Questionable Utility of Galactomannan Testing for Diagnosis of Exserohilum rostratum Infection

LETTER

With interest, we read the paper by Korem et al. describing a case of Exserohilum rostratum sinusitis from which serum galactomannan (GM) levels were elevated upon diagnosis (1). They further demonstrated that the extract of the E. rostratum clinical isolate yielded positive GM results comparable to those generated by Aspergillus fumigatus. The authors concluded that the GM testing, approved by the U.S. Food and Drug Administration, could be a useful diagnostic tool to aid in the early detection of E. rostratum infection.

E. rostratum is the major causative agent for the recent nationwide meningitis outbreak due to the contaminated methylprednisolone acetate injections (2, 3). We previously reported the mycological procedures and histopathological features for the identification of E. rostratum in an index outbreak patient at the Johns Hopkins Hospital (4). To examine the utility of GM testing as an aid in the detection of the E. rostratum meningitis outbreak, we first performed in vitro testing of the E. rostratum clinical strain isolated from the index patient for its GM content using the methods described by Korem et al. (1), with slight modifications. Briefly, a fresh inoculum of 0.4 × 10⁴ to 5 × 10⁴ spores per ml was grown for 72 h at 25°C (room temperature) and at 37°C in Sabouraud dextrose broth (Becton, Dickinson, Sparks, MD, USA) on a rotary shaker. The culture suspensions were centrifuged at 2,000 × g, and then the supernatants were filtered (0.2-μm pore size). The filtrates, undiluted and diluted to 10⁻³ with saline, were tested for GM (Platelia Aspergillus antigen [Ag] kit; Bio-Rad Laboratories, Inc., CA, USA). With a cutoff value for the GM index (GMI) of 0.5 and compared to results for the Aspergillus fumigatus positive control, the E. rostratum clinical isolate yielded low positive GM results with the undiluted extract but produced negative GM results with the 10⁻³ dilution extract (Table 1).

TABLE 1.

Galactomannan levels in fungal extracts and CSF

Sample	GM index value for sample with:
	No dilution	10−3 dilution
Fungal extracts
Aspergillus fumigatus (clinical isolate)
37°C	19.45	3.21
25°C	19.45	2.16
Exserohilum rostratum (index case)
37°C	1.02	0.24
25°C	0.82	0.13
Sabouraud dextrose broth	0.24
CSF
Case 1	0.10
Case 2	0.10
Case 3	0.11

We subsequently tested GM in cerebrospinal fluid (CSF) from three outbreak patients (Table 2). According to the Centers for Disease Control and Prevention (CDC) case definitions for this outbreak, the first case was defined as a proven case based on a history of exposure and a positive Exserohilum PCR assay (2). Cases 2 and 3 were defined as probable cases based on a history of exposure and the subsequent development of meningitis. All cases tested negative by the GM assay (Table 1).

TABLE 2.

Demographic and CSF characteristics of cases^a

Case (reference)	Age (yr)	Symptoms	Day of symptom onsetb	Day of voriconazole initiationb	Day CSF collectedb	No. of CSF WBC (cells/mm3)	% PMN, % lymph, % mono in CSF	CSF protein (mg/dl)	CSF glucose (mg/dl)	CSF fungal culture result	CSF fungal PCR result (CDC)	CSF (1,3)-β-d-glucan (pg/ml)
Index (4)	51	HA, diplopia, ataxia	∼7	NA	10	850	84, 15, 1	153	36	Posc	NA	NA
1	97	Lethargy, ataxia	∼80	90	90	47	95, 4, 1	116	36	Neg	Pos	797
2 (6)	55	HA, ISP, blurred vision	∼7	45	57	1,224	69, 13, 18	164	55	Neg	Neg	2,396
3	41	HA, photophobia	∼14	38	38	353	38, 48, 14	85	43	Neg	Neg	1,524

^aAll cases were female. HA, headache; ISP, injection site pain; WBC, white blood cells; PMN, polymorphonuclear cells (neutrophils); lymph, lymphocytes; mono, monocytes; NA, not available; Pos, positive; Neg, negative.

^bNumber of days after epidural steroid injection date.

^cCulture grew Exserohilum rostratum at day 10.

Laboratory diagnosis of the recent multistate outbreak of fungal meningitis predominantly caused by E. rostratum has been challenging. Neither culture nor a PCR assay demonstrated a satisfactory sensitivity in detecting this E. rostratum infection (5). Korem et al. proposed a potential use of GM testing as an aid in the diagnosis of E. rostratum infection based on their findings that showed that the organism produced GM. However, it has been our experience that the E. rostratum clinical strain isolated from our index patient produced little GM content. As a result, the GM testing was not sensitive enough, as evidenced by the negative results from testing GM in CSF from three outbreak patients. Instead, we detected high levels of (1,3)-β-d-glucan (BDG) in the CSF of the three patients (Table 2). CSF BDG testing is more useful for the diagnosis of E. rostratum infection (5, 6).