Severe hypocalcemia and seizures after normalization of pCO₂ in a patient with severe bronchopulmonary dysplasia and permissive hypercapnia

Introduction

Alkalosis, both respiratory and metabolic, has been related to a decrease in total and ionic blood Ca (iCa²⁺) [1], which may condition, in severe cases, tetany and seizures [2]. In chronic patients, capillary blood samples are commonly used, although they seem to be less accurate in infants than in adults [3].

Case presentation

We present a 46 weeks postmenstrual age (PMA) male infant with a history of extreme prematurity and severe bronchopulmonary dysplasia. He was born at 25 weeks of gestational age (GA) and presented a mild respiratory distress syndrome, being resuscitated with non-invasive ventilation (NIV) and a maximum FiO₂ of 0.4. The initial evolution was favorable with NIV support and a FiO₂ between 0.21 and 0.35. Later on, he developed a hemodynamically significant patent ductus arteriosus that did not respond to two cycles of ibuprofen, requiring intubation and surgical closure at 26 days of life. Thereafter, he evolved in a torpid way, with four extubation failures, mainly due to hypoxemia, hypercarbia, and a significant increase in work of breathing, and three episodes of ventilator-associated pneumonia due to Pseudomonas Aeruginosa, Enterobacter Aerogenes and Klebsiella Pneumoniae. During NIV, hypercapnia was allowed, with capillary pCO₂ of up to 80 mmHg and pH >7.20. He received loop diuretics to decrease pulmonary edema and improve lung compliance until 43 weeks PMA, when they were withdrawn due to severe hyponatremia.

Food tolerance was good most of the time, although the patient developed osteopenia of prematurity. At 45 weeks PMA he was on full feed through orogastric tube. The levels of total Ca and Phosphorus were 9.4 and 4.2 mg/dL, respectively, and Alkaline Phosphatase was 991.0 U/L. At that time the patient was receiving oral supplements of Ca and P, and 1,200 IU of vitamin D Per os per day. Postnatal weight, length and head circumference growth were restricted (percentile <3).

At 46 weeks PMA, due to clinical and radiological worsening, with progressively higher FiO₂ requirements of up to 0.9, and arterial blood gas with pH 7.37, pCO₂ 97 mmHg, pO₂ 59 mmHg, HCO₃ ⁻ 56.1 mmol/L and BE +25 mmol/L, he was intubated, and assisted ventilation was initiated with PIP 25–30 cmH₂O, PEEP 7 cmH₂O, FiO₂ 60%, RR 50 rpm, with additional adjustment to maintain an ETCO₂ around 60 mmHg. At that time the iCa²⁺ was 4.4 mg/dL (Figure 1, Panel A), with intravenous (IV) infusion of calcium gluconate 300 mg/kg/day (27.9 mg/kg/day of calcium element). Capillary blood gas analysis 4 h after intubation showed pH 7.68, pO₂ 29 mmHg, pCO₂ 37 mmHg, HCO₃ ⁻ 43.7 mmol/L, BE +14.5 and iCa²⁺ 3.81 mg/dL. About 20 h after intubation, the patient exhibited hypertonic crisis, clonic jerks in the extremities and erratic eye movements. An infectious work-up screening was performed, and antibiotic and anticonvulsant treatments were started. A control blood gas showed pH 7.55, pCO₂ 42 mmHg, HCO₃ ⁻ 36.8 mmol/L, EB +12.9 mmol/L and iCa²⁺ 2.36 mg/dL. After optimizing sedoanalgesia, ventilatory support, and increasing the IV Ca infusion, the blood gas values and iCa²⁺ normalized, remitting the seizures.

Figure 1:

(A) Clinical and gasometrical evolution during the first 70 h after intubation (note that, for reasons of space and clarity in the graph, the time intervals are not homogeneous). (B) Comparison of the correlation between pH and iCa²⁺ in arterial and capillary blood samples. For the elaboration of this graph, 43 gasometrical controls (28 capillary and 15 arterial) were used from the patient fourth and fifth months of life, including the 70 h post-intubation described in the case report. The linear regression and the figure were carried out with the free software “R” (R Core Team [2020]. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL https://www.R-project.org/).

Discussion

At the beginning of the eighties, Watchko et al. [4] observed, in hyperventilated neonates with persistent pulmonary hypertension, an inverse relationship between pH and iCa²⁺, with a reduction of 0.42 mg/dL of iCa²⁺ for each increase of 0.1 units in pH. In that study, the patients were term neonates who were hyperventilated as a strategy to reduce pulmonary hypertension, with an average reduction of pCO₂ from 35 to 17 mmHg and an increase in pH from 7.39 to 7.59. These changes were accompanied by a mean reduction in iCa²⁺ from 3.7 to 2.6 mg/dL. The authors do not report acute neurological complications in this group of seven patients, although all were paralyzed and two died, presumably due to causes derived from their underlying condition. The reduction in iCa²⁺ observed was twice that reported for adults, but similar to other studies in neonates after bicarbonate infusion [5].

In our patient, the reduction in pCO₂ did not constitute a true hyperventilation, since the patient was taken to “normocapnia”. However, the speed of the gas exchange modifications could not be accompanied by the necessary metabolic adjustments, mainly at the renal level, so that, ultimately, a predominance of organic bases was produced, along with the corresponding imbalances at the ionic level [6]. It is well known that the changes in plasma iCa²⁺ secondary to metabolic pH modifications are more pronounced than those produced by variations in pCO₂ [2]. Despite little evidence, carbonic anhydrase inhibitors have been used to counteract diuretic-induced metabolic alkalosis in neonates with chronic respiratory failure [7]. The acute development of the events in our patient did not allow us to consider such therapy.

In addition, when we compared the correlation between the pH and iCa²⁺ in arterial and capillary blood gases (Figure 1, Panel B), we found that it was significantly greater in arterial samples, with the reduction in iCa²⁺ twice that of capillary samples for each increase of 0.1 pH units (0.8 vs. 0.4 mg/dL, respectively). Other studies have also shown higher iCa²⁺ values (+0.47 mg/dL) in capillary samples compared to arterial samples [8]. These findings suggest that, in chronic patients in whom arterial cannulation is usually no longer available, although arterial puncture could signify a greater discomfort to the patient, perhaps arterial blood gas should be of choice, occasionally, at times when important clinical changes are taking place, in order to anticipate or prevent severe complications.

Conclusions

Our findings highlight the need for close monitoring of acute metabolic changes to prevent the appearance of alkalosis and hypocalcemic tetany, by avoiding the abrupt reduction of pCO₂ in chronic patients with permissive hypercapnia. Although arterial samples are rarely necessary, they could be superior to capillaries to improve the precision of this control in selected cases.