Management of incidental pulmonary nodules

The Fleischner Society guidelines for the management of incidental pulmonary nodules found on CT scans was just updated, involving international input from radiologists, pulmonologists, surgeons, pathologists (see pulm nodule incidental guidelines radiol2017 in dropbox, or doi:10.1148/radiol.2017161659​).

Details:

--the guidelines refer to incidental pulmonary nodules found by CT scan in those >35yo, not for patients at high risk (eg in those with cancer who might havemets, or those getting CTs for screening purposes)

--the minimum threshold size leading to  recommendation for nodule follow-up is if the estimated cancer risk is >1% (arbitrarily chosen)

--follow-up CTs should use low-radiation techniques, no more than 3 mGy in a standard sized person, in order to reduce radiation exposure, esp in patients likely to receive many of them. 

--Single solid nodules

   -- <6 mm (a larger size than prior guidelines): no further follow-up studies (grade 1C, strong recommendation, low- or very-low quality of evidence). Those at higher risk (see below), such as suspicious morphology or upper lobe location has optional recommendation to repeat CT but not before 12 months (small risk by waiting this long and earlier study might provide false reassurance)

   -- 6-8mm and low clinical risk (see below): follow-up 6-12 months, depending on size, morphology and patient preference (grade 1C, strong recommendation, low- or very-low quality of evidence​). usually one follow-up exam is sufficient, though optional one at 18-24 months. For those at higher risk, this additional 18-24 month follow-up is recommended.  (strong recommendation, moderate quality of evidence​)​. average risk of cancer 0.5-2%

   -- >8mm: same recommendation independent of risk: CT at 3 months, PET/CT, or tissue sampling (grade 1A, strong recommendation, high quality of evidence). average risk of cancer 3%

--Multiple solid nodules

   --dominant nodule <6mm: same as with single (Grade 2B, weak recommendation, moderate quality of evidence​)​.

   --6-8 mm: same as with single, but initial follow-up CT at 3-6 months (optional follow-up at 18-24 months if low risk, recommended if high risk). Grade 1B, strong recommendation, moderate quality of evidence​​.

   -->8 mm: same as 6-8mm

--Single subsolid nodules

   --ground glass nodule: <6mm, no routine follow-up; >6mm, CT at 6-12 months to confirm persistence, then every 2 years for 5 years​. (grade 1B; strong recommendation, moderate-quality evidence). Approximately 10% grow and 1% progress to adenocarcinoma, in a study done in Asian population.

   --part solid nodule: <6mm, no routine follow-up; >6mm, CT at 3-6 months to confirm persistence (could be infectious and resolve). If solid component remains <6mm, annual CT for 5 years  (grade 1C; strong recommendation, low- or very-low-quality evidence). If solid component >6mm, highly suspicious for cancer (grade 1B; strong recommendation, moderate quality evidence.)

--Multiple subsolid nodules

   --<6mm: CT at 3-6 months. If stable, consider repeat at 2 and 4 years (grade 1C; strong recommendation, low- or very-low-quality evidence)

   -->6mm: CT at 3-6 months. if persistent, consider diagnosis of multiple primary adenocarcinomas (grade 1C; strong recommendation, low- or very-low-quality evidence). subsequent management depends on most suspicious nodule.

--risk factors for malignancy:

    --nodule size is the dominant risk factor

    --nodules are classified as solid, pure ground glass, and part-solid. BUT huge inter- and intra-observer disagreements (correct classification of nodules as solid or sub-solid was found in only 58% of cases!!). Those with marginal spiculation are more likely to be malignant (OR 2.2-2.5), though this is not really a binary finding, and no threshold of the degree of spiculation has been defined

    --nodule location: upper lobes, and esp right upper lobe, more likely to be cancer, with odds ratio of about 2.0  Adenocarcinomas and mets are more likely peripheral, and squamous cell cancers are more near the hila. Small nodules inperifussural or subpleural areas often are lymph nodes

    --nodule multiplicity: increased risk of cancer as number of nodules increases from 1 to 4, but decreases with >4 (more likely prior granulomatous infection)

    --nodule growth rate: solid cancers double their volume (a 26% increase in diameter) in 100-400 days; subsolid nodules (eg, primary adenocarcinomas) have average doubling times of 3-5 years: hence the longer follow-up time for those with subsolid nodules

    --emphysema/fibrosis: emphysema is independent risk factor for cancer. Old studies found about 3x increase. NLST  (National Lung Screening Trial, which led to current screening recommendations for smokers) found incidence of 25 cancers/1000 screened in those with emphysema and 7.5/1000 in those without. Both emphysema-predominant COPD and increasing severity of centrilobular emphysema increase the risk of cancer. As does pulmonary fibrosis (espidiopathic pulmonary fibrosis, with HR 4.2)

    ​--age/demographics: rare <40yo, and increases each decade of life. Women may be at higher risk in a few studies: esp if lower BMI and lower educational level, but also a higher cancer risk in women with nonsolid nodules. Family history  is a risk factor both in smokers and never-smokers, with RR 1.5-1.8 if an affected sibling. Also higher in black men and native Hawaiian men at lower levels of smoking. 

    --tobacco/exposures. 10- to 35-fold increased risk in smokers. passive smokers also with increased risk, but less so. smoking mostly associated with squamous cell cancers. incidence of adenocarcinomas is increasing over time, esp for female nonsmokers, but unclear effect of smoking on this and smoking not included as a risk factor for adenocarcinomas. other inhaled carcinogens noted as cancer risk factors include asbestos, uranium and radon. [silicosis may be, but not shown conclusively, per my reading. similar with beryllium exposure. and, no doubt, others]

--the categories of risk used in this guideline are from the American College of Chest Physicians (these are spelled out, with a mathematical equation including the variety of risk factors, in http://journal.publications.chestnet.org/article.aspx?articleid=1685303​ ):

    --low-risk: estimated cancer risk (<5%), found typically in those of young age, less smoking, smaller nodule size, regular margins, and location other than upper lobe

    --high risk (>5%), more often if older age, heavy smoking, larger nodule size, irregular or spiculated margins, and upper lobe location

--one not-so-uncommon clinical situation is the patient who has an abdominal CT, where the CT finds a small lung nodule (<6mm), but only the lower part of the lungs is visualized. If patient is low risk, no further follow-up recommended. If intermediate size (6-8mm), follow-up CT of the complete chest after 3-12 months depending on clinical risk.  If larger nodule or suspicious characteristics, full chest CT right away for further evaluation

Commentary:

--with increasing use of CT scans, lots of lung nodules are found. in US adults between 2006-14, more than 4.8 million had at least one chest CT, with >1.5 million nodules identified, and 63K lung cancers diagnosed within 2 years.

--I would personally include as higher risk any patients with industrial exposures esp if there is evidence of distorted lung parenchyma (eg fibrosis, as by silicosis), especially since other causes of fibrosis are higher risk (emphysema, idiopathic pulmonary fibrosis). and I would be more inclined to follow them more closely, as well as caution them about avoiding any other exposures more aggressively (smoking, etc).

--they do not mention HIV in this guideline (they refer to immunocompromise as creating higher likelihood of infection, no mention of cancer), but there are observational studies of 2- to 4-fold increased risk of lung cancer at younger age and lower smoking exposure: I had a patient who was an elite controller, who had an undetectable viral load, high CD4 count off any antiretroviral meds, minimal smoking history, but who died in his late 40s from lung cancer. My review of the literature found this phenomenon not so uncommon, even in patients with good immunologic response to antiretrovirals.  so I would add HIV, controlled or not, as potential risk factor

--I am very concerned about radiation exposure (as per many prior blogs, see http://gmodestmedblogs.blogspot.com/search/label/radiation ). Above they mention that the repeat CT scans done should produce only 3 mGy of radiation exposure. LDCT, low-dose CT used for screening smokers, is 1.5 mSv, which from my search is the same as 1.5 mGy, whereas a diagnostic chest CT is about 7-8 mGy, and a chest xray about 0.1 mGy. So the recommended radiation dose for the follow-up repeat CT is less than 1/2 that of a regular diagnostic CT. I'm not sure exactly what that means. Is the follow-up CT different from a diagnostic CT?? Or are they recommending new CT scanners which deliver less radiation, but may not be available in many places (and might limit the generalizability or utility of their algorithm, since more cancers may be created by the higher radiation exposure)? There was an article about a new Toshiba CT scanner which delivers around 4 mGY: see https://www.technologyreview.com/s/510861/ct-scanner-delivers-less-radiation/​ , which may be what they are referring to....

--BUT, one concern I have is that the additional radiation exposure from multiple CTs (even from lower radiation ones) might have an even higher risk of causing cancer in patients with baseline abnormal lungs. The data on radiation exposure and cancer, from what I can find, is largely mathematical modeling based on people with normal lungs (eg see radiation exposure and lung cancer JAMA 2007 in dropbox, or Einstein AJ JAMA 2007; 298: 317, as well as http://gmodestmedblogs.blogspot.com/2014/11/cardiac-imaging-and-radiation-exposure.html , http://gmodestmedblogs.blogspot.com/2013/07/ct-scanning-in-kids-and-radiation.html   ). My guess, though not addressed in anything I have seen, is that those with diseased lungs are at higher risk of radiation-related lung cancer, and that risk may be much higher than estimates from the current mathematical models of people with normal lungs (eg, maybe the underlying lung pathology is associated with inflammation and fibrosis which is associated with significant chromosomal damage, etc, which puts the lungs at higher cancer risk from further damage by radiation??? sort of a multiple-hit theory??)

So, I think this guideline is helpful for us in primary care. Given the rather low bar to get CT scans these days and the frequent finding of difficult-to-interpret incidental “abnormalities” found, it is useful to have some sense of how to interpret and follow the findings, and why. of course, there are real concerns about the radiation exposure, but at least these guidelines are more lenient than prior ones (larger size cutpoint of when to do followup CTs, and less aggressive followup than before)​