The major focus of this study was to quantitatively interpret the uptake of fluorescent nanoparticles in three different tumor lines and to determine which correlates are most important for driving the permeability of the tumor. Each of the relative fluorescence uptake values reported in were normalized by normal tissue values from the same mouse. Ratiometricdata interpretation allows comparison of the signals between animals. The data in shows thatall measurements had a median value higher than unity, indicating thatfluorescence nanoparticles always entermore into the tumor relative to normal tissue, as would be expected for permeability-based delivery in the tumor.
The rate of uptake of nanoparticles, indicated by the slope of the line in , shows increasing fluorescence over time for the A431 tumorand relatively flat uptake for 9L and with a decreasing slope for AsPC-1 tumors. These tumors were again chosen for their disparate vascular and collagen content values, and the difference in uptake kinetics indicates the existence of a dissimilar drug delivery dynamics in these tumor lines. The differences in uptake observed would be driven by the vascular permeability, but uptake over several minutes would also be driven by the interstitial pressure, which would in turn be affected by the vascular density and collagen density. So these major driving factors were assessed for their correlation with the uptake value, using both individual tumor data as well as average tumor line data.
A key part of this study is the illustration that relationships between the driving forces for drug delivery are not well interpreted with a single tumor line, but rather are better elucidated by comparing data between tumor lines. These results are seen in , showing that most of the parameters are at least partially correlated to each other. The observed drug dynamics for subcutaneous tumor lines A431 and AsPC-1 are the most different of the three, and the effects seem to be well interpreted from the pathophysiological parameters measured here. The 9L tumor line is moderate in comparison to the extremes of AsPC-1 and A431 for vascular density and collagen content. Higher vessel density in A431 tumors relative to AsPC-1 is correlated to a more pronounced drug accumulation, and the lower fraction of collagen content of A431 tumors likely pose a significantly weaker barrier to the penetration of the nanoparticles into the tumor and lead to the lower interstitial pressure values observed. Just based upon these two tumor lines, one might guess that the impact of vascular area dominates drug accumulation; however, further inspection shows that the 9L tumor line, despite the comparable vessel density, has significantly smaller drug accumulation than the A431 tumor. This observation, together with the comparable collagen densities of the 9L and AsPC-1, suggests that the elevated collagen expression in the 9L tumor acts as either a barrier to nanoparticle delivery or as a booster of IFP, which itself is known to be a prominent barrier to nanoparticle delivery.
Figure 6 Dependence of the fluorescence is plotted with respect to the analyzed pathophysiological parameters (a) vascular density, (b) interstitial pressure, and (c) collagen density. The data from each tumor type grouped as a single point with the standard error (more ...)
IFP is thought to limit vascular permeability because the pressure difference from within the vessel to outside the vessel drives the convection process across the endothelium. In the past few decades there have been several ways to attempt interstitial pressure measurement of tumors, including: 1) “wick-in-needle” (WIN) measurement, 2) micropipettes, 3) micropore chambers for average IFP[14
], and 4) fiber optic pressure sensing devices. Each of these methods has its advantages and limitations [40
], and all are invasive with arguably high variability due to systematic bias from placement errors. Thus, accurately measuring IFPis controversial. The absence of significant difference between the recorded IFP values of A431 and 9L could be attributed to the general controversy that surrounds measuring IFP values, regardless of methods used [41
]. All the methods currently used in measuring IFP, involve invasive punctures that can potentially alter the tumor[41
]. Despite these measurement limitations, the finding that tumors with highest IFP had lowest drug penetrance is convincing
One of the most confounding factors in this type of analysis is the fact that the parameters are all likely interrelated. A linear correlation test aimed at showing the dependence of the drug uptake on any one of the given parameters, regardless of tumor type, failed to verify the existence of any significant correlation, although statistically significant values are observed when the tumor data is interpreted as average numbers, as presented here in . This result is attributed to inherently complex simultaneous dependence of drug delivery in each tumor on multiple pathophysiological parameters. However, once the tumors were binned according to tumor type, as shown in , a very strong linear correlation with R2=0.98 between the observed means of the average temporal fluorescence ratio and percent collagen content was seen, . Additionally, the correlation was strong, R2 =0.71, for means of the average temporal fluorescence ratio and IFP, . The correlation was not as strong for means of the average temporal fluorescence ratio and percent vascular density, . These values are summarized in .
Table 2 Tabulated R2-values from correlation test, summarizing the strength of the dependence of the drug uptake on the pathophysiological parameters, and the functional dependence of theparameterson each other, when the data is binned with respect to tumor type. (more ...)
As discussed in the introduction section and illustrated in , the complex interplay in tumor growth is inter-related, but one general model for this is that neovascular growth, epidermal growth and stromal growth all together contribute to a number of the features observed. As tumors grow interstitial pressure increases, and this relationship would be consistent with the significant trends of . This pressure then ultimately limits drug penetration, consistent with the data in .
In this paperbulk uptake of nanoparticles relative to the pathophysiological parameters was assessed through correlation analysis, and while individual tumor lines did not exhibit significant trends, summary average values between tumor lines show expected model trends. Overall, a high presence of collagen appears to pose the most prominent barrier to drug delivery, although it is not clear if it is the collagen itself or the collagen controlled interstitial pressure that imposes the real barrier, or even if there are other inter-related factors related to this. Additionally; high interstitial fluid pressure strongly correlated with the decrease in average drug uptake in tumor lines. It was further shown that barriers to drug delivery can prevent the penetration of nanoparticles even to the tumor line, 9L, with relatively high vasculature. Generally, the data is consistent with a model in which the permeability into the tumor is inhibited by higher interstitial pressure, which in turn shuts down the vasculature and is controlled by collagen growth. These effects are correlative, but appear to be interrelated and further elucidation would require larger numbers of tumor models for better correlation analysis, or through mechanistic interventions to test the role of the individual parameters in drug penetration.