Imaging intratumoral convection: pressure-dependent enhancement in chemotherapeutic delivery to solid tumors

Abstract

Purpose: Low-molecular weight (LMW) chemotherapeutics are believed to reach tumors through diffusion across capillary beds as well as membrane transporters. Unexpectedly, the delivery of these agents seems to be augmented by reductions in tumor interstitial fluid pressure, an effect typically associated with high-molecular weight molecules that reach tumors principally through convection. We investigated the hypothesis that improved intratumoral convection can alter tumor metabolism and enhance the delivery of a LMW chemotherapeutic agent to solid tumors.

Experimental design: For this purpose, we applied 31P/19F magnetic resonance spectroscopy (MRS) and magnetic resonance spectroscopic imaging (MRSI) to examine the influence of type I collagenase on tumor bioenergetics and the delivery of 5-fluorouracil (5FU) to HT29 human colorectal tumors grown s.c. in mice.

Results: Collagenase effected a 34% reduction in tumor interstitial fluid pressure with an attendant disintegration of intratumoral collagen. Neither mice-administered collagenase nor controls receiving PBS showed changes in (31)phosphorus MRS-measured tumor bioenergetics; however, a time-dependent increase in the content of extracellular inorganic phosphate (Pi(e)) was observed in tumors of collagenase-treated animals. (31)Phosphorus MRSI showed that this increase underscored a more homogeneous distribution of Pi(e) in tumors of experimental mice. (19)Fluorine MRS showed that these changes were associated with a 50% increase in 5FU uptake in tumors of experimental versus control animals; however, this increase resulted in an increase in 5FU catabolites rather than fluoronucleotide intermediates that are required for subsequent cytotoxicity.

Conclusions: These data indicate that the modulation of convective flow within tumors can improve the delivery of (LMW) chemotherapeutics and show the potential role for noninvasive imaging of this process in vivo.

Figure 1

Type I collagenase effects a marked reduction in TIFP associated with a corresponding reduction in intratumoral collagen. A: TIFP relative to the time of PBS (□) or 0.1% collagenase (▲) injection. TIFP values are reported @ 10 minute intervals and are normalized to the pressure assessed 10 minutes prior to injection. A transient increase in TIFP was observed immediately following injection presumably associated with the increase in vascular volume. B-C: Composition of representative tumor sections from PBS (B) or collagenase (C) treated mice stained for type I collagen.

Figure 2

Collagenase does not significantly alter tumor bioenergetics but affects a significant increase in extracellular Pi. ³¹PMR spectroscopic evolution of phosphorus metabolites in the tumor- bearing legs of mice administered PBS (A) or 0.1% collagenase (B). Serial ³¹PMR spectra from representative animals are shown. The start of each spectral acquisition relative to the time of injection is indicated. The inset demonstrates resolved metabolite peaks including (1) phosphoethanolamine, (2) phosphocholine, (3) Pi_major, (4) Pi_minor, (5) phosphocreatine, (6) γ-NTP, (7) α-NTP, (8) nicotinamide adenine dinucleotide, (9) diphosphodiesters, (10) β-NTP.

Figure 3

³¹PMRS-measured bioenergetic parameters in the tumor bearing legs of mice relative to the time of PBS (□) or 0.1% collagenase (▲) injection. Data are reported @ 10 minute intervals and are normalized to measurements made 10 minutes prior to injection. (A): Time course of the mean NTP/Pi_major ratio. (B): Time course of the mean Pi_minor levels quantitated as area under the curve (AUC).

Figure 4

Collagenase improves the spatial homogeneity of extracellular Pi. 2D- ³¹PMR spectroscopic images and related parametric maps demonstrating the regional distribution of phosphorus metabolites in tumor bearing legs of mice administered PBS (A) or 0.1% collagenase (B). Representative images acquired with an in-plane resolution of 2.0×2.0 mm and registered to morphologic reference images are shown along with corresponding parametric maps of local Pi_minor levels. The contours of the tumor are outlined in each spectroscopic image.

Figure 5

Collagenase increases intratumoral concentrations of 5FU. ¹⁹FMRS-determined 5FU pharmacokinetics in tumors of experimental or control mice. Spectra were acquired with a 10 minute temporal resolution. The fitted data were normalized to an external reference standard and reported as the normalized signal amplitude (AUC). (A): Serial measurements of mean 5FU levels demonstrate its accumulation to significantly higher levels in tumors of collagenase-treated mice (▲) as compared to PBS-treated controls (□) (**, P=0.0185). Linear regressions of these time course data are shown for both groups (y=4.92-0.086x, r²=0.993 and y=3.99-0.081x, r²=0.991 respectively). (B): Serial measurements of mean FNuc levels in tumors of collagenase- (▲) and PBS-treated (□) mice. (C): Serial measurements of mean FUPA levels in tumors of collagenase- (▲) and PBS-treated (□) mice.

Figure 6

Collagenase does not influence the growth delay of tumors treated with 5FU relative to 5FU-treated controls. Mean tumor growth curves for subcutaneous HT29 tumors growing in the legs of control (untreated, ▲) mice or mice receiving collagenase only (□), PBS + 5FU (▲) or collagenase + 5FU (□). The data are presented as transformed Gompertz tumor volumes plotted against time. The best fit linear regression lines with slope -α for these data are indicated where α is the specific growth rate (y=1.22-0.07x, r²=0.974, y=1.30-0.07x, r²=0.982, y=1.27-0.03x, r²=0.952 and y=1.25-0.03x, r²=0.975, respectively).