Severe ischaemic gangrene of scalp in an extreme preterm: a fatal case of combined aetiology

Abstract

A 24+5-week preterm neonate with a severe scalp lesion was admitted to the neonatal intensive care unit (NICU) after caesarean section due to maternal chorioamnionitis (MC). An Arabin pessary had been inserted in addition to a previous cervical cerclage due to cervix insufficiency at 21+5 weeks of pregnancy (wp). At 23+5 wp, preterm rupture of membranes was evidenced. Both devices were kept to provide fetal viability. On 24+4 wp, she developed MC. Urgent caesarean section was performed. Transvaginal manual manipulation was required during the procedure. On NICU, she presented severe shock which required high-dose vasopressors and blood products. Following surgical repair, a bilateral grade IV intracranial haemorrhage was evidenced. Subsequently, it was agreed to withdraw life support. We hypothesise that MC and local infection could have acted as predisposing factors, with the presence of a pessary in the setting causing uterine contractions and its manipulation acting as a precipitating factor.

Background

Caesarean section, pessaries or vaginal progesterone is a generalised accepted intervention intended to prevent preterm birth in cases of maternal cervical insufficiency.

Complications of those devices are deemed to be low. In this article, we describe a very dramatic case of severe scalp lesion in an extreme preterm whose mother had Arabin pessary (AP) placed for cervical incompetence and posteriorly presented with preterm premature rupture of the membranes (PPROM) and maternal chorioamnionitis (MC).

We believe this article will help the reader to overweight the benefit and risks of those devices in this fragile population. We will carefully examine the difficult dilemma that arises specifically in ‘near-viability gestational age (GA) scenarios’, particularly in cases where suspicion of MC and 0ligohydramnios (OH) arise.

Case presentation

A 24+5-week preterm neonate was admitted to the neonatal intensive care unit (NICU) after emergency CS due to preterm labour (PL) and MC.

Pregnancy was closely followed up at obstetric clinic due to a history of recurrent mid-trimester spontaneous abortions because of cervix insufficiency. The mother was 29 years old and she was otherwise healthy. Laboratory test results were normal, and TORCH (Toxoplasma, Rubella, Citomegalovirus, Herpes simplex and Human Inmunodeficiency virus) screening was negative for acute infections; combined screening performed at 12 weeks of pregnancy (wp) revealed low risk of neural tube defects and low risk of chromosome 13, 18 and 21 aneuploidies.

On the 14th wp, cervical cerclage (CC) was performed due to a short cervix (length 27 mm) on cervical ultrasound (CUS). At 21+5 wp, she was admitted due to progressive shortening of cervix on CUS with amniotic sac prolapse across the internal cervix orifice. Vaginal and cervical swabs yielded negative microbiological results. An AP device was inserted with good response. Vaginal progesterone and indomethacin were prescribed. The patient remained asymptomatic until 23+5 wp when PL along with PPROM was evidenced. She was then started on intravenous antibiotics, atosiban infusion and intramuscular betamethasone. Decision to keep both CC and AP in place was taken due to extreme low GA and was agreed by the mother. She was closely clinically monitored with daily blood tests and fetal scans. The initial amniotic fluid index (AFI) was 5. On 24+4 wp, she developed mild fever, uterine contractions, sero-hematic vaginal discharge along with leucocytosis and raised C reactive protein (CRP). AFI dropped to 1, indicating severe OH. Fetal presentation was evaluated as cephalic. Neuroprotection with magnesium sulfate was administered. At 24+5 wp, a cephalic right upper limb presentation was noted on vaginal examination across the AP; therefore, urgent CS was indicated. During the procedure, following removal of the device, transvaginal manual manipulation by a third operator was required to allow extraction of the fetus across caesarean surgical incision.

A newborn girl baby was born with no respiratory effort and pallor; her heart rate was 80 beats per minute. Initial positive pressure ventilation was required, with fractional inspired oxygen increasing up from 30% to 100%. Apgar scores at minutes 1, 5 and 10 were 6, 8 and 8 respectively. At 3 min, she started irregular breathing with normalised heart rate. However, she was noted to be extremely pale. At that time, a severe scalp lesion with exposure of almost entire right calotte was observed. She was then intubated and transferred to NICU.

On admission, she was connected to mechanical ventilation on moderate settings; an umbilical vein catheter was inserted, and norepinephrine infusion was started and titrated up to 0.5 μg/kg/min to maintain adequate mean blood pressures. Admission venous blood gas showed a pH of 7 with partial pressure of carbon dioxide (Pco₂) of 7.6 kPa and lactate of 10 mmol/L. Blood results revealed a severe anaemia with haemoglobin of 65 g/L and haematocrit of 22.7%, with a platelet count of 115×10⁹/L. Clotting tests showed a prothrombin time (PT) of 20.8 s, activated partial thromboplastin time (aPTT) of 93.7 s, Quick Index of 41% and fibrinogen of 283 mg/dL. Packed red cells and frozen plasma were urgently transfused.

White cell count was 41.2×10⁹/µL, with a neutrophil count of 30.2×10⁹/µL. CRP was 105 mg/L and procalcitonin was 19.3 ng/mL. Intravenous cefotaxime and amikacin were started. Admission swaps and blood cultures were taken and yielded negative microbiological results.

The scalp lesion (figure 1) extended to 70% of right parieto-temporo-occipital bone surface; skin borders were clean but friable and were carefully sutured by the surgical team with good result (figure 2). Following surgical repair, an urgent brain ultrasound was performed, which evidenced a bilateral grade IV intracranial haemorrhage (figure 3).

Figure 1.

Scalp lesion on admission.

Figure 2.

Scalp lesion following surgical repair.

Figure 3.

Head ultrasound showing grade IV intracranial haemorrhage.

Investigations

Admission venous blood gas showed a pH of 7 with Pco₂ of 7.6 KPa and lactate of 10 mmol/L. Blood results revealed a severe anaemia with haemoglobin of 6.5 g/dL and haematocrit of 22.7%, with a platelet count of 115×10³/µL. Clotting tests showed a PT of 20.8 s, aPTT of 93.7 s, Quick Index of 41% and fibrinogen of 283 mg/dL.

White cell count was 41.2×10³/µL, with a neutrophil count of 30.2×10³/µL. CRP was 105 mg/L and procalcitonin was 19.3 ng/mL.

Urgent brain ultrasound was performed by the neonatology and radiology team, which evidenced a bilateral grade IV intracranial haemorrhage

Differential diagnosis

Differential diagnosis would include aplasia cutis, epidermolysis bullosa, linear scleroderma ‘sword stroke’ and focal dermal hypoplasia (Goltz syndrome). In our opinion, the suggested mechanism of injury matches the described medical history and clinical signs noted.

Outcome and follow-up

In light of the bleak neurological prognosis in combination with extreme prematurity and severe septic and haemorrhagic shock, the decision to withdraw intensive care support was taken and ultimately accepted by parents.

Discussion

CC, AP or vaginal progesterone is a generally accepted intervention intended to prevent preterm birth in cases of maternal cervical insufficiency. The effectiveness and safety of those treatments either alone or in combination have been tested in multiple prospective studies, clinical trials and meta-analysis.^1–6 However, some recent meta-analyses have found no significant benefit of these therapies towards preterm birth prevention.^7–9 It is beyond the scope of this article to explore this matter. Its complications are deemed to be low^{5 6} and are normally limited to mechanical disturbances and local or generalised infections.¹⁰

There are several factors that may have contributed to the development of such a severe scalp in our patient. First, uterine pressure can reach up to 80–100 mm Hg during the second stage of labour.¹¹ This pressure is normally well tolerated, but, especially in cases of prolonged labour, mild contusions, abrasions, fat necrosis, pressure necrosis and ischaemic gangrene can occur in term newborns.⁵

Morykwas et al¹² and Matthews¹¹ reported two cases of full-thickness scalp necrosis after prolonged unsupervised labour. Complete ischaemia resulted in a full-thickness loss of scalp within the compression ring. We believe that the occurrence of scalp in our patient may not be solely explained by 24 hours of labour period alone.

Second, local infection related to prolonged PPROM and MC may have also played an important role. Kulshrestha et al¹³ reported two cases of ischaemic gangrene of scalp in two neonates with PPROM; in one of the cases, Kleibsella sp was isolated from pus discharged from scalp. Kalra et al¹⁴ reported a case of a large scalp defect secondary to scalp cellulitis and osteomyelitis of skull bones. Sadly, no swap sample was taken from the scalp wound in our patient. Nevertheless, despite the fact that maternal placenta culture and neonatal swaps and cultures were negative, the presence of MC and raised neonatal inflammatory markers may indicate a contribution of local subcutaneous scalp tissue infection increasing subdermal fragility, promoting its disruption in response to minimal trauma.

Finally, the presence of AP, acting as a foreign body against which the cranium was forced by unstoppable labour uterine pressure and requirement of transvaginal manipulation, undoubtedly played a major role as a primary mechanism of action. Severe OH probably contributed to an increased pressure effect of fetal cranium against AP with subsequent development of subcutaneous ischaemia.

Following a deep bibliography review, we have only found one similar case¹⁵ of a 33-year-old pregnant woman with cervix insufficiency in which AP was placed at 22+1 weeks of gestation (wg). PPROM along with PL occurred at 24+4 wg; antibiotics and atosiban were started. The device was not retrieved until 26+4 wg when persistent PL and severe OH were identified. In this case, delivery was vaginal. It is noticeable that authors refer to difficulties while extracting the AP and acknowledge some degree of local traumatic effect during extraction. Despite there being no bacterial growth from the area of scalp, the placenta yielded positive culture for Escherichia coli. In our case, unfortunately cultures were not taken by the surgical team during closure and parents did not consent to necropsy.

Regarding prognosis, most case reports communicate general good outcomes regarding local healing of scalp lesions and neurological outcomes. Sadly, our case presented with severe intracranial haemorrhage, coagulopathy and septic shock, leading to fatal consequences; the scalp lesion was only part of a myriad of complications of a severe vertical septic picture in an extreme preterm.

It is difficult not to question whether CC and AP should have been removed earlier. Current recommendations suggest removing these devices at 37 wg and in cases of uncontrolled preterm labour, PPROM or active vaginal bleeding.^{1 15} However, in cases of fetal GA below 24 weeks, considered as non-viable in some centres, this decision becomes significantly complex.

This case was profusely discussed in perinatal multidisciplinary meeting (MDT), bearing in mind the risks and benefits of maintaining or removing the device. Based on initial non-viable GA and the absence of neither raised infection markers nor severe OH, AP was left in place to avoid the delivery of a non-viable fetus. The mother was adequately informed; she understood the risks and consented.

It is important to note that there has been previous successful experiences in our centre with three cases with similar maternal settings in which AP use successfully bridged non-viable fetuses towards a prolongation of delivery up to 25–26 wg, achieving optimal neonatal outcomes. In this case, as preterm labour reappeared and MC was evidenced, CC and AP were urgently removed. Nevertheless, in the light of this fatal outcome and after a careful review of the literature, we would now advocate for an earlier removal of AP provided that fetal maturation and neuroprotection have been completed and a ‘viable GA’ has been achieved (this approach might differ between centres). Particularly in cases of MC and OH, these devices should be immediately and carefully retrieved as soon as a decision of fetal delivery or extraction is taken.

Patient’s perspective.

We fully understood the risk of obstetric strategy and felt that the perinatal team always advised us based on our baby’s best interest. I can only feel gratitude to all neonatology team, as they showed a high degree of human values, especially when they allowed me to come to neonatal intensive care unit on my own bed and hold my baby and cuddle her at the end of her short life. I am also thankful to the obstetric team for their loving and committed professional care during the postnatal period at the obstetric unit

Learning points.

• Severe obstetric gangrenous scalp injury is a rare complication and is rarely described in the literature.

• Its aetiology can be multifactorial: prolonged labour, preterm premature rupture of the membranes along with maternal chorioamnionitis (MC) and local fetal cephalic subcutaneous tissue infection could act as a predisposing factor; the presence of an Arabin pessary in the setting of severe 0ligohydramnios (OH) against which uterine contractions may force a vulnerable preterm cranium would play a major role, and finally the manipulation of the device in these settings could precipitate the occurrence of the scalp.

• Careful and close follow-up and early removal of those devices once a ‘viable gestational age’ of fetus has been achieved are advisable. These devices should be immediately retrieved in cases of high suspicion of maternal chorioamnionitis and OH.