Neurological Manifestations in Whipple's Disease: A Scoping Review of Clinical Features, Laboratory Findings, and Neuroimaging Characteristics of 111 Cases

ABSTRACT

Background

Whipple disease (WD) is a rare, potentially fatal infection caused by Tropheryma whipplei (TW). Central Nervous System involvement (CNS‐WD) is a severe complication of WD with a challenging diagnosis, due to limited data integrating neurological features, imaging, cerebrospinal fluid analysis, PCR, and intestinal histopathology. Prompt recognition and treatment are essential to improve patient outcomes.

Methods

A scoping review of the literature from 2000 to 2025 was conducted, including all CNS‐WD case reports and series, to outline its clinical spectrum and diagnostic workup.

Results

Included studies were 89, accounting for 111 CNS‐WD cases. Most patients were male (79.8%), with a mean age of 53.6 years. Notably, 13.2% presented with CNS involvement in the absence of systemic WD. Pathognomonic signs were rare (13/111, 11.7%). Most patients (82.9%, 92/111) presented with two or more neurological symptoms, the most frequently reported being altered mentation, followed by ataxia, eye movement abnormalities, and other movement disorders. CSF was abnormal in 74.7% (56/75) of patients and TW PCR was positive in 75% (48/64); 64% (55/86) of cases showed intestinal histopathology consistent with WD. TW PCR on gastrointestinal biopsies was positive in 67.9% (19/28) cases. CNS imaging revealed abnormalities in 85.4% (82/96) of patients, with no distinctive pattern.

Conclusions

CNS‐WD diagnosis remains challenging due to nonspecific symptoms, variable imaging, and occasional discordance among CSF, TW PCR or biopsy findings. Our review suggests that CNS‐WD should be considered in the differential diagnosis of unexplained neurological syndromes, even in patients with atypical clinical or radiological profiles, supporting broader indications for TW testing.

1. Introduction

Whipple's disease (WD) is a rare but potentially fatal disorder caused by Tropheryma whipplei (TW). Most people who are infected with TW become asymptomatic carriers [1, 2, 3, 4, 5, 6], while a small subset of predisposed individuals with underlying immunological abnormalities develop WD [7, 8]. Historically, only the classical form of WD was recognized, primarily affecting the gastrointestinal tract and presenting with diarrhea, weight loss, and arthralgia [9]. The typical course of WD begins with an initial symptomatic phase, with intermittent arthralgia and/or arthritis lasting approximately 6–8 years [10, 11], followed by the classical phase, which is characterized by gastrointestinal manifestations (chronic diarrhea and abdominal pain), as well as weight loss, fever, lymphadenopathy and joint symptoms. In the late disseminated phase, the infection may involve any organ, including the central nervous system (CNS) [7]. Frequently, in its initial presentation, WD is misdiagnosed as seronegative arthritis [10, 11, 12], leading to inappropriate immunosuppressive therapy that may accelerate TW dissemination and increase the risk of CNS involvement [13].

Over the past decades, advances in molecular biology techniques have permitted the detection of TW in various biological samples, including the cerebrospinal fluid (CSF) [14], refining clinical diagnosis and expanding the recognized disease spectrum [11, 15, 16, 17]. CNS‐WD is now increasingly described; it is estimated to occur in 21%–43% of patients with other clinical manifestations of WD (WD with secondary CNS involvement) [11, 18, 19], but it can also present as an isolated neurological disease (WD with primary CNS involvement), in approximately 4% of WD cases [10, 19]. However, there is no universally accepted clinical and diagnostic definition of CNS‐WD. Although pathognomonic signs of WD‐CNS have been identified—namely oculomasticatory myorhytmia (OMM) and oculofacioskeletal myorhythmia (OFSM)—only a minority of patients present with them [20]. Previous reviews, which are limited by small sample size, indicated that most affected individuals present with nonspecific neurologic symptoms, including cognitive decline [21], motor impairment, and oculomotor dysfunction [18, 22, 23, 24] accompanied by diverse CSF abnormalities [18, 22, 23] further complicating diagnosis. Similarly, neuroimaging findings have been reported to include a broad and heterogeneous spectrum of abnormalities [25]. Nevertheless, systematic data integrating CSF profiles, T. whipplei PCR across multiple biological specimens, and intestinal histopathological findings in patients with CNS WD remain lacking.

Diagnosing WD per se and as well as WD with CNS involvement remains challenging, and delayed recognition is common [26].

CNS‐WD carries a poor prognosis, with approximately 25% mortality and an additional 25% of patients developing severe long‐term neurological impairment [27]. Hence, early recognition and prompt antibiotic treatment are crucial for improving patient outcomes [28].

This review summarizes the clinical spectrum and key diagnostic features of CNS‐WD, based on recent literature, aiming to facilitate earlier and more accurate identification of CNS‐WD.

2. Methods

This was a scoping review of the literature. The study protocol was registered with Prospero (https://www.crd.york.ac.uk/PROSPERO/home) on 1st February 2022 (ID CRD42022308202).

2.1. Eligibility Criteria

We reviewed case reports and case series describing patients with CNS‐WD diagnosis, published between January 2000 and June 2025, and restricted to English, French, Spanish and Italian. We included WD cases presenting with neurologic symptoms, either with primary or secondary CNS involvement. We included only cases with a diagnosis of WD based on: (i) pathognomonic neurologic signs; (ii) positive T. whipplei PCR in at least one relevant biological specimen (e.g., blood, urine, synovial fluid, vitreous humor); (iii) PAS‐positive staining in intestinal biopsies or lymph node biopsy. We excluded cases describing patients with pre‐existing comorbidities causing neurologic symptoms that could confound the clinical presentation; WD patients with CNS manifestations not related to direct parenchymal involvement such as embolism from T. whipplei endocarditis, seizures secondary to ionic changes; patients with isolated posterior ocular involvement such as posterior uveitis, retinitis, or optic neuritis and cases in which the diagnosis was based solely on PAS‐positive staining in brain biopsy, given its poor specificity.

2.2. Information Sources, Search Strategy, and Selection Process

We searched MEDLINE (PubMed), EMBASE, and The Cochrane Library on August 20th, 2020, with updates on June 30th, 2025. The full search strategy and keywords used for this review are provided in Appendix S1.

Search results were combined, duplicates removed, and all records imported into Rayyan software for screening [29]. Two authors independently screened titles and abstracts, followed by full‐text assessment for eligibility. Data extraction was performed independently by two authors (L.M. and D.T.) using a predesigned Excel file (version 2402, Microsoft Corporation).

2.3. Data Synthesis

Detailed information was collected on demographics (age at onset of neurological symptoms, and sex), prior immunosuppressive therapy, clinical presentation (systemic and neurological symptoms and signs), diagnostic findings (CSF characteristics, histopathology, T. whipplei PCR results from various biological samples, neuroimaging), and outcomes (death, improvement or stabilization).

We classified cases according to diagnostic certainty, distinguishing confirmed from possible CNS‐WD and evaluating signs and symptoms according to the pattern of involvement, distinguishing primary (isolated) or secondary (systemic) CNS involvement, using the following definitions:

• –Confirmed CNS‐WD: Presence of pathognomonic neurologic signs (oculomasticatory myorhythmia [OMM] or oculofacioskeletal myorhythmia [OFSM]) and/or a positive T. whipplei PCR in CSF or brain tissue
• –Possible CNS‐WD: Presence of neurological symptoms together with at least one of the following findings: (i) PAS‐positive staining in lymph node or intestinal biopsies; (ii) positive T. whipplei PCR in another relevant biological specimen (e.g., blood, urine, synovial fluid, vitreous humor)
• –CNS‐WD with primary CNS involvement: Isolated neurological WD without systemic manifestation
• –CNS‐WD with secondary CNS involvement: Classical form of WD with secondary brain involvement

2.4. Statistical Analysis

Descriptive statistics were used to summarize the extracted cases. Results were reported as counts and percentages and organized by key domains, including general characteristics, clinical presentation, diagnostic findings, and outcomes. Categorical variables were summarized using absolute frequencies and percentages, whereas continuous variables were summarized using means and standard deviations (SD). Group comparisons—between confirmed and possible CNS‐WD, and between CNS‐WD with primary and secondary CNS involvement—were performed using Fisher's exact test or Pearson's chi‐squared test for categorical variables, and the Wilcoxon rank‐sum test for continuous variables. All statistical tests were two‐sided and conducted at a 5% significance level. Data analysis and figure generation were performed using R statistical software (version 4.5.1).

3. Results

The search strategy permitted to retrieve 299 articles published between January 2000 and June 2025. After removal of 18 duplicate records, 157 articles were excluded based on title and abstract screening, and an additional 35 were excluded after full‐text evaluation. Ultimately, 89 studies were included in the final analysis, accounting for 111 cases of CNS‐WD. The study selection process is illustrated in Figure 1.

FIGURE 1.

Flow chart.

3.1. Demographic, Risk Factors, Clinical Characteristics, and Outcome

A total of 111 cases of CNS‐WD were included in the analysis (Table 1). The mean age at disease presentation was 53.6 years (SD, 12.11; range, 22–80 years), and 79.8% (87/109) of patients were male. A history of immunosuppressive therapy before the onset of neurological symptoms was reported in 18.2% (18/99) of patients. Based on clinical manifestation and laboratory findings, 70 (51.5%) patients were classified as having confirmed CNS‐WD, and 41 (48.5%) as possible CNS‐WD. The pattern of CNS involvement could not be determined for five patients. Among the remaining 106 cases, CNS‐WD with primary CNS involvement was documented in 13.2% (14/106), while CNS‐WD with secondary CNS involvement was identified in 86.8% (92/106). Confirmed WD was reported in 78.6% (11/14) of CNS‐WD cases with primary CNS involvement and in 58.7% (54/92) of those with secondary CNS involvement. Overall, the most frequent non‐neurological manifestations were arthralgia and gastrointestinal symptoms, respectively 50.9% (54/106) and 47.2% (50/106), followed by weight loss (46.2%, 49/106) and fever (25.5% (27/106) (Figure 2)). Overall, 9.4% (10/106) of patients died, whereas 90.6% (96/106) showed clinical improvement or stabilization after treatment.

TABLE 1.

Demographics, risk factor, clinical manifestation, and outcomes of patients with Whipple Disease and Central Nervous System Involvement (CNS‐WD), according to diagnostic certainty (N = 111) and type of neurologic involvement (N = 106), 2000–2025.

	CNS‐WD according to
	Diagnostic certainty (N = 111)					Type of neurologic involvement (N = 106)
	NA	Overall (N = 111 a )	Confirmed CNS‐WD (N = 70 a )	Possible CNS‐WD (N = 41 a )	p b	NA	Overall (N = 106 a , b )	Secondary CNS‐WD (N = 92 a )	Primary CNS‐WD (N = 14 a )	p c
Demographics
Age, years d	0				0.852	0				0.762
Mean (SD)		53.64 (12.11)	53.44 (12.79)	53.99 (11.02)			54.17 (12.00)	54.05 (11.66)	54.93 (14.51)
Min, Max		22.00, 80.00	22.00, 80.00	28.00, 75.00			22.00, 80.00	22.00, 76.00	28.00, 80.00
Sex	2				0.396	2				0.730
Female		22/109, 20.2%	12/68, 17.6%	10/41, 24.4%			21/104, 20.2%	19/90, 21.1%	2/14, 14.3%
Male		87/109, 79.8%	56/68, 82.4%	31/41, 75.6%			83/104, 79.8%	71/90, 78.9%	12/14, 85.7%
Risk factors
Immunosuppressive therapy e	12	18/99, 18.2%	12/64, 18.8%	6/35, 17.1%	0.843	7	18/99, 18.2%	18/85, 21.2%	0/14, 0.0%	0.067
CNS‐according to diagnostic certainly
Confirmed CNS‐WD		70/111 (63.1%)	70/70, 100%	0	—		65/106, 61.3%	54/92, 58.7%	11/14, 78.6%	0.155
Possible CNS‐WD		41/111 (36.9%)	0	41/41 (100%)	—		41/106, 38.7%	38/92, 41.3%	3/14, 21.4%
CNS‐WD according to neurologic involvement
Primary CNS‐WD	5	14/106, 13.2%	11/65, 16.9%	3/41, 7.3%	0.155	0	14/106, 13.2%	0	14/14, 100%
Secondary CNS‐WD		92/106 86.8%				0	92/106, 86.8%	92/92, 100%	0
Systemic manifestations
Gastrointestinal	5	50/106, 47.2%	30/65, 46.2%	20/41, 48.8%	0.792	0	50/106, 47.2%	50/92, 54.3%	0/14, 0.0%	< 0.001
Arthralgia	5	54/106, 50.9%	31/65, 47.7%	23/41, 56.1%	0.399	0	54/106, 50.9%	54/92, 58.7%	0/14, 0.0%	< 0.001
Weight loss	5	49/106, 46.2%	30/65, 46.2%	19/41, 46.3%	0.985	0	49/106, 46.2%	49/92, 53.3%	0/14, 0.0%	< 0.001
Fever	5	27/106, 25.5%	13/65, 20.0%	14/41, 34.1%	0.104	0	27/106, 25.5%	27/92, 29.3%	0/14, 0.0%	0.019
Cutaneous	5	13/106, 12.3%	7/65, 10.8%	6/41, 14.6%	0.555	0	13/106, 12.3%	13/92, 14.1%	0/14, 0.0%	0.208
Ocular	5	11/106, 10.4%	5/65, 7.7%	6/41, 14.6%	0.330	0	11/106, 10.4%	11/92, 12.0%	0/14, 0.0%	0.352
Clinical course	5				0.742	5				0.613
Death		10/106, 9.4%	7/67, 10.4%	3/39, 7.7%			10/101, 9.9%	8/88, 9.1%	2/13, 15.4%
Improvement or stabilization f		96/106, 90.6%	60/67, 89.6%	36/39, 92.3%			91/101, 90.1%	80/88, 90.9%	11/13, 84.6%

Note: Bold p‐values indicate statistical significance (p < 0.05).

Abbreviations: CNS, Central Nervous System; CNS‐WD, Whipple Disease and Central Nervous System Involvement; NA, Not Available.

^a n/N, %.

^b For 5 patients the type of neurological involvement was not available.

^c Wilcoxon rank sum test; Pearson's Chi‐squared test; Fisher's exact test.

^d Age at diagnosis.

^e Before the onset of neurologic manifestations.

^f Includes both clinical improvement and stabilization after treatment.

FIGURE 2.

Frequency of non‐neurologic manifestations among 111 patients with Whipple Disease and Central Nervous System Involvement (CNS‐WD), 2000–2025.

No significant differences were observed between confirmed and possible CNS‐WD and between primary and secondary CNS‐WD with respect to demographic characteristics, risk factor, clinical presentation, and clinical outcome (Table 1). Mortality did not differ statistically (p = 0.613) between patients with primary CNS involvement (2/13, 15.4%) and those with secondary CNS involvement (8/88, 9.1%).

3.2. Neurologic Manifestations

At presentation, 17.1% (19/111) of patients with CNS‐WD had one neurologic symptom, 37.8% (42/111) had two or three symptoms, and 45% (50/111) had more than three symptoms (Table 2). The most common clinical feature was altered mentation, including cognitive, behavioral, and alertness disturbances, reported in 74.8% (83/111) of cases. Ataxia, imbalance, or clumsiness were reported in 40.5% (45/111) of cases, followed by eye movement disorders in 37.8% (42/111), movement disorders in 31.5% (35/111), sleep disorders in 21.6% (24/111), and dysarthria or other speech impairments in 19.8% (22/111). Myorhythmia was reported in 18.9% (21/111), upper motor neuron deficit in 17.1% (19/111) of cases, OMM or/and OFSM, both pathognomonic signs of CNS‐WD, in 11.7% of patients (13/111), focal neurologic deficit in 9.9% (11/111).

TABLE 2.

Neurologic manifestations in patients with Whipple Disease and Central Nervous System Involvement (CNS‐WD), according to diagnostic certainty (N = 111) and type of neurologic involvement (N = 106), 2000–2025.

	CNS‐WD according to
	Diagnostic certainty (N = 111)					Type of neurologic involvement (N = 106)
	NA	Overall (N = 111 a )	Confirmed WD‐SNC (N = 70 a )	Possible CNS‐WD (N = 41 a )	p b	NA	Overall (N = 106 a )	Secondary CNS‐WD (N = 92 a )	Primary CNS‐WD (N = 14 a )	p b
Quantity
Number of symptoms	0				< 0.001	0				0.020
1		19/111, 17.1%	5/70, 7.1%	14/41, 34.1%			19/106, 17.9%	15/92, 16.3%	4/14, 28.6%
2–3		42/111, 37.8%	27/70, 38.6%	15/41, 36.6%			40/106, 37.7%	39/92, 42.4%	1/14, 7.1%
> 3		50/111, 45.0%	38/70, 54.3%	12/41, 29.3%			47/106, 44.3%	38/92, 41.3%	9/14, 64.3%
Type
Altered mentation c	0	83/111, 74.8%	54/70, 77.1%	29/41, 70.7%	0.453	0	80/106, 75.5%	70/92, 76.1%	10/14, 71.4%	0.742
Ataxia/imbalance/clumsiness	0	45/111, 40.5%	34/70, 48.6%	11/41, 26.8%	0.024	0	41/106, 38.7%	32/92, 34.8%	9/14, 64.3%	0.035
Eye movement disorders	0	42/111, 37.8%	28/70, 40.0%	14/41, 34.1%	0.539	0	39/106, 36.8%	36/92, 39.1%	3/14, 21.4%	0.201
Movement disorders	0	35/111, 31.5%	23/70, 32.9%	12/41, 29.3%	0.695	0	33/106, 31.1%	29/92, 31.5%	4/14, 28.6%	> 0.999
Sleep disorders	0	24/111, 21.6%	16/70, 22.9%	8/41, 19.5%	0.679	0	24/106, 22.6%	22/92, 23.9%	2/14, 14.3%	0.732
Dysarthria and other speech impairments	0	22/111, 19.8%	16/70, 22.9%	6/41, 14.6%	0.294	0	21/106, 19.8%	16/92, 17.4%	5/14, 35.7%	0.146
Myorhythmia	0	21/111, 18.9%	18/70, 25.7%	3/41, 7.3%	0.017	0	20/106, 18.9%	18/92, 19.6%	2/14, 14.3%	> 0.999
Upper motor neuron deficit	0	19/111, 17.1%	10/70, 14.3%	9/41, 22.0%	0.301	0	17/106, 16.0%	13/92, 14.1%	4/14, 28.6%	0.234
OMM and/or OFSM d	0	13/111, 11.7%	13/70, 18.6%	0/41, 0.0%	0.002	0	12/106, 11.3%	11/92, 12.0%	1/14, 7.1%	> 0.999
Focal neurologic deficits	0	11/111, 9.9%	9/70, 12.9%	2/41, 4.9%	0.208	0	11/106, 10.4%	8/92, 8.7%	3/14, 21.4%	0.159
Hypothalamic disorders	0	9/111, 8.1%	7/70, 10.0%	2/41, 4.9%	0.481	0	7/106, 6.6%	4/92, 4.3%	3/14, 21.4%	0.047
Intracranial hypertension	0	2/111, 1.8%	1/70, 1.4%	1/41, 2.4%	> 0.999	0	2/106, 1.9%	2/92, 2.2%	0/14, 0.0%	> 0.999
Autonomic disorders	0	1/111, 0.9%	0/70, 0.0%	1/41, 2.4%	0.369	0	1/106, 0.9%	1/92, 1.1%	0/14, 0.0%	> 0.999
Seizures	0	1/111, 0.9%	1/70, 1.4%	0/41, 0.0%	> 0.999	0	1/106, 0.9%	1/92, 1.1%	0/14, 0.0%	> 0.999
Not specific	0	46/111, 41.4%	30/70, 42.9%	16/41, 39.0%	0.692	0	45/106, 42.5%	41/92, 44.6%	4/14, 28.6%	0.259

Note: Bold p‐values indicate statistical significance (p < 0.05).

Abbreviations: CNS‐WD, Whipple Disease and Central Nervous System Involvement; NA, Not Available.

^a n/N, %.

^b Wilcoxon rank sum test; Pearson's Chi‐squared test; Fisher's exact test.

^c Altered mentation, including cognitive, behavioral, and alertness disturbances.

^d Pathognomonic signs permitting classification of the confirmed CNS‐WD.

The number of neurologic symptoms differed significantly according to the diagnostic certainty (p < 0.001) (Table 2). Among patients with confirmed CNS‐WD, 7.1% (5/70) reported one symptom, while 54.3% (38/70) reported more than three symptoms. In contrast, among patients with possible CNS‐WD, 34.1% (14/41) reported one symptom and 29.3% (12/41%) reported more than three symptoms. Excluding the two pathognomonic signs (OMM and OFSM), which were observed exclusively in confirmed cases ataxia, imbalance, or clumsiness and myorhythmia were significantly more frequent in confirmed than in possible cases. Ataxia, imbalance, or clumsiness was reported in 48.6% (34/70) of confirmed cases compared with 26.8% (11/41) of possible cases (p = 0.024). Similarly, myorhythmia was observed in 25.7% (18/70) of confirmed cases versus 7.3% (3/41) of possible cases (p = 0.017). No significant differences were observed for the other neurologic symptoms (Table 2).

Also, the number of neurologic symptoms reported differed significantly according to the type of CNS involvement (p = 0.020) (Table 2). Among patients with CNS‐WD with secondary CNS involvement, 16.3% (15/92) reported one symptom, 42.4% (39/92) with two to three symptoms, and 41.3% (38/92) with more than three symptoms. In contrast, among patients with primary CNS involvement, 28.6% (4/14) reported one symptom, 7.1% (1/14) reported two to three symptoms, and 64.3% (9/14) reported more than three neurologic symptoms. When comparing patients with primary versus secondary CNS involvement, ataxia, imbalance, or clumsiness were reported significantly more frequently in cases with primary CNS involvement (64.3%, 9/14) than in those with secondary CNS involvement (34.8%, 32/92; p = 0.035). Hypothalamic disorders were significantly more common in primary CNS‐WD (3/14, 21.4%) than in secondary CNS‐WD (4.92, 4.3%, p = 0.047). No relevant differences were found for the other neurological symptoms (Table 2). The frequency of combined neurological symptoms, expressed as percentages among 111 patients with CNS‐WD, is shown in Figure 3.

FIGURE 3.

Patterns of neurological symptom combinations among 111 patients with Whipple Disease and Central Nervous System Involvement (CNS‐WD), 2000–2025.

3.3. Laboratory Findings and Imaging

The analysis of laboratory findings and imaging was conducted in 106 patients with a specified type of neurologic involvement (Table 3). CSF analysis was available for 75 of the 106 cases (70.7%) and abnormal findings were reported in 74.7% of these patients (56/75). Among patients with abnormal CSF results, the median cell count was 34 cells/mm³ (interquartile range, 17–88), with a predominance of lymphocytes, and the median protein concentration was 146.00 mg/dL (interquartile range, 72.00–250.00). PCR testing for T. whipplei was performed on CSF samples in 64 patients (60.4%), resulting positive in 75% (48/64) cases. Among patients with cognitive impairment CSF PCR resulted positive in 74.1% (40/54) and in 100% (6/6) of patients with hydrocephalus.

TABLE 3.

Laboratory in patients with Whipple Disease and Central Nervous System Involvement (CNS‐WD), overall and according to type of neurologic involvement (N = 106), 2000–2025.

	NA	Overall (N = 106 a )	Secondary CNS‐WD (N = 92 a )	Primary CNS‐WD (N = 14 a )	p b
CSF	31				0.219
Normal		19/75, 25.3%	15/66, 22.7%	4/9, 44.4%
Abnormal		56/75, 74.7%	51/66, 77.3%	5/9, 55.6%
CSF cells (number/μL)	77				0.863
Median (Q1, Q3)		34.00 (17.00, 88.00)	34.00 (17.00, 88.00)	235.50 (7.00, 464.00)
Min, max		6.00, 35,000.00	6.00, 35,000.00	7.00, 464.00
CSF proteins (mg/dL)	79				0.108
Median (Q1, Q3)		146.00 (72.00, 250.00)	141.50 (72.00, 226.00)	202,700.00 (202,700.00, 202,700.00)
Min, max		47.00, 202,700.00	47.00, 107,600.00	202,700.00, 202,700.00
PCR for T. whipplei , positive
CSF	42	48/64, 75.0%	39/53, 73.6%	9/11, 81.8%	0.716
Stool	97	8/9, 88.9%	8/9, 88.9%	0/0, NA%	> 0.999
Saliva	100	5/6, 83.3%	5/6, 83.3%	0/0, NA%	> 0.999
Blood	85	15/21, 71.4%	11/17, 64.7%	4/4, 100.0%	0.620
Cerebral biopsy	89	14/17, 82.4%	11/12, 91.7%	3/5, 60.0%	0.191
Gastrointestinal biopsy	78	19/28, 67.9%	17/25, 68.0%	2/3, 66.7%	> 0.999
Histology, compatible with WD
Cerebral biopsy	84	18/22, 81.8%	14/15, 93.3%	4/7, 57.1%	0.077
Gastro‐intestinal biopsy	20	55/86, 64.0%	53/78, 67.9%	2/8, 25.0%	0.023

Note: Bold p‐value indicate statistical significance (p < 0.05).

Abbreviations: CNS, Central nervous system; CSF, Cerebrospinal fluid; NA, Not available.

^a n/N, %.

^b Pearson's chi‐squared test or Fisher's exact test. The median cell count was calculated exclusively for abnormal CSF samples. The median protein level was calculated exclusively for abnormal CSF samples.

Thirty‐six patients had positive T.whipplei PCR on CSF and available CSF biochemical analysis. CSF findings were normal in 9 patients (25.0%) and abnormal in 27 patients (75.0%). PCR testing for T. whipplei was also positive in 88.9% of stool samples (8/9), 83.3% of saliva samples (5/6), and 71.4% of blood samples (15/21). PCR performed on biopsy specimens was positive in 82.4% of brain tissue samples (14/17) and in 67.9% of gastrointestinal biopsy samples (19/28). Histologic findings suggestive of WD were reported in 81.8% of cerebral biopsy specimens (18/22), and in 64.0% of intestinal biopsy specimens (55/86).

When comparing patients with primary versus secondary CNS involvement, those with secondary CNS involvement had a higher proportion of gastrointestinal histology compatible with WD (67.9% [53/78] vs. 25.0% [2/8] p = 0.023).

CNS imaging, including computed tomography (CT) or magnetic resonance imaging (MRI), revealed abnormal findings in 85.4% of cases (82/96). A wide range of abnormalities was observed, with no pathognomonic nor characteristic lesions identified. In the majority of cases, lesions were described as widespread (61/96, 63.5%), whereas in 9.4% of cases (9/96) the lesion was focal and described as a mass. Atrophy was reported in 22/96 cases (22.9%), and hydrocephalus in 9/96 cases (9.4).

An overview of neurological manifestations and diagnostic findings—including abnormalities on CSF chemical–physical analysis, cerebral imaging abnormalities, and positivity for T. whipplei PCR in CSF—among 111 patients with CNS‐WD reported between 2000 and 2025 is illustrated in Figure 4. Figure 5 provides an overview of the diagnostic findings among the 111 patients with CNS‐WD, together with the classification of CNS‐WD as possible or certain.

FIGURE 4.

Patient‐level distribution of neurological manifestations and diagnostic findings among 111 patients with Whipple Disease and Central Nervous System Involvement (CNS‐WD), 2000–2025.

FIGURE 5.

Patient‐level distribution of diagnostic findings among 111 patients with Whipple Disease and Central Nervous System Involvement (CNS‐WD), 2000–2025.

4. Discussion

This is a comprehensive review of 89 studies including 111 patients diagnosed with WD and CNS involvement. Most patients were males (79.8%), in line with previous reports [11, 19, 22, 23, 24]. The mean age at diagnosis was 53.6 years, slightly lower than that reported in earlier studies (55–56 years) [11, 19, 22]. However, the age distribution in our sample was broad, ranging from 22 to 80 years, so CNS‐WD should be considered a possibility almost at any age.

Sixty‐three percent of cases were confirmed WD. Most patients had secondary rather than primary CNS involvement (86.8% and 13.2%, respectively). Patients who had received immunosuppressive therapy prior to the onset of neurological manifestations were 18.2%; none of them had primary CNS involvement. This observation confirms that immunosuppressive therapy, administered for misdiagnosis, may facilitate dissemination of T. whipplei , leading to secondary neurological involvement [13].

In our cohort, patients with CNS‐WD presented a lower frequency of gastrointestinal and articular symptoms and weight loss than previously reported [18, 24], highlighting the need to consider WD even in the absence of its typical clinical profile.

Among patients with CNS‐WD and secondary CNS involvement, 83.7% presented with at least two associated neurological symptoms, consistent with previous observations [18]. The most frequently reported manifestations in this category were altered mentation (76.1%), eye movement disorders (39.1%), gait and coordination disturbances (34.8%), and movement disorders (31.5%), in line with findings from prior analyses of smaller case series [18, 22, 23]. Nonspecific neurological symptoms were frequent (44.6%), whereas pathognomonic signs, such as OMM and OFSM, were observed in only 12% of patients.

PCR for T. whipplei on saliva and stool is recommended as a first‐line, noninvasive screening tool for WD [30], whereas more invasive procedures, such gastrointestinal biopsies, lumbar puncture and cerebral biopsy (when CNS involvement is suspected) mightfollow [7]. In our cohort of patients with secondary CNS involvement, the most sensitive diagnostic test was PCR performed on stool samples (88.9%), followed by PCR on saliva (83.3%). However, PCR performed alone on mucosal samples (e.g., saliva or stool) is known to have a relatively low positive predictive value, which may improve when the bacterial load is high [30]. Meyer et al. also suggested a PCR Ct value cutoff [31] to differentiate between patients with WD from asymptomatic carriers. Unfortunately, data on Ct values were not available in the vast majority of the included cases, so we could not evaluate this aspect.

Notably, blood PCR was positive in 64.7% of patients, supporting its potential role as a complementary diagnostic tool in disseminated disease [7]. Among patients with CNS‐WD and secondary CNS involvement, histological findings suggestive of WD were identified in 67.9% of gastrointestinal biopsies, and PCR for T. whipplei was positive on gastrointestinal tissues in 68% of cases. It has been previously reported that patients with classic WD may display normal small bowel histology without the typical PAS‐positive staining in 9.4% of cases (18/191) [30]. In our cohort, the proportion of negative histological results was even higher. Worth noting, in the majority of the cases the immunopathological findings were not detailed, and immunohistochemistry was not reported. Therefore, this aspect could not be fully addressed. 77.3% of patients with secondary CNS involvement had pathological cerebrospinal fluid findings and 73.6% had a positive CSF PCR for T. whipplei . Overall, clinicians should be aware that normal CSF findings on routine chemical and physical analysis, and even a negative T. whipplei PCR, do not fully exclude the possibility of CNS‐WD. This is consistent with the observations of Compain et al. [23]. In fact, when cerebral biopsy was performed (22 patients), histology was suggestive of WD in 81.8% of cases, and PCR for T. whipplei was positive in 82.4%. Worth noting, we did not gather information about the type of PCR, which was not specified in most reports. This adds some uncertainty about the interpretation of accuracy of the PCR results. For instance, more recent molecular methods such as the quantitative real‐time PCR for the detection of specific 16S–23S rRNA intergenic spacer regions, repeated gene sequences, or the rpoB gene have higher specificity and sensitivity than previous, conventional methods [32].

The proportion of patients with primary CNS‐WD (13.2%) in our study exceeded the one reported in previous studies (4%) [10, 19]. This difference may reflect advancements in diagnostic techniques and increased clinician awareness, resulting in a higher detection rate in recent years; nevertheless, the potential influence of publication bias cannot be excluded. More than half of all cases had more than three neurologic symptoms, and gait and coordination disturbances were more frequent in primary CNS WD compared to secondary CNS involvement (64.3 vs. 34.8%, p = 0.035). OMM and OFSM were reported in only one patient, emphasizing that CNS‐WD is possible even in the absence of typical extra‐cerebral manifestations or pathognomonic features.

Two subgroups of patients with CNS‐WD are of particular interest from the clinical point of view, because their condition is potentially treatable, thus proper diagnosis and treatment might have a substantial impact on their lives; these are patients presenting with neuropsychiatric manifestations—including cognitive, behavioral, and alertness disorders, as well as psychiatric disturbances—and patients with hydrocephalus. In such cases, clinicians should consider WD, as previously suggested [14, 25, 26], and perform a lumbar puncture with CSF Tropheryma whipplei PCR testing, irrespective of CSF testing abnormalities (overall 1/4 of CNS‐WD individuals have normal CSF).

Neuroimaging was abnormal in 85.4% of cases, showing a wide range of nonspecific findings, with no pathognomonic nor characteristic lesions. This is consistent with previous reports, which have failed to identify a specific neuroimaging pattern associated with CNS‐WD [25]. Indeed, CNS‐WD may present with highly variable radiological features, including focal or widespread lesions, or even normal imaging. When present, the most commonly reported abnormality in the literature is increased T2/FLAIR signal intensity, observed in more than 80% of cases, typically involving the midbrain, hypothalamus, thalamus, periaqueductal gray matter, quadrigeminal plate, corticospinal tracts, and supratentorial regions [25, 33, 34, 35, 36]. Contrast enhancement is often minimal or transient, particularly in the acute phase, and lesions typically do not show restricted diffusion on DWI, a feature that may aid in the differential diagnosis from stroke, abscess, or metastatic lesions. A detailed analysis of MRI lesion characteristics was not performed.

Given the wide spectrum of radiological lesions, clinicians and radiologists should maintain a high index of suspicion for WD in patients with atypical brainstem or diencephalic lesions, particularly in the absence of diffusion restriction. However, considering the potential overlap with other inflammatory, neoplastic, or demyelinating conditions, the diagnostic process should integrate imaging findings with clinical, laboratory, and, when necessary, histopathological data.

This study has several limitations. First, there is no uniform definition of CNS involvement in WD in the medical literature; to address this, we applied strict predefined criteria of inclusion for study purposes. Specifically, we decided to exclude cases of CNS‐WD where the diagnosis was based on unvalidated tests or tests with poor specificity, such as PAS‐positive staining on brain biopsy, in order to minimize diagnostic uncertainty. This approach may have resulted in the exclusion of some true CNS‐WD cases, particularly those reported from settings where molecular testing was not widely available and PAS staining represented the only feasible diagnostic method. Second, data quality was limited by reliance on published case reports. Given the multisystemic nature of WD, cases were reported across different medical specialties, which may have introduced inconsistencies, particularly in the description of neurological manifestations and neuroimaging findings. Third, we included only symptomatic patients with parenchymal CNS involvement, excluding cases with unrelated CNS manifestations, myelopathies, or isolated posterior ocular involvement. While this approach aimed to improve cohort homogeneity, it may have introduced a selection bias. Fourth, in patients with prolonged disease courses and multiple CSF, MRI, or CT evaluations, only the most severe findings were recorded, potentially overestimating clinical, biological, or radiological severity. Fifth, we acknowledge that some relevant aspects could not be fully addressed (e.g., type of PCR assay, Ct values) due to limited or unavailable data. Finally, limited information on follow‐up duration and relapse prevented assessment of recurrence risk across different CNS‐WD phenotypes. Despite these limitations, this work includes a larger number of cases and a broader scope of analysis than previous reviews focusing on CNS‐WD.

5. Conclusion

CNS‐WD is a rare but severe condition, with a challenging diagnosis. In fact, it is characterized by heterogeneous clinical, laboratory, and radiological features, and the frequent absence of pathognomonic signs. Primary CNS involvement accounted for nearly one‐seventh of our cases, underscoring the clinical relevance of this poorly understood disease subtype. Our findings support a stepwise diagnostic approach integrating clinical suspicion, molecular testing on noninvasive samples, cerebrospinal fluid analysis, and targeted tissue biopsy when appropriate. Neuroimaging findings are frequent but nonspecific, reinforcing the need to interpret radiological abnormalities within the broader clinical and microbiological context. Early inclusion of WD in the differential diagnosis of unexplained neurological syndromes, particularly when accompanied by systemic manifestations, may help prevent inappropriate immunosuppressive therapy, reduce disease dissemination, and improve outcomes. Future prospective and multicenter studies are needed to better define the natural history, optimal diagnostic algorithms, and long‐term outcomes of CNS‐WD, especially in patients with primary CNS involvement.

Author Contributions

Lucia Moro: conceptualization, writing – review and editing, methodology, data curation, investigation, writing – original draft. Domenicantonio Tropepi: conceptualization, writing – review and editing, data curation, investigation. Dora Buonfrate: methodology, writing – review and editing, project administration. Cristina Mazzi: formal analysis, methodology. Giovanni Foti: writing – review and editing. Andrea Fittipaldo: data curation, methodology, writing – review and editing. Andrea Angheben: conceptualization, investigation, writing – review and editing, supervision. Anna Beltrame: conceptualization, writing – review and editing, investigation, supervision.

Funding

This work was supported by the Italian Ministry of Health—“Fondi Ricerca Corrente” to IRCCS Sacro Cuore Don Calabria hospital under the project L3P1.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Appendix S1

Data Availability Statement

The data that support the findings of this study are openly available in Zenodo at https://zenodo.org/.