Respiratory depression is still one of the most undesirable side effects of using opioids. Therefore, good alternatives for classic opioids are needed especially in mild or moderate pain situations when non-steroidal anti-inflammatory drugs (NSAIDs) are contraindicated. The opioids with μ-opioid receptor activity are associated with clinical respiratory depression . Tramadol is a centrally acting opioid with a low affinity for μ-opioid receptors. Tramadol has recently been registered as an analgesic in Finland, Sweden, Denmark and Great Britain. It enhances the antinociceptive effect of the descending inhibitory pathway in the spinal cord by inhibition of monoaminergic re-uptake . Because of its low affinity for μ-opioid receptors, tramadol has been claimed not to depress respiration [2-4].
Fractional inspired-expired oxygen concentration(F(I-E)O2) difference is the most sensitive and novel method for early detection of hypoventilation and hypoxia . It has been successfully used for evaluation of respiratory effects of opioids .
We therefore compared the respiratory effects of tramadol and pethidine, a classic opioid, in a standardized clinical setting with special reference to F(I-E)O2 difference in a randomized, placebo-controlled and double-blind study.
Patients and methods
Thirty-six ASA Grade I or II patients gave their in-formed consent to participate in the study approved by the Ethics Committee of the hospital. Obese patients (body mass index greater than 30) or patients with pulmonary or cardiac disease were excluded. The patients were premedicated with an oral diazepam mixture 0.2 mg kg−1 30 min before arrival in the operating theatre. An intravenous(i.v.) infusion of Ringer's acetated solution was started. Non-invasive monitoring of arterial pressure, ECG, and pulse oximetry was instituted. The pharynx was anaesthetized with 30 mg of lignocaine 4% spray. After glycopyrronium 3 μg kg−1, anaesthesia was induced with propofol. Suxamethonium 1.5 mg kg−1, preceded by alcuronium 0.03 mg kg−1, was given to facilitate tracheal intubation. Before intubation the trachea was anaesthetized with 4% lignocaine 40 mg spray. The lungs were manually ventilated with 70% nitrous oxide in oxygen with halothane until the return of spontaneous respiration. Halothane was then adjusted to reach an end-tidal concentration of 0.3%. Thereafter, the patients were allocated by selection using a sealed envelope to receive either tramadol 0.6 mg kg−1 (tramadol group), pethidine 0.6 mg kg−1 (pethidine group) or physiological saline (placebo group). A trained nurse not participating in the study prepared the study drugs. The drugs were diluted in the same volume of 10 mL given to each patient 0.1 mL kg−1. The study solution was injected i.v. within 30 s when the respiratory parameters had been constant for 10 min and the uptake of nitrous oxide was completed (F(I-E)N2O=0). After the trial drugs, ventilation was assisted if needed until spontaneous breathing returned with an end-tidal carbon dioxide concentration (PetCO2) less than 8.0 kPa. The study was performed before the start of surgery.
The expired gas passed a non-rebreathing valve. The respiratory parametres were monitored by side-stream spirometry (Datex® Ltd, Finland).
The respiratory parameters: F(I-E)O2, PetCO2, end-tidal halothane, respiratory rate, minute volume, tidal volume and pulse oximetry (SpO2) were measured and recorded by a chart recorder before and 1, 2, 3, 4, 5, 10, 15, 20, 25, and 30 min after injection of the study drugs. Arterial blood was sampled for blood gas analysis immediately before and 30 min after the injection of the study drug as a single bolus. Thereafter, the patients received vecuronium and anaesthesia was continued as required for surgery.
The changes within a group were analysed with one-way analysis of variance with Bonferroni's correction. The differences between the groups were tested with two-way analysis of variance. The results are expressed as mean and 95% confidence intervals (95% CI). A P-value less than 0.05 was considered statistically significant.
The three groups, 12 patients in each, were comparable in age, sex, height, weight and body mass index (Table 1). The study drugs were injected on average 19 min (95% CI, 17-21 min) after administration of propofol 1.5-2.7 mg kg−1. There were no differences in the dose of propofol in the three groups.
F(I-E)O2 and SpO2
The F(I-E)O2 tended to increase slightly in placebo and tramadol groups (P=0.28 and 0.10, respectively). The F(I-E)O2 increased significantly in the pethidine group (P<0.001) and was elevated throughout the trial (Fig. 1). The F(I-E)O2 was significantly higher in the pethidine than in the placebo or tramadol groups from the 3-min study point to the end of the trial (P<0.01). No changes occurred in SpO2(mean 97) in any of the groups.
PetCO2 and arterial blood gases
PetCO2 remained stable during the study period both in the placebo and tramadol groups. In the pethidine group, PetCO2 increased significantly from base-line (Fig. 2). The mean base-line pH of arterial blood was similar at the 30 min measurement in the placebo and tramadol groups (7.37) but decreased from a mean of 7.37 (7.35-7.38) to a mean of 7.34 (7.31-7.36) in the pethidine group (P<0.05). Arterial oxygen tension remained at base-line levels during the trial. Arterial PCO2 remained stable in the placebo and tramadol groups but increased significantly in the pethidine group from 6.2 (5.7-6.5) to 6.9 (6.3-7.4) kPa (P<0.05).
Two minutes after administration of pethidine two patients developed apnoea. Their lungs had to be manually ventilated for 10 min until spontaneous respiration returned. The respiratory rate decreased significantly from base-line in the pethidine group (P<0.001) but remained constant in the placebo and tramadol groups (Fig. 3). The difference between pethidine and the two other groups was statistically significant throughout the trial.
Minute ventilation and tidal volume
The minute ventilation did not change in the placebo and tramadol groups. However, in the pethidine group it decreased by 50% within 2 min and did not reach base-line value during the trial (P<0.001) (Fig. 4). The tidal volume remained stable in all the groups.
The systolic arterial blood pressure decreased on average 7, 6, and 5 mmHg and heart rate 10, 11 and 8 beats per min in the placebo, tramadol and pethidine groups, respectively (NS). No adverse effects on the cardiovascular system were noted in any group during the trial.
In this study setting, the effect of tramadol on respiration was similar to that of placebo. On the contrary, pethidine induced a significant respiratory depression involving all respiratory parametres.
Linko and Paloheimo  demonstrated elegantly that F(I-E)O2 is the most sensitive measure for the detection of hypoventilation. It predicts hypoxaemia earlier than changes in SpO2. The fractional oxygen difference is also a more sensitive indicator of hypoventilation than PetCO2. The is partially due to the fact that CO2 accumulation is a rather slow process , but this ratio has not gained the popularity it deserves. Accordingly, in our study, the F(I-E)O2 with pethidine increased whereas no changes could be noted in SpO2. With the sensitive F(I-E)O2 method the effects of tramadol were similar to those of the placebo.
This model for studying respiratory depression by opioids is a modification of the method described by Vickers and colleagues . Our study in anaesthetized patients was conducted in the pre-operative phase in order to avoid disturbing surgery. Also, the irritating effect of the endotracheal tube on respiration was minimized with a local anaesthetic sprayed into the pharynx and trachea before intubation. The respiratory effects of propofol can be considered to be minimal because the time interval from propofol to administration of the trial drugs was ≈ 19 min . The concentration of halothane was kept as low as possible. Therefore, respiratory depression by anaesthetic regimen was eliminated as seen in our placebo group.
The potency ratio between tramadol and pethidine has been estimated to be 0.94 . Therefore, we chose the same dose for both drugs (0.6 mg kg−1).
Forty-two to 60% of pethidine  and 20% of tramadol  is plasma bound. The pH of the serum or plasma can markedly affect opioid plasma binding [8, 9]. In our study the base-line arterial pH was normal in all groups because of the study plan. In the pethidine group pH was initially normal but decreased significantly during the trial, although the pH of arterial blood remained within the normal range. Therefore, we can assume that there was no change in the free fraction of study drugs during the study.
All opioids acting via μ-receptors cause a dose-related depression of respiration via a mechanism probably involving a decrease in the sensitivity of the respiratory centre to CO2. Decreased respiratory rate, increased PetCO2 and a decrease in minute volume of ventilation have been considered indicators of hypoventilation . In our study pethidine caused a significant change in all these measurements. Respiratory rate and minute volume of ventilation decreased below base-line values one minute after pethidine injection. PetCO2 was significantly elevated after 4 min. In the study by Vickers and colleagues, with a similar scheme , the mean respiratory rate decreased significantly with tramadol 0.5 mg kg−1 compared with placebo. However, this response was not dose dependent and there was no change in end-tidal cartoon dioxide tension. They used a higher end-tidal halothane concentration (1.0%) which may explain the difference in the results of our study and those of Vickers . The analgesic effect of i.v. tramadol is seen 15 min after its administration to healthy volunteers . Therefore, our observation time (30 min) can be considered appropriate.
Tramadol is suggested for the treatment of moderate acute or chronic pain including post-operative and obstetric pain as well as in pain of various other origins . The results of our study suggest that tramadol may be a safer opioid with respect to respiratory depression than pethidine, a classic opioid. Haemodynamic changes were minimal and similar after all drugs in the present trial.
It can therefore be concluded that tramadol does not depress respiration in spontaneously breathing anaesthetized patients. An equianalgesic dose of pethidine caused significant respiratory depression compared with tramadol in this setting.
The authors are indebted to Ms Marja Friman, Ms Päivi Vättö and Ms Ulla Kohvakka for their skilful assistance. The financial support from Orion corporation, Kuopio, Finland is gratefully acknowledged.
1 Paakkari P, Paakkari I, Sirén A-L, Feuerstein G. Respiratory and locomotor stimulation by low-doses of dermorphine, a Mu 1 receptor-mediated effect. J Pharmac Exp Ther
2Lee CR, McTavish D, Sorkin EM. Tramadol: a preliminary review of its pharmacodynamic and pharmacokinetic properties, and therapeutic potential in acute and chronic pain states. Drugs
3Vickers MD, O'Flaherty D, Szekely SM, Read M, Yoshizumi J. Tramadol: pain relief by an opioid without depression of respiration. Anaesthesia
4 Houmes R-JM, Voets MA, Verkaaik A, Erdman W, Lachmann B. Efficacy and safety of tramadol vs. morphine for moderate and severe postoperative pain with special regard to respiratory depression. Anesth Analg
5 Linko K, Paloheimo M. Inspiratory end-tidal oxygen content difference: A sensitive indicator of hypoventilation. Crit Care Med
6Hamunen K. Ventilatory effects of morphine, pethidine and methadone in children. Br J Anaesth
7Cummings GC, Dixon J, Kay NH et al. Dose requirements of
ICI 35,868(Propofol, 'Diprivan') in a new formulation for induction of anaesthesia. Anaesthesia
8 Wood M. Plasma drug binding: Implications for anesthesiologists. Anesth Analg
9 LaRosa C, Morgan DJ, Mather LE. Pethidine binding in whole blood: methodology and clinical significance. Br J Clin Pharmacol
10 Tabatabai M, Kithata LM, Collins JG. Disruption of the rhythmic activity of the medullary inspiratory neurons and phrenic nerve by fentanyl and reversal with nalbuphine. Anesthesiology
11 Parth P, Madler C, Morawetz RF. Characterization of the effect of analgesics on the assessment of experimental pain in man. Pethidine and tramadol in a double blind comparison. Anaesthesist