Introduction
An incidental pulmonary nodule (IPN) is a single, well-defined process in the lungs
that is found incidentally and does not exceed 3 cm in diameter [1]
[2]
[3]. The increased use of high-resolution cross-sectional imaging in recent decades
has significantly increased the detection rate of IPNs [4]. In the Netherlands, too, the identification of IPNs in chest CTs has steadily increased
over the past decade and was associated with more stage I lung cancer diagnoses [5].
The vast majority of IPNs in clinical CT examinations are benign, but a very small
proportion turn out to be lung cancer. In Germany, approximately 57 000 people develop
lung cancer each year. Lung cancer is one of the most prognostically unfavorable tumors,
which is reflected in a low relative 5-year survival rate of around 21 percent in
women and 15 percent in men, as reported in Germany in 2019 [6]. Survival rates in lung cancer vary significantly depending on the stage of the
disease. Since patients with lung cancer often do not report any complaints in the
early stages, the disease is often discovered late and unexpectedly. Native low-dose
computed tomography (LDCT) detects lung cancer at earlier stages than chest radiography,
and leads to a reduction in lung cancer-related mortality in both structured screening
programs for the high-risk population [7]
[8]
[9]
[10]
[11] and consistent follow-up of IPNs [12].
However, while only high-risk groups meet the inclusion criteria for lung cancer screening,
a broader population would benefit from consistent follow-up of IPNs [4]. This is very important because a large number of patients with lung cancer do not
meet the usual inclusion criteria of an early detection program for the high-risk
population [13]
[14]. For example, more than 10 % of lung cancer cases occur in patients who have never
smoked [15]. Furthermore, the participation rate of the high-risk group in an early detection
program is often low [16]
[17]
[18]. For example, in Mississippi in the USA, 38 % of cancer diagnoses were made in a
structured IPN program for consistent guideline-compliant follow-up of IPNs, compared
to 8 % in the screening program for the high-risk population and 54 % with symptoms
in the clinic (clinic group). Approximately 51 % of patients with lung cancer diagnosed
through the IPN program did not meet the inclusion criteria of the screening program
for the high-risk population. Furthermore, a better 5-year survival compared to the
clinic group could be demonstrated [19]. Therefore, consistent IPN follow-up would make an additional contribution to the
population in addition to the success of screening [19].
Although there are already many national and international recommendations for action
[1]
[2]
[3], in reality, they are often not known, communicated or implemented [4]
[20]
[21]
[22]. In addition to recommendations for action, several technical approaches are now
available to optimize detection, risk assessment, and follow-up.
This article provides a review of the current national and international recommendations
for action, the typical pitfalls, and possible solutions for more effective follow-up.
Main part
What are incidental pulmonary nodules and how common are they?
Pulmonary nodules are deemed incidental if they are discovered by chance during other
examinations. These may include examinations of neighboring organs and structures
in which the lungs have been partially screened too, e. g., CT of the shoulder or
MRI of the spine. Exceptions are examinations with oncological or infectious indications,
as this increases the likelihood of nodules. They are usually largely roundish dense
nodules up to and including 30 mm in diameter that are at least partially surrounded
by lung tissue. For larger dense nodules, the term space-occupying lesion is used.
Depending on their radiation transparency, these nodules are referred to as solid
or subsolid. Subsolid nodules are again divided into ground-glass opacity nodules
and partially solid nodules, in which the underlying lung structure can be fully or
partially delineated ([Fig. 1]). Furthermore, there must be no atelectasis, a plump hilum or pleural effusion,
or other evidence of advanced intrathoracic tumors [1]
[2]
[3]
[23].
Fig. 1 Morphological classification of the nodules depending on their radiation transparency
into (a) solid, (b) ground-glass and (c) semi-solid.
The increasing detection of IPNs in recent years has been associated with more frequent
imaging and improved techniques [4]. For example, between 2006 and 2012 in the USA, the annual number of chest CTs increased
from 1.3 % to 1.9 % in all adults, while the frequency of identification of nodules
increased from 24 % to 31 % in all examinations performed [4]. Moreover, a higher level of awareness after the initial publication of the Fleischner
criteria in 2005 may have further contributed to this [24]. Studies from France and China have shown a more frequent occurrence with increasing
age, in men and in smokers or those exposed to smoke, or in people with lung disease
[25]
[26]. Nevertheless, IPNs are often also discovered in individuals who do not meet the
usual inclusion criteria of lung cancer screening. Nodules were detected in 8.5 %
of polytrauma examinations, of which over 80 % required follow-up according to the
Fleischner criteria [27]. Even in a young cohort aged between 18 and 24, an incidence of clinically relevant
nodules of 0.6 per 1,000 person years could still be determined [28].
Risk stratification of pulmonary nodules
The risk of malignancy of a nodule and the general condition of the patient generally
serve as the basis for further management. On the one hand, morphological criteria
or characteristics of the nodule are used for the malignancy risk; on the other hand,
independent risk factors can also be taken into account to statistically estimate
the risk.
With regard to the nodule criteria, size is the dominant factor for malignancy [29]. Growth behavior can also provide crucial information regarding the etiology of
a nodule. For example, solid nodules with volume doubling times of approximately 50–400 days
and subsolid nodules between 3–5 years are suspected of being malignant. Faster growth
is more indicative of an inflammatory event [30]
[31]. Persistent subsolid nodules usually correspond to precursors of adenocarcinoma
with very slow growth. New or growing solid parts of a partially solid nodule are
highly likely to be malignant. Spiculation or pleural involvement are also typical
morphological malignancy criteria [29]. Nodules on caverns or cysts represent a typical malignant manifestation – usually
of adenocarcinoma [32]
[33]
[34]
[35]
[36]
[37]. Furthermore, localization in the upper lobe, 1–4 nodules [29], or concomitant pulmonary fibrosis and emphysema are associated with an increased
risk of malignancy [38]. However, it is important to distinguish nodules that are undoubtedly benign such
as calcified granulomas, apical calluses, or fatty hamartomas.[29]. Even larger (> 6 mm) perifissural, subpleural, or juxtapleural nodules usually
represent benign lymph nodes as long as they are smooth, oval, or triangular, and
should not be checked [2]. However, as soon as morphological abnormalities such as spiculation, retraction
of the pleura, or a history of lung cancer are present, follow-up after 6–12 months
is recommended by the Fleischner Society [2].
By taking into account further epidemiological information such as age, gender, ethnicity,
family history of lung cancer, or information regarding exposure to noxious agents
such as smoking, the individual risk of malignancy can be calculated using statistical
models, including in particular, the Mayo Clinic model, the Brock model (CT) or the
Herder model (CT+PET) [29]
[39]. However, it should be noted that the models were developed based on cohorts with
high pre-test probabilities, meaning that corresponding deviations may occur in patients
with IPNs and a lower overall risk [29]
[39]
[40]. Nevertheless, the recommendations for action of the Fleischner Society, the British
Thoracic Society (BTS), and the current S3 guideline on lung cancer recommend the
use of IPN risk calculators [1]
[2]
[3].
Although it has been shown that nodules > 5 mm can be detected quite reliably with
MRI [41]
[42]
[43] and that a sensitivity and specificity comparable to PET-CT can be achieved using
MR diffusion and MR perfusion imaging [44]
[45]
[46], the BTS guidelines do not recommend an MRI malignancy assessment if a PET-CT is
available (BTS).
S3 guideline on lung cancer
In the revision of the S3 guideline on lung cancer published in December 2022 (version
2.1), the chapter on IPNs was modified and supplemented. Similarly, the current S3
guideline (version 2.2 dated July 2023) refers to the already established international
recommendations for action of the Fleischner Society and the British Thoracic Society
[1]
[2]
[3]. Regarding applicability, it should be noted that the BTS criteria cover all nodules
occurring in patients aged 18 years and over, which have not been pathologically found
to be lung cancer or metastasis. They therefore also apply to lung carcinoma screening.
For screening examinations, the Fleischner criteria and the S3 guideline refer to
the use of the Lung-RADS classification of the American College of Radiology [47]. Furthermore, the Fleischner criteria should only be used in people aged 35 or over,
without known or suspected tumors or immunosuppression. The S3 guideline uses a combination
of these inclusion criteria with a minimum age of 18 years, without known pre-existing
malignancy or immunocompromisation. Furthermore, although it applies to multiple nodules,
it does not apply to disseminated nodules, without this being discussed in more detail
([Fig. 2]). Very small nodules (solid or subsolid) < 5 mm (< 80 mm3), benign nodules (e. g., calcified or fatty), or nodules in patients whose general
condition does not allow diagnostic confirmation or treatment should not be clarified
according to the S3 guideline ([Fig. 3]). Furthermore, any prior imaging should be used to assess the growth behavior. For
nodules (solid or subsolid) ≥ 5 mm and ≤ 8 mm (≥ 80 mm3 and ≤ 250 mm3), follow-up checks should be performed. Follow-up intervals are 3, 6–12, and 18–24
months. Follow-up after 3 months serves primarily to exclude inflammatory changes
in subsolid nodules. In the case of persistent subsolid nodules, check-ups should
be performed over a duration of 3–5 years due to the usually slower growth behavior.
The BTS guidelines recommend follow-ups after 1, 2, and 4 years; for partially solid
nodules, the Fleischner Society even recommends annual follow-ups ([Fig. 4]). Furthermore, if subsolid nodules are found, the size of any solid part that may
form should be used to estimate the probability of malignancy, depending on the patient’s
age, smoking status, peripheral eosinophilia, history of lung carcinoma, and the radiomorphology
of the nodules. Longer intervals generally allow for a more accurate estimation of
the growth behavior by calculating the volume doubling time (VDT). Patients with nodule
growth of < 25 % per year (< 2 mm increase in diameter) or a VDT of > 600 days or
with a limiting general condition can be exempted from follow-up. In the case of faster
growth with a VDT of < 400 days (≥ 2 mm increase in diameter) or formation or increase
in the solid part of a partially solid nodule, definitive pathological clarification
should be sought. Solid nodules > 8 mm to ≤ 30 mm can also be monitored if the risk
of malignancy (Brock model) is < 10 %. For nodules with an initial risk of malignancy
> 10 %, diagnostic confirmation using PET-CT should be offered in accordance with
the S3 guideline, provided that the nodule is above the detection threshold of PET-CT.
The risk can then be re-evaluated using the Herder model, which takes into account
the FDG avidity of a nodule. If the risk of malignancy is still > 10 %, definitive
histological clarification is recommended. However, check-ups may continue to be performed
in cases of a high puncture risk or patient preference. If there is a very high risk
of malignancy > 70 %, resection with rapid incision can be considered even without
prior pathological confirmation. In case of inoperability, nonsurgical ablative or
radiotherapeutic treatment can also be performed. An online calculator for the Brock
model, the Herder model, and for the VDT is available free of charge at https://www.brit-thoracic.org.uk/quality-improvement/guidelines/pulmonary-nodules/pn-risk-calculator/.
Fig. 2 A comparison of the inclusion criteria for the management of incidental pulmonary
nodules of the S3 guideline for lung carcinoma, the Fleischner criteria, and BTS guidelines.
Fig. 3 Initial diagnostic algorithm for solid incidental pulmonary nodules (IPN). Fig. 3
is based on data from the current S3 guideline on prevention, diagnosis, treatment,
and follow-up care of lung carcinoma (July 2023) [1].
Fig. 4 Initial diagnostic algorithm for subsolid incidental pulmonary nodules (IPN) based
on data from the current S3 guideline on prevention, diagnosis, treatment, and follow-up
care of lung carcinoma (July 2023) [1]. Supplementary recommendations of the Fleischner Society and the BTS guidelines
in red [2]
[3].
Further diagnostic and therapeutic courses of action for suspicious nodules should
in principle be decided in a multidisciplinary manner, with the involvement of pneumology,
thoracic surgery, and radiology, and according to the patient’s wishes.
The size of a nodule should ideally be determined semi- or fully-automatically using
volumetry, as this has been shown to be reproducible and more sensitive compared to
size progression [48]
[49]
[50]
[51]. However, different software can lead to significant differences in volumetry, so
the same algorithm should always be used for follow-ups [52]
[53]. If this technical tool is not available, diameters are still specified in the abovementioned
guidelines. It should be noted that the recommendations for action in the S3 guideline
and the Fleischner criteria use the arithmetic mean of the longitudinal and transverse
diameter of the nodule in the same transverse, coronary, or sagittal CT reconstruction.
The BTS guidelines use the maximum diameter of the three spatial planes.
The S3 guideline does not address the question of how to handle IPNs in cases of pulmonary
parenchyma that is not completely detected. In the case of a medium-sized (6–8 mm)
lump, the Fleischner Society recommends monitoring the entire chest after an appropriate
interval (3–12 months depending on the clinical risk). If a nodule is large or looks
very suspicious, a complete chest CT examination is recommended [2].
CT examinations used for the detection of nodules or their follow-up should always
be examined with a native low-dose protocol that complies with the specifications
of the Federal Office for Radiation Protection for the early detection of lung cancer
[54]: An isotropic spatial resolution of 1 mm or less must be achieved. Only in this
way can the images be viewed equally from all sides, and only in this way is volumetry
of lesions that are only a few millimeters in size possible with sufficient accuracy
and reproducibility. For the LDCT scan, a maximum CT dose index (CTDI) of 1.3 mGy
is permitted (based on the standard patient of 80 kg, 175 cm, BMI 26). This value
can and should be significantly lower on modern devices. An important dose reduction
measure in LDCT is the use of patient-specific prefilters (e. g., tin or silver),
which can be appropriately selected via a filter change mechanism.
How are the guidelines implemented?
Although established national and international recommendations for action already
exist, there are numerous indications that these are insufficiently implemented. In
some cases, radiologists, pulmonologists, or other specialists are not aware of the
recommendations for action [55]
[56], or, despite knowing them, they are not correctly applied [57]
[58]. This can lead not only to missed or delayed follow-ups. Premature follow-ups can
also be problematic due to greater inaccuracy in estimating growth behavior and excessive
accumulated radiation exposure. In addition, unnecessary invasive diagnostics or nuclear
medicine examinations could be avoided. However, even simple measures, such as displaying
the recommendations for action at the radiologistsʼ workstations [59] or attaching a description of the respective malignancy risk [60]
[61] to the CT report templates [21] and findings [62], could lead to improved adherence on the part of the radiologists or referring physicians.
Although hundreds of thousands of IPNs are discovered on CT scans each year, follow-up
care appears to be inadequate in most of the newly discovered nodules, with follow-up
rates ranging from 29 % to 39 % [22]
[63]
[64], which raises the question of why approximately 2 out of 3 patients with IPNs do
not receive adequate radiological and clinical follow-up care [22]. These results are interesting in that most radiological reports (up to 68 %) recommended
a follow-up examination of the pulmonary nodules [22], indicating that in many patients with potential early stage cancer, adequate examination
of the lump is not performed. Several pitfalls have been described in the literature
that may be responsible for inadequate treatment of IPNs, which shows that while radiologists
initiate the process of treatment of IPNs by documenting them in the radiological
report, they are not solely responsible for the fact that IPNs are often neglected.
Diverse healthcare providers and patients are also important factors in the success
of IPN management [65].
Software for nodule detection and structured follow-up
Automated AI-based nodule detection has developed significantly over the past few
years, so that there are now several commercial software solutions with FDA or CE
labels. However, these techniques are not yet widely used. Ideally, the nodules are
automatically detected, volumetrized, categorized, and their malignancy risk is estimated
using the model-based Brock score, for example. In addition to the statistical risk
models mentioned above, there are promising approaches in which nodule characterization
is carried out using quantitative image analysis (radiomics) or deep learning algorithms
[66]. The latter has already managed to achieve a high level of sensitivity and specificity
similar to that of experienced radiologists [67]
[68]
[69]
[70]. AI-based software is also making inroads in the areas of speech recognition, structured
report generation, and image reconstruction.
An important component of improving the description of the nodules is incorporating
important nodule characteristics into the radiological findings template. This significantly
increased the complete description of the nodules from 12 % to 47 % [61]. Optimal integration into the radiological work process is always very important
for all the techniques mentioned. Only in this way can the available technical possibilities
be applied in clinical routine.
In this regard, there are various communication and follow-up systems between radiologists,
referring physicians, and patients, which can be used to check the timely implementation
of follow-ups and, otherwise, to send automated reminders to referring physicians
and patients [71]
[72]
[73]. By implementing the Radiology Result Alert and Development of Automated Resolution
(RADAR), the timely follow-ups could be significantly improved from 64.5 % to 84.3 %
[72].
In recent years, specialized IPN clinics and Lung Nodule Management programs have
been established in the USA. The special feature of these facilities is the Lung Navigator,
a person who plans the coordinated procedure for each patient and provides the patient
with important information as the contact person. Other specialized personnel support
the patients and processes in the facilities [74]. This, in addition to an increase in compliance in combination with a structured
screening program, also resulted in a stage shift of lung cancer [12].
