on 21-Oct-2022 (Fri)

Definite classic WD was defined by positive periodic acid-Schiff (PAS) staining and probable WD by specific positive TW polymerase chain reaction (PCR) of intestinal specimens.

Cases of TW infection diagnosed from 1995 to 2010 were identified in three US referral centers and from 1995 to 2015 in one

Development of a polymerase chain reaction (PCR) assay in 1991 [8] allowed expansion of diagnostic testing to a multitude of tissues and body fluids [9–14]

PCR precedes diagnostic histologic changes, is highly sensitive for diagno- sis, and is useful in monitoring treatment response [15, 16]

Several reports and large series describe persistent predominance in middle-aged men, latency of diagnosis, and risk of relapse with development of CNS manifestations [3, 4, 22–26]

Cases required a clinical history compatible with TW infection and positive PCR result from at least one site or clear histologic evidence of TW by positive periodic acid-Schiff (PAS) stain- ing macrophages in clinically involved tissue [25–27]

Three classifications for TW infection were used in this study: definite classic WD, probable classic WD, and local- ized WD

While previously considered pathognomonic, pos- itive PAS staining macrophages from small bowel biopsies are not entirely specific and may have a high false positive rate, as they can be present in other infections, such as Myco- bacterium Avium [28]. Definite classic WD was therefore defined by both a positive PAS and positive PCR from small bowel biopsies [26].

Probable classic WD was defined by positive PCR with either no or a negative PAS result, or positive PAS and no or a negative PCR result from small bowel biopsies. Cases of positive PCR in the setting of a negative PAS have been attributed to the patchy localization of bacteria in the gut, leading to a sampling error that has a larger negative impact on PAS due to smaller sample size and poorer sensitivity [29].

Stool and saliva samples were not collected or processed for PCR

We present a total of 33 cases of TW infection: 29 (88%) confirmed by PCR, and four cases diagnosed by clinical symptoms and PAS positivity without PCR testing.

No cases diagnosed by PAS had a negative PCR from the same site

Eighteen cases were classified as classic WD (55%) and 15 cases as extraintestinal localized TW infection [CNS (n = 7), articular (n = 4), vitreitis (n = 1), pericarditis (n = 1), bacte- remia (n = 1), and myositis (n = 1)]

The majority of TW infection (both classic and localized) occurred in males (82%) (Table 2). Median age at diagnosis was 53 years (range 11–75) and was similar between groups

Interval to diagnosis varied but was not significantly longer in localized disease.

Therapy was directed at inflam- matory arthropathies in the majority of cases (8 of 14, 57%) and also at pericarditis, Lyme disease, Crohn’s disease, and chronic inflammatory pachymeningitis. The indication for prior use of corticosteroids in the 2 remaining cases was not clear.

The 14 patients on immunosuppressives for a prior diag- nosis either had a lack of response or worsening of symp- toms with these medications. The majority of patients on prednisone experienced improvement on the steroid, specifi- cally with joint pains and ocular dryness, and felt that their symptoms flared during a prednisone taper.

Weight loss was the most common clinical feature (22 of 33, 67%) and was present in 17 of the 18 classic cases (94%). Quantitative data were available in 14 with a median of 11 kg (interquartile range (IQR) 7–16.5). Weight status was documented in 8 cases of localized disease, and weight loss was present in 5 of 8 (63%) with a median of 15 kg (IQR 11–17). Of the 22 patients with documented weight loss, 13 (59%) had diarrhea: 10 classic WD and 3 localized TW infection

Fever, diarrhea, asthenia, and adenopathy were found in roughly half of patients with classic WD, and arthritis/ arthralgias occurred in nearly two-thirds (64%) of localized disease. Arthritis/arthralgias were present in 9 of 18 (50%) cases of classic WD

Abdominal pain was noted in 10 cases with the majority (7 of 10, 70%) in classic WD

Ascites was a rare finding (2 cases).

Melanoderma—hyper- pigmentation of the skin—was seen in three cases, all classic WD. Melanoderma is viewed as a classic symptom that is now rarely reported due to earlier diagnosis [25]

Neurologic symptoms (cranial nerve deficits, demen- tia, memory impairment, encephalopathy, seizures, ataxia, paresthesia, vertigo, diplopia, proximal muscle weakness, oculomasticatory myorhythmia) were present in 21 of 33 (64%). Eleven (52%) had localized disease, and 10 (48%) had classic WD.

Oculomasticatory myorhythmia was noted in only one case, which was classic WD (Table 3)

Four patients had the classic combination of fever, diar- rhea, abdominal pain, and arthralgias: two males and one female with classic WD and one male with localized disease

An additional 11 cases were noted to have 3 of the 4 classic findings.

The most common laboratory finding was anemia (Table 4). Median anemic hemoglobin value was 10.5 g/dL over- all (range 7.4–12.5) and was 9.4 (range 7.4–12.5) and 11.3 (range 9.1–12.3) for classic and localized disease

Erythrocyte sedimentation rate values were available in 23 cases and were elevated in 11 (48%). C-reactive protein (CRP) values were available in 11 cases and were elevated in 9 (82%). CRP was elevated in all cases of classic disease with CRP data.

The majority of cases were diagnosed by PCR, 29 of 33 (88%); 15 of 18 (83%) and 14 of 15 (93%) of classic and localized cases, respectively (Table 5).

Positive PAS was present in 11 of 15 (73%) of cases of WD with positive PCR (Fig. 1)

Twenty-three of 33 (70%) had multi- ple sites tested at the time of diagnosis, and 11 of 33 (33%) had multiple sites with positive findings for TW infection

Of 15 localized cases, 6 (43%) had negative intestinal evaluation: articular infection (n = 1), bacteremia (n = 1), pericarditis (n = 1), CNS infection (n = 3). The remain- ing 9 cases had no bowel investigation within 4 weeks of diagnosis.

Blood PCR was performed in 17 cases at the time of case defining testing and was negative in 9 of 17 (53%) overall, 5 of 11 (40%) classic WD, 2 of 3 (67%) CNS, 0 of 1 (0%) vitreitis, 1 of 1 (0%) articular infection, and 2 of 3 (100%) pericarditis (2 cases of pericarditis occurred in the setting of classic WD).

Of 30, 24 (80%) showed a clear clinical response to antibiotic therapy, 2 of 30 (7%) no clinical response or further deterioration, and 4 of 30 (13%) equivocal clinical response after median follow-up of 14 months (IQR 2–39).

Two of the four cases with equivocal clinical response were CNS disease treated with intravenous (IV) ceftriaxone and either oral trimetho- prim–sulfamethoxazole (TMP-SMX) or ceftibuten, and the two other cases were classic WD treated with ceftriaxone and TMP-SMX.

The majority, 23 of 32 (72%), received a combination of IV ceftriaxone and oral TMP-SMX.

Lack of response to therapy or further deterioration occurred in two patients, one with classic WD and the other in the patient that deferred therapy. The lack of response in the classic WD case may represent immune reconstitution inflammatory syndrome (IRIS) as opposed to treatment failure, due to the initial response to pred- nisone; however, there was no confirmation of successful treatment through a negative tissue PCR, which is required for diagnosis of IRIS [31]

All cases treated with TMP-SMX monotherapy showed clear clinical response to therapy including classic disease (n = 4), CNS infection (n = 1), articular infection (n = 1), and bacteremia (n = 1). Minocycline was used to treat one case of articular infection with good clinical response

Exposure to immunosuppressing agents was common, but it is not clear whether exposure to these agents represents an independent risk factor or is merely a clinical misstep before diagnosis

Arthritis and arthralgias were more common in extraintestinal disease than in classic WD

Proximal muscle weakness, altera- tions in memory, gait, personality, and level of conscious- ness may be underappreciated findings and do not reliably predict positive diagnostic testing from CSF

Strikingly, neurologic symptoms were present in 64% of total cases in the US cohort as compared to 22 and 24% in the European cohorts. This discrepancy is likely due to a difference in definition of neurologic symptoms. While our study evalu- ated for the same neurologic symptoms as the two European cohorts, we also included unique symptoms such as vertigo, diplopia, and ataxia. When looking just at the symptoms reported in the European cohorts (dementia, personality changes, myoclonus, oculomasticatory myorhythmia, and hypothalamic changes), the frequency of neurologic symp- toms in our study fell to 33.33%

A previous study showed excellent sensitivity and spec- ificity of PCR [15], and the false positive rate of PCR in the gastrointestinal tract has been reported at 6% (95% confidence interval 5–8%) [33], but in our cohort, blood PCR showed a sensitivity of only 47% (Table 5) and is likely unhelpful in excluding the diagnosis. Therefore, there is a need for aggressive clinically focused diagnostic evaluation.

Diagnosis of localized TW infection can prove difficult, and TW immunohistochemistry is a sensitive and specific method that can be used to confirm extraintestinal PAS positive tissue, which is not specific to TW infection [34]. Additionally, recent developments in PCR, such as utilizing primers targeted against repetitive sequences and combining rapid cycling with fluorescence-based detec- tion in a closed tube, have led to increased sensitivity and specificity of PCR. This improved diagnostic accuracy has led to the proposed use of noninvasive PCR testing on saliva and stool samples as an initial screening test [35]

Salivary PCR, however, has been noted to have false posi- tives. These have been attributed to similarities between TW and oral bacteria and the TW gene targeted by the primer [36]. This can be potentially minimized with the use of several or more specific primers, and the require- ment to test both stool and saliva.

Additionally, there are asymptomatic carriers of TW, both in the saliva and stool, which can lead to false positives. This effect is mitigated by the low prevalence of carriers in the general popula- tion, and the sensitivity of high bacterial load in the stool for infection [37]. The positive predictive value of having both a positive stool and saliva PCR has been reported as 95.2% and the negative predictive value of either test being negative as 99.2% [37].

While neurologic symptoms are a classic feature in classic WD and attributed to CNS involvement, which has been identified in 90% of brain and spinal cord speci- mens from infected patients and carriers, isolated neuro- logic infection is less common [29].

Five of the 7 individuals with isolated neurologic infection underwent CSF PCR, and all 5 were positive. Additionally, the single cerebral biopsy performed was PAS positive. This is consistent with a prior study [38].

The use of TMP-SMX was further supported by a 2010 rand- omized controlled trial on 40 individuals that found that two weeks of IV meropenem or IV ceftriaxone as initial therapy followed by oral TMP-SMX was equally effica- cious in achieving clinical resolution and remission for a minimum of almost 6 years [42].

French data revealed that the molecular target for TMP is absent in TW, and in vitro studies revealed mutations in TW causing resistance to sulfa agents [43–46]

These in vitro results were supported by a retrospective review of 29 patients with TW infec- tion which found no treatment failures in 13 individuals treated with doxycycline and hydroxychloroquine, and 100% treatment failures in the 14 individuals treated first line with TMP-SMX [43]

We did not observe clinical relapse or primary treat- ment failures sufficient to recommend a doxycycline and hydroxychloroquine-based regimen over a TMP-SMX-based regimen.

Lifetime susceptibility to relapse despite successful treat- ment has been described and is likely possible given the underlying genetic alterations of the immune system that predispose individuals to infection. CNS relapse can be severe and lead to death. Given this, a course of 1 year of doxycycline and hydroxychloroquine followed by a lifetime of doxycycline has been proposed [48].

Follow-up testing demonstrated PCR response in nearly all cases where available. Prior investigation showed that residual histologic changes can persist for months after clini- cal response to treatment [15].

However, patients were more likely to relapse if PCR positivity persisted after clinical response [15]

The limited follow-up data in our series indicate that PCR becomes negative between 1 and 4 weeks after initiation of therapy.

With recent antibiotic exposure for more than 1 week in duration less than 4 weeks in the past, histopathologic examination may be preferred over PCR

The textbook presentation of classic WD was infrequent in this case series, and the majority of affected individuals may not present with the classic combination of diarrhea, weight loss, abdominal pain, and arthralgias