Effects on Cancellous Bone in the Metaphysis

Abstract

Cancellous bone has high metabolic activity compared to many other bone compartments and can be affected not only by changes in physeal activity but also by perturbations in homeostasis caused by changes in physiology or on-target pharmacology. Examples of several types of resulting morphologic findings were presented; if known, the pathways causing morphologic changes were discussed.

Cancellous bone, particularly the metaphysis, is metabolically active (particularly during phases of rapid growth) and is responsive to a wide variety of pharmacologic agents. Familiarity with normal bone processes and responses is necessary to differentiate between normal variation and a test-article effect. Several examples of bone effects in metaphyseal cancellous (trabecular) bone are presented here to illustrate the variety of potential findings in this bone compartment resulting from on-target and off-target effects of several test articles.

All procedures performed on animals were in accordance with regulations and established guidelines and were reviewed by an Institutional Animal Care and Use Committee or through an ethical review process.

Basic Review of Some Structural and Cellular Topographic Features of the Metaphysis

Sexual dimorphism in the rat is recognizable in the metaphyses of long bones even in rapidly growing rats (Riesenfeld 1977). Due to their higher growth rate, male rats have thicker growth plates than age-matched females. Female rats have more cancellous bone particularly in the center of the bone and also deeper in the metaphysis than their male counterparts. The metaphyseal cortex, however, is thicker in males than in females.

There are 5 common sites used to evaluate cancellous bone in toxicology studies. The proximal tibia has a very consistent orientation of the growth plate, spongiosa, and cancellous bone. The distal femur has a much more variable orientation of the growth plate, spongiosa, and cancellous bone. The sternum has less bone growth than long bones and is better suited for evaluation of hematopoiesis. The costochondral junction of the rib stays active in large animal species longer than the long bone growth plates. Lumbar vertebrae contain more cancellous bone in older mice than in their long bones (Gunson, Gropp, and Varela 2013).

Cellularity of the metaphyseal compartment consists of 6 main cell types: osteoblasts form lamellar bone on primary and secondary spongiosa as well as on endocortical surfaces. Osteoclasts prune unneeded calcified cartilage emanating from active growth plates, control the amount of cancellous bone in the mid to deep metaphysis, and narrow the forming cortical shell on the periosteal surface at the “cutback zone” to produce the final diaphyseal bone diameter. Osteocytes are embedded in cancellous bone which provide cell signaling to respond to changes in mechanical load, local microdamage, and mineral homeostasis. Bone lining cells, located on quiescent cancellous surfaces, can differentiate into osteoblasts. Hematopoietic cells in the bone marrow space can have remarkable effects on metaphyseal bone after cell injury, and mesenchymal stem cells can differentiate to osteoblasts or adipocytes depending on cell signaling.

Bone modeling, which by definition refers to bone formation in locations without previous bone resorption, occurs in 3 locations in the metaphysis in rapidly growing animals: as part of longitudinal bone growth from endochondral ossification at the physeal/metaphyseal interface, at the cutback zone that forms and controls the final diameter of the diaphyseal cortical bone, and in the mid- to deep metaphysis controlling the amount of cancellous bone in these regions. In these locations, bone formation and resorption processes are both present, although not precisely coupled topographically. For example, at the cutback zone, osteoclasts are active on the periosteal surface decreasing the outer portion of the forming cortical shell, while osteoblasts are very active on the endocortical surface forming the inner cortical shell which will become the narrower diaphyseal cortical shell.

Bone remodeling, a process in which bone resorption is followed by coupled bone formation in the same location, occurs in basic multicellular units (BMUs) in large animals (dogs, nonhuman primates, minipigs, ferrets, etc.). Bone remodeling is a response to cell signaling to repair macro- or microdamage, to respond to changes in biomechanical loading, or in response to systemic needs for substances found in bone (such as calcium, phosphorus, etc.). One visible feature of normal BMUs is a membrane or “canopy” which separates the cells of the BMU from the microenvironment of the adjacent hematopoietic compartment; these structures are formed by osteoblast-like cells and are associated with capillaries.

Examples of Various Test Articles on Cancellous Bone

Parathyroid hormone (PTH) has markedly different effects on bone depending upon whether the exposure is intermittent (anabolic) or continuous (catabolic). Intermittent PTH exposure acts on bone through canonical wnt signaling promoting osteoblast differentiation from mesenchymal stem cells and bone lining cells. Continuous PTH exposure results in a greatly increased receptor activator of nuclear factor-κB ligand to osteoprotegerin ratio resulting in increased osteoclast formation and increased bone resorption (Evenepoel, Bover, and Ureña Torres 2016). Two images, one of each type of PTH exposure, were shown. With the anabolic response to intermittent PTH exposure, trabecular bone volume and connectivity were largely restored in an ovariectomized rat model (Gowen et al. 2000). With the catabolic response to continuous PTH exposure, osteoclasts may remove so much cancellous bone that defects in the trabecula (“tunneling resorption and/or “trabecular splitting”) can be readily appreciated in a standard hematoxylin and eosin decalcified section (Gunson, Gropp, and Varela 2013).

Inhibition of Dickkopf-1 leads to decreased osteoclast formation (Pinzone et al. 2009). In the example shown, very few osteoclasts were visible at the physeal/metaphyseal interface, resulting in dramatically increased primary and hence secondary spongiosa in the metaphysis due to the lack of normal primary spongiosa “pruning” by osteoclasts. In this example, however, the cutback zone was normal unlike the case with some antiresorptive agents (bisphosphonates) in rapidly growing animals.

An opposite effect on osteoclasts after administration of an unnamed test article was shown next. There were increased osteoclasts at physeal/metaphyseal interface, resulting in increased cartilage core “pruning” and very few short primary spongiosa; in some places, there was a notable gap between the physis and thicker secondary spongiosa. It was uncertain whether the thicker secondary spongiosa represented an independent effect on osteoblasts or a compensatory biomechanical response to the missing primary spongiosa. The mechanism of this effect is not known.

Sodium–glucose cotransporter type 1 inhibition has been previously reported as an off-target effect of high doses of sodium–glucose cotransporter type 2 inhibitors in growing rats. The microscopic appearance of this off-target effect includes increased numbers of primary spongiosa, increased metaphyseal trabecula with retained cartilage cores, and misshapened cortical bone, all due to decreased osteoclast function (Samadfam et al. 2010). It has also been reported that some of these effects can also be attributed to the source of carbohydrate (glucose vs. fructose) in rat diet resulting in increased calcium absorption in the intestinal tract (Tirmenstein et al. 2013).

Antineoplastic agents often have cytotoxic effects on the hematopoietic compartment in the bone marrow which can also affect nearby bone cells. An example of acute bone necrosis 4 days after the first dose of a cytotoxic agent was shown followed by an image which demonstrated the visible effects of cyclic dosing with the same test article as seen by linear zones of thicker metaphyseal cancellous bone separated by zones which contained little to no bone. The criteria for bone necrosis include associated bone marrow and adipocyte necrosis, bone lacunae which are either empty or contain pyknotic osteocyte nuclei, sufficient size (≥500 μm²), altered staining (pallor), and associated bone cellular response (increased osteoblast and osteoclast activity; Allen and Burr 2008). An example of a more chronic focus of bone necrosis from the sternum of a dog was used to illustrate these criteria.

Somatostatin has dramatic effects on the metaphysis of growing animals by ceasing endochondral ossification (via the growth hormone/insulin-like growth factor 1 axis), which results in no production of primary spongiosa (and hence no longitudinal bone growth) and increased adiposity in the metaphyseal bone marrow likely from an increased production of bone marrow adipocytes from mesenchymal stem cells (Hardouin, Pansini, and Bernard Cortet 2014).

In summary, when evaluating metaphyseal and cortical bone, it is important to be aware of species, age, and sex differences and use sex- and age-matched controls; be aware of bone site differences in cellularity, structure, and metabolism. The metaphysis is an area of high metabolic activity, particularly in rapidly growing animals commonly used in toxicity studies; realize that numerous pathways affect bone cells; keep in mind the tremendous cross-talk between osteocytes, osteoblasts, and osteoclasts; and use descriptive International Harmonization of Nomenclature and Diagnostic Criteria (INHAND) terminology for recording findings (Fossey et al. 2016).