What ever the Ab you have used you have to count the vessels in the tumour area. This is what I am doing.

Are you just interested in the number of vessels, or diameter changes as well. If you are specifically looking at tumor tissue, it may be more relevant to just consider vessels within the tumor, and/or compare to surrounding tissue. I also think the latter (vessel area/tumor area) is more appropriate to gather information about neovascularization and vasodilation. Although you could also quantify number of angiogenic sprouts and number of vessels larger than a particular diameter to capture this info.

As far as software, you could either just count manually (which is tedious and time consuming) or find a way to segment vessels. ImageJ is free and can do some segmentation, or as I have done, you can write a program in MAtlab to segment and quantify particular parameters (though this is not trivial).

Hope this helps.

Thank you Sarah for the elaborated answer. We are counting manually we did tried a software northern eclipse but did not work.

@Sarah Ruderman

Thanks for your reply. Which plugin do you use in ImageJ for the segmentation?

I use Nikon's Elemnt software, it is expensive but really verssatile to do these measurements.

Fiji (fiji.sc) should have a number of different segmentation possibilities. It is a highly annotated version of ImageJ (and is also free) that is groomed toward biomedical research.

Aside from that, it may be worthwhile to use multiple vascular markers on test slides to verify that your staining is actually capturing all of the vessels you are interested in. Alternative markers that work well for FFPE are CD31 (SZ31 monoclonal from Dianova) or Griffonia simplicifolia isolectin B4 (Invitrogen or Vector labs are good sources).

A third thing to consider is that tumor vasculature can be highly variable within different regions of the tumor for certain types of tumors. For example, a necrotic core can be largely avascular. So, it's important to make sure that you are consistent on where you are making your measurements and at what stage you are making them. Also given that tumor can have very unusual capillary diameters, you could easily get larger vascular area in cross section but a relatively normal capillaries/mm2 measurement, which gives you slightly different information. So, if you can automate the process, it would be worthwhile to do both measurements, since they are not interchangeable.

Hi Joshua, thanks for your answer.

This is really useful information. I was thinking about the different regions within a tumour. I'm comparing different mouse genotypes so they will have various degrees of tumour area size, and some will have tumour and normal tissue mixed together.

I noticed a number of obvious blood vessels around the individual tumour nodules so I was wondering whether you would only quantify the number of vessels within the nodule or also those just around the nodule, and if so how far would you go from the nodule?

While I am less familiar with the spatial differences in tumor microvasculature (most of my work is in ischemic muscle tissue, which also has large spatial variation), the regions inside and outside of a tumor nodule will likely have structurally different microvasculature. Within the tumor, the environmental cue will be substantially different from matrix composition to growth factors, and the resulting inbalance between stabilizing ECM components and growth factors is believed to induce the abnormal appearance of tumor microcirculation. From perfusion studies in our lab using contrast enhanced ultrasound (mouse flank model of glioma), we tend to see a ring of high perfusion just around the edge of the tumor. While it's difficult to match those measurements exactly with histological sections, it seems like that the highly perfused areas are both within the tumor nodule and in the surrounding region (though with our glioma tumors, there is a lot of invasiveness, so are likely some amount of tumor cells in that tissue as well). It would be worthwhile to find an identifying edge within the tumor and then set an educated, but ultimately arbitrary boundary on either side of that edge for analysis of angiogenesis both within the tumor nodule and outisde of the nodule (just basic visualization of stained tissue should help with choosing that distance). I expect the changes microvascular structure in the regions outside of the nodule will likely be more driven by hemodynamic forces than directly secreted factors from the tumor and vice versa for within the tumor.

You can try to use Adobe photoshop CS5 to quantify CD31 stained areas. 

you can also try using Scion Image, MetaMorph

Microvessel parameters MVD, VC, VCSA, %TVCSA, and TVBD were

were calculated using the following relations:

•

MVD = (Total number microvessels in high power 5 fi

elds)/(Total area of the 5 fields)

•

VCSA = {π ·(VC)2/4}.

•

%TVCSA = MVD × VCSA × 100.

•

TVBD = π ×MVD×VC

Correlation of microvessel parameters in invasive ductal carcinoma breast and fibroadenomas: a morphometric study | Request PDF. Available from: https://www.researchgate.net/publication/308758543_Correlation_of_microvessel_parameters_in_invasive_ductal_carcinoma_breast_and_fibroadenomas_a_morphometric_study [accessed Aug 24 2018].