Increased intrapulmonary shunt and alveolar dead space post-COVID-19

Abstract

Increased intrapulmonary shunt (Q_S/Q_t) and alveolar dead space (V_D/V_T) are present in early recovery from 2019 Novel Coronavirus (COVID-19). We hypothesized patients recovering from severe critical acute illness (NIH category 3-5) would have greater and longer lasting increased Q_S/Q_t and V_D/V_T than patients with mild-moderate acute illness (NIH 1-2). Fifty-nine unvaccinated patients (33 males, aged 52 [38-61] yr, body mass index [BMI] 28.8 [25.3-33.6] kg/m²; median [IQR], 44 previous mild-moderate COVID-19, and 15 severe-critical disease) were studied 15-403 days postacute severe acute respiratory syndrome coronavirus infection. Breathing ambient air, steady-state mean alveolar Pco₂, and Po₂ were recorded simultaneously with arterial Po₂/Pco₂ yielding aAPco₂, AaPo₂, and from these, Q_S/Q_t%, V_D/V_T%, and relative alveolar ventilation (40 mmHg/[Formula: see text], VArel) were calculated. Median [Formula: see text] was 39.4 [35.6-41.1] mmHg, [Formula: see text] 92.3 [87.1-98.2] mmHg; [Formula: see text] 32.8 [28.6-35.3] mmHg, [Formula: see text] 112.9 [109.4-117.0] mmHg, AaPo₂ 18.8 [12.6-26.8] mmHg, aAPco₂ 5.9 [4.3-8.0] mmHg, Q_S/Q_t 4.3 [2.1-5.9] %, and V_D/V_T16.6 [12.6-24.4]%. Only 14% of patients had normal Q_S/Q_t and V_D/V_T; 1% increased Q_S/Q_t but normal V_D/V_T; 49% normal Q_S/Q_t and elevated V_D/V_T; 36% both abnormal Q_S/Q_t and V_D/V_T. Previous severe critical COVID-19 predicted increased Q_S/Q_t (2.69 [0.82-4.57]% per category severity [95% CI], P < 0.01), but not V_D/V_T. Increasing age weakly predicted increased V_D/V_T (1.6 [0.1-3.2]% per decade, P < 0.04). Time since infection, BMI, and comorbidities were not predictors (all P > 0.11). VArel was increased in most patients. In our population, recovery from COVID-19 was associated with increased Q_S/Q_t in 37% of patients, increased V_D/V_T in 86%, and increased alveolar ventilation up to ∼13 mo postinfection. NIH severity predicted Q_S/Q_t but not elevated V_D/V_T. Increased V_D/V_T suggests pulmonary microvascular pathology persists post-COVID-19 in most patients.NEW & NOTEWORTHY Using novel methodology quantifying intrapulmonary shunt and alveolar dead space in COVID-19 patients up to 403 days after acute illness, 37% had increased intrapulmonary shunt and 86% had elevated alveolar dead space likely due to independent pathology. Elevated shunt was partially related to severe acute illness, and increased alveolar dead space was weakly related to increasing age. Ventilation was increased in the majority of patients regardless of previous disease severity. These results demonstrate persisting gas exchange abnormalities after recovery.

Keywords: alveolar dead space; hyperventilation; intrapulmonary shunt; post-COVID-19; ventilation perfusion mismatch.

Graphical abstract

Figure 1.

Alveolar-arterial differences for Po₂ (AaPo₂) and Pco₂ (aAPco₂) across the indicated combinations of shunt and dead space, each from 0 to 50%. The grid can be used to estimate both shunt and alveolar dead space for any combination of AaPo₂ and aAPco₂. Note that the position of each point on the grid is dependent on a number of ancillary variables including pH, Hb, Hb P50, temperature, base excess, V̇e, cardiac output, V̇co₂, and V̇o₂, with all except cardiac output and Hb P50 automatically available from either the expired gas or arterial blood samples used to measure AaPo₂ and aAPco₂. Reproduced by editorial request, and modified from Wagner et al. (10). aAPco₂, arterial- alveolar partial pressure difference for carbon dioxide; AaPo₂, alveolar-arterial partial pressure difference for oxygen; V̇o₂, oxygen consumption in L/min; V̇co₂, carbon dioxide production in L/min.

Figure 2.

Individual data for 59 post-COVID-19 patients NIH 1-2 (44 patients with low-medium severity, open blue circles) and NIH 3-5 (15 patients with severe-critical severity, enclosed red squares) for arterial Po₂ (A), arterial Pco₂ (B), alveolar Po₂ (C), and alveolar CO₂ (D). **P < 0.005 compared to low-medium severity. Black horizontal line represents median value. COVID-19, 2019 Novel Coronavirus; NIH, National Institutes of Health.

Figure 3.

Individual data for 59 post-COVID-19 patients NIH 1-2 (44 patients, low-medium severity, open blue symbols) and NIH 3-5 (15 patients, severe-critical severity, closed red squares) showing AaPo₂ mmHg (A), aAPco₂ mmHg (B), shunt (C), and alveolar dead space (D). **P < 0.008 compared to low-medium severity. Black horizontal line represents median values, dotted lines represent 95% confidence levels for shunt (5%), and alveolar dead space (10%) from healthy subjects (11). aAPco₂, arterial- alveolar partial pressure difference for carbon dioxide; AaPo₂, alveolar-arterial partial pressure difference for oxygen; COVID-19, 2019 Novel Coronavirus; NIH, National Institutes of Health.

Figure 4.

Individual data from 7 healthy subjects (closed black triangles; A and C) and for 59 post-COVID-19 patients (B and D) NIH severity 1–2 (44 patients; open blue circles) and NIH severity 3–5 (15 patients; closed red squares) showing: measured AaPo₂ mmHg and aAPco₂ mmHg (A and B) and calculated shunt and alveolar dead space values (C and D). Dotted lines represent 95% confidence levels for shunt (5%), and alveolar dead space (10%) from healthy subjects (11). Note that the healthy subjects’ data lie within, or close to, the 95% confidence levels, indicating technical adequacy of the AaPo₂ and aAPco₂ measurements. aAPco₂, arterial- alveolar partial pressure difference for carbon dioxide; AaPo2, alveolar-arterial partial pressure difference for oxygen; COVID-19, 2019 Novel Coronavirus; NIH, National Institutes of Health.

Figure 5.

Individual data for 59 subjects for Bohr-Enghoff dead space [calculated as $\left(\frac{\text{P}{\text{a}}_{\text{C}{\text{O}}_2}\hspace{0.17em}-\hspace{0.17em}\text{P}{\text{E}}_{\text{C}{\text{O}}_2}}{\text{P}{\text{a}}_{\text{C}{\text{O}}_2}}\right)\hspace{0.17em}$× 100] plotted against measured alveolar dead space as described in the text (A) and AaPo₂ from the standard alveolar gas equation, where $\text{P}{\text{A}}_{{\text{O}}_2}$ = $\text{P}{\text{I}}_{{\text{O}}_2}$ ₋ $\text{P}{\text{a}}_{\text{C}{\text{O}}_2}$/R + $\text{P}{\text{a}}_{\text{C}{\text{O}}_2}$ × $\text{F}{\text{I}}_{{\text{O}}_2}$ × (1−R)/R plotted against AaPo₂ using measured $\text{P}{\text{A}}_{{\text{O}}_2}$ as described in the text (B). Lines of identity are in red, linear regression lines in black. Note that Bohr-Enghoff dead space measurements are much greater than the measured alveolar dead space measurement and AaO₂ gradient calculation underestimates the true AaO₂ gradient. AaPo₂, alveolar-arterial partial pressure difference for oxygen; $\text{F}{\text{I}}_{{\text{O}}_2}$, fraction of inspired oxygen; $\text{P}{\text{a}}_{\text{C}{\text{O}}_2}$, partial pressure of arterial carbon dioxide; $\text{P}{\text{A}}_{{\text{O}}_2}$, partial pressure of alveolar oxygen; $\text{P}{\text{I}}_{{\text{O}}_2}$, partial pressure of inspired oxygen.

Figure 6.

Individual data for 59 post-COVID-19 patients for shunt (A) and alveolar dead space (B) plotted against time since acute SARS-CoV-2 infection for post-COVID-19 patients with low-medium severity (44 patients, NIH-1 and 2, open blue circles) and severe-critical severity (15 patients, NIH-3-5, closed red squares). Dotted lines represent 95% confidence levels for shunt (5%), and alveolar dead space (10%) from healthy subjects (11). COVID-19, 2019 Novel Coronavirus; NIH, National Institutes of Health.

Figure 7.

Individual data for 59 post-COVID patients NIH severity 1–2 (44 patients, open blue circles), and NIH severity 3–5 (15 patients, closed red squares) and methodological controls (7 subjects, black triangles) for relative ventilation plotted against $\text{P}{\text{a}}_{{\text{O}}_2}$ (A), intrapulmonary shunt (B), $\text{P}{\text{a}}_{\text{C}{\text{O}}_2}$ (C), alveolar dead space (D), and $\text{P}{\text{a}}_{\text{C}{\text{O}}_2}$ plotted against alveolar dead space (E). For comparison in A, historical data from healthy controls for Wagner et al. [closed black circles (19)], Torre-Bueno et al. [open black circles (21)] and Hammond et al. [closed black squares (20)] are also plotted. Vertical line 5% shunt (B), $\text{P}{\text{a}}_{\text{C}{\text{O}}_2}$= 40 mmHg (C), 10% alveolar dead space (D), $\text{P}{\text{a}}_{\text{C}{\text{O}}_2}$= 40 mmHg (E). Curved lines in C represents V_Arel = 40/$\text{P}{\text{a}}_{\text{C}{\text{O}}_2}$, V_Arel = 40/($\text{P}{\text{a}}_{\text{C}{\text{O}}_2}$ − 5), and V_Arel = 40/($\text{P}{\text{a}}_{\text{C}{\text{O}}_2}$ − 10) if $\text{P}{\text{A}}_{\text{C}{\text{O}}_2}$ = $\text{P}{\text{a}}_{\text{C}{\text{O}}_2}$_, and horizonal line is at V_Arel= 1.0. Horizontal line in E is at alveolar dead space = 10%. COVID-19, 2019 Novel Coronavirus; NIH, National Institutes of Health; $\text{P}{\text{a}}_{\text{C}{\text{O}}_2}$, partial pressure of arterial carbon dioxide; $\text{P}{\text{A}}_{\text{C}{\text{O}}_2}$, partial pressure of alveolar carbon dioxide; $\text{P}{\text{a}}_{{\text{O}}_2}$, partial pressure of arterial oxygen; V_Arel, 40/$\text{P}{\text{A}}_{\text{C}{\text{O}}_2}$= relative alveolar ventilation.