Modern Issues In Planning

Introduction

Planning has changed in the modern era, and it is now a combination of clinical knowledge, drawing skills, and Informatics. The issues of excellence deal with how these are actually combined together into a cohesive practice model that can withstand medical and legal scrutiny.

The underlying issue is that the decision making that results from a patient assessment is to be electronically documented in a discoverable way that is predictable.

Emerged Paradigm that are now entrenched

The basis of the record making that describes the consultation has not changed. There is still the requirement to document the definition of the patient's disease and the subsequent conclusions that result from the oncologist's assessment of the patient and the patient's reaction to the offer of treatment according to their assessment of benefit and risk.

The changes that have occurred are the requirement to document this material not only in prose form, but also itemised within the information system. The usefulness of the prose documentation is insufficient to act as a validated informatics repository of information enabled for reporting and analysis. Either the prose is analysed manually (cost ++, accuracy -), or by natural language processing (cost +, accuracy —). The requirements to meet the modern need for data worthy of analysis can only be met by the real-time recording of parameters by the staff undertaking tasks. To put a not too fine point on it - ONCOLOGISTS HAVE TO TAKE RESPONSIBILITY FOR RECORDING THEIR DATA LIKE EVERYONE ELSE. So if there is a box in the OIS that is labelled "Diagnosis", the oncologists have to fill it in.

Why the oncologist? You ask. There are several reasons:

  1. the oncologist discovers it first when the patient comes for consultation, and it is easiest for them to enter it so that it is available to all downstream staff & activities.
  2. the oncologist 'owns' this data. By this, I mean that the professional group who analyses data according to things like diagnosis, histopathology and stage is the oncologists.
  3. If the oncologists are going to use the data, it makes sense for them to enter it and quality assure it. Now the fact is that oncologists do this already. All literature reports contain data collated, quality assured and analysed by the oncologist. The new electronic paradigms permit oncologists to enter data into the OIS and then pull summary spreadsheets from the stored data. The need to read the notes each time a new study is proposed has disappeared IF the oncologists are storing their normal workflow electronically as they go.
  4. When anyone has difficulty coding a patient's disease profile, a question comes to an oncologist thereby establishing the oncologist as the gold standard for the data.

While there is not wide acceptance of this in oncologist ranks, it is difficult to deny on any grounds other than 'I am too busy' or 'i don't have the expertise'. Frankly, I have met few oncologists who actually are being honest when they say this. Not that they would admit it!

If you accept the premise that entry of oncological data should be electronic then there are other things that follow. The recording of electronic data is not random, there is a science behind it. There is a knowledge structure to oncological thinking that we learn during training. This structure has to be imposed on the data collection process so that the collected data can be reconstructed.1

Over the last 10-15 years, all (most?) radiation departments have adopted 'electronic' records, whether in MOSAIQ/ARIA/etc or a hospital EMR (EPIC/PowerChart/etc). The issue is whether an oncologist can use their system to record the oncological data and then extract it so that it can be used to generate literature data. This has been achieved with MOSAIQ. I am unaware of whether it has been achieved with ARIA, EPIC or PowerChart.

The same issues operate in the planning realm also. There are ICRU definitions which are remarkably concordant with Informatics principles and thankfully are standardised across the professional - well, mostly.

GTV

The GTV was reasonably settled as the 'visible' tumour. However, this has been altered because of the use of multiple imaging modalities, there is a need for modifications to the GTV. This has been described in the AAPM Task Force Report on Standardised Naming. We could settle for the GTV if metadata could be attached to the name. This would specify:

  • the imaging modality used (e.g., CT/MRI/PET/eye),
  • accuracy of image fusion with the base planning CT,
  • the contrast setting (e.g., Lung window),
  • the use of contrast (e.g., +C/-C),
  • any auto-contouring threshold (e.g., 0, 800),
  • the name of the person contouring.

In the future, there will be several GTVs, each with a different set of metadata. What will also change is the manner of drawing the GTV. In the future, there will be more useful tools to draw the outlines of the GTV according to density gradations using software algorithms that grow or reduce to match density values. There are manual tools to do this, but the likely steps will be to outline a 'typical' area of a tumour and then let the algorithm grow in 3D to match the area. Current Model-Based Segmentation does this already but the process only works well where high contrast planes exist and where the resultant 3D shape is similar to the predetermined 3D model shape. This works fairly well for lung, brain, brainstem, kidneys, bladder and prostate, which are all known size, shape and delineation. The outcome in parotids and neck node areas is far more inaccurate and problematic.

CTV

The future of the CTV is determined by the desire for adaptive radiotherapy. If the patient is re-scanned at any point for the purposes of re-voluming and re-planning, there is the need to reproduce the contours of the original plan. It is impossible to do this unless ONE of the volumes has been anatomically based. By a process of exclusion, this HAS to be the CTV. The GTV is the bit of the scan which is NOT anatomy! The PTV is a geometric expansion in 3D space, so while the same number can be used, the starting volume has to be adapted.

So the future of the CTV is that it will be an anatomical volume. The likelihood is that the oncologist will be required to name the oncological regions of risk. This will need to be contoured by name and then assigned to a dose. The naming and subsequent contouring of an anatomical area allow two things to occur. The oncologist's decision making is explicit and can be quality assured. Secondly, the contours drawn can then be correlated with the name of the anatomical site to verify the oncologist's execution.

At present, when I contour a H&N case, I define a contour called "LN_HN_1a1b2a2b345_rp_fb_p_pa_o_R". Explicitly this means lymph nodes of the "Head & Neck comprising Levels 1a, 1b, 2a, 2b, 3, 4, 5, retropharyngeal, faciobuccal, parotid, post-auricular and occipital on the right". After reviewing the case, I then adjust the contour name to include the areas that I think are 'at microscopic risk'. So for a T2N3M0 Tonsil SCC, I might end up with LN_HN_1b2a2b345_rp_R and LN_HN_2a2b34_L. Then I draw these nodal contours2 on the planning scan.

When the patient is re-scanned for adaptive therapy, the GTV may be less distinct or even missing, but the anatomical boundaries in the patient, while in a different position, will be identifiable and able to be contoured. For this reason CTV IS KING!.

PTV

This volume is a geometric construct to which the radiation dose is aimed. To construct it, you need to know the expected physiological movement and set up error of the site, and the intended dose. The expansion occurs around the CTV which determines the clinical risk area, which alternately surrounds the visible cancer volume.

In terms of naming, the PTV is therefore named as PTV7000 for instance.

Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-NonCommercial-NoDerivs 3.0 License