Identification, validation and biological characterisation of novel glioblastoma tumour microenvironment subtypes: implications for precision immunotherapy.

BACKGROUND: New precision medicine therapies are urgently required for glioblastoma (GBM). However, to date, efforts to subtype patients based on molecular profiles have failed to direct treatment strategies. We hypothesised that interrogation of the GBM tumour microenvironment (TME) and identification of novel TME-specific subtypes could inform new precision immunotherapy treatment strategies. MATERIALS AND METHODS: A refined and validated microenvironment cell population (MCP) counter method was applied to >800 GBM patient tumours (GBM-MCP-counter). Specifically, partition around medoids (PAM) clustering of GBM-MCP-counter scores in the GLIOTRAIN discovery cohort identified three novel patient clusters, uniquely characterised by TME composition, functional orientation markers and immune checkpoint proteins. Validation was carried out in three independent GBM-RNA-seq datasets. Neoantigen, mutational and gene ontology analysis identified mutations and uniquely altered pathways across subtypes. The longitudinal Glioma Longitudinal AnalySiS (GLASS) cohort and three immunotherapy clinical trial cohorts [treatment with neoadjuvant/adjuvant anti-programmed cell death protein 1 (PD-1) or PSVRIPO] were further interrogated to assess subtype alterations between primary and recurrent tumours and to assess the utility of TME classifiers as immunotherapy biomarkers. RESULTS: TMEHigh tumours (30%) displayed elevated lymphocyte, myeloid cell immune checkpoint, programmed cell death protein 1 (PD-1) and cytotoxic T-lymphocyte-associated protein 4 transcripts. TMEHigh/mesenchymal+ patients featured tertiary lymphoid structures. TMEMed (46%) tumours were enriched for endothelial cell gene expression profiles and displayed heterogeneous immune populations. TMELow (24%) tumours were manifest as an 'immune-desert' group. TME subtype transitions upon recurrence were identified in the longitudinal GLASS cohort. Assessment of GBM immunotherapy trial datasets revealed that TMEHigh patients receiving neoadjuvant anti-PD-1 had significantly increased overall survival (P = 0.04). Moreover, TMEHigh patients treated with adjuvant anti-PD-1 or oncolytic virus (PVSRIPO) showed a trend towards improved survival. CONCLUSIONS: We have established a novel TME-based classification system for application in intracranial malignancies. TME subtypes represent canonical 'termini a quo' (starting points) to support an improved precision immunotherapy treatment approach.

Elasticity of arterioles and venules in postmortem human lungs.

The elasticity and branching order of noncapillary microscopic blood vessels less than 100 microns diam were studied in human lungs obtained 7-30 h postmortem, using a silicone elastomer method that selectively filled pulmonary arterioles or venules. The lungs were inflated to 10 cmH2O pressure and a gradient of transmural vascular pressure of 0-17 cm H2O, from lobe base to apex, was established in the silicone-filled vascular system. Histological materials were obtained after airway fixation by formaldehyde solution and analyzed for vessel diameter in the branching order of 1, 2, and 3, with the smallest noncapillary vessel designated as order 1, in accord with the Strahler system. The change in vessel diameter within a branching order at different levels of transmural pressure is a derived measure of vascular elasticity expressed as compliance coefficient alpha, alpha Values are 0.128, 0.164, and 0.210 micron/cmH2O or 0.682, 0.472, and 0.354%/cmH2O, respectively, of orders 1-3 for arterioles and 0.187, 0.215, and 0.250 micron/cmH2O or 0.992, 0.612, and 0.424%/cmH2O, respectively, of orders 1-3 for venules. The percent is normalized with D0, which is the value of diameter (D) when the transmural pressure is zero. These data are compared with those for the cat where alpha = 0.274 for similar juxta-alveolar vessels.

A new theory of pulmonary blood flow in zone 2 condition.

In pulmonary blood flow, if arterial pressure (Pa) and pleural pressure (Ppl) were fixed, the flow increases with decreasing venous pressure (Pv) only when venule pressure (Pven) greater than airway pressure (PA) (i.e., in zone 3). When Pven less than or equal to PA (i.e., in zone 2), with Pa fixed, the flow decreases with decreasing Pv. The pressure-flow relationship has a hysteresis loop. This phenomenon can be explained by conservation of mass and momentum and the morphology and material properties of the lung, including the observation that reseparation of adhered cells requires an extra force. The key mathematical observation is that the solution h = 0 (h being the blood sheet thickness in the interalveolar septa) can coexist with the solution h not equal to 0 in zone 2 condition, resulting in "patchy" filling of the alveolar walls. When h = 0, the sheet is collapsed and endothelial cells adhere. Experimental results show that the adhered endothelial cells do not reseparate by raising Pv in zone 2 but can be accomplished under zone 3 condition.

Zero-stress states of human pulmonary arteries and veins.

The zero-stress states of the pulmonary arteries and veins from order 3 to order 9 were determined in six normal human lungs within 15 h postmortem. The zero-stress state of each vessel was obtained by cutting the vessel transversely into a series of short rings, then cutting each ring radially, which caused the ring to spring open into a sector. Each sector was characterized by its opening angle. The mean opening angle varied between 92 and 163 degrees in the arterial tree and between 89 and 128 degrees in the venous tree. There was a tendency for opening angles to increase as the sizes of the arteries and veins increased. We computed the residual strains based on the experimental measurements and estimated the residual stresses according to Hooke's law. We found that the inner wall of a vessel at the state in which the internal pressure, external pressure, and longitudinal stress are all zero was under compression and the outer wall was in tension, and that the magnitude of compressive stress was greater than the magnitude of tensile stress.

Vascular impedance analysis in dog lung with detailed morphometric and elasticity data.

On the basis of experimentally measured morphometric and elasticity data and model-derived mean pressure-flow conditions, we attempt a theoretical modeling of pulsatile flow in the whole lung. In the model we use the "elastic tube" for arteries and veins, and the vascular impedance in arteries and veins follows Womersley's theory and electric analogue. We employ the "sheet-flow" theory to describe the flow in the capillaries and to obtain the microvascular impedance matrix. The characteristic impedance of each order along the vascular tree, the input impedance at the capillary entrance and exit, and the pulmonary arterial input impedance at the main pulmonary artery are computed under certain physiological conditions. Using the pulsatile flow model, we investigate the effects of arterial vascular obstruction on pulmonary vascular impedance. The model-derived data are compared with the available experimental results in the literature.

Morphometry of the human pulmonary vasculature.

The morphometric data on the branching pattern and vascular geometry of the human pulmonary arterial and venous trees are presented. Arterial and venous casts were prepared by the silicone elastomer casting method. Three recent innovations are used to describe the vascular geometry: the diameter-defined Strahler ordering model is used to assign branching orders, the connectivity matrix is used to describe the connection of blood vessels from one order to another, and a distinction between vessel segments and vessel elements is used to express the series-parallel feature of the pulmonary vessels. A total of 15 orders of arteries were found between the main pulmonary artery and the capillaries in the left lung and a total of 15 orders of veins between the capillaries and the left atrium in the right lung. The elemental and segmental data are presented. The morphometric data are then used to compute the total cross-sectional areas, blood volumes, and fractal dimensions in the pulmonary arterial and venous trees.

Morphometry of the dog pulmonary venous tree.

The biophysical approach to the study of blood flow in the pulmonary vasculature requires a detailed description of vascular geometry and branching pattern. The description of the pulmonary venous morphometry in the dog is the focus of this paper. Silicone elastomer casts of a dog lung were made and were used to measure the diameters, lengths, and branching pattern of the pulmonary venous vasculature. The anatomic data are presented statistically with a diameter-defined Strahler ordering scheme, a rule for assigning the order numbers of the vessels on the basis of a diameter criterion. The asymmetric branching pattern of the pulmonary venous vasculature is described with a connectivity matrix. Results show that for the dog's right pulmonary venous tree 1) a total of 11 orders of vessels lay between the left atrium and the capillary bed; 2) the average ratios of the diameter, length, and number of branches of successive orders of veins were 1.701, 1.556, and 3.762, respectively; and 3) a fractal description of the tree geometry resulted in diameter and length fractal dimensions of 2.49 and 2.99, respectively. The morphometric data were used to compute the cross-sectional area, vascular volume, and Poiseuillean resistance in the venous vessels.

Speed of stress wave propagation in lung.

The speed of stress waves in the lung parenchyma was investigated to understand why, among all internal organs, the lung is the most easily injured when an animal is subjected to an impact loading. The speed of the sound is much less in the lung than that in other organs. To analyze the dynamic response of the lung to impact loading, it is necessary to know the speed of internal wave propagation. Excised lungs of the rabbit and the goat were impacted with water jet at dynamic pressure in the range of 7-35 kPa (1-5 psi) and surface velocity of 1-15 m/s. The stress wave was measured by pressure transducer. The distance between the point of impact and the sensor at another point on the far side of the lung and the transit time of the stress wave were measured. The wave speed in the goat lung was found to vary from 31.4 to 64.7 m/s when the transpulmonary pressure Pa-Ppl was varied from 0 to 20 cmH2O where Pa represents airway pressure and Ppl represents pleural pressure. In rabbit lung the wave speed varied from 16.5 to 36.9 m/s when Pa-Ppl was varied from 0 to 16 cmH2O. Using measured values of the bulk modulus, shear modulus, and density of the parenchyma, reasonable agreement between theoretical and experimental wave speeds were obtained.

Effect of chronic thromboembolism on the pulmonary artery pressure-flow relationship in dogs.

To understand the hemodynamic alterations associated with chronic thromboembolic pulmonary hypertension, the large pulmonary arteries of mongrel dogs were chronically obstructed with lysis-resistant thrombi. Pulmonary hemodynamics were experimentally measured and described by multipoint pulmonary arterial pressure (PAP) vs. flow plots. In nine anesthetized chronically embolized dogs, but not in six control dogs, the PAP-flow line shifted significantly upward in a parallel fashion by 4.2 +/- 0.7 mmHg. The postembolic pulmonary circulation was further characterized by predictions from a morphometric-based elastic tube and sheet flow model of the canine pulmonary circulation. After model validation with the preembolic PAP-flow data, the derived postembolic PAP matched the in vivo results to within 1 mmHg. A detailed analysis of the model-derived PAP drop revealed that the PAP-flow line shift can be accounted for by a novel fixed resistor in the largest obstructed pulmonary artery.

Trauma of lung due to impact load.

A quantitative evaluation of lung injury due to impact loading is of general interest. Hemorrhage and edema are the usual sequelae to traumatic pulmonary impact. To gain some quantitative understanding of the phenomena, we perfused excised rabbit lung with Macrodex at isogravimetric condition and monitored lung weight continuously after impact. It is shown that a factor of importance is the rigidity of the surface on which the lung rests. The rate of lung weight increase is smaller if the lung was 'freely' supported on a soft cloth, more if it was supported on a rigid plate. This suggests the influence of stress wave reflection. The critical condition correlates with the initial velocity of impact at the surface of the lung, or with the maximum deflection. For a freely supported lung, the rate of lung weight increase was 22% of the initial total lung weight per h after impact when the impact velocity was 11.5 ms-1, 30% when the velocity was 13.2 ms-1, several 100% at 13.5 ms-1, signaling massive lung injury. Since the velocity of sound in rabbit lung is 33.3 ms-1 when the inflation (transpulmonary) pressure is 10 cm H2O, the critical velocity of 13.5 ms-1 corresponds to a Mach number of 0.4. The maximum surface displacement of the lung is almost linearly proportional to the initial velocity of impact. The exact cause of edema and hemorrhage is unknown; we hypothesize that it is due to tensile stress in the alveolar wall caused by the impact.

Stochastic simulation of human pulmonary blood flow and transit time frequency distribution based on anatomic and elasticity data.

The objective of our study was to develop a computing program for computing the transit time frequency distributions of red blood cell in human pulmonary circulation, based on our anatomic and elasticity data of blood vessels in human lung. A stochastic simulation model was introduced to simulate blood flow in human pulmonary circulation. In the stochastic simulation model, the connectivity data of pulmonary blood vessels in human lung was converted into a probability matrix. Based on this model, the transit time of red blood cell in human pulmonary circulation and the output blood pressure were studied. Additionally, the stochastic simulation model can be used to predict the changes of blood flow in human pulmonary circulation with the advantage of the lower computing cost and the higher flexibility. In conclusion, a stochastic simulation approach was introduced to simulate the blood flow in the hierarchical structure of a pulmonary circulation system, and to calculate the transit time distributions and the blood pressure outputs.

A continuum model for pressure-flow relationship in human pulmonary circulation.

A continuum model was introduced to analyze the pressure-flow relationship for steady flow in human pulmonary circulation. The continuum approach was based on the principles of continuum mechanics in conjunction with detailed measurement of vascular geometry, vascular elasticity and blood rheology. The pulmonary arteries and veins were considered as elastic tubes and the "fifth-power law" was used to describe the pressure-flow relationship. For pulmonary capillaries, the "sheet-flow" theory was employed and the pressure-flow relationship was represented by the "fourth-power law". In this paper, the pressure-flow relationship for the whole pulmonary circulation and the longitudinal pressure distribution along the streamlines were studied. Our computed data showed general agreement with the experimental data for the normal subjects and the patients with mitral stenosis and chronic bronchitis in the literature. In conclusion, our continuum model can be used to predict the changes of steady flow in human pulmonary circulation.

Mechanical properties of human lung parenchyma.

In order to have a detailed analysis of the distribution of stresses in the lung, one needs to understand the mechanical behavior of the lung material. For the stress-strain relationship of human lung, the present state of the art is that the form of the constitutive equations is known, but associated material constants are unknown. In this study, biaxial loading experiments were done on specimens of excised cadaver lung parenchyma without the effects of large blood vessels, bronchi, and pleura. Curves of strain vs. stress were recorded. A non-linear form of strain energy function was used to examine the stress-strain relationship. This relationship fits the experimental data well. The analysis based on data from 11 specimens of excised human lung parenchyma yielded that the physical constants are C/delta = 3.06 +/- 0.84 K x dyn/cm2, alpha = 4.47 +/- 1.94, and beta = -4.20 +/- 2.55.

Effect of velocity of distribution on red cell distribution in capillary blood vessels.

Through the use of simulated model experiments, data on blood cell distribution into a bifurcating capillary blood vessel are obtained. The results show that the movement of red blood cells at a bifurcation point is influenced by the difference in velocities of flow in the daughter branches. If the velocity of flow in one branch is slower than that in the other, the hematocrit decreases in the slower branch and increases in the faster branch. For velocity ratios sufficiently smaller than a certain critical value, the hematocrit ratio can be expressed by a linear relationship, (H1/H2) - 1 = a[v1/v2) - 1], in which v1, v2 and H1 H2 denote the particle velocities and tube hematocritis in the branches 1 and 2, respectively, and a is a dimensionless contant dependent upon a number of factors, the most important of which are the ratio of cell diameter to tube diameter, the shape and rigidity of the pellets, and the hematocrit in the feeding tube. For velocity ratios beyond a critical value, nearly all the cells flow into the faster branch. The smaller the feeding-tube hematocrit is, the smaller is the critical velocity ratio at which this phenomenon occurs.

Inversion of Fahraeus effect and effect of mainstream flow on capillary hematocrit.

The well known Fahraeus effect (1929) states that if whole blood is allowed to flow from a large reservoir into a small circular cylindrical tube, the hematocrit in the tube is smaller than that in the reservoir, and the smaller the tube, the smaller will be the tube hematocrit. This is interpreted as a feature of particulate flow. We find that this relationship is not monotonic in a model experiment in which gelatin particles (circular disks) are suspended in a silicone fluid to simulate blood. When the diameter of the underformed cell is equal to or greater than the tube diameter, the volume fraction of the cells in the tubes increases to a value equal to or greater than that in the reservoir. Thus the Fahraeus effect has a point of inversion. Additional experiments show that the hematocrit in the tube could be greatly influenced by the flow condition outside the entrance of the tube. If the tube is perpendicular to the main direction of flow in the reservoir (as is the case of Barbee and Cokelet's experiment, or in most arteriole-capillary junctions), the velocity gradient and the velocity of flow in the reservoir just outside the entrance to the tube affects the hematocrit in the tube.

Elasticity of small pulmonary veins in the cat.

The distensibility of pulmonary veins of cats, in the diametric range of 100-1200 micrometer, was studied as a function of the venous pressure p upsilon and pleural pressure p PL, while the alveolar air pressure was maintained at zero (atmospheric). The resulting percentage changes in diameter normalized with respect to the diameter at delta P of 10 cm H20 (D10) are expressed as function of delta = p upsilon - p PL. It was found that in most cases the diameter varies linearly with delta P: D/D10 = 1 + alpha (p upsilon - p PL) where alpha is the compliance coefficient. The results show that smaller veins of the cat are more compliant than larger veins. For example, when pleural pressure is -10 cm H2O, the values of alpha for vessels in the ranges of diameters of 100-200 micrometer, 200-400 micrometer, 400-800 micrometer and 800-1200 micrometer are, respectively, 2.05, 1.44, 1.08 and 0.71 percent per cm H2O or Pa-1. The effects of lung inflation on the distensibility of pulmonary veins are also studied. Our results show that for vessel sizes in the range of 400-800 micrometer and 800-1200 micrometer the compliance constant alpha is not affected by the inflation of the lungs (changes in pleural pressure to more negative values). For smaller veins in the size ranges 100-200 micrometer and 200-400 micrometer our results show an increase in compliance from 2.05 to 2.79 and from 1.44 to 2.01 percent per cm H20 or Pa-1, respectively, when pleural pressure is changed from -10 to -15 cm H20. When the pleural pressure is more negative than -15 cm H20, however, the compliance of the vessels in the foregoing two size ranges is observed to decrease.

Analysis of blood flow in cat's lung with detailed anatomical and elasticity data.

Recently, a complete set of data on the branching pattern of the cat's pulmonary arterial and venous trees and the elasticity of these blood vessels was obtained in our laboratory. Hence it becomes possible for the first time to perform a theoretical analysis of the blood flow in the lung of an animal based on a set of actual data on anatomy and elasticity. This paper presents an analysis of steady flow of blood in cat's lung. The effect of the vessel elasticity is embodied in the "fifth-power law" and the "sheet-flow" theory. The theory yields the pressure-flow relationship of the whole lung, the longitudinal pressure distribution, and the transit time of blood in the capillaries. These results are compared with available experimental data in the literature.

Comparison of theory and experiment in pulsatile flow in cat lung.

A mathematical model of pulsatile flow in cat lung based on existing morphometric and elastic data is presented and validated by experimental results. In the model, the pulmonary arteries and veins were treated as elastic tubes, whereas the pulmonary capillaries were treated as two-dimensional sheets. The macro- and microcirculatory vasculature was transformed into an analog electrical circuit. Input impedances of the pulmonary blood vessels of every order were calculated under normal physiological conditions. Pressure-flow relation of the whole lung was predicted theoretically. Experiments on isolated perfused cat lungs were carried out. The relation between pulsatile blood pressure and blood flow was measured. Comparison of the theoretically predicted input impedance spectra with those of the experimental results showed that the modulus spectra were well predicted, but significant differences existed in the phase angle spectra between the theoretical predictions and the experimental results. This latter discrepancy cannot be explained at present and needs to be further investigated.

Microcirculation impedance analysis in cat lung.

An analysis of pulsatile microcirculation in cat lung, with special attention to the pulmonary microvascular impedance, is presented. A theoretical calculation is made on the basis of a complete set of experimental data on the morphology and elasticity of cat's pulmonary capillary sheets. The transfer matrix of the pulmonary microvascular impedance is obtained. The input impedance at the capillary entrance and exit are determined. The input impedance at the pulmonary arterial trunk is compared under various physiological conditions. It is shown that although the impact of pulmonary microcirculation on the relationship between the steady mean flow and pressure in the pulmonary arteries and veins is decisively large, the influence of the alveolar microcirculation on the input impedance at the pulmonary arterial trunk is small.