CMV in Perilymphatic Fluid (2005) PDF

The Laryngoscope Lippincott Williams & Wilkins, Inc. © 2005 The American Laryngological, Rhinological and Otological Society, Inc. Cytomegalovirus in the Perilymphatic Fluid Paul W. Bauer, MD; Mojgan Parizi-Robinson, PhD; Peter S. Roland, MD; Subramanian Yegappan, MD Objective: The incidence...

The Laryngoscope Lippincott Williams & Wilkins, Inc. © 2005 The American Laryngological, Rhinological and Otological Society, Inc. Cytomegalovirus in the Perilymphatic Fluid Paul W. Bauer, MD; Mojgan Parizi-Robinson, PhD; Peter S. Roland, MD; Subramanian Yegappan, MD Objective: The incidence of congenital cytomega- hearing loss (SNHL) are genetic in origin.1 Unidentified lovirus (CMV) infection is approximately 1% of neo- genetic mutations or unrecognized viral pathogens are nates. Ninety percent of congenitally infected infants thought to cause the remaining 40% to 50% of cases. are “asymptomatic;” they have no signs or symptoms The incidence of congenital cytomegalovirus (CMV) at birth. The prevalence of congenital CMV in the infection is approximately 1% (1/100) of neonates.2 Most profoundly deaf population and the pathogenesis of deafness from CMV are unknown. The objective of congenitally infected infants (90%) are “asymptomatic;” this study is to determine whether CMV can be dem- they have no signs or symptoms at birth. The prevalence onstrated and quantified in perilymphatic fluid of pa- of SNHL in asymptomatic, congenital CMV ranges from tients with congenital CMV infection and sensorineu- 7% to 15%, of which 23% will have bilateral profound ral hearing loss (SNHL) using a quantitative real-time SNHL.3 In general, recent literature reports that CMV polymerase chain reaction (QRTPCR). Study Design: causes 20% to 30% of all cases of congenital SNHL.4 Prospective case series. Methods: Perilymphatic fluid Little is known about the pathogenesis of CMV- was collected at the time of cochlear implantation mediated HL. Animal models of CMV infection suggest from children with known or radiologic evidence of that either a local inflammatory response or direct cyto- congenital CMV infection and analyzed for the pres- ence of CMV using QRTPCR. Blood was collected and pathic effects of the virus in the inner ear may play a role.5 analyzed for CMV using QRTPCR, serology, and cul- In human temporal bone sections, CMV has only been ture. CMV was quantified in perilymphatic fluid and demonstrated in endothelial structures of the endolym- compared with that present in the patient’s blood. phatic space and has been theorized to reach the inner ear Results: Perilymphatic fluid and blood was collected by hematogenous spread through the stria vascularis.6 from six children. QRTPCR was positive for CMV in Cochlear implantation has created a new means by the perilymphatic fluid of four patients. Blood ana- which to investigate potential etiologies of congenital lyzed with QRTPCR, and culture was negative in all SNHL. Perilymphatic fluid is exposed and removed as a patients. Conclusions: CMV can be demonstrated and quantified in perilymphatic fluid using QRTPCR. Re- standard part of the procedure. CMV has been cultured finements in our technique and sampling of perilym- from perilymph extracted from cadaveric temporal phatic fluid from a large population of children with bones.7,8 Viral culture is a highly specific technique for congenital SNHL and unknown etiology can deter- detecting CMV, with a positive predictive value of 100%; mine the prevalence of CMV-mediated profound HL. however, the volume of specimen available affects the Key Words: CMV, cytomegalovirus, hearing loss, per- sensitivity of the assay.9 The volume of perilymph in the ilymphatic, sensorineural, SNHL. human cochlea is approximately 50 ␮L.10 The volume of Laryngoscope, 115:223–225, 2005 perilymph that can be extracted from the cochlea in vivo is unknown. Quantitative real-time polymerase chain reac- INTRODUCTION tion (QRTPCR) to detect a segment of the CMV genome Genetic research indicates that as many as 50% to could provide a highly specific and sensitive means of 60% of cases of congenital nonsyndromic sensorineural detecting the presence of virus in perilymphatic fluid. Our objective is to determine whether CMV can be demon- Presented at the Annual Meeting of the Triological Society, Phoenix, strated and quantified in perilymphatic fluid of patients AZ, May 3, 2004. with known congenital CMV infection using QRTPCR. From the Division of Pediatric Otolaryngology (P.W.B.), Department of Otolaryngology—Head and Neck Surgery; the Department of Pathology (M.P.-R.), University of Texas Southwestern Medical Center; the Depart- ment of Otolaryngology—Head and Neck Surgery (P.S.R.); and Molecular MATERIALS AND METHODS Pathology, Department of Pathology (S.Y.), University of Texas Southwest- ern Medical Center, Dallas, Texas, U.S.A. During the precochlear implant evaluation process, six chil- Editor’s Note: This Manuscript was accepted for publication July 15, dren were identified: four with known congenital “asymptomatic” 2004. CMV and two with radiologic evidence on their brain magnetic Send Correspondence to Dr. Paul W. Bauer, Department of Otolar- resonance imaging (MRI) of findings consistent with congenital yngology–Head and Neck Surgery, University of Texas Southwestern Med- viral infection. One patient with known congenital CMV has ical Center, 5323 Harry Hines Boulevard, Dallas, TX 75390 –9035, U.S.A. undergone sequential bilateral cochlear implantation. The Uni- E-mail: [email protected] versity of Texas Southwestern Medical Center Institutional Re- DOI: 10.1097/01.mlg.0000154722.55044.fc view Board approved the project (#1002–566). Laryngoscope 115: February 2005 Bauer et al.: CMV Perilymphatic Fluid 223 Sample Acquisition ital viral infection. These children were originally seen All children involved in this study received standard of care with no known etiology to their HL. pediatric cochlear implantation including a mastoidectomy with The findings from perilymphatic fluid and blood are facial recess approach into the middle ear. The cochleostomy was summarized in Table I. The concentration of viral targets created initially, preserving the endosteum. The scala tympani found in the perilymphatic fluid using QRTPCR varied. was entered with a 27 gauge needle on a 1 mL syringe, and the The concentration of CMV in blood using QRTPCR and perilymphatic fluid was aspirated and placed in a sterile con- tainer. A 5 mL blood sample was also collected. the blood CMV culture was negative in all cases; there- fore, blood was not a contaminant to account for our pos- itive perilymph findings. Sample Processing Patient 2 only had radiologic evidence suggesting Perilymphatic fluid and blood QRTPCR. The perilym- phatic fluid specimen was initially centrifuged (6,000 rpm) to CMV preoperatively; it was not diagnosed at birth. How- remove surgical contaminants (blood) in the precipitate. To detect ever, patient 2’s concentration of CMV was higher than the presence of CMV and compare and quantify the amount of would be expected in a positive blood sample (⬎50,000). virus with that present in the blood, DNA was extracted from the Patient 5 also had radiologic evidence suggesting CMV perilymphatic fluid. The DNA sequence specific for CMV was preoperatively; however, the results of their IgG and IgM then detected and quantified using the icycler iQ multi-color data indicates that they have never had a CMV infection RTPCR detection system. This technique uses a fluorescent probe of any type. (TaqMan probe) labeled with a reporter and a quencher that binds to the target DNA. During amplification, the Amplitaq Gold polymerase cleaves the reporter from the quencher and thus DISCUSSION fluorescence is detected. The cycle threshold at which fluores- cence is detected above threshold is proportional to the amount of We have demonstrated and quantified CMV in the PCR product made. The amount of PCR product amplified can perilymphatic fluid of four patients. Two patients with a then be compared with that present in the patient’s blood. The known history of congenital CMV had negative perilym- same QRTPCR was conducted on a sample of the patient’s blood phatic fluid. In patient 1, perilymphatic fluid from the serum. first implanted ear contained the CMV virus; however, the virus was not detected in the second ear implanted 1 year CMV Blood Culture and Serology later. It is unknown whether this reflects an error in our Children’s Medical Center of Dallas microbiology and serol- technique or is evidence that the virus is no longer ogy laboratory was used to complete standard CMV blood culture present. The duration of time that CMV can persist in the and immunoglobulin (Ig) serology. perilymphatic fluid after congenital infection has not been studied. In a report on a 14-year-old patient with congen- RESULTS ital cytomegalic inclusion disease and sensorineural HL, Seven perilymphatic fluid and blood samples were CMV could not be cultured from the perilymphatic fluid at collected from six children. The volume of perilymphatic the time of autopsy.11 fluid collected was less than 50 ␮L in all case. Patients 1, The small volume of fluid that can be extracted from 3, 4, and 6 had known “asymptomatic” congenital CMV. the cochlea at the time of implantation does not provide a These four children were diagnosed at birth with positive volume sufficient for culture; in addition, it may affect the urine or blood cultures and positive serum PCR. At our DNA yield during DNA extraction. As the volume of the institution, children who fail their birth hearing screen sample decreases, the likelihood that sampling error can are immediately tested for CMV. Patients 2 and 5 had occur increases. In addition, cellular debris such as blood radiologic findings on their MRI consistent with a congen- cells in the specimen can alter the results of PCR. Even TABLE I. Patient Data. Perilymph Age QRTPCR Blood CMV CMV Blood Patient (mo) (viral target/mL) QRTPCR IgG* IgM† Culture 1a 12 368 Neg. ⬎12.00 0.32 Neg. 1b 24 Neg. Neg. — — — 2 19 184,000 Neg. ⬎15.8 0.28 Neg. 3 12 Neg. Neg. 12.6 0.22 Neg. 4 54 945 Neg. 1.96 0.39 Neg. 5 27 Neg. Neg. 0 0 Neg. 6 48 58,000 Neg. ⬎12.00 0.26 Neg. *Interpretive guide for IgG: ⬍0.9 ⫽ negative (no significant level of detectable CMV antibody); ⬎1.10 ⫽ positive (antibody to CMV detected which may indicate a current or previous infection). †Interpretive guide for IgM data: ⬍0.9 ⫽ negative (no significant level of detectable CMV antibody); ⬎1.10 ⫽ positive (antibody to CMV detected which may indicate a current or recent infection). QRTPCR ⫽ quantitative real-time polymerase chain reaction; CMV ⫽ cytomegalovirus; Ig ⫽ immunoglobulin. Laryngoscope 115: February 2005 Bauer et al.: CMV Perilymphatic Fluid 224 though the sample is initially centrifuged to separate the BIBLIOGRAPHY clear perilymph from surgical contaminants, the volume 1. Marlin S, Garabedian EN, Roger G, et al. Connexin 26 gene of fluid that must then be extracted without disturbing the mutations in congenitally deaf children: pitfalls for genetic precipitant is very small. The same QRTPCR was con- counseling. Arch Otolaryngol Head Neck Surg 2001;127: ducted on a sample of the patient’s blood, and the blood 927–933. 2. Demmler GJ. Infectious Diseases Society of America and was cultured for CMV to assure that surgical contami- Centers for Disease Control. Summary of a workshop on nants were not creating false-positive results. surveillance for congenital cytomegalovirus disease. Rev Histologically, CMV only infects the endolymphatic Infect Dis 1991;13:315–329. compartment.7,11,12 The presence of CMV in the perilym- 3. Fowler KB, McCollister FP, Dahle AJ, et al. Progressive and phatic fluid is supported by previous reports documenting fluctuating sensorineural hearing loss in children with asymptomatic congenital cytomegalovirus infection. J Pe- that CMV can be cultured from the perilymphatic fluid diatr 1997;130:624 – 630. postmortem.7,8 It is possible that our samples were con- 4. Barbi M, Binda S, Caroppo S, et al. A wider role for congenital taminated with endolymphatic fluid. All of our patients cytomegalovirus infection in sensorineural hearing loss. have profound SNHL as documented on auditory brain- Pediatr Infect Dis J 2003;22:39 – 42. stem response testing and standard audiometric testing. 5. Keithley EM, Woolf NK, Harris JP. Development of morpho- logical and physiological changes in the cochlea induced by We assume that their HL is secondary to congenital CMV. cytomegalovirus. Laryngoscope 1989;99:409 – 414. The concentration of viral targets in our samples 6. Strauss M. A clinical pathologic study of hearing loss in varied greatly. In the immune-suppressed population congenital cytomegalovirus infection. Laryngoscope 1985; quantification of CMV DNA copy numbers in plasma sig- 95:951–962. nificantly distinguishes patients at risk of developing 7. Davis LE, Johnsson LG, Kornfeld M. Cytomegalovirus laby- rinthitis in an infant: morphological, virological, and im- CMV retinitis.13 It has also been shown in vitro that the munofluorescent studies. J Neuropathol Exp Neurol 1981; CMV/endothelial cell interaction is both qualitatively and 40:9 –19. quantitatively dependent on the virus strain.14 Patient 2 8. Davis LE, James CG, Fiber F, McLaren LC. Cytomegalovirus and 6 have viral target concentrations in their perilymph isolation from a human inner ear. Ann Otol Rhinol Laryn- that are higher than the concentration of CMV typically gol 1979;88(3 Pt 1):424 – 426. found in positive blood serum. So why then do patients 1 9. Lazzarotto T, Guerra B, Spezzacatena P, et al. Prenatal di- agnosis of congenital cytomegalovirus infection. J Clin Mi- and 4 have comparatively lower concentrations? A limita- crobiol 1998;36:3540 –3544. tion of QRTPCR is the inability to identify the specific 10. Thorne MB, Salt A. Cochlear fluid space dimensions for six genetic strain of virus. The CMV strain can effect the species derived from reconstructions of three-dimensional phenotypic expression. Recent work has shown that intra- magnetic resonance images. Laryngoscope 1999;109: genic variability may be a complicating factor in molecular 1661–1668. 11. Rarey KE, Davis LE. Temporal bone histopathology 14 years epidemiologic studies.15 Our variation in viral concentra- after cytomegalic inclusion disease: a case study. Laryngo- tion can be a reflection of different genetic strains or may scope 1993;103:904 –909. relate to the small sample volume. 12. Davis GL, Spector GJ, Strauss M, Middlekamp JN. Cytomeg- alovirus endolabyrinthitis. Arch Pathol Lab Med 1977;101: CONCLUSION 118 –121. 13. Rasmussen L, Zipeto D, Wolitz RA, et al. Risk for retinitis in CMV can be demonstrated and quantified in perilym- patients with AIDS can be assessed by quantitation of phatic fluid using QRTPCR. Much of our work is currently threshold levels of cytomegalovirus DNA burden in blood. speculative. Genetic advances have shed new light on the J Infect Dis 1997;176:1146 –1155. etiology of congenital SNHL. The incidence of CMV- 14. Waldman WJ, Sneddon JM, Stephens RE, Roberts WH. En- mediated SNHL, however, remains unknown. Refine- hanced endothelial cytopathogenicity induced by a cyto- ments in our technique and sampling of perilymphatic megalovirus strain propagated in endothelial cells. J Med Virol 1989;28:223–230. fluid from a large population of children with congenital 15. Rasmussen L, Geissler A, Winters M. Inter- and intragenic SNHL and unknown etiology can determine the preva- variations complicate the molecular epidemiology of hu- lence of CMV-mediated profound HL. man cytomegalovirus. J Infect Dis 2003;187:809 – 819. Laryngoscope 115: February 2005 Bauer et al.: CMV Perilymphatic Fluid 225