What is the recommended course of action for a thyroid biopsy with atypical Bethesda III results?

Management of Bethesda III (AUS/FLUS) Thyroid Nodules

For Bethesda III thyroid nodules, repeat FNA is the recommended initial approach, but the threshold for proceeding to diagnostic lobectomy should be lower than traditionally taught, as malignancy rates range from 27-38% in surgical series—substantially higher than the originally estimated 5-15% risk. 1, 2, 3

Initial Diagnostic Approach

Repeat Fine Needle Aspiration

• Repeat ultrasound-guided FNA is the standard first step after an initial Bethesda III diagnosis, as this can yield a more definitive cytologic diagnosis in 42-62% of cases 4, 5
• When repeat FNA is performed, approximately 43% will be reclassified as benign, 38% remain AUS/FLUS, and 14% are upgraded to suspicious or malignant categories 3
• The conclusive result rate from repeat FNA increases with higher ultrasound risk stratification: 58% for low suspicion nodules versus 89% for high suspicion nodules 6

Critical Caveat on Repeat FNA

• Even with two consecutive Bethesda III diagnoses, malignancy risk remains 26%, suggesting repeat FNA has limited ability to rule out cancer in this category 3
• Age under 30 years is the strongest predictor of malignancy and should lower your threshold for surgery over repeat FNA 1

Risk Stratification Using Ultrasound Features

High-Risk Ultrasound Features Warranting Surgery

The following sonographic features significantly increase malignancy risk and should prompt surgical consultation rather than repeat FNA 2, 6:

• Marked hypoechogenicity (58% malignancy rate) 2
• Taller-than-wide shape (50% malignancy rate) 2
• Punctate echogenic foci/microcalcifications (49% malignancy rate) 2
• Lobulated or irregular margins (46% malignancy rate) 2
• ACR TI-RADS 5 classification (55% malignancy rate) 2

Integration with TI-RADS/K-TIRADS

• K-TIRADS high suspicion nodules with Bethesda III cytology have nearly 100% malignancy risk and should proceed directly to surgery 6
• K-TIRADS intermediate suspicion nodules show 88% malignancy rate 6
• K-TIRADS low suspicion nodules have 60% malignancy rate, but this still exceeds traditional estimates 6

Molecular Testing as Alternative to Repeat FNA

Role of Molecular Markers

• Molecular testing with Afirma GSC or Thyroseq v3 can effectively rule out malignancy with a false negative rate of only 0.6% over 3 years of surveillance 7
• Nodules testing benign/negative on molecular testing can be safely observed with ultrasound surveillance every 6-12 months 7
• The single false negative case was a 2.7 cm minimally invasive Hürthle cell carcinoma that demonstrated sonographic growth during surveillance 7

When Molecular Testing is Most Useful

• Consider molecular testing for patients who strongly prefer to avoid surgery and have nodules without high-risk ultrasound features 7
• Molecular testing is particularly valuable when repeat FNA remains indeterminate (second Bethesda III result) 7

Surgical Decision Algorithm

Proceed Directly to Diagnostic Lobectomy When:

1. Patient age <30 years (strongest clinical predictor of malignancy) 1
2. Nodule size ≥4 cm (size alone warrants aggressive evaluation) 8
3. High-risk ultrasound features present (ACR TI-RADS 5, marked hypoechogenicity, microcalcifications, taller-than-wide) 2, 6
4. TSH >4.5 mIU/L (associated with 35% malignancy rate) 2
5. Clinical risk factors: history of head/neck radiation, family history of thyroid cancer, rapidly growing nodule, or suspicious cervical lymphadenopathy 4, 9
6. Two consecutive Bethesda III results (26% malignancy risk persists) 3

Consider Repeat FNA or Molecular Testing When:

• Patient age >30 years with low-intermediate risk ultrasound features 1
• Nodule <2 cm without high-risk sonographic features 2
• Patient strongly prefers to avoid surgery and accepts surveillance requirements 7

Important Practice Considerations

Malignancy Rates Higher Than Expected

• The actual malignancy rate in Bethesda III nodules is 27-38% in surgical series, not the 5-15% originally proposed by the Bethesda classification 1, 2, 3
• This discrepancy is particularly pronounced in tertiary cancer centers where malignancy rates reach 38% 3
• Guidelines recommending routine observation or repeat FNA may underestimate cancer risk in many practice settings 3

Histologic Subtypes to Anticipate

• Papillary thyroid carcinoma (follicular variant and classical) accounts for 66% of malignancies in Bethesda III nodules 2
• K-TIRADS effectively predicts papillary carcinoma but poorly predicts follicular carcinoma, as 80% of follicular cancers show low suspicion ultrasound features 6
• This limitation means that even "low-risk" appearing nodules cannot be dismissed without tissue diagnosis 6

Surveillance Protocol if Non-Operative Management Chosen

• Ultrasound surveillance every 6-12 months for nodules with benign molecular testing 7
• Any sonographic growth during surveillance should prompt surgical resection, as this was the only scenario where molecular testing missed malignancy 7
• Median safe surveillance period is 34 months based on current data 7

Common Pitfall to Avoid

• Do not rely on repeat FNA alone to exclude malignancy—even after multiple benign repeat FNAs, a substantial cancer risk persists in nodules with high-risk clinical or sonographic features 3
• The presence of multiple high-risk ultrasound features should override reassuring cytology and prompt surgical consultation 2, 6