In some experiments for investigate drug effects, we are generally using mice which have 100 mm3 tumor volume. This was a general process within ethical rules. I wonder how was this general process determined? Is there have any reason for using 100 mm3 tumor volume when it's starting therapy? Any reference is appreciated.
this is, by experience, a reliable tumor volume for your in vivo experiments for two reasons:
1) Consider the difference in body size between human and mice.
A tumor volume of 100 mm3 represents a good example of a tumor that you would normally treat in human. Actually, is even much bigger, but is it difficult to measure xenografs whose size is below an average of 100 and you can not actually assess a qualitatively good in vivo study unless you know your starting point.
2) It si very very difficult to observe treatment response when a xenografts riches high size.
I hope this helps.
best
Monica
this is, by experience, a reliable tumor volume for your in vivo experiments for two reasons:
1) Consider the difference in body size between human and mice.
A tumor volume of 100 mm3 represents a good example of a tumor that you would normally treat in human. Actually, is even much bigger, but is it difficult to measure xenografs whose size is below an average of 100 and you can not actually assess a qualitatively good in vivo study unless you know your starting point.
2) It si very very difficult to observe treatment response when a xenografts riches high size.
I hope this helps.
best
Monica
Another consideration is vasuclarization. Tumor growth is slow until a blood supply is recruited and then more rapid. Somewhere in the 100 to 200 mm3 size is when growth becomes more rapid and i.v. administered drugs become accessible to tumor cells.
I second Monica's responses. If you're going to try and evaluate the effect of a treatment on behavior (here, growth) of established tumors, then you have to choose a point in the process of tumorigenesis where you feel that you can examine the tumors with sufficient precision and/or minimal random variance (in this case, measurement error). This could be determined empirically, i.e. someone could examine the variance in tumor measurements over time (likely the variance decreases as the tumor size increase), and perhaps the variance becomes dramatically smaller when the tumors reach a particular size. This could depend on injection site, injection depth (s.c. or s.d.) and technique/skill.
Also you have to bear in mind that for your specific situation, ideally you'd like to be able to measure both increases (in the vehicle group) and decreases (in the therapeutic group) in the volume of an established tumor with acceptable precision, and as a function of time. Perhaps "100 mm3" is really just a convenient biomarker for your choice of tumor cell/injection conditions, which represents a known "mid" point, where you know the tumor will grow to a maximum-acceptable size Y within X time. This choice provides some confidence that any results are likely due to the intervention, and not simply that the tumor cells failed to thrive for some other reason (i.e. the tumors weren't really established in the treatment animals, for some other reason).
I agree with many points above. But wanted to add in my own words.
- There is often some variation in injection of cells, and this helps normalize and keep error bars smaller.
- Maybe you forgot to inject that mouse, or the cell spilled out. This way you know there were tumour cells there before treatment. Often you are expecting tumours to grow. Of course if you expect reduced growth rates you may get reduced take rate but then there is not point to treat the mouse.
- Also I think its a good idea to cull the mice before or max of 1000mm3, I have been to several conferences where larger sizes were displayed and this is cruel to the mouse and unnecessary. Also you rarely see good phenotype changes at the molecular level the larger the tumours, that's why people often have a cull point earlier on for a portion of mice.
Another some considerations are:
1. Type of tumor model. The aggressive tumor model may be need treatment earlier.
2. What formulas are used for tumor volume calculation? There are three formulations:
1) V = (L x W x W)/2
2) V = (4/3) × π × (L/2) × (L/2) × (H/2)
3) V = π/6 (L × W × H)
Where V is tumor volume, W is tumor width, L is tumor length and H is tumor high (or D, D is tumor depth). In the beginning, tumor H is less than W. The tumor volume in formulas 3) is smaller than formulas 1) & 2), but it is more accurate and closer tumor weight.
3. Depend on test drugs
@Monica Bartucci, William H Gmeiner, Jason M Fritz, Jessica Lilian Bell and Guozhong Qin:
Hi, thank you for your useful information. So the size of tumor volume was determined mainly by experience. In consideration of the vascularization, the body size, the reliability of data, the tumor models and the drugs, we choose 100~200 mm3 for treatment initiated. Got it, thank all of you very much!.
I do agree with the above answers, but rather than the size, I think that the delay between engraftment and treatment is most critical. You must allow the tumor cells to nest properly and start growing indeed before initiating the treatment if you want to be relevant, because the longer you wait, the better is the tumor micro-environement which is going to mimic an actual tumor in human. Actually 100 mm^3 is usually considered as the time when you can detect the tumor by palpation and measure it with a caliper with an acceptable precision. If you use another way to monitor tumor growth (e.g., imaging) then you can afford injecting much fewer tumor cells at the very beginning, wait for the cells to properly nest, and start treating, whereas tumors are far below 100 mm^3 and still invisible upon palpation actually,. It makes great experiments because they going to last months (rather than weeks when you inject standarly 10^6 cells s.c.). Once again, rather than reaching a certain size, usually a 7-12 days delay must be observed to disciminate positive from negative xenografts, identify the animals with homogenous size of tumors (otherwise big SD will prevent your stat to be significant eventually), and be sure that the tumors have settled. The limiting point is the way you detect the tumors - the other limiting point is that there is apparently a cut-off in the number of tumor cells to be injected, based upon my own experience below 70 000 cells getting a batch of homogenous tumors at 10 days is tricky, I suppose because you need enough stem cells at the beginning to kick the growth, or because cell mortality when handling the cells prior to grafting has a greater impact. About the various formulas to transform voulumes into weight, they are all basically based upon the assumption that most tumors are spherical - I don't think that one formula is better than another one. As long as you still use the same way to make your measurements, it's going to be fine and you can test and compare your different groups. I hope this helps? Best,
Hi, Joseph, thank you very much for your useful information. Your experience for establishing xenograft tumor models and suggestions are helpful for my experiment.
Hi Kelong,
In general, the treatment is given to the mice when the volume of the tumor is in the range of 100 to 350 mm3. Because with such volumes it is possible to have a reliable tumor growth (the control group of mice that did not receive treatment would exhibit a tumor volume significantly greater than the group that received anti-tumor treatment).
Kalthoum
In addition to the suggestion to inject cells, we usually transplant tumors from one mouse into the other. This can easily be done by making a small incision. Tumors are obtained from a mouse with a good growing tumor taking out and cut into small pieces (2 x 2 x 2 mm) of viable tissue. These tumors usually grow more easily than tumor established from cells. This is also the procedure when deriving patient derived xenografts or keeping them established. We prefer this method since oyu also transpant part of the tumor microenvironment.
Frits Peters
Hi, Prof. Peters, thank you for your suggestion. For some types of tumor cells which are hard to form into tumors in nude mice, the method you described is very helpful!.
Kelong
Just one more comment: this starting point for tumor treatments has also been chosen in view of animal welfare. The initial size is sufficient for proper detection, but it also allows further tumor growth within a time window sufficient to assess the potential therapeutic impact of your drug/treatment before the tumor size can cause significant discomfort to the animals. So, ethical rules were also taken into consideration when this size was determined.
Hi Kelong,
I completely agreed with all above answers.
I just need to mention that using 100mm3 size will also reduce the variability of in-vivo experiments, which is very important on latter stage when you will do analysis and applying statistic.
If a mice won't show tumor a growth properly or not in comparable to others, we won't use that mouse for treatment experiment as it will just add more variation and bias
I would also recommend to do randomization before doing the treatment
May be a suggestion for analysis(irrelevant to your question): if you tumor is slow growing in general, then its good to show the change in tumor size, keeping initial tumor size as 100% and compare it between vehicle and drug-treated group
good luck for your experiment
Good point indeed. Keeping variability at its minimum is critical when handling xenografted mice- otherwise your data will be just worthless, not to mention confounding factors interfering with the final interpretation of your data, and poor stats eventually. In some labs I have been in touch with (ie, big pharmas), beside tumor mass, PI starts by evaluating tumor growth pace of each animal over a short period of time as a preliminary observation. Randomization for treatments or different dosing is only performed next among mice displaying tumors of homogenous size, plus similar tumor growth. Any animal with slower or faster growth prior to starting treating is not included in the experiment. Because usually sample size are kept as little as possible to comply with the 3R rule, big error-bars are public ennemy n°1 -
Apart from restrictions imposed by your IACUC (institutional animal care and use committee in the US jargon) or equivalent on tumor size/burden per animal, you need to consider the nature of the experiment. For example, a typical standard growth delay assay might look at the effects of a drug in terms of the time it takes for a tumor to e.g., double in size relative to the same time in the vehicle treated controls. The expected end size of the tumors must also be within appropriate volumes - and this is usually related to the starting size.
With respect to numbers (also raised in this thread) - common mistakes are to use a small number (n=5 is very common but is often inadequate) that may not be sufficient to test the hypothesis, or to use the wrong statistical test to analyze the data. It is important to get a statistical power analysis done before you start, so that you use the minimum number of animals needed to power the test to give you an answer that is most likely to be interpretable and correct.
If you are not an expert, find a friendly biostatistician and they will help you with the study design and analysis. They will likely want to know what the goal is of the study/experimental endpoint(s), what you know about the growth characteristics of the tumor model, whether you expect to have censored data (e.g., lose some animals to toxicity), etc.
I hope this is helpful...Bob
You're welcome. Not sure if I uploaded it (and its a bit out of date by now, since it was written in 1997!!) but there's a Breast Cancer Research and Treatment review (vol 46: pp255-278) we wrote that addresses some of the experimental design issues for these types of studies. Maybe I'll get round to updating it one of these days. Some of it should still be useful (hopefully).
If you want to read it and cant find it on my ResearchGate profile, let me know and I'll see if there's a copy around somewhere for you...Bob
Hi, Prof. Clarke, I have got the paper from your RG profile. I will read it in detail. And I think it would very helpful for my research design. Thank you very much!
The size of 100 is good to heterotopic xenograft in cancer animal model, but in ortothopic xenografts, some time, 100 is no possible to obtain. How it will be possible 100 at colon, stomach, prostate, ovary cancer models, specially if the research is done with Balb c nu/nu, a very little murine? Other consideration is the body % area, size of tumor and the drug test concentration into the tumor. Is it possible reach the sufficient concentration in the tumor center with big size tumor? So, there are questions no answer yet.
I know this is an old thread but I have seen that when using Syngeneic mouse models (CT26/ BALBc), there is a heterogeneity in tumor size and therefore my SD values are pretty big as the drug works best on smaller tumors which makes sense. I know I could inject a lot of animals and choose the ones that have similar tumor size. I just wanted to know what you guys feel about Injecting the animals, randomizing them and starting the treatment even if the tumors aren't palpable, say on Day 3 after tumor injection. With CT26/CTLA4, I always get 100% success with getting tumors but they are of variable sizes.
You always have a control grupo, aren't you?
If yes, you could:
1. Standardize by Ctl value or
2. Make a polynomial curve to compare in a period of time. Use a caliper to do the measures, dayly.
You should always treat when the tumor reachs 100mm3, because in this size it has the same conditions: cells number, nutrition capacity and hypoxia tumoral are identical.
100 mm^3 is a soft standard. We wait until they reach from 50-150 mm^3 before beginning treatment (http://altogenlabs.com/xenograft-models/breast-cancer-xenograft/t-47d-xenograft-model/) in randomized study groups. The real reason is that you need sufficient volume for numerical calculations, but in the end if your chemotherapeutic agent is effective then the initial volume won't have that significant of an impact on comparative effectiveness.
- Badges
- Science topic
- Similar topics
- Medicine
- Oncology
- Cancer Research
- Cancer Biology