Clinical and laboratory predictors and prevalence of immune reconstitution inflammatory syndrome in patients with Whipple's disease

Whipple's disease (WD) is a rare and potentially fatal infectious disease caused by Tropheryma whipplei. It is characterized by a long prodromal phase that mimics a rheumatological disease, often leading to immunosuppressant treatment. Immune reconstitution inflammatory syndrome (IRIS) is currently the most important complication of WD, requiring prompt recognition and treatment as it can be fatal. However, epidemiological data on IRIS are scarce. We aimed to identify the clinical and laboratory predictors of IRIS at WD diagnosis and to evaluate whether the prevalence of IRIS has changed over time.

Whipple's disease (WD) is a rare disease caused by Tropheryma whipplei (T.2][3] The natural history of WD is characterized by a long prodromal phase characterized by migratory peripheral seronegative arthritis mimicking rheumatological disease. 4It is therefore quite common for patients with WD to be misdiagnosed as a rheumatological condition at the early phase of the disease and treated with immunosuppressants, including biologics. 5,6[6][7] In the past, cachexia and neurological involvement were the major causes of death in patients with WD.0][11] However, data on the epidemiology of IRIS in patients with WD are scarce. 8In 2010 we and other researchers showed that IRIS affected 10%-20% of patients with WD and that more than 10% of those with IRIS died. 8,12In the following decade, the prevalence, incidence, and severity of IRIS may have been influenced by several factors, such as an increased use of biologic immunosuppressants, and increased clinician awareness about the prevention and treatment of this complication.However, data is lacking on whether the prevalence figures of IRIS have changed over the last decades. In this study we aimed to identify the clinical manifestations and laboratory biomarkers at the time of WD diagnosis that predict subsequent development of IRIS (clinical study) and to investigate whether the epidemiology of IRIS has changed over time (epidemiological study).

| Study population and setting
This was a single-center study conducted in the University of Pavia (Pavia, Italy) based on a retrospective evaluation of clinical and laboratory data of 45 consecutive patients with WD (10 females and 35 males) at a mean age of 52 ± 11 years who were followed up at our institution between January 2000 and December 2021.Patients seen in our institution before September 2011 were previously described. 12Patients were divided into two groups according to whether or not they developed IRIS during the follow-up period.At the time of diagnosis of WD, none of these patients were affected by further oncological, hematological, hepatic, or rheumatological conditions that could have influenced their clinical manifestations and laboratory parameters.
The diagnosis of WD was made based on duodenal biopsy specimens showing a typical pattern with periodic acid Schiff (PAS)positive macrophages infiltrating the mucosa and/or positive polymerase chain reaction (PCR) for T. whipplei in the samples obtained from extragastrointestinal sites. 4,13In accordance with Feurle et al's study, 8 the diagnosis of IRIS was made based on an initial clinical response of symptoms within 3 weeks of antibiotic treatment followed by a recurrence of systemic or local inflammation for at least 1 week.In all patients, recurrence of WD while IRIS was manifested was excluded by means of histological examination and negative PCR for T. whipplei.Other hospital-acquired infection was also carefully excluded.For the purpose of this study, we classified IRIS into minor and major forms according to the absence or presence of at least one of the following criteria: steroid-dependency requiring further immunosuppressants, hospital admission due to IRIS, or death of the patient.

| Data collection
For each patient the demographic and clinical data at the time of diagnosis of WD were retrospectively collected from medical records, including clinical form of WD (classic or localized), 4 age, sex, body mass index (BMI), weight loss, diarrhea, fever, arthralgia, neurological and cardiologic involvement, lymphadenopathy, immunosuppressive therapy before the diagnosis of WD, and laboratory test results.
For the epidemiological part of the study, patients were divided into two groups according to the year of WD diagnosis.The first group consisted of patients diagnosed between 2000 and 2011 and the second group diagnosed during 2012-2021.

| Statistical analysis
Statistical analysis was performed using R version 4.1.2(R Foundation for Statistical Computing, Vienna, Austria; https://www.R-project.org/).Categorical variables were expressed as numbers and percentages, and were compared among groups by using the Fisher's exact test.Continuous variables were expressed as mean ± standard deviation for normally distributed variables or median and interquartile range (IQR) for skewed data.Continuous variables were compared among groups using Student's t-test or Mann-Whitney U-test, when appropriate.Patients with missing data were excluded from relevant analyses.No imputation of data was performed.
For variables that differed significantly between patients who developed IRIS and those who did not, the receiver operating characteristic (ROC) curves were computed to evaluate the discriminatory ability of parameters and the area under the ROC curve (AUROC) and 95% confidence interval (CI) were calculated.The Youden's index was used to identify the optimal cut-off value for discriminating between patients who developed IRIS from those who did not.A two-sided P value of less than 0.05 was considered statistically significant for all analyses.

| Clinical study
All 45 patients with WD were affected by a classic form of WD, of whom 10 (22.2%) developed IRIS.The median follow-up duration was 106 months (IQR 59-142 months).The majority of the patients were treated with ceftriaxone followed by trimethoprim-sulfamethoxazole; a few of them were then switched to doxycycline.All patients who developed IRIS were males, although this was not statistically significant (P = 0.09).Tables 1 and 2 summarize the clinical characteristics and laboratory test results in patients who developed IRIS compared to those who did not.At the time of WD diagnosis, erythrocyte sedimentation rate (ESR) and platelet (PLT) count were significantly higher in WD patients who did not develop IRIS.BMI was higher in patients who developed IRIS.Although there was no statistical significance, C-reactive protein (CRP) showed an increasing trend in patients who subsequently did not develop IRIS.However, previous immunosuppressive therapy was confirmed to be more common in WD patients who developed IRIS (Table 1).Table 3 summarizes the various immunosuppressants used in these patients before the accurate diagnosis of WD was reached; unfortunately, the precise duration of immunosuppressive treatment was unknown.The ROC analysis identified ESR ≤46 mm/h (AUROC 0.88, 95% CI 0.72-1.00)and PLT count ≤327 Â 10 9 /L (AUROC 0.85, 95% CI 0.70-1.00)as the optimal cut-off values to identify patients with WD at the highest risk of developing IRIS during treatment (Figures 1   and 2).F I G U R E 1 Erythrocyte sedimentation rate of 46 mm/h or lower is the optimal cut-off value to identify patients with Whipple's disease who subsequently developed immune reconstitution inflammatory syndrome from those who did not (AUROC 0.88, 95% CI 0.72-1.00)

| Epidemiological study
A platelet count of 327 Â 10 9 /L or lower is the optimal cut-off value to identify patients with Whipple's disease who subsequently developed immune reconstitution inflammatory syndrome from those who did not (AUROC 0.85, 95% CI 0.70-1.00)

| DISCUSSION
This was a single-center retrospective study that primarily identified clinical and laboratory predictors of IRIS in patients with WD, and described trends in the prevalence figures of IRIS in WD patients over two decades.
The main finding of our work was that patients with WD who present at the time of diagnosis with a lesser degree of systemic inflammation have the highest risk of developing IRIS.We also provided cut-off values of ESR and PLT count, which may help identify patients at a high risk of developing IRIS since the time of WD diagnosis who require a more rigorous follow-up and treatment.Interestingly, recent biological data have shown that at the time of WD diagnosis, patients who subsequently developed IRIS had a lower level of mucosal inflammatory chemokines. 11Although it is very tempting to consider our results as the clinical counterpart of these biological results, such parallelism is extremely difficult and needs further studies.
Our second finding was that the prevalence of IRIS in WD has remained stable over two decades (22.2%), although we observed a higher incidence of major forms of IRIS requiring additional immunosuppression or hospital admission in the second decade.A possible explanation for this finding may lie, on one hand, in a greater awareness about IRIS by clinicians, leading to timely recognition and management, and on the other hand in an increased use of immunosuppressants prior to WD diagnosis, which may have counterbalanced each other.In our daily clinical practice, newly diagnosed WD patients have almost invariably been treated with immunosuppressants and at our center we adopted an empirical approach of initially maintaining and then very gradually withdrawing immunosuppressant therapy.Since it is impossible to use this strategy with biologic immunosuppressants, we usually replace them with steroids, which can gradually be tapered.
In 2009 we studied HLA alleles in WD in a multicenter European study, we found that 87.5% of patients with WD complicated by IRIS were HLA-DQ6-positive. 14While in the present study we found HLA-DQ6 in only 50% of the detected WD patients with IRIS.Due to the retrospective nature of our study and the small sample size, we could not further evaluate the role of genetic factors in the development of IRIS in WD.
6][17] It is an infectious disease that, especially in its initial phase, resembles many rheumatological conditions.The vast majority of patients with WD in their prodromal phase are quite often misdiagnosed as having seronegative arthritis and treated with immunosuppressive therapy. 5,6Once the diagnosis of WD is finally reached, there is then a tendency to suspend this immunosuppressive therapy and start antibiotic treatment.However, the sudden withdrawal of immunosuppressants represents the major trigger for the development of IRIS, especially in its severe forms. 8Knowledge of this complication is, therefore, of paramount importance in order to prevent it and appropriately treat it.Since an evidence-based treatment for patients with IRIS is still lacking, treating these patients is challenging.
Antibiotics do not need to be changed, and previous immunosuppressive therapy must be maintained and tapered very slowly.A wait-and-see approach is probably best in the mildest forms. 9Nonsteroidal anti-inflammatory drugs can be added on demand.Steroids must be initiated in patients on biologics or in the rare case of patients developing IRIS despite no previous therapy with immunosuppressants. 8,9High-dose steroids should also be prescribed in the more severe forms and tapered very slowly.In the few steroiddependent patients, in whom symptoms recur at low steroid dosage, thalidomide or other immunomodulatory drugs have been prescribed. 8,9The prompt start of these therapies, which may seem paradoxical in infectious diseases, is effective in the majority of patients. 8,9r study had some clear limitations.The small sample size is the main limitation of our study and is linked to the extreme rarity of this condition, as is exemplified by a recent American study conducted at four referral centers for small bowel diseases where only 33 cases of T.
whipplei infection, including only 18 cases of frank WD, were collected over 20 years. 16Another American registry-based study found a prevalence of WD of 4.6 per 1 million hospitalizations. 17A previous study by our group found a prevalence in northern Italy of three WD cases per 1 million inhabitants. 15Although our sample of 45 patients with WD is certainly worthy of consideration from a clinical point of view, it is nevertheless undeniable that, from a statistical point of view, 45 patients are insufficient to perform a more extensive statistical analysis.Due to the limited sample size, we could not develop a predictive score that, taking into consideration all these parameters, would allow a stratification of the patients according to their risk of developing IRIS.Finally, while lower ESR and PLT count are indeed markers of the subsequent development of IRIS, they cannot provide information on its severity.A close follow-up of these patients is therefore mandatory.
In conclusion, ESR, PLT count, and obviously previous immunosuppressive therapy identify WD patients at higher risk of IRIS.In the future, these parameters should be taken into account to guide treatment and follow-up of patients with WD.
All patients signed informed consent before any invasive procedure, both for clinical and research purposes.The study was approved by the Ethics Committee of San Matteo Hospital Foundation (Pavia, Italy) according to the 1975 Declaration of Helsinki (6th revision, 2008).

Finally
Abbreviations: NS, not significant; SD, standard deviation.
During the first decade(2000-2011), 27 patients (including five females and 22 males) at a mean age of 54 ± 11 years were diagnosed with WD, of whom 6 (22.2%) developed IRIS.As already described, a major form developed in one patient who died.12While during the second decade (2012-2021), WD was diagnosed in 18 patients (including five females and 13 males) at a mean age of 50 ± 10 years, of whom four (22.2%) developed IRIS.Major forms of IRIS were more common during the second decade (100% [4/4] vs 16.7% [1/6], P = 0.047), although none were fatal.Two of these four patients with IRIS in the second decade required hospital admission and azathioprine was started in three out of the four patients.Of these four patients, IRIS resolved in two and markedly improved in the other two.In the majority of patients, IRIS occurred in the first few weeks just after the diagnosis of WD and the initiation of the antibiotics.In only one patient IRIS developed at 8 months after the diagnosis of WD.T A B L E 3 Immunosuppressive drugs used before the diagnosis of Whipple's disease Immunosuppressive drugs (n, %) Patients with IRIS (n = 10) Patients without IRIS (n = 35) Most patients were treated with more than one drug.Abbreviation: IL, interleukin; IRIS, immune reconstitution inflammatory syndrome; NS, not significant; TNF, tumor necrosis factor.