Note: This guideline had been updated. The National Guideline Clearinghouse (NGC) is working to update this summary. The recommendations that follow are based on the previous version of the guideline.
Note from the National Guideline Clearinghouse (NGC): This guideline is a part of a larger document, Pulmonary (Acute & Chronic) which, in addition to "Lung cancer and cancer of the pleura," includes the following chapters:
Initial Evaluation of Chronic Cough
Lung cancer is largely, although not completely, a preventable condition. However, the use of beta carotene, vitamin E, retinoids, N-acetylcysteine, aspirin, budesonide, cyclooxygenase-2 (COX-2) inhibitors, and 5-lipoxygenase (5-LOX) inhibitors have not been shown to be of value in the prevention on a primary, secondary, or tertiary level. Selenium and organosulfurs are currently being studied for chemoprevention but the results are still unknown and these agents cannot be recommended for use at the current time, except as part of a study program.
As the earlier stages of lung cancer are generally more amenable to treatment than later stages, it seems reasonable to perform screening evaluations, especially in high-risk individuals. Studies using surveillance chest x-rays and sputum cytologies have not demonstrated sufficient value to allow for a recommendation in these individuals. Surveillance computed tomography (CT) studies of the chest have demonstrated suggestions of value, but the results of sufficient numbers of screened individuals is not yet available to allow for a recommendation. Further, the use of CT imaging, by itself, raises the concern of radiation delivery, especially in the susceptible individual. For this reason, current studies in high-risk groups utilize low-dose radiation CT scanning (LDCT). Major issues that need to be addressed when reviewing these studies are:
- The over-all mortality of the two groups (individuals surveilled vs. those not subject to surveillance)
- The disease-free interval in these two groups
- The surgical stage of the individual in these two groups
- The rates of false-positive results and the risk/cost of the evaluation of those individuals who have an abnormality detected on surveillance
A frequently encountered problem in these studies is that the time from diagnosis to death is often prolonged in the group that had the cancer discovered by surveillance techniques, raising the issue of effectiveness of the surveillance technique. Thus, if the total mortality is identical, then the only benefit of the surveillance was to allow for an earlier diagnosis, with no appreciable difference in the outcome of the patient. It needs to be remembered that lung cancer is present for years prior to the time of diagnosis and that the time from inception until detection represents the greater time interval than the time interval from diagnosis until death. For example, with a doubling time of 65 days, lung cancer is present for approximately 4 years before being detectable by CT imaging whereas it is generally present for 6.2 years prior to "normal" detection and 7.5 years prior to death.
Other surveillance techniques include sputum analyses for biomarkers, the presence of volatile organic compounds in the exhaled air, and screens for deoxyribonucleic acid (DNA) alterations. The value of these tests is undergoing research at the current time and their use cannot be recommended.
All lung cancers should be staged using the following protocol:
- T stage by histologic type, histologic grade, tumor size, location, pleural involvement, lymphovascular invasion, mediastinal or chest wall extension, and the resection margins
- N stage by lymph node (hilar/bronchial), mediastinal (location/station), and distant sites
- M stage by the presence of metastases
General principles: When approaching the diagnosis of possible lung cancer, it is generally easier to approach the condition based on the location of the suspected lesion rather than by the cell type. From a simplistic standpoint, lung cancer presents either in the lung parenchyma or in the airways. In the lung parenchyma, it usually presents as a solitary pulmonary nodule (SPN), less commonly as a mass lesion, as a lesion abutting the pleural space, or as multiple nodules. (Multiple nodules generally imply metastatic cancer.) In the airways, it is referred to as an endobronchial tumor although there is generally parenchymal invasion as well. This differentiation by site can also be described as central (endobronchial) vs. peripheral (SPN). The diagnostic approaches will be different for the parenchymal lesion as opposed to the endobronchial lesion. For example, a sputum cytology specimen would be more likely to demonstrate malignant cells with the more central, endobronchial tumor than with the peripheral, parenchymal tumor.
||Dyspnea, hemoptysis, cough
||None, signs of metastatic disease
||Normal, mass lesion in central chest, obstructive atelectasis or pneumonia
||Solitary pulmonary nodule
|Diagnostic tests of higher yield
||Sputum cytology, bronchoscopy
||Squamous cell, small cell
Sputum cytologies oftentimes represent the easiest way to confirm either the suspicion or diagnosis of lung cancer. The value of the test depends on a variety of criteria, with reported overall sensitivities ranging from 0.42 to 0.97 and specificities ranging from 0.68 to 1.00. Factors that make this test more likely to be accurate include the presence of hemoptysis and a central/airway tumor; conversely, the reliability in parenchymal nodules is lower. Further, the greater the number of specimens submitted, the greater the accuracy, with sensitivities of 0.68 for one specimen, 0.78 for two specimens, and 0.85 to 0.86 for three or more specimens. Similar statements can be made about flexible fiberoptic bronchoscopy (FOB), for similar reasons. In patients with central tumors, the overall sensitivity of FOB in establishing a diagnosis is 0.88, with the sensitivity of a directed forceps biopsy being 0.74 and 0.48 to 0.59 for washings and brushings respectively. For a peripheral nodule, the sensitivity of a transbronchial biopsy is 0.57 (most studies used fluoroscopic guidance) with the number of specimens taken affecting the accuracy (ranging from 0.45 for one specimen to 0.70 with six specimens); the sensitivity of the brush biopsy in these cases is 0.54 and the sensitivity of bronchoalveolar lavage (BAL) is 0.43.
The choice of diagnostic test is determined by the likelihood of a positive result and this depends on the presumptive diagnosis and the risk factors for the presence of cancer. These tests include:
- Sputum cytology
- Observation with serial CT studies
- Transthoracic, percutaneous needle biopsies
- Surgical biopsies performed by way of a thoracoscopic or open approach
In a review of studies of individuals with lung cancer, representative results of diagnostic tests were:
- Positron-emission tomography (PET) scan — for lung nodules — sensitivity 87% with specificity of 83%; for mediastinal lymph nodes — sensitivity 74% with specificity of 85%
- Dynamic CT with nodule enhancement — for lung nodules — sensitivity 96% with specificity of 75%; for mediastinal lymph nodes — plain CT — sensitivity 51% with specificity of 85%
- CT-guided needle biopsy — nondiagnostic results in 44% of patients with benign nodules (median value) and 8% of patients with malignant nodules (median value); diagnostic results in, approximately, 90% of patients with malignant nodules. The average risk of pneumothorax was 26.5% with 5% of patients requiring a chest tube. Factors that increase the risk of pneumothorax include the presence of emphysema, the location (i.e., central vs. peripheral), and the size of the SPN. Using CT imaging for guidance of the needle is slightly more accurate than the using for fluoroscopy (sensitivities of 0.92 vs. 0.88)
- Bronchoscopy — diagnostic yield is 10% to 50% with variables affecting the yield including:
- The size of the nodule
- The presence of air bronchograms in the nodule
- The presence of malignant vs. benign disease
- The location of the nodule
- The simultaneous use of assistive devices such as CT or endobronchial ultrasound guidance or electromagnetic navigation
- The number of specimens taken (as discussed above)
- The presence of an airway extending into the lesion (sensitivities of 0.60 to .025 for those nodules without this feature)
- The size of the nodule (0.34 for nodules <2 cm to 0.63 for nodules >2 cm)
- The use of a post-bronchoscopy sputum sample (0.35)
- The use of ultrasound guidance (an increase in diagnostic yield by up to 82%)
- The use of autofluorescence and laser light allows for detection of carcinoma in situ.
- Thoracoscopic biopsies — sensitivity and specificity approaching 100% with an associated mortality rate of 1%.
- The use of ultrathin bronchoscopy guided by virtual bronchoscopy (increase in success rate from 43 ± 16% to 94 ± 7.9%)
Special staining is oftentimes used in lung tumors to identify cell type. In individuals with a pleural based tumor, a variety of stains are available to help distinguish an adenocarcinoma (primary or metastatic) from mesothelioma. The "gold standard" is the ultrastructural features identified on transmission electron microscopic evaluation. Similarly, a variety of immunohistochemical assays can be used to distinguish small-cell vs non-small cell cancers and primary from metastatic cancers affecting the lung. Of value is the ability to distinguish bronchoalveolar carcinoma (BAC) adenocarcinomas from those without this component. (BAC is a subtype of adenocarcinoma that grows along the luminal surface of alveolar septae with no evidence of parenchymal, vascular, or pleural invasion.)
For individuals with suspected Stage IV disease (metastatic) or small-cell lung cancer (SCLC), the diagnostic test of choice is the one that can be performed with the least risk to the patient, as the goal is merely to confirm the diagnosis cytologically or (preferred) pathologically. This might involve, for example, the sampling of a peripheral but easily biopsied lymph node.
Solitary pulmonary nodule: The recommended diagnostic approach for the SPN starts with an imaging study. Many of these tumors will be discovered serendipitously, as they generally do not present with symptoms such as hemoptysis or dyspnea. They are generally encountered, for example, on a pre-surgical or other type of screening chest x-ray (CXR). The first step in the evaluation of these patients is to review the old x-rays to determine if the SPN was present in the past and, if so, if there has been any change in the size. The nodule itself is evaluated for other features that help determine malignancy vs. benignancy. Features of a SPN that are more commonly associated with benignancy are:
- Presence of calcium in the nodule
- Long doubling time (usually ≥2 years although tumors with longer histologic doubling times would be expected to have longer radiographic doubling times, sometimes exceeding the 2 year value; therefore, inspect the nodule for any growth)
- Smooth vs "shaggy" or indistinct borders (only a relative factor since as many as 20% of the latter group can be malignant)
- Size ≤4 mm on CT scanning
Of concern when reviewing studies of these patients is the nature of the population being studied, i.e., individuals with risk factors vs. the general population, as the incidence of malignancy vs. benignancy varies with the population studied. The addition of PET scanning conveys good sensitivity with only fair specificity. Other features that suggest benignancy are age of the patient and the absence of risk factors.
If the patient is a candidate for surgery, then further evaluation will depend on:
- Patient preference (for example, if the individual would never have lung surgery or take chemotherapy, even in the face of an established diagnosis of cancer, it would be pointless to pursue the evaluation)
- The general medical status of the individual (for example, if the patient would not tolerate lung surgery due to infirmity, significant other medical conditions, advanced age, or significant underlying pulmonary disease, then the goal is merely to establish a cytological diagnosis)
Diagnostic testing for these individuals varies and depends on factors such as:
- The desire to have an established diagnosis prior to surgical intervention
- The presence of risk factors
- The size of the SPN
- The general medical status of the individual (those individuals closer to marginal status for lung resection generally require only confirmation of the cancer and the possible presence of metastases prior to surgery)
- The results of non-interventional evaluations such as the CT scan or the PET scan
- The risk/benefit ratio of the proposed diagnostic test (for example, a transthoracic percutaneous needle biopsy is associated with a higher risk if the lesion is considered highly vascularized as demonstrated by CT imaging with dynamic contrast)
- If the SPN is presumed benign but a diagnosis of the benign disease is necessary (for example, an infectious process)
The American College of Chest Physicians (ACP), in review article, recommends that if an individual has a non-diagnostic study of an SPN that was <2 cm in diameter, further testing be performed if the suspicion of cancer remains.
Pleural-based tumors: In individuals with a pleural-based tumor and an effusion, it is generally presumed that the tumor is already metastatic (either primary lung or from a different organ) to allow for the effusion, although the rare case of post-obstructive pneumonia/atelectasis in the individual with significant obstruction of an airway with associated pleural effusion may be seen. The evaluation of the pleural-based tumor is by a directed percutaneous needle biopsy. In the individual with pleural effusion, blind biopsies are diagnostic in approximately 50% of cases. When three separate effusions are analyzed for malignant cells, the results are positive for malignancy in approximately 80% of cases. Thoracoscopic-guided biopsies are more accurate with sensitivities between 0.80 and 0.99 and specificities between 0.93 and 1.00 with a negative predictive value between 0.93 and 0.96. The most common cause of inaccuracy is of the presence of mesothelioma.
Paraneoplastic syndromes: Lung cancers can present with a variety of paraneoplastic syndromes that should not, by themselves, prevent therapeutic intervention with curative intent.
Scar carcinomas: It has long been known that some individuals with scars on their lungs are at increased risk of developing lung cancer. The postulated etiology is unrelieved, chronic inflammation. Carcinomas have been found in individuals with tuberculomas, mycetomas, pneumoconiosis, and interstitial fibrosis. A recent cohort study of 66,863 individuals identified scars on chest radiographs in 5,041 subjects. The hazard ratio was found to be 1.5 (95% confidence interval 1.2 to 1.8) with increasing incidences over time.
Evaluation of the Individual Prior to Resectional Surgery
Clearly it is of no value to cure the cancer but leave the patient a pulmonary cripple. Thus, the individual with (or with suspicion of) lung cancer may need pre-operative clearance, depending on his/her general or pulmonary status. The ACP made recommendations regarding this evaluation and provided graded recommendations for a variety of interventions. These recommendations include issues related to the patient's general medical status (primarily cardiac). The pulmonary status variables included symptoms and, especially, results of pulmonary function testing. Age alone is not a contraindication to curative surgery.
The search for the presence of metastatic disease is dependent on certain characteristics, although it is recommended that all patients have CT imaging of the chest that includes views of the upper abdomen (to include the liver and adrenal glands). Following a review of the recent medical literature, the ACP recommended evaluation for metastatic disease if the following features are present:
- Weight loss >10 lbs
- Skeletal pain
- Neurologic symptoms that included headache, syncope, seizures, extremity weakness, or recent change in mental status
- Lymphadenopathy (>1 cm)
- Superior vena cava syndrome
- Bone tenderness
- Hepatomegaly (>13 cm span)
- Focal neurologic signs
- Soft tissue mass
- Laboratory abnormalities:
- Hematocrit (<40% in men and <35% in women)
- Elevated liver enzymes (alkaline phosphatase, gamma-glutamyltransferase, or serum glutamic oxaloacetic transaminase [SGOT])
The ACCP (American College of Chest Physicians) recommended that, regardless of the absence of the above abnormalities, those individuals with clinical state IIIA or IIIB disease undergo evaluation to determine the presence of distant metastases. It also makes intuitive sense that those individuals having borderline general medical or pulmonary status pre-operatively also have an evaluation for metastases and that the closer the individual is to a prohibitive surgical risk, the more important will be this evaluation.
Elements of the metastatic evaluation depend on the potential/probable site of the metastasis. Tests used for screening include:
- Head — CT or magnetic resonance imaging (MRI)
- Bone — blood testing, bone scanning, or whole body PET scanning
- Other — blood testing, whole body PET scanning, or abdominal/pelvic CT scanning
In those individuals who have abnormal imaging results consistent with metastatic disease, biopsy confirmation (or overwhelming clinical evidence) should precede their rejection from an attempt at a surgical cure. For potential mediastinal lymph node involvement (as determined either by lymph node enlargement on CT scanning or by PET scanning [note, this is an either/or situation; in other words, an enlarged lymph node on CT scanning needs evaluation even if the PET scan is negative]), the initial step is a biopsy. Reasonable approaches include mediastinoscopy, EUS-NA (esophageal endoscopic ultrasound with needle aspiration), TBNA (transbronchial needle aspiration), EBUS-NA (endobronchial ultrasound with needle aspiration), TTNA (transthoracic needle aspiration), VATS staging (video-assisted thoracic surgery), Chamberlain procedure (also known as an anterior mediastinotomy), and extended cervical mediastinoscopy. Each procedure carries its own accuracy rate although these rates are determined by several factors including the size and number of abnormal lymph nodes, the skill of the operator, and the location of the lymph node. If the needle biopsy is non-diagnostic, then a mediastinoscopy should be performed.
In summary, the approach to the patient with suspected lung cancer is:
- Review the old CXR.
- Determine if the individual is likely to be at risk for lung cancer.
- Perform CT scanning of the chest and upper abdomen.
- Possibly supplement with PET scanning.
- Perform a diagnostic test.
- Recommend a therapeutic procedure.
The general rule is to assume that a neoplasia exists if the clinical suspicion is raised based on an abnormal test result such as a sputum cytology or imaging study or if certain signs or symptoms are present such as hemoptysis in an individual with risk of cancer.
The goal of treatment in an individual with lung cancer is to cure the cancer. In most circumstances, this will not be possible. The choice of initial treatment depends on many factors, but most especially the pre-surgical stage of the cancer, the patient’s ability to withstand the therapeutic intervention, and the choice made by the patient regarding the type of treatment from the options presented. The choice of therapeutic intervention starts with dividing the tumor into small-cell and non-small cell cancer and then, for non-small cell cancers, by the clinical stage. (Further, it needs to be appreciated that cancer treatment is an ever evolving discipline.)
A summary of recommendations for non-small cell cancer treatment offered by the ACCP is:
- Carcinoma in situ — treated with a variety of endobronchial techniques such as cryotherapy, photodynamic therapy, electrocautery, laser therapy, and brachytherapy
- Stage I — surgical resection is recommended. Adjuvant chemotherapy is not recommended for Stage IA or Stage IB patients in whom the cancer was amenable to complete resection. Post-operative radiotherapy is not recommended.
- Stage II — surgical resection is recommended followed by adjuvant chemotherapy. Post-operative radiotherapy is not recommended.
- For Stage I and Stage II patients who are either not surgical candidates or who prefer not to have surgery, curative intent fractionated radiotherapy is recommended.
- For Stage III patients found to have occult N2 disease (i.e., Stage IIIA2), resect the primary tumor and metastases, if feasible, followed by adjuvant chemotherapy. Post-chemotherapy radiotherapy is also recommended. Concurrent chemotherapy and radiotherapy is not recommended.
- For Stage IIIA3 patients (local metastases identified pre-operatively), the recommended treatment is either surgery (although not a pneumonectomy) followed by platinum-based combination chemoradiotherapy or platinum-based combination chemoradiotherapy alone.
- For Stage IIIA4 patients, concurrent chemoradiotherapy is recommended without surgery.
- Stage IV — chemoradiotherapy depending on the performance status of the patient
VATS is an acceptable alternative to open thoracotomy, although this modality requires a surgeon experienced with this technique. A lobectomy (or greater removal) is generally preferred over a sublobar resection (such as a wedge or segmentectomy) and a sleeve resection is preferred over a pneumonectomy.
The ACCP recommendations for small-cell lung cancer cover topics such as the use of PET scanning for staging, the use of prophylactic cranial irradiation, surgical intervention, treatment of mixed histology tumors, the differentiation of treatment based upon limited-stage disease vs. extensive-stage disease, second-line treatment, and the treatment for the patient with relapsed/progressive disease.