Strongyloides stercoralis colitis in a patient positive for human T-cell leukaemia virus with rheumatoid arthritis during an anti-rheumatic therapy: a case report

Abstract

An elderly woman with rheumatoid arthritis (RA) presented with a chief complaint of abdominal pain and diarrhoea while undergoing treatment with low-dose corticosteroids and abatacept. Endoscopic and histopathological findings revealed manifestations of ulcerative colitis (UC). An intermediate dose of corticosteroids and 5-aminosalicylic acid were administered. Abatacept was discontinued; the anti-TNF biologic, golimumab, was administered for treatment of both RA and UC. However, colitis worsened in response to this therapeutic regimen. Colonoscopy revealed severe mucosal lesions; larvae were detected in samples taken from multiple shallow mucosal ulcers. The patient was diagnosed with Strongyloides stercoralis colitis based on the results of an anti-parasite antibody test and examination of the larval DNA. Furthermore, serology revealed a positive test for antibodies against human T-cell leukaemia virus type 1 (HTLV-1). Immunosuppressive treatment was terminated; ivermectin was administered, which resulted in improvements in colitis symptoms within a few weeks. There are several published reports describing S. stercoralis colitis as a lethal mimic of UC. Corticosteroid and anti-TNF therapies have been reported as among the major risk factors associated with strongyloidiasis in patients with HTLV-1 infection. Therefore, HTLV-1 and Strongyloides infections may be considered in cases of new-onset gastrointestinal symptoms during immunosuppressive therapy, particularly in HTLV-1-endemic regions.

Keywords:

• Strongyloidiasis
• parasite infection
• human T-cell leukaemia virus type 1
• rheumatoid arthritis
• biologics

Introduction

Strongyloides stercoralis is an intestinal nematode that is the etiologic agent of strongyloidiasis; chronic infection affects ∼30–100 million people worldwide [1]. Strongyloidiasis is an endemic disease in tropical and subtropical regions with a prevalence of 20% in some regions; this infection is endemic to the southern regions of Japan, including Kyushu and Okinawa. The disease may be asymptomatic in immunocompetent patients or may manifest with occasional stomachaches, intermittent diarrhoea and constipation, bloating, nausea and loss of appetite. Infection with human T-cell leukaemia virus type 1 (HTLV-1) is also a known risk factor associated with strongyloidiasis [1]; the prevalence of strongyloidiasis is reported to be high among HTLV-1 carriers [2]. Among the potential mechanisms, HTLV-1 infection may modify the immune response against S. stercoralis infection [3,4]. As such, immunocompromised patients, including those experiencing malnutrition and those undergoing bone marrow transplantation and/or immunosuppressive therapies including corticosteroids, chemotherapy and biologics, are at increased risk for strongyloidiasis [1]. Several published case reports describe exacerbations of S. stercoralis infection in patients with rheumatoid arthritis (RA) who were undergoing treatment with anti-rheumatic therapies, including methotrexate, corticosteroids and anti-TNF biologics [5–8].

Severe strongyloidiasis is characterized by multi-organ manifestations, including pleuritis, pneumonia, and colitis [9,10]. Strongyloides colitis may present as a lethal mimic of ulcerative colitis (UC); a low index of suspicion can lead to misdiagnosis with fatal consequences [11,12]. Completely different therapeutic regimens are used to treat Strongyloides colitis and UC; as such, an accurate diagnosis is required in order to identify the appropriate treatment for each condition.

In this report, we present the case of a patient with RA who developed S. stercoralis colitis after the administration of anti-TNF therapy together with an intermediate dose of corticosteroid. This therapeutic regimen was initiated for the treatment of presumptive UC based on endoscopic and pathological findings observed at the onset of colitis. This patient was seropositive for anti-HTLV-1 antibody as determined by serological testing after the diagnosis of S. stercoralis colitis was made. Taken together, the combined impact of corticosteroids, biologic agents and HTLV-1 may have promoted the development of S. stercoralis colitis. Nonetheless, it remains unclear whether a screening test for HTLV-1 antibody should be performed in patients with RA who are in need of anti-rheumatic therapies including immunosuppressants or biologics; this may be particularly helpful with respect to evaluating the risks associated with strongyloidiasis, particularly among those living in endemic areas.

Case presentation

A 71-year-old Japanese woman with a past medical history of long-standing (26 years) of RA. She had continued anti-rheumatic therapy in Miyazaki prefecture. She presented with a chief complaint of six months of gradually increasing abdominal pain and diarrhoea while undergoing treatment with low-dose corticosteroids and abatacept. Her RA was treated in the past with various disease-modifying anti-rheumatic drugs, including salazosulfapyridine and methotrexate; these drugs were discontinued because of inadequate response or adverse events. Disease activity associated with RA had been well-controlled for the past 5 years with a treatment regimen that included low-dose prednisolone (PSL; 7 mg/day) together with abatacept. Six months prior to presentation, she experienced gradually worsening gastrointestinal symptoms, including abdominal pain and diarrhoea. Her laboratory examination included a white blood cell count at 7487/μl with eosinophils mildly elevated at 18% (1415/μl); serum C-reactive protein (CRP) was measured at 0.34 mg/dl. Colonoscopy was notable for inflammatory mucosal lesions, including fine erosions, multiple shallow small ulcers, and map-like redness from the caecum to the splenic flexure (Figure 1(A)). Histopathological examination of the mucosa revealed profound infiltration with eosinophils and granulation tissue, including foreign body-type of giant cells; however, no larvae were detected in intestinal mucosal specimens (Figure 1(B)). As such, she was diagnosed with UC, as both peripheral blood eosinophilia and mucosal infiltration with eosinophils were considered to reflect disease activity associated with UC. An intermediate dose of PSL (20 mg/day) and 5-aminosalicylic acid were administered. Furthermore, abatacept was discontinued; golimumab (GLM), an anti-TNF biologic, was initiated for the combined treatment of both RA and UC. At 8 weeks of treatment, the patient reported a fever of 38 °C together with vomiting. Her laboratory findings included elevated serum CRP at 3.56 mg/dl and increasing numbers of eosinophils in peripheral blood, at 37.7% (2950/μl). Colonoscopy revealed exacerbation of the mucosal inflammatory lesions, including multiple ulcerative lesions with white plaques that extended to the descending colon from the splenic flexure (Figure 2(A)). Numerous larvae were isolated from ulcerative lesions (Figure 2(B)). The patient was re-diagnosed with S. stercoralis colitis; the parasite was identified by antibody testing and examination of the larval DNA contents. The prevalence of S. stercoralis infection has been reported to be high among carriers of HTLV-1 [3,13]. As such, we considered the possibility of HTLV-1 coinfection in this patient. Serological tests results revealed that the patient was positive for anti-HTLV-1 antibody. HTLV-1 proviral loads (PVLs) of more than 4.0 copies/100 peripheral blood mononuclear cells (PBMCs) have been reported to be predisposing factors associated with the development of adult T-cell leukaemia (ATL) [14]. The HTLV-1 PVL in this patient was comparatively low, at 0.11 copies/100 PBMCs; as such, the risk of ATL was considered to be low. The patient was instructed to eliminate all food intake; intravenous hyperalimentation was provided to treat the high fever, right abdominal pain and vomiting. Ivermectin (0.2 mg/kg/day) was administered once a week and anti-rheumatic treatment was withdrawn. The signs and symptoms of colitis improved within several weeks, and the number of eosinophils decreased to 567/μl. After the Strongyloides colitis resolved, administration of GLM resumed; this maintained her RA disease activity in remission without recurrence of colitis. We used her extensively regarding sources of Strongyloides infection, such as the possibility of skin contact with contaminated soil. The patient was unable to recall any such events; as such, the routes of Strongyloides infection remained unclear.

Figure 1. Colonoscopy findings. (A) Colonoscopy revealed inflammatory mucosal lesions, including fine erosions, multiple shallow small ulcers, and map-like redness from the caecum to the splenic flexure at the onset of colitis. (B) Biopsy specimen of the colonic mucosa is notable for a dense inflammatory infiltrate including many eosinophils and several foreign body type of giant cells (at the arrows; original magnification ×132. Electronic camera lens magnification is ×3.3. Microscope objective lens magnification is ×40).

Figure 2. Exacerbation of colitis and a single larva isolated from an ulcerative lesion. (A) Follow-up colonoscopy revealed exacerbation of the inflammatory mucosal lesions in response to administration of golimumab and an intermediate dose of corticosteroid. Numerous ulcerative lesions with white plaques that extended from the splenic flexure to the descending colon were detected. (B) A larva of Strogyloides stercolaris isolated from an ulcerative lesion (lesions circled with broken lines in A).

Discussion

Biologics, corticosteroids and HTLV-1 infection have been identified as risk factors for strongyloidiasis [1,5,15]. It was difficult to distinguish between Strongyloides colitis and UC on endoscopy prior to treatment with GLM and an intermediate dose of corticosteroid.

There are several case reports that describe the emergence of UC in patients with RA [16–18]. It can be difficult to impossible to differentiate between UC and other forms of colitis, particularly those primarily or secondarily related to RA and/or drug-induced colitis associated with RA treatments. For example, Motohashi et al. [19] reported the cases of two patients with RA who developed UC during treatment with abatacept. In this case report, termination of abatacept and administration of anti-TNF therapy resulted in resolution of abatacept-induced UC. The patient presented in our case underwent treatment with abatacept at the onset of colitis, and both endoscopy and histopathological examinations supported the diagnosis of UC. As such, abatacept was discontinued and GLM therapy was initiated for the concomitant treatment of both RA and UC. However, initiation of both GLM together with an intermediate dose of PSL resulted in clear exacerbation of Strongyloides colitis; larvae were ultimately identified on endoscopy in association with UC-like mucosal lesions.

Several reports indicated that clinical and pathologic features of Strongyloides colitis may mimic those of UC [11,12]; this is a major source of diagnostic error in cases of Strongyloides colitis [11]. Several reports recommended that careful examinations led to findings associated with strongyloidiasis, including intact filariform larvae in the crypts and in the inflamed lamina propria [20–24]. Stool examination was also recommended in order to detect larvae and ova when diagnosing Strongyloides colitis [20–24]. In our case, initial mucosal specimens revealed UC-like findings, including small epithelioid granulomata and profound infiltration with eosinophils in the absence of larvae. Follow-up endoscopy revealed numerous larvae in the ulcerative lesions. Peripheral blood eosinophilia and mucosal infiltration with eosinophils have been reported in patients with active UC [25]. Aggressive medical therapy, including administration of anti-TNF modalities, is recommended more frequently for patients with UC and PBE vs. those with UC but without PBE. Therefore, we first diagnosed this patient with severe UC, which is typically treated with anti-TNF therapy, depending on the extent of PBE. However, the administration of both an intermediate dose of PSL and anti-TNF therapy resulted in the exacerbation of Strongyloides colitis and an increasing number of larvae. The number of eosinophils also increased in response to this therapeutic regimen. Strongyloides colitis can result not only from anti-TNF therapy but also in response to corticosteroids [8,15]. Since Strongyloides colitis may mimic the characteristics of UC when evaluated by endoscopy, this diagnosis should be considered in any case of new-onset colitis that develops in patients on anti-rheumatic therapy with corticosteroids, immunosuppressants and/or biologics.

Several reports suggested that co-infection with Strongyloides occurs frequently among carriers of HTLV-1 when compared with anti-HTLV-1-seronegative individuals [13,26,27]. A recent meta-analysis reported an odds ratio of 2.48 for concurrent infection with HTLV-1 in patients diagnosed with strongyloidiasis [1]. S. stercoralis hyperinfection syndrome, which is one of fatal outcome of this disease, is common in HTLV-1 co-infected patients [28]. However, the mechanisms underlying the susceptibility of HTLV-1 carriers to S. stercoralis infection are not fully understood. HTLV-1-infected T cells can behave in ways reminiscent of T helper (Th) 1-like cells and can autonomously produce interferon (IFN)-γ. In HTLV-1 carriers can suppress the parasite-specific Th2 response, which results in a comparatively low eosinophil count, low levels of parasite-specific IgE, as well as lower IL-4 and IL-5 levels and a higher regulatory T-cell count in peripheral circulation [4,29–31]. Coinfection with HTLV-1 results in a switch from Th2 to Th1 responses in patients with strongyloidiasis [32,33]. Therefore, IFN-γ produced by HTLV-1-infected T cells may play crucial role in decreasing the Th2 response in the setting of a Strongyloides infection. Spontaneous production of IFN-γ has been reported in carriers of HTLV-1 with high PVLs [34]. Interestingly, the HTLV-1 PVLs associated with strongyloidiasis/HTLV-1 co-infection were reported to be higher than those in HTLV-1 asymptomatic carriers [29,30,35]. High HTLV-1 PVLs may correlate with the severity of strongyloidiasis [30,35]. It is possible that HTLV-1 PVLs may have an impact both on the development and the severity of strongyloidiasis. Although the HTLV-1 PVL was low in the present case, she had received anti-rheumatic therapies, including corticosteroids and biologics. These treatments have also been described as risk factors for the development of severe strongyloidiasis [1,33]. Therefore, the concurrent HTLV-1 infection and also the therapeutic agents might have served to promote the development of severe S. stercoralis infection in our case.

In conclusion, physicians should strongly consider strongyloidiasis in patients presenting with new-onset gastrointestinal symptoms and colonoscopy findings suggestive of UC while on immunosuppressive treatment; different strategies are required for treatment of these two very different disorders. In addition, since HTLV-1 infection has been reported to be one of the risk factors for Strongyloides infection, screening tests for anti-HTLV-1 antibodies may considered to be performed prior to initiating an immunosuppressive therapy regimen for rheumatic diseases including RA for patients residing in HTLV-1-endemic areas.

Patient consent

Written informed consent for the publication of this report was obtained from the patient.

Author contributions

YH and KU are the principal investigators of this case who drafted the manuscript. HM, AK, YH, and KU performed clinical assessment and data collection. EN and HM performed parasite antibody test and DNA examination. TH and AO provided support in drafting the manuscript. YH and KU submitted this manuscript.

Acknowledgments

The authors thank Professor Yujiro Asada of the University of Miyazaki, Pathophysiology Section, Department of Pathology, for help in preparation of both figures and legends of pathological findings of mucosal lesion. He contributed revision of manuscript and important intellectual content as pathologist. We would like to thank Enago (https://www.enago.jp/) for English language editing.

Conflict of interest

None.

Additional information

Funding

This report did not receive any funding support from public or private organizations.