Insights into ventilation–gas exchange coupling in chronic thromboembolic pulmonary hypertension

To the Editor:

Exercise intolerance due to excessive ventilation and dyspnoea are fundamental clinical features of patients with pulmonary vascular diseases [1]. Among these diseases, chronic thromboembolic pulmonary hypertension (CTEPH) is associated with the largest increase in exercise ventilation [2]. Understanding the mechanism(s) underlying excess exercise ventilation in CTEPH is clinically relevant when designing evidence-based therapeutic and rehabilitative strategies to improve patients' symptoms and quality of life.

Excess exercise ventilation in CTEPH has been traditionally ascribed to increased “wasted” ventilation, i.e. hypoperfusion of well-ventilated alveoli. In fact, the physiological dead space fraction of tidal volume (V_D/V_T ratio) at peak exercise not only predicts CTEPH after pulmonary embolism [3] but also improves after clinical [4] and surgical [5] treatment. As the end-tidal carbon dioxide tension (P_ETCO2) diminishes when ventilation is excessive relative to perfusion [6], increased V_D/V_T has been mechanistically linked to inordinately low P_ETCO2 in CTEPH [7]. A P_ETCO2 lower than arterial (a) carbon dioxide tension (P_CO2) (i.e. a positive P_(a-ET)CO2 gradient) [6] also suggests impaired pulmonary perfusion as a potential explanation for low exercise P_ETCO2in these patients [8].

Surprisingly, however, seminal studies comparing lung absorption of multiple inert gases showed no or a limited increase in V_D/V_T despite marked pulmonary arterial obstruction [9]. This prompts an alternative (or complementary) explanation for a low exercise P_ETCO2 in CTEPH: reduced alveolar P_CO2 (hyperventilation) [10]. In fact, heightened neural drive has been found in pulmonary vascular diseases secondary to increased chemosensitivity [11] and/or higher afferent stimuli from “central” baro- and mechano-receptors [12]. Thus, alveolar hyperventilation under the stress of exercise may also explain the relatively lower P_ETCO2 in CTEPH. A hyperbolic correlation between increasing minute ventilation (V′_E)/carbon dioxide production (V′_CO2) and decreasing arterial carbon dioxide tension (P_aCO2) in CTEPH is also suggestive of heightened neural drive and consistent with the alveolar equation [13].

There is another intriguing feature of the P_ETCO2 response that brings additional uncertainty about the meaning of a low exercise P_ETCO2 in CTEPH. Similar to pulmonary arterial hypertension (PAH) [9], P_ETCO2 paradoxically increases (instead of further decreasing as in normal subjects) as soon as CTEPH patients enter the recovery phase. In this context, evaluation of P_aCO2 (or arterialised capillary (P_cCO2) as its surrogate) [8] and P_(c-ET)CO2 across the exercise-to-recovery transition might shed light on the mechanisms underlying the P_ETCO2 behaviour during recovery in CTEPH. Thus, we reasoned that during recovery: 1) if P_cCO2 remains stable or further decreases despite higher P_ETCO2, a narrower P_(c-ET)CO2 would suggest improved V_D/V_T (scenario 1); 2) conversely, if P_cCO2 and P_ETCO2 increase proportionally leading to stable P_(c-ET)CO2, lower neural drive would explain higher P_ETCO2 (scenario 2); 3) but if a narrower P_(c-ET)CO2 develops as a consequence of P_cCO2, increasing less than P_ETCO2, improved V_D/V_T plus lower neural drive would explain a higher P_ETCO2 (scenario 3).

We therefore measured P_cCO2 (arterialised ear lobe blood), P_(c-ET)CO2 and plasma lactate 1 min before peak incremental exercise, at peak and in recovery at every minute up to the fifth minute of unloaded cycling. Responses from 10 patients (age 54–78 years, five males, mean±se pulmonary arterial pressure 51.5±9.7 mmHg, pulmonary wedge pressure 9.2±3.4 mmHg, stable segmental perfusion defects after 6 months of anticoagulation) were contrasted with those from eight age- and sex-matched healthy subjects. As expected, patients had a lower peak oxygen uptake, V′_CO2 and arterial oxygen saturation measured by pulse oximetry (S_pO2) (p<0.05). Lactate corrected for peak work rate was higher in patients (0.14±0.05 versus 0.08±0.04 mEq·L⁻¹·W⁻¹; p<0.01). These findings were associated with lower peak P_ETCO2 and positive P_(c-ET)CO2 (figure 1b).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1

a) Ventilation, carbon dioxide (CO₂) output, b) capillary carbon dioxide tension (P_cCO2) and capillary end-tidal carbon dioxide tension (P_ETCO2), and c) their differences during exercise and recovery in patients with chronic thromboembolic pulmonary hypertension and healthy controls. Data are presented as mean±se. *: p<0.05 for inter-group comparisons at a given time-point.

In line with our previous observations, whereas recovery P_ETCO2 consistently increased throughout recovery in patients, it remained lower than peak exercise in controls (figure 1b). This increased P_ETCO2 was associated with a faster decline in V′_E compared with V′_CO2 in patients; conversely, V′_E recovered at a slower rate than V′_CO2 in controls (figure 1a). Lactate remained similarly increased in both groups; conversely, S_pO2 was lower in patients up to the third minute of recovery (91.1±3.5% versus 96.4±2.0%; p<0.05). In controls, P_cCO2 further decreased up to the second minute and remained stable thereafter, i.e. a negative P_(c-ET)CO2 was found for most of recovery. In patients, P_cCO2 increased throughout recovery; however, as it increased less than P_ETCO2, P_(c-ET)CO2 systematically decreased in CTEPH (figure 1c). Under the logical assumption that recovery reversed the abnormalities seen during exercise, scenario 3 seems in line with the suggestions of Delcroix et al. [10] and Naeije and van de Borne [11] that both high V_D/V_T and heightened neural drive are related to excess exercise ventilation in CTEPH.

As pointed out by the late eminent physiologist Brian J. Whipp, the transition between different rates of metabolic demand (e.g. exercise–recovery) provides a unique window to understand the complexities linking tissue gas exchange to pulmonary ventilation [14]. Thus, when V′_CO2 slowly decreased after exercise in our patients, V′_E did not lag behind as in normal subjects. Conversely, V′_E decreased consistently faster than V′_CO2 in patients (figure 1a). The sudden decrease in V′_E and resulting elevation in P_cCO2 (figure 1b) suggest an abrupt reduction in neural drive in these patients. Sluggish muscle gas exchange kinetics due to a slower rate of phosphocreatine replenishment and/or slower cardiocirculatory dynamics during recovery may have contributed to a more protracted V′_CO2 in patients [14]. It is also conceivable that carbon dioxide released during exercise has been slowly washed-out from carbon dioxide reservoirs, which are likely to be larger in chronically hypocapnic patients [15]. In fact, resting P_aCO2 and exercise-induced decrements in P_cCO2 were closely related to V′_CO2 kinetics in patients (r=0.85 and r=0.88, respectively; p<0.001). Although additional studies are needed to clarify the underlying mechanism(s), it is clear that V′_E did not “wait” for V′_CO2, which resulted in post-exercise decreases in the V′_E/V′_CO2 ratio, thereby contributing to increase P_ETCO2 in CTEPH.

Another key finding relates to a P_(c-ET)CO2 decrease after exercise in patients, which seems consistent with post-exercise improvement in V_D, likely to be secondary to enhanced pulmonary perfusion and ventilation/perfusion matching [6, 7]. Zhai et al. [13] found that while V′_E/V′_CO2 (assumed as an index of V_D/V_T behaviour) for a given P_aCO2 was increased in CTEPH compared with PAH, the difference disappeared when V′_E/V′_CO2 was expressed as a function of P_ETCO2. This suggests a greater contribution of increased V_D to the excessive exercise ventilation (and low P_ETCO2) in CTEPH compared with PAH during exercise.

In conclusion, increased “wasted” ventilation (high V_D/V_T) is not the only explanation for excess exercise ventilation in CTEPH: heightened neural drive also contributes to these abnormalities. Interventions able to decrease both V_D/V_T and neural drive might prove particularly valuable in this patient population.