HSC Section 8_April 2017

grading system, 13 (68.4%) were classified as level 4; 2

(10.5%), level 1a; 2 (10.5%), level 3b; and 2 (10.5%), level

2b. The most common study country of origin was Japan (n

= 6, 31.6%), followed by Sweden (4, 21.1%), and 1 each

(5.3%) of England, Germany, Hungary, Italy, Mexico,

Netherlands, South Korea, Turkey, and the United States.

The diagnostic and reporting criteria varied. The most

common criteria set used was the American Academy of

Otolaryngology—Head and Neck Surgery Committee on

Hearing and Equilibrium guidelines (n = 4), followed by the

1972 American Academy of Ophthalmology and

Otolaryngology criteria (n = 3) and the Japan Society for

Equilibrium Research for MD guidelines (n = 3). The

remainder reported a constellation of symptoms and various

diagnostic tests as inclusion criteria for MD.

Eight medications with diuretic properties were analyzed.

Six (31.6%) studies investigated isosorbide, 5 (26.3%) hydro-

chlorothiazide, 2 (10.5%) acetazolamide, 2 (10.5%) chlorthali-

done, and 1 (5.3%) each of betahistine or hydrochlorothiazide,

chlorthalidone-acetazolamide, hydrochlorothiazide-triamterene,

and nimodipine. Betahistine does not have known diuretic

properties but was included, as it was a treatment arm in a

study investigating hydrochlorothiazide. The dosing, therapy

duration, and follow-up period varied widely (

Table 2

). The

outcome measurements varied from unstructured reports of

symptomology, ancillary testing (electrocochleography, gly-

cerol tests, vestibular testing), and audiology variables to

formal reporting of accepted guidelines, including the 1985

Academy of Otolaryngology—Head and Neck Surgery

Committee on Hearing and Equilibrium guidelines, 1972

American Academy of Ophthalmology and Otolaryngology

standards, and the Japan Society for Equilibrium Research for

MD guidelines (

Table 3

Eight studies (42.1%) reported a degree of objective

hearing improvement. Six (31.6%) reported mixed hearing

results, while the remainder demonstrated no or inconclu-

sive hearing results (

Table 4

). Fifteen (79.0%) studies

reported improvement in vertigo symptoms, with 2 (10.5%)

reporting mixed results and 2 (10.5%) reporting no vertigo

outcomes (

Table 5

). Ten (52.6%) studies reported no side

effects as a result of therapy. Three (15.8%) studies specifi-

cally reported abdominal discomfort, and the remainder

reported different constellations of side effects (

Table 6

Discussion

MD is a common condition with occasionally debilitating

symptoms. As the pathophysiology is not well understood,

devising effective therapeutic strategies has been a challen-

ging task. Acute symptom exacerbations can be treated with

benzodiazepine and steroid regimens. Physicians may offer

more invasive procedures for refractory cases, including medi-

cal labyrinthectomy with transtympanic gentamicin injections,

endolymphatic sac decompression, surgical labyrinthectomy,

and surgical vestibular nerve section. Combinations of low-salt

diet restriction and oral diuretics have largely been used as

maintenance therapy. Despite the use of diuretics for MD, lit-

erature reviews have consistently demonstrated lack of evi-

dence of effectiveness. The Cochrane Group published a

thorough systematic review of diuretic efficacy in MD patients

in 2006 and an update in 2010; however, no RCTs met their

rigorous standards for a systematic review.

We realized an

opportunity to broaden inclusion criteria to all study designs

beyond RCTs to evaluate the potential benefits of diuretic

therapy.

We found 19 studies of Level of Evidence 4 or higher,

per the Oxford Centre for Evidence-Based Medicine, with

considerable heterogeneity in patient population, study

design, diuretic type and dosage, follow-up time, and out-

comes. Level of Evidence 4 includes studies that are either

case series or poor-quality cohort and case-control studies.

This heterogeneity precluded formal meta-analysis. Only 4

RCTs were included in our review.

8-11

One RCT directly

compared hydrochlorothiazide with betahistine with no pla-

cebo arm, and both were found to improve vertigo.

Betahistine does not have known diuretic properties but was

included, as it was a treatment arm in comparison with

hydrochlorothiazide. One RCT with a placebo arm investi-

gated Dyazide (hydrochlorothiazide and triamterene) and

found significant improvements in vestibular symptoms but

not in hearing or tinnitus.

Another RCT with a placebo

arm investigated hydrochlorothiazide and found statistically

significant improvements in both hearing loss and vestibular

symptoms but no significant improvement in tinnitus.

Last,

a low-quality RCT compared isosorbide with 2 control

compounds—methyl acid dihydroergotoxine and vitamin

B3—and found significant improvements in headache and

tinnitus.

The remainder of the studies were retrospective

case series,

12-23

case-control studies,

24,25

and 1 prospective

case series.

As the studies analyzed are of mostly ‘‘low’’ quality with

heterogeneous methods for reporting outcomes, caution

must be exercised in drawing conclusions from the

Figure 1.

PRISMA-P

flow diagram for evaluation of identified

studies.

Otolaryngology–Head and Neck Surgery 154(5)