TABLE 3.

Results of Metaregression Analysis of PET/MRI for Diagnosis of Locoregional HNSCC

Diagnostic indicators
CovariateCategory estimateNumber of studiesSensitivityPSpecificityPI2 (%)
Study subjectsOne HNSCC subtype2100%
(95% CI, 99%–100%)		0.0098%
(95% CI, 95%–100%)		0.6881
Various HNSCC* subtype1088%
(95% CI, 81%–95%)		95%
(95% CI, 92%–98%)
Study designProspective594%
(95% CI, 87%–100%)		0.6095%
(95% CI, 91%–100%)		0.260
Retrospective794%
(95% CI, 83%–100%)		96%
(95% CI, 93%–100%)
MRI fusionSimultaneous992%
(95% CI, 85%–99%)		0.2796%
(95% CI, 93%–99%)		0.530
Retrospective396%
(95% CI, 89%–100%)		94%
(95% CI, 85%–100%)
MRI acquisitionWith DWI382%
(95% CI, 61%–100%)		0.0893%
(95% CI, 85%–100%)		0.0333
Without DWI995%
(95% CI, 91%–100%)		97%
(95% CI, 94%–99%)
Reference standardPathologic590%
(95% CI, 78%–100%)		0.5589%
(95% CI, 80%–98%)		0.9267
Clinical and pathologic795%
(95% CI, 89%–100%)		98%
(95% CI, 96%–99%)
QUADAS-2 riskPresence597%
(95% CI, 92%–100%)		0.0696%
(95% CI, 91%–100%)		0.2616
Absence789%
(95% CI, 79–99)		96%
(95% CI, 93%–99%)
Publication year4.870.690.9659

  • ↵* Various HNSCC refers to studies evaluating PET/MRI in patients with single primary HNSCC, regardless of its anatomic origin (more than one anatomic origin). Studies evaluating PET/MRI in patients with single primary HNSCC of similar anatomic origin, such as nasopharyngeal HNSCC, are referred to as one HNSCC subtype.