This is the first study to use MRI to assess the appearance of the lymph nodes and spleen in various immuno-deficient and wild-type mouse strains. Given the substantial differences in the cellular composition of the immune systems in these mice, and reports from a small number of pathological studies which noted differences in lymph node morphology between different mouse strains, we hypothesized that the MRI appearance of the lymphoid tissues would also be disparate in the immuno-deficient and wild-type mouse strains we examined. MRI revealed some considerable differences in the appearance of lymph nodes in the different mice. Most notably, some lymph nodes in nude and SCID mice appeared with a region of signal hyperintensity in bSSFP images. By comparing the bSSFP images with T1- and T2-weighted SE images we were able to determine that the high signal intensity in the bSSFP images is most likely due to the presence of fluid within the node. Fluids have long T1 and T2 relaxation times. In T1-weighted images tissues with long T1 relaxation times appear dark, while in T2-weighted images tissues with long T2 relaxation times appear bright. Contrast in bSSFP images is related to T2/T1 and tissues with similar T1 and T2 values, like fluids (and fat), will have T2/T1 close to 1, which results in high signal intensity. It is also worth noting that when nude mice were followed over time with MRI, those nodes that appeared with a region of signal hyperintensity did not change in appearance during the imaging experiment.
There are relatively few papers that describe the anatomical or cellular features of normal lymph nodes in immune-deficient mice. In a paper by Sainte-Marie and Peng 
, 8 types of lymph nodes, in 7 nude and 4 C57Bl/10 mice, were carefully assessed by histopathology. They showed that in nude mice the absence of the thymus greatly inhibits the development of the lymphocyte population in the cortex, creating a small cavity. In addition, in some mice a cyst of variable size was found in the axillary or brachial nodes, often near the hilus 
. These findings are similar to our observations in nude mice by MRI, where fluid filled cavities were observed in the axillary, brachial and inguinal nodes.
A histological examination of the immune organs in SCID mice, by Ge et al., showed that the lymph nodes had no clear cortex and appeared to be totally devoid of lymphocytes 
. These modifications of lymph node anatomy could also be expected to result in features such as cavities or cysts. Our histology demonstrated that nodes which appeared with a region of high signal intensity in MR images, had a cavity within them, supporting the notion of a fluid-filled node.
Exposure of immuno-deficient mice to potentially pathogenic organisms must be restricted. Specific pathogen free (SPF) housing systems, often referred to as barrier facilities, are commonly used to house immuno-deficient mice and employ sterilization of feed, bedding, water, and cages along with the use of filter-top or individually ventilated caging systems and strict adherence to aseptic techniques for animal handling. Lymph nodes may enlarge when immune cells react to pathogen, such as virus or bacteria, due to proliferation of lymphocytes. Swollen glands, common to many illnesses, are an example of nodes enlarging in response to a pathogen. The immuno-deficient mice used in our longitudinal imaging studies leave the barrier facility for the first scanning session (day 7 after arrival) and thereafter were housed within an external barrier in ventilated cages. They were transported to/from the external barrier and the MRI facility for the next two scans (days 14 and 28). We hypothesized that the transfer between barriers and time spent outside of the barrier in the MRI facility, would result in changes in the size of the lymph nodes. To test this we imaged wild-type, nude and SCID mice at three time-points and measured the lymph node volumes. Changes in the lymph node volumes were measured for all mice. The largest changes were observed in the nude mice, however, the node volume in nude mice was found to be the most variable for all scans. In SCID mice the node volume was actually observed to decrease over time. However, there was no trend for increasing node size with number of times imaged (or number of transfers out of the barrier).
Even though the lymph nodes in these immuno-deficient mice are underdeveloped, or rudimentary, many studies show that the lymph nodes are a frequent site of cancer metastases 
. In fact several studies suggest that lymph node metastases are more common in the mice with the more severe immunodeficiencies 
. Dewan et al. have reported that the rate of metastasis of human breast cancer cells (MDA-MB-231) is much higher in NOG mice, compared to NOD/SCID mice inoculated in the same way 
. This included metastasis to the regional lymph nodes. It is interesting that, even though lymph nodes in NOG mice are not obvious at dissection, and not visible in MR images, the rudimentary node tissue still provides a suitable microenvironment for metastatic growth.
When lymph nodes are abnormal they increase in size 
. Enlarged lymph nodes are readily visible in MRI. In fact, traditionally, MRI of the lymphatic system has been focused on conventional anatomical imaging whereby enlargement of lymph nodes is considered the primary diagnostic criterion for disease. Secondary architectural and pathological changes are also often apparent on MRI. It is therefore important to recognize that, in the different mouse strains we imaged, the size and appearance of the lymph nodes is quite variable in healthy animals and the lymph node volumes change over time in both wildtype and immuno-deficient mouse strains. A change in the lymph node size, as measured by MRI, in these mice should not be considered evidence of disease without additional validation.
In diseased lymph nodes the tissue is sometimes homogenized so that the cortical and medullary areas are no longer differentiated 
. Necrosis may lead to accumulation of fluid (pus) within nodes, and can cause a fluid filed cavity. It is therefore very important to recognize that some diseased nodes can appear hyperintense in bSSFP and T2-weighted images (or with low signal in a T1-weighted image) and that this has the potential to be confused with normal lymph nodes in non-tumor bearing immuno-deficient mice.
The MRI appearance of the spleen in the different mouse strains was also notable. The spleens of wild-type and nude mice were quite similar in size and MRI appearance; a distinct architectural pattern is observed in MR images of the spleen in wild-type and nude mice. In SCID and NOG mice, the images show a smaller spleen (3–5x smaller in volume) devoid of pattern and in the case of the NOG mouse acellular. These differences in the MRI appearance likely reflect the impaired development of the spleen in these mice.
The normal spleen is composed of what is known as (non-lymphoid) red pulp and (lymphoid) white pulp. Red pulp consists of connective tissue and many splenic sinuses that are engorged with blood, giving it a red appearance. It functions to filter the blood and is a storage site for red blood cells 
. It is also a reserve site for monocytes, which upon injury or disease leave the spleen and migrate to tissues for repair 
. The high blood content causes it to appear with very low signal intensity in bSSFP images. White pulp consists of lymph nodules (germinal centers), composed of follicles, and periarteriolar lymphoid sheaths. White pulp is rich in lymphocytes 
SCID mice contain a defect preventing the functional development of T- and B- lymphocytes 
. Ge et al. have shown by histology that SCID mice have relatively empty splenic follicles 
. It is because of these deficits the spleen in SCID mice is rudimentary in appearance and function. By MRI it appears to have no tissue contrast within it, suggesting minimal structural features. The NOG mouse is a SCID mouse strain that has multi-functional defects in NK activity, macrophage function, complement activity and dendritic cell function, in addition to lacking T- and B-cells. In all NOG mice examined by MRI the spleen appeared black reflecting an absence of signal likely due to an absence of cellularity.
In summary, this paper investigates an important technical aspect of mouse body imaging, namely the differential appearance of lymph nodes and spleens in 4 commonly used strains of experimental mice (C57Bl/6, nu/nu, CB-17 SCID, and NOG). These strains of mice are widely used for cancer research, and imaging is often used to identify metastasis to lymph nodes and distant organs. The use of these mice is not standardized; different laboratories use certain mice for a variety of reasons, and differences in the appearance of the lymph nodes across the strains can be a source of confusion in data interpretation. We have shown that there are particular features within the nodes of some mice that can mimic the appearance of pathology. We have found that changes in the size of nodes, in healthy mice, that occur with repeated imaging fall within the typical range of node sizes which show variability. By presenting knowledge of the normal MRI appearance of the lymphoid organs in healthy, immuno-deficient and immuno-competent mice we provide information that will help to avoid data misinterpretation and to advance the field.