The Acute Chest Syndrome in Cameroonian children living with sickle cell disease

This study showed that ACS accounted for 6.2 % of SCD children hospital admissions with almost 2.1 SCD patients presenting ACS per month. This proportion is comparable to the 1.9 ACS episodes/month among hospitalized SCD children in Bruxelles [15], and higher than the 1.4, 0.3 and 0.2 ACS episodes/month respectively reported in Brazzaville, Antananarivo and French Guiana [20‐22]. However, our prevalence of ACS is lower than the 10–20 % rate of hospital admissions reported by Miller and Gladwin in their review [6]. Our prevalence may be an underestimate of the real burden of ACS among our children suffering from SCD for some reasons. First, hemoglobin electrophoresis is not systematically performed in our milieu, and many parents neither know their status nor that of their kids. As we considered only known SCD patients, some unknown SCD patients may have presented this condition and escaped from enrolment. Second, due to parents’ financial constraints, we are used to performing just one chest x-ray during hospitalization. As a result, we could have missed all those whose first chest x-ray was normal but who developed ACS later on, given that radiographic findings in case of ACS may progress over time [9, 11, 23]. If it is true that a positive chest x-ray is the key element to define the disease [12], there have been some claims that a single negative chest x-ray cannot exclude the disease [23‐25], taking into account that clinical assessment may appear inadequate to identify ACS [11]. It has been also suggested that every SCD patient presenting with fever must undergo a chest x-ray [11, 24].

Consequently, close clinical monitoring alongside serial radiographic evaluation (without ignoring the risk of radiation) is necessary. In fact, it has been shown that ACS may develop 1 to 3 days after admission for VOC as there may exist a close relationship between ACS and VOC [6, 11, 15, 24]. This is truer for adults than children, as the latter usually present with ACS at admission [26]. In our study for instance, only 2 children (9.5 %) developed ACS 2 days after admission, and pains were concomitantly associated in 52.4 % of cases. Some risk factors involved in the development of ACS during hospital stay have been identified such as the male sex, past medical history of ACS, thoracic pains at entry and use of morphine during hospitalization [15, 27].

Although we did not assess risk factors of developing ACS, we observed a male predominance and relatively low levels of HbF, in line with the literature [6, 16, 17]. We also found that fever (90.5 %), cough (81 %) and thoracic pains (28.6 %) were the main symptoms in keeping with previous reports [9, 21, 28], these symptoms being mainly in favor of pulmonary infections. Despite the fact that we did not have our patients’ baseline leukocyte counts (explained by the fact that many of our patients are not regularly followed-up and are from financially-limited families), we noticed very high leukocyte counts [32479 (17862) mm³], especially neutrophils [23476 (11543) mm³], associated with elevated serum CRP levels [228.2 (132.6]. These elements are also strongly suggestive that infection may initiate or precipitate the development of ACS among our patients, corroborating previous reports elsewhere [15, 16]. However, though blood culture was not performed in all patients, no germ was identified. This result was also reported by Rucknagel et al. [25] who incriminated rib infarction as the main etiology of ACS. But without invasive procedures (bronchoscopy with broncho-alveolar lavage), it is very difficult to rule out conclusively a bacterial infection as a cause of ACS. Microorganisms that have been identified associated with ACS include Streptococcus pneumonia, Chlamydiae pneumonia, Mycoplasma pneumonia, influenza virus A H1N1, parainfluenza virus, respiratory syncytial virus and coronavirus among others [15, 21, 23].

Besides, we observed that radiological abnormalities of the lower lobes (90.5 %) were the prevailing ones, mirroring previous findings [23, 28]. Contrariwise, other studies showed that the main radiological localizations of ACS were the upper lobes in children, and the lower ones in adults [9, 20]. While the upper lobes predominance has been associated with an infectious etiology especially in children, the lower or multi-lobes predominance has been linked to pulmonary thrombosis and fat embolism [9, 16, 20] or rib infarction [25]. Our results are therefore suggestive that the pathogenesis of ACS in our setting is also multifactorial. Further well-designed studies with large sample sizes are warranted to better elucidate the etiology of ACS in our setting.

In line with the literature [6, 19, 22], our management of ACS included broad-spectrum antibiotherapy, hydration, analgesics, supplemental oxygen and transfusion, but none of these supportive care impacted the duration of hospital stay. Actually, only patients presenting respiratory distress were placed on oxygen, mainly due to limitation of resources. Likewise, we transfused only SCD patients who had a hemoglobin level ≤ 7 g/dl given that blood safety remains an issue of major concern in the milieu [29]. Although transfusion did not reduce the hospital stay, it significantly increased the hemoglobin level (p = 0.039), perhaps improving therefore the clinical state of our patients. Early transfusion of SCD patients presenting with ACS should be encouraged in SSA settings, especially transfusion of packed red blood cells instead of whole blood to reduce transfusion-related adverse reactions, despite recurrent blood shortages that occur in the region [30]. Nonetheless, more studies are needed to underpin this suggestion with robust scientific evidence and indicate which threshold should be considered to transfuse SCD children suffering from ACS.

The mean duration of hospitalization we found (6.8 days) is comparable to the 7 days-duration reported by Bertholdt et al. [15], but a bit higher than findings from Vichinsky et al. (5.4 days) [9] and lower than what has been reported by Vichinsky et al. in another study [23] and Hunald et al. [21]: more than 10 days and 12 days respectively. Introduction of other supportive care in our practice like bronchodilators and incentive spirometry as elsewhere [6, 19], along with systematic oxygen supplementation and early blood transfusion could substantially reduce the duration of hospital stay. Furthermore, the use of dexamethasone deserves some close considerations. In fact, there is body of evidence bolstering that low dose dexamethasone (0.3 mg/kg/12 × 4 doses) significantly diminished the duration of hospital stay, the need for blood transfusion, oxygen and analgesics use, and duration of fever, without a rebound effect [31] which was observed with high dose methylprednisolone in cases of VOC [32].

One of our patients died, hence a mortality rate of 4.8 % which concurs the 4 % percentage obtained by Bertholdt et al. in Belgium [15]. This was an inadequately-vaccinated and unregularly-followed female SCD patient aged 18 months, with a HbF level of 18.9 %. She was brought to consult for fever, cough and respiratory distress, and the chest x-ray revealed medium and lower right lobes alveolar consolidations. Titers of LDH and CRP were respectively 10366 IU/l and 414 mg/l. She was placed on a triple antibiotherapy (ceftriaxone, gentamycin and azithromycin), oxygen supplementation, and was transfused as well. But she died at day 4 of hospitalization in a context of severe sepsis.

Unfortunately, the retrospective design and small sample size constitute some weaknesses of the present study. Besides, generalization of our results may be impeded by the use of data collected from only one hospital centre. Another limitation is that biological data were incomplete due to parents’ financial limitations, as the study was not sponsored. Nonetheless, in this particularly difficult context, we carried-out serious and reliable data collection and analyses, and have shown that despite the difficulties in managing our patients, the outcome is relatively similar when compared with resource-rich settings in terms of duration of hospital stay and mortality rate. Eventually, and to the best of our knowledge, this is the first study conducted in our milieu targeting ACS burden and presentations.