Sickle related events following cardiac catheterisation: risk implication for other invasive procedures

Sickle cell disease (SCD) is an autosomal recessive disease caused by a single point mutation in beta-globin gene which results in the substitution of glutamic acid by valine. Morbidity and mortality in SCD arises from recurrent acute vaso-occlusive painful crises (VOC), acute chest syndrome (ACS), cerebrovascular events, splenic and renal dysfunction. Additionally, 6–11% of the sickle cell population have pulmonary hypertension (PH) (Fonseca, et al 2012, Parent, et al 2011, Sharma, et al 2013). Cardiac catheterisation (CC) is an important part of the evaluation of patients with SCD, and is performed to confirm PH, assess the hemodynamic effects of treatment, evaluate patients for a heart or lung transplantation, and for intracardiac biopsies (Callan and Clark 2016). Procedure related complications have been reported previously, but this report did not include patients with SCD (Hoeper, et al 2006). These complications consist of hematoma, pneumothorax, arrhythmias, hypotension, or vagal reaction; rarer complications include pulmonary infarction or pulmonary arterial rupture, from prolonged inflation of the balloon catheter in the wedge position. These event rates reported by Hoeper et al (adverse event rate, 1.1%; mortality rate, 0.05%) were lower than previously reported (1.7% and 4.2% respectively) (Fuster, et al 1984).

In patients with SCD, hemodynamic results of the CC have been previously reported (Gladwin, et al 2004, Mehari, et al 2013). However, the post-procedure rates of SCD related adverse events (SCD-AEs) have not been well described. We therefore performed a retrospective analysis of patients with SCD who underwent CC under National Heart, Lung, and Blood Institute institutional review board approved protocols from 2005 to 2016. CC typically included left and right heart catheterization under moderate sedation via transjugular, transfemoral, and/or transradial access, with nitric oxide vasodilator provocation for those with mean pulmonary artery pressures >25 mm Hg. Routine pre-procedure hydration were not implemented. Patients were monitored in post-anesthesia care unit and transitioned to a telemetry inpatient unit post-CC. Red blood cell (RBC) transfusion within one month prior CC, hemoglobin levels 1 week pre- and 1 week post-CC, use of medications (e.g. hydroxyurea, PH specific, or narcotics), length of hospital stay, and clinical events surrounding CC (e.g. pain, VOC, development of acute respiratory symptoms from pulmonary vasculopathy or cardiomyopathy) were captured from electronic medical records (Table 1, Supplemental table).

Table 1.

Clinical characteristics of patients undergoing cardiac catheterization

Variable	Total number of procedures (n=110) Total patients (n=76)
Median Age	46 years (range 15–84)
Gender  Male  Female	37 (48%) 39 (52%)
Repeat procedures	2 procedures: 11 3 procedures: 6 4 procedures: 2 5 procedures: 1
Disease phenotype  HbSS  HbSC  HbS/β0 thal  HbS/β+ thal	68 (89%) 6 (8%) 1 (1.3%) 1 (1.3%)
Medications  Pulmonary hypertension specific  Hydroxyurea  Long acting opioid  Short acting opioid	19 (25%) 44 (58%) 35 (46%) 48 (63%)
Red cell transfusion  No  Pre-procedure  Pre- and post-procedure	50 (66%) 18 (24%) 4 (5.3%)

HbSS, homozygous for sickle haemoglobin; HbSC, compound heterozygous for haemoglobin S and C; HbS/β0 thal, compound heterozygous for haemoglobin S and null expression β-globin; HbS/β+ thal, compound heterozygous for haemoglobin S and partial expression β-globin

76 patients underwent 110 CC procedures (Table 1). No post-CC hematoma, pneumothorax, arrhythmia, hypotention, vagal reaction, pulmonary infarction, pulmonary arterial rupture, or death were observed in the follow-up period. PH were diagnosed in a total of 19 patients (25%); in seven patients, PH was newly diagnosed through serial CC. These 19 patients received therapy with sildenafil (n=15), bosentan (n=1), sildenafil and bosentan (n=2), or a combination of tadalafil, treprostinil, and bosentan (n=1). Responses to PH therapies have been reported previously (Machado, et al 2011, Weir, et al 2017).

A total of 10 patients (13%) experienced 11 SCD-AEs (10%) within two weeks of CC. Twenty patients had two or more CC, four patients had one SCD-AE after one of the CC. Five of 19 patients with PH vs five of 57 patients without PH, had SCD-AEs post-CC. These complications also could be broadly separated into those having only pain or those having pain associated with respiratory symptoms (Table 2). Among the five patients experiencing only pain, three received short-acting opioids and another two received both short- and long-acting opioids prior to CC. None of them received RBC transfusion or were diagnosed with PH. In contrast, among the five patients having pain associated with respiratory symptoms, two received transfusions peri-CC and four patients were initiated on sildenafil therapy post-CC. These findings suggested that the respiratory symptoms could be related to underlying PH as previously reported (Rich, et al 1987).

Table 2.

Sickle-related exacerbations following cardiac catheterization

Hydroxyurea	PH medication	Hb level within 1 week pre-CC	RBC transfusion pre-CC	Hb level within 1 week post-CC	RBC transfusion post-CC	Length of stay >3 days	Clinical event
No	No	8.0	No	7.3	No	Yes	Pain
Yes	No	8.5	No	7.0	No	No	Pain
Yes	No	9.7	No	9.9	No	No	Pain
No	No	9.9	No	not reported	No	No	Pain*
No	No	10.8	No	10.6	No	No	Pain*
No	No	10.8	No	10.5	No	No	Headache
No	Post-CC	8.0	No	7.0	No	Yes	Pain, respiratory failure
No	Post-CC	8.2	No	8.0	No	No	Readmission 19 days post-CC for respiratory failure
No	Post-CC	8.2**	Yes	10.2	No	No	Pain, respiratory infection
Yes	Pre- and post-CC	9.3**	Yes	10.4	Yes	Yes	Heart failure, respiratory distress#
No	Post-CC	10.1	No	not reported	Yes	No	Dyspnea, chest pain

PH, pulmonary hypertension; CC, cardiac catheterisation; RBC, red blood cell; Hb, haemoglobin

^*These two separate events occurred in the same patient;

^**= on monthly exchange transfusions;

^#= on sildenafil after first CC 5 years prior to this CC

Seven patients (with seven procedures) had haemoglobin levels below 7 g/dL pre-CC, and none had SCD-AE post-CC. Three received RBC transfusions at physician discretion: one pre-, one post-, and one pre- and post-CC (data not shown). Another 35 patients had haemoglobin levels between 7.0 and 8.0 g/dL. Two of them experienced SCD-AEs, and none received RBC transfusions (Table 2). The remaining patients had haemoglobin >8.0 g/dL, and several patients experienced SCD-AEs post-CC. Although pre-procedure transfusion prior to CC did not appear to reduce SCD-AEs, most patients had haemoglobin levels of ≥7 g/dL, indicating this cutoff reflected the lower limit of the hemoglobin range typically observed in patients with SCD.

Next, we compared haemoglobin levels before and after CC in the 10 patients with post-CC exacerbations. Three patients had decreased haemoglobin levels, three had increased levels as a result of RBC transfusions, and the remainder had no change (Table 2). It was unclear if the change in haemoglobin levels contributed to the development of SCD-AE. The role of transfusion in close proximity to CC in altering cardiopulmonary dynamics for the eventual PH diagnosis was also unclear. Although a large number of CC were performed, this study represented the experience of a single center and uses a retrospective design to evaluate SCD-AE which limited the generalisability of these findings.

In summary, CC related complications that were directly attributable to the procedure itself were very low in patients with SCD but the rates of post-procedure SCD-AEs were 10%. The stress of CC could have triggered acute VOC or pain, especially in patients who were already taking narcotics. Likewise, patients with suspected PH or previous diagnosis of PH were more likely to develop respiratory symptoms following CC. Finally, it is reasonable to expect similar SCD-AE in SCD patients undergoing similar invasive procedures of short procedural duration like CC, such as central venous catheter placement, endoscopy, liver biopsy, arthroscopy, or oocyte retrieval.

To prepare patients with SCD prior to an invasive procedure, it seems prudent to formulate individualized treatment, including titration of hydroxyurea, consideration for pre- or post-CC RBC transfusion to minimize symptoms of anaemia, optimization of chronic pain management, and careful assessment of respiratory or other symptoms.

Figure 1:

Clinical symptoms, hydroxyurea treatment, and transfusion details

* 2 separate complications documented in the same patient, 4 months apart

** started sildenafil after CC

*** was taking sildenafil before CC and continued after CC

RBC, red blood cell; CC, cardiac catheterization; RF, respiratory failure; PH, pulmonary hypertension