The tissue diagnostic instrument

INTRODUCTION

The tissue diagnostic instrument (TDI) was redesigned from the bone diagnostic instrument^{1, 2} so as to measure tissue mechanical properties subcutaneously and in vivo with additional probe assemblies and an adjustable compliance (Fig. 1). It consists of a thin probe assembly that can penetrate skin and soft tissue to reach deep tissues. The disposable, sterilizable probe assembly consists of an outer reference probe made from a 23 gauge hypodermic needle and an inner test probe made from stainless steel wire ranging from 175 to 300 μm in diameter and from 2 to 90 mm in length. Since friction between the test probe and the reference probe increases with length, it is desirable to use only the length needed to access the desired tissue location. The test probe is held in a nickel tube that couples to a magnet, which in turn is linked to a force generator. During operation the force generator oscillates the probe within the tissue of interest and concurrently measures the force and displacement. The maximum values for force and displacement are 12 N and 600 μm. The probe is typically operated at a frequency of 4 Hz because this is rapid enough to allow hand holding yet sufficiently slow to allow easy decoupling of the elastic and viscous response of the tissue (see supplementary material³ for more details including force and displacement ranges.)

MEASUREMENTS

We first illustrate TDI use in human spinal disks that are composed of a thick outer ligament (annulus fibrosus) and a central swelling hydrogel (nucleus pulposus). Spinal disk degeneration can be the underlying cause of back pain leading to significant morbidity and societal expense. Intervertebral disks are one of the most highly loaded tissues in the body, and consequently material property insufficiency can lead to damage accumulation, inflammation, and pain. Disk degeneration is currently diagnosed using imaging techniques, such as magnetic resonance.⁴ Unfortunately, these methods can only indirectly suggest disk mechanical properties, which currently cannot be measured in vivo.

Using image-guided, percutaneous placement [Figs. 2a, 2b], disk material properties can now be measured safely in vivo using a type N probe assembly [Fig. 1c]. The novel sharpening of the reference probe for type N probe assembly decreases the problem of tissue being caught between the test probe and the reference probe during insertion and thus decreases the friction between the test probe and the reference probe. The friction between the test and reference probes is typically ∼0.02 N. The specific value is recorded by the software before testing a sample and is removed from the samples’ force versus distance plot before analysis. The force versus displacement data are plotted in real time and recorded digitally [Figs. 2c, 2d]. The slope of the force versus displacement curve provides a measure of disk elasticity: in the case of a simple spring, the slope would be the spring constant. The energy dissipation in the force versus displacement curve is the area inside the curve and is a measure of the viscous behavior. Viscosity is absent from a simple spring yet is large for a purely viscous material such as petroleum jelly, which has an elasticity near zero.

There are significant differences (p values of <0.01) in slope (N∕m) and energy dissipation (μJ) between the annulus fibrosus and the nucleus pulposus (Fig. 2). The annulus fibrosus has both higher slope and energy dissipation. These results are representative of our measurements on 11 disks: the slope and energy dissipation are always greater in the annulus than the nucleus. This observation of higher slope or stiffness in the annulus is consistent with previous experiments measuring the compressive properties of both annulus and nucleus.^{5, 6} However, a precise comparison with established mechanical data is not readily available because mechanical testing in this manner at high spatial resolution has not been possible previously. Because these properties are known to change with age and degeneration, an eventual goal would be to determine whether in vivo measures of annulus and nucleus material properties provide novel data that improve back pain diagnosis and treatment. We note that the 23 gauge needle is consistent with the recent recommendation⁷ that a spinal needle smaller than or equal to 22 gauge should be used to prevent postsurgery leakage.

An epithelial tissue such as the mammary gland is an example of the softest tissue that can be probed with the current TDI. Figure 3 shows a paired-comparison of the mammary glands #2∕3 and #4 from normal Friend Virus B, mice and tumors arising in the matched mammary glands of their MMTV-PyMT+∕- (Mouse Mammery Tumor Virus-Polyoma Middle T Antigen) littermates. These data are representative of the data in an ongoing study of various tumors. The results of that study are beyond the scope of this paper, but we can report that all 30 tumors are stiffer (have higher slope) than all 15 normal mammary glands in tissue site-matched and age-matched mice. Normal murine mammary glands have elastic modulus below 1 kPa as measured with a conventional rheometer and with the TDI device. This value is comparable with our calibration curves on polyacrylamide (PA) gels (see supplementary material) that demonstrate TDI sensitivity below 1 kPa. Normal and transformed human breast tissue is considerably stiffer than mouse tissue⁸ and is therefore well within the range of the TDI. Tissue stiffness increases in many breast cancers. To quantify stiffness and improve breast tumor detection, imaging modalities such as sonoelastography and Magnetic Resonance elastography have been used.^{9, 10} The TDI offers a tractable and economical approach to measure breast stiffness in situ with millimeter resolution. Our preliminary trials on human breast tissue from cadavers showed detectable variations in mechanical properties between different locations in the same specimen with a spatial resolution of 2 mm (data not shown). Based on these observations an eventual goal would be to use the TDI for localization of human breast cancer in situ. A foreseeable clinical application includes using the device to define margins of affected tissue and sites for biopsy. The TDI measurement could easily be combined with biopsy; the test probe could be withdrawn into the reference probe to collect a biopsy sample after mechanical testing. We are currently investigating the molecular mechanism of how the mechanical properties contribute to breast cancer; the TDI can be applied to clarify the molecular link between matrix materials properties of tissues and tumor risk (i.e., breast cancer in women with mammographically dense breasts^{11, 12}).

Prior to clinically apparent symptoms of osteoarthritis, the material quality and mechanical function of cartilage matrix are compromised.¹³ The ability to noninvasively probe the material quality of this stratified tissue will complement and extend current diagnostic capabilities.¹³ Furthermore, detection of cartilage degeneration early in osteoarthritis may increase the success of therapeutic intervention. To that end, the ability of the TDI to distinguish the elastic modulus of synthetic materials with moduli comparable to cartilage was validated against well-established methods including atomic force microscopy, nanoindentation, and bulk stress relaxation (Fig. 4). In addition to accurately measuring the elastic modulus of PA gels with a range of moduli from 0.2 to 1 MPa, the TDI could measure the elastic modulus of a stiff gel that was inferior to a compliant gel, demonstrating its ability to noninvasively evaluate a stratified material [Fig. 4d]. When applied to cartilage, the TDI readily discriminated between a young healthy cadaveric specimen and an old degenerated surgical specimen that were probed in situ [Fig. 4f].

Human dentin is an example of the hard tissue that can be probed with our current device (Fig. 5). One unsolved problem for practicing dentists is deciding when a sufficient depth has been reached when drilling to remove carious dentin from a cavity. One proposed solution has been the development of a new experimental polymer bur (from SS White Burs, Inc., Lakewood, NJ), which is designed to remove soft decaying dentin but blunt on harder healthy dentin and thus self-limit the tissue amount removed. Here we show the properties of the remaining dentin after a first excavation by such a bur and then after a second excavation with a second polymer bur of the same type (Fig. 5). Next was a third excavation with a #4 round carbide bur and finally a cavity preparation into presumably sound dentin using a #330 carbide bur. Note that even after the cavity preparation, the dentin did not have the full elastic modulus of the healthy dentin. Our primary focus was on relative values as excavation proceeded. The elastic modulus, as calculated from force versus displacement curves generated by the TDI and analyzed using a modified Oliver and Pharr¹⁴ method, for the dentin left in the cavity by the polymer bur was below that for healthy dentin far from the cavity [Fig. 5c]. Please see the appendix for the details of the modified method. The hardness, as calculated from force versus displacement curves generated by the TDI and analyzed using a modified Oliver and Pharr¹⁴ method, for the dentin left in the cavity by the polymer bur was below that for healthy dentin far from the cavity [Fig. 5c]. One-way analysis of variance gives values of P<0.0001 for the elastic modulus and P=0.0008 for the hardness, indicating that the variation among means is significantly greater than expected by chance. Thus the TDI has the potential to quantify the properties of dentin left in a cavity and could be used to study the outcome of various treatment strategies for how much degenerated dentin is removed before filling the cavity.

The absolute value for the elastic modulus of our “healthy dentin” is well within the range of existing measurements but below the value of 20–25 GPa recommended in a recent critical reevaluation of the literature.¹⁵ The reason is probably the storage of the teeth in water for weeks before measurement.¹⁶ To our knowledge, the TDI is the first instrument that can measure elastic modulus and hardness inside irregularly shaped, fully hydrated dentin cavities. It could be used for research projects without further modification. For individual clinical use, a smaller, less expensive version with an angled probe would be desirable. The experiments on dentin reported here build on a rich history of measuring mechanical properties with indentation methods.^{14, 17, 18} Of special interest is recent work modeling size effects with finite element analysis¹⁹ because extensions of work such as this may lead to a more quantitative understanding of TDI measurements on soft tissue as well as hard tissue.

The hard tissue, bone, is of particular interest medically because of the growing incidence of debilitating bone fracture as our population ages.²⁰ Changes in bone material properties are believed to play a role in fracture risk.^{21, 22, 23} With the top screw backed off, as discussed in the supplementary materials,³ the TDI functions as a Bone Diagnostic Instrument,^{1, 2} which may, after clinical tests, prove useful in quantifying the component of bone fragility due to degraded material properties. Figure 6 shows tests of the TDI, working as a bone diagnostic instrument, on a living patient to determine if the procedure is painful or results in complications. Neither this patient nor the others tested to date experienced any pain beyond the initial “stick” when the local anesthesia was injected. There have been no complications.

DISCUSSION OF NEW POSSIBILITIES

It might, in the future, be possible for the TDI to measure the interaction forces between antibody coated test probes and tissues. This would allow measurements of single molecule interactions as is currently achieved with an atomic force microscope.²⁴ Rupture forces in the range of 20–140 pN have been measured for many receptor-ligand interactions with single-cell force spectroscopy.²⁵ With these interaction forces, we can make order of magnitude estimates of forces we might find when trying to rotate or translate a test probe that had bound to a tissue with many molecular bonds in parallel. Assuming a molecular density of 1 molecule∕10 nm², an interaction force per molecule of 50 pN, a coated region of area of 4×10⁻⁶ m² (the exposed area of the type D probe), and a fractional binding of 1%, we would get a force of 50 pN∕molecule×4×10⁻⁶ m²×1 molecule∕10 nm²×.01=200 mN. The present lower limits of sensitivity of the TDI for forces come from the friction between the test probe and the reference probe, of the order of 10 mN, and from the force noise in our force transducer, of the order of 5 mN in a 1 kHz bandwidth. Thus forces of the magnitude that could be expected from molecular interactions with coated tips should be measurable. A big problem could be nonspecific interaction masking specific interactions. A proof of concept experimental approach to overcoming this masking effect would be to use a test probe coated on just one side that was exposed to the tissue under test through a window in the wall of a closed-end reference probe. The difference in the forces between the test probe and the tissue under test for the coated versus uncoated side could be measured. This could naturally be extended with multiple coatings on multiple strips on the test probe, each exposed one by one through a slit in the wall of the closed-end reference probe. We emphasize, however, that proof of concept experiments will be necessary to evaluate this potential application of the TDI.

It is important to note that though the present instrument is able to make basic measurements in a wide range of tissues (almost all tissues in the human body from very soft breast tissue to hard, mineralized tissues), it is in a very early stage of development. More versatile instruments with more measurement modalities, such as mentioned above, and more user convenience features, such as wireless operation, are possible. Specialized instruments for specific measurements in specific tissues could be developed at a small fraction of the cost of the fully versatile instrument. The device could also be modified to assess materials properties of various bioengineered artificial three dimensional tissues.^{26, 27}