Dirk Van Gestel's PhD defense | Helical tomotherapy for the treatment of head and neck cancer. An exploratory study in UZA/ZNA

Last modified: 
1/2/2016

Disclaimer

The motives for my public transcription of public PhD defenses can be read at "Why I transcribe PhD defenses".

The transcriber is not, in any way other than the love for wisdom, involved in this research area. No financial benefits were given to him.

The speakers have first been contacted for to correct or clarify transcription or grammatical errors via an interactive audio transcript. Additional comments, corrections or suggestions from the speakers or third parties; are specified as such, and are always welcomed. One can always contact me for any problems or suggestions. Comments can also be made at the bottom of this page, and will appear after the transcriber's review. Presentations, the doctoral or other manuscripts would also provide this public archive with additional value.

Information in collapsable frames (like this one), footnotes, hyperlinks and oblique and (some parenthesized) clarifications were not present in the original defense. The grapheme tilde: ~, means incomprehensible or bad recording.

Meta-information

The thesis itself can be consulted physically in Belgium @ University of Antwerp Library.

Title: Helical tomotherapy for the treatment of head and neck cancer. An exploratory study in UZA/ZNA.

As defended at the University of Antwerp on 20/1/2015. Conflict of interest: I have found no mention of any financial funds declared in Dirk Van Gestel's PhD thesis, nor during the public defense. However, when consulting the publications which make up the thesis, at least 3 of them mention that Dirk Van Gestel received fees from &/or declares conflict of interest via the company Accuray/Tomotherapy®.1 Considering that the research of the PhD is heavily affiliated with the company Accuray/Tomotherapy®, I consider it would have been a better practice to declare these fundings. I guess that this company was also physically represented during the defense, since Dirk Van Gestel spoke of "the company", while referring to a section in the room.

HNC by continent, incidences and mortality

Estimated cases in 2012, sorted by continent, of many types of head and neck cancer; larynx, lip, oral cavity, nasopharynx and other pharynx:2
Estimated cancer deaths in 2012, sorted by continent, of the same types of HNC:3
Sigrid Stroobants: Professor Christopher Martin Nutting external jury, can you open the discussion?

Christopher Martin Nutting external jury

Sure. Thank you very much indeed for the very nice presentation. I was having some difficulty with the Flemish, but your slides were understandable. So I am gonna start off quite often asking you 2 or 3 questions about your introductory chapters4 first and then go to the data chapters5 in a few minutes. You talk about the changing pattern of head-neck patients over the last 10 years or so, with HPV becoming a prominent feature of cancer patients at the present time. But you don't have any HPV data on your patients in your outcome studies?

HPV can cause HNC

HNC can be caused by human pappillomavirus. Most (>90%) HPV-associated head and neck squamous cell cancers are caused by 1 virus type: HPV 16, the same type that leads to HPV-associated anogenital cancers.6 Other major causes of HNC are smoking and heavy alcohol consumption, although other important causes are known, amongst which chewing areca nuts.
No, that's true, but for the first study7, anyway, it was too early: it was 2003-2007. So we, first of all, I think, we weren't even aware of the problem. And for the second study8, 2008-2009 we started to do this HPV determination, but it was not in routine, so we couldn't draw any line on it, so. But I think nowadays it is important to do that and to evaluate the present of the HPV. Because, not for the moment because there is not enough evidence, but it will certainly become an important prognostic, and maybe even predictive factor in the future. Ok, thank you. What do you think it will effect in sense of treatments in the future? I think, as we know that patients with HPV have a better prognosis and doing better, maybe we can just de-intensify, make the treatments somewhat lighter and more adept to the individual patients.

HPV-related oropharyngeal cancers are better treatable

Compared with the traditional smoking-associated head and neck squamous cell carcinoma, HPV-related oropharyngeal carcinoma has a favorable natural history and responds better to treatment. Consequently, patients with this cancer have better long-term survival than those with HPV-unrelated head and neck squamous cell carcinoma (e.g. 5-year overall survival rate of >80% versus ±40% for patients with stage III-IV tumors), and hence they are more likely to experience chronic therapy-induced morbidity.9
What about the radiation treatment? Do you think it will affect the radiation treatment for patients, or is it all about chemotherapy? I think these kind of tumors respond better to radio therapies, so maybe we can think about de-escalation, especially in the elective lymph node regions, I think we can try, as we already did here in Belgium, try to go to a lower dose, for example, of 40 Gray, or even lower. But that is hypothetical of course. Okay, good answer. One of the other topics you touch on in the other introductory chapters4 is particle therapy, and protons. What will be their role in the future? I left this out of my conclusive slide, to avoid this kind of question. I think particle therapy, proton therapy especially, will have some impact. Because, people maybe don't know, but, so, when you see the dose ... the depth curve of particle therapy, of protons, then you see that after the deliverance of the dose at the level of the target volume, there is no dose left anymore. So there is no exit-dose, if you want to, so if you use particle therapy if you use protons, then the contra-lateral side will be spared very nicely. So, the global toxicity will be less. Now, if you look at the situation for head and neck cancer, quite often you have to irradiate bilaterally, so, then I think the use of protons will already be less. Another requirement is that you have a gantry, for the protons. Not all centers have that for the moment. That's also a problem, because if you don't have a gantry, I think the benefit of protons will be very, very small. And I don't know whether your question also is about carbon ions? No. So I'd like just to move on to chapter 47, to your step-and-shoot planning and results. You describe a 2-phase IMRT technique, whereas a lot of people starting early on preferred to use a simultaneous integrated boost approach. Can you just discuss for me the pros and cons of the 2 approaches, and why you chose to do what you did?

Step-and-shoot

IMRT treatments based on the step-and-shoot principle include usually several segments with a limited number of monitor units, e.g. <10 MUs. When segments like these are used, it is essential that the performance of the accelerator is within the required tolerance as soon as possible after start-up. During start-up, depending on the accelerator design, the energy can differ shortly from the intended energy. This can have an effect on all kinds of beam characteristics like dose per MU, PDD and beam profiles. It is important to determine the start-up characteristics during commissioning and to check these regularly in the QA.10
If I have to be very honest, it was not my decision to do it in this way. But I think on the other hand there is evidence to keep it in the normal fractionation, and then give it together with chemotherapy. It has been shown that whenever you are accelerating, then your outcome will be better, unless you are giving chemotherapy together with your hyper-fractionation or your accelerated fractionation. So, as we are you using a lot of chemotherapy, and especially in the second study8, because in the first study7 it was still, I think 50%. Whenever you are giving chemotherapy, you shouldn't accelerate, or hyper-fractionate. But for the other patients, and especially for the smaller tumors, I think this population might have been candidate for hyper-fractionation or accelerated radiotherapy. So am I right to thinking that, for the largest group in the early study 17, that they received 2 Grayfraction for each fraction of radiotherapy, and had another phase of IMRT on top of that, to go up to 70 Gray?

IMRT

IMRT [Intensity-modulated radiation therapy], an advanced form of external beam irradiation, is a type of 3D conformal radiotherapy (3D-CRT). It represents one of the most important technical advances in RT since the advent of the medical linear accelerator. (...) 3D-CRT/IMRT requires the use of 3D treatment planning capabilities, such as defining target volumes and organs at risk in 3D by drawing contours on cross-sectional images (i.e., CT [computed tomography], MRI [magnetic resonance imaging]) on a slice-by-slice basis as opposed to drawing beam portals on a simulator radiograph.11
To 70 Gray, yes. Ok. And it's slightly paradoxical, that it tend to be the poorer conditions patients, rapidly proliferating tumors, who got the simultaneous integrated boost. Yea, that was in an attempt to do something more for these patients. But it's true that ... And were patients with lateralized tumors, like small tonsillolith tumors or oral cavity tumors, also treated by IMRT in the first study7? Yes, but not so very much. I think I precise it somewhere. I don't know it by heart how much or how many. There was a few, a minority. Yes, it was a minority, because at that moment we still had to make the choice between the IMRT, who was very labor intensive, and the simple 2- and 3D techniques. We used 3D techniques quite often, and so, I think, for these unilateral radiations, you can very well do that with a 3D technique. I don't say that it will be exactly the same dose distribution, but I think you can also do that with classical 3D. With your experience now, what would you choose to do with lateralized tumors? Tomotherapy. Tomotherapy.

TomoTherapy®

Tomotherapy®, presented by one of its co-founders, Thomas Rockwell Mackie Wikipedia, ResearchGate, University of Wisconsin, in a "promotional" video:
►
Isn't the low dose to the contra-lateral side an issue? That depends of the planner. If you plan well, you can give whatever dose you want. You can limit the dose that you want. So, I think that's very nice on tomotherapy, that is: if you really know the system, you can draw your dose where you want to draw it. But if you don't know the system well, you are able to do some very big mistakes. Yea, right. And also in the selection of patients in the first clinical cohort, it was a minority of patients who received IMRT at that time. Yes. In study 17, 78 patients out of the total over 1000 patients had IMRT. How were those patients selected then for IMRT? It was quite randomly. I didn't make these decisions myself. But, I think certainly, at the beginning, we chose the very difficult patients. So whenever we were not able to get the good dose distribution with the classical treatment possibilities, then we putted an IMRT on it. So, that is also the reason why you find some sinus ~ tumors and some nasal nasopharyngeal tumors, because at that moment we said: ok, we cannot get ... especially in the nasopharyngeal tumors we weren't able to get the good dose, at the level of the base of skull. So that moment, we said: ok, this will be an indication for: we can do better with IMRT. But we were very reluctant towards IMRT at that moment. Ok. And I'd like to sort a bit more about the final part of your discussion, just from the first cohort. About the different prognosis of your patients. Other groups have so much better results then your patients, who are similar to other European groups. Do you make any hypotheses to why patients in the U.S. might be, apparently, doing better with the similar treatment? I think there are quite some explanations for that. First of all, these patient group-settings in the U.S. are highly selected. Often these centers ask some extra money, so only the better patients will be treated. Just very maybe, there might be a bigger sub-population of HPV-positive tumors. And what we also see in the study of James A. Bonner University of Alabama, because it is about this study that we discussed that: that the people generally are in better states in the U.S. group than in the European group. The WHO classification is better for the U.S. people. Do you know what the instance of HPV in North America, compared to Europe, at this time quite was? I have no idea. Do you have an idea? Yea. It's out there. What do you think it is? Yes, at that moment 30%? Probably about 30-50% in Europe and 70-80% in the U.S. So I think, to my view, that is probably the largest confounding factor of the North American series versus the European series. Also, these are community centers; I am not sure, there would have been some patient selection, but whether that or not that was advantageous or not. Ok. So, I just want to move on to chapter 612, which is your multi-institutional planning study. I mean just to manage to achieve that study, with different hospitals, planning the same patients and that was an academic exercise, but inevitably, how did you manage this collaboration? Now, the biggest work has been done by the lovely lady of the national cancer institute, so. But I think in the Netherlands they are cooperating much more between centers, there are not so many centers as in Belgium. I think there are only 15 centers for 15,000,000 people in the Netherlands, why here in Belgium we talk about 25 centers for about 11,000,000 patients. So, the centers are much much bigger, and they communicate much better between the centers. There are also much better national guidelines in the Netherlands. So, in Belgium, it's a little difficult, you know, we have Flemish people, we have the French-speaking people and people from Brussels, so the communication is not that good. For example, we have a head and neck group in Flanders, but we don't communicate with Brussels or ... ~ too much for me. So each of the centers plan the same patients, and the same PTV's were planned?

PTV

There are 3 main volumes in radiotherapy planning:
  • Gross tumor volume (GTV): the position and extent of gross tumor, i.e. "what can be seen, palpated or imaged."
  • Clinical target volume (CTV): "contains the GTV, plus a margin for sub-clinical disease spread which therefore cannot be fully imaged".
  • Planning target volume (PTV): "allows for uncertainties in planning or treatment delivery".13
Their hierarchical sizes can be clarified in a simple diagram:


GTVCTVPTV


Yes, were identically. So I did the contouring of the same 5 patients as I mentioned ... were contoured and sent to the different centers by CD-rom, at that moment. So each center planned as well a classical static beam IMRT, as a rotational ~ So the only variations were the planning system on the planner, in the local center; so it is controlled. Good. There is a comment on page 10214, that you used collimated rotation to reduced tongue-and-groove effects. Just remind me how that ... explain that to me, please. If you put your collimator just like ye said, and you have your different leaves, and there is a leakage, especially when 2 leaves are coming together. And if you put your beam just, like at 0 degrees, then there is the leakage always at the same place. So if you treat your patients like that, you will have some higher doses, some leakage doses there, and then every other cm, there were 2 leaves coming together, so that's the reason. And how does the collimator twist avoids that? Because then you will smear it out over the total volume. Perfect. Ok. On page 108 you talk about quite a lot of variation in low dose radiation: figure 415. And that seems to be one of the major differences that you found in different centers. ~ reduce volumes of healthy tissue, receiving doses above 30 Gray ~ IMRT. I am sorry, can you repeat the question? Page 108: you talk about the amounts of normal tissue being radiated to some moderate doses ... and demonstrate figure 415. Would you explain that figure15 to me, how you got to that conclusion? No, I can not. I have to be honest. I think this is the figure15 that has been made by our colleague physicist, and I have to admit that, this one, like that, no I cannot explain. It seems to me the different centers are quite variable in terms of the doses. Yea, there are some differences, but that is because all these centers have their different approaches and their different ways of planning, so. As a consequence of the data that you present in chapter 612, was there a change, amongst the centers, after this? Yes, we also describe. So, in fact, what happened was that different centers sat together multiple times during this procedure, and they all made their IMRT planning more or less in the same way. So the centers who were having a better performance were teaching the other centers, so that they all ... So there was a change there ~. Yes. Good. Then, in chapter 716 you start to move on to the comparison of the tomotherapy versus the other SmartArc techniques. You talk about the tomotherapy systems producing smaller volumes than some of the other planning systems. What was the magnitude of that difference? I think, for most of the organs, the difference was quite acceptable. The difference was mainly, I think, with the EclipseTM system. Which, there, the volumes were significantly smaller in the EclipseTM study. And probably this is due to the calculation system, as sometimes you take the outside of the voxel, and another time the middle of a voxel and another time the inside of the voxel. The problem is then: if you compare the studies for dose, then you are comparing different doses for different volumes, and that's ... So it's an inherent problem of importing CT-data in different ways. Yes. You think that is an important consideration ~? I think not for the big volumes, but for the small volumes it might be a problem. Because, there, if you have a small volume where you find a lot of variation, then you have to put some questions on the dose you found. But, I think, whenever these dose differences are very big, then they will compensate. I don't think that the volume problem will be such a big magnitude at the moment. Sigrid Stroobants: Maybe a final question? Ok, last one. It's a question about when you planned your planning study, you were anticipating relatively small differences between different planning systems: how they perform. But you chose to only study 5 patients. Yes, the problem was I contoured ... In your discussion, you say ... Yes, I contoured for 10 patients. But you have seen that the study has been done with other centers, so the national cancer institute and also Leuven cancer institute; and especially at the Netherlands cancer institute, they were not eager to plan 10 patients. It was just too much extra work for them. So, I was already happy to have 5. Because, personally, you could say that you were inevitably gonna find not much difference between the importance of the planning systems with relatively small numbers. Yea, that's a critic that is coming back all the time. But, when you see the sometimes very big differences between the planning system; and when these differences are, in 1 planning system, pretty stable. Then you can think this will be a statistically significant difference, and that even with small number of patients, you can draw conclusions of that. Sigrid Stroobants: Are you satisfied with the answers? Yes. Some of them. Sigrid Stroobants: Ok, some of them. Professor Eric Félix Lartigau, can I ask you to proceed?

Eric Félix Lartigau external jury

Thank you very much. First I would like to thank the university of Antwerp for this very kind invitation, and I am very pleased to be here, and to be with this very distinguished jury. So, Dirk Van Gestel defender, I must say, I have been mostly interested by 3 chapters in your thesis: chapter 612, chapter 716 and chapter 817, dealing with comparison, how to implement better dose distribution, timing. So, I may have a few questions on these 3 chapters, and then we may go to more: future, what's next? And I would like to spend a few minutes with you on what's next. And I think you are prepared for that. So, coming back on these chapter 612, chapter 716 and 817. There is surprise you didn't say on the monitor units issue, because you have been clearly demonstrating that, by the comparison mostly, I think in chapter 716, at the comparison between the different systems, there is a huge difference in monitor units delivery, according to the machines and the system. Is it for you on a, I would say technical and clinical perspective, a big deal? If any, if you are able, on chapter 817, to demonstrate that you can reduce the amount of monitor units, there is still big difference between TomoHelicalTM delivery and the rest of the systems. Is this something for you, to make a deal around? Do we have to implement new tools to ... Is it something you are considering? Yes, I think it's important, the number of monitor units. Because, and that's the reason, because we think: if we give more monitor units, there is more risk of secondary, eventually, of induced secondary cancers, because of the bigger low dose, so. I think monitor units from tomotherapy, and monitor units from the other systems is, as you say in English: apples and oranges. Because, and that's what we describe in chapter 518. Tomotherapy, you are just working in another way. So, the total amount of monitor units is so high, because you keep on giving money to it each time again. But if you look at the very localized, very small space. And this place will only be treated during, let's say; if you have 4 times overlap, than you treat a single tumor cell, or a single healthy cell; you only treat during 48 seconds, so something like that. So the mere monitor units on an individual base, will be much lower, and I think it's just the kind of system makes that you need so many monitor units. But I don't think the total load of monitor units on your patients is so much different. Yea, I would agree on that. But I think it's a very important question. So I think it will be nice, in any next step, or any new study, to have a look on that. Because, with tomo making a strict correlation between the monitor units, the way the delivery with the multileaf is working; and what you call the integral dose; is quite tricky. And I think that is something which is not exactly the same as it is with different systems. So, it is, I mean, it's very often preserved as something going down at the side of the tomo. That's just something going differently. It's different, I don't think it's ... Yea, so maybe going that way. Because, as you know there are criticisms and limits to use tomotherapy, maybe, in children. Would you elaborate on that? Is it for you a good system to be used in children, because of that kind of questioning? Or can we safely use, for example, for medulloblastoma, where it's so easy to treat children? I know it, we are out of the head-neck, but for you: if you can elaborate a bit on this? I think for children indeed, the question is different. Because, then, there, the dose you give with tomotherapy will have smeared out a lot more than with the other techniques, and you can wonder whether if you are giving a cranio-spinal axis in a classical way, whether this total amount of dose would be important than with tomotherapy. You will certainly will have less high dose zone with tomotherapy, but in compensation you will have a higher low dose zone. I think, it is not really clear, for today, which dose is the most problematic for the secondary cancer. It is the very low dose zone, or is it the intermediate dose zone? And more and more, I think, we shift to the idea that it might be the intermediate dose zone, and then tomotherapy is doing better again. To make a long answer short, is that I think that, for children, whenever possible, there, I think, that proton therapy is indicated, because then you can avoid this exit dose. So, if you have the choice between tomotherapy and protons, then you go for protons. If you don't have protons, then I think that tomotherapy might do better than the classical systems, because you don't have this problem of the gap, you go much faster, so intra-treatment movement will be lower, and things like that, so, yes. Yes, with all the technical limits of protons, but you've been already answering to Christopher Martin Nutting external jury on that specific question. On chapter 817, which is a very interesting chapter because the title is good: "Is time priceless?", I mean: do we have to really run after time, when we do these relatively difficult cases. You are comparing different systems, you have been using TomoEDGETM. You are not elaborating on the VoLOTM. Is VoLOTM for you, too which was interesting implementing on the tomotherapy, did it change something? I will come on TomoEDGETM later on, but just one word of VoLOTM, does it change something for you? Outside of the speed of the optimization?

VoLOTM

A planning technique may calculate dose distributions ... (...) with the VoLOTM algorithm using a convolution/superposition algorithm. VoLOTM features high-speed parallel processing for both dose calculation and optimization, every 10 iterations recalculating and integrating a "full dose".19
It was a little bit faster, and I had the impression I had to change a little bit my way of working, in that the penalties are somewhat more different. You have to be let aggressive on your target volumes, so there is also a difference on the floor, with VoLOTM. And then there is one thing I do not really explain well with VoLOTM ... or with the TomoEDGETM system, that's: if you take a look at the TomoEDGETM system, your conformity index will be less, than with the other techniques. And there I have no explanation of that.

Radiation conformity index (RCI)

The radiation conformity index is determined as the ratio between the target volume (PTV) and the irradiated volume.20
It was my next question, and it's ... Sorry. No, no. You just anticipate. Thank you. I felt you coming here. It's page 14721. No, I fully agree. Why is the mean conformity index is worse with TomoHDTM Fast? Why? With TomoHDTM Fast? With the TomoHDTM Fast? Yea. Because there, you are treating with 5 cm, so ... Yes, but did you check with the one of the 2.5 cm? If you go down to the 2.5 cm. I mean: what's the ratio in between the speed, the size of the leaves, and the conformity index? Because you have taken the 5 cm to get speed, but if you play with that, how does it impact on the conformity index, if you use the 2.5 cm, or ...? Did you test that? No. Ok. That could be interesting too, because of you have a very strict statement saying that index is worse, and I would love to see what exactly the impact of the multileaf and the ~, in that specific issue. Yea, I think it will be very interesting to look at that, but for the moment ... we didn't do that. Ok, I think that could be ... So, once again, I was very interested with these chapters: chapter 612, chapter 716 and chapter 817; and there was something which was missing, but hopefully you came back on that point at the end, which was: quality assurance. And on the last slide you had a very bold quality delivery, so I love that. Could you elaborate a bit on that. Because, I honestly think, you know, honestly speaking, that's something which was missing, and I think that's one of the big pluses of the tomo, which is the easy quality assurance and the full quality assurance. What is missing today on that? I touched, in several chapters, I touched a little bit the quality assurance. But it's true, I don't know how it is in France, but in Belgium, physicists and physicians are quite often living separately together, if you understand what I say. So, quality assurance is physicist's matter and the physicians are going on the rest, so. But I think, and it's my opinion, that quality assurance indeed is very important, and maybe I left it out because you could write another thesis on that. No, yea, that's right. But you have been elaborating on adaptive treatment. And, I think, if we go to adaptive treatment, and if we want to be as much on line as possible, this is a critical part. And that's where, maybe, physicists and physicians can be back together working, because if we want to do real full adaptive treatment, we have really to consider, on both sides, what will be the best compromise. Because there will be compromises to do.

Adaptive radiotherapy

Adaptive radiotherapy is defined as "changing the original radiation treatment plan (by modifying either beam apertures or intensity patterns) during a course of fractionated radiotherapy to account for the temporal changes in anatomy (weight loss, tumor shrinkage, internal organ motion) or changes in tumor biology or function (hypoxia, proliferation)".22
So what is for you tomorrow? And we enter the tomorrow questions. What is for you, tomorrow, the best adaptive treatment in the head-neck? And how would you go to adapt? In your thesis you are pointing to studies you are participating to, but: you, if you have a dream on your side, what would be tomorrow the perfect adaptive treatment in head and neck cancer? I think that answer is very easy. So, if you have the software, the good software, who is able to do recalculation, just based on your re-contouring, and recalculation based on your daily MV of kV(kilovoltage)-CT scan, and you can implement your adaptation right away, I think that's the best way to do it. And that, maybe ... I know they are working on that, and probably it will be in the future, that we can do an adaptive treatment on a daily base. What the questions with adaptation is: do we adapt on CTV's, do we adapt on organs at risk? What could be your best way to ... where and how far would you adapt? I think adaptations on your organs at risk, is safe. That you can do adaptations on your target volumes, that's another question. Because the problem there is that we don't know whenever a tumor or a lymph node shrinks, whether there are tumor cells or remaining ... on the spot, or that they are shrinking concentrically. So, and that is the big question, so as we don't know that for the moment, I think you cannot consider to adapt on your target volume. Unless you see, for instance, a muscle that's moving inwards, so that you can crop out these muscles, because unless you think there was invasion of the muscle at start of course, but otherwise you can crop for this bony anatomy and these changes in air cavity and ... I think, there you can adapt. But, for the rest, I think you have to very careful to adapt on the level of your target volume. Yea, yea. I would fully agree on that. But we have to very careful. Because, we have, even today, some people advocating to re-adapt on tumor volume, and I think that's something we are not ready to do. In the same track: treatment adaptation; you have clearly demonstrated that by combining systemic treatment and IMRT, dynamic IMRT, you have more toxicity. And that's something everybody knows. So, next step would be to keep on with getting some kind of very active treatment, and combining whatever we combine to IMRT, there will be more acute toxicity. Do you see treatment adaptation, on the same way as we have been just talking about, can be a good way to decrease acute toxicity? Or do we have something else to do, in order to keep on with active treatment, and try to solve the acute toxicity issue? What's the best way to go? I think if we go for less toxicity, then I don't think adaptation will make a big thing. You can do it after a few weeks, but after 2, 3 days you will not see ... We know that for the ~parental transcripts~(?), for instance that it starts shrinking after 10 days, 2 weeks. So we could imagine to do an adaptive plan at that moment, but the big problem there is the cut-off value. Which value or you going to ... or are you going to do an adaptive plan for every patient: that is hardly possible. So I think there you have to make choice. And then I think, what seems interesting to me is, that you just try to limit the toxicity in the low doses zone. So I already mentioned the doses de-escalation study, where you give less to your elective lymph nodes or the possibility to use sentinella procedures, so that you can just leave out parts of your elective lymph node. So, I think, at that moment, you are doing a better job than with the adaptation. The adaptation, I would use it like we use it now, for instance, in dose painting by numbers, where you give a very very high extra dose. You give your proven historical dose, you always give ... then you give an extra dose, very localized, up to the tolerance limit of the surrounding tissues. And you're just going for the greater surgeries, ~ for destroying the tumor, instead of to go for a proliferative cell death. Yea, whatever. But even so, that would be difficult to do it without slightly indulging the low dose to the surrounding tissue, so there will be a tricky ratio to define. And I don't know, honestly speaking, that's why I ask you to the question, I don't know which is the best way to go to. To much more adaptive, to go to less active treatment with less acute toxicity? But, that may be the tricky question for us in Europe, because we have still a lot of knowledge ~ patients. And for these patients, we still have to go to the high dose side. My last question would be on one of your last remarks, on: we need to randomize. So I always love to randomize. But, just like the English people would say: but. There is a big but, so you have been putting on one of your slides: next would be randomization. Tell me what will be the study, please, Dirk Van Gestel defender? But I think, at that moment, you just have to randomize between rotational techniques and static beam IMRT techniques. And what would be the endpoint, clinically speaking? I think the endpoint will mainly be toxicity. We've seen you've been able to demonstrate with combined treatment, systemic treatment and ~, a real difference in acute toxicity between static and dynamic delivery? I think that at certain levels, and certainly at the level of the swallowing structure, I think you can demonstrate differences, yes. Well, I am not so sure. That's why we need randomized trials. Yes, but that's why it will be very difficult towards such randomized study. Yes, I know. But that's the next point in my discussion: we need to work together, more together. I've read an article where, I think it was the initiative of Shalin J. Shah ResearchGate, University of Texas, LinkedIn, where he said, ok: we have 45 or so much, 45 organizations who have signed this contract, why are they not working together, these 45 centers? Then you have your study running, not in a month, but in a year instead of in 5 years. Because they are maybe working all in Belgium, so, no I am joking. I am joking. It's a pity that Vincent Grégoire copromotor, absent is not here. Well. Ok, so thank you again, and I give you the micro back. Sigrid Stroobants: You are satisfied with the answers? I am. Sigrid Stroobants: Ok. So professor Andrew Maas, can you proceed?

Andrew Maas internal jury

Thank you. For the benefit of our external guests, I'll continue in English, if you allow me. Dear candidate, I really enjoyed following the progress of your studies, during you doctorate period, when I was involved, in the background; and I particularly enjoyed reading your thesis. To me, of course, much of it, was completely un-understandable. But I at least recognize this type of pictures as a, I would say a CT-scan; but for you it's a radiotherapy device. I would like to focus and zoom in on a number of aspects, a number of details, looking at figures, numbers, morality and design in a way. So I would like to look with you at chapter 47, to begin with, on page 67. There you state in line 5: "After a median follow-up of 18.7 months (range 4 days to 51.7 month)". I thought: good gracious, what a wide range. What has he done? How can you loose a patient after 4 days? And then a thought strapped me: that patient probably died. Is that correct? Yes this patient died ~, shortly after the beginning. Yes. And now I think you are in trouble, because if you look at follow-up, a patient who is dead at 4 months, will be dead at 3 months and will be dead at 6 months, I presume. Yes, absolutely. So is this the correct way to look at things, and express things? I would say, the shortest range is the patient who survived with the shortest follow-up. Would you think I have a point there? Or do you say: no, no, I disagree? It depends on the definition of your follow-up, and I think we defined that as, from the start of radiotherapy. So at that moment, I think you have to stick to your definition, and keep it in that way. But I agree that, and I also discuss that with other people, I agree that there are a lot of problems, troubles in this study. This is a suboptimal study, I agree with you. But, at that time, it was also the material I had, so there was not much to do about that. But I agree. If I would do a study, I would do it prospectively, this is a retrospective ... I don't think that it's so difficult to do prospective study on this kind of patients. And then it would already avoid a lot of these troubles that we have now. Well, perhaps that's one part of the story. The other part is that the way you've looked at this follow-up, I would say, maybe confounds the way you express you results. If I take you a few pages further to page 6923, there you describe the percentages of complications that are noted. And my question there is: were these percentages related to the overall population, or to the number of survivors, so only in survivors? Take this case of the patient who died on day 4: he can never develop a late toxicity. No, that's true. So for late toxicity, he would be a negative. No. I'm just ... Yes? And that's kind of tweaking your data in a very positive way, is it not? Or, did you ... so how, first of all: how did you relate percentages: to the total population or only to survivors? I don't really remember of this study, because this is already 5, 6 years old, but I know ... Ok, well, in that case, let me take you to a more recent study, on page 165, in chapter 98. There we took only the patients who survived. Only the patients who survived. The recent study, we did it like that. We ... You're sure about it? I am 100% sure. It is even noted on page 165, the figure 524 and figure 625. You see there the one asterisk, in table 625, there is noted that "patients who died within 6 months after radiotherapy were excluded from late toxicity analysis". Ok. Excellent. So you learned as you went on, yea ... It was a learning curve, yes. It was a learning curve. I was also intrigued ... I am sticking a little bit on this chapter 47 at the moment. I was intrigued on page 68 by figure 126, and I had to put on my glasses to read the legends in it. But if I get it properly, interpreting this Kaplan-Meier curve, the number of patients with symptom-free survival ... or a symptom-free period, I can't read it here in the dark exactly, is higher than the overall survival. Lights brighten. Ah, that's better: the loco-regional control; that's higher than the survival. And I had a little bit difficulty in understanding that. Could you help me there? I think that it was due to the fact that the number of these patients died of non-tumoral causes. As you can see in ... So a dead patient, dying of non-tumoral causes, you would consider as a loco-regional control? That's a good way of achieving good results. I ... yea, ok. At that moment I questioned the statistician and they told me it was possible so. It was professor Joost Weyler University of Antwerp. I will ask him the question again. I think it's an interesting hypothesis. Ok. I will question him. Ok. Now, I would like to jump with you to chapter 612. Here, let me just get the details right. How many patients did you enter? 5 patients. And how many centers participated? 5 centers. And how many authors did you have? Somewhat more. I counted 8. Yea, it's possible. And I really think the university here must be extremely happy with your productivity. Because, having seen this, I thought: well, a manuscript with 5 patients from 5 countries. That's only 1 per country, then I started to understand it. But then I continued reading, and on continued reading, I came to chapter 716, and I found table 127 on page 121; and that's identical to the table28 in chapter 612. So then I suddenly got this hallucination maybe, or this revelation, whatever you call it: that is the same population you describe as in chapter 612, is that correct? That's very correct. And then, I did read through everything, so I reached chapter 817. Yes, and there you find the same patients. And there I found the same patients. Because these are planning studies, so you ... So you managed to get 3 different publications29 out of 5 patients? No, I don't think that it's ... Yea, if you see it like that, but that's just the basic populations you are examining. But what's important here is the result, the dosimetric results, and not the results of your patient. So you are not examinating these patients, these patients are just a middle to get your results. Yea, I understand, but ... I don't see, I don't see ... Let me ask the next question: why ... or the question in the other way around: what would be the added value of 5 other patients? For chapter, let's say, for chapter 817. Well, it depends what're are looking at. I mean: how certain are you that results obtained in 1 patient, in 1 center, are completely reproducible? Yes, I'll agree: you've done it in 5 different centers, and they were fairly equal; but they all use different techniques. And I've always learned ... Well, we can say, if you've seen a case, you can talk about my personal experience; but it's still only 1 patient. No, your remark is very interesting, because if you send these articles for publication, then they say: "oh, there are too many different centers, so we're not sure that it really was always the same planner, and that you did put the same accents". And when you send them an article, done by one center, then they say: "yea, but maybe the center is not right, and you are not using the right goals". So we can go for hours and hours, ok. And I understand that 5 is maybe not enough, I'm sure. I'm willing to continue for hours and hours, but I don't think we'll be allowed to. No, be my guest. Thank you very much. Sigrid Stroobants: Are you satisfied with the answers? With some of them, yes. Sigrid Stroobants: With some of them. Ok. So then, to proceed, I have also a few questions.

Sigrid Stroobants internal jury

In your summary, you compare the results of chapter 47 and chapter 98 between static beam and helical tomotherapy. And the differences weren't that great. Maybe less than you had hoped for, and you explained a little bit this was due to different patients inclusions. But also the fact that the other volumetric beam therapies are also more rapid. My question is then, and certainly the situation here in Antwerp is intriguing this question: is there still a place for tomotherapy in the regular radiotherapy department? Given the fact that it takes longer, it's more costly. So is there still a place for this machine? I think there are 2 questions. Let me ask you 2 questions. Which quality of treatment do you want for yourself? Of course I want the best cure. There is a place for tomotherapy then. So, is it's as simple as that? And every ... I think not for the very simple cases. But, I think, for the difficult cases, it's just way ... it's easier to plan, it is easy to administer, and you're sure ... I think, because I didn't spoke a lot about quality assurance But we did a hell, a lot, of quality assurance. Our physicists did a very very good job. So, you get ... what you see is what you get, with tomotherapy. And I don't know with the other techniques, but we are implementing now other rotational techniques, and we have a lot of troubles to get good dose distribution, to have the good quality, to ... So ... So, to understand, that for tomotherapy, it's easier to plan? The quality assurance is easier then for the RapidArc® console? I think it's easier. You have to ask that to physicists, if you want to know that in detail. But I think it's easier. And also the planning, it's very easy. It takes no 10 minutes. So it's not a system that you would reserve, let's say, for an academic center with specialized service, but is really a routine machine? I don't think so. I don't think so. I think it becomes more and more a routine machine. The only problem; and I look at the company ; is the price. It's still a bit more priceless than the ... it's more expensive than the other techniques. But if you take that apart, it's clear that you go for tomotherapy. And if you have to take in account ... and that's what I was referring to as quality-driven radiotherapy. We're not discussing quality driven radiotherapy, now we're discussing how we can get the best quality for the lowest price. And that's difficult if it's going about cure and about cancer patients. Yea, and I agree with that, that the cure is probably the most important item. And then I was a little bit shocked by the answer on of the questions of professor Eric Félix Lartigau external jury, if you would design a randomized trial, you would not go for quality assurance, but for toxicity. Why is that? Because I think, maybe, you make a better statement if you can prove that also the quality assurance proves? Why is it so difficult to design a randomized trial with the primary endpoint being quality assurance? Because, I think that, just by giving better dose distribution, we will not be able ... as you saw in the slides to the dose distribution of the different target volumes, is not that different between these different techniques. The big problem in head and neck cancer is the recurrences within the high dose, so if you wanna tackle this, then you have to make a randomized trial with dose with dose escalation, ok. But that's a another ... that's an extra step we have to make. If you just want to keep it easier, and to make it ... to show benefit of it, for the moment, I think it's easier to prove that your toxicity level will be lower, because that you can do in the normal setting. You don't have to adapt much into your department, while this dose painting is still very experimental. So ... So the major reason is, because of this dose painting is still experimental? Yes, that's right. So, then, if we move forward from dose painting, we come of course to biological doses, and then imaging. So how do you think we can help you in that, with imaging? And how far; or what is the work we still need to do, before you really can use this, for instance, in a clinical trial? I think there is a lot of work that you can do for us, to help us, with imaging. It already starts with the delineation of your tumor. Now, for the moment; we are making a reference CT-scan, we're making the fusion with the PET-scan, we're making the fusion with the MRI; but even on these examinations, you often have blurry images. Your PET-scan, you know, you have this 5 millimeter in-certainty, because of the PET-... ~ range, hé? Yea, the PET-range. For MRI, sometimes you see your tumor very nicely, and sometimes it's really very very blurred. Also, superficial tumors are very hard to see, on imaging. So all these things can be done better, can give us extra resources to better treat your patients. So, that's for the delineation. And then, afterwards, if you're going for adaptive treatment; I think, for the moment we're using FDG-method. I think, maybe FDG is not the ideal tracer to go, and to see where the most resistant tumors are. So, maybe we should go for hypoxic markers, or for proliferation markers. So, I think, at that level too, there is still some work to do.

FMISO as an hypoxic marker in HNC

Positron emission tomography (PET) imaging with [F-18] fluoromisonidazole (FMISO) has been validated as a hypoxic tracer. Head and neck cancer exhibits hypoxia, inducing aggressive biologic traits that impart resistance to treatment. Delivery of modestly higher radiation doses to tumors with stable areas of chronic hypoxia can improve tumor control.30

Proliferation markers in HNC

Cell proliferative activity has been extensively investigated in head and neck tumors. Ki67/MIB-1 immunostaining, tritiated thymidine or bromodeoxyuridine labeling indices, DNA S-phase fraction, proliferating cell nuclear antigen expression, potential doubling time and analysis of the nucleolar organizer region associated proteins (AgNORs) have shown significant correlation with prognosis in 4806 cases of tumors of the oral cavity, salivary glands, pharynx and larynx. However, this was not observed in 2968 other reported cases. (...) Provided that large and homogeneous series are evaluated by standardized methods, cell proliferative activity can still be regarded as an inexpensive and reliable prognostic factor in head and neck tumors. (...) Proliferation markers can be classified into three main categories: (i) growth fraction markers; (ii) markers of specific phases of the cell cycle; and (iii) cell cycle time markers.31
And, in practice, how would you do that? Because, for instance, if you repeat, for instance an FMISO scan hypoxia, you see that the hypoxic volume changes over time. And if you want to boost, how will you do that? Would you boost on the initial hypoxic volume? Would you repeat the scan during the treatment, and then adapt accordingly? Because that's still unclear to me: how; even if you would have this tracer available; how you would design a trial that includes this ~? It's very though, but maybe you have some ideas? No, no, I think, I agree with you, that for the moment there are no clear answers how to proceed on this. I think there are several studies ongoing to see which marker might be better, but I think we don't know for the moment. And that's a big problem in the dose painting: that we don't know how to evaluate the results; to see which sets, or which regions we have to tackle. And I think that's a big problem. So it's a new PhD-thesis for another candidate. More than one, I think. More than one. So, also: what are the advantages, I think, of tomotherapy, is the fact that you have this on-board CT-scan, that you can do image guided treatment. I was struck also, when I read your thesis, that currently you only use it for repositioning. But a lot of the possibilities are still not taken into account, because of software problems. But is it only the lack of software? Or if this would be available: is it so time consuming if you have to re-adapt all the time, that this will be something that will be very hard to do, or? I think there are a number of problems, for the moment. The biggest problem is the fact that the quality of the CT-scan, the MVCT scan, is not good enough to very precisely delineate our volumes. And then, eventually, you could cope that by matching your initial dosimetric CT-scan to your ... so you could cope that to match the initial volume with the, for instance, the PET-volume; and then afterwards redo the PET scan and refuse the examination and see what it gives. But this is all quite experimental for the moment, so ... And time consuming. And very time consuming. That's the first problem, the other problem is the gray-scaling, so the Hounsfield units who are quite variable ~, so you have to recalibrate, if you want to do that. So, for the moment, the only really option for adaptation is just to restart all over again: so make another CT-scan, a planning CT-scan, and redo the contouring. Ok, for the moment, with the software we have, where we can do fusions. Non-rigid fusions make it much easier to redo a contouring, but still it is very time-consuming. So, as I already said, quite often it's ... if there is a big difference, then, ok, everybody agrees: ok, we should replan. If it's a small difference, then, from experience, I know, we rather close our eyes, and say: ok, we'll check out tomorrow again. And then the day after: ok, the difference is a bit less. And then the next day it's even worse, and then ... The problem is now also with this 3D-imaging, that we know much more than before. So things that passed easily in earlier days, now you see that it's a problem, and that it's not correct anymore. And, also, for instance if your spinal cord is not matching with your tumor anymore; so I mean that your tumor has changed in position, in regard to your spinal cord; what are you going to do? You cannot let it pass. But, on the other hand, is it critical or is it not critical? So, you try to see whether your safety marges are still sufficient or not, and then you have to make a decision: ok, I restart all over again, or I continue ... And then, this is a very difficult decision. So, we're waiting for the physicists to develop the system, with on-line, daily on-line adaptive treatment.

EPIDs

On-line electronic portal imaging devices have been developed for acquiring megavoltage images during patient treatment. Megavoltage images, obtained in digital format with such a device, are then used for further analysis, mainly for determining set-up errors. The image information can, however, also be related to the dose delivered to the EPID, yielding dose information in a plane instead of in one or a few points.32
Ok, so there is still a way to go before it's really personalized care in radiotherapy. Maybe in the future we are going there. Ok, I have no further questions. I don't know if the promotor Jan Vermorken promotor has any questions?

Jan Vermorken promotor

Oh well. Sigrid Stroobants: That's unusual. That's very unusual. That's a good reason to start with that ... I was just ... everything is in line and understandable between the first study7 and the second study8. I was just wondering ... the post-operative cohort patients: it surprised me that you have a better local control when you used thomotherapy. And that is true for the post-operative patient group, but also those who are being treated definitively, without operation. And still the survival is worse. Now, you don't say a lot about that, because ... and most probably you can not give a good explanation. I was just wondering, just to hear you say, it is not because with tomotherapy you induce second tumors, by which they might die. And I expect it's not. Having more distant metastasis may coincide with a better local control, that's possible. But I don't understand how you percentage your second primary tumors. Can you just say a few words about that? No,it's just that ... I'm not sure, but one of the hypotheses is that, as you already ... there are two reasons, in my opinion. The first is that these patients have already been operated before, so your ~ follow-up time in fact, or after the first treatment is bigger than for the other cohort. So this in fact is not 3 years, but probably 3 years and 4 months. Ok, that's not a major difference, but I think the major difference might be the fact that you are inducing, with your operation, you're inducing, once again, a huge bunch of growth factors ... mobilizing a huge bunch of growth factors, who can only influence your primary tumors, your second primary tumors; which already are there, but still very small at that moment. You will push them and make them grow faster because of this extra ... And is this not the case then in the first group? But in the IMRT group, you mean? Yea. I think the IMRT group, the follow up was not so long. Ok, we say that we have ... the overall survival age 3 years; but the patients themselves were only, the follow-up was only about 17 months. So, while, this study is 46 months. So I think their follow-up should be better and longer. Yea. And it's not induced, because of the influence of tomotherapy? Yea, but then I don't understand why, in the definitive group you don't have these same numbers. In the definitive group you only have 5%, so, it's ...In the other group it's 17%, so ... Sigrid Stroobants: Professor Paul Van de Heyning, also?

Paul Van de Heyning copromotor

It is really highly unusual that every promotor ... Normally we don't have time enough to raise all our questions, but during your discussion there came a very interesting point of: do you feel that it would be interesting, that when a patient is being assessed by the radio therapists and the contours have been drawn and the dose distributions have been calculated, that it should be rediscussed at the display of an oncological committee? Would that add value, to have that kind of discussions? I am sorry for all the colleagues that have to do that job, if you say yes now, of course. No, no, there are two answers. I think: yes: it would add something. But no: you should do that before your calculation of the dose. The contouring, you should do that; at that moment you should discuss, and that would really add something, when you contour with the surgeon, with the radiologist, and that's what I try to do as much as possible. And what we need. And what we need, yea.

Footnotes

  • 1. As can be found in the following on-line articles (yet not in the corresponding chapters of the physical thesis):
    • Chapter 5: ▸ http://theoncologist..., Dirk Van Gestel e.a., The potential of helical tomotherapy in the treatment of head and neck cancer, The oncologist 84, p. 367-374, 2013.
    • Chapter 6: ▸ http://www.ro-journa..., Andrea Holt e.a., Multi-institutional comparison of volumetric modulated arc therapy vs. intensity-modulated radiation therapy for head-and-neck cancer, Radiation oncology 8, p. 26, 2013.
    • Chapter 9: ▸ http://theoncologist... / https://www.research..., Helical Tomotherapy in Head and Neck Cancer. A European Single-Center Experience, The Oncologist 20 (3), 3/2015, p. 279-290.
  • 2.http://globocan.iarc..., Jacques Ferlay e.a., Globocan 2012 v1. Cancer incidence and mortality worldwide. IARC Cancerbase 11, Lyon: International agency for research on cancer, 2013.
  • 3.http://globocan.iarc..., Jacques Ferlay e.a., Globocan 2012 v1. Cancer incidence and mortality worldwide. IARC Cancerbase 11, Lyon: International agency for research on cancer, 2013.
  • 4. a. b.
    • Chapter 1: general introduction.
    • Chapter 2:
      http://www.touchonco..., Dirk Van Gestel e.a., Technological advances in external beam radiotherapy for head and neck cancer, Oncology & hematology review (US) 9 (2), p.109-114, 2013.
    • Chapter 3:
      http://theoncologist..., Vincent Grégoire e.a., Intensity-modulated radiation therapy for head and neck carcinoma, The oncologist 12 (5), p. 555-564, 2007.
  • 5.
    • Chapter 4: ▸ http://www.birpublic..., Dirk Van Gestel e.a., Retrospective analysis of intensity modulated radiotherapy results in patients with head and neck cancer. A European single-center experience, British journal of radiology 84, p. 367-374, 2011.
    • Chapter 5: ▸ http://theoncologist..., Dirk Van Gestel e.a., The potential of helical tomotherapy in the treatment of head and neck cancer, The oncologist 84, p. 367-374, 2013.
    • Chapter 6: ▸ http://www.ro-journa..., Andrea Holt e.a., Multi-institutional comparison of volumetric modulated arc therapy vs. intensity-modulated radiation therapy for head-and-neck cancer, Radiation oncology 8, p. 26, 2013.
    • Chapter 7: ▸ http://www.ro-journa..., Dirk Van Gestel e.a., RapidArc, SmartArc and TomoHD with classical Step and Shoot and Sliding Window IMRT in an oropharyngeal cancer treatment plan comparison, Radiation oncology 8, p. 37, 2013.
    • Chapter 8: fast helical tomotherapy in head and neck cancer treatment. Is time priceless? (Submitted).
    • Chapter 9: ▸ http://theoncologist... / https://www.research..., Helical Tomotherapy in Head and Neck Cancer. A European Single-Center Experience, The Oncologist 20 (3), 3/2015, p. 279-290.
  • 6.http://natap.org/201..., Shanthi Marur e.a., HPV-associated head and neck cancer: A virus-related cancer epidemic. Review, The Lancet Oncology 11, p. 781-789, 6/5/2010.
  • 7. a. b. c. d. e. f. g. h. i. j. Chapter 4: ▸ http://www.birpublic..., Dirk Van Gestel e.a., Retrospective analysis of intensity modulated radiotherapy results in patients with head and neck cancer. A European single-center experience, British journal of radiology 84, p. 367-374, 2011.
  • 8. a. b. c. d. e. Chapter 9: ▸ http://theoncologist... / https://www.research..., Helical Tomotherapy in Head and Neck Cancer. A European Single-Center Experience, The Oncologist 20 (3), 3/2015, p. 279-290.
  • 9.http://www.ncbi.nlm...., Kian Ang & Erich Sturgis, Human papillomavirus as a marker of the natural history and response to therapy of head and neck squamous cell carcinoma, Seminars in Radiation Oncology 22 (2), p. 128-142, 4/2012.
  • 10.http://www.estro.org..., Markus Alber e.a., Guidelines for the verification of IMRT, Brussels: ESTRO, p. 45, 2008.

    Used with permission.
  • 11.http://www.ncbi.nlm...., Intensity modulated radiation therapy collaborative working group, Intensity-modulated radiotherapy. Current status and issues of interest 51 (4), p.880-914, 15/11/2001.
  • 12. a. b. c. d. e. f. g. h. Chapter 6: ▸ http://www.ro-journa..., Andrea Holt e.a., Multi-institutional comparison of volumetric modulated arc therapy vs. intensity-modulated radiation therapy for head-and-neck cancer, Radiation oncology 8, p. 26, 2013.
  • 13.http://www.ncbi.nlm...., Neil Burnet e.a., Defining the tumour and target volumes for radiotherapy, Cancer imaging 4 (2), p. 153-161, 2004.
  • 14. Page 102 reads: "The collimator angle was typically set to a value between 10° and 30° (or alternatively between 330° and 350°) to avoid tongue-and-groove effects."

    As can be found in chapter 6: ▸ http://www.ro-journa..., Andrea Holt e.a., Multi-institutional comparison of volumetric modulated arc therapy vs. intensity-modulated radiation therapy for head-and-neck cancer, Radiation oncology 8, p. 26, 2013.
  • 15. a. b. c. d. Figure 4: averaged absolute volume difference between DVHs for healthy tissue and p-value for pooled data: averaged absolute volume difference between DVHs for IMRT and VMAT for healthy tissue for the pooled data (black) and stratified by institute (color see legend). The p-values shown were obtained for the pooled data using a paired two-sided Wilcoxon signed ranked test.
    As can be found in chapter 6: ▸ http://www.ro-journa..., Andrea Holt e.a., Multi-institutional comparison of volumetric modulated arc therapy vs. intensity-modulated radiation therapy for head-and-neck cancer, Radiation oncology 8, p. 26, 2013.
  • 16. a. b. c. d. e. f. Chapter 7: ▸ http://www.ro-journa..., Dirk Van Gestel e.a., RapidArc, SmartArc and TomoHD with classical Step and Shoot and Sliding Window IMRT in an oropharyngeal cancer treatment plan comparison, Radiation oncology 8, p. 37, 2013.
  • 17. a. b. c. d. e. f. g. Chapter 8: fast helical tomotherapy in head and neck cancer treatment. Is time priceless? (Submitted).
  • 18. Chapter 5: ▸ http://theoncologist..., Dirk Van Gestel e.a., The potential of helical tomotherapy in the treatment of head and neck cancer, The oncologist 84, p. 367-374, 2013.
  • 19. ▸ Dirk Van Gestel, Helical tomotherapy for the treatment of head and neck cancer. An exploratory study in UZA/ZNA, ~: University of Antwerp, 20/1/2015.
  • 20.http://www.redjourna..., Tommy Knöös e.a., Volumetric and dosimetric evaluation of radiation treatment plans: radiation conformity index, Radiation oncology 42 (5), p. 1169-1176, 1/12/1998.
  • 21. As can be found in chapter 8:
    ▸ Fast helical tomotherapy in head and neck cancer treatment. Is time priceless? (Submitted).
  • 22.http://www.jmp.org.i..., Image-guided radiation therapy. Physician's perspectives, Journal of medical physics 37 (4), p. 174-182, 2012.
  • 23. As can be found in the discussion-section of chapter 4: ▸ http://www.birpublic..., Dirk Van Gestel e.a., Retrospective analysis of intensity modulated radiotherapy results in patients with head and neck cancer. A European single-center experience, British journal of radiology 84, p. 367-374, 2011.
  • 24. Table 5: late toxicity related to therapy in group A.

    As can be found in Chapter 9: ▸ http://theoncologist... / https://www.research..., Helical Tomotherapy in Head and Neck Cancer. A European Single-Center Experience, The Oncologist 20 (3), 3/2015, p. 279-290.
  • 25. a. b. Table 6: late toxicity related to therapy in group B.

    As can be found in Chapter 9: ▸ http://theoncologist... / https://www.research..., Helical Tomotherapy in Head and Neck Cancer. A European Single-Center Experience, The Oncologist 20 (3), 3/2015, p. 279-290.
  • 26. Figure 1: treatment results in patients receiving intensity-modulated radiotherapy as a definitive non-surgical approach.
    As can be found in chapter 4: ▸ http://www.birpublic..., Dirk Van Gestel e.a., Retrospective analysis of intensity modulated radiotherapy results in patients with head and neck cancer. A European single-center experience, British journal of radiology 84, p. 367-374, 2011.
  • 27. Table 1: patient characteristics.
    As can be found in chapter 7: ▸ http://www.ro-journa..., Dirk Van Gestel e.a., RapidArc, SmartArc and TomoHD with classical Step and Shoot and Sliding Window IMRT in an oropharyngeal cancer treatment plan comparison, Radiation oncology 8, p. 37, 2013.
  • 28. Table 1: patient characteristics.

    As can be found in chapter 6: ▸ http://www.ro-journa..., Andrea Holt e.a., Multi-institutional comparison of volumetric modulated arc therapy vs. intensity-modulated radiation therapy for head-and-neck cancer, Radiation oncology 8, p. 26, 2013.
  • 29. Chapter 6: ▸ http://www.ro-journa..., Andrea Holt e.a., Multi-institutional comparison of volumetric modulated arc therapy vs. intensity-modulated radiation therapy for head-and-neck cancer, Radiation oncology 8, p. 26, 2013. Chapter 7: ▸ http://www.ro-journa..., Dirk Van Gestel e.a., RapidArc, SmartArc and TomoHD with classical Step and Shoot and Sliding Window IMRT in an oropharyngeal cancer treatment plan comparison, Radiation oncology 8, p. 37, 2013. Chapter 8: fast helical tomotherapy in head and neck cancer treatment. Is time priceless? (Submitted).
  • 30.www.thegreenjournal...., Kristi Hendrickson, Hypoxia imaging with F-18 FMISO-PET in head and neck cancer. Potential for guiding intensity modulated radiation therapy in overcoming hypoxia-induced treatment resistance, Radiotherapy and oncology 101 (3), p. 369-375, 12/ 2011.
  • 31.http://annonc.oxford..., Achille Pich, Prognostic relevance of cell proliferation in head and neck tumors, Annals of oncology 15 (9), p. 1319-1329, 2004.
  • 32.http://www.estro.org..., Markus Alber e.a., Guidelines for the verification of IMRT, Brussels: ESTRO, p. 34, 2008.

    Used with permission.

Add new comment