Thanks, Saeed, this is useful! What about the issue that the tumor center can be necrotic and living tumor cells are closer to the periphery?

I think think it is not important radioation dose ti the tumor isocenter or to the tumor periphery. It is important that you are going to give how much dose to which part of the tumor. General, I think terapeutic dose should cover all tumor that can see in radiologic film.

It is not clear nor is there a universal answer to this. Saeed is quite right that in normal clinical practice you aim for 100% of the tumour to get full dose. (For many cases there are treatment guidelines limiting how much the planned dose is allowed to vary over the tumour.) This is a simple technique but it is robust and can be safely done within the current clinical workflow.

However it is an assumption and is treating the tumour as homogeneous. Parts of some tumours may be more radioresistant (like the necrotic core example you gave). Standard structural imaging is unlikely to be enough to let you map out this radioresistance. Really you need functional imaging and possibly even multiple rounds as the treatment progresses.

As another complication people have yet to really measure the efficiency of "shell dose" plans. ie high doses shaped around the periphery to kill the tumor by damaging it's vascular supply rather than the tumor cells themselves.

Thanks, Anthony and Chunzhi! Have there been any studies trying to test which BED - central or peripheral - is a better predictor of tumor control probability? I am wondering because for some techniques the difference between these BEDs can be say 20%, but for others even 50%.

The "ideal" dose distribution could be planned starting from the best possible knowledge of the tumor characteristics. The kind of information can be obtained from PET or MRI In particular with Contrast Enhanced MRI is now possible to know which region is hypoxic, necrotic, well perfused and peripheral normal tissue [1]. With protons or heavy ions radiation the answer would be: try to eliminate the hypoxic region, which is generally (but not always) in the central part of the tumor. The hypoxic region is the true "head of the snake" which, with chemical messages, controls the tumor growing, vessels recruitment and so on [2]. Unfortunately with low LET radiations the dose required to destroy this region would be so high that it could severely damage adjacent tissues and organs. For this reason several dose “enhancer”, like hyperthermia or specific drugs can be useful to make perfused the hypoxic region. Fractioning the radiation dose is another way to try to overcome the problem. My personal opinion [3] is that the very best treatment would be to use High Intensity Focused Ultrasound (HIFU) to destroy the hypoxic region of the tumor and complete the treatment with X-Rays on a larger field. In this way an additional benefit could be obtained: around the “ablated” region, HIFU creates an Hyperthermia field that could potentiate the radiation effect.

[1] R. Stoyanova, K. Huang, K. Sandler, H. Cho, S. Carlin, P. B. Zanzonico, J. A. Koutcher, and E. Ackerstaff. Mapping Tumor Hypoxia In Vivo Using Pattern Recognition of Dynamic Contrast-enhanced MRI Data. Transl.Oncol. 5 (6):437-447, 2012.

[2] M. W. Dewhirst, Y. Cao, and B. Moeller. Cycling hypoxia and free radicals regulate angiogenesis and radiotherapy response. Nat.Rev.Cancer 8 (6):425-437, 2008.

[3] G. Borasi, G. Russo, F. Alongi , A. Nahum, G.C. Candiano, A. Stefano , M.C. Gilardi, and C. Messa. High-intensity focused ultrasound plus concomitant radiotherapy: a new weapon in oncology? Journal of Focused Ultrasound 1 (6):1-4, 2013.

Thank you, Giovanni! Your answer makes a lot of sense. However, there appear to be studies (e.g. Joern Wulf et al, Dose-response in stereotactic irradiation of lung tumors. Radiotherapy and Oncology, 11/2005; 77(1):83-7, Quynh-Thu Le, Results of a phase I dose-escalation study using single-fraction stereotactic radiotherapy for lung tumors. Journal of thoracic oncology, 11/2006; 1(8):802-9) which suggest that peripheral doses may be most important for tumor control. I would be grateful for your (and everyone's) comments!

Thanks Igor!

The result you quote may be explicated by considering that the central region of the tumour would be hypoxic. In this case it may be little (or not at all) sensitive to low LET radiation. The same may be true if that region would be necrotic or a mixing of the two. It would be of the maximum interest to have information of the extension of these regions in the patient tumour. Unfortunately the Authors of the first study didn't do (or report) such a study. I don't have access to the second paper.

All my best,

Gianni

There is dose inhomogeneity within the tumour and in clinical prescription practice ! During computer-aided treatment planning, the tumour centre is most often aligned to the machine isocentre - often the point of dose normalization (= 100%) .The radiation oncologists often prescribe the dose to that POINT (e.g. 76 Gy to isocentre). However, the dose is not uniform throughout the clinical target volume (CTV) and an isodose shell may be used (e.g. 95% (of 76Gy) isodose shell to encompass the CTV). A more recent parameter is D95 - a check point on the integrated dose-volume histogram curve. D95 is the dose (in Gy) that 'covers' 95% of the CTV volume. 95% of this volume gets at least the D95 dose.Finally, radiobiological issues (e.g. hypoxia) may lead to an advantageous hotter region at the tumour centre (as used in radiosurgery). In the future, it is likely that non-uniform doses in the tumour may be optimized as we image internal tumour growth and radioresponse characteristics. Not all tumour cells are the same and their local environment also differs. This summarizes many of the earlier statements by other contributors whom I thank. J2B

Thanks, everyone! I suppose the question comes down to the following - do most failures of tumor control result from: (1) inability to kill the most radioresistant (e.g. hypoxic) cells (which may be located close to the center), or (2) insufficient dose to small parts of the tumor (e.g. "projections" from the periphery) which can grow back. What do you think?

Also, for Giovanni and anybody elso who may be interested - attached is the Le et al. paper I mentioned earlier.

Thanks Igor and anybody. The idea of 'projections' it's highly stimulating and may explain the bad results of too localized therapies like IORT as the only radiation therapy in breast cancer. I'll come back next week after reading the paper.

Nice weekend to anyone.

Gianni

The question is not so much about killing the "central" or "peripheral" regions of a tumour. It is more about killing cells that are radioresistant (e.g. hypoxia) or most likely to repopulate the tumour (e.g. rapid doubling time), wherever they are! Giovanni rightfully suggest ways to image or map out these regions for regional dose 'boosts' using optimized forms of energy and radiosensitizers.

A related question is "where do tumours recur most often clinically and do these match the regions of inadequate radiation dose?". With advances in 3D/4D/molecular imaging and 3D/4D dose sculpting (IMRT), this question will be resolved during the next 5-10 years...and lead to a new strategy of "inhomogeneous dose to inhomogeneous tumours"... There is evidence of 'dominant' lesions within prostate tumours already but the mapping with MRI, PET, and CT is discordant so far.Time will tell.

J2

Thank you, Jerry!

I would be grateful if you provide references about the evidence for 'dominant' lesions within prostate tumors you mentioned.

First question is about radiobiology of hypoxia.

All theories about hypoxiaare based on hydrolyse of H20 and linear quadratic model

Several factors must have a role in a new approach:

- Dose gradient with some techniques like IMRT

- Angiogenesis

- Nanoparticles

We just have started....

Good explanation from different people. I would like add bit more.

Generally in treatment planning system of radiation therapy workflow, the center of the tumor is considered as the beam isocenter for the radiation delivery. Even though the cancer is spread till periphery of the marked(segmented) tumor. In the case of photon radiation distribution, the dose looks somewhat in homogeneous and in case of protons or ions, it shows homogeneity. Like how the intensity of the earthquake tremors gradually decreases as it moves from its epicenter, the quantity of the dose distribution also decreases as we move from isocenter to periphery. Technically these layers of dose is called isodoses which represents different effective doses like 20mSv(at the isocenter) -> 16mSv -> 10mSV ->5mSv(at periphery) etc.

The above explanation depicts an ideal situation. But treatment planning softwares allow to place the radiation beam isocenter even at periphery also. More information can be obtained about doses from ICRU reports from http://www.icru.org/ about the radiation quantities

Dear Friends, I read again with more attention the papers quoted by Igor: (Le, J.Thoracic Oncol, 2006 and Wulf, Radiot.Oncol, 2005). Both refer to stereotactic irradiation of lung tumors. Both relate a better outcome to the higher dose to the tumor periphery : that means that the central part of the tumor receives an even higher dose (1/65% or 1/80% higher) that the in the periphery. The second paper states an equivalence in terms of BED: TCD50 with 94 Gy at the isocenter or 50 Gy at PTV margins. In addition, by using a the “stepwise multiple regression” they state that the BED at the PTV margins is the only significant factor. As before, that means that an even higher dose is given to the tumor center. In other words, from the two papers we conclude that an higher dose to the center and an adequate covering of the tumor region is a good strategy, even in presence of some disomogeneity related to the stereotactic irradiation (and taking into account the toxicity). The conclusion that “The dose at the periphery is more important than the central one” would, in my opinion, require a different kind of irradiation, in which the dose in the periphery is increased while the central dose is decreased. In other words the highest dose should be inside a “shell” around the tumor. That means a totally new irradiation strategy and this kind of experiment, of course, need to be done first on mice.

Note: in the “stepwise regression” is necessary to evaluate and report the correlation factors between all the variables (in our case, in particular, between the dose in the center and at periphery)

Thanks a lot, Clemens! Your references are indeed highly relevant - I will read with interest!

Thanks a lot, Clemens. I'll read the papers you mentioned.

Gianni

I can add that, ablated the central part of the tumor, supposed hypoxic, its peripheral region, generally well oxygenated, can be irradiated without irradiating the center (already ablated), by using tumor "tangential" fields. These tumor tangential fields centered in the periphery, where tumor infiltrate the healthy tissue, are narrower (and, eventually, at low dose) give a much reduced "integral" dose to the whole region, with a great sparing of possible, radiation induced, damages to the patient.