Small Pulmonary Artery and Vein Volumes Independently Predict Oxygen Desaturation in Smokers.

Background: To personalize the care for persons with smoking-related lung disease, a thorough understanding of its etiology is essential. The role of pulmonary vessels remains poorly understood. Living at high altitude provides a natural model to investigate the effects of low oxygen levels on pulmonary vessels. This study aims to evaluate the relationship between living at high altitudes and small pulmonary vein and artery volumes. We hypothesize that small vein and artery volumes were independently associated with living at high altitude. Methods: We quantified small pulmonary vein and artery dimensions (ᴓ < 1 mm) on computed tomography (CT) down to 0.2 mm in diameter and normalized the dimensions by body surface area. In 8931 current and former smokers participating in the COPDGene study, we used multivariate regression models corrected for clinical and technical confounders. Results: 1262 residents (14.1%) were defined as high-altitude residents (~1600 m, Denver, CO, USA). Compared to lower-altitude residents, the high-altitude residents had a higher age (62.0 ± 9.1 vs. 59.6 ± 9.0 years), more pack-years smoked (46.8 vs. 44.1) and a lower FEV1% predicted (64.6 ± 32.4% vs. 76.8 ± 25.2%). Both mean small artery volume (4.09 ± 0.89 mL/m2 vs. 3.85 ± 0.90 mL/m2) and mean small vein volume (2.96 ± 0.53 mL/m2 vs. 2.67 ± 0.53 mL/m2) were higher in high-altitude residents. Multivariate linear regression showed that, in those without COPD, high-altitude residents have a higher small vein volume (0.129 mL/m2, p < 0.001) and higher small artery volume (0.170 mL/m2, p = 0.001) compared to lower-altitude residents. There was no significant association in residents with COPD. Conclusions: In current and former smokers without COPD, higher small pulmonary vein and artery volumes were associated with living at high altitude, independent of lung disease or technical CT parameters. A potential cause includes vascular remodeling due to an elevated need for blood oxygen transport, which becomes concealed when COPD develops.

Higher small pulmonary artery and vein volume on computed tomography is associated with mortality in current and former smokers

1. This study was conducted to determine adrenomedullin (AM) action sites in the pulmonary vascular bed and the relation between its vasodilator effects and vascular tone. Moreover, an examination was made into whether calcitonin gene-related peptide (CGRP) receptors mediate pulmonary vasodilatations induced by AM. To this end, we directly measured internal diameter (i.d.) changes in small pulmonary arteries and veins (100-1100 microns i.d.) by use of an X-ray television system on the in vivo cat lung. 2. Under control (resting vascular tone) conditions, AM injections into the left main pulmonary artery caused dose-related i.d. increases in both small arteries and veins. The mean i.d. increase of the 100-1100 microns arteries (4 +/- 1, 11 +/- 2, and 17 +/- 2% with 0.01, 0.1, and 1 nmol kg-1 AM, respectively) was significantly larger than that for the veins (1 +/- 1, 5 +/- 2, and 7 +/- 2% with 0.01, 0.1 and 1 nmol kg-1 AM, respectively) whatever the injected dose of AM. 3. When unilobar hypoxia (5% O2) had decreased the i.d. of the 100-1100 microns arteries and veins by 16 +/- 3 and 6 +/- 3%, respectively, AM (0.1 nmol kg-1) was able to induce significantly larger i.d. increases in the arteries (28 +/- 3%) and veins (11 +/- 3%) than those under control conditions. 4. The AM-induced i.d. response pattern in the serially connected pulmonary arteries was quite different from that induced by CGRP; AM caused a greater increase in smaller vessels (100-500 microns) than in larger vessels (500-1100 microns). In the case of CGRP, a greater increase was observed in the larger vessels. 5. CGRP8-37 (100 nmol kg-1, i.v., followed by a continuous infusion of 0.2 nmol kg-1 min-1) had no significant effect on the i.d. increase induced by AM (0.1 nmol kg-1) in any serial segments of the arteries and veins. 6. The results indicate that, in the cat, AM induces greater vasodilatation in small pulmonary arteries and lesser vasodilatation in small veins, the maximum dilatation being in the more peripheral arterial segment (100-500 microns). The vasodilator effect of AM was enhanced when vascular tone was elevated. The data suggest that the AM-induced pulmonary vasodilatation is not mediated by CGRP receptors but by its own specific receptor.

The contractile response of ring segments of large, medium, and small pulmonary arteries and veins of the dog to histamine, norepinephrine, and serotonin have been studied. The maximum contractile response to these drugs was normalized with respect to the maximal response obtained in stimulation with 127 mM K+. The small pulmonary artery was more reactive to histamine, norepinephrine, and serotonin when compared with large and medium pulmonary arteries. The medium and large pulmonary artery showed no difference in reactivity to histamine. However, the mean effective dose (ED50) values for these agonists among the different segments of pulmonary arteries showed no significant difference. The small and medium pulmonary veins demonstrated increased reactivity to histamine, but not norepinephrine and serotonin. The ED50 values also indicated that both small and medium veins were more sensitive to histamine when compared with the large pulmonary vein. The log concentration percent response curves for both small and medium pulmonary veins were displaced leftward (increased sensitivity) with respect to that for the large pulmonary vein. However, the reactivity and sensitivity to histamine between medium and small pulmonary veins were no different. The reactivity and sensitivity of different segments of pulmonary veins to norepinephrine and serotonin showed no significant differences among them. We conclude that histamine and other vasoactive substances, which are directly or indirectly related to mast cell degranulation, exert pharmacological effects on the pulmonary vasculature which possesses differential responsiveness at various levels of the vascular tree.

Shareable abstractCigarette smoking is related to increased small pulmonary vein and artery volumes. These findings highlight the potential of CT imaging to assess smoking-related vascular changes in diseases like COPD and pulmonary hypertension.https://bit.ly/4ljnU6q

Objective: To establish a standardized method for isolated pulmonary artery and vein rings with different diameter, pressure and length, which could provide a more scientific method for in vitro study of pulmonary vessel diseases. Methods: Male SD rats were anesthetized, and the right ventricular systolic pressure were measured. Small pulmonary artery and vein rings with 200-400 μm in diameter and 2 mm in length were prepared by dissecting pulmonary arteries and veins. The pulmonary vessel rings were mounted in the organ bath by 2 stainless steel wires with diameter of 40 μm. Then the internal circumference of the vessel rings was increased gradually with 100 μm per step. At the same time the vascular tension was recorded by the Myograph System and Acknowledgement data acquisition system, and subsequently the passive length-tension exponential curve was made. The initial tension of the rings was set, equilibrated for another 30 min, and then stimulated with 60 mmol/L KCl 3 times, and the best contractile reactivity was achieved. The contractile reactivity of pulmonary artery rings and endothelial integrity were detected by exposure to 1 μmol/L phenylephrine(PE) and 10 μmol/L acetylcholine(Ach), while the contractile reactivity of pulmonary vein rings was detected by exposure to 1 μmol/L U46619 and 10 μmol/L papaverine. Results: The contraction and relaxation effects of the pulmonary artery rings reached 0.39 mg and 92% when they were stimulated by 1 μmol/L PE and 10 μmol/L Ach. The contraction and relaxation effects of pulmonary vein rings were up to 0.13 mg and 84% when they were exposed to 1 μmol/L U46619 and 10 μmol/L papaverine, respectively. Conclusion: Pulmonary artery and vein rings with appropriate basal tension and optimal vasodilator activity were prepared, and a standardized method of tension experiment for isolated pulmonary artery and vein rings established.

A few reports have provided histopathological insight into pulmonary hypertension developing after antitumor chemotherapy. In general, plexogenic pulmonary arteriopathy is a commonly observed finding in patients with severe pulmonary hypertension. We herein report a novel pathological finding that may characterize the histopathological change occurring in patients with pulmonary hypertension after chemotherapy for malignancy. Lung biopsy or autopsy was performed in 7 patients with pulmonary hypertension that developed during or after chemotherapy between 2006 and 2013 to examine the pulmonary vascular changes or to determine the cause of death. Pathological findings included in situ thrombosis in the small pulmonary arteries in 4 of 7 patients. In 2 of 4 patients, pulmonary hypertension was controlled by anticoagulants and antithrombotic agents. One patient who had organized thrombi attained spontaneous remission with oxygen therapy. The other patient died of sudden cardiopulmonary arrest during chemotherapy. Autopsy showed complete occlusion of the peripheral small pulmonary arteries and veins by thrombi. These results demonstrate that in situ thrombosis in the small pulmonary arteries could cause pulmonary hypertension after chemotherapy.

Background: In COPD, the risk of adverse outcomes associated with chest computed tomography (CT)-based metrics of pulmonary vascular disease including loss of small pulmonary arterial vascular volume or pruning, increased right ventricular volume and pulmonary artery dilation, are well described. The relevance of these imaging features in smokers without COPD is not known. We investigated the association of CT features along a pulmonary vascular continuum with death in smokers without COPD. Methods: We analyzed subjects from the Genetic Epidemiology of COPD (COPDGene) study Phase 1 visit with available vascular volumes, PA to aorta (PA/Ao) diameter and right /left ventricle (RV/LV) volume measurements with FEV1/FVC ≥ 0.7 and FEV1 ≥ 80% (n=2,530). Cox proportional hazard models were used to compare the standardized predictive ability of each vascular feature with all-cause mortality; the proportional hazards assumption was checked to ensure against violation. Results: The prevalence of current and former smokers was similar (50.6% and 49.4%) with median age 57 [12.3] years and 34.4 [23.5] pack-year exposure; subjects had minimal emphysema (%LAA-950) (1.1[2.6]%). At 15-year follow up, overall survival in the quartile with the least pruning to that with the most pruning, was 92.1% vs. 96.2% (p&amp;lt;0.0001). In a multivariable model adjusted for age, sex, race, body mass index, smoking status (current or former), smoking pack-years, FEV1, 6-minute walk distance, supplemental oxygen use, modified Medical Research Council (mMRC) score ≥ 2, %LAA-950, congestive heart failure and heart attack, each standard deviation decrease in small vessel volume (&amp;lt;5mm2) (more pruning) was associated with a 29% increased risk of death (HR 1.29, 95% CI 1.14, 1.31; p&amp;lt;0.0001). In an additional model inclusive of the stronger extra-parenchymal vascular predictor of death (PA/Ao over RV/LV), pruning remained significant. Conclusions: Among smokers without COPD on spirometry and with minimal emphysema on imaging, loss of distal pulmonary vascular volume is associated with increased risk of death. As pulmonary vascular injury may precede development of COPD and is associated with pulmonary hypertension (PH), these results question whether small vascular volume may be a measurable therapeutic target for disease prevention and progression.

IntroductionThe influence of anthropometrics, sex, and age at diagnosis on artificial intelligence (AI)–derived pulmonary blood volumes (PBV) from computed tomography pulmonary angiography (CTPA) remains poorly characterized. These physiological and biological determinants may affect PBV-based pulmonary hypertension (PH) prediction models.MethodsAn AI-based segmentation model quantified pulmonary artery and vein volumes in a secondary CTPA analysis from the Cambridge PH Registry. PBV were modelled as functions of anthropometrics, sex, and age at diagnosis, adjusting for pulmonary vascular resistance (PVR), PH diagnostic category (group 1, 2, 3 or 4 PH, or no PH), and number of cardiac comorbidities. The impact of PBV normalization strategies, including anthropometrics, on sex-related differences was assessed. Multivariable linear regression evaluated associations between PBV and invasively measured PVR and cardiac output (CO), and the incremental predictive value of anthropometrics.Results376 patients (median age 60 years; 57% female) investigated with right heart catheter were included: 120 pulmonary arterial hypertension, 30 group 2 PH, 79 group 3 PH, 102 chronic thromboembolic PH, and 45 without PH. Pulmonary artery volume increased with height, weight, body mass index (BMI), and body surface area (BSA) (all p < 0.001), with a stronger height-related effect in males. Older diagnosis age associated with larger pulmonary artery volume, particularly those with higher weight, BMI, and BSA (all p < 0.001). Pulmonary vein volume also increased with anthropometrics, but with older age disproportionately increasing volumes in females with higher weight, BMI, and BSA. PBV normalization to anthropometrics (height, weight, and BSA) attenuated sex-related differences (p < 0.001). In hemodynamic models, larger pulmonary artery volume and lower pulmonary vein volume were independently associated with higher PVR, while larger pulmonary vein volume was associated with higher CO (all p < 0.001). Inclusion of anthropometrics, particularly BSA, significantly improved prediction of PVR (F = 12.8, p < 0.001) and CO (F = 23.8, p < 0.001).ConclusionAI-derived PBV from CTPA are shaped by complex interactions between anthropometrics, sex, and diagnosis age, with distinct effects on pulmonary arterial and venous compartments. Accounting for these determinants is essential for translating AI-quantified PBV into clinically intuitive PH prediction or phenotyping models.

1. Noradrenaline induces a meagre vasoconstriction in small muscular pulmonary arteries compared to large conduit pulmonary arteries. We have examined whether this may be partially related to differences in the beta-adrenoceptor-mediated vasorelaxation component and, in particular, beta-adrenoceptor-mediated NO release. 2. Noradrenaline induced a bell-shaped concentration-response in large (1202+/-27 microm) and small (334+/-12 microm) pulmonary arteries of the rat. In large arteries tension increased to 95.6+/-1.8% of 75 mM KCl (KPSS; n=8) at 2 microM, above which tension declined. The response in small arteries was meagre (12+/-1.5% KPSS, n=9), peaking at 0.2 microM. N(G)-monomethyl-L-arginine (L-NMMA; 100 microM) abolished the decline in tension induced by higher concentrations of noradrenaline in large arteries, and increased maximum tension (117+/-3.5% KPSS, n=5, P<0.05). In small arteries peak tension doubled (22.0+/-3.4% KPSS, n=6, P<0.01), but still declined above 0.2 microM. 3. Propranolol (1 microM) abolished the decline in tension at higher concentrations of noradrenaline in both groups, but increased tension substantially more in small (37.4+/-3.7% KPSS, n=5, P<0.001) than in large arteries (112.2+/-3.7% KPSS, n=9, P<0.05). In the presence of L-NMMA, propranolol had no additional effect on large arteries, whereas in small arteries there was greater potentiation than for either agent alone (67.8+/-5.9% KPSS, n=4). 4. Beta-adrenoceptor-mediated relaxation was examined in arteries constricted with prostaglandin F2alpha (50 microM). In the presence of propranolol isoprenaline caused an unexpected vasoconstriction, which was abolished by phentolamine (10 microM). In the presence of phentolamine, isoprenaline caused a maximum relaxation of 43.3+/-2.1% (n=6) in large, and 49.0+/-4.5% (n=6) in small arteries. L-NMMA substantially reduced relaxation in large arteries (7.4+/-1.5%, n=6, P<0.01), but was less effective in small arteries (26.8+/-5.8, n=5, P<0.05). 5. Atenolol (beta1-antagonist, 5 microM) reduced relaxation to isoprenaline (large: 34.8+/-4.5%, n=5; small: 35.0+/-1.9%, n=6), but in combination with L-NMMA had no additional effect over L-NMMA alone. ICI 118551 (beta2-antagonist, 0.1 microM) reduced isoprenaline-induced relaxation more than atenolol (large: 18.0+/-4.6%, n=6, P<0.05; small: 25.6+/-10.7%, n=6, P<0.05). ICI 118551 in combination with L-NMMA substantially reduced relaxation (large: 4.8+/-2.6%, n=9; small: 6.5+/-3.6%, n=5). 6. Salbutamol-induced relaxation was reduced substantially by L-NMMA in large arteries (control: 34.7+/-6.4%, n=6; +L-NMMA: 8.3+/-1.3%, n=5, P<0.01), but to a lesser extent in small arteries (control: 50.9+/-7.5%, n=6; +L-NMMA: 23.0+/-0.7%, n=5, P<0.05). Relaxation to forskolin was also partially antagonized by L-NMMA. 7. These results suggest that the meagre vasoconstriction to noradrenaline in small pulmonary arteries is partially due to a greater beta-adrenoceptor-mediated component than in large arteries. Beta-mediated vasorelaxation in large arteries was largely NO-dependent, whereas in small arteries a significant proportion was NO-independent. Noradrenaline stimulation was also associated with NO release that was independent of beta-adrenoceptors.

Hypoplasia of the small pulmonary arteries in total anomalous pulmonary venous connection with obstructed pulmonary venous drainage

Differential actions of the prostacyclin analogues treprostinil and iloprost and the selexipag metabolite, MRE-269 (ACT-333679) in rat small pulmonary arteries and veins

Background and Aims Pulmonary hypertension (PH) is highly prevalent in the hemodialysis (HD) patient population. Right heart catheterism remains the gold standard for PH diagnosis and etiological stratification – this makes a comprehensive investigation of PH challenging in these patients. The PEPPER study suggested that postcapillary PH is the most common form of PH in HD patients, as the result of volume overload and left ventricular dysfunction. We hypothesized that novel quantitative imaging-derived biomarkers, such as pulmonary vessel volume and pulmonary artery volume, would improve our insight on the relationship between PH, volume status and left ventricular dysfunction in HD patients. In this study, we explored the combined role of noncontrast chest CT and echocardiography to investigate PH in a sample of HD patients. Method Study participants underwent noncontrast chest CT and doppler echocardiography on a non-HD day. To avoid potential confounders, chronic hemodialysis patients with previously diagnosed chronic lung disease, cancer and infections were excluded, and smoking history was limited to 20 packs/year. Pulmonary vessel volume was automatically segmented and measured using commercial software (VIDA Diagnostics Inc., Coralville, USA). Total pulmonary artery (PA) volume was segmented manually from CT, including 25 mm of the main, left and right pulmonary arteries starting from the bifurcation; volumes were calculated using a combination of in-house software (3D Quantify, Robarts Research Institute, London, Ontario, Canada; MATLAB MathWorks, Inc., Natick, Massachusetts, USA). PA volume and pulmonary vessel volume were indexed by body surface area (BSA), to correct for body size. Left atrial volume and PA systolic pressure were measured from doppler echocardiography according to current clinical guidelines. Associations between quantitative imaging biomarkers and demographics were assessed with Pearson and Spearman correlation, as appropriate. Linear fitting was performed with linear regression. Results Five HD patients were studied. Two patients had PA systolic pressure ≥ 35 mmHg. Preliminary analysis showed a nonlinear trend correlation between PA systolic pressure and pulmonary vessel volume/BSA (Panel A), PA systolic pressure and pulmonary artery volume/BSA (Panel B). Additionally, pulmonary vessel volume showed a significant, positive linear correlation with total pulmonary artery volume (Panel C) and left atrial volume (Panel D). Conclusion Preliminary correlations between pulmonary vessel volume, pulmonary artery volume, left atrial volume and PA systolic pressure suggest that intravascular volume and left ventricular dysfunction may play a significant role in determining PH in HD patients. Quantitative imaging allows screening for PH and provides additional, noninvasive, and relevant clinical information on the pathophysiology of PH in HD patients. Correlation for PA Systolic Pressure (mmHg) with pulmonary vessel volume/BSA and total PA volume/BSA (Panels A and B, respectively). Correlation for pulmonary vessel volume with left atrial volume and total PA volume (Panels C and D, respectively).

Using an X-ray television system on the in vivo cat lung, we directly measured internal diameter (ID) changes in the small pulmonary arteries and veins (100-1,100 microns ID) in response to 5, 15, and 40 ppm nitric oxide (NO) inhalations. We also measured to what extent 40 ppm NO inhalation can attenuate large ID constrictions at the different serial segments of the small vessels due to unilobar anoxic (0% O2) exposure. Under normoxic conditions, 5-40 ppm NO inhalations significantly increased the ID of both arteries and veins less than approximately 900 microns dose dependently but caused no significant, or only slight, ID increases in the vessels larger than this, if any at all. The ID increase in the serially connected arteries was nonuniform (4-18, 8-28, and 7-35% with 5, 15, and 40 ppm NO inhalations, respectively), whereas that for the veins was relatively uniform (4-9, 6-17, and 7-18% with 5, 15, and 40 ppm NO, respectively). The maximum ID increase occurred in the 200- to 500- and 200- to 700-microns arteries in response to 5-15 and 40 ppm NO, respectively. Unilobar anoxic exposure significantly decreased the ID of the 100- to 700-microns arteries and veins, but not the ID of the other-sized vessels. The ID decrease in the serially connected arteries was nonuniform (13-29%) but relatively uniform in the veins (8-12%). The maximum ID decrease occurred in the 200- to 300-microns arteries. However, adding 40 ppm NO to the lobe completely eradicated the ID decreases at all segments of the arteries and veins and, instead, caused significant ID increase (11-21%) in the arteries and (10-12%) in the veins. The data indicate that, according to dosage, 5-40 ppm NO inhalations cause selective dilation of approximately 100- to 900-microns pulmonary arteries and veins, particularly the 200- to 700-microns arteries. During anoxic exposure, the vasodilator effect of NO is preserved and can completely reverse the marked pulmonary vasoconstriction.

In this edition of the Equine Veterinary Journal, three epidemiological investigations address the impact of exercise-induced pulmonary haemorrhage (EIPH) on Thoroughbred racehorses. In prospective studies, Sullivan et al. followed the racing career of 744 Australian Thoroughbreds after a single post-race endoscopic examination 1, while Morley et al. enrolled 1000 South African Thoroughbreds to compare the race performance of EIPH-positive and -negative horses 2. These investigations found that EIPH, even an EIPH score of 1 2, affects earnings. In contrast, Preston and coworkers used Hong Kong Jockey Club records to examine the racing careers of 822 Thoroughbred geldings imported from New Zealand. This investigation, in which horses were examined multiple times (median 15), often by different veterinarians, found no effect of EIPH on length of racing career 3. Despite differences in experimental design, a topic which will be vigorously debated by epidemiologists, the balance of these investigations is that EIPH affects performance. The EIPH prevalence was similar in the three locales and, although one might question the value of the single examination used by Sullivan et al. to predict racing career 1, Preston et al. confirmed that the majority of horses maintained a consistently low EIPH grade on repeated examination 3. However, the EIPH score was less consistent in horses that had at least one serious bleed during their career. The findings in these papers raise the issues of disease progression in EIPH and of the value of EIPH score as an indicator of the presence and severity of lung lesions. A review of the pathogenesis of EIPH is pertinent to these questions. Thoroughbred racehorses have been selectively bred for speed. Densely packed muscle mitochondria provide a very high maximal oxygen consumption (220 ml/kg bwt/min), requiring a huge cardiac output (240–450 l/min or more) for lung to muscle oxygen transport. The high cardiac output is achieved by a heart beating about 4 times per second with a stroke volume 1.7 l or more. At this rapid heart rate, the cardiac cycle lasts only 0.25 s, so the ventricles must receive the 1.7 l in ∼0.125 s, which requires a very high ventricular filling pressure that is provided by the left atrium. Pertinent to EIPH, left atrial pressure reaches 70 mmHg or more during high-speed exercise. To maintain the driving pressure for blood flow through the lung, the pulmonary arterial pressure also has to increase to 90–140 mmHg. Given that the pulmonary capillaries lie between the pulmonary artery and the left atrium, their pressure is at least 80 mmHg (see Poole and Erickson 4 for review). In 1993, West and coworkers proposed that this pressure was causing pulmonary capillary stress failure, which they concluded was the major cause of EIPH 5. While there is little doubt that capillary failure is the ultimate cause of bleeding, it is a transient condition and resolves when the pressure is reduced. However, the lesions that are present in the lungs of horses with EIPH suggest that much more is going on than simple reversible stress failure 6-8. It is important to distinguish the high pulmonary arterial pressure that occurs in horses during intense exercise from that occurring in the approximately 15 per 1,000,000 people with idiopathic pulmonary hypertension 9. Idiopathic pulmonary hypertension is due to pulmonary arterial obstruction by varying degrees of hypertrophy/hyperplasia of the media and/or intima of the pulmonary arteries. Vascular resistance is increased in the arteries, which are upstream from the capillaries, consequently left atrial and, therefore, pulmonary capillary pressure remain within almost normal limits. In exercising horses, in contrast, the increased pulmonary arterial pressure is originating to a large degree downstream from the capillaries as a consequence of the elevated left atrial pressure. As blood backs up into the lungs, venous and then capillary pressure increase, similar to left heart failure 10. The lesions of EIPH reflect that situation. The bilateral gross lesions are characterised by varying degrees of blue-black discolouration of the pleural surface that is a consequence of haemosiderin accumulation. Haemosiderin is accompanied by fibrosis of the alveolar and interlobular septa and of the pleura, which makes the lung firm to the touch and unable to deflate fully when the lungs are excised. Vasculogenesis, probably from the bronchial circulation, also occurs within the lesions and may provide a low-resistance conduit between the systemic circulation and pulmonary capillaries within the lesion. Most importantly, there is striking remodelling of the small pulmonary veins (100–200 μm outer diameter) characterised mainly by accumulation of adventitial collagen and, in some vessels, smooth muscle hyperplasia. In the most severely affected vessels, the vascular lumen is significantly reduced 6-8. Given that remodelled veins can be mistaken for pulmonary arteries, earlier investigations may have described similar vascular changes using terms such as 'vascular lesions typical of hypertension' 11 or 'medial thickening of arterioles' 12. Venous remodelling can occur without haemosiderin, but the inverse rarely occurs; fibrosis never occurs without the other 2 lesions 8. The EIPH lesion distribution (most extensive in the caudodorsal region and less common more cranially and ventrally) matches the distribution of pulmonary blood flow 13, 14. Venous remodelling typically occurs when there is elevated venous transmural pressure and/or increased shear stress on the endothelium. An example familiar to most readers occurs when saphenous veins are used to replace blocked arteries in the human heart or extremities. In order to withstand the higher pressure, the saphenous veins remodel; medial and intimal thickening occur in response to pressure and endothelial shear stress, respectively 15. We have proposed that remodelling of small pulmonary veins of the racing horse is a response to the high flow and/or pressure to which the small veins are exposed during intense exercise 6-8. As its wall thickens, the small vein becomes less compliant 16 and, therefore, unable to distend to accommodate the high blood flow of exercise. This downstream obstruction further raises the capillary pressure, which increases the tendency to stress failure. In this scenario, the first-affected regions of lung should be those in which flow is greatest, i.e. parts of the dorsocaudal region where local blood flow can be up to 20-fold higher than in cranioventral regions 13, 14. As veins become obstructed in one region, however, flow will be diverted to adjacent regions, setting off the same scenario and gradually spreading the remodelling forwards and downwards from the initial site. Venous remodelling is likely to begin when horses first enter training. Lesions occur in unraced Thoroughbreds in training 12, and elevated left atrial pressure does not require racing speed but occurs during speeds used in gallops. An individual horse's degree of remodelling will be determined by the frequency and duration of the 'high-pressure event (HPE)' (i.e., gallops, breezes or races) and the individual horse's vascular sensitivity to the HPE. Some low-sensitivity animals may remodel little even after many HPEs and never bleed; others with very high sensitivity may remodel with only a few HPEs and bleed copiously on their first race. The majority of horses will fall in the middle range of the normal distribution of sensitivity to HPE. These may be the animals with grade 1, a score at which they are less likely to win 2 but their duration of career is unlikely to be affected 1, 3. In the most sensitive of this middle range of horses (grade ≥2), performance will be affected by the lesions in their lungs and, as the EIPH lesions expand, earning potential will reduce or horses will race at a lower class. Although this hypothesis must remain untested until the extent of pulmonary lesions in racehorses can be quantified, the following observations are supportive: EIPH lesions occur in trained but unraced animals 12; raced horses have stiffer small pulmonary veins than unraced horses 16; and there is evidence that EIPH is related to the number of lifetime starts rather than age 17. Is it reasonable to expect the EIPH score to correlate with lesion severity? Almost certainly, for the horses with grade 0 no lesions would be expected. Given that an EIPH score of 1 is associated solely with a reduced chance of winning but not of placing 2, it might be difficult to detect lesions other than capillary stress failure. It is likely that most horses with grade 2 or higher EIPH have lung lesions, microscopic if not gross. By the time a horse's EIPH cannot be managed, they probably all have significant gross lesions occupying 30% or more of the caudal lung lobe 8. We have noted considerable variation in the extent and severity of the lesions in horses retired because of EIPH, but an accurate history of the onset and progression of bleeding was not available. The lesions are very complex, so that blood could be originating from regions currently undergoing remodelling or from neovascularisation in lesions as described above. With regard to the source of neovascularisation, the arterialisation may be proceeding from the bronchial arteries 18 into an extensively remodelling capillary bed 19. Given that the bronchial artery is a systemic artery, the blood entering the neovascularised lesion and its adjacent capillaries may be at higher pressure and may or may not be able to escape from within the lesion in order to become visible on endoscopy. It is not surprising to us, therefore, that given the pathology of EIPH lesions, Preston et al. found considerable variation in EIPH scores in horses with high EIPH score at least once in their careers 3. To understand the pathogenesis and progression of the disease, it will be essential to conduct more investigations of lung pathology in horses with a variety of EIPH scores at various stages of their careers. In order to make a valid assessment of the extent and severity of lung lesions, it will be necessary to use a statistically valid but time-consuming sampling method, such as that used by Williams et al. 8. From the clinical viewpoint, it will be vital to develop ways to determine the extent of lesions in vivo. If that were possible, then EIPH-affected horses could be evaluated before making recommendations concerning their future career. The more difficult task will be preventing the development of lung lesions, because this will require fewer or less intense HPEs, which will require altered training methods. Alternatively, it may be possible to select against horses that are very sensitive to HPEs by not breeding from animals that develop high EIPH scores. The authors have no conflicts of interest. The investigations of EIPH-associated lung pathology described herein were supported by grants from the Grayson-Jockey Club Research Foundation.