Page 18 - Delaware Medical Journal - December 2016
P. 18
TABLE 2: Clinical Validity of ThyroSeq (2,3) and Afirma (18)
TEST
AUS/FLUS
FN
ThyroSeq
PPV-77
PPV-83
NPV-97
(Based on 23% cancer prevalence)
Retrospective/Prospective Single Center Pathologists Not- Blinded to Molecular Results
PPV-96
(Based on 27% cancer prevalence)
Retrospective/Prospective Single Center Pathologists Not- Blinded to Molecular Results
Afirma
PPV-38
PPV-37
NPV-95
(Based on 24%
cancer prevalence)
Prospective
Multi-center Pathologist Blinded to Results
NPV-94
(Based on 25% cancer prevalence)
Prospective
Multi-center Pathologist Blinded to Results
pathology.
pathology of benign adenomatoid nodule. One patient with a Met gene mutation
Calculated prevalence of cancer for
percent). The author previously published his individual cancer prevalence for
DISCUSSION
Patients displayed a strong preference for molecular testing instead of diagnostic indeterminate thyroid nodules. All patients in this series, 25/25 (100 percent) with indeterminate thyroid nodules elected
or diagnostic thyroid surgery. Sixteen
on ThyroSeq testing elected active surveillance rather than surgery or repeat
for a mutation are observed clinically and
with ultrasound every six to 12 months. Their ultimate rate of malignancy may be determined over time.
Calculated prevalence of cancer for
published his cancer prevalence for Lower cancer prevalence in an individual
a patient without mutation for the brief duration of clinical and ultrasound follow- up. Surgical truth was not established for the patient with no mutation. Surgical truth is also generally not established
5 percent false negative rate.7 The 19 percent prevalence of malignancy for indeterminate thyroid nodules in this study among patients with no mutation and un-operated nodules is unlikely, even with limited follow-up. These data for the study period are consistent with the reported 2,3 More than half of the tested patients with indeterminate thyroid nodules avoided diagnostic thyroid surgery and have been followed carefully with clinical exams and ultrasound. Still, some data suggest that portion of thyroid cancers.9 Thus, further validation will require longer follow up and larger clinical trials. Cost analysis studies are also needed as the cost of the ThyroSeq test is $3,200.
has been reported to have a 90 percent
370
Del Med J | December 2016 | Vol. 88 | No. 12
