Features of Resolving and Nonresolving Indeterminate Pulmonary Nodules at Follow-up CT: The NELSON Study

Zhao YR, Marjolein AH, et. al.

Journal club 6/3/2014

Background

66% of participants in CT screening trials have at least 1 pulmonary nodule
Most indeterminate nodules benign

granulomatous or infectious lesions
benign lymph nodes

Purpose: to retrospectively identify features allowing prediction of disappearance of solid, indeterminate intraparenchymal nodules

The NELSON Study

Nodule management based on volume and volume doubling time

Population

7557 participants
46-76 years old (mean age 59)
high risk for lung cancer

current or former smokers
> 15 pack years

CT protocol

1.0 mm slice thickness, low dose
baseline, 1 year, 3 years
If indeterminate nodule found,
followup scan at 3 months

No growth or < 25% growth:

regular screening

> 25% growth or volume doubling time < 400 days:

referred for further workup

Image Reading

Read twice independently

1st by one of 13 radiologists 1-20 yrs experience
2nd by one of 2 radiologists at least 6 yrs experience

If discrepancy

3rd radiologist > 15 yrs experience arbitrated

Discrepancies between 1st and 2nd reading found in 43 lesions

Volume calculation

Syngo Lungcare software package
3D template generated after clicking nodule with automated volume measurements

In 86% of >4000 nodules, double reading gave same volume
Differences >15% found in 4% of nodules

If volumes differed, 2nd reader measurements used

Nodule characteristics

Indeterminate size:

50 - 500 mm^3 (4.6 - 9.8 mm)

Considered benign if no growth after 2 years
or benign histology
Considered resolved if disappeared on followup CT
Peripheral defined as outer 1/3 of lung
Margins defined as smooth vs.
non-smooth (lobulated, irregular, spiculated)

Results: Figure 1

Results: Table 1

Results: Table 2

Results: Table 3

Nodule resolution by size

Results: Table 4

Moment of resolution by nodule size

Results: Figure 2 (smooth, round)

Results: Figure 2 (lobulated)

Discussion

In NELSON, 10.7% of participants had solid, intraparenchymal nodule with volume 50-500 mm^3

10.1% disappeared at followup

Though most do not resolve, 75% of those that do resolve can be identified at 3-month repeat CT

Discussion

Nodules with non-smooth edges disappeared more frequently than those with smooth edges

However, non-smooth edges also more frequently found in malignant nodules

No distinction can be made between resolving and nonresolving malignant nodules based on nodule edge

Discussion

The rate and speed of disappearance higher for larger nodules (>= 8 mm)

However, rate of malignancy also higher

May be explained by benign conditions like inflammation commonly showing nodules with a larger diameter

Limitations

Precise time of nodule resolution could not be ascertained

No histology for resolving nodules

Generalizability to incidental nodules found on non-screening exams

Conclusions

About 10% of solid intraparenchymal nodules of intermediate size resolve during followup

Short term followup CT valuable for detecting fast-growing nodules and for identifying 3/4 of resolving nodules

Resolving pulmonary nodules share CT features with malignant nodules.