Hello again....

Welcome to you web surfers - sorry the text is a bit busy on the screen. I'll try make it more user friendly as time goes on........ (Ed)

Sorry to those readers who have just received their copies of this Newsletter in the post who find that the pages are garbled. A new copy will be dispatched ASAP.

Here is the second newsletter from SCRAP reporting on the recent meeting held in Nottingham. Thanks go out to Tony Bedford and all the staff at Nottingham City Hospital for all their hard work in making the meeting run very smoothly indeed. Thanks also to Shakilah Patval from RMH Fulham Road for helping to collect all the grubby £5 notes on the door from the delegates. The number of delegates attending this year's event was slightly down. I partly blame myself for not managing to circulate a detailed programme beforehand which might have enticed more punters to attend. However, I also think that brachytherapy, which formed the bulk of the meeting, is viewed as the slightly poor relation to the more glamourous IMRT (and related 3D planning techniques) and is written off as an uninteresting branch of radiotherapy physics. I would like to think that this meeting, and the renewed interest in prostate seed implants, have gone some way in reviving people's interest and enthusiasm in the subject.

The Report

The meeting for 1999 was held in the Post Graduate Education Centre at Nottingham City Hospital on Friday 2^nd July 1999. Some 35 delegates attended and the meeting was supported by an excellent turn out of manufacturers - so thanks also go to: ADAC, DHA, Prowess, Helax, Nucletron, Qados, CMS, and Varian BrachyVision. The meeting followed the format of recent years with 4 presentations given in the morning and a further 2 in the afternoon. This was followed by the roundtable discussion and finally, for those interested, the obligatory tour of the department.

The first presentation was given by Julian Perks, lately of RMH Fulham Road but now working at St. Bartholomew's Hospital. He gave a very succinct explanation of the use of ImageFusion - a piece of commercial software, linked to the Xknife suite of programmes, used for matching CT and MR images in the brain. Here, MR images are used to visualise tumours in the brain and the resulting target volume is then transferred to a set of CT scans. Why is ImageFusion necessary? Well, MR gives very good soft tissue contrast allowing the tumour to be outlined very easily. However, for accurate dosimetry it is necessary to have CT scans where the images are relatively distortion free and include electron density data. Julian stated that it was essential to have the electron density information, especially for tumours arising in the base of brain. If the head is assumed to be composed of water alone, errors of some 7% can be introduced. Distortion is not such a problem as it is of the order of 1% in the head coil of the MR scanner and so ImageFusion does not need to correct for this. To commission the system RMH Fulham Road set up a trial of 15 patients who did not receive treatment based on the volumes localised by the new method. The results showed a very good match between the old and new methods but highlighted the need for pre-alignment of the CT scan by a radiologist.

So, the CT and MR images are loaded onto the system, bony windows are set and three anatomical match points are identified - these points can be anywhere within the brain but should be easily identified on the two sets of scans. Two of these points are usually taken as the centre of the orbits. The system will then "fuse" the 2 data sets by a process of iteration and the resulting matched image sets can be assessed by using the overlay function on bony or soft tissue landmarks - the tentorium is a good landmark to use. Once accepted, the fused image is saved and imported into Xknife for GTV volume drawing. The GTV is then transferred, via some in-house software, on to CadPlan where it is grown into a PTV and planned using fixed conformal fields. At RMH Fulham Road ImageFusion is used for base of skull tumours, acoustic neuromas, meningiomas, pituitary adenomas and to identify critical organs such as the optic chiasm. Future work envisaged was to use functional imaging and to use the system in other parts of the body.

The second presentation was given by Julia Nevinson from the Wessex Radiotherapy Centre Southampton. She gave a talk entitled "The use of a purpose made PMMA phantom for quality control and dose verification in MDR Selectron brachytherapy in the treatment of carcinoma of the uterine cervix." Julia described the design of a PMMA (perspex) phantom which has been used to evaluate Nucletron's Plato brachytherapy software as applied to the dosimetry of the MDR selectron in treating cervical cancer. The phantom's size represents a small pelvis and contains the standard cervix reference dose points i.e. points A and B, bladder, rectum and small bowel. Each reference point consists of a removable plug which can be loaded with either radiopaque markers for use with CT and diagnostic x-rays or TLD microrods for dose verification. The centre of the phantom is removable and will accept the standard Nucletron cervix applicators.

The phantom has been used to obtain comparative dosimetry between traditional point dose calculations - based on orthogonal radiographs - and that obtained using CT brachy. This comparison showed good agrement, with calculated differences of less than 2.5% between the two methods, allowing Southampton to pursue a separate research study into the use of CT image based dosimetry. It should be noted that the high dose gradients at the point A positions lead to differences of up to 5% which highlights the difficulty in accurately digitising these points. Indeed, to assess the variability in dose point and source point reconstruction and the resulting dose calculations, data from different operators using the same set of films and the same operator using various sets of films has been compared. Reconstruction point accuracy was found to be within 3mm and the calculated percentage dose variation was within 6%.

Initial TLD dose verification revealed a discrepancy between the calculated and measured doses at the reference points. The measured dose at point A was significantly less than expected. This error was reduced when an energy correction was utilised. The TLD had been calibrated at 6MV but when the sensitivity of the chips was measured using a caesium film badge source it was found that a 12% energy correction to the calibration factor needed to be applied.

The portable, robust features of the phantom offers scope for a UK wide audit of LDR/MDR dosimetry. A "trial" audit was undertaken at RMH Fulham Road. Pelvic dose points and selectron applicator source positions reconstructed from orthogonal radiographs and the calculated percentage doses fell within the variation already noted above. Data from both hospitals show a similar trend in measured to calculated ratios. Across the data, the measured dose is consistently less than that calculated. The dose at point A was found to be some 10% less than expected while the dose at the other points varied between 5-10% less than expected. This discrepancy still needs to be resolved although no account has been taken of the PMMA to water conversion, the planning system does not take into account absorption within the applicator walls or neighbouring sources and TLD accuracy is limited to 5%.

If anyone is interested in participating in an LDR/MDR Selectron audit please contact:

Julia Nevinson
Department of Radiotherapy Physics
Wessex Radiotherapy Centre
Royal South Hants Hospital
Brintons Terrace
St. Mary's Road
SOUTHAMPTON SO14 0YG
Tel: +44 (0) 1703 825384
Email: jkn@soton.ac.uk

The next presentation, "Planning Techniques in I-125 Prostate Implants" was given by Carolyn Richardson from Cookridge Hospital, Leeds. She reported that since 1995 Cookridge Hospital has been using the "Seattle" technique, developed by Blasko et al., for ultrasound guided perineal prostate implants using I-125 seeds. The prostate gland is permanently implanted with up to 110 I-125 seeds, usually Rapid Strand seed trains, using needles passed through the perineum. The type of patients which are suitable for this technique are where there is no extracapsular spread seen on a bone scan, CT or MRI, the gland is less than 50cm³, the cancer has a Gleeson score of less than 7/8 and a PSA of less than 30, there has been no previous TURP and there is a moderate to good urinary outflow.

The placement of the seeds is planned using a pre-treatment rectal ultrasound scan where an implant template image is superimposed onto the scans. The ultrasound images are transferred to the planning system via video, DICOM or by digitising the outlines manually. The seed positions are planned following pre-determined guidelines , within a series of limitations as to the spacing of the needles and the placing of the seeds within those needles. The pre-plan aims to deliver a minimum of 145Gy to the periphery of the prostate following the TG43 dose calculations (this is equivalent to 160Gy as calculated in the past). The urethral and rectal doses are minimised. Some of the limitations found are: it may be necessary to put seeds outside the prostate gland, the dose tothe urethra may exceed 150% (~220Gy), it may not be possible to cover the entire prostate - aim to cover 95% - and finally, seeds may fall out (not such a problem when using Rapid Stands). The pre-plan can be evaluated by viewing the isodoses, looking at individual points and by calculating the DVH's of the critical structures. Before the implant is carried out the plan needs to be checked by a second physicist and assessed by the oncologist, the seeds need to be ordered (which could take up to 6 weeks) and a needle loading chart needs to be produced for use in the theatre.

After the implant it is possible to carry out a post-plan evaluation. This includes assessing how good the technique is at that point, how good that particular implant was and whether the technique as a whole is improving. The implant is evaluated by taking a CT scan some 4 - 6 weeks after the event, when any oedema has disappeared, and using the planning system to reconstruct the seed positions. This can be a very time consuming task as perhaps 110 seeds need to be identified either by hand or by using an auto-locate routine within the planning computer.

The final talk before lunch was given by Alison Vinall from The Middlesex Hospital in London (part of the UCL group of hospitals). She gave a brief presentation on "Early experience of endovascular brachytherapy at the Middlesex Hospital." At the time, the system had only been running for some 8 weeks and was centred around treating the femoral arteries using the microSelectron HDR. A catheter is introduced into the lumen and an integral balloon is inflated to ensure that the source will be located within the centre of the vessel. The presciption point is 0.5mm into the lumen wall - some 2 - 3mm from the source. A dose of 15Gy is prescribed at 3mm with a treatment length varying from 3 to 22cm. The Target-2 planning system is used because at these very short distances the dose anisotropy, due to the source size and internal absorption and attenuation, can produce large errors in the dose calculation if it not taken into account. Plato BPS v14 is now available and this also takes the anisotropy into account. As the lumen changes shape along its length the dose will vary with a minimum of perhaps 8Gy rising to a maximum of 32Gy to the surface.

After lunch, Barry Baily from the Cromwell Hospital in London, gave a presentation entitled "Commissioning a Gamma Knife" - a voyage of discovery assisted by his fellow physicist Ian Paddick! Barry showed some very interesting slides of the initial refurbishment and construction of the Gamma Knife facility at the Cromwell - how to fit a quart into a small London mews pint pot!! The machine itself weighs some 18 tonnes and houses 201 Co-60 sources with a total activity of 6200Ci. The Gamma Knife is marketed by Eleckta and has treated in excess of 100,000 patients worldwide - it seems that it is only in the UK that this device is viewed with suspicion!!

The actual commissioning of the machine is carried out with the help of Eleckta. The sources have to be aligned correctly and this is checked using Gafchromic film. The film is sent away to be analysed as a laser densitometer is required to scan the film. Once the sources are correctly installed and positioned, the accuracy of dose delivery is within 0.5mm. The dose rate at the "isocentre" is 3.6Gy/minute.

Patients have a relocation frame attached to their skull's under a local receive a single fraction of 8Gy or between 15 and 20Gy if the lesion is an obvious metastasis. The largest lesion which can be treated has a 4cm diameter. The Gamma Knife works rather like an external brachytherapy source. The sources are focussed into a very small volume - depending on the size of applicator used. The applicators look rather like very chunky collinders. So, to treat a large tumour requires that the patient's head is moved very slightly between "shots" until the whole tumour is treated. The resulting "multi-isocentre" distribution will not be homogenous but will look similar to a brachytherapy seed implant. A single shot treatment will take of the order of 15 minutes while for a 2-3cm metastasis may require 5-6 shots with a total treatment time of 25 minutes.

Planning the treatment may take a great deal longer. A plan may be the result of an analysis of 100 MR and 60 CT images and may take between an hour to 3 or 4 hours to complete. The planning system assumes that the head is composed entirely of water.

The last presentation was given by Gerry Lowe from Mount Vernon Hospital. It was entitled: "Choosing and commissioning a 3D treatment planning system - The Mount Vernon experience with ADAC Pinnacle 3TM."

Mount Vernon has just commissioned a new planning system, and, as it is the first ADAC system in the UK, it was thought that sharing the experience might help others. The selection process had 2 phases. Firstly there was a one-day demonstration from the 7 systems with a functionality questionnaire completed. 4 systems were short-listed, and these companies gave a further presentation. All relevant professionals were involved in the process, with a core group of physicists, doctors and radiographers. An evaluation form was given to those involved asking for peoples opinions - with these weighted according to how much of the process they attended. The functionality questionnaire was in 2 parts. A qualitative functional assessment designed to highlight errors in calculations for clinical situations, and a review of the available features. The functional assessment covered wedge and plain beams, heterogeneities, missing tissue, blocks etc. This showed up some interesting features that might be worth knowing as a background. The available features covered contouring, planning options, DVH's, radiobiology, output format, electron and brachytherapy and documentation and training. The short listed companies were looked at in more detail, covering features such as volume growing, graphics, MLC, DRR and compensator handling. The user friendliness of the system and viability of the company were also considered. A spreadsheet was used to assess opinion on the features. It should be noted that staff were very impressed with the professionalism and quality of the companies and systems. There was not much between those short-listed - and site requirements influence the final choice. The ADAC system was chosen, as its algorithm was perceived to be very accurate, and the system was strong on outlining, volume growing, compensator and MLC handling. However - assessment of algorithms was obviously very difficult. The presentation continued with how the system and 3D beam model was set up. The documentation was all machine readable as HTML files - but also came with PowerPoint presentations that were very useful for step by step modelling. Beam modelling, though complex, is broken down into manageable steps. Starting with a pre-stored date base of spectra, depth doses are match by interpolating between the spectra given. The energy spectrum, electron contamination and in field profiles can all be adjusted. This is not to say there weren't problems. The modelling in general was fine - but could do with some bigger ranges on parameters. Also the flattening filter was assumed to be conical - so a compromise model was needed. The selection process worked very well at Mount Vernon, and they were happy with the result. This procedure was helpful in making a choice between similar systems that have a wealth of features and could be adopted by other radiotherapy centre's about to tender for a planning system. Gerry Lowe or Mary Oatey can be contacted for further information.

The round table discussion:

The first topic discussed was financial inducement for sites hosting SCRAP meetings. Currently those involved contribute time and resources for free - so keeping costs down. Considering the quality of venues and meals provided we consider that increasing the registration fee (currently still at £5) so that the department can receive a small donation is not unreasonable. We will see how this affects those willing to host a meeting, and those willing to attend.

The next item was the name "SCRAP" (Systems for Computing Radiotherapy Plans). While there may be a lot of affection for the name, there is some concern that sending mail to "The SCRAP Representative" may cause confusion - visions of post ending up in the refuse department!! So, if anyone has another snappy acronym, perhaps stressing medical physics more, please send them to Phil Tapper and any interesting ones will be published in the next Scrap Newsletter.

The SCRAP web site is up and running - and well worth a visit - although its address has changed. Please surf your way to:

www.radphysics.f9.co.uk/scraphom.htm.

Not only does the site contain information about SCRAP, it also contains the SCRAP database and you should be able to read this newsletter online too. The database lists equipment and planning systems in use at each site and could be a very useful source of information for people wanting to know about a certain system or looking for advice. If the information on the database is incorrect or has changed please let Phil know.

The thorny topic of DICOM compatibility raised its head again. This has been an ongoing concern, and rightly so. In theory it is in everyone's interest to have compatibility, and manufacturers work towards this. But it is also advantageous for manufacturers to ensure that customers only use their products. There has been a vastly increased use of combining modalities, with only minor problems, and it has to be said that manufacturers have worked hard to make improvements. However, there is still work to be done. It was felt that conformance statements could contain more information - though it was argued that a state of "information overload" may be reached. Often "plugs" don't fit, and certain equipment is still not DICOM compliant. A few people complained that some newer scanners were not always better than their predecessors. Do the improvements merit the extra cost? People's views on this would be appreciated. Attention was drawn to the increased use of the web - which SCRAP sees as a very useful tool. Software is now being downloaded from the web - freeware, shareware and patches. Maybe SCRAP could give information and links to useful sites on its web pages? If you have a useful site that others may be interested in - again, let SCRAP know.

The Y2K problem, which one way or another will be soon be over, had an airing again. Of course no one knows the effect this will have, and several people are having to work on the 1st, 2nd and/or 3rd January. However, the general opinion is that departments are fairly well prepared and have contingency plans in place. It might be worth repeating here though: make sure all work is backed up and secure.

Phil Tapper and David Dommett

December 1999

Letters Spot?

With regard to the above discussion I wondered if this newsletter could become a forum for debate. If anyone has any points they would like to raise with the group but would rather not do it during the meeting perhaps you could send me a letter for publication in the next newsletter. Please send ideas, thoughts, comments to:

Phil Tapper
SCRAP Secretary
Radiation Physics Department
The Harley Street Clinic
Oncology Unit
81 Harley Street
LONDON W1N 1DE

Y2K's Meeting

So, we come on to the next meeting. At present we do not have any firm bookings for the next meeting. If anyone would like to host the next meeting then please get in touch.

Phil Tapper

20th October 1999