Dynamic contrast-enhanced MR imaging to assess physiologic variations of myometrial perfusion

Abstract

Purpose

To prospectively evaluate the ability of dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) to assess physiological microvascular states in normal myometrium.

Materials and methods

Eighty-five women (62 women of reproductive age, 23 postmenopausal) undergoing DCE-MRI of the pelvis were included. Microvascular parameters for the inner and outer myometrium were analysed using a pharmacokinetic model. These parameters were tissue blood flow (F), blood volume fraction (V _b), permeability–surface area product (PS), interstitial volume fraction (V _e) and lag time (Dt).

Results

In the women of reproductive age, the inner myometrium displayed higher F and PS, lower V _b and V _e, and longer Dt than the outer myometrium (p = 0.02, p = 0.01, p = 0.005, p = 0.03 and p = 0.01, respectively). The inner myometrium presented microvascular variations during the menstrual cycle with a pre-ovulatory peak followed by a fall reaching a nadir of F and V _b about 4 days after ovulation. Compared with women of reproductive age, in the postmenopausal state, F and V _b decreased in the outer myometrium, while PS, V _e and Dt increased (p < 0.0001, p = 0.001, p = 0.001, p = 0.03 and p = 0.0004, respectively).

Conclusion

DCE-MRI is a non-invasive technique that can measure variations of myometrial microcirculation, and thereby be potentially useful to help characterize the role and states of the myometrium in assisted reproductive therapy.

Appendix

In the two-compartment model displayed in Fig. 2, measurements are translated into quantities of contrast agent in voxels of tissue as follows:

$$\begin{gathered} {{\mu _{{\text{art}}} } \mathord{\left/ {\vphantom {{\mu _{{\text{art}}} } k}} \right. \kern-\nulldelimiterspace} k} = \frac{{Q_{{\text{art}}} }}{{V_{{\text{art}}} }} + n_{{\text{art}}} = C_{{\text{b,art}}} + n_{{\text{art}}} = q_{{\text{art}}} + n_{{\text{art}}} \hfill \\ {{\mu _{\text{T}} } \mathord{\left/ {\vphantom {{\mu _{\text{T}} } k}} \right. \kern-\nulldelimiterspace} k} = \frac{{Q_{\text{T}} }}{{V_{\text{T}} }} + n_{\text{T}} = \frac{{Q_{\text{c}} + Q_{\text{e}} }}{{V_{\text{T}} }} + n_{\text{T}} = q_{\text{c}} + q_{\text{e}} + n_{\text{T}} = q_{\text{T}} + n_{\text{T}} \hfill \\ \end{gathered}$$

where μ values are the measures of the contrast enhancement S(t) − S(0) in arterial and tissue regions, S is the MRI intensity at a given time t, and the indices indicate the arterial (art), tissue (T), capillary (c) and interstitial (e) zones; V values are the volumes of the regions and Q the contrast agent quantities in these regions, and q values are the equivalent quantities per volume of the corresponding total measured region. Noise is represented by n and the linear coefficient between contrast enhancement and concentration by k. C _b,art is the contrast agent concentration in the artery.

Intermediate plasma variables are defined in the blood compartment for a more readable pharmacokinetic model: the tissue plasma volume V _p = (1 − H _c) × V _b and the tissue plasma perfusion F _p = (1 − H _c) × F _b, where H _c is the haematocrit in small vessels. Then, the incoming plasma concentration can be determined, at each time t, by:

$$C_{{\text{p,art}}} \left( t \right) = {{q_{{\text{art}}} \left( {t - Dt} \right)} \mathord{\left/ {\vphantom {{q_{{\text{art}}} \left( {t - Dt} \right)} {\left( {1 - H_a } \right)}}} \right. \kern-\nulldelimiterspace} {\left( {1 - H_a } \right)}}$$

where Dt is the artery to tissue delay and H _a the haematocrit in large arteries. The outgoing plasma concentration is:

$$C_{{\text{p,vein}}} \left( t \right) = q_{\text{c}} \left( t \right)/v_{\text{p}}$$

where t (min) is the minimal arteriole to veinule delay and v _p is the plasma volume fraction defined by v _p = V _p/V _T. The pharmacokinetic two-compartmental model can then be written as follows:

$$\left\{ \begin{gathered} \frac{{dq_{\text{c}} }}{{dt}} = F_{\text{p}} \times \left( {C_{{\text{p,art}}} \left( t \right) - C_{{\text{p,vein}}} \left( t \right)} \right) - \frac{{PS}}{{v_{\text{p}} }} \times q_{\text{c}} + \frac{{PS}}{{v_{\text{e}} }} \times q_{\text{e}} \hfill \\ \frac{{dq_{\text{e}} }}{{dt}} = \frac{{PS}}{{v_{\text{p}} }} \cdot q_{\text{c}} - \frac{{PS}}{{v_{\text{e}} }} \cdot q_{\text{e}} \hfill \\ \end{gathered} \right.$$

where dq/dt is the time derivation of the quantity q of the agent in the capillary (c) and interstitial (e) compartments and PS is the permeability–surface product and where the interstitial volume fraction v _e is defined by v _e = V _e/V _T. When C and q are replaced by μ, the factor k, found in each term of the equation, is nullified.