Epidemiology

Lung cancer is the most common cause of death from cancer in the industrialized world. There are 37,000 new patients per year in the UK and 170,000 per year in USA. Lung cancer is rarely diagnosed below 40 years, but the incidence rises steeply thereafter with a peak at 75–84 years. The median age at diagnosis is 60 years. The life time risk is 1 in 13 in men and 1 in 20 in women in the UK. Lung cancer is responsible for 25% of all cancer deaths in the UK and 28% in USA.

Aetiology

Smoking

Smoking accounts for up to 92% of lung cancer deaths in men and 80% in women. The age-adjusted relative risk of developing lung cancer for people who smoke more than 20 cigarettes per day is 20 times that compared with life-long non-smokers. This risk remains the same until approximately 5 years after smoking cessation and even 15–20 years after stopping smoking there is 2–3 times the risk of developing lung cancer, compared to life-long non-smokers.

Passive smoking causes an average 24% increase in the incidence of lung cancer compared to non-smokers living with non-smoking partners. Passive smoking causes about 3% of the lung cancers.

Pathogenesis and pathology

Pathogenesis

Lung cancer arises in the mucosa of the bronchi or more rarely in the lung parenchyma. It has been hypothesized that invasive lung cancer is the endpoint of a multi-step process as a result of cumulative genetic abnormalities involved in the regulation of epithelial cell growth and division caused by carcinogens in tobacco smoke.

Pre-invasive lesions

The two well-known pre-invasive lesions are atypical adenomatous hyperplasia (AAH) and diffuse idiopathic pulmonary neuroendocrine cell hyperplasia (DIPNECH). AAH manifests as localized ground glass areas of less than 5 mm on high resolution CT scan and progresses to invasive adenocarcinoma at an unknown rate. DIPNECH manifests as small nodules and thickened bronchiolar walls associated with a mosaic pattern of air trapping on CT scan, and is a precursor of pulmonary carcinoid.

Box 9.1 shows a classification of lung tumours, which is based on light microscopic assessment of patterns of differentiation. The most important distinction is between small cell carcinoma (SCLC: 15–25% of lung cancers) and other carcinomas, collectively called non-small cell carcinoma (NSCLC). These subgroups behave differently in their presentation, natural history and response to treatment.

Presentation

More than 90% of patients are symptomatic at presentation of which approximately 30% are due to tumour, 30% are due to non-specific systemic symptoms (anorexia, weight loss or fatigue) and 30% due to metastatic disease. Five percent of patients are asymptomatic and are detected incidentally.

Respiratory symptoms

Cough is the most common presenting symptom (70%), followed by dyspnoea (60%), chest pain/discomfort (50%) and haemoptysis (41%).

Other symptoms

Symptoms due to compression of recurrent laryngeal nerve (hoarse voice), phrenic nerve (breathlessness), brachial plexus (shoulder pain and/or arm pain, sensory changes and muscle wasting in the distribution of C8 and T1–2 nerves) and sympathetic plexus (Horner’s syndrome) can occur. Four percent of lung cancer patients present with superior venacaval obstruction (SVCO) and is most common with small cell lung cancer. Pleuritic chest pain and pleural effusion occurs in pleural involvement. Pericardium involvement can present with pericardial effusion and rarely, tamponade. Rarely, dysphagia due to massive nodal disease compressing the oesophagus can occur.

Bone pain is the presenting symptom in up to 25% patients. Liver involvement usually leads to pain if there is massive metastasis. Clinical adrenal insufficiency is rare. Brain metastases at diagnosis occur in up to 10% and present with features of raised intracranial tension, focal neurologic deficits, fits and mental and personality changes.

Paraneoplastic syndromes occur in up to 10% of patients. Box 9.2 lists the paraneoplastic syndromes. SCLC is the most common type associated with paraneoplastic neurologic syndromes. Hypercalcaemia either due to excess parathyroid hormone (PTH) or PTH-related peptides is common in squamous cell carcinomas (15%). Hyponatremia occurs in up to 10% of patients, which is either due to syndrome of inappropriate secretion of ADH (SIADH) or to atrial natriuretic factor. SIADH is commonly associated with SCLC.

Box 9.2
Paraneoplastic manifestations of lung cancer

Endocrine

Syndrome of inappropriate ADH secretion (SIADH)¹

Humoral hypercalcaemia of malignancy ²

Ectopic Cushing’s syndrome ¹^,³

Ectopic hCG syndrome ⁴

Growth hormone excess ³

Cardiovascular

Thromboplebitis ²

Arterial thrombosis

Marantic endocarditis

Haematological

Haemolytic anaemia

Red cell aplasia

Thrombocytosis

Others

Nephrotic syndrome

Hypouricaemia

Investigations and staging (Figure 9.1)

Evaluation of a patient with suspected lung cancer is to:

• establish histologic type

• define extent of disease (staging)

• assess fitness to undergo the best appropriate treatment.

Figure 9.1 Evaluation of lung cancer.

Imaging

Chest X-ray (Figure 9.2A) may show lung lesion with or without lymphadenopathy. There can be associated lung collapse, pleural effusion, synchronous nodules or pericardial effusion. Bone metastases or bony destruction due to direct invasion can also been seen.

Figure 9.2 Imaging in lung cancer.

Contrast enhanced CT scan of chest, including upper abdomen (3–10% patients show asymptomatic liver and/or adrenal metastasis – Figure 9.2B&C) should be performed prior to further diagnostic investigations including bronchoscopy.

Lymph nodes of >1 cm in short axis diameter, a central low intensity suggesting necrosis and rounding of the contour of a hilar node where it meets the lung margin suggest malignant lymph nodes. CT is not reliable in assessing lymphadenopathy, to distinguish T3 from T4 disease and in demonstrating chest wall invasion, which all are important in making treatment decision.

CT scan is useful in assessing resectability. Encasement of proximal pulmonary arteries/veins, gross mediastinal involvement by tumour, widespread mediastinal lymphadenopathy and distant metastasis on CT scan suggest an unresectable tumour (96% accuracy). Tumour invasion of central pulmonary artery and vein, involvement of main stem bronchus and tumour extension across the major fissure (anywhere on the left side, above the minor fissure on right), all suggest the need for pneumonectomy for complete resection.

CT scan of brain is done if there is clinical suspicion of brain metastasis or patient is planned to undergo curative treatment.

Bone scan is indicated if there is bone pain or isolated raised alkaline phosphatase.

Tissue diagnosis

Flexible bronchoscopic biopsy can be obtained in 60–70% of lung cancers (central tumours). Sampling using multiple techniques (biopsy, brushings and washings) gives the highest diagnostic yield (sensitivity of 83–88%).

CT guided percutaneous needle aspiration/biopsy – is useful in peripheral tumours. Core biopsies improve the sensitivity compared with aspirates. Complications include bleeding and pneumothorax.

Fine needle aspiration is suitable for sampling lymph nodes, skin nodules and liver and adrenal metastases. Pleural effusion requires aspiration, biochemical analysis and cytology together with multiple pleural biopsies.

Sputum cytology has wide variation in sensitivity (10–97%) and should be reserved for patients with central tumours who are unable to tolerate or unwilling to undergo bronchoscopy or other invasive procedures.

Staging

Stage determines prognosis and guides treatment. TNM staging of lung cancer is given in Box 9.3. Small cell lung cancer can also be staged using a simple staging classification (Box 9.4). SCLC can be classified as TNM staging or using the simplied.

Box 9.3
Staging of non-small cell lung cancer

Stage Ia

T1N0	Tumour ≤3 cm diameter

Stage Ib

T2aN0M0	Tumour >3 cm or ≤5 cm OR
	Tumour of ≤5 cm with involvement of visceral pleura, main bronchus ≥2 cm distal to the carina or partial atelectasis

Stage IIA

T2bN0M0	Tumour >5 cm or ≤7 cm
T1-T2aN1M0	Metastasis in ipsilateral hilar or peribronchinal nodes

Stage IIB

T2bN1M0
T3N0M0	Tumour of >7 cm OR
	Tumour invading chestwall, diaphragm, phrenic nerves, mediastinal pleura or parietal pericardium OR
	Tumour <2 cm from carina, atelectasis of entire lung or nodule in the same lobe

Stage IIIA

T4N0-1M0	Tumour invading mediastinal structures or nodule in different ipsilateral lobe
T1-3N2M0	Metastases in ipsilateral mediastinal and/or subcarinal nodes

Stage IIIB

T4N2M0
anyT N3M0	Metastases in contralateral hilar or mediastinal nodes
anyT N3M0	Metastases in scalene or supraclavicular nodes

Stage IV

M1a	contralateral lung nodule/pleural nodule/malignant effusion distant metastasis
M2b

Box 9.4
Staging of small cell lung cancer

Limited stage (stage I-III)

Tumour confined to the hemithorax of origin, the mediastinum, and the supraclavicular nodes, that can be encompassed within a tolerable radiation therapy port.

Extensive stage (stage IV)

Tumor that is too widespread to be included within the definition of limited-stage disease above. Patients with distant metastases (M1) are always considered to have extensive stage disease.

Mediastinal staging in non-small cell lung cancer

Accurate distinction of stages II, IIIA and IIIB is important in deciding optimal surgical treatment in the potentially operable patients. Stage II disease is treated with lobectomy or pneumonectomy with mediastinal sampling or dissection. Stage IIIA disease may be treated with neoadjuvant chemotherapy followed radical surgery or radical chemoradiotherapy, whereas IIIB disease is unresectable. This necessitates both an accurate distinction of T3 from T4 tumours and an accurate staging of nodal status. A classification of mediastinal nodes is shown in Figure 9.3.

Figure 9.3 The classification of mediastinal nodes, patterns of nodal spread and method of choice of staging.

Methods of mediastinal staging

Positron emission tomogram (PET) (Figure 9.4) – PET is scan based on the principle of differential uptake of 2-[fluorine-18]-fluoro-2-deoxy-D-glucose (FDG) by cancer cells. FDG is metabolized at a higher rate by tumour cells than normal cells and the areas of increased activity can be detected by a scanner. FDG-PET has higher sensitivity (87% vs. 68%), specificity (91% vs. 61%) and accuracy (82% vs. 63%) at detecting mediastinal disease than CT scan. PET also has a high negative predictive value in exclusion of N2/N3 disease, which helps to omit further mediastinal staging in case of a negative PET. Overall, there is 20% improvement in accuracy of PET over CT for mediastinal staging of NSCLC.

Figure 9.4 PET scan in lung cancer.

PET also detects distant metastases. The incidence of extra thoracic metastases on PET which are not suspected by CT scan is 9 to 11%. CT false-positive adrenal nodules have been correctly identified as negative on FDG-PET.

One of the limitations of PET is false-positive ‘hot spots’ in the mediastinum due to inflammatory processes (in up to 13 to 17%). It is less accurate when the lesion is <1 cm and in tumours of low metabolic activity such as carcinoid and bronchoalvelolar carcinoma.

PET scan is recommended for all NSCLC patients being considered for radical treatment. Patients with no lymphadenopathy on CT scan but with PET scan positive nodal disease needed mediastinoscopy prior to definitive treatment as 50% of patients can have false positive PET scan.

MRI scan – has limited role in staging. This may be useful to evaluate vascular and vertebral body invasion and to assess integrity of brachial plexus in Pancoast’s tumour.

Endobronchial ultrasound (EBUS) – EBUS is able to penetrate up to 5 cm and can identify lymph nodes and vessels. It is helpful in visualizing paratracheal (stations 2 & 4) and peribronchial lymph nodes (see Figure 9.3) and enables EBUS guided transbronchial needle aspiration (Figure 9.5). Sonographic features of malignant lymph node are round shape, sharp margins and hypoechoic texture. The advantage of EUS over CT and PET is that it can characterize lymph nodes smaller than 1 cm. It can also determine the depth and extent of tumour invasion of tracheobronchial lesions and helps to define the relationships to the pulmonary vessels and the hilar structures. EBUS has been found to predict the lymph node staging correctly in 96% of cases, with a sensitivity of 95% and specificity of 100%.

Figure 9.5 Endobronchial ultrasound (A) and its use for FNA (B) of an enlarged lower paratracheal node (C&D).

Endo-oesophageal US (EUS) with fine needle aspiration is useful to assess sub-aortic (station 5), subcarinal (station 7), and paraoesophageal (station 8) lymph nodes as well as the upper retroperitoneum, which are not accessed by mediastinoscopy. It is envisaged that EUS-FNA and EBUS will not replace but will complement surgical techniques like mediastinoscopy.

Transbronchial needle aspiration (TBNA) – TBNA using a flexible fibre optic bronchoscope has a sensitivity of 78% and high specificity (approaching 100%) for identifying mediastinal metastases. However, the diagnostic yield is operator dependent.

Mediastinoscopy is generally regarded as the gold standard for preoperative mediastinal evaluation. CT scan is helpful in guiding selection of patients for mediastinoscopy prior to surgery. Cervical mediastinoscopy allows better access to contralateral lymph nodes. It is useful in evaluating paratracheal (2 & 4), scalene and inferior mediastinal nodes (7, 8 & 9). It is not useful in evaluating aortic nodes (5 & 6). Mortality rates from mediastinoscopy are negligible, but morbidity rates especially arrhythmias are reported to be 0.5–1%. It has to be borne in mind that up to 30% of patients with lung carcinoma who have a negative mediastinoscopy prove to have mediastinal nodal metastasis at surgery.

Video-assisted thoracoscopy (VATS) is useful to evaluate aortic (5 & 6), paraoesophageal (8) and pulmonary ligament nodes (9). It is also useful in assessing the pleural cavity and obtaining pleural biopsies.

Management of lung cancer

Patients presenting with symptoms suggestive of lung cancer need urgent evaluation (Figure 9.1) and referral to a team specializing in the management of lung cancer. Treatment depends on type of lung cancer (NSCLC vs. SCLC), stage, performance status, co-morbidities and cardiopulmonary reserve.