Infectious adverse events in patients treated with biotherapies in a tropical Amazonian context, French Guiana

Abstract

Background

Data on infectious complications associated with biologic therapies in tropical settings are scarce. French Guiana, an Amazonian French overseas territory, combines an important burden of infectious diseases with a high use of biologic therapies. Our objective was to assess the incidence, spectrum, and risk factors for infections in patients treated with biotherapies for inflammatory and autoimmune diseases in French Guiana.

Methods

We performed a multicentre retrospective study of adult patients receiving biotherapies between January 2021 and December 2023 in the French Guiana University Hospital (including its main site in Cayenne, its three local hospitals and 15 remote health centres). The primary outcome was the occurrence of any infectious event. Secondary outcomes included the severity of infection and identification of clinical and epidemiological risk factors.

Results

146 patients (mean age 44.1 years, 61.6% female) were included. 48.0% of patients included had at least one infectious event. Most infections were mild (90.4%) and significantly associated with concomitant immunosuppressants (35.7%) or with remote health centres follow-up (11.4%). The tropical context revealed a distinct epidemiology, with five cases of dengue fever, two cases of scabies and one histoplasmosis. The majority (55%) of events occurred under first-generation biotherapies (notably rituximab and anti-TNF agents). The most frequent type of infection was cutaneous (27.2%), particularly in patients with a dermatological indication.

Conclusions

Biotherapies appear generally well tolerated in this tropical setting, but patients living in remote areas or with concomitant immunosuppressant therapies remain vulnerable. Specific awareness should be dedicated to secondary cutaneous infections in patients treated for dermatological indications.

Clinical trial number

Not applicable.

Supplementary information

The online version contains supplementary material available at 10.1186/s12879-026-13390-8.

Background

Autoimmune and inflammatory diseases are characterized by inappropriate activation of the immune system directed against self-antigens. The underlying mechanisms are heterogeneous and may result from hyperactivation or failure of regulatory signalling pathways [1, 2]. Initially managed with immunomodulatory agents such as conventional synthetic disease-modifying antirheumatic drugs (CsDMARDs) and/or immunosuppressants (including corticosteroids, cyclophosphamide, azathioprine, and mycophenolate mofetil), recent advances have led to the development of targeted biologic therapies. These agents exert specific modulatory effects on the immune pathways [2].

Since their introduction in the early 2000s, biologic therapies have demonstrated a substantial efficacy. However, these drugs are associated with an increased risk of infections, even when used as corticosteroid-sparing agents [3–5]. The infectious risk varies across different classes of biologic agents, and is frequently confounded by concurrent or prior use of other immunosuppressants [4]. Reported infections are primarily bacterial, including reactivation of latent tuberculosis (TB) [5–7]. Viral reactivations, particularly herpes simplex virus (HSV) and hepatitis B virus (HBV), have also been observed [6, 8–10]. Infections predominantly affect the respiratory, urinary, and cutaneous systems, with varying degrees of severity [3–6].

French Guiana is a French overseas department located in South America. This territory is characterized by an equatorial climate, a French universal health care system and sharp socioeconomic disparities. Indeed, more than half of the population lives below the poverty threshold [11]. The territory is provided with three main coastal hospitals (Cayenne, Kourou, and Saint-Laurent du Maroni) while all healthcare in the hinterland is provided by three small local hospitals and fifteen remote health centres (due to the absence of private practice). The rainforest hinterland is mostly inhabited by autochthonous populations who maintain a traditional way of life in small forest settlements, and illegal gold miners living on very harsh conditions [11–14].

Infectious diseases represent a major public health issue in French Guiana, with high incidence rates of malaria, Human immunodeficiency virus (HIV), Human T-lymphotropic virus type 1 (HTLV1), Q fever and tuberculosis [13, 14]. Specific infections due to the Amazonian ecosystem, such as leprosy [15], histoplasmosis, Chagas disease, cutaneous leishmaniasis [16–18], Buruli ulcer [19, 20] and arboviral diseases are also prevalent [13, 21]. Neglected Tropical Diseases [22, 23] are particularly significant in remote and border areas, where the Amazonian infectious exposome, the traditional way of life and barriers to healthcare access can lead to severe presentations [12, 14, 24].

An increasing number of French Guiana patients are being treated for autoimmune and inflammatory diseases, leading to a growing use of biotherapies. However, there is currently no available local data regarding infectious complications in tropical settings. Most safety studies on biotherapies are conducted in high-income and temperate countries, where the epidemiology and prevalence of infectious pathogens are considerably different [3–6, 12, 13]. It is possible that different safety data would be collected in the equatorial, highly infectious setting of French Guiana.

The primary objective of this study was to assess the incidence of infectious adverse events in patients treated with biologic therapies in French Guiana. Secondary objectives were to evaluate the severity of infectious events and identify clinical and epidemiological risk factors of infection.

Methods

This retrospective study included all patients who received biologic therapy for autoimmune or inflammatory indications between January 2021 and December 2023 in all sites of the French Guiana University Hospital. Patients who did not have at least one recorded consultation after biotherapy initiation were considered lost to follow-up and excluded.

The primary outcome (PO) was the occurrence of any infectious event. Secondary outcomes (SO) included the severity of and identification of clinical and epidemiological risk factors. We defined infectious event as the occurrence of any clinical infection corresponding to infectious categories defined in the CTCAE classification [25], whether antibiotics were prescribed or not. Asymptomatic abnormal lab tests were not included. The severity was categorized according to the CTCAE classification, “severe” events corresponding to CTCAE grade 3 (hospitalization), 4 (life-threatening, ICU transfer) and 5 (death).

Data were extracted from medical records. The following variables were collected: sex, age, primary immunosuppression, indication for biologic therapy, concomitant use of immunosuppressants, site of inclusion (main site in Cayenne, local hospitals and remote centres), date of first consultation, and date of first infectious event. The following comorbidities considered leading to primary immunosuppression were: chronic organ failure, diabetes, HIV infection, nephrotic syndrome, systemic lupus erythematosus (SLE), and Sjögren’s disease, hypogammaglobulinemia. First-generation biotherapies (anti-tumour-necrosis factors (TNF) and rituximab) were distinguished from newer drugs.

Categorical variables were described using frequencies and percentages, and continuous variables using means and medians. Associations between infectious events and variables were evaluated using logistic regression models for binary outcomes (Pearson’s chi-squared test or Fisher’s test if less than five occurrences). Odds ratios (ORs) with 95% confidence intervals (CIs) were estimated for each potential risk factor. Variables with p < 0.2 in the univariable analysis were included in a descending unconditional stepwise multivariable logistic model. p values < 0.05 for the Wald test in the final multivariable model were considered significant.

Risk factors which were significant in the multivariate model were then used to compare the time to first event in patients exposed (cases) or not exposed (controls) to this factor. The Kaplan-Meier method was used, with the time to occurrence of first infectious event as time-to-event outcome. Stata® software (Statacorps, College Station, TX, USA) was used for all statistical analyses.

Results

General results

A total of 146 patients were included (61.6% female, mean age 44.1 years). Among them, 12/146 (8.2%) were followed in remote health centres. Most prescriptions originated from the department of Dermatology (46.7%), followed by Rheumatology (13.0%), Gastroenterology (11.6%) and Immunology/Internal Medicine (10.3%) (Table 1, Additional File 1). Main indications were atopic dermatitis and psoriasis (18.5% each), followed by inflammatory bowel disease (8.9%) and rheumatoid arthritis (RA) (8.2%) (Additional File 2). The most frequently prescribed biotherapies in the overall study population were dupilumab (20.3%), rituximab (18.6%), and infliximab (12.4%) (Table 1). In the remote health centres, prescribed biotherapies included guselkumab (5 patients, 41.7%), dupilumab (4, 33.3%), secukinumab (2, 16.7%) and infliximab (1, 8.3%). At baseline, 26.0% had a primary immunosuppression (PI) status and 27 (18.5%) were on concomitant immunosuppressant therapy.

Table 1.

Population characteristics of patients with biotherapies and distribution of prescribed biotherapies, BIORISK study. 2021-2023. French Guiana (n: 146 patients and 177 biotherapies)

Baseline characteristics	Number of patients (n=146)	%
Gender
Male	56	38.4
Female	90	61.6
Age at the beginning of follow-up
< 20 years	21	14.4
20-45 years	52	35.6
46-65 years	54	37.0
>65 years	19	13.0
Primary immunosuppression1	38	26.0
Diabetes	18	12.3
Organe failure	14	9.6
HIV	2	1.4
Others 2	19	13.0
Prescribed biotherapy by speciality
Dermatology	68	46.7
Rheumatology	19	13.0
Gastroenterology	17	11.6
Immunology/Internal Medicine	15	10.3
Neurology	13	8.9
Nephrology	11	7.53
Pneumology	3	2.0
Prescribed biotherapy3	Number of biotherapies (n=177)	%
Dupilumab	36	20.34
Rituximab	33	18.64
Infliximab	22	12.4
Ustekinumab	14	7.91
Adalimumab	11	6.21
Guselkumab	11	6.21
Secukinumab	8	4.52
Etanercept	7	3.95
Belimumab	6	3.39
Abatacept	5	2.82
Others 4	27	15.25

¹Among the 38 patients with immunosuppression, some had several factors (diabetes and HIV, for example)

²Chronic organ failure, diabetes, HIV infection, nephrotic syndrome, systemic lupus erythematosus (SLE), and Sjögren’s disease, hypogammaglobulinemia

³177 different biotherapies were prescribed for 146 different patients, some patients may have received several biotherapies

⁴Risankizumab, tocilizumab, tofacitinib, golimumab, obinituzumab, upadacitinib, certolizumab, brodalumab, caplacizumab, eculizumab, ocrelizumab, omalizumab, sarilumab, vedolizumab

Infectious events

Infectious events in included patients

Concerning the distribution of follow-up time, visual inspection suggested non-normality and the median was therefore calculated at 15.6 months (IQR: 6–23). During follow-up, 114 infections were reported in 70/146 patients (48.0%). Among the 144 infectious events, 103/114 (90.4%) were mild, with eleven severe cases (grade 3: 1/11 (9.0%); grade 4: 6/11 (54.5%), and grade 5: 4/11 (36.3%)). Among the 70 infected patients, 25/70 (35.7%) had recurrent events, including 20/70 (28.6%) with 2–3 events and 5/70 (7.1%) with >3 events. The infection rate was 0.50 event per patient-year. Among the 70 infected patients, 19/70 (27.1%) had a primary immunosuppressive status, 25/70 (35.7%) were receiving an immunosuppressant therapy. First-generation biotherapies were involved in 55/114 infections (48.2%), including 31/114 (27.2%) under rituximab. First-generation drugs were involved in 7/11 (64%) of serious events, mainly with rituximab (five events).

The most common types of infections were cutaneous/mucosal (27.2%), followed by flu-like syndromes (20.1%) associated with Covid-19, influenza or dengue fever; digestive (14.0%), ENT (ear, nose, and throat) (13.1%), respiratory (8.0%), and urinary (6.1%) syndromes (Table 2). Identified pathogens were bacterial (42.1%), viral (36.8%), fungal (10.5%), and parasitic (1.8%). Concerning tropical infections, we reported five cases of dengue fever. Most dengue (3 cases) and flu-like syndromes (7 cases) occurred in Rheumatology patients. Two cases of scabies and one case of disseminated histoplasmosis and pulmonary tuberculosis were also reported. On the other hand, there was no case of Q fever, Buruli ulcer, leptospirosis, leishmaniasis, leprosy, malaria or any other Neglected Tropical Disease endemic in French Guiana.

Table 2.

Distribution and type of infectious events in the study population, by involved organ and documented pathogen, BIORISK study, 2021–2023, French Guiana (n = 114)

Infectious event	n = 114
Cutaneous-mucosal	31
• HSV initial infection/recurrence	3
• Shingles	3
• Chickenpox	2
• Scabies	2
• Others/undocumented:  ○ dermatophytosis (7)  ○ diabetic foot infection (3)  ○ non-necrotizing dermohypodermitis (3)  ○ abscess, cellulitis (2)  ○ ecthyma, furunculosis (2)  ○ central venous implantable port infection (1)  ○ necrotizing dermohypodermitis (1)  ○ resolutive gingivitis under antibiotics (1)  ○ seborrheic dermatitis (1)	21
Systemic	29
• Influenza	8
• Dengue fever	5
• Bacteraemia:  ○ Extended-spectrum beta-lactamase Klebsiella pneumoniae (1)  ○ Methicillin-resistant Staphylococcus aureus (1)  ○ Methicillin-resistant Staphylococcus epidermidis (1)  ○ Pseudomonas aeruginosa (1)	4
• COVID-19	3
• Histoplasma capsulatum	1
• Mycobacterium tuberculosis (latent tuberculosis infection: LTBI)	1
• Others/undocumented: flu-like syndrome	7
Digestive	16
• Clostridium difficile enteritis	2
• Helicobacter pylori gastritis	1
• Others/undocumented:  ○ Gastroenteritis (6)  ○ Anoperineal abscess (2)  ○ Fungal digestive infection (2)  ○ Bacterial digestive infection (1)  ○ Febrile diarrhoea (1)  ○ Lithiasis cholangitis (1)	12
Respiratory	9
• Legionella	1
• Pneumococcus pneumonia	1
• Mycobacterium tuberculosis	1
• Others/undocumented:  ○ Pneumonia (3)  ○ Bronchitis/tracheobronchitis (3)	6
Urinary	7
Escherichia coli (5) Enterococcus faecalis (1) Pseudomonas aeruginosa (1)	7
Others	2
Undocumented: pharyngitis (8), sinusitis (4), external otitis (2), conjunctivitis (2), osteitis (2), rhinitis (1); dental abcess (1), dental infection (1); vaginal yeast infection (1)	2

Infectious events by medical specialty

Infectious events (PO) and their characteristics (SO) per specialty are detailed in Additional File 3. Concerning each medical specialty, Nephrology had the most important proportion of patients with infectious events, (7, 63.6%), followed by Rheumatology (10, 52.6%), Gastro-enterology (8, 47.1%), Immunology (7, 46.7%), Neurology (6, 46.1%). Dermatology was the speciality with the lowest proportion of patients with infectious events (31, 45.6%), though this speciality, due to its large number of patients, represented almost half of all infected participants. Immunology/internal medicine presented the highest proportion of serious events (6 events, 37.5%), while dermatology presented only two severe events out of 42 (4.8%). However, there was no significant difference in the ratios of infectious events between specialities. The most frequent locations of infections were cutaneous-mucosal in Dermatology (50%), digestive in Gastro-enterology (41.2%) and systemic in Immunology (50%), Nephrology (55.6%) and Rheumatology (42.1%).

Patients included in the remote health centres were all treated for dermatological indications (7 cases of psoriasis, 4 atopic dermatitis and one leprosy reaction). Thirteen infections were reported (11.4% of total events), mostly dermatological (7, 53.8%).

Risk factors of infectious events

In univariate analysis (Table 3), there was no association between the occurrence of infectious events and the following factors: age, sex, diabetes, HIV status, medical specialty/indication and primary immunosuppression. On the other hand, a concurrent treatment with an immunosuppressant drug was significantly associated with a higher risk of infectious event. In multivariate analysis (Table 3), follow-up in remote health centres (OR 5.1; 95% CI [1.2- 21.4], p = 0.028) and concurrent immunosuppressant therapy (OR 7.9; 95% CI [1.8–33.8)], p = 0.005) were significantly associated with the occurrence of infectious events.

Table 3.

Association between infectious events and study population characteristics in general population: univariate and multivariate analysis, BIORISK study, 2021–2023, French Guiana (n = 114)

Variable	n with event	(%)	n without event	(%)	Total		Univariate analysis			Multivariate analysis
						Odds Ratio (95% CI)		p-value	Odds Ratio (95% CI)		p-value
Gender								0.6950			–
Male	28	0.50	28	0.50	56	–			–
Female	42	0.47	48	0.53	90	0.8	(0.4- 1.8)		–
Age under 20								0.362			–
Yes	12	0.57	9	0.43	21	1.5	(0.5- 4.4)		–
No	58	0.46	67	0.54	125	–			–
Age under 65								0.661			–
Yes	10	0.53	9	0.47	19	1.2	(0.42–3.7)		–
No	60	0.47	67	0.53	127	–			–
Remote health centres								0.050			0.028
Yes	9	0.75	3	0.25	12	3.6	(0.8- 21.3)		5.1	(1.2- 21.4)
No	61	0.46	73	0.54	134	–			–
Diabetes								0.411			–
Yes	7	0.39	11	0.61	18	0.7	(0.2- 1.9)		–
No	63	0.49	65	0.51	128	–			–
HIV								0.953			–
Yes	1	0.50	1	0.50	2	1.9	(0.01- 86.4)		–
No	69	0.48	75	0.52	144	–			–
Rheumatology								0.661			–
Yes	10	0.53	9	0.47	19	1.2	(0.4–3.7)		–
No	60	0.47	67	0.53	127	–			–
Nephrology								0.278			–
Yes	7	0.64	4	0.36	11	2	(0.4- 9.7)		–
No	63	0.47	72	0.53	135	–			–
Dermatology								0.594			–
Yes	31	0.46	37	0.54	68	0.8	(0.4- 1.7)		–
No	39	0.50	39	0.50	78	–			–
Immunology								0.917			–
Yes	7	0.47	8	0.53	15	0.9	(0.3- 3.2)		–
No	63	0.48	68	0.52	131	–			–
First-generation biotherapy								0.269			0.101
Yes	34	0.53	30	0.47	64	1.4	(0.7- 2.3)		3.6	(0.8- 16.2)
No	36	0.44	46	0.56	82	–			–
Primary immunosuppression								0.768			–
Yes	19	0.50	19	0.50	38	1.2	(0.5- 2.5)		–
No	51	0.47	57	0.53	108	–			–
Immunosuppressant therapy								0.000			0.005
Yes	25	0.93	2	0.07	27	21.0	(4.8- 188.2)		7.9	(1.8–33.8)
No	44	0.37	75	0.63	119	–			–

Survival analysis

In the survival analysis, the time to first infectious event was analysed in each group with and without three risk factors: concurrent immunosuppressant therapy and follow-up in remote health centre group (significant in multivariate analysis) and treatment with first-generation biotherapy (highly involved in the literature [26, 27].

Thus, the median time to first infectious event was five months in patients with concurrent immunosuppressant therapy, versus ten months without concurrent drug (Fig. 1). At ten months, 75% of patients receiving immunosuppressant therapy had experienced an infection, compared to fifteen months in the non-associated group. The median time to first event was eight months in patients followed in the remote health centres, versus nine months in patients followed in Cayenne (Fig. 2). The median time to infection was nine months in the group of patients treated with firs-generation biotherapy, versus ten months in the control group, with 75% of patients infected at ten and twelve months, respectively.

Fig. 1.

Kaplan Meier curve – survival analysis estimation of free infectious event (PO) in patients with vs. Without concurrent immunosuppressant therapy in general population. BIORISK study. 2021–2023. French Guiana (n = 146). Ass_IS: associated immunosuppressant therapy; no_ass_IS: no associated immunosuppressant therapy

Fig. 2.

Kaplan Meier curve – survival analysis estimation of free infectious event in remote health centres versus coastal hospitals – follow up patients. BIORISK study. 2021–2023. French Guiana (n = 146). (RHC: remote health centres)

Discussion

In this multicentre cohort study conducted in French Guiana, nearly half of the patients receiving biotherapies (48%) experienced at least one infectious event, of which approximately 10% were severe (grade 3: 0.9%, grade 4: 5.3%; grade 5: 3.5%). While these figures are consistent with previous international cohorts [4, 28] they also highlight several specific features related to the Amazonian context. Two factors were significantly associated with the occurrence of infections: the concomitant use of immunosuppressant therapy and a follow-up in remote health centres.

The most frequent events were mild and cutaneous infections (dermatophytosis, mild bacterial infections) (27.2%), flu-like syndromes (COVID-19, influenza, dengue) (25%), and gastrointestinal infections, mainly infectious colitis (14%). Common infections were frequently identified, including primary or recurrent herpes simplex virus (HSV), chicken pox and non-necrotizing dermo-hypodermatitis. Parasitic infections like scabies were also found (6.5% of cutaneous infections).

Dengue fever is an emerging tropical arboviral infection, responsible for important morbidity and mortality, both in endemic countries and in international travellers [29]. Therefore, the spread of biotherapies in tropical countries and among travellers could lead to increased cases in patients at risk of severe complications [3, 12, 30]. In our study, five documented dengue were reported. However, we did not report any lethal case, despite 2/5 being under another immunosuppressant drug. Though dengue fever can be associated with significant mortality [29], none of our five patients fulfilled the criteria for severe dengue fever and all of them were treated in outpatient consultation.

Tuberculosis is another infectious agent often associated with immunosuppressant drugs and first biotherapies [7, 27, 31]. In our study, we reported one case of pulmonary tuberculosis. Although one case of active tuberculosis was observed in our study, the limited sample size does not allow extrapolation of the infectious risk associated with biologic therapy in French Guiana. High or middle-income settings with high tuberculosis endemicities should be the target of specific guidelines to prevent an increased risk of tuberculosis reactivation. In the WHO recommendations [32], LTBI screening before initiating non-anti-TNF biotherapies does not appear to be necessary, except in individuals at high risk for tuberculosis. However, all guidelines agree on mandatory screening in countries with high TB endemicity [32–34]. IGRAs were likely performed for all our patients prior to treatment, though this data was not always recorded and a precise LTBI rate cannot be determined.

Concerning other tropical diseases, one case of lethal disseminated histoplasmosis was reported in a patient followed for a severe leprosy reaction treated with oral steroids, infliximab and apremilast. Severe leprosy reactions (reversal reaction (RR) and erythema nodosum leprosum (ENL)) are typically managed with corticosteroids. Biotherapies, mainly anti-TNF agents (especially infliximab), are recommended for refractory RR as steroid-sparing agents [35–38]. Concerning histoplasmosis, this fungal infection is the leading cause of opportunistic infection in people living with HIV in French Guiana, with an average incidence of 9.41 per 1,000 person-years between 1992 and 2021 [39]. However, in this study, none of the two HIV-infected patients presented histoplasmosis. No case of Buruli ulcer were observed. Although French Guiana reports the highest number of cases in the Americas, the disease remains rare [20]. Exposure to rural and coastal marshy environments was underrepresented in our cohort, which may partly explain the absence of Buruli ulcer cases. We did not report other tropical diseases such as leptospirosis, cutaneous leishmaniasis, leprosy, Q fever or malaria. Though endemic of French Guiana, these diseases are mostly transmitted in rural forested areas and primarily affect gold miners [14, 16] who were not represented in our predominantly urban cohort.

Two cases of scabies were reported among patients followed for psoriasis. Involved biotherapies were guselkumab and ustekinumab. One patient was followed in a remote health centre. A case of Norwegian crusted scabies under adalimumab (anti-TNF) for psoriatic arthritis was reported in 2015 [40]. Additional cases of Norwegian scabies have been described in patients receiving TNF-α, IL-17, and IL-23 inhibitors. The proposed mechanism involves TNF and IL-17 dysregulation induced by these biotherapies, especially increased by concomitant corticosteroid use or methotrexate therapy [41, 42]. The scabies incidence in our study (1,370 per 100,000 inhabitants) was lower to the figures observed in French Guiana, where it can be as high as 2000–4000/ 100 000 inhabitants in remote areas or slums [43, 44]. Therefore, the risk of scabies among our patients did not seem higher than in the general population of French Guiana.

Patients from the Dermatology cohort represented a large part of the infectious events recorded in our study, although most of them were mild. Half of the infections were cutaneous (21/42, 50%), mainly dermatophytosis, benign bacterial infections, recurrent VZV/HSV, and scabies. This high infection rate may be partly explained by the high proportion of patients included from the remote health centres in the dermatological cohort (12/68, 17.6%). A follow-up in these centres was associated with a higher frequency of infectious events. The predominance of cutaneous infections may also be related to the high endemicity of scabies or the hot and humid climate, known to favour fungal infections. Another likely explanation is that patients with underlying skin diseases are predisposed to secondary cutaneous infections, sometimes complicated by superinfection due to traditional remedies [45].

Follow-up in remote health centres was associated with a higher infection risk. However, all cases in remote centres were mild, mostly cutaneous (53.8%), and no deaths occurred. Populations living in remote or border areas of French Guiana often face harsh conditions, including limited access to clean water and electricity, combined with a hot and wet rainforest environment which facilitates the spread of infectious diseases [12]. Behavioural factors associated with a traditional lifestyle (hunting, fishing and gathering) can cause increased exposure to skin barrier breaches (wounds, insect bites, impetigo …).

Concomitant use of immunosuppressant therapy emerged as a strong predictor of infection in our cohort. Median infection-free survival was significantly shorter in patients with concomitant immunosuppressant drug (five months vs. ten months), and all of them developed an infection within 21 months. These findings are consistent with previous studies, particularly in patients with rheumatoid arthritis treated with methotrexate in combination with biologic therapies [46]. In French Guiana, L. Cachera and al. [47] analysed infectious diseases in kidney transplant recipients, all receiving immunosuppressant therapy. Infections occurred in 42/82 patients (51%), with 15% requiring ICU transfer and 7% dying. Endemic tropical infections accounted for 6% of cases, including dengue, disseminated histoplasmosis, Coxiella burnetii, and Exophiala oligosperma. This overall infection rate was comparable to the 48% of infectious events observed in our biotherapies cohort. However, our death rate (2.7%) was clearly lower.

In our cohort, first-generation biotherapies were involved in 64/146 patients (0.97 event/patient) (0.63 event per patient-year). Seven out of ten serious events (63.7%) and two out of four deaths (50%) occurred in patients treated with one of these drugs. Although old biotherapies such as rituximab and anti-TNF appeared to be associated with a higher number and greater severity of events, no statistically significant association was identified, possibly due to a limited sample size.

Our study has several limitations. Some retrospective data were based on undocumented clinical reports, exposing the study to recall or classification bias. The study design itself was retrospective, with potential differences in quality of data collection between urban centres and remote centres. Febrile syndromes without microbiological documentation also represent a source of diagnostic uncertainty. Estimated incidences for infectious diseases were often based on a limited number of occurrences. Nevertheless, this study has several strengths: it is the first to explore infectious risk associated with biotherapies in French Guiana, and it specifically includes patients followed in remote and often underserved areas. The study was multicentric, spanned three years, and focused on a vulnerable, diverse population that remains largely underrepresented in the literature.

Conclusion

The use of biologic therapies in this Amazonian setting appears to overall well tolerated, particularly with new-generation drugs. Infectious events were frequent but most of them were mild. Use of concomitant immunosuppressant therapy and a location in remote areas were associated with a higher risk of events. While overall rates were consistent with international cohorts, the tropical context revealed local epidemiology specificities. Targeted biotherapies probably offer a more favourable safety profile than immunosuppressant drugs, which makes them particularly valuable for tropical areas exposed to high infectious risk. However, the environmental context of each patient should be evaluated, as patients maintaining a traditional way of life in rural settings seem to be more exposed to infections. Parasitic, bacterial, fungal and viral secondary infections should be regularly expected and screened in patients treated for dermatological indications.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Abbreviations

ASp

Axial Spondyloarthritis

ANCA

Antineutrophil cytoplasmic antibody

CI

Confidence interval

Cs-DMARD

Conventional synthetic disease-modifying antirheumatic drug

ENL

Erythema nodosum leprosum

ENT

Ear, nose, and throat

GCA

Giant cell arteritis

HBV

Hepatitis B virus

HIV

Human immunodeficiency virus

HSV

Herpes simplex virus

HTLV-1

Human T-lymphotropic virus type 1

IBD

Inflammatory bowel disease

ICU

Intensive care unit

II

Initial infection

ITP

Immune thrombocytopenic purpura

ISAI

Immunosuppressed and/or associated immunosuppressants

JAK

Janus Kinase

LTBI

Latent tuberculosis infection

MSSA

Methicillin-sensitive staphylococcus aureus bacteremia

MRSA

Methicillin-resistant staphylococcus aureus bacteremia

MGN

Membranous glomerulonephritis

NETs

Neutrophil extracellular traps

OR

Odds ratio

PI

Primary immunosuppression

PO

Primary outcome

RA

Rheumatoid arthritis

RHC

Remote health centres

RR

Reversal reaction

ROS

Reactive oxygen species

SAPHO syndrome

Synovitis, acne, pustulosis, hyperostosis, osteitis syndrome

SLE

Systemic lupus erythematosus

SO

Secondary outcome

TB

Tuberculosis

TNF

Tumour-necrosis factors

TTP

Thrombotic thrombocytopenic purpura

VZV

Varicella-zoster virus

WHO

World Health Organization

Author contributions

DPL, HA, DL, TG, AA, MS, MB, CM, ND, GG collected and recorded data on occurring infectious events and medical histories in their respective departments; RP gathered and analysed all data, ran the statistical tests, wrote the draft; RB conceived the study, supervised the statistical analysis and reviewed the manuscript. All authors read an approved the final manuscript.

Funding

This work did not receive any specific funding.

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Human ethics and consent to participate

This study was conducted in line with the Helsinki declaration. This study was approved by the ethics/research directorate of the French Guiana University Hospital and was published on Health Data Hub with the name BIORISK. Informed consent was collected for all patients. According to French legislation (Law n° 2012–300, passed on 5^th March 2012), no further legal clearance was needed. The BIORISK database contained only medical data corresponding to the clinical variables exposed in the study, and all data were anonymized. This study does not contain any identifying patient data and consent for publication is not applicable.

Competing interests

RB has received fees for scientific expertise and reimbursements for transport and accommodation from Sanofi and Novartis, unrelated to the present study. The other authors declare no competing interests.