Diffusing capacity for carbon monoxide

INTRODUCTION — A test of the diffusing capacity of the lungs for carbon monoxide (DLCO, also known as transfer factor for carbon monoxide or TLCO), is one of the most clinically valuable tests of lung function. The technique was first described 100 years ago [1-3] and applied in clinical settings many decades later [4-6]. The DLCO measures the ability of the lungs to transfer gas from inhaled air to the red blood cells in pulmonary capillaries. The DLCO test is convenient and easy for the patient to perform. The ten seconds of breathholding required for the DLCO maneuver is easier for most patients to perform than the forced exhalation required for spirometry.

Standards for DLCO instruments, performance of the test, and calculation of the results were initially published by the American Thoracic Society in 1987, and updated by the American Thoracic Society and European Respiratory Society in 2005 and in 2017 [7-9]. The indications for DLCO measurement and the interpretation of the results will be discussed here. The use of other pulmonary function tests in the evaluation of respiratory disease in adults and children is discussed separately.

●(See "Office spirometry".)

●(See "Overview of pulmonary function testing in adults".)

●(See "Overview of pulmonary function testing in children".)

DEFINITIONS

●DLCO – The diffusing capacity for carbon monoxide (DLCO) is also known as the transfer factor for carbon monoxide or TLCO. It is a measure of the conductance of gas transfer from inspired gas to the red blood cells.

●VA – The alveolar volume (VA) can be considered the number of contributing alveolar units and is measured during the single breath DLCO by use of a tracer gas (eg, helium).

●KCO – The carbon monoxide transfer coefficient (KCO is approximately kCO/barometric pressure in mL/minute/ mmHg/L) is often written as DLCO/VA. It is an index of the efficiency of alveolar transfer of carbon monoxide.

●kCO – The permeability factor (kCO, minute^-1) is the rate constant for alveolar-capillary CO transfer (the logarithmic rate of decay of the alveolar CO concentration).

PHYSIOLOGY — The diffusing capacity of the lungs for carbon monoxide (DLCO) is designed to reflect properties of the alveolar-capillary membrane, specifically the ease with which oxygen moves from inhaled air to the red blood cells in the pulmonary capillaries. The DLCO can be affected by factors that change the membrane properties and also by changes in hemoglobin and capillary blood volume. The process of carbon monoxide (CO) uptake can be simplified into two components: membrane conductance (D_M) and reactive conductance, the chemical reaction between CO and hemoglobin. D_M reflects the properties of diffusion across the alveolar capillary interface. Uptake of CO by hemoglobin depends on the reaction rate (R) and the pulmonary capillary blood volume (Vc). The two components occur in series and when conductances occur in series, they are added as reciprocals. This relationship can be expressed as:

Diseases in which the uptake of oxygen is reduced cause parallel decreases in the uptake of CO, as measured by the DLCO. Older textbooks suggest that thickening of the alveolar-capillary membrane (in interstitial lung disease) and loss of alveolar membrane surface area (in emphysema) are the primary causes of a low DLCO. However, subsequent experimental data suggest these and most other diseases that influence the DLCO do so by reducing the volume of red blood cells in the pulmonary capillaries. The total volume of blood in the lungs in healthy adults at rest is less than 150 mL. Diseases in which the alveolar-capillary surface area is reduced (eg, idiopathic pulmonary fibrosis and emphysema) lead to a reduction in the blood volume in the lungs.

The volume of blood in the pulmonary capillaries and the DLCO are increased in the following circumstances:

●When pulmonary capillaries are recruited, as occurs during exercise

●When the patient is in the supine position

●During a Mueller (reverse Valsalva) maneuver

●When a left-to-right cardiac shunt is present

INDICATIONS — There are multiple clinical indications for measurement of diffusing capacity of the lungs for carbon monoxide (DLCO), and there are no contraindications or adverse effects (table 1) [10]. In general, the DLCO is used to identify the cause of dyspnea or hypoxemia, monitor disease progression in interstitial lung disease, and identify pulmonary hypertension in patients at risk, such as those with systemic sclerosis.

Obstructive disease

●The DLCO is an excellent index of the degree of anatomic emphysema in smokers with airways obstruction. A low DLCO correlates highly with a low mean density of lung tissue on lung CT scan and with the degree of anatomic emphysema [11-13].

●Patients with COPD and disproportionately low DLCO values are at increased risk for group 3 pulmonary hypertension [14]. The severity of the DLCO impairment is associated with increased mortality [15].

●Smokers with airways obstruction but normal DLCO values usually have chronic "obstructive" bronchitis but not emphysema.

●Patients with airway obstruction from asthma typically have normal or high DLCO values [16].

●Patients with cystic fibrosis have a normal DLCO until their disease becomes very severe [17,18].

●Patients with bronchiolitis obliterans usually have a reduced DLCO and may have airflow limitation on spirometry.

Low DLCO values have been associated with increased symptoms, limitations in exercise capacity, increased risk of exacerbation, and increased risk of death in COPD [19-23]. (See "Chronic obstructive pulmonary disease: Diagnosis and staging".)

Restrictive disease — The DLCO helps in the differential diagnosis of restrictive lung disease, which is identified by reduced total lung capacity (TLC) and vital capacity (VC). A low DLCO combined with reduced lung volumes suggests interstitial lung disease (ILD) [24,25]. A normal DLCO associated with low volumes is consistent with an extrapulmonary cause of the restriction, such as obesity, pleural effusion or thickening, neuromuscular weakness, or kyphoscoliosis, although clinical evaluation and chest imaging are needed for confirmation.

Another common application of the DLCO is for detection of mild (early or preclinical) interstitial lung disease in high-risk patients, including those with [24,26-33]:

●Sarcoidosis

●Hypersensitivity pneumonitis (extrinsic allergic alveolitis)

●Chest irradiation and cancer chemotherapy

●Use of drugs known to have pulmonary toxicity (eg, amiodarone, bleomycin, nitrofurantoin)

●Rheumatic disease (eg, systemic sclerosis)

The normal range for DLCO is wide; as a result, it is much more sensitive to obtain a baseline DLCO test prior to therapy and subsequently perform a follow-up examination. Changes in follow-up values of DLCO in patients being treated for interstitial lung disease are often more sensitive to improvement or worsening than are changes in TLC or VC [34].

Pulmonary vascular disease — An abnormal DLCO may be due to pulmonary vascular disease in those patients with chronic dyspnea but normal spirometry and lung volumes. Reduction in the DLCO is an indicator of [35-42]:

●Chronic recurrent pulmonary emboli or chronic thromboembolic pulmonary hypertension (see "Epidemiology, pathogenesis, clinical manifestations and diagnosis of chronic thromboembolic pulmonary hypertension", section on 'Clinical features')

●Idiopathic pulmonary arterial hypertension (see "Clinical features and diagnosis of pulmonary hypertension of unclear etiology in adults")

●Pulmonary vascular involvement from rheumatic diseases and vasculitides (eg, systemic sclerosis, systemic lupus erythematosus, mixed connective tissue disease) (see "Overview of pulmonary complications of systemic sclerosis (scleroderma)", section on 'Pulmonary function testing' and "Mixed connective tissue disease", section on 'Pulmonary arterial hypertension')

●Heart failure (see "Heart failure: Clinical manifestations and diagnosis in adults")

Prior to lung resection surgery — The DLCO can be helpful in predicting patients at higher risk for postoperative complications following lung resection for lung cancer and lung volume reduction surgery (LVRS). A very low DLCO increases the risk of postoperative morbidity and mortality in patients with lung cancer, although there is no consensus regarding the threshold below which surgery is contraindicated [43-45]. A low DLCO (≤20 percent predicted) suggests diffuse homogeneous emphysema and has been identified as a major factor for poor outcomes following LVRS [46,47]. (See "Preoperative physiologic pulmonary evaluation for lung resection" and "Lung volume reduction surgery in COPD", section on 'Primary endpoints'.)

Disability evaluation — Measurement of the DLCO is also used for disability evaluation in patients with severe COPD or interstitial lung disease. A DLCO below 40 percent predicted (or less than 9 mL/min per mmHg) may qualify a patient for total disability according to Social Security standards, whereas American Medical Association and American Thoracic Society guidelines use a threshold value of 45 percent predicted. (See "Evaluation of pulmonary disability".)

Need for oxygen therapy — A low DLCO (eg, ≤50 percent of predicted) is the major predictor of oxygen desaturation during exercise, so a DLCO test may be worthwhile as a screening test in patients presenting with dyspnea on exertion [48-50]. However, a normal DLCO does not exclude the possibility of desaturation on exertion [48], and demonstration of arterial oxygen desaturation, using either pulse oximetry or arterial blood gas analysis, is required prior to initiating long-term oxygen therapy. (See "Long-term supplemental oxygen therapy".)

METHODOLOGY — Almost all pulmonary function laboratories use a single-breath method, in which the patient is seated, has a nose clip and mouthpiece in place, and is instructed to adhere to tidal volume breathing without deep inspirations until instructed.

Preparation — In preparation for DLCO testing, patients should adhere to the following [8]:

●No cigarette smoking on the day of the test (or note the timing of the last cigarette smoked).

●No supplemental oxygen for at least 15 minutes prior to and during the test: use of supplemental oxygen can decrease DLCO by approximately 0.35 percent per mmHg change in arterial oxygen tension (PaO₂). The DLCO test cannot be performed in patients who are unable to discontinue supplemental oxygen for at least 15 minutes.

●DLCO maneuvers are frequently performed after administration of short-acting bronchodilators in the interval between pre and post-bronchodilator spirometry [51,52]. Studies have demonstrated that albuterol (salbutamol) has no significant effect on DLCO in normal control participants or in patients with airflow obstruction, including patients with COPD. The updated standards no longer recommend against use of bronchodilators before DLCO testing [9].

DLCO maneuver — The single breath DLCO maneuver begins with a full exhalation to residual volume (RV), the mouthpiece is connected to the test gas (0.3 percent carbon monoxide [CO], tracer gas [eg, 10 percent helium or 0.3 percent methane], oxygen, and nitrogen), and the subject inhales rapidly to total lung capacity in <4 seconds (figure 1) [8,53]. Following a 10±2 second breath hold, the subject exhales quickly and completely to RV. An alveolar sample of the exhaled gas is collected immediately following dead space washout and analyzed for calculation of the dilution of the tracer gas and the uptake of CO.

Most DLCO instruments are unable to measure the DLCO when the patient's vital capacity is less than approximately 1.5 L. While newer equipment with real time gas analyzers may provide a means to measure DLCO in patients with lower vital capacities, results may be less reliable for very low vital capacities. (See 'Future directions' below.)

Calculation of DLCO and VA — DLCO testing equipment calculates the DLCO and VA from the measured volumes and fractions of inspired and expired CO and tracer gas (such as helium) during the DLCO maneuver. While modern PFT equipment performs these calculations, understanding the individual components and their relationships can help in interpreting the results. The following equations describe how the measurements obtained during a single breath DLCO test are used to calculate the DLCO [54].

During breath holding, CO leaves alveolar gas at an exponential rate. The rate constant kCO is the measured logarithmic change in CO concentration per minute. VA is the alveolar volume (accessible during a 10 second breath hold), which is calculated by knowing the fractional concentration of the tracer gas in the inhaled and exhaled gas and also volume of gas inhaled.

The VA is determined from the dilution of the known volume and fractional concentration of the tracer gas using the relationship, F₁V₁ = F₂V₂, which can be solved for V₂ = V₁F₁/F₂.

VA = V_I x (F_I tracer/F_A tracer)

●F_A tracer = alveolar (exhaled) fraction of tracer gas (eg, helium)

●F_I tracer = inspired fraction of tracer gas (eg, helium)

●V_I = volume of inspired gas

The product of VA and the rate constant kCO yields V_CO, which is the uptake of CO:

VA x kCO = V_CO (mL/min)

Dividing both sides of the equation by P_B (barometric pressure – water pressure) gives the DLCO:

(VA x kCO)/(P_B-P_H2O) = V_CO/P_B-P_H2O = DLCO mL/min per mmHg

As KCO = kCO/(P_B-P_H2O), the relationship can be simplified to:

DLCO (gas exchange capacity) = KCO (efficiency per lung unit) x VA (number of lung units).

The KCO (carbon monoxide transfer coefficient) is reported as DLCO/VA on many PFT reports. This is because the units of KCO are units representing diffusion (mL/min) per unit of volume (liter). It is important to note that the DLCO/VA does not represent a correction factor for alveolar volume as the relationship is not independent of lung volume. Understanding the relationships between KCO, VA, and DLCO can aid in interpreting test results.

Quality of testing — In 2017, the European Respiratory Society and American Thoracic Society published revised standards for single-breath carbon monoxide uptake in the lung [8,9]. The major changes in the updated standards address systems using rapidly responding gas analyzers for carbon monoxide and the tracer gas.  

The criteria for an acceptable DLCO maneuver have also been revised: the inspiratory volume (VI) should be ≥90 percent of the largest measurement for vital capacity (VC). A maneuver may be deemed acceptable if the VI is <90, but ≥85 percent of the largest VC and the alveolar volume (VA) is within 200 mL or 5 percent (whichever is greater) of the highest VA among acceptable maneuvers [9]. At least 85 percent of the test gas VI must be inhaled in <4 seconds. The calculated breath-hold time should be 10±2 seconds. There should be no evidence of a Mueller or Valsalva maneuver based on observation of the patient during the maneuver and review of the mouth pressure tracing.

In classic systems, the first 0.75 to 1 L is discarded as dead space gas and then a sample gas volume of 0.5 to 1 L is collected for analysis. In newer systems that use real-time gas analyzers, the standards require measurement of dead space washout; collection of a sample as low as 85 mL can be analyzed immediately after the dead space washout. After a minimum of four minutes, the test is repeated; a longer interval may be needed if the patient has obstructive airways disease.

To meet the criteria for repeatability, at least two acceptable tests must be within 2 mL/min per mmHg (0.67 mmol/min per kPa) of each other [9]. Not more than five tests should be done in the course of one session, as five tests can decrease the measured DLCO by 3 to 3.5 percent [9]. The updated standards provide an algorithm to grade the quality of the test.

The American Thoracic Society recommends expressing the DLCO in mL (at standard temperature, pressure, and dry [STPD])/min per mmHg, while the European Respiratory Society uses SI units mmol/min per kPa [8]. Values in SI units are multiplied by 2.987 to obtain values in traditional units.

ADJUSTMENTS — The diffusing capacity of the lungs for carbon monoxide (DLCO) changes with anemia, carboxyhemoglobin levels, altitude, and lung volume, so adjustments for variances in these factors may be needed prior to interpretation.

Anemia — Anemia, with the resultant decrease in carbon monoxide carrying capacity, reduces the DLCO. As an example, loss of blood from a healthy adult male leading to a decrease in hemoglobin from 16 down to 8 g/dL results in an approximately 30 percent decline in DLCO. This would be reported as a mildly reduced DLCO. Therefore, if the patient is at risk for anemia or known to have anemia, the laboratory should obtain a recent hemoglobin value in order to report both the standard and adjusted percent predicted DLCO (calculator 1 and calculator 2) [8,9].

The hemoglobin adjustment estimates a value for percent predicted DLCO that would be comparable to transfusion of the patient to a normal hemoglobin (14.6 g/dL for adult men and 13.4 g/dL for women and for children under age 15). The DLCO falls about 9 percent during menstruation, although this effect is not completely explained by changes in hemoglobin concentration [55].

Carboxyhemoglobin and cigarette smoking — The carboxyhemoglobin (COHb) level may be elevated in the blood if the patient was smoking just prior to the DLCO measurement [56,57]. An increase of 1 percent in COHb results in a proportionate 1 percent decrease in the measured DLCO. Most laboratories merely ask the patient to refrain from smoking for four hours prior to the test, but do not measure COHb or adjust the measured DLCO for the small effect of an increased CO "back pressure" [57]. Smoking cessation results in a mean DLCO increase of 2 to 4 mL/min per mmHg within a few days [58].

High altitude — If the laboratory is located at high altitude, the ambient, alveolar, and arterial oxygen concentrations are lower than at sea level. The lower arterial oxygen concentration results in less competition for CO binding to hemoglobin, increased CO uptake, and higher measured DLCO compared to a test done at sea level.

Patients living at high altitude also have slightly higher hemoglobin levels, resulting in increased DLCO values. In order to correct for these effects, laboratories at higher altitudes may choose to use the DLCO reference equations from the study of Crapo and coworkers, which was performed at high altitude [59].

Volume correction — In the past, the term DLCO/VA (also known as KCO) was misinterpreted as a correction factor for low lung volume, leading to potential misinterpretation of DLCO results [8,9,60]. The nonlinear relationship between KCO and lung volume precludes KCO from being a correction factor for DLCO when lung volumes are reduced. Instead, DLCO/VA (KCO) reflects alveolar CO uptake efficiency at a given volume [54]. (See 'Calculation of DLCO and VA' above.)

DLCO/VA and DLCO are affected by changes in lung volume (eg, DLCO falls and KCO rises as lung volume becomes smaller in healthy subjects) [60]. However, applying DLCO/VA as the correction factor is a misinterpretation of this value. Instead, it is important to understand the relationship between the measured values KCO and VA and how these may provide insight as to the underlying physiologic state and disease process [8,54,60-63]. Incomplete alveolar expansion, diffuse versus localized loss of alveolar units, and poor alveolar mixing are some of the factors that warrant consideration. This is described in greater detail below. (See 'Relationship between DLCO and KCO (DLCO/VA)' below.)

Nonetheless, the average VA of the tests used to generate the reported DLCO should be reported. In addition, the volume inspired (VI) should be >90 percent of the largest vital capacity (VC) to show that any reduction in VA is not due to poor inspiratory effort [8]. (See 'Methodology' above.)

INTERPRETATION — Without a previous diffusing capacity of the lungs for carbon monoxide (DLCO) test, the patient's result is interpreted by comparing it with the normal distribution using the z-score, which is more consistent across age and sex compared with percent of predicted [64]. Using the z-score, threshold values for elevated, mild, moderate, and severe disease have been recommended as a replacement for percent of predicted (table 2). For DLCO values that are close to the lower limit of the normal range (eg. z-score -1.5 to -1.645 or between 75 and 80 percent of predicted), the correlation with the presence or absence of clinical disease is less well-defined. For reports that have not yet changed to using z-scores or are in the process of transitioning and present both interpretation strategies, an example of severity classification using percent of predicted is also provided (table 2). The difference between the percent predicted approach and the z-score approach is most evident among older individuals, who may be classified with moderate-severe defects using the percent predicted approach despite falling only slightly outside the normal range. (See 'Reference equations' below and "Selecting reference values for pulmonary function tests" and "Evaluation of pulmonary disability", section on 'Pulmonary function testing'.)

Disorders associated with an abnormal DLCO are summarized in the table (table 3).

Low DLCO due to cigarette smoking — The DLCO is substantially lower in current smokers than in nonsmokers, a factor that may be important to integrate into clinical interpretations, such as when estimating the degree of impairment secondary to asbestos exposure. Since the reduction of DLCO partially resolves after smoking cessation, the reduction of DLCO in current smokers cannot be attributed entirely to emphysema and can be influenced by increased carboxyhemoglobin levels [56,58,65]. (See 'Carboxyhemoglobin and cigarette smoking' above.)

Low DLCO with obstruction — In long-term cigarette smokers, a low DLCO with airways obstruction is usually due to emphysema. The subsequent rate of decline in FEV₁ is predicted by the degree of airflow obstruction and airway hyperreactivity [66]. There is little evidence that a lower DLCO predicts increased morbidity or mortality from COPD (for a given baseline FEV₁ and degree of airway hyperreactivity).

In COPD, the loss of DLCO usually occurs after the decline in FEV₁, so that the percent predicted DLCO is usually higher than the percent predicted FEV₁. If the DLCO is severely reduced with only mild airflow obstruction, causes of a low DLCO other than emphysema should be considered.

Cystic fibrosis and alpha-1 antitrypsin deficiency should be considered in children and young adults with obstruction and a low DLCO. This pattern can also be seen in adults with bronchiolitis obliterans, bronchiectasis, and lymphangioleiomyomatosis [67].

Low DLCO with restriction — A reduced DLCO in patients with low lung volumes, indicating a restrictive pattern, suggests interstitial lung disease or pneumonitis. Mild airflow obstruction may be superimposed on the restriction, due to narrowing of the terminal airways by the interstitial process. This pattern of low DLCO plus both restriction and "small airways" obstruction is also seen in:

●Sarcoidosis (stage II through IV)

●Asbestosis

●Miliary tuberculosis

●Heart failure

In patients with heart failure, restriction and low DLCO are due to alveolar filling, while mild obstruction can be due to edema of small airways.

Normal DLCO with restriction — Among patients with a restrictive ventilatory defect, a normal DLCO suggests an extrapulmonary cause of the restriction, such as pleural effusion, pleural thickening, neuromuscular weakness [68], or kyphoscoliosis.

Low DLCO with normal spirometry — Disorders to consider with an isolated decrease in DLCO include:

●Pulmonary vascular disease (mild to severe decrease in DLCO), such as chronic recurrent pulmonary emboli, idiopathic pulmonary arterial hypertension, and pulmonary vascular involvement with rheumatic diseases and vasculitides (see "Treatment and prognosis of pulmonary arterial hypertension in adults (group 1)")

●Early interstitial lung disease (eg, mild to moderate decrease in DLCO before the vital capacity falls below the lower limit of the normal range)

●Anemia (see 'Anemia' above)

●Hepatopulmonary syndrome [69,70] (see "Hepatopulmonary syndrome in adults: Prevalence, causes, clinical manifestations, and diagnosis")

●An increased carboxyhemoglobin level due to cigarette smoking (see 'Carboxyhemoglobin and cigarette smoking' above)

Increased DLCO — Disorders to consider when the DLCO is near or above the upper limit of the normal range include the following [16,71,72]:

●Obesity

●Asthma

●High altitude

●Polycythemia

●Pulmonary hemorrhage

●Left-to-right intracardiac shunting

●Mild left heart failure (due to increased pulmonary capillary blood volume)

●Exercise just prior to the test session (due to increased cardiac output)

●Supine position; Mueller maneuver (inhalation against closed glottis decreases intrathoracic pressure and increases blood return to the lungs)

Of course, laboratory errors, such as entry of a low height or age, a leak around the mouthpiece, or an inaccurate CO analyzer, should also be considered. (See 'Quality control' below.)

Relationship between DLCO and KCO (DLCO/VA) — The DLCO is the product of the measured values for the KCO (carbon monoxide transfer coefficient) and the VA (alveolar volume). The KCO is commonly reported as the DLCO/VA in PFT reports [54,60-62]. The following examples demonstrate the effect of certain clinical situations on KCO (DLCO/VA) results (table 4):

●Incomplete lung expansion – In patients who have neuromuscular disorders, kyphoscoliosis, or inadequate inspiration due to poor test performance, the KCO (DLCO/VA) is elevated [68]. The increase in KCO is due to an increase in the surface to volume ratio for diffusion per alveolus as the alveoli become smaller because of the relationship of higher surface area to volume at lower lung volumes. The reduction in DLCO is proportionately less than the reduction in VA, so the ratio DLCO/VA is actually increased.

●Pneumonectomy – For patients who have undergone pneumonectomy, but do not have lung disease in the remaining lung, the VA is decreased due to discrete loss of alveolar units. Blood flow is diverted to the remaining lung and the KCO (DLCO/VA) is usually increased. As a result, the DLCO is slightly decreased.

●Emphysema – In emphysema, the DLCO is reduced by loss of gas exchanging surface due to alveolar capillary damage and the KCO is low. While actual lung volumes are often normal or supranormal, the measured VA may be low due to incomplete mixing of gases during the breath hold time. Incomplete mixing of gases may be indicated by a reduced VA/total lung capacity (TLC) ratio (normal >85 percent).

●Interstitial lung disease (ILD) – In ILD, the DLCO is decreased by diffuse alveolar capillary damage. The VA is low due to the loss of aerated alveoli. The KCO (DLCO/VA) is often reduced to a lesser extent than the DLCO as ILD is typically inhomogeneous with some diversion of blood flow from more diseased units to those that are less affected [73]. This circumstance can be misleading if the KCO is interpreted as DLCO "corrected for alveolar volume" and clinicians are falsely reassured by the "normal" KCO value.

●Pulmonary vascular disease – In pulmonary hypertension, the DLCO is reduced. The VA is typically normal, and the KCO (DLCO/VA) is reduced due to impairment at the alveolar-capillary interface.

USING DLCO TO MONITOR DISEASE COURSE — DLCO can also be used to monitor the course of disease or response to treatment. If a patient had a previous DLCO test of high quality in the same laboratory, the change in DLCO from the baseline or most recent DLCO (rather than the percent predicted) should be used to determine progression or regression of disease [74]. In pulmonary function laboratories with excellent quality control procedures, a change in absolute value greater than 4 mL/min per mmHg is outside the range of measurement "noise" (ie, the short-term, within-subject variability). (See 'Quality control' below.)

QUALITY CONTROL — Achieving consistent results between and within laboratories remains a difficult problem. When healthy subjects are tested in different laboratories, the results (both absolute diffusing capacity of the lungs for carbon monoxide [DLCO] and percent predicted) can vary widely [75,76]. Even when tested in the same laboratory a few days later, DLCO results from healthy subjects may vary as much as 8 mL/min per mmHg, although variance from the best laboratories is about 2 to 4 mL/min per mmHg [77,78]. The short-term variability of DLCO in patients with lung disease (without interventions) is likely to be somewhat larger than that of healthy "biologic controls."

Each laboratory should test the DLCO of at least one healthy technician weekly, and compare it to their previous results. When interpreting the significance of a reported change in DLCO, it is prudent to ask the laboratory for recent results of repeatability studies.

Standards committees from the American Thoracic Society and the European Respiratory Society have published DLCO instrument and test methodology standards [8,9], but no third-party testing of commercially available DLCO instruments has been done to see which models meet these standards. DLCO results are very sensitive to errors in carbon monoxide (CO) analyzers, which drift over time if not closely maintained. In 2003, an advance in DLCO quality was made with the commercial availability of a device to validate DLCO results using calibration gases, syringes, valves, and software [79,80].

REFERENCE EQUATIONS — A variety of reference equations have been used to calculate predicted values for DLCO, resulting in wide variations in the predicted DLCO for the same patient tested in different laboratories. The Global Lung Initiative published reference values for DLCO based on population data from over 12,000 asymptomatic, lifelong nonsmokers from White populations from Australia, New Zealand, the Netherlands, Bulgaria, the United States, Canada, Mexico, Spain, Greece, and Italy [81-85]. A limitation of these reference values is the lack of data from individuals who are not from these countries; the extent to which racial and ethnic differences impact DLCO is unclear.

FUTURE DIRECTIONS — Manufacturers of pulmonary function testing equipment have developed technology to incorporate real time gas analyzers. Advantages of this capability include continuous measurement of carbon monoxide (CO) and the tracer gas (eg, methane) and also the use of small sample volumes [86]. This technology may allow for visual assessment of dead space clearance to ensure alveolar gas sampling and provide a means to measure DLCO among patients with lower vital capacities [8,9,86].

Over the past several decades, measurement of pulmonary diffusing capacity using nitric oxide (DLNO) has been introduced to provide additional information about the alveolar membrane diffusing capacity [87,88]. Since NO has a much faster rate of reaction with hemoglobin and the red blood cell resistance approaches zero, DLNO remains unaffected by fluctuations in pulmonary capillary blood volume and reflects the alveolar-capillary membrane diffusing capacity. The DLNO/DLCO ratio can be measured in a single maneuver and is inversely related to the thicknesses of the alveolar membrane and capillary sheet [87,89]. At present, this measurement is not widely available in clinical laboratories and is mostly used in research settings.

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Pulmonary function testing".)

SUMMARY AND RECOMMENDATIONS

●Physiology – The diffusing capacity of the lungs for carbon monoxide (DLCO) measures the ability of the lungs to transfer gas from inhaled air in the alveoli to the red blood cells in pulmonary capillaries. (See 'Physiology' above.)

●Indications – There are multiple clinical indications for measurement of DLCO (table 1). In general, the DLCO is used to identify the cause of dyspnea or hypoxemia, monitor disease progression in interstitial lung disease, and identify pulmonary hypertension in patients at risk, such as those with systemic sclerosis. (See 'Indications' above.)

●Preparation – In preparation for DLCO measurement, patients should avoid cigarette smoking on the day of the test (or note the timing of the last cigarette smoked) and discontinue any supplemental oxygen for at least 15 minutes prior to testing. (See 'Preparation' above.)

●Performance

•Method – The DLCO maneuver begins with a full exhalation to residual volume (RV), the mouthpiece is then connected to the test gas (eg, 0.3 percent carbon monoxide [CO], 10 percent tracer gas [eg, helium]), and the subject inhales rapidly to total lung capacity (figure 1). Following a 10±2 second breath hold, the subject exhales quickly and completely to RV. An alveolar sample of the exhaled gas is then analyzed for calculation of the dilution of tracer gas and uptake of CO. (See 'Methodology' above.)

•Quality control – When assessing the adequacy of the inspiratory effort, the inspired volume (VI) should be ≥90 percent of the largest measurement for vital capacity (VC). A maneuver may also be considered acceptable if the VI is within 85 percent of the largest VC and the VA is within 200 mL or 5 percent (whichever greater) of the largest VA from other acceptable maneuvers (figure 1). (See 'Volume correction' above.)

●Interpretation – The causes and severity of increases and decreases in DLCO are listed in the tables (table 2 and table 3). (See 'Interpretation' above.)

●Diagnostic utility – Measurement of the DLCO is useful in a number of clinical settings (table 1):

•Obstructive diseases – The DLCO can narrow the differential diagnosis of airway obstruction that persists after inhaling a bronchodilator. The DLCO is decreased in patients with emphysema, bronchiolitis obliterans, and lymphangioleiomyomatosis, whereas it tends to be normal or high in patients with asthma. (See 'Obstructive disease' above.)

•Restrictive diseases – The DLCO can also narrow the differential diagnosis of restrictive lung disease. A low DLCO combined with reduced lung volumes suggests interstitial lung disease. In contrast, a normal DLCO associated with low lung volumes suggests an extrapulmonary cause of the restriction, such as pleural effusion, pleural thickening, neuromuscular weakness, or kyphoscoliosis. (See 'Restrictive disease' above.)

•Pulmonary vascular disease – Disorders to consider when an isolated decrease in the DLCO is detected include the pulmonary vascular diseases (eg, pulmonary hypertension, thromboembolic disease), early interstitial lung disease, anemia, hepatopulmonary syndrome, and carboxyhemoglobinemia. The predicted DLCO can be adjusted for the presence of anemia (calculator 1 and calculator 2). (See 'Low DLCO with normal spirometry' above.)

●Use in disease management

•Objective measure of COPD severity – Among patients with COPD, a low DLCO predicts oxygen desaturation during exercise and long-term outcome of lung volume reduction surgery. (See 'Need for oxygen therapy' above and 'Prior to lung resection surgery' above.)

•Monitoring disease course – A change in the DLCO is a good index of disease progression or response to therapy. A change greater than 4 mL/min per mmHg is likely a clinically significant difference in pulmonary function labs with policies and procedures to maintain consistency of results. (See 'Using DLCO to monitor disease course' above.)

ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges Dr. Paul Enright, who contributed to earlier versions of this topic review.