Diffusing capacity for carbon monoxide
Diffusing capacity for carbon monoxide (DLCO), also known as the transfer factor for carbon monoxide (TLCO), is a pulmonary function test that quantifies the lung's ability to transfer gas from the alveoli to the pulmonary capillary blood, specifically measuring the uptake of carbon monoxide (CO) per minute per unit of pressure gradient, typically expressed in ml/min/mmHg.[1] It serves as a surrogate for oxygen transfer efficiency, providing insight into the integrity of the alveolar-capillary membrane and pulmonary blood volume.[2] DLCO is particularly valuable for diagnosing and monitoring conditions affecting gas exchange, such as emphysema, interstitial lung disease, and pulmonary vascular disorders.[1]Fundamentals
Definition and Purpose
The diffusing capacity of the lung for carbon monoxide (DLCO) is a key pulmonary function test that quantifies the lung's ability to transfer gas from the alveoli to the pulmonary capillary blood. It represents the rate at which carbon monoxide (CO) is taken up by the blood per unit of pressure difference between the alveolar air and the blood, typically expressed in milliliters of CO per minute per millimeter of mercury (mL CO/min/mmHg). This measurement provides an indirect assessment of the efficiency of gas exchange across the alveolar-capillary membrane, encompassing both the diffusive properties of the membrane itself and the volume of blood available for gas binding in the pulmonary capillaries.[3][4] The primary purpose of DLCO testing is to evaluate the integrity of the alveolar-capillary interface and the adequacy of pulmonary capillary blood volume, which are critical for effective oxygenation of blood. Clinically, it aids in the diagnosis and monitoring of various lung diseases, including restrictive disorders such as interstitial lung disease and pulmonary fibrosis, as well as certain obstructive conditions like emphysema, where reduced DLCO values indicate impaired gas transfer. By isolating the diffusive component of gas exchange, DLCO complements other spirometric measures like forced vital capacity and helps differentiate between parenchymal lung damage and vascular issues, such as pulmonary hypertension.[3][5] DLCO was first described in 1915 by Danish physiologist Marie Krogh, who selected CO as the test gas due to its minimal endogenous production in the body and its high affinity for hemoglobin, which ensures that uptake is primarily limited by diffusion rather than reaction rates. This innovation allowed for the first quantitative evaluation of pulmonary diffusing capacity in humans. Over the subsequent century, the test has evolved into a standardized procedure, with volumes reported under standard temperature and pressure, dry (STPD) conditions to ensure comparability across measurements. Corrections for factors such as hemoglobin concentration and altitude are routinely applied to normalize results and account for physiological variations that influence CO binding.[6][7][4]Physiological Role in Gas Exchange
The diffusing capacity for carbon monoxide (DLCO) serves a key physiological role in evaluating the lung's capacity for gas exchange during respiration, particularly by approximating the diffusing capacity for oxygen (DL_O2). Both CO and O₂ are diffusion-limited gases that adhere to Fick's law of diffusion across the alveolar-capillary membrane, allowing DLCO to provide insight into the lung's ability to transfer O₂ from alveoli to pulmonary capillary blood. In contrast, CO₂ exhibits much higher membrane solubility—approximately 20 times that of O₂—resulting in faster diffusion and making its transfer primarily perfusion-limited rather than diffusion-limited. According to Fick's law predictions, CO transport across the lung is about 83% that of O₂ due to differences in molecular weight and solubility.[8] DLCO integrates contributions from the alveolar-capillary membrane and the pulmonary blood, expressed through its primary components: the membrane diffusing capacity (Dm), which reflects the surface area available for diffusion and the thickness of the blood-gas barrier, and the blood reaction component (θVc), where θ represents the rate at which CO reacts with hemoglobin and Vc denotes the volume of blood in the pulmonary capillaries. The overall DLCO is related to these by the Roughton-Forster equation: \frac{1}{\text{DLCO}} = \frac{1}{\text{Dm}} + \frac{1}{\theta \text{Vc}} This formulation, established in seminal work on gas transfer kinetics, demonstrates how reductions in membrane area or increases in barrier thickness diminish Dm, while alterations in hemoglobin availability affect θ and thus the blood component.[1][8] At its core, DLCO derives from Fick's law of diffusion, which quantifies the steady-state rate of gas transfer (V̇gas) as: \dot{V}_{\text{gas}} = A \cdot D \cdot \frac{(P_1 - P_2)}{T} Here, A is the effective surface area of the alveolar-capillary interface, D is the diffusion coefficient specific to the gas (influenced by its solubility and molecular size), (P₁ - P₂) is the partial pressure gradient driving diffusion, and T is the effective thickness of the diffusion barrier. To derive DLCO, this law is applied along the length of the pulmonary capillary, where gas partial pressure in blood approaches equilibrium exponentially with transit time; the overall capacity is then the ratio of uptake rate to the mean driving pressure, standardized for a low test gas concentration to minimize back-pressure effects. This integration reveals DLCO as a conductance measure (mL/min/mmHg) that captures the lung's integrated diffusive efficiency for poorly soluble gases like O₂ and CO.[1][8] Despite its utility, DLCO measurements are subject to physiological limitations related to hemoglobin binding dynamics. In anemia, reduced hemoglobin concentration lowers θ, decreasing the measured DLCO and causing it to underestimate the true membrane diffusing capacity (Dm) unless adjusted using established correction formulas. Conversely, high cardiac output states often increase DLCO due to greater pulmonary capillary blood volume (Vc), enhancing the blood component of gas transfer. These factors highlight the need to interpret DLCO in the context of systemic influences on blood reactivity and flow.[1][9][8][3]Measurement Techniques
Standard Procedure
The standard procedure for measuring diffusing capacity for carbon monoxide (DLCO) employs the single-breath technique, as outlined in the 2017 ERS/ATS technical standards.[10] Patient preparation begins with ensuring no fasting is required, though heavy meals should be avoided immediately prior to testing to prevent discomfort.[10] Patients are instructed to abstain from smoking for at least 24 hours beforehand to minimize elevated carboxyhemoglobin levels that could interfere with CO uptake measurements.[11] The test is typically performed in a standardized sitting position, with the option to repeat in the supine position if assessing positional changes in gas transfer; a nose clip and mouthpiece are applied for an airtight seal.[10] Supplemental oxygen, if used, should be discontinued for at least 10-15 minutes prior, provided it is clinically safe.[1] The procedure commences with the patient exhaling to residual volume. They then rapidly inhale a test gas mixture containing approximately 0.3% carbon monoxide, 10% helium (as an inert tracer gas), 21% oxygen, and the balance nitrogen, reaching total lung capacity (TLC) within 2-4 seconds, with inspired volume comprising at least 85-90% of vital capacity.[10] The breath is held for a standardized 10 seconds (±2 seconds), avoiding straining maneuvers such as Valsalva or Müller to ensure uniform distribution.[10] Exhalation follows smoothly and rapidly to residual volume, with the initial 0.75-1.0 liter (dead space) discarded; an alveolar sample of 0.5-1.0 liter is then collected and analyzed for CO and helium concentrations using a gas analyzer or chromatograph.[10] This process is repeated after a 4-minute interval to allow CO washout. Test reproducibility requires at least two acceptable maneuvers, with the two closest DLCO values within 2 mL·min⁻¹·mmHg⁻¹ (or 0.67 mmol·min⁻¹·kPa⁻¹) of each other, or within 10% for values below 20 mL·min⁻¹·mmHg⁻¹; up to five attempts may be made, but no more than two are averaged for reporting.[10] Corrections are applied for environmental factors, including conversion to body temperature and pressure saturated (BTPS) conditions, adjustments for barometric pressure and water vapor, ambient temperature, and the inspired volume to estimate alveolar volume (VA).[10] Safety considerations emphasize the low toxicity of the CO dose, which typically increases carboxyhemoglobin by only 0.5-1% per test, posing minimal risk to most patients.[10] Absolute contraindications include recent pneumothorax or known carbon monoxide toxicity, while relative contraindications encompass acute myocardial infarction within the past month, severe hypertension, or inability to follow instructions due to cognitive impairment.[1] Potential side effects are rare but may include transient lightheadedness or mild hypoxia. For non-cooperative patients unable to perform breath-holding, a rebreathing method serves as an alternative, involving continuous rebreathing of the gas mixture for 30-60 seconds in a closed circuit to achieve equilibration, though it requires different equipment calibration and is less commonly used in routine clinical settings.[10]Equipment and Calculations
The measurement of diffusing capacity for carbon monoxide (DLCO) requires specialized pulmonary function testing equipment to ensure accurate gas analysis, volume measurement, and data processing. Gas analyzers are essential for detecting carbon monoxide (CO) and the inert tracer gas, typically helium (He), with rapidly responding gas analyzers (RGAs) preferred in modern systems for their ability to provide continuous measurements during the breath-hold maneuver. These analyzers must meet performance criteria including a 0–90% response time of ≤150 ms, nonlinearity of ≤1% of full scale, accuracy within ±1%, and drift of ≤10 ppm for CO or ≤0.5% for the tracer gas over 30 seconds. Common types include non-dispersive infrared (NDIR) sensors for CO and thermal conductivity detectors for helium, though mass spectrometers offer high precision for simultaneous multi-gas analysis in advanced setups.[10][4] Spirometers integrated into the system measure inspired and expired volumes, with flow sensors accurate to ±2% across -10 to +10 L/s and volume accuracy of ±75 mL when calibrated with a 3-L syringe. These devices operate under body temperature and pressure saturated (BTPS) conditions to reflect physiological lung volumes. Computer software is critical for real-time signal alignment (gas and flow data digitized at ≥100 Hz with ≥14-bit resolution), automated calculations, and quality grading, often incorporating global lung function initiative (GLI) reference equations for interpretation. Dead-space volume in the equipment must be minimized to <200 mL for adults to avoid dilution errors.[10][4] The primary calculation for DLCO derives from the single-breath method, quantifying CO uptake as the product of alveolar volume (VA) and the transfer coefficient (KCO). Specifically, \text{DLCO} = \frac{V_A}{t_{BH}} \times (P_B - P_{H_2O}) \times \ln\left(\frac{F_{I_{CO}}}{F_{A_{CO}} \times \frac{F_{A_{Tr}}}{F_{I_{Tr}}}}\right) where V_A is alveolar volume in liters (BTPS), t_{BH} is breath-hold time in seconds, P_B is barometric pressure in mmHg, P_{H_2O} is water vapor pressure in mmHg, F_{I_{CO}} is the inspired CO fraction, F_{A_{CO}} is the alveolar CO fraction at end-breath-hold, and F_{A_{Tr}} and F_{I_{Tr}} are the alveolar and inspired tracer gas fractions, respectively. This logarithmic form accounts for the exponential decay of CO during breath-holding, yielding DLCO in mL·min⁻¹·mmHg⁻¹ (or mmol·min⁻¹·kPa⁻¹ in SI units). An approximation for short breath-holds uses a linear difference: DLCO ≈ [VA × (F_{A_{CO,0}} - F_{A_{CO}})] / [ (P_B - P_{H_2O}) × t_{BH} / 760 ], but the full equation is standard for precision.[10][4] Alveolar volume (VA) is estimated via single-breath dilution of the inert tracer gas (commonly 0.3% helium), reflecting the accessible gas-exchanging lung volume. The formula is V_A = (P_B - P_{H_2O}) \times V_I \times \frac{F_{I_{He}} - F_{A_{He}}}{F_{A_{He}}} where V_I is the inspired volume in liters (BTPS). This assumes negligible initial helium in the lung and corrects for dead-space contributions; in rapid gas analyzer systems, VA may incorporate mass balance from continuous exhaled tracer data for enhanced accuracy. VA is reported under BTPS conditions and typically ranges 4–6 L in healthy adults, directly influencing DLCO as it scales with lung size.[10][4][3] Corrections adjust raw DLCO for physiological and environmental factors to standardize results. For hemoglobin (Hb), which affects CO binding, the correction is \text{DLCO}_{corr} = \text{DLCO} \times \frac{1.7 \times \text{Hb}}{10.22 + \text{Hb}} for adult males and adolescents (standard Hb = 14.6 g/dL), or \text{DLCO}_{corr} = \text{DLCO} \times \frac{1.7 \times \text{Hb}}{9.38 + \text{Hb}} for adult females and children under 15 years (standard Hb = 13.4 g/dL), with Hb measured in g/dL. This accounts for anemia or polycythemia, as DLCO varies linearly with Hb concentration in the R-state. For carboxyhemoglobin (COHb), back-pressure from endogenous CO reduces uptake; correction uses \text{DLCO}_{corr} = \text{DLCO} \times \left(1 + \frac{F_{A_{CO_b}}}{560}\right) where F_{A_{CO_b}} is the end-expiratory baseline CO fraction (often 0.5–1% in smokers), requiring co-oximetry measurement. Lung volume corrections are indirect, adjusting VA for predicted total lung capacity minus dead space if discrepancies arise, though no universal formula exists beyond ensuring full inspiration. Barometric pressure adjustments approximate DLCO scaling with (0.505 + 0.00065 × PB) for PB in mmHg. These are applied post-measurement using patient-specific data like recent Hb and COHb levels.[10][4][12] Quality control follows ATS/ERS guidelines to ensure test acceptability and repeatability. Maneuvers are acceptable if there are no leaks (verified by volume stability), full inspiration achieves ≥90% of vital capacity (or ≥85% if VA is within 5% of the maximum), inspiration time is <4 seconds for ≥85% of VI, breath-hold duration is 8–12 seconds (10 ± 2 s), and exhalation sample collection is ≤4 seconds without Valsalva or glottis closure. At least two acceptable tests are required, with DLCO values repeatable within 2 mL·min⁻¹·mmHg⁻¹ (0.67 mmol·min⁻¹·kPa⁻¹) or <10% of the average. Systems grade maneuvers A–E based on criteria met, with A (all criteria) or B (minor deviations) usable for reporting the mean. Daily calibration with certified gases and biological controls (e.g., healthy subjects) maintains accuracy.[10][4]Underlying Mechanism
Principles of Diffusion
The diffusion of gases across the alveolar-capillary membrane in the lungs follows Fick's law of diffusion, which quantifies the rate of gas transfer as directly proportional to the surface area available for diffusion, the diffusion coefficient of the gas (a measure of its solubility and molecular mobility in the tissue), and the partial pressure gradient across the membrane, while being inversely proportional to the thickness of the diffusion barrier.[13] This principle underpins the measurement of diffusing capacity for carbon monoxide (DLCO), where the overall transfer rate reflects the integrated effects of these biophysical factors in pulmonary gas exchange.[8] In healthy lungs, CO diffusion is predominantly limited by the alveolar-capillary membrane rather than by the blood's capacity to bind CO, owing to the gas's high affinity for hemoglobin, which facilitates rapid uptake once CO reaches the red blood cells. This membrane limitation is described by the Roughton-Forster equation, which partitions DLCO into the membrane diffusing capacity (DM) and the blood component (θCO × Vc, where θCO is the rate of CO uptake by blood and Vc is pulmonary capillary blood volume):\frac{1}{\text{DLCO}} = \frac{1}{\text{DM}} + \frac{1}{\theta_{\text{CO}} \cdot \text{Vc}}
Rearranging yields DM as
\text{DM} = \frac{\text{DLCO} \cdot \theta_{\text{CO}} \cdot \text{Vc}}{\theta_{\text{CO}} \cdot \text{Vc} - \text{DLCO}}
This formulation highlights how membrane properties dominate under normal conditions, with blood resistance contributing minimally to the total resistance for CO.[14] Historically, the diffusion properties of CO in lung tissue were quantified using Krogh's constant (K), defined as the product of the diffusion coefficient and solubility, representing the tissue's permeability to the gas; for CO, this value is approximately 0.0008 mL/min/mmHg/cm², as determined in early physiological studies of pulmonary diffusion.[15] This constant provides a foundational metric for modeling gas permeation through lung parenchyma and informs modern interpretations of DLCO variability. Endogenous CO production from heme breakdown by heme oxygenase generates a small back pressure in the pulmonary capillaries (typically 0.3–0.5 mmHg partial pressure), which diminishes the effective partial pressure gradient for exogenous test CO and thereby reduces the measured DLCO by approximately 5–10% compared to the true diffusing capacity.[4] This effect is routinely accounted for in clinical measurements through corrections based on estimated or measured carboxyhemoglobin levels.
Rationale for Using Carbon Monoxide
Carbon monoxide (CO) is selected for measuring diffusing capacity due to its exceptionally high affinity for hemoglobin, which is approximately 210 to 250 times greater than that of oxygen. This strong binding ensures that CO uptake is primarily limited by the diffusion properties of the alveolar-capillary membrane rather than by pulmonary blood flow or hemoglobin saturation levels, allowing for a more isolated assessment of the lung's membrane diffusing capacity.[3] Additionally, CO exhibits low solubility in plasma, with a Bunsen solubility coefficient of 0.018 mL gas per mL plasma per atm, which minimizes the amount dissolved in the blood and emphasizes the role of trans-membrane diffusion over simple dissolution. Endogenous CO levels in the pulmonary capillaries are negligible, typically near zero partial pressure, creating a driving pressure gradient that approximates the inspired CO partial pressure minus zero, thereby simplifying the calculation of diffusing capacity without interference from back pressure.[16][4] In contrast, using oxygen directly for such measurements is complicated by the variable saturation of hemoglobin along the capillary, which depends on local oxygen partial pressure and can confound the assessment of pure diffusion capacity. CO avoids this issue by maintaining near-complete binding due to its affinity, providing a purer evaluation of membrane conductance as a surrogate for oxygen transfer.[3] To mitigate potential toxicity, the test employs a low CO concentration of 0.3% in the inspired gas mixture, resulting in a minimal increase in carboxyhemoglobin levels, typically less than 0.5% per test, which is safe and well below thresholds associated with physiological impairment.[1][4]Influencing Factors
Factors That Decrease DLCO
Several pathological conditions and physiological states can lead to a reduction in diffusing capacity for carbon monoxide (DLCO), primarily by impairing the alveolar-capillary membrane (Dm), pulmonary capillary blood volume (Vc), or the reaction rate of CO with hemoglobin (θ).[1] In emphysema, a key component of chronic obstructive pulmonary disease (COPD), the destruction of alveolar walls results in a loss of surface area available for gas exchange, directly reducing Dm and overall DLCO. This structural damage limits the effective diffusion pathway, with DLCO often markedly reduced in moderate to severe cases compared to predicted values.[17] Interstitial lung diseases (ILDs), such as idiopathic pulmonary fibrosis, cause thickening of the alveolar-capillary membrane due to inflammation and fibrosis, which increases the diffusion distance for gases and impairs Dm. This leads to a progressive decline in DLCO, serving as an early marker of disease severity.[1][17] Anemia reduces θ, the rate at which CO binds to hemoglobin in red blood cells, because lower hemoglobin concentrations decrease the blood's capacity to carry CO, thereby lowering measured DLCO. This does not reflect intrinsic lung impairment, and correction formulas adjust raw DLCO values for hemoglobin levels to estimate true pulmonary diffusing capacity.[18][12] Pulmonary hypertension diminishes Vc through vascular remodeling, vasoconstriction, and capillary destruction, which reduces the effective blood volume participating in gas exchange and thus decreases DLCO. This effect is particularly pronounced in pulmonary arterial hypertension, where capillary bed loss correlates with the severity of right ventricular strain.[18][19] Other factors include transient reductions during high-altitude exposure, where hypocapnia from hyperventilation and acute hypoxia can impair diffusion dynamics, leading to a measurable drop in DLCO. Additionally, post-exercise fatigue causes a temporary 10-15% decrease in DLCO, attributed to pulmonary interstitial edema or reduced capillary recruitment following strenuous activity.[20][21]Factors That Increase DLCO
Polycythemia, characterized by an elevated red blood cell count and hemoglobin concentration, increases diffusing capacity for carbon monoxide (DLCO) primarily by enhancing the rate of CO binding to hemoglobin, denoted as θ in the Roughton-Forster equation, which directly elevates the θVc component where Vc represents pulmonary capillary blood volume.[1] This physiological adaptation improves the overall capacity for gas transfer across the alveolar-capillary membrane, with studies reporting elevations up to approximately 20% in affected individuals compared to normal hemoglobin levels.[22] In endurance-trained athletes, regular exercise training leads to structural adaptations in the pulmonary vasculature, including an expanded pulmonary capillary blood volume (Vc), which contributes to a 10-15% higher DLCO at rest and during submaximal exercise intensities compared to sedentary individuals.[23] These changes arise from increased capillary recruitment and density in the lungs, optimizing gas exchange efficiency without significant alterations in membrane diffusing capacity (DM) at baseline.[24] The enhanced Vc allows for greater blood flow through the alveolar region, facilitating improved CO uptake and reflecting the lungs' adaptation to sustained high cardiac outputs during training.[25] Mild asthma can result in supranormal DLCO values, often exceeding 120% of predicted norms, due to hyperinflation that enlarges alveolar surface area and promotes capillary recruitment through inflammatory processes.[26] This hyperinflation, stemming from air trapping and increased lung volumes, effectively increases the effective diffusion surface for CO, distinguishing mild asthma from other obstructive diseases where DLCO is typically reduced.[27] The associated mild inflammation further enhances pulmonary blood flow, amplifying the transfer factor without compromising overall ventilatory function.[28] Left-to-right intracardiac shunts, such as those in atrial or ventricular septal defects, elevate DLCO by augmenting pulmonary blood flow, which distends and recruits additional pulmonary capillaries, thereby increasing Vc and the rate constant Kco (DLCO/VA).[3] This hyperkinetic circulation redistributes blood volume more evenly across the lung fields, reducing diffusion resistance and boosting overall gas exchange capacity, often resulting in DLCO values well above predicted levels.[29] The mechanism aligns with the Roughton-Forster model, where heightened Vc lowers the 1/θVc term, enhancing total DLCO.[3] The supine position increases DLCO by 10-20% compared to the sitting posture in healthy individuals, primarily due to gravity-dependent pooling of blood in the pulmonary vasculature, which expands Vc particularly in dependent lung regions.[30] This postural shift improves alveolar-capillary contact and uniformizes blood flow distribution, minimizing ventilation-perfusion mismatches that occur in upright positions.[31] In endurance-trained athletes, the effect is more pronounced at rest, reflecting their greater baseline vascular compliance.[31]Clinical Interpretation
Normal Values and Variability
The diffusing capacity of the lung for carbon monoxide (DLCO) in healthy adults typically ranges from 20 to 30 mL/min/mmHg, with predicted values higher in men (mean approximately 28 mL/min/mmHg) than in women (mean approximately 21 mL/min/mmHg) when adjusted for age and height.[32] These reference values are derived from large cohorts using standardized single-breath techniques and vary based on demographic factors, with equations incorporating age, sex, and height to generate individual predictions.[4] DLCO declines gradually with age, at a rate of approximately 2-3% per decade after age 30, reflecting physiological changes in alveolar surface area and pulmonary capillary blood volume.[33] Variability in DLCO measurements is inherent, with intra-subject reproducibility typically within 5-10% on repeat testing under controlled conditions, while inter-laboratory differences can reach up to 15% due to equipment and procedural variations.[4] Natural variation is also influenced by factors such as sex, height, and ethnicity; for instance, African Americans exhibit predicted DLCO values about 10-15% lower than Caucasians of similar age, height, and sex, necessitating ethnicity-specific reference equations. The GLI-2017 equations are derived from Caucasian data, with adjustments recommended for other ethnicities, and ongoing research develops multi-ethnic references.[34][4] In clinical practice, modern interpretation employs z-scores to assess deviation from predicted norms, with the lower limit of normal (LLN) defined as 1.64 standard deviations below the mean, corresponding to the 5th percentile in healthy populations and aiding in the identification of abnormality.[4] Adjustments are routinely applied to raw DLCO measurements for hemoglobin concentration (using sex-specific equations to account for anemia effects), carboxyhemoglobin levels (correcting for back-pressure, particularly in smokers, at ~1% reduction per 1% COHb), and alveolar volume (VA) to normalize for lung size.[4] The Global Lung Function Initiative (GLI-2017) provides comprehensive reference equations incorporating these adjustments.| Age Group (years) | Predicted DLCO (mL/min/mmHg) - Males (height 175 cm) | Predicted DLCO (mL/min/mmHg) - Females (height 165 cm) |
|---|---|---|
| 20-29 | 30-31 | 23-24 |
| 30-39 | 29-30 | 22-23 |
| 40-49 | 28-29 | 21-22 |
| 50-59 | 27-28 | 20-21 |
| 60+ | 25-27 | 19-21 |