Brain oxygen monitoring: in-vitro accuracy, long-term drift and response-time of Licox- and Neurotrend sensors

BACKGROUND

Oxygen tension sensors have been used to monitor tissue oxygenation in human brain for several years. The working principals of the most frequently used sensors, the Licox (LX) and Neurotrend (NT), are different, and they have never been validated independently for correct measurement in vitro. Therefore, we tried to clarify if the two currently available sensors provide sufficient accuracy and stability.

METHOD

12 LX oxygen tension sensors and NT sensors were placed into a liquid-filled tonometer chamber. The solution was kept at 37 +/- 0.2 degrees C and equilibrated with five calibration gases containing different O(2)- and CO(2)-concentrations. After equilibration, readings were taken for each gas concentration (accuracy test). Afterwards, the sensors were left in 3% O(2) and 9% CO(2) and readings were taken after 24, 48, 72, 96 and 120 hours (drift test). Thereafter, a 90% response time test was performed transferring sensors from 1% to 5% oxygen concentration and back, using pre-equilibrated tonometers.

FINDINGS

All Licox oxygen probes [12] were used for this study. Two of 14 Neurotrend sensors did not calibrate, revealing a failure rate of 14% for NT. Oxygen tension during the accuracy test was measured as follows: 1% O(2) (7.1 mmHg): LX 6.5 +/- 0.4, NT 5.3 +/- 2.3 mmHg, 2% O(2) (14.2 mmHg): LX 12.9 +/- 0.6, NT 12.1 +/- 2.2 mmHg, 3% O(2) (21.4 mmHg): LX 19.8 +/- 0.7, NT 19.4 +/- 2.4 mmHg, 5% O(2) (35.8 mmHg): LX 33.4 +/- 1.0 mmHg, NT 33.5 +/- 2.9 mmHg, 8% O(2) (57.0 mmHg): 53.8 +/- 1.5, NT 53.6 +/- 3.3 mmHg. After 120 hours in 3% O(2) (21 mmHg), LX measured 19.8 +/- 1.9 mmHg, NT 17.9 +/- 4.7 mmHg. 90% response time from 1% to 5%/5% to 1% oxygen concentration was 129 +/- 27/174 +/- 26 sec for LX, 55 +/- 19/98 +/- 39 sec for NT.

CONCLUSIONS

Both systems are measuring oxygen tension sufficiently, but more accurately with LX probes. NT sensors read significantly lower pO(2) in 1% O(2) and show an increasing deviation with higher oxygen concentrations which was due to two of twelve probes. A slight drift towards lower oxygen tension readings for both sensors but more pronounced for the NT does not impair long-term use. NT measures pCO(2) and pH very accurately.

Enhancing accuracy of magnetic resonance image fusion by defining a volume of interest

We compared the registration accuracy for corresponding anatomical landmarks in two MR images after fusing the complete volume (CV) and a defined volume of interest (VOI) of both MRI data sets. We carried out contrast-enhanced T1-weighted gradient-echo and T2-weighted fast spin-echo MRI (matrix 256 x 256) in 39 cases. The CV and a defined VOI data set were each fused using prototype software. We measured and analysed the distance between 25 anatomical landmarks in predefined areas identified at levels L(1)-L(5) corresponding to defined axial sections. Fusion technique, landmark areas and level of fusion were further processed using a feed-forward neural network to calculate the difference which can be expected based on the measurements. We identified 975 landmarks for both T1- and T2-weighted images and found a significant difference in registration accuracy ( P<0.01) for all landmarks between CV (1.6+/-1.2 mm) and VOI (0.7+/-1.0 mm). From cranial (L(1)) to caudal (L(5)), mean deviations were: L(1) CV 1.5 mm, VOI 0.5 mm; L(2) CV 1.8 mm, VOI 0.4 mm; L(3) CV 1.7 mm, VOI 0.4 mm; L(4) CV 1.6 mm, VOI 0.6 mm; and L(5) CV 1.6 mm, VOI 1.6 mm. Neural network analysis predicted a higher accuracy for VOI (0.05-0.15 mm) than for CV fusion (0.9-1.6 mm). Deviations due to magnetic susceptibility changes between air and tissue seen on gradient-echo images can decrease fusion accuracy. Our VOI fusion technique improves image fusion accuracy to <0.5 mm by excluding areas with marked susceptibility changes.

Rehabilitation of hearing and communication functions in patients with NF2

Most patients with neurofibromatosis type 2 (NF2) lose hearing either spontaneously or after removal of their neurofibromas. The patient may benefit from conventional hearing aids if, due to modern microsurgery and intraoperative monitoring the integrity of the cochlea and the 8th nerve is preserved. With lost auditory function but preserved electrical stimulibility of the 8th nerve a cochlear implant may be appropriate. But if the patients have no remaining 8th nerve to stimulate, there is no benefit from cochlear implants. Until some years ago, vibrotactile aids, lip-reading, and sign language have been the only communication modes available to these patients. With auditory brain stem implants it is now possible to bypass both the cochlea and the 8th nerve and to stimulate the cochlear nucleus directly. Stimulation of the devices produces useful auditory sensations in almost all patients. Testing of perceptual performance indicated significant benefit from the device for communication purposes, including sound-only sentence recognition scores and the ability to converse on the telephone. Also lip-reading is significantly improved with brain stem implants. The successful work of an auditory brainstem program center depends very much on the close interdisciplinary collaboration between the Departments of Neurosurgery and ENT-surgery. In the future new developments like speech processing strategies and new designed electrodes accessing the complex tonotopic organization of the cochlear nucleus may further improve rehabilitation in these patients who would have been deaf some years ago.

High risk of intraoperative awareness during cardiopulmonary bypass with isoflurane administration via diffusion membrane oxygenators

In cardiac surgery with the aid of extracorporeal circulation (ECC), inhalation anaesthetics can be administered via the oxygenator. Until the recent advent of a new type of diffusion membrane oxygenator, we routinely added the inhalation agent, isoflurane, to the gas flow of a microporous capillary membrane-type oxygenator. Applying this procedure to the diffusion-type oxygenators, the depth of anaesthesia appeared to be affected, which manifested itself through unusually high intraoperative perfusion pressures. This observation led to a prospective randomized study comprising 60 patients and two models of a microporous capillary membrane oxygenator, as well as two models of a diffusion membrane oxygenator. Simultaneous isoflurane concentration measurements at both the gas inlet and outlet ports of the oxygenators showed that, whereas in the microporous capillary-type oxygenators the isoflurane administered was reduced by about 50% during the passage of gas through the device, there was only a minimal transfer of isoflurane in the diffusion-type membrane oxygenators.

High risk of intraoperative awareness during cardiopulmonary bypass with isoflurane administration via diffusion membrane oxygenators

In cardiac surgery with the aid of extracorporeal circulation (ECC), inhalation anaesthetics can be administered via the oxygenator. Until the recent advent of a new type of diffusion membrane oxygenator, we routinely added the inhalation agent, isoflurane, to the gas flow of a microporous capillary membrane-type oxygenator. Applying this procedure to the diffusion-type oxygenators, the depth of anaesthesia appeared to be affected, which manifested itself through unusually high intraoperative perfusion pressures. This observation led to a prospective randomized study comprising 60 patients and two models of a microporous capillary membrane oxygenator, as well as two models of a diffusion membrane oxygenator. Simultaneous isoflurane concentration measurements at both the gas inlet and outlet ports of the oxygenators showed that, whereas in the microporous capillary-type oxygenators the isoflurane administered was reduced by about 50% during the passage of gas through the device, there was only a minimal transfer of isoflurane in the diffusion-type membrane oxygenators.

The translaminar approach to canalicular and cranio-dorsolateral lumbar disc herniations

INTRODUCTION

The interlaminar approach is the standard procedure for most disc herniations in lumbar spine surgery. However, in cranially extruded disc herniations including canalicular herniations, partial or complete facetectomy is necessary with increased risk of postoperative spinal instability. We present the translaminar technique which allows a more direct and less destructive operative approach.

METHODS

30 patients using the translaminar fenestration were analysed by a postoperative follow-up of 6 weeks and one year. The mean-age was 57.2 years. For resection of the disc herniation, a small round or oval fenestration (6-8 mm) in the hemilamina, craniomedially to the facet joint, was performed. No patient received a partial or total facetectomy.

RESULTS

The majority of affected discs were at the L4-L5 level (53%). An extruded fragment was found in 28 patients (93%). In 5 patients bleeding from epidural veins complicated the intra-operative course. In 50% the nerve root was visually exposed. 15 patients (50%) had an intervertebral discectomy additional to the fragment excision. One patient was re-operated on after 10 days because of persisting radicular pain by using the same translaminar approach. 28 patients showed complete or nearly complete relief of radicular pain. Using this approach we have seen no major complication or clinical instability during a follow-up of at least one year.

CONCLUSIONS

The translaminar approach is an effective and minimally invasive technique in both canalicular and cranio-dorsolateral disc herniations. It gives an additional possibility to avoid partial removal of the facet joints, can be performed in all lumbar segments and preserves structures important for segmental spinal stability. The approach allows access to the extruded disc fragment and intervertebral disc space comparable to classical approaches and is a frequently used operative technique in our department.

Intra-Operative electromyographic monitoring of extra-ocular motor nerves (Nn. III, VI) in skull base surgery

BACKGROUND

Extraocular motor nerves (Nn. III, IV, VI) are at risk of damage during skull base surgery. A new recording technique was employed in 18 patients suffering from various skull base tumours in order to extend intra-operative EMG monitoring to the extra-ocular muscles.

METHODS

Selective intra-operative EMG recordings were obtained from extra-ocular muscles by placement of single-shafted bipolar needle electrodes under the guidance of B-mode ultrasound to visualise the needle tip within the target muscle in the orbital cavity.

FINDINGS

Following bipolar electrical stimulation, the oculomotor nerve (N.III) was intra-operatively identified in 5 out of 7 cases, and the abducens nerve (N.VI) in 12 out of 18 cases. Postoperative (3-6 months) oculomotor nerve function remained unchanged in 5 and improved in 2 patients. No permanent deterioration was observed. Abducens nerve function deteriorated in two patients and improved in one case, but remained unchanged in 15 cases. No side effects occurred. There was neither any distinct relation of ocular motor nerve function to the kind and extent of SMA ("spontaneous muscle activity") patterns, nor could such relationship be detected with concern to neurophysiological parameters (latencies, amplitudes) of electrically evoked CMAP ("compound muscle action potentials").

INTERPRETATION

The EMG technique proposed proved to be mainly effective as a mapping tool for intra-operative localisation and identification of ocular motor nerves in skull base surgery. However, the predictive value of conventional neurophysiological parameters for clinical outcome, seems to be rather poor. Further studies on a larger number of patients are therefore required to develop new quantification techniques which enable an intra-operative prediction of ocular motor nerve deficits. Further efforts are also necessary to extend this technique to the trochlear nerve.

Intra-operative electromyographic monitoring of the lower cranial motor nerves (LCN IX-XII) in skull base surgery

The functional preservation of lower (motor) cranial nerves (LCN) is endangered during skull base surgery. Intra-operative EMG monitoring of the LCN IX-XII was investigated in 78 patients undergoing 80 operations on various skull base tumors with regard to technical feasibility and clinical efficacy. Ongoing 'spontaneous muscle activity' (SMA) and 'compound muscle action potentials' (CMAP) following supramaximal bipolar stimulation were intra-operatively recorded applying needle electrodes into the soft palate (CN IX: n=76), the vocal cord (CN X: n=72), the trapezius muscle (CN XI: n=18), and the tongue (CN XII: n=71). From 24/22/8 cases with LCN IX/X/XII deficits (despite monitoring) only 5/6/4 remained unchanged (3-6 months postoperative). An irreversible plegia of the LCN IX/X/XII occurred in three (1/1/1) patients. In 7/6/1 patients postoperative (3-6 months) LCN IX/X/XII function was better than preoperatively. In all patients accessory nerve function remained unchanged. 'Pathological' SMA of the LCN IX/X/XII occurred in 12/16/8 cases, but in only 6/5/3 cases corresponded to postoperative LCN deficits. Corresponding 'pathological' SMA patterns were found in 18/17/5 out of 24/22/8 cases with postoperative LCN IX/X/XII dysfunction. Reproducible CMAP of LCN IX/X/XI/XII could be recorded in 59/56/11/32 patients. Approximate 'normal' values were calculated and compared to (very few) data so far given in the literature. Electromyographic monitoring proved to be a safe tool for the intra-operative identification and localization of the LCN contributing to their anatomical and functional preservation. The predictive value of standard neurophysiological parameters for functional outcome, however, is limited.

The fronto-orbital osteotomy as plastic-reconstructive approach to the anterior and middle skull base

INTRODUCTION

A combined extra-intracranial access for the operative exploration of tumours of the anterior and middle skull base is indicated when the tumour extends intracranially and simultaneously into the nasal cavity, the paranasal sinuses or the orbit.

METHODS

Two standardized modifications of the fronto-orbital osteotomy, the fronto-orbito-nasal and the fronto-orbito-zygomatic osteotomy, allow safe removal of skull base tumours in these locations. In extensive skull base tumours, a modified bilateral fronto-orbital-zygomatic osteotomy can be used.

RESULTS

Between February 1993 and July 2000 skull base tumours in 111 patients were resected using the presented methods. The most frequent tumour type was meningioma in 29 cases. Complications were encountered in 13 cases (11.7%).

CONCLUSION

The advantages over other approaches are good extra- and intracranial overview and minimal cerebral trauma. Additional transfacial incisions are not usually necessary. Exact repositioning of the fronto-orbital segments leads to optimal aesthetic results.

cAMP response element modulator (CREM): an essential factor for spermatogenesis in primates?

CREM is a cAMP-related transcription factor and alternate promotor usage and splicing generate repressor and activator transcripts of CREM within the testis. CREM activators are highly expressed in post-meiotic haploid germ cells and are essential for spermatid maturation in the mouse model as revealed by gene-targeting studies. Analysis of testicular CREM expression in rodent and monkey species, and in men yielded a highly comparable pattern thus suggesting that CREM is of general importance for spermatid development in the mammalian testis. Also, many CREM target genes have been identified in haploid germ cells. Studies in men with spermatogenic disturbance and spermatid maturation arrest demonstrated abnormal CREM expression and altered splicing events. Collectively, the data strongly argue for an essential role of CREM during spermatid maturation in primates.

CREM activator and repressor isoforms in human testis: sequence variations and inaccurate splicing during impaired spermatogenesis

cAMP responsive element modulator (CREM) activators are specifically expressed in haploid germ cells prior to cell elongation and are essential for spermatid development in mice. Recent studies indicate that CREM activators are also involved in the process of spermatid maturation in men. Unlike the activators, CREM repressors were suggested to be absent from adult mouse and dog testes. The present work investigates CREM transcripts in human testis with normal (n = 4) and impaired spermatogenesis (n = 2). Two activator transcripts could be identified corresponding to the tau2 isoform with and without exon gamma. Interestingly, four CREM repressors could be isolated from specimens with complete spermatogenesis. These were gamma-repressor (exons B, E, F, H, I(b)), CREM DeltaC-F, beta (exons B, G, H, I(b)), CREM DeltaC-G, beta (exons B, H, I(b)), and CREM DeltaC-G, alpha (exons B, H, I(a), I(b)). These isoforms were also present in cynomolgus monkey testes. A novel CREM splice variant (CREM DeltaC-H) was detected in a specimen with spermatid maturation defect. Beyond that, inaccurate CREM splicing, giving rise to inactive transcripts, was encountered in a specimen with impaired spermatogenesis. In conclusion, several CREM repressor transcripts are present in adult human testes, and altered transcript splicing is associated with impaired spermatogenesis.

Cloning and expression analysis of testis-specific cyclic 3', 5'-adenosine monophosphate-responsive element modulator activators in the nonhuman primate (Macaca fascicularis): comparison with other primate and rodent species

The cAMP-responsive element modulator (CREM) gene encodes a transcription factor that is essential for spermatogenesis. In mouse testis, several CREM repressors and activators have been identified. In contrast to the situation for the mouse, however, little is known about CREM isoforms in the primate testis. We analyzed CREM isoforms and mRNA expression in a clinically relevant primate model, the cynomolgus monkey (Macaca fascicularis). A cDNA library was generated from monkey testis; and two activator isoforms (tau2 with and without exon gamma) were identified, which displayed high sequence identity to mouse and human isoforms. The insertion of exon gamma was observed for the first time in the primate testis. CREM activator expression was confined to the testis, where it was seen in late pachytene spermatocytes and round spermatids in specific spermatogenic stages, as revealed by in situ hybridization. Comparison of the mRNA and the recently described protein expression indicated a lack of translational delay of CREM expression. Comparative analysis of testicular CREM expression by reverse transcription-polymerase chain reaction yielded several transcripts in the rat, mouse, hamster, and marmoset; two transcripts in cynomolgus and rhesus monkeys; and one transcript in men. These findings suggest an evolutionary trend from multiple activator isoforms to a single activator transcript in men.

Symptomatic intraspinal air entrapment

A 59-year-old man who had undergone the removal of a lipoma of the thoracic spine presented with progressive weakness of the lower limbs when lumbar puncture followed drainage of a subcutaneous collection of cerebrospinal fluid. Computed tomography showed entrapped intraspinal air which compressed the spinal cord. This rare, but serious complication can occur in a patient with altered intrathecal pressure following spinal surgery.

Round spermatids show normal testis-specific H1t but reduced cAMP-responsive element modulator and transition protein 1 expression in men with round-spermatid maturation arrest

During spermiogenesis, histones are replaced by transition proteins, which in turn are replaced by protamines. The TNP1 gene-encoding TP1 (transition protein 1) protein contains a cAMP-responsive element (CRE) that serves as binding site for the CRE modulator (CREM). To gain further insight into the complex regulation of nucleoprotein exchanges in haploid spermatids and its potential role for spermatogenic impairment, we studied the gene expression of testis-specific histone H1t, CREM, and TNP1 in testicular biopsies from men with normal spermatogenesis (n = 20) and with round spermatid maturation arrest (n = 16). During normal spermatogenesis, H1t messenger RNA (mRNA) was present in 86.2%+/-8.7% of pachytene spermatocytes (stages III-V), whereas H1t protein was synthesized in 83.5%+/-13.0% of pachytene spermatocytes (stages III-V) and persisted in 95.2%+/-3.1% of spermatids (steps 1-5). CREM mRNA was detectable in 74.2%+/-9.4% of pachytene spermatocytes (stages IV-V) and in 78.7%+/-10.0% of spermatids (steps 1-3). CREM protein was synthesized in 81.2%+/-14.2% of spermatids (steps 1-3). TNP1 mRNA was present in 80.0%+/-13.5% of spermatids (steps 2-4), whereas TP1 protein was synthesized in 89.7%+/-5.3% of spermatids (steps 3-4). In men with round spermatid maturation arrest, spermatids only develop to step 3 of differentiation. These spermatids were devoid of both CREM and TP1 but did contain H1t. These results indicate that TP1 is likely to be an important parameter in the histone-to-protamine exchange and in the initiation of spermatid elongation. CREM is involved in the regulation of TNP1 gene expression and consequently plays a vital role in the correct differentiation step from round spermatids to mature spermatozoa.

High-fat dairy product consumption increases delta 9c,11t-18:2 (rumenic acid) and total lipid concentrations of human milk

Conjugated octadecadienoic acids (18:2, conjugated linoleic acids) have been shown to be anticarcinogenic and may influence growth and nutrient partitioning. The delta 9c,11t-18:2 isomer (rumenic acid, RA) is most common in both food sources and human tissues. To determine if maternal diet can influence milk RA concentration, breastfeeding women (n = 16) were enrolled in a 3-wk crossover study. Women initially consumed minimal amounts of food containing RA during week 1, then were assigned randomly to consume diets rich in high-fat dairy foods (and thus RA) during week 2 or 3. Milk was collected by complete breast expression twice during each experimental week. Current and chronic RA intakes were estimated by 3-d dietary records and food frequency questionnaires, respectively. Estimated chronic RA intakes ranged from 49 to 659 mg/d. Dietary RA intake was greater during the high compared to the low dairy period (291 +/- 75 vs. 15 +/- 24 mg/d, respectively; P < 0.0001). Milk contained more RA during the high than the low dairy period (13.5 +/- 0.1 vs. 8.2 +/- 0.4 mumol/g lipid, respectively; P < 0.0001). Milk lipid concentration was influenced by diet, such that lipid concentration was greater during the high than the low dairy period (46.6 +/- 5.0 vs. 38.3 +/- 1.6 mg/g milk, respectively; P < 0.05). Additionally, multiple regression analyses suggested that body mass index was the primary predictor of milk RA and lipid concentrations. In summary, these data indicate that both lipid and RA concentrations of human milk can be influenced by diet.

Germ cell-specific cyclic adenosine 3',5'-monophosphate response element modulator expression in rodent and primate testis is maintained despite gonadotropin deficiency

cAMP response element modulator (CREM) is an important component of the cAMP-mediated signaling pathway and is essential for differentiation of haploid male germ cells. In the rodent, testicular expression of CREM is believed to be controlled by FSH. We studied the expression pattern of CREM and gonadotropic control in the nonhuman primate and rodent testis. Adult cynomolgus monkeys (Macaca fascicularis) received daily either vehicle or the potent GnRH antagonist (ANT) cetrorelix for periods of 25 and 56 days. Rats were also exposed to vehicle or ANT for periods of 14 and 42 days. ANT treatment suppressed pituitary gonadotropin secretion, reduced testis size, and altered spermatogenesis. A rabbit polyclonal antibody raised against recombinant CREM tau and reacting with CREM alpha, -beta, -gamma, -tau1, and -tau2 at similar affinities was used for immunocytochemistry and Western blotting. CREM expression was seen in round spermatids, with highest levels during spermatogenic stages V-VII, but declined with progression of spermatid development in the primate. Similar observations were made for the rat testis. Thus, CREM expression was maximal at the onset of acrosome formation and was low or undetectable upon initiation of spermatid elongation in both species. A weak, but specific, CREM signal was seen in mid- to late pachytene spermatocytes and during meiotic division in both species. After ANT exposure, the germ cell- and stage-specific pattern of CREM expression was quantitatively retained at all time points and in both species. Northern and Western blot analysis confirmed the maintenance of testicular CREM expression despite 25 days of ANT treatment. A retrospective immunocytochemical analysis of rat testes 14 days posthypophysectomy revealed CREM signals in round spermatids. These findings demonstrate that the testicular expression of CREM is not entirely dependent on gonadotropic hormones but, rather, on the maturational stage of haploid round germ cells.

Technical developments in intra-operative monitoring for the preservation of cranial motor nerves and hearing in skull base surgery

Although EMG recordings from mimic muscles have become the standard for intra-operative facial nerve monitoring, few data are available concerning other motor cranial nerves (MCN). Auditory brainstem responses (ABR) are a proven tool for intra-operative hearing preservation, but have their limitations, suggesting the application of supplementary methods. This paper describes new developments of MCN and cochlear nerve monitoring in skull base surgery. Up to 2 x 8 EMG channels were recorded after bipolar stimulation of MCN using concentric coaxial probes. A special software enabled event-dependent registrations of all signals exceeding a definable threshold level. Selective recordings from masticatory muscles (N.V) were obtained using rectangular Teflon-insulated needle electrodes. For oculomotor (Nn.III/ VI) nerve recordings bipolar needle electrodes were precisely placed by orbital ultrasound guidance. Lower cranial nerves were monitored inserting needle electrodes into the soft palate (N.IX), tongue (N.XII) and vocal muscles (N.X) during laryngoscopy using a special applicator. For ABR recordings, click stimuli (95 dB HL) were applied monaurally through insert earphones. Electrocochleography was simultaneously recorded as a near-field potential without averaging after promontory (transtympanic) electrode placement using otomicroscopy. Regarding the ABR biosignal, a characteristic response pattern was detected following bipolar electrical stimulation of the auditory nerve possibly useful for its intra-operative identification.

Heparin-coated equipment reduces the risk of oxygenator failure

BACKGROUND

The development of an abnormal pressure gradient (APG) across the oxygenator is the most common cause of oxygenator failure during cardiopulmonary bypass. This necessitated changing the oxygenator in 4 patients in this series. A retrospective analysis of conditions predisposing to APG was performed.

METHODS

One thousand nine hundred fifty-nine operations with cardiopulmonary bypass were performed in adults. A range of membrane oxygenators was used subject to availability; 769 oxygenators were heparin-coated and 1,190 were uncoated. The pressure gradient across the oxygenator was measured under standardized conditions. An APG was defined as a gradient of greater than twice the mean.

RESULTS

An APG occurred in 44 uncoated and 3 heparin-coated oxygenators (p < 0.001). The mean age was higher for the APG group (p < 0.001). Fibrin deposits in the arterial line filter were noted in 45 patients. Logistic regression revealed that only fibrin deposition in the arterial line filter and the use of uncoated oxygenators were significantly associated with APG.

CONCLUSIONS

We conclude that a heparin-coated oxygenator effectively prevents APG. This adds significantly to the safety of open heart operations.