Gastrointestinal Monitoring Using Measurement of Intramucosal PCO2

The gastrointestinal tract is a major contributor to the pathogenesis of the systemic inflammatory response syndrome, sepsis, and multiorgan dysfunction in critical illness [1]. Evidence suggests that this is due to gastrointestinal malperfusion, with consequent damage to the lining of the gut and loss of the barrier preventing bacterial penetration. With its unique blood flow distribution, the gastrointestinal tract is especially vulnerable to ischemia and hypoxia. Therefore, monitoring of the perfusion of the gastrointestinal tract is important in critical illness. Under clinical conditions, the direct determination of the perfusion of the gastrointestinal tract is not feasible. Thus, Fiddian-Green et al. [2,3] introduced gastric tonometry, an indirect measurement of intramucosal PCO2 (PiCO (2)) within the lumen of the gastrointestinal tract, which provides information on gastrointestinal blood flow. A decrease of the calculated intramucosal pH (pHi) that results from an increase in PiCO2 can give information on the adequacy of metabolism and on the perfusion of the gastrointestinal mucosa. Gastrointestinal tonometry has been established as the only minimally invasive procedure for gastrointestinal monitoring routinely used under clinical conditions. In this review, we describe newer techniques for the determination of PiCO2 and compare these with conventional gastric tonometry. In addition, this article deals with the physiological background of gastric tonometry, as well as with several problems and misconceptions of this method. Finally, we provide a new hypothesis for the interpretation of data obtained by PiCO2 measurement. Conventional Gastric Tonometry Because of its physiological properties, CO2 can easily diffuse through various tissues and milieus. In the 1960s, Bergofsky [4] and Dawson et al. [5] showed that the diffusion of CO2 between the mucosa of a viscous organ and its lumen continues until an equilibrium is reached. Thus, the CO2 partial pressure in the mucosa (PiCO2) of a hollow viscous organ can be determined indirectly from the intraluminal pCO2[6]. The first report on the intraluminal pCO2 determination in the gastrointestinal tract was published in 1968 by Hamilton et al. [7], but it was not until 1982 that Fiddian-Green et al. [2] introduced the concept of gastric tonometry or "hollow viscous tonometry." The nasogastric tonometer was introduced 2 yr later.1 (1) Fiddian-Green RG. A sensitive and specific test for intestinal ischemia using silastic tonometers [abstract]. Eur Surg Res 1984;16:32. Conventional nasogastric tonometry is an intermittent and indirect method for the determination of the intraluminal pCO2. Saline is instilled into a balloon that is permeable to CO2 on the tip of a nasogastric tonometer probe. The saline is allowed to equilibrate with the intraluminal CO2 of the stomach and is then aspirated from the tonometer balloon and the PCO2 is determined by using a standard blood gas analyzer. The PiCO2 is obtained from the PCO2 value using a time-dependent correction factor that adjusts for the equilibrium time. Fiddian-Green et al. [2] did not consider the PiCO2 to be the most important variable for providing information on the gastrointestinal perfusion; instead, they used it to calculate the pHi value. The calculation of the pHi uses the Henderson-Hasselbalch equation: Equation 1 In addition to PiCO2, this Equation requiresinformation on the intramucosal bicarbonate concentration (cHCO3-), a variable that cannot be obtained clinically. Fiddian-Green et al. [2] recommend using the arterial cHCO3- instead of the intramucosal cHCO3- based on the assumption that the intramucosal cHCO3- in the gastrointestinal capillary region equals the arterial cHCO3-. Fiddian-Green et al.'s [2] pHi concept and the conventional technique of nasogastric tonometry are associated with two drawbacks: 1. A methodological error inherent in the indirect discontinuous PiCO (2) measurement with a nasogastric tonometer and a standard blood gas analyzer. 2. Difficulties with the interpretation of the calculated pHi as an indicator of gastrointestinal malperfusion. Several authors [9-12] have shown that using a standard blood gas analyzer, the PCO2 determination of an unbuffered aqueous solution as recommended in gastric tonometry can vary by up to 57.5% depending on the type of blood gas analyzer used [9]. Using a buffered tonometer solution, e.g. a phosphate buffer [9,10], the precision of the determination can be increased. However, several other methodological problems remain, including the handling of the tonometer and the CO2 sample, the long equilibration periods, and mistakes from equilibration [13]. These methodological insufficiencies have led to the development of better techniques for the determination of the PiCO2. New Methods for Determining PiCO2 The new methods for determining PiCO2 are summarized in Table 1. Salzman et al. [14] developed a balloonless air technique for the indirect intraluminal measurement of PiCO2 with a gas sample obtained from the gastrointestinal tract via a gastric probe. Likewise, Guzman and Kruse [15] used air as the tonometer medium, but they still used the conventional nasogastric tonometer. The introduction of air not only resulted in an improvement in precision, but also allowed the continuous measurement of PiCO2 via a closed loop recirculation system. Because the PiCO (2) is determined online by the infrared sensor of a capnograph, Guzman and Kruse called their method capnometric recirculating gas tonometry.Table 1: Comparison of Different Methods for Intramucosal PCO2 determinationThe Tonocap[trade mark sign] System (Datex, Helsinki, Finland) represents a similar concept of gas tonometry. In an automated procedure, air instilled into the balloon of a nasogastric tube is aspirated after a predefined equilibration time (5-10 min), then subjected to PCO2 determination by capnography. Creteur et al. [16] showed that this technique can improve the precision of PiCO2 determination significantly compared with conventional gastric tonometry. These methods determine PiCO2 indirectly, outside the gastrointestinal tract, but there is also a direct method for the continuous intraluminal PiCO2 measurement [17]. This method uses a fiberoptic CO2 sensor that is placed in the lumen of the gastrointestinal tract. A green light is passed down a 60-cm fiber glass to an acrylamid gel colored with Phenol Red at the fiber tip. This acrylamid gel is fixed in bicarbonate buffer and coated with an ion-resisting membrane that is permeable to CO2. If CO2 diffuses through the acrylamid gel, the pH and, thus, the absorption characteristics of the Phenol Red are changed. Depending on these pH changes, the absorption of the green light is altered. After proper calibration, this technique provides direct information on the PCO2 in the lumen of the gastrointestinal tract. In vitro validation has shown that within 9 min, the fiberoptic CO2 sensor will continuously detect a known PCO2 value with an error of +/- 3.5% (Figure 1) [17]. This is a major improvement compared with the conventional indirect techniques, which are not only discontinuous and indirect, but also have errors of +/-12.5%, even after 90 min. The complete CO2 equilibration inside the tonometer balloon depends on time and the tonometer fluid used (air or saline). As a consequence, capnometric recirculating gas tonometry or the Tonocap[trade mark sign] System must use correction factors (Table 1 shows the various techniques for the intramucosal PCO2 determination).Figure 1: PCO2 differences (measured PCO2 minus predefined PCO2) (mean n = 12). Data shown represent the continuous PCO2 measurement with the fiberoptic CO2 sensor in water ([horizontal bar]) and in humidified air ([triple dash]). The closed squares ([black square, horizontal bar, black square]) represent the measurements with the conventional nasogastric tonometer after 30, 60, and 90 min of equilibration. The measurement with the fiberoptic CO2 sensor introduced into the saline-filled balloon of the conventional nasogastric tonometer was determined in a single case ([quadruple dash]). Original Figure fromKnichwitz et al. [17].The continuous measurement of PiCO2 has many advantages over the discontinuous methods [24]. Depending on the set limits of PiCO2, it will automatically alarm the clinician when the value is beyond normal range. It can detect short-term changes in PiCO2 and can therefore be used for the early detection of gastrointestinal perfusion disturbances [15-17]. Finally, it offers another possibility. It is known that intragastric CO2 production from duodenal bicarbonate reflux interferes with the conventional measurement of intramucosal PCO2. The continuous PiCO2 measurement offers the chance to differentiate between ischemia and such a reflux. With bicarbonate reflux present, the aspiration of the gastric content will rapidly alter the produced signal. Sensitivity or Specificity: pHi or PiCO2? An important issue has been the question of whether the directly determined PiCO2 or the calculated pHi should be considered the primary variable in the interpretation of gastrointestinal perfusion disorders. Fiddian-Green1 states that for the detection of gastrointestinal malperfusion, the pHi is more sensitive than PiCO2. Perfusion disorders cause an increase in organ dysfunction and thus can influence outcome in critical illness. In 1987, Fiddian-Green and Baker [19] demonstrated that a reduction in pHi below 7.32 was a more sensitive indicator of organ dysfunction after cardiac surgery than arterial pH, blood pressure, cardiac output, or urine output (Table 2). In this article, the determined sensitivity was 100%. The specificity-that is, the identification of the healthy as such-was rather low (55.85%). Gastrointestinal acidosis was detected in all cases, but only 55.85% of patients tested had suffered organ failure.Table 2: Sensitivity and Specificity of the Variables Intramucosal pH (pHi), Intramucosal PCO2 (PiCO2), and Arterial Base Excess (art BE) with Respect to Predicting Organ Failure, Mortality, or pHiOther authors [20-23] have confirmed the high sensitivity and the low specificity in detecting organ failure or mortality in various subgroups of patients with different pHi limits and various time periods between pHi determinations (Table 2). Friedman et al. [22] showed that the PiCO2 was associated with a higher specificity but a lower sensitivity compared with the pHi in detecting mortality in the critically ill. These opposite findings are not surprising, because the pHi is calculated from a local gastrointestinal variable (PiCO2) and a systemic variable (arterial cHCO3-) (Table 2). In a study published by Boyd et al. [24], the arterial base excess, a variable comparable to the arterial cHCO3-, was able to detect a pHi decrease below 7.32 with low sensitivity and high specificity (Table 2). These authors demonstrated the important mathematical dependency between the systemic (arterial) cHCO3- and the pHi. Without a local gastrointestinal PiCO2 increase, an isolated arterial cHCO3- reduction can result in a calculated pHi decrease and thus indicate gastrointestinal acidosis where none exists. Either PiCO2 or cHCO3- alone is inferior to pHi as a predictive variable of mortality. In critical illness, pHi is very sensitive for estimating outcome. However, it is by no means a sensitive variable for detecting gastrointestinal malperfusion, contrary to what is often assumed. The Physiologic Basis of PiCO2 Measurement Hypoxia and ischemia can be detected indirectly using metabolic variables. In addition to a reduction in pH and lactic acidosis, an increase in PCO2 can be a sign of malperfusion. With a PCO2 increase, two pathophysiological mechanisms play an important role [25]: 1. CO2 accumulation: The reduced washout of CO2 due to an impaired blood flow causes a CO2 accumulation in the tissues. This CO2 is eliminated with reperfusion and then exhaled by the lung. 2. CO2 production under the condition of anaerobic energy metabolism: During anaerobic energy production, hydrogen ions (H+) are generated from two metabolic pathways. With anaerobic glycolysis, the metabolism of glucose to lactic acid forms two hydrogen ions (D - glucose [right arrow] 2ATP + 2lactate- + 2H (+)). The adenylate kinase reaction transforms adenosine triphosphate (ATP) to energy and leads to the intracellular generation of hydrogen, which is not used for the resynthesis of ATP. Equation 2 With hypoxia, this increased production of H (+) results in an increased production of CO2 (H+ + HCO3- [left and right arrow] H2 O + CO2). The CO2 accumulation and anaerobic CO2 production are related pathways, but their relative importance differs depending on the circumstances. Only the regional CO2 accumulation forms the basis for a regional PiCO2 measurement. Without CO2 accumulation, no PiCO2 increase can occur. This basic postulate is of utmost importance for the interpretation of gastrointestinal PiCO2. The Significance of Increased Intramucosal CO2 Production The early detection of anaerobic energy metabolism that can result in irreversible tissue damage is of major clinical interest. The key point of the PiCO (2) measurement is the relation between blood flow reduction and CO2 increase. In theory, considering constant CO2 production, any twofold increase in arteriovenous CO2 in a certain region represents a reduction of that region's perfusion by 50%. Under these circumstances there should be a linear relation between blood flow and PiCO2. However, with ischemia or hypoxia, the circumstances are less predictable and less constant, resulting in an additional CO2 increase, as shown in Figure 2[26]. These results were obtained in experiments on six pigs with stepwise reduction of the superior mesenteric arterial blood flow at 10-min intervals. The resulting PiCO2 changes were detected with a fiberoptic CO2 sensor placed within the ileal lumen. After a period with only minor PiCO2 changes, there was a dramatic PiCO2 increase, together with a blood flow reduction of approximately 60%. This finding supports the observations by Zhang and Vincent [27] and Schlichtig and Bowles [28], who demonstrated a similar critical limit of blood flow, which, if surpassed, produced a sudden PCO2 increase. This sudden CO2 change reflects a metabolic decompensation.Figure 2: Individual changes (n = 6 pigs) of the ileal intramucosal PCO2 (PiCO (2)) measured by using the fiberoptic CO2 sensor ([black square, horizontal bar, black square]) with stepwise reduction of the superior mesenteric artery bloodflow (BFsMA). Original Figure fromKnichwitz et al. [26].The PiCO2 continuously measured with the fiberoptic CO2 sensor can detect this threshold between compensated and decompensated mesenterial perfusion. Schlichtig and Bowles [28] postulated that this sudden CO2 increase is caused mainly by anaerobic energy production [28]. However, in their experimental model, the reduction of systemic oxygen delivery did not result in regional, isolated mesenteric ischemia, but in a global ischemia, a condition that has an impact on the HCO3-CO2 buffer system. The amount of CO2 produced below this threshold of sudden CO2 increase is unknown. The HCO3-CO2 Buffer System On first sight, the HCO3-CO2 buffer with its pK of 6.1 and blood pH of 7.40 seems to be a rather and system. is the result of the constant HCO3- production and CO2 which causes an between HCO3- and (2) with a of The Equation the HCO3-CO2 buffer the of produced CO2 by the is of These circumstances are called an system. A condition that results in a CO2 accumulation is as a closed system. Under conditions, the addition of 2 results in a pH of as only the of the Equation In under closed conditions, the produced CO2 is not and thus the of the Because the HCO3-CO2 buffer cannot under these conditions, these changes result in a decrease of the pH to of the buffer Equation The closed with CO2 accumulation has a lower bicarbonate buffer because the of the is not A new is between CO2 and HCO3- with reduced of With this closed the buffer an important role because it the The closed is by HCO3- compared with the system. PiCO2 in and The between and closed is of major importance for the PiCO2 measurement. a of experiments to this influence on the (Figure shows the PiCO2 determined by using a fiberoptic CO2 sensor in the ileal lumen of a After measurements were gastrointestinal ischemia was by a reduction of the blood flow in the superior mesenteric artery for min after which the gut was for min. This represents a closed because the reduced perfusion did not an washout of the produced CO2. the arterial PCO2 (Figure bar, black square, horizontal bar]) constant at approximately a sudden PiCO2 increase in the intestinal mucosa (Figure bar, black square, horizontal This PiCO2 increase was most caused by CO2 accumulation with additional CO2 production from anaerobic energy The concentration in the mesenteric blood was the PCO2 was determined in a outside the gut the superior mesenteric a but PCO2 increase was (Figure black square, systemic and local changes under mesenteric ischemia by a reduction of the superior mesenteric artery blood flow for min. measurement of a min = 30, min = mesenteric min = mesenteric in intramucosal PCO2 (PiCO bar, black square, horizontal mesenteric PCO2 black square, and arterial PCO2 ([triple black square, dash]). in intramucosal pH bar, black square, horizontal and mesenteric black square, and arterial ([triple black square, bicarbonate concentration from et al. finding was the min of the PCO2 in the superior mesenteric its value after the of the PiCO2 increased at (Figure This can be by the various of the PCO2 determinations that represent different tissue Because of its intraluminal the PiCO2 measurement can detect in the which is vulnerable to In this reperfusion and, thus, CO2 washout are by and capillary In the superior mesenteric PCO2 represents only CO2 washout and, as that allowed proper of the showed of a finding that has been by The specific of the gastrointestinal tract an important role in the of the of the The artery of the to the tip of the and then into a capillary to perfusion. The artery very to the The are only oxygen can the concentration of As a oxygen can the tip of the and the Depending on the of blood flow reduction and the of the tissue of the and resulting in an even reduced CO2 washout with CO2 detect these pathophysiological changes, the PiCO2 should be The pHi value (Figure right bar, black square, horizontal bar]) calculated using PiCO2 and mesenteric cHCO3- (Figure right black square, a acidosis with a pHi of However, because mesenteric cHCO3- changes not be under such closed conditions, this pHi decrease from the local intestinal PiCO2 increase. The are different under conditions, a that we introduced with The results are shown in in Figure A was placed on for min, by an period of min. During a lactic acidosis with a concentration of and an arterial pH of were The blood flow in the superior mesenteric artery was not compared with Under these conditions, the CO2 produced from anaerobic metabolism was and eliminated by the membrane was no change in mesenteric or arterial PCO2 (Figure CO2 was eliminated from the and was by of Therefore, a arterial (Figure right black square, and mesenteric cHCO3- reduction Figure right black square, from to was Because the PiCO2 did not the calculated pHi decrease (Figure right bar, black square, horizontal bar]) from the cHCO3- of ileal tissues from this showed no of the gastrointestinal The led to a systemic acidosis that was associated with a local intestinal This acidosis was less than that under closed conditions, did not cause irreversible or capillary and thus did not the CO2 The continuous measurement of PiCO (2) under showed no PiCO2 systemic and local changes under for min. measurement of a min = 30, min = min = after in intramucosal PCO2 bar, black square, horizontal mesenteric PCO2 black square, and arterial PCO2 ([triple black square, dash]). in intramucosal pH bar, black square, horizontal and mesenteric black square, and arterial ([triple black square, bicarbonate concentration PiCO2, cHCO3-, and pHi for and closed can be from other these it is that the of the gastrointestinal using only the pHi is In the clinical it is often to differentiate between and closed as within a single of the a gastrointestinal perfusion under conditions. Depending on the illness, closed are introduced when CO2 washout is by impaired capillary The detection of this time point must be the most important of any PiCO2 measurement. Only continuous PCO2 measurement the determination of the time point when CO2 accumulation in the mucosa This time point to the time at which a certain is from perfusion and thus represents a closed system. This is also the time point at which tissue damage is to perfusion is The of PiCO2, cHCO3-, and pHi that are summarized in Table this it is that the determination of the pHi by not provide information on gastrointestinal malperfusion. The tissue CO2 accumulation, and PiCO2 increase of local ischemia and CO2 accumulation, the calculated pHi can decrease as a result of a systemic Thus, only the PiCO2 can provide the information to detect the change from an to a closed to the Variables Intramucosal PCO2 (PiCO2), Intramucosal pH (pHi), and Arterial (art at the Under the several have the influence of various on a mathematical from the of systemic cHCO3- and local PiCO2. these two has led to an increase in sensitivity the outcome of critical illness. it not any information on a gastrointestinal perfusion many have used the pHi as an indicator of impaired gastrointestinal perfusion. The only variable that can provide such information on gastrointestinal blood flow is PiCO2, PiCO2 is a value that can only be used to an impaired capillary perfusion from a normal perfusion a PiCO (2) With impaired gastrointestinal perfusion and increased PiCO2 the of any can only be by a decrease in PiCO2 caused by CO2 washout with capillary A provide data on PiCO cHCO3-, and pHi and give into With no for the monitoring of gastrointestinal a of are used gastric tonometry, pHi and PiCO2 measurement. This has led to that be eliminated if the method were after the primary variable that it to PiCO2 measurement. PiCO2 measurement can be summarized as 1. The PiCO2 is the primary variable of this gastrointestinal monitoring technique and an only for the hollow viscous organ in which it is An increase in PiCO2 suggests a local with CO2 accumulation in the region of the gastrointestinal tract. 2. The PiCO2 measurement not the gastrointestinal blood flow. It information on normal or impaired perfusion. Data from experiments that there is a threshold of impaired perfusion at approximately of the gastrointestinal blood flow. reduction in blood flow causes a increase in PiCO2 with additional CO2 interpretation of PiCO2 of metabolic changes of of and PCO2 causes a PiCO2 increase [17]. The interpretation of any obtained PiCO2 value additional information on the between intramucosal and arterial PCO2 and closed the measurement of PiCO2 and the arterial cHCO3-. The calculated pHi can as a sensitive outcome It regional perfusion With all on gastrointestinal perfusion over the have used conventional discontinuous nasogastric tonometry. Because this technique can only provide information on the pHi over a certain many short-term are Only the introduction of continuous such as the fiberoptic CO2 the online detection of important PiCO2 several have clinical with a continuous but this technique can be recommended for clinical additional that a normal and which to an improvement of gastrointestinal perfusion are