Excessive bleeding is a known complication of surgery in patients with osteogenesis imperfecta. Disseminated intravascular coagulation (DIC) may result from many factors including sepsis, fat embolus, transfusion reactions and obstetric catastrophe. In the case described the aetiology of fibrin consumption is unclear and may have been a result of a combination of factors. We suggest that the dramatic cutaneous manifestation may be because of DIC in association with the collagen defect underlying the bleeding disorder in this patient.
A 32-year-old woman with osteogenesis imperfecta type III complicated by invagination of the base of the skull by the cervical spine (basilar impression) presented for correction of a progressive cervico-thoracic kyphosis sufficiently advanced to prevent normal mouth opening; the interdental gap was 1.2 cm. Four years previously she had developed signs of brainstem compression and had undergone surgery. At this time an elective tracheostomy was deemed necessary, prior to anterior decompression via a Le Fort maxillary osteotomy, with excision of the odontoid peg and posterior occipito-cervical fusion with a Ransford loop in order to stabilize the cervical spine.
On this occasion there was insufficient access to carry out elective tracheostomy. The patient was unable to exhale through the nose but she was able to inhale through the right nostril. Following anaesthesia of the nasal mucosa by the application of cocaine hydrochloride (10%) solution on cotton pledgets, the patient was sedated with midazolam (10 mg) and awake fibreoptic intubation was attempted. It was not possible to pass the fibreoptic laryngoscope which had a diameter of 6 mm through the right nostril into the nasopharynx (although passage of an 18-gauge nasogastric tube through the right side was possible). Oral fibreoptic intubation was then attempted. The oropharyngeal mucosa was anaesthetized with lignocaine 4% solution introduced through the port of the fibreoptic laryngoscope, but the glottis could not be visualized. The patient became increasingly distressed and further sedation was maintained with an infusion of propofol (0.5 mg kg−1 h−1). With the patient breathing spontaneously, a size 3 laryngeal mask was inserted with some difficulty, the tube being compressed to pass it through the interdental gap. The tube was then split immediately proximal to the connection to allow insertion of the fibreoptic laryngoscope whilst the patient continued to breathe a 50% oxygen/air mixture. Intubation was eventually achieved using a size 6 tracheal tube, the entire intubation procedure taking over 2 h.
Total intravenous (i.v.) anaesthesia was then induced with fentanyl (0.2 mg), propofol loading dose (1.0 mg kg−1) and propofol in fusion (initial rate 10.0 mg kg−1 h−1). Invasive monitoring was instituted; arterial pressures were transduced from a 22-gauge cannula in the right radial artery and central venous pressures from a triple lumen catheter in the right femoral vein.
The patient was placed prone and surgery proceeded with full spinal cord monitoring. The propofol infusion rate was reduced to 5 mg kg−1 h−1 after 30 min and thereafter adjusted according to clinical needs. During surgery additional fentanyl (0.05 mg) was administered as appropriate. Cervico-thoracic osteotomy at the level of T1 was carried out and further posterior fusion of the spine to T3 was achieved by 'piggy-backing' a Hartshill rectangle onto the existing Ransford loop and bone grafting from the iliac crest.
Surgery lasted 90 min during which time the measured blood loss was 3300 mL; the patient's estimated blood volume was 3200 mL. Peroperative fluid replacement consisted of Hartmann's solution (2000 mL), hydroxyethyl starch (1500 mL) and packed red cells (5 units). Her cardiovascular system was stable on arrival in the intensive care unit (ICU) where her lungs were electively ventilated mechanically. Four units of fresh frozen plasma were given empirically as measured blood loss exceeded the patient's estimated circulating volume. At this time her blood pressure was 110/60 mmHg, CVP was 4 mmHg and heart rate 90 beats min−1. Her temperature measured from a rectal probe was 36.4°C and arterial blood gas results were as follows: pH 7.31, PaO2 11.1 kPa, PaCO2 5.9 kPa, BE 3.0 mmol L−1, and SpO2 94% (FiO2 40%). Two hours after arriving in the ICU her blood pressure started to decrease and she became tachycardic, heart rate increasing to 120 beats min−1. Considerable oozing from the cervical incision and iliac crest was apparent and the patient looked pale. The haemoglobin concentration, which had been 9.8 g dL−1 on arrival in the ICU had decreased to 4.9 g dL−1 and the patient was given 4 units of ORh negative packed cells as there was no time to crossmatch blood. As bleeding continued from both surgical incisions, secretions from the endotracheal tube and nasogastric tube became bloodstained and there was frank epistaxis. This was accompanied by the sudden development of florid petechiae, peri-orbitally and across the shoulders initially, but extending down the torso within minutes (Fig. 1). A clinical diagnosis of DIC was made, and this was vigorously treated with cryoprecipitate, fresh frozen plasma and platelets, as well as packed red cells. This diagnosis was confirmed by coagulation studies: prothrombin time 36 s (control 15 s), KCT 103 s (control 35 s), fibrinogen 0.73 g L−1 (normal range 2–4 g L−). The platelet count had decreased from 95 × 109 L−1 to 5 × 109 L−1. An infusion of dopamine at 1 μg kg−1 min−1 was started to maintain renal perfusion but her blood pressure continued to decrease. Inotropic support was required and an infusion of adrenaline at 0.01 μg kg−1 min−1 was commenced and adjusted as necessary. Within 12 h coagulation indices were normal, PT 15 s, platelet count 101 × 109 L, and no further inotropic support was required. Eighteen hours post-operatively chest radiography showed bilateral pulmonary infiltrates in keeping with a diagnosis of adult respiratory distress syndrome (ARDS) and required 70% oxygen to maintain a PaO2 of 8.4 kPa (pH 7.28, PaCO2 6.54, BE 3.5 mmol L−1).
In spite of these complications, the patient began to improve within 3 days, with radiological and clinical evidence that the ARDS was resolving. Eight days post-operatively the patient's trachea was extubated and 20 days post-operatively she was discharged home.
Osteogenesis imperfecta is an inherited connective tissue disorder, a result of abnormal Type 1 collagen synthesis [1,2]. Type 1 collagen is the predominant protein of bone matrix and the cardinal clinical feature of osteogenesis imperfecta is osteopenia associated with recurrent fractures and skeletal deformity. Although classically osteogenesis imperfecta is said to comprise a triad of brittle bones, blue sclerae and deafness, there is considerable variation in the severity of the condition and a large number of associated clinical findings [3,4]. Given that the gene for Type 1 collagen is one of the largest genes sequenced, there are many potential sites for mutation and therefore the potential for considerable phenotypic variability . A classification system has been devised by Sillence  based on mode of inheritance (dominant or recessive), age of onset (congenital or late), abnormal dentition and other characteristics, and has provided a framework for prognosis and a basis for further biochemical studies [2,4].
A subgroup of patients with osteogenesis imperfecta develop basilar impression, or invagination of the base of the skull by the cervical spine [7–9]. This results in a 'tam-o'-shanter' skull, or 'crane a rebord' and a characteristic appearance; domed forehead, temporal bulge, overhanging occiput and protuberant mandible. The deformity of the skull base results in neurological deficit secondary to brain stem compression, distortion of the vertebral arteries and hydrocephaly. Although a number of other clinical conditions are associated with basilar impression , in osteogenesis imperfecta the condition tends to be progressive in spite of surgery [7,11,12].
This patient was classified as having osteogenesis imperfecta Type III with progressively deforming bones although she had neither multiple fractures nor blue sclerae. She was of short stature (150 cm) exacerbated by her kyphosis, and normal dentition. She was deaf in her left ear and complained of increasing deafness on the right. She had the typical facies of basilar impression in osteogenesis imperfecta with short neck, overhanging temporal and occipital bones and prominent mandible. Four years previously she had developed signs of brainstem compression with episodes of tetraparesis lasting several hours and associated with nausea, vomiting, nystagmus and ataxia. Her surgical management at this stage has already been discussed in the literature . Her neurological deficit following initial surgery took the form of mild tetraparesis and pathologically brisk reflexes but no sensory deficit. Despite this surgery, which had served to prop up the skull at the back, the face had continued to drop at the front; this, in association with the kyphosis developing at the cervico-thoracic junction, had led to severely restricted mouth opening as her chin was arrested by the sternum and the left mandible impinged on the clavicle. For 2 years the patient had been unable to eat solid food and she was limited to taking liquidized meals through a straw. The second surgical intervention was undertaken to improve mouth opening and her ability to eat normally. Whilst reviews of anaesthesia for osteogenesis imperfecta suggest the possibility of difficulty in tracheal intubation , none describe the anatomical problems of basilar impression, the lack of access to the trachea for tracheostomy, or suggest the degree of difficulty exhibited by our patient. The possibility of cervical or mandibular fracture add to the hazards of attempting intubation in a patient whose prominent occiput makes positioning difficult.
Bleeding problems are a well recognized feature of osteogenesis imperfecta and there are several reports of excessive bruising [14–16] and of slow generalized oozing from wound sites following surgery [17,18]. It has been reported that up to two thirds of patients with osteogenesis imperfecta exhibit easy bruising and of these, increased capillary fragility occurred in 35%, decreased platelet retention occurred in 23% and decreased factor VIII occurred in 23%. In addition, there was deficient collagen-induced platelet aggregation in a number of patients. Morton  reported excessive bleeding in a patient with osteogenesis imperfecta who underwent Le Fort osteotomy; although that patient reported bruising easily, no bleeding problems had occurred during previous surgery. Morton concluded that, in the light of the absence of any detectable abnormality being identified in his, and other reported cases, the underlying collagen defect in osteogenesis imperfecta would result in friable tissues, small blood vessels which are unable to constrict adequately and defective platelet response and would be sufficient to produce excessive bleeding at operation. The initial response to blood vessel damage is constriction and platelet plug formation, the exposed collagen attracting platelets which release ADP which causes further platelet aggregation. Abnormalities at this stage of clotting have been reported frequently in osteogenesis imperfacta [14–17]. Morton  was able to exclude DIC as a cause of bleeding in his patient and certainly this is not a common occurrence in osteogenesis imperfecta—although a single fatal case of DIC occurring during spinal surgery has been reported . Petechiae and ecchymoses are not a feature of peri-operative bleeding in osteogenesis imperfecta although a case of spontaneous recurrent purpuric episodes occurring in a patient with osteogenesis imperfecta has been reported . That patient was found to have abnormal platelet function but a normal pattern was obtained on thromboelastography.
There was no report of disproportionate bleeding during her first surgical procedure and our patient denied any history of easy bruising. Persistent oozing was seen in the early post-operative period, the extent of which was initially underestimated. The cause of the bleeding diathesis was probably multifactorial and certainly, the incriminating factors in the development of DIC in this patient were numerous. Although the exact cause of DIC in this patient could not be identified, it is important to state that incompatibility of transfused red cells was specifically excluded. Petechial haemorrhage is not a manifestation of acute DIC. The administration of blood and aggressive blood component therapy served to limit the condition before irreversible end organ damage occurred.
This case demonstrates the anaesthetic problems, surgical management and post-operative complications of patients with osteogenesis imperfecta. Bleeding diatheses of any aetiology may occur and this should be anticipated so that aggressive management may be instituted early and avert fatality.
1 Hollister DW. Molecular basis of osteogenesis imperfecta
. Curr Probl Dermatol
2 Byers PH. Disorders of collagen biosynthesis and structure. In: The Metabolic Basis of Inherited Disease,
6th Edn. New York: McGraw Hill, 1989.
3 Sillence D. Osteogenesis imperfecta
: An expanding panorama of variants. Clin Orthopaed
4 Smith R, Francis MJO, Houghton GR. The Brittle Bone Syndrome: Osteogenesis Imperfecta.
London: Butterworths, 1983.
5 Solomons CC, Millar EA. Osteogenesis imperfecta
—new perspectives. Clin Orthopaed Rel Res
6 Sillence DO, Senn A, Danks DM. Genetic heterogeneity in osteogenesis imperfecta
. J Med Genet
7 Harkey HL, Crockard HA, Stevens JM, Smith R, Ransford AO. The operative management of basilar impression in osteogenesis imperfecta
8 Frank E, Berger T, Tew JM. Basilar impression and platybasia in osteogenesis imperfecta
tarda. Surg Neurol
9 Turek SL. Orthopaedics,
3rd Edn. Philadelphia: Lippincott, 1977: 309–313.
10 Taylor AR, Charkavorty BC. Clinical syndromes associated with basilar impression. Arch Neurol
11 Kurimoyo M, Ohara S, Takaku A. Basilar impression in osteogenesis imperfecta
tarda. J Neurosurg
12 Pozo JL, Crockard HA, Ransford AO. Basilar impression in osteogenesis imperfecta
: a report of three cases in one family. J Bone Joint Surg
13 Partridge BL. Skin and bone disorders. In: Katz J, Benumof J, Kadis L, eds. Anaesthesia and Uncommon Diseases,
3rd Edn. Philadelphia: W.B. Saunders Co., 1983: 685–686.
14 Siegel BM, Friedman A, Schwartz SO. Haemorrhagic disease in osteogenesis imperfecta
. Am J Med
15 Solomons CC, Hathaway WE, Millar EA. Platelet metabolism in osteogenesis imperfecta
. J Bone Joint Surg
16 Endo Y, Mamiya S, Nitsu H. Haemorrhagic diathesis in van der Hoeve's syndrome. Arch Intern Med
17 Sperry K. Fatal intraoperative haemorrhage during spinal fusion surgery for osteogenesis imperfecta
. Am Forensic Med Pathol
18 Morton MG. Excessive bleeding
after surgery in osteogenesis imperfecta
. Br J Oral Maxillofac Surg