Advice

Evidence review

Evidence review

Clinical and technical evidence

Regulatory bodies

Two incidents were identified in the US Food and Drug Administration (FDA) database: Manufacturer and User Device Facility Experience (MAUDE), both of which were attributed to system malfunctions. Neither of the incidents resulted in any patient being harmed. In the first, Oncentra Prostate v3.09 malfunctioned 6 times during treatment, resulting in a delay to the patient's second brachytherapy fraction. In the second incident, a radiotherapy treatment plan was optimized twice with the dose volume histogram‑based optimization algorithm, with inconsistent results. The version of the system used during this incident is not known (MAUDE Adverse Event Report: Nucletron BV Oncentra Prostate [SWIFT] treatment planning).

Clinical evidence

Many of the advanced radiotherapy planning features of the system included in v4.x were initially licensed in v3.x, so this evidence review focuses on the Oncentra Prostate v3.x and above. A literature search identified 1 study of the Oncentra Prostate v4.x and 3 of the Oncentra Prostate v3.x:

  • Gomez‑Iturriaga et al. (2014, table 4)

  • Pokharel et al. (2013, table 5)

  • Choudhury et al. (2014, table 6)

  • Adamczyk et al. (2013, table 7).

In addition, 7 conference abstracts investigating the Oncentra Prostate system were identified, although the version used was not always reported. These abstracts are summarised and presented in table 3.

Feasibility studies

Three feasibility studies of the Oncentra Prostate system were identified. All 3 of these were small, single‑centre studies analysing technical parameters of radiotherapy planning, mainly the radiation dose delivered to the prostate and the surrounding healthy organs at risk of radiation exposure.

Pokharel et al. (2013) investigated the effectiveness of the hybrid inverse planning and optimization (HIPO) algorithm for real‑time HDR brachytherapy in 20 men. The HIPO algorithm is a feature of v3.x and v4.x of the Oncentra Prostate system. It is a user‑independent optimisation algorithm, which automatically calculates the best distribution of the applicators (dwell positions and dwell times) to create the radiotherapy treatment plan. Pokharel et al. compared the HIPO algorithm with the graphical optimization (GRO) algorithm, which needs manual adjustment of the isodose lines (these are virtual lines which the software draws on the clinical images to show the areas of tissue that receive different levels of the prescribed radiation dose) to create the radiotherapy treatment plan. The study found that HIPO can provide the same prescribed dose to the prostate as GRO, but with a statistically significant reduction in the dose received by the surrounding organs at risk (urethra, bladder and rectum).

Adamczyk et al. (2013) presented a small, single‑centre study of prostate cancer in 15 men. The effect of 3 different automatic optimisation algorithms on treatment plan quality during 3D‑conformal real‑time HDR brachytherapy was investigated. The 3 algorithms were geometrical optimization, inverse optimization and blind inverse optimization, which are all available with Oncentra Prostate v3.0.9. The authors concluded that the blind inverse optimization algorithm achieved adequate radiation dose for the prostate and the surrounding organs at risk of radiation damage. Although the geometrical optimization algorithm delivered smaller doses to the urethra, this was also associated with a lower therapeutic dose to the prostate.

Gomez‑Iturriaga et al. (2014) reported results from a single‑centre study on the effect of an intraoperative MRI/transrectal ultrasound (TRUS) fusion procedure in 9 men with extracapsular extension prostate cancer (clinical tumour stage 3a) who had real‑time HDR brachytherapy. Real‑time MRI/TRUS fusion can be done using the 'Fusion for all modalities' module of the Oncentra Prostate v4.x. The authors analysed men with intermediate‑ and high‑risk prostate cancer who had a single HDR fraction of 15 Gy followed by EBRT at a dose of 37.5 Gy delivered in 15 fractions over 3 weeks. For each participant, 2 virtual treatment plans were developed based on the MRI/TRUS fusion images. These treatment plans were only generated to test the capabilities of the software, and were not used for treatment. According to their findings, MRI contributed additional information for the prediction of extracapsular extension disease and the authors concluded that it should be used for high‑risk patients having HDR brachytherapy. One of the authors is also the principal investigator of an ongoing prospective study investigating the feasibility of MRI/TRUS fusion during dose‑escalation prostate brachytherapy (ClinicalTrials.gov identifier: NCT01909388), but it should be noted that the Gomez‑Iturriaga et al. (2014) study is not a preliminary report of the data from this ongoing prospective trial.

Cohort study

Choudhury et al. (2014) provides a large UK‑based prospective series reporting patient‑reported outcomes and health‑related quality of life for patients having HDR brachytherapy with the Oncentra Prostate v3.x and hypofractionated EBRT. All patients had androgen‑deprivation therapy before and after their radiotherapy treatment. They had a single brachytherapy fraction of 12.5 Gy and 15 fractions of EBRT at 37.5 Gy, starting 2 weeks after brachytherapy. Patients completed postal questionnaires at various time points before and after their treatment. There were no statistically significant associations between dosimetric parameters and patient‑reported outcomes. The radiotherapy treatment schedule used was associated with a temporary effect on health‑related quality of life and acceptable rates of urinary and bowel side effects.

Table 3 Summary of abstracts reporting data on the Oncentra Prostate system

Study

Population

Intervention

Outcome measure

Conclusion

Challapalli et al. (2011)

30 men with high risk localised/ locally advanced prostate cancer.

Oncentra Prostate (version number was not described).

All men had EBRT (46 Gy) + HDR (12.5 Gy).

Treatment time and dose coverage of the CTV.

Real‑time ultrasound planning was a safe and well‑tolerated method for increasing the prescribed radiation dose to the prostate whilst not exceeding the dose tolerance of the surrounding organs at risk.

Filipowski et al. (2011)

190 men with stage T1–T2c prostate cancer, mean age=68 years, mean PSA=11.069 ng/ml, Gleason score range 3–7 .

SWIFT 2.11.8 and Oncentra Prostate v3.0.9 and v4.0.

60 men who had EBRT (30 Gy) with HDR (30 Gy) and 130 with HDR (45 Gy) only.

Early treatment side effects and radiation dose received by the CTV and organs at risk.

The majority of patients tolerated the treatment well with acceptable levels of acute side effects.

Kabacinska et al. (2013)

77 men (no information on patient characteristics was provided).

Oncentra Prostate v3.2.3.

Most men had EBRT (no information on dose in Gy) with HDR (20 Gy) and the rest had HDR (31.5 Gy) only.

Radiation dose received by the CTV and organs at risk.

The use of anatomy‑based inverse optimization followed by graphic optimization during the radiotherapy treatment plan with the system fulfils almost all brachytherapy dosimetric recommendations of ESTRO and ABS guidelines.

Kanikowski (2013)

103 men (no information on patient characteristics was provided).

SWIFT and Oncentra Prostate (version number was not described).

37 men had EBRT (50 Gy) with HDR (1 session at 15 Gy), 36 men had EBRT (46 Gy) with HDR (2 sessions at 10 Gy) and 30 men had HDR (3 sessions at 15 Gy) only.

Radiation dose received by the CTV and organs at risk.

Although the same prostate dosimetric parameters were achieved with all 3 different radiotherapy treatment schedules, the HDR brachytherapy schedule using 2 brachytherapy fractions resulted in higher doses received by the organs at risk and lower doses by the prostate.

Kunhiparambath et al. (2011)

5 men (no information on patient characteristics was provided).

Oncentra Prostate (version number was not described).

Radiation dose received by the PTV and organs at risk.

Single objective dose volume histogram‑based inverse planning optimization is superior to volume‑based geometric optimization.

Laviraj et al. (2011)

4 men with (no information on patient characteristics was provided).

Oncentra Prostate (version number was not described).

Dosimetric parameters.

Better dose coverage of the prostate and almost the same dose to healthy organs at risk were achieved with the multisolution dose volume histogram‑based algorithm in comparison with the multisolution variance‑based optimization algorithm.

Schwarz et al. (2010)

36 men with intermediate‑ or high‑risk prostate cancer.

Oncentra Prostate (version number was not described).

All men had EBRT (50.4 Gy) with HDR (9 Gy).

Radiation dose received by the CTV and organs at risk.

Oncentra Prostate Nucletron brachytherapy system was successfully implemented in daily routine. The 3D‑treatment optimization showed excellent dose parameters in accordance to the radiotherapy treatment plan objectives.

Abbreviations: ABS, American Brachytherapy Society; CTV, clinical target volume; ESTRO, European Society for Radiology and Oncology; PTV, planning target volume.

Table 4 Summary of the Gomez-Iturriaga et al. (2014) single‑centre feasibility study

Study component

Description

Objectives/ hypotheses

To evaluate the impact of intraoperative MRI/TRUS fusion in patients with cT3a prostate cancer treated with high‑dose‑rate real‑time brachytherapy.

Study design

A retrospective, single‑centre, feasibility study.

Setting

Spain

Inclusion/exclusion criteria

Inclusion criteria:

  • proven adenocarcinoma of the prostate;

  • clinical (MR imaging) stage T3a disease and

  • without clinical or radiographic evidence of metastases

Primary outcomes

All outcomes were presented as mean figures.

  • Prostate dose volume histograms

  • Extracapsular extension (ECE)

  • Maximal urethral and rectum dose

Statistical methods

Comparisons of the mean values between the 2 plans were done using the paired t‑test. Significance was defined as a probability value less than 0.05, and no adjustment was made for multiple comparisons.

Participants

9 consecutive patients participated in this study, mean age=68 years (range 60–78), intermediate‑risk=6, high‑risk=2, very high risk=1.

Results

Mean radial distance of ECE was 3.6 mm (SD: 1.1).

No significant differences were found between prostate V100, V150, V200, and OARs DVH‑related parameters between the plans.

Mean values of ECE V100, V150, and V200 were 85.9% (SD: 15.1), 18.2% (SD: 17.3) and 5.85% (SD: 7) respectively when the doses were prescribed to the PTVUS.

Mean values of ECE V100, V150, and V200 were 99.3% (SD: 1.2), 45.8% (SD: 22.4) and 19.6% (SD: 12.6) respectively when doses were prescribed to PTVMR

(p=0.028, p=0.002 and p=0.004 respectively).

Conclusions

TRUS/MRI fusion provides valuable information for prostate brachytherapy, allowing delivery of a higher dose and better target coverage of extracapsular disease in patients with clinical stage T3a.

Abbreviations: CI, confidence interval; DVHO, dose‑volume histogram–based optimization; GRO, graphical optimization; HIPO, hybrid Inverse treatment planning; ITT, intention to treat; n, number of patients; RR, relative risk; TRUS, transrectal ultrasound.

Table 5 Summary of the Pokharel et al. (2013) single‑centre feasibility study

Study component

Description

Objectives/ hypotheses

To investigate the effectiveness of the HIPO planning and optimization algorithm for real‑time prostate HDR brachytherapy.

Study design

A retrospective, single‑centre, feasibility study.

Setting

USA‑based.

Recruitment period: March 2007 to October 2009.

Inclusion/exclusion criteria

No specific criteria were presented.

Primary outcomes

  • PTV

  • PTV to OARs (urethra, rectum, bladder, and normal tissue)

  • HI

  • COIN.

Statistical methods

The paired student t‑test (p<0.05) was used to make statistical comparisons of different dosimetric quality indices of treatment plans optimized by different optimization algorithms.

The statistical comparisons were:

  • HIPO1 vs GRO

  • HIPO1 vs HIPO2 

  • HIPO1 vs DVHO.

The comparison between HIPO and DVHO with the same weighting factors was carried out to investigate the importance of dwell position optimization.

Participants

20 patients: the authors do not provide any further information regarding patient characteristics.

Results

The PTV receiving 100% of the prescription dose (V100) was 95.38% with HIPO and 97.56% with GRO. The mean dose and minimum dose to 10% volume for the urethra, rectum, and bladder were all statistically lower with HIPO compared with GRO using the paired student t‑test at 5% significance level.

Conclusions

HIPO can provide treatment plans with comparable target coverage to that of GRO with a reduction in dose to the critical structures.

Abbreviations: CI, confidence interval; COIN, conformal index; DVHO, dose‑volume histogram–based optimization; GRO, graphical optimization; HI, homogeneity index; HIPO, hybrid inverse treatment planning; ITT, intention to treat; n, number of patients; RR, relative risk.

Table 6 Summary of the Choudhury et al. (2014) single‑centre cohort study

Study component

Description

Objectives/ hypotheses

To investigate the effect of HDR brachytherapy combined with EBRT on patient‑reported outcomes and health‑related quality of life.

Study design

A prospective, single‑centre, cohort study.

Patients completed postal questionnaires after an initial consultation (baseline), immediately before attending for HDR brachytherapy (pre‑treatment), and after EBRT at 6 weeks and 6, 12, 18, 24 and 36 months.

Setting

The Christie NHS Foundation Trust, UK.

Recruitment period: July 2008 to March 2010.

Inclusion/exclusion criteria

Patients presenting with histologically confirmed intermediate‑ or high‑risk prostate cancer according to the D'Amico classification were considered eligible for HDR brachytherapy combined with hypofractionated EBRT. All patients were staged using cross‑sectional pelvic imaging and an isotope bone scan to exclude metastases.

Exclusion criteria: prostate‑specific antigen>100, IPSS>20, history of previous transurethral resection of the prostate, and contraindications to general anaesthesia.

No specific prostate volume constraints were applied within this group of patients.

Primary outcomes

Patient‑reported toxicity data:

  • IPSS

  • LENT‑SOMA questionnaires.

Health‑related quality of life data:

  • EPIC questionnaire.

Statistical methods

The mean and median IPSS were calculated at each time‑point. LENT‑SOMA data were presented as an overall mean and median score for each anatomical subscale in addition to the proportion of patients reporting a score of 2 for a single question within each subscale.

The EPIC questionnaire was analysed using the mean, median and interquartile range for the urinary, bowel, sexual and hormonal domains and their subdomains. A change of≤10% in the EPIC score was considered to be clinically significant.

Median scores at each time‑point were compared from baseline using the Wilcoxon matched‑pairs signed‑rank test and a 2‑sided p‑value≤0.01 (Bonferroni correction made for multiple comparisons).

Spearman's rank coefficient was used to investigate any associations between patient clinical factors and dose parameters at HDR planning and subsequent toxicity.

Participants

95 men with intermediate‑ or high‑risk prostate cancer, median age=68 (range=51–78) years, median prostate‑specific antigen=16.7 (range=0.29–90) ng/ml.

Results

95 men had an HDR boost of 12.5 Gy followed by EBRT delivered as 37.5 Gy in 15 sessions over 3 weeks.

The IPSS peaked 6 weeks after radiotherapy (median=9).

The LENT‑SOMA bladder/urethra mean baseline score was 0.35 and peaked 6 weeks after radiotherapy (mean=0.59).

Difficulties with urinary flow and frequency were the most common reported symptoms.

LENT‑SOMA rectum/bowel mean scores at baseline were 0.24 and peaked after 6 months (mean=0.37).

Bowel urgency was the most common reported toxicity.

EPIC urinary scores returned to baseline values at 6 months and bowel median scores recovered after 24 months.

There were no statistically significant associations between patient or dosimetric parameters and patient‑reported outcomes.

Conclusions

A combined HDR boost and hypofractionated EBRT regimen offers a well‑tolerated method of dose escalation with acceptable levels of patient reported

toxicity.

Abbreviations: CI, confidence interval; , EPIC, Expanded Prostate Cancer Index Composite; IPSS, International Prostate Symptom Score; ITT, intention to treat; LENT‑SOMA, Late Effects in Normal Tissues‑Subjective, Objective, Management and Analytic scales; n, number of patients; RR, relative risk

Table 7 Summary of the Adamczyk et al. (2014) single‑centre feasibility study

Study component

Description

Objectives/ hypotheses

To present the effect of different optimization algorithms (BIO and GO) on treatment plan quality during 3D‑conformal real‑time HDR brachytherapy.

Study design

A retrospective, single‑centre, feasibility study.

Setting

Poland.

Inclusion/exclusion criteria

The authors provided no information on inclusion/exclusion criteria.

Primary outcomes

Differences between dose distributions were tracked using: D90, V100, V200, Dmax (for prostate); D10, Dmax (for urethra); D10, V100, Dmax (for rectum).

Statistical methods

The analysis of each index was done by dividing the data into 3 groups depending on the algorithm used in the optimization process. Statistical differences between groups were verified using t‑test and Wilcoxon's test.

Differences between groups were considered significant if the p‑value was <0.05.

Participants

15 patients (no information provided on patient characteristics).

Results

The analysis of mean values of D90 and V100 in the prostate showed that inverse algorithms gave the best results (mean D90 was 12.1% greater for BIO and 9.3% greater for IO compared with GO, mean V100 was 8.2% greater for BIO and 6.3% greater for IO compared with GO).

From a clinical point of view, GO diminished the doses in the PTV and urethra in all analysed parameters. The lowest mean doses in the rectum were achieved for plans optimized with IO and BIO (mean D10: 61.2% for GO, 58.1% for IO, 58.0% for BIO; mean Dmax: 92.8% for GO, 85.1% for IO, 83.6% for BIO).

Conclusions

Application of the BIO algorithm led to clinically best dose parameters for PTV and the rectum. Use of GO led to smaller doses in the urethra, which was however associated with a dose decrease in the PTV.

Abbreviations: BIO, blind inverse optimization; CI, confidence interval; D10, dose covering 10% of the urethral or rectal volume; D90, the dose (in Gy or as a percentage of the prescription dose) that covers 90% of the prostate volume; Dmax, the maximum dose received by the prostate and organs at risk; GO, geometrical optimization; ITT, intention to treat; n, number of patients; PTV, planning target volume; RR, relative risk; V100, the fractional volume of the prostate that receives 100% of the prescription dose; V200, the fractional volume of the prostate that receives 200% of the prescription dose.

Recent and ongoing studies

One ongoing trial using the Oncentra Prostate v4.x was identified (ClinicalTrials.gov identifier: NCT01909388). The BRAPOST trial (Dose Escalation to Dominant Intraprostatic Lesions With MRI‑TRUS Fusion HDR Prostate Brachytherapy) started in July 2013 and at the time of writing is recruiting patients. The estimated completion date is July 2016.

Costs and resource consequences

No published evidence on resource consequences of the Oncentra Prostate was identified in the systematic review of evidence.

The Oncentra Prostate v4.x is intended to replace the use of CT imaging in the planning and delivery of HDR brachytherapy treatment, and this would have the greatest resource and planning consequences for brachytherapy units that currently do not use real‑time single‑step brachytherapy. These consequences include potentially decreased procedural time and releasing CT and MRI imaging facilities for use by other patients.

The manufacturer states that 7 NHS clinical oncology departments are currently using the Oncentra Prostate system.

Strengths and limitations of the evidence

Real‑time ultrasound‑guided HDR brachytherapy with the Oncentra Prostate system is designed for men who have been diagnosed with intermediate‑ or high‑risk localised prostate cancer. Although the patients included in the Gomez‑Iturriaga et al. (2014) and Choudhury et al. (2014) studies were men with intermediate‑ or high‑risk localised prostate cancer, Pokharel et al. (2013) and Adamczyk et al. (2013) do not provide any information regarding the risk status of the men included in their studies. As a result, it is unclear whether their study populations match the intended use of the system.

The Pokharel et al. study reported the results of a retrospective, single‑centre study of 20 men with prostate cancer. It should be noted that the small sample size and parametric statistical methods employed may reduce the statistical power of this study. The publication made no mention of data distribution or normality. Given the small sample size, these factors should have been considered or non‑parametric testing should have been used. The primary outcome measures of the study investigated the effectiveness of the HIPO algorithm in technical aspects of radiotherapy treatment planning. The absence of long‑term follow‑up means that only limited conclusions can be drawn on the effect of the achieved dosimetric parameters with the HIPO algorithm (as compared with GRO) on patient‑related outcomes including survival, biochemical control or health‑related quality of life measures. Although some data suggest a relationship between radiation dose and patient outcome measures, including survival, the strength of these conclusions is often limited by the small sample sizes of these studies, the incompleteness of the data, and the presence of bias in non‑randomised studies (van Tol‑Geerdink 2006).

The Adamczyk et al. (2013) study presented a single‑centre, small sample size (n=15) study investigating the effect of 3 different optimisation algorithms on radiotherapy treatment planning. The authors provided no information on either patient characteristics or the exact treatment that patients had. Adamczyk et al. failed to demonstrate statistically significant differences between all outcomes, which may be a result of a small sample size.

Choudhury et al. (2014) presented their single‑centre experience using a combination of EBRT and real‑time HDR brachytherapy (a boost approach), describing the toxicity and quality of life effects associated with the combination. Compared with the other studies reviewed in this briefing, this prospective single‑centre study had a larger sample size (n=95) and presents outcome measures relating to patient‑reported quality of life. There were no statistically significant associations between clinical characteristics or dosimetric parameters and patient‑reported outcomes. The patient‑reported outcome data were strengthened because the authors used a validated questionnaire (LENT‑SOMA) for their analysis, assessed the outcomes at various time points, and had a long‑term follow‑up. However, patient‑reported measures may introduce bias because of overestimation or subjectivity. The authors also noted that although questionnaires return rates were over 50% at every time point, they were not always fully completed. In this study, p‑values were adjusted to reduce the risk of false‑positive results arising simply because a large number of comparisons were made. However, this will also increase the likelihood of false‑negative results. Details of the correction and of the p‑values prior to correction were not provided.

The study by Gomez‑Iturriaga et al. (2014) was limited, because their analysis was of a small number of men and included only dosimetric parameters. Nevertheless, their prospective study was planned to report patient‑related outcomes on acute side effects, treatment tolerability and efficacy (assessed by biochemical control, MRI and biopsy results) with long‑term follow‑up at 12, 24 and 30 months after treatment.

A number of abstracts were also identified in the literature search. These abstracts did not report full information on study design or patient characteristics. It also was not always possible to confirm which version of the Oncentra Prostate was used in each study. The abstract by Filipowski et al. (2011) was the only one of these to specify that v4.x was used, but this version was not used for the entirety of the study. Generally, all abstracts included in this briefing reported improvements in radiotherapy treatment planning dosimetric parameters associated with the use of the Oncentra Prostate system. However, only limited data on patient‑related outcomes were available. Overall, the abstracts reported that HDR brachytherapy with the Oncentra Prostate was safe and well tolerated. Two deaths were reported among the 190 men enrolled in the Filipowski et al. study, but the authors provided no information as to the causes.

The Filipowski et al. abstract highlighted the need for longer follow‑up data, specifically related to survival and toxicity. This applies equally to all of the studies reviewed in this briefing.

In general, the studies included in this briefing focused on technical aspects of treatment delivery rather than patient outcomes. Choudhury et al (2014) is the only study to address clinical outcomes, but even in this case treatment tolerability rather than efficacy is measured.

The manufacturer claims that ultrasound treatment planning of real‑time single‑step HDR brachytherapy using the Oncentra Prostate system results in a reduction in the length of the procedure time compared with a 2‑step process. However, no published evidence was found to support this. The use of automated optimisation algorithms can potentially reduce the time needed to find the best radiotherapy treatment plan and improve the dosimetric parameters (Pokharel et al. 2013, Adamczyk et al. 2013), but data on the actual impact on the overall procedure time are not available.

The majority of identified evidence relates to previous versions of the Oncentra Prostate rather than the currently available v4.x. No published evidence directly comparing the performance of the v4.x with previous versions has been identified. As a result, conclusions on improved dosimetric or patient‑related outcomes between the new and older versions cannot be formed.