Results and discussion

Among 22 immunocompromised patients treated due to strongyloidiasis, five (22.7%) developed SHS. All were autochthonous males at a median age of 64 years (37–81), residing in central continental Croatia, including four from rural areas, with no travel history. The basic demographic features and history of the patients are shown in table 1.

Table 1. Demographic characteristics and medical history of patients with Strongyloides stercoralis hyperinfection treated at the University Hospital for Infectious Diseases ‘Dr. Fran Mihaljević’ from January 2010 to June 2021.

^a The fifth R-CHOP (rituximab, cyclophosphamide, doxorubicin, vincristine, prednisone) regimen – ten days before symptoms.

^b Methylprednisolone was discontinued four weeks before the onset of symptoms which started six days after resection of the small bowel due to stenotic and fistulizing Crohn's disease.

^c Maintenance therapy.

The median symptom duration before diagnosis was 1.5 (0.5–3) months. Chronic haematologic patients predominated. All patients were anti-HIV and the latest SARS CoV-2 negative. Patients were not tested for HTLV-1. Clinical features of patients are shown in table 2.

Table 2. Clinical characteristics of patients with Strongyloides stercoralis hyperinfection treated at the University Hospital for Infectious Diseases ‘Dr. Fran Mihaljević’ from January 2010 to June 2021.

^a All patients were anorexic.

^b Until death.

All patients were described as cachectic, had anorexia and nausea, and recorded a recent weight loss ranging from 5 to 30 kg. However, four out of five patients were severely hypoalbuminemic (serum albumin range: 13.3–27.7 g/l). Nonspecific gastrointestinal symptoms were present in four out of five of the patients, which progressed to ileus in two of them. All patients had secondary infections caused by faecal microbiota. All bacterial isolates were susceptible to routinely tested antibiotics, and all patients were treated with adequate antimicrobial therapy. Patients 3 and 5 were treated in the intensive care unit, and patient 5 was mechanically ventilated and dialysed. E. coli was isolated from his blood and cerebrospinal fluid (CSF), and CSF examination was consistent with bacterial meningitis. Although no eosinophils or parasite larvae were found in the CSF, disseminated strongyloidiasis in the central nervous system (CNS) was suspected. This patient with severe consciousness impairment (GCS 7) died of a massive bilateral ischaemic cerebrovascular accident after 20 days of antimicrobial and 14 days of combined antiparasitic treatment without clinical improvement.

In all patients, strongyloidiasis was diagnosed accidentally, by analysing the upper gastrointestinal biopsy sample or bronchial aspirate. The concentration method subsequently used for the detection of cysts and ova in three stool samples proved positive in four out of five patients. In three deceased patients, motile rhabditiform larvae were seen in the native wet-mount stool samples, even during antiparasitic treatment, which was not the case in the surviving patients. Their abundant presence was recorded during antihelmintic therapy in a stool sample from patient 5 (see supplementary video 1). Patients 3 and 5 were serologically positive for S. stercoralis. Peripheral blood eosinophilia was absent in two out of five patients, both of whom died. Despite combined antibiotic, antifungal and antihelmintic therapy, three out of five (60%) patients died.

All of our patients confirmed contact with soil in their medical history – four older patients lived in rural areas where they engaged in gardening and the youngest urban patient was probably exposed to soil during playtime with his pet dog on the ground in suburban surroundings. Changes in agricultural activities and improvement in environmental and personal sanitary conditions have contributed to human strongyloidiasis becoming a very rare disease in the region. However, recently diagnosed primoinfections confirm current endemicity of this parasite in the soil in central continental Croatia (Balen Topić et al., Reference Balen Topić, Marjanović, Tomasović and Sviben2021). Therefore, immunosuppressed residents and foreign travellers to this region should be advised to take contact precaution measures while patients undergoing planned immunosuppression and organ donors from Croatia should be screened for strongyloidiasis.

The long symptom duration before diagnosis and the accidental finding of the parasite without previous clinical suspicion in our immunosuppressed patients suggest that clinicians are unfamiliar with the SHS/DS and that severe disease forms may be underdiagnosed. But could clinical suspicion even be raised early in a region with a very low incidence of symptomatic strongyloidiasis?

According to our results, although aggravating gastrointestinal symptoms predominated, they were not specific enough to raise an early clinical suspicion. It seems that recent significant inappetence, weight loss and consecutive hypoalbuminemia are the most common clinical signs in gradually accelerating strongyloidiasis, progressing to SHS and secondary infections caused by gut flora.

A significant diagnostic obstacle, especially important in most severe cases, could be the absence of the peripheral blood eosinophilia, recorded in two out of three of our deceased patients. Besides deflection of clinical suspicion from parasitosis, leading thereby indirectly to delay in diagnosis, many reports have shown that lack of eosinophilia presents an ominous sign in strongyloidiasis, as it reflects weak or deficient Th2 response. The pivotal role of eosinophil-mediated cytotoxicity against filariform larvae of S. strongyloides has been visualized in vivo in the sputum of patients with mantle cell lymphoma and SHS (Incani et al., Reference Incani, Hernández and González2010). The enzyme immunoassay detecting S. stercoralis serum antibodies with reported sensitivity of up to 95% in the general population should be used as a screening tool in immunosuppressed patients, too. However, the sensitivity of this diagnostic method in immunosuppressed patients with severe SHS/DS is unknown. As observed in one of the three of our tested patients, cases of most severe forms of strongyloidiasis in immunosuppressed patients with negative serology can be found in the literature (Rodriguez et al., Reference Rodriguez, Abraham and Williams2015; Lier et al., Reference Lier, Tuan, Davis, Paulson, McManus, Campbell, Peaper and Topal2020b). In immunosuppressed patients, the stool microscopic examination (including the Baermann test) can also reveal negative results (Rodriguez et al., Reference Rodriguez, Abraham and Williams2015). It seems that in immunosuppressed patients, even if there was an early suspicion, clinicians could be easily misguided by negative results of common laboratory results and screening tests.

In a setting of frequent corticosteroid use during the COVID-19 pandemic, the diagnostic circumstances may be further complicated, primarily by the SARS-CoV-2-mediated eosinopenia, which was found in 71.7% of COVID-19 patients with pneumonia (Xie et al., Reference Xie, Ding, Han, Yin, Lu and Zhang2021). A case of vanished chronic eosinophilia in a patient with chronic strongyloidiasis, during acute COVID-19, before introducing corticosteroid therapy has been reported (Stylemans et al., Reference Stylemans, Van Cauwelaert, D'Haenens and Slabbynck2021). A case of severe DS during acute COVID-19, with zero eosinophils at admission and negative initial serum antibodies and stool analyses for ova and parasites, has been described (Lier et al., Reference Lier, Davis and Topal2020a). A nationwide survey in Spain showed that only 18% of the 121 centres included screening for strongyloidiasis in SARS-CoV-2 patients. There were 227 strongyloidiasis cases identified, among which four developed SHS and one patient died (Rodríguez-Guardado et al., Reference Rodríguez-Guardado, Álvarez-Martínez, Flores, Sulleiro, Torrús-Tendero, Velasco and Membrillo2021). However, most of the reported SHS/DS developed during the COVID-19 reconvalescence phase – on average 22.8 days and up to two months after hospital discharge in patients previously treated with corticosteroids (Gautam et al., Reference Gautam, Gupta, Meher, Siddiqui and Singhai2021; Pereira et al., Reference Pereira, Mastandrea, Medeiros, Gryschek, Paula and Corral2021). To accelerate diagnosis, we agree with the authors who emphasize the value of an early upper (and sometimes even lower) gastrointestinal endoscopy and microscopic examination of gastrointestinal and/or respiratory tract samples in diagnosing SHS (Al-Sajee & Al-Hamdani, Reference Al-Sajee and Al-Hamdani2010; Rodriguez et al., Reference Rodriguez, Abraham and Williams2015; Lier et al., Reference Lier, Tuan, Davis, Paulson, McManus, Campbell, Peaper and Topal2020b; Bdioui et al., Reference Bdioui, Bchir, Missaoui, Hmissa and Mokni2021).

Besides diagnostic issues, there are many therapeutic controversies. Peroral ivermectin, which is considered the first-choice drug, is often not immediately available, as was the case in the first four of our patients, which is crucial for critically ill patients. Furthermore, ivermectin is registered only for intestinal strongyloidiasis. According to a meta-analysis which included 1147 patients, peroral ivermectin showed a higher parasitological cure rate than albendazole and thiabendazole (74–84% versus 48% and 69%, respectively). However, these data refer to patients with chronic intestinal strongyloidiasis (Henriquez-Camacho et al., Reference Henriquez-Camacho, Gotuzzo, Echevarria, White, Terashima, Samalvides, Pérez-Molina and Plana2016). Due to the lack of data, the efficacy of ivermectin and albendazole in SHS/DS is difficult to assess. Given the high mortality of critical SHS/DS, it seems prudent to administer a combination of both drugs pending clinical improvement and parasitological negativization. The report of a successful treatment of previously intractable disseminated strongyloidiasis with a combination of oral ivermectin, partially in double-than-recommended dosage of 400 μg/kg/day, and albendazole for 14 days can be found in the literature (Pornsuriyasak et al., Reference Pornsuriyasak, Niticharoenpong and Sakapibunnan2004). Certainly, albendazole must be added to ivermectin if disseminated CNS infection is suspected or proven, as albendazole (unlike ivermectin) crosses the blood–brain barrier. The optimal duration of antiparasitic therapy in severe cases has not been determined, nor has the duration of empirical antibiotic treatment preventing secondary bacterial infections after the introduction of antiparasitic therapy. According to the clinical course of disease, the duration of antiparasitic therapy should be individually tailored. The World Gastroenterology Organisation recommends, especially in critically ill patients, to continue antiparasitic therapy for two weeks after negativization of previously positive samples (stool, urine, respiratory samples) (Farthing et al., Reference Farthing, Albonico and Bisoffi2018). Although controlled trials are lacking, the clinical experience articulates against early discontinuation of antibiotic therapy after introducing antiparasitic treatment (Kow & Hasan, Reference Kow and Hasan2020; Lier et al., Reference Lier, Davis and Topal2020a, Reference Lier, Tuan, Davis, Paulson, McManus, Campbell, Peaper and Topalb).

Even though there is no published evidence of S. stercoralis resistance to antiparasitic drugs, some patients fail to clear the parasite from the gut despite antiparasitic therapy. As observed in our cohort, it is strongly associated with a lethal outcome. Recent research suggests an association between pre-treatment microbiome community composition or enterotype and therapeutic outcomes in patients with soil-transmitted nematodes (Trichuris trichiura and hookworm) (Schneeberger et al., Reference Schneeberger, Gueuning, Welsche, Hürlimann, Dommann, Häberli, Frey, Sayasone and Keiser2022). Presumably, failure to clear the parasite could have multiple causes, which remain to be elucidated.

The triggering role of rituximab was so far suspected in one case of SHS in a patient with mantle cell lymphoma (Incani et al., Reference Incani, Hernández and González2010). As rituximab maintenance therapy was the only immunosuppression in one of our patients, we can confirm this observation. However, besides the possible triggering role of a recent ileum resection in one patient, corticosteroids in our case series represent the leading risk factor for SHS/DS.

In conclusion, lack of clinical suspicion and significant delay in diagnosis have been observed in our case series of the most severe strongyloidiasis forms. Although being beyond diagnostic recommendations, an early upper gastrointestinal endoscopy with pathohistology could accelerate diagnosis in immunosuppressed patients with nonspecific gastrointestinal symptoms, weight loss and systemic, potentially recurrent infections caused by gut flora of an unknown pathogenesis. The reasons for persistence of parasites in the gut despite antihelmintic therapy should be further investigated. The choice and duration of antiparasitic therapy should be tailored individually. Education of the clinicians, especially those dealing with immunosuppressed persons in the low incidence countries, should be conducted.