PURPOSE: Pulmonary arteriovenous malformations (PAVMs) cause hypoxemia due to right-to-left (R-L) shunting. Erect blood oxygen saturation (SpO2) is a better predictor of R-L shunt than supine (Thompson et al, Chest 1999). Additionally 90% of PAVMs occur in patients with hereditary hemorrhagic telangiectasia, when anemia frequently coexists secondary to nasal and/or gastrointestinal bleeding. However most PAVM patients do not present with dyspnea (Shovlin et al Thorax 2008). We hypothesised that dyspnea, when present, may be better explained by other physiological factors.
METHODS: In 161 previously unstudied patients, presenting with CT-proven PAVMs between 06/05 and 07/10, retrospective analysis of self-reported exercise tolerance at presentation, age, gender, smoking history, body mass index, FEV1, VC, KCO, SpO2, pulse rate, hematologic, and biochemical indices were performed. Blinded to physiological measurements, two investigators assigned patients to the Medical Research Council dyspnea scale, with individuals undertaking regular/intense sport reclassified as Grade “0”. For statistical analyses of pulse rate and SpO2, the average of four separate measurements after 7, 8, 9 and 10 minutes at rest in both supine and erect postures were calculated. Statistical analyses were performed using STATA IC version 11 (Statcorp, Texas).
RESULTS: 44 (27.3%) of the population were dyspneic (MRC grade ≥2), reducing to 18.6% when 14 patients with severe co-existent cardiopulmonary pathologies were excluded. Orthodeoxia (SpO2 fall of ≥2% on standing) was present in 51/161 (31.7%). The dataset of 109 patients with no severe co-existing cardiopulmonary disease and full blood tests was examined. Low supine SpO2, but not erect, emerged as a significant predictor of dyspnea grade (odds ratio -0.064 [-0.11, -0.02], p=0.004). However, supine SpO2 was no more significant than other non-PAVM parameters of increased age (p<0.001); increased supine pulse rate (p=0.004) or low serum iron (p=0.004). Prospective cardiopulmonary exercise testing is in progress to dissect these relationships further (NRES 11/H0803/9).
CONCLUSIONS: These data imply that it is unusual for PAVM-induced hypoxemia alone to account for significant dyspnea.
CLINICAL IMPLICATIONS: Additional reversible factors should be sought for PAVM patients with symptomatic breathlessness.
DISCLOSURE: The following authors have nothing to disclose: Vatshalan Santhirapala, Hannah Tighe, James Springett, Heidi Wolfenden, Mike Hughes, James Jackson, Claire Shovlin
No Product/Research Disclosure Information