes: "effective volume" and "preferred Lyman". DVHs for competing treatment plans were derived from a sample DVH by varying dose uniformity in a high dose region so that the obtained cumulative DVHs intersected. Treatment plans were ranked according to the calculated NTCP values. RESULTS: Whenever the preferred Lyman scheme was used to reduce the DVH, competing plans were indistinguishable as long as the mean dose was constant. The effective volume DVH reduction scheme did allow us to distinguish between these competing treatment plans. However, plan ranking depended on the radiobiological model used and its input parameters. CONCLUSIONS: Dose escalation will be a significant part of radiation treatment planning using new technologies, such as 3-D conformal radiotherapy and tomotherapy. Such dose escalation will depend on how the dose distributions in organs at risk are interpreted in terms of expected complication probabilities. The present study indicates considerable variability in predicted NTCP values because of the methods used for DVH reduction and radiobiological models and their input parameters. Animal studies and collection of standardized clinical data are needed to ascertain the effects of non-uniform dose distributions and to test the validity of the models currently in use.     

Rectal planning risk volume correlation with acute and late toxicity in 3-dimensional conformal radiation therapy for prostate cancer

The purpose of this study was to evaluate rectum motion during 3-Dimensional conformal radiation therapy (3D-CRT) in prostate cancer patients, to derive a planning volume at risk (PRV) and to correlate the PRV dose-volume histograms (DVH) with treatment complications.This study was conducted in two phases. Initially, the PRV was defined prospectively in 50 consecutive prostate cancer patients (Group 1) who received a radical course of 3-D CRT. Then, the obtained PRV was used in the radiotherapy planning of these same 50 patients plus another 59 prostate cancer patients (Group 2) previously treated between 2004 and 2008. All these patients' data, including the rectum and PRV DVHs, were correlated to acute and late complications, according to the Common Toxicity Criteria (CTC) v4.0.The largest displacement occurred in the anterior axis. Long-term gastrointestinal (GI) complications grade ≥ 2 were seen in 9.2% of the cases. Factors that influenced acute GI reactions were: doses at 25% (p 5 0.011) and 40% (p 5 0.005) of the rectum volume and at 40% of the PRV (p 5 0.012). The dose at 25% of the rectum volume (p 5 0.033) and acute complications ≥ grade 2 (p 5 0.018) were prognostic factors for long-term complications. The PRV DVH did not correlate with late toxicity. The rectum showed a significant inter-fraction motion during 3D-CRT for prostate cancer. PRV dose correlated with acute gastrointestinal complications and may be a useful tool to predict and reduce their occurrence.     

Comparison of dose-volume histograms and dose-wall histograms of the rectum of patients treated with intracavitary brachytherapy.

The correlation between dose values from dose-volume histograms (DVHs) and dose values from dose-wall histograms (DWHs) of the rectum tissue of patient with uterine cervix cancer was determined. The minimum dose in 2 cm3 in the high-dose region of the DVH is a good estimate of the dose in the rectum wall.     

CyberKnife radiotherapy for localized prostate cancer: rationale and technical feasibility

There is a clear dose response for localized prostate cancer radiotherapy and there probably is a radiobiological rationale for hypo-fractionation. Combining the two should maximize tumor control and increase the therapeutic ratio. This study examines the rationale and technical feasibility of CyberKnife radiotherapy (a robotic arm-driven linear accelerator) for localized prostate cancer. Its ability to deliver non-coplanar non-isocentric arcs can yield maximally conformal isodoses. It is the only integrated system capable of target position verification and real-time tracking during delivery of conformal stereotactic radiotherapy. Inverse planning with the CyberKnife is used to design a course of radiotherapy for localized prostate cancer. Fiducial markers within the gland are used to verify organ position and track organ motion via an orthogonal pair of electronic x-ray imaging devices and provide real-time feedback correction to the robotic arm during delivery. Conformal isodose curves and dose volume histograms (DVH) are used to compare with an optimized Intensity-Modulated Radiotherapy (IMRT) plan actually delivered to the study patient based upon CT scan-derived organ volumes. The CyberKnife can produce superior DVHs for sparing of rectum and bladder and excellent DVHs for target coverage compared with IMRT, and possesses dose heterogeneities to the same degree as IMRT plans. Because of the significantly longer delivery times required it would be best suited for hypo-fractionated regimens. Such dose regimens might allow for biologically equivalent dose escalation without increased normal tissue toxicity. Since the CyberKnife can verify organ position and motion and correct for this in real-time it is the ideal means of achieving such excellent DVHs without a compromise in doses to normal tissues. These capabilities are essential if one contemplates hypo-fractionated regimens with large dose-per-fraction sizes (>5Gy to 10Gy) and dose-escalation.     

On the use of margins for geometrical uncertainties around the rectum in radiotherapy planning.

BACKGROUND AND PURPOSE: To derive planning organ at risk volume (PRV) margins for the rectum and to analyse the impact of such margins on rectum dose volume histograms (DVHs). PATIENTS AND METHODS: Weekly repeat computer tomography (CT) scans of 19 bladder cancer patients acquired during a conformal radiotherapy course were registered with the corresponding planning CT scans. From these scans, the internal rectal motion was quantified, and the margins that had to be added to the rectum contour in the planning scan to encompass the observed span of rectum motion were determined. These margins were compared to the margins derived using a recent PRV margin recipe. To illustrate the impact of margins on rectum DVHs, the margins were applied in treatment plans of six prostate cancer patients. RESULTS: Altogether 141 CT scans were analysed. On average 24% of the repeat scan rectum volume was displaced outside the planning scan contours, and wall movements of up to 30 mm were observed. Margins of 16 mm anterior and 11 mm posterior encompassed all rectal motion except for the two most displaced rectum walls in each of these directions, in 89% of the patients. Using a recently published statistics-based recipe, margins of 6 mm anterior and 5 mm posterior accounted for the systematic rectum variation, i.e. the average wall position, in 90% of the patients. Adding anterior margin only caused consistent increases (up to 20%) in the fraction of the volume inside the high-dose region (40-70 Gy) compared to the DVH of rectum only. When using both anterior and posterior margins only small shifts (<5%) in the volume fractions were observed. CONCLUSIONS: Rectum PRV margins of 5-6 mm will encompass the systematic component of rectum motion, while margins up to 16 mm are required to also account for most of the random variation. Use of anterior margins only caused large shifts in the DVHs in the clinically significant dose range, while only minor shifts were seen when using both anterior and posterior margins.     

The role and strategy of IMRT in radiotherapy of pelvic tumors: Dose escalation and critical organ sparing in prostate cancer

PURPOSE: To investigate the intensity-modulated radiotherapy (IMRT) strategy in dose escalation of prostate and pelvic lymph nodes. METHODS AND MATERIALS: Plan dosimetric data of 10 prostate cancer patients were compared with two-dimensional (2D) or IMRT techniques for pelvis (two-dimensional whole pelvic radiation therapy [2D-WPRT] or IM-WPRT) to receive 50 Gy or 54 Gy and additional prostate boost by three-dimensional conformal radiation therapy or IMRT (3D-PBRT or IM-PBRT) techniques up to 72 Gy or 78 Gy. Dose-volume histograms (DVHs), normal tissue complication probabilities (NTCP) of critical organ, and conformity of target volume in various combinations were calculated. RESULTS: In DVH analysis, the plans with IM-WPRT (54 Gy) and additional boost up to 78 Gy had lower rectal and bladder volume percentage at 50 Gy and 60 Gy, compared with those with 2D-WPRT (50 Gy) and additional boost up to 72 Gy or 78 Gy. Those with IM-WPRT (54 Gy) also had better small bowel sparing at 30 Gy and 50 Gy, compared with those with 2D-WPRT (50 Gy). In NTCP, those with IM-WPRT and total dose of 78 Gy achieved lower complication rates in rectum and small bowel, compared with those of 2D-WPRT with total dose of 72 Gy. In conformity, those with IM-WPRT had better conformity compared with those with 2D-WPRT with significance (p < 0.005). No significant difference in DVHs, NTCP, or conformity was found between IM-PBRT and 3D-PBRT after IM-WPRT. CONCLUSIONS: Initial pelvic IMRT is the most important strategy in dose escalation and critical organ sparing. IM-WPRT is recommended for patients requiring WPRT. There is not much benefit for critical organ sparing by IMRT after 2D-WPRT.     

Impact of volume and location of irradiated rectum wall on rectal blood loss after radiotherapy of prostate cancer

PURPOSE: To identify dose-volume parameters related to late rectal bleeding after radiotherapy for prostate cancer. MATERIALS AND METHODS: Clinical complication data from a randomized trial were collected and linked to the individual dose-volume data. In this trial, patients with prostate cancer were treated with either conventional (with rectangular fields) or three-dimensional conformal radiotherapy to a dose of 66 Gy. Patient complaints, including rectal blood loss, were collected for 199 patients, using questionnaires. Absolute and relative dose-volume histograms (DVHs) of the rectal wall (with and without the anal region) were calculated with and without rectal filling. A proportional hazard regression (PHR) model was applied to estimate the probability of any rectal blood loss within 3 years, as a function of several DVH parameters. In a multivariable analysis, dose-volume parameters were tested together with patient- and treatment-related parameters (age, smoking, diabetes, cardiovascular disease, tumor stage, neo-adjuvant androgen deprivation, conformal vs. conventional and rectal bleeding during treatment). RESULTS: The estimated incidence of any and moderate/severe rectal bleeding at 3 years was 33% and 8%, respectively. Differences between the conventional and conformal technique were small and not significant. The analysis of relative DVHs of the rectal wall (with and without the anal region), showed significant (p < 0.01) relations between the irradiated volume and the probability of rectal blood loss within 3 years for dose levels between 25 Gy and 60 Gy. This relationship was shown in subgroups defined by dose-volume cutoff points as well as in the PHR model, in which a continuously rising risk was seen with increasing volumes. For absolute DVHs and DVHs of the rectum including filling, less or no significant results were observed. The most significant volume-effect relation (p = 0.002) was found at 60 Gy for the rectum wall excluding the anal region. The probability of rectal bleeding increased from 10% to 63% when the irradiated rectum volume at 60 Gy increased from 25% to 100%. Other factors. including age, smoking, diabetes, cardiovascular disease, tumor stage, neo-adjuvant androgen deprivation, conformal vs. conventional, rectal bleeding during treatment, rectum length. and whole rectum volume. did not have a significant effect in the multivariable analysis. When controlling for the volumes at 60 Gy, the volumes at lower dose levels (25-55 Gy) were no longer significant (p = 0.5). CONCLUSIONS: For any rectal bleeding within 3 years, an overall incidence of 33% was observed for patients treated to 66 Gy. For this endpoint, a volume-effect relation was found for DVH parameters of the relative rectal wall volume. This relationship appeared to be most significant for the rectum without the anal region and for the higher dose levels (50-60 Gy).     

Bone marrow doses and leukaemia risk in radiotherapy of prostate cancer.

BACKGROUND AND PURPOSE: As more and more patients with prostate cancer are cured and survive with only minor chronic morbidity, other potentially treatment related morbidity, in particular second cancers, becomes an urgent problem which may influence decisions on treatment strategy and treatment plan optimisation. Epidemiological data suggest a radiotherapy associated risk of AML in prostate cancer patients of approximately 0.1% in 10 years. The aim of the study was to determine the range of bone marrow doses from different treatment plans and in different patients in order to develop criteria for optimisation of treatment plans in conformal radiotherapy of prostate cancer to further minimise the small risk of secondary leukaemia. MATERIALS AND METHODS: Doses to the pelvic bone marrow were calculated for eight different plans used in radiotherapy of prostate cancer to determine the variability of bone marrow doses in radiotherapy of prostate cancer. Computer tomography (CT) slices of the entire pelvic region of an Alderson phantom were acquired and transferred to the TPS. Critical bone marrow structures were outlined in each slice. Different treatment plans were evaluated on this phantom and dose-volume histograms (DVH) for the pelvic bone marrow were obtained. Similarly, the DVH for the bone marrow of 14 patients who received conformal radiotherapy for prostate cancer was determined. RESULTS: Mean total bone marrow doses ranged from 3.4 to 5.6 Gy in the phantom study. Approximately 99% of the mean dose to the total bone marrow comes from the dose to bone marrow located in the pelvic bones and lumbar vertebrae. Mean bone marrow doses of 14 patients given the same conformal radiotherapy plan ranged from 3.5 to 7.7 Gy. CONCLUSIONS: No correlation was found between the rectum normal tissue complication probability (NTCP) and the mean bone marrow dose. This means that in the process of treatment planning, exposure to both critical organs, the rectum as well as the bone marrow, should be minimised independently to arrive at the optimal treatment plan.     

Potential role of intensity modulated proton beams in prostate cancer radiotherapy

PURPOSE: The present study was undertaken to assess the potential benefit of intensity modulated (IM) proton beams in optimizing the dose distribution to safely escalate the tumor dose in prostate cancer radiotherapy. METHODS AND MATERIALS: Four treatment plans were compared in a prostate cancer patient aiming to deliver 81 Gy to the target: 1) conformal 18 MV X-rays, 6-fields; 2) 214 MeV protons, 2-fields; 3) IM 15 MV X-rays, 5-fields; and 4) 177-200 Mev IM protons, 5-fields as in Plan 3. In addition, IM methods were used to further escalate the tumor dose to 99 Gy. Dose-volume histograms (DVH) were used to physically compare the treatment plans. DVH data were also used to obtain normal tissue complication probabilities (NTCP) for the rectum, bladder, femoral heads, and tumor control probabilities. RESULTS: Although the planning target volume dose distribution was satisfactory with the four treatment plans, the homogeneity was slightly reduced in both X-ray plans (IM and standard) and the low-to-medium doses delivered to all organs at risk, and other normal tissues were significantly reduced by both proton plans. For a prescribed dose of 81 Gy, only the IM X-ray and IM proton plans both succeeded in predicting an acceptably low NTCP for the rectum (<5%, Grade 3). The integral nontarget dose was significantly reduced with IM proton beams (i.e., 3.1, 1.3, and 1.7 times less than Plans 1, 2, and 3, respectively). When escalating the dose to 99 Gy, no additional improvement between IM protons and IM X-ray beams was observed. CONCLUSION: Both IM X-ray and proton beams were able to optimize the dose distribution and comply with the goal of delivering the highest dose to the target while reducing the risk of severe morbidity to acceptable levels. The main advantage compared to IM X-rays was that IM protons succeeded in significantly reducing the low-to-medium dose to the nontarget tissues and achieved a small improvement in planning target volume (PTV) dose heterogeneity.     

Dose-volume histograms

A plot of a cumulative dose-volume frequency distribution, commonly known as a dose-volume histogram (DVH), graphically summarizes the simulated radiation distribution within a volume of interest of a patient which would result from a proposed radiation treatment plan. DVHs show promise as tools for comparing rival treatment plans for a specific patient by clearly presenting the uniformity of dose in the target volume and any hot spots in adjacent normal organs or tissues. However, because of the loss of positional information in the volume(s) under consideration, it should not be the sole criterion for plan evaluation. DVHs can also be used as input data to estimate tumor control probability (TCP) and normal tissue complication probability (NTCP). The sensitivity of TCP and NTCP calculations to small changes in the DVH shape points to the need for an accurate method for computing DVHs. We present a discussion of the methodology for generating and plotting the DVHs, some caveats, limitations on their use and the general experience of four hospitals using DVHs.     

Rectal sequelae after conformal radiotherapy of prostate cancer: dose-volume histograms as predictive factors.

PURPOSE: To identify clinically relevant parameters predictive of late rectal bleeding derived from cumulative dose-volume histograms (DVHs) of the rectum after conformal radiotherapy of prostate cancer. MATERIALS AND METHODS: One hundred and nine patients treated with 3D conformal radiotherapy between 1/1994 and 1/1996 for localized prostate cancer (clinical stage T1-T3) were available for analysis. All patients received a total dose of 66 Gy/2 Gy per fraction (specified at the International Commission on Radiation Units and Measurements ICRU reference point). DVHs of the contoured rectum were analyzed by defining the absolute (aV) and relative (rV) rectum volume that received more than 30% (V30), 50% (V50), 70% (V70), 80% (V80), 90% (V90) and 100% (V100) of the prescribed dose. Additionally, a new aspect of DVH analysis was investigated by calculation of the area under the DVH-curve between several dose levels (area under the curve (AUC)-DVH). DVH-variables were correlated with radiation side effects evaluated in 3-6 months intervals and graded according to the EORTC/RTOG score. The median follow-up was 30 months (12-60 months). RESULTS: Univariate and multivariate stepwise Cox-Regression analysis including age, PTV, rectum size, rV100, rV90, rV80, rV70, rV50 rV30 and aV30 to aV100 were calculated. Late rectal bleeding (EORTC/RTOG grade 2) was significantly correlated with the percentage of rectum volume receiving > or = 90% of the prescribed dose (rV90) (P = 0.007) and inversely correlated in a significant way with the size of contoured rectum (P = 0.006) in multivariate analysis. In our series, a proportion of the rectum volume > or = 57% were included in the 90%-isodose (rV90 > or = 57%) in one half of the patients, with an actuarial incidence of 31% of late rectal bleeding at 3 years. In the other half of the patients, when rV90 < 57%, the 3-year actuarial incidence was 11% (P < 0.03). CONCLUSION: Our data demonstrate a dose-volume relationship at the reference dose of 60 Gy ( approximately 90% of the prescribed dose) with respect to late rectal toxicity. The rV90 seems to be the most useful and easily obtained parameter when comparing treatment plans to evaluate the risk of rectal morbidity.     

Finding dose-volume constraints to reduce late rectal toxicity following 3D-conformal radiotherapy (3D-CRT) of prostate cancer.

BACKGROUND AND PURPOSE: The rectum is known to display a dose-volume effect following high-dose 3D-conformal radiotherapy (3D-CRT). The aim of the study is to search for significant dose-volume combinations with the specific treatment technique and patient set-up currently used in our institution. PATIENTS AND METHODS: We retrospectively analyzed the dose-volume histograms (DVH) of 135 patients with stage T1b-T3b prostate cancer treated consecutively with 3D-CRT between 1996 and 2000 to a total dose of 76 Gy. The median follow-up was 28 months (range 12-62). All late rectal complications were scored using RTOG criteria. Time to late toxicity was assessed using the Kaplan-Meyer method. The association between variables at baseline and > or=2 rectal toxicity was tested using chi(2) test or Fisher's exact test. A multivariate analysis using logistic regression was performed. RESULTS: Late rectal toxicity grade > or=2 was observed in 24 of the 135 patients (17.8%). A 'grey area' of increased risk has been identified. Average DVHs of the bleeding and non-bleeding patients were generated. The area under the percent volume DVH for the rectum of the bleeding patients was significantly higher than that of patients without late rectal toxicity. On multivariate analysis the correlation between the high risk DVHs and late rectal bleeding was confirmed. CONCLUSIONS: The present analysis confirms the role of the rectal DVH as a tool to discriminate patients undergoing high-dose 3D-CRT into a low and a high risk of developing late rectal bleeding. Based on our own results and taking into account the data published in the literature, we have been able to establish new dose-volume constraints for treatment planning: if possible, the percentage of rectal volume exposed to 40, 50, 60, 72 and 76 Gy should be limited to 60, 50, 25, 15 and 5%, respectively.     

Comparison of the conventional 'box technique' with two different 'conformal' beam arrangements for prostate cancer treatment.

PURPOSE: To quantify the possible advantages arising from the use of 'conformal' radiotherapy of localized prostate cancer, and to compare the dose distributions obtained with two different 'conformal' techniques. PATIENTS AND METHODS: Twelve patients with localized prostate cancer were enrolled in the study. For each patient, three techniques were planned: the standard 'box technique' (A), a four-fields 'conformal' technique (B), and a 6-fields conformal technique (C). For each of the 36 3D plans, dose-volume histograms (DVH) were obtained, along with the mean, maximum and minimum doses for the clinical and planning target volumes (CTV, PTV) for the rectum, the bladder, and the femoral heads. The resulting data were compared. RESULTS: On average, the standard technique resulted in the exposure of a significantly larger bladder volume to the higher doses; a similar, but less remarkable difference has been observed for the rectal volume. The coverage of the PTV appears to be significantly more homogeneous with the two conformal techniques. CONCLUSIONS: The results presented here add to the evidence available in the literature and suggest a possible advantage of both the conformal techniques over the standard 'box technique' for the treatment of localized prostate cancer. The 6-field conformal technique does not seem superior to the four field one.