Physical Examination, Complementary Tests, and Functional Status Assessment

This patient presented with significant (>10% from baseline) involuntary weight loss with decreased appetite. These are nearly always signs of a serious medical or psychiatric illness. One study evaluated 154 such patients and found that 36% had cancer, particularly gastrointestinal, lung, lymphoma, renal, and prostate cancers. Twenty-three percent of cases remained unexplained despite extensive evaluation, and remaining patients had primarily other gastrointestinal or psychiatric diseases.¹ Our patient had no evidence of mental illness or another extrathoracic medical disorder to explain weight loss and fatigue. However, he did have moderate chronic obstructive pulmonary disease (COPD), which could be responsible for pulmonary cachexia. The mechanisms that lead to muscle atrophy and weight loss in patients with advanced pulmonary disease are not well understood, but muscle disuse, low-level chronic inflammation, and oxidative stress all appear to contribute to an imbalance between protein degradation and synthesis.² Lung cancer could also explain this patient’s symptoms. Weight loss and tumor burden, however, may not be closely related, because both increased energy expenditure and reduced energy intake may be present; in lung cancer, these appear to be mediated by enhanced production of cytokines, including tumor necrosis factor (TNF) and interleukin (IL)-6.³

Our patient’s CT scan showed a 3-cm right upper lobe (RUL) mass and a 1.7-cm right lower paratracheal lymph node (LN). Computed tomography (CT) scanning is usually affordable and is often performed to define the nature of a pulmonary abnormality, to assess potential mediastinal or hilar involvement, and to assist with the clinical diagnosis of suspected lung cancer. Reasons for choosing one sampling approach over another, for example, are governed primarily by anatomic factors (e.g., the location and size of a lung mass or lymph nodes) rather than by metabolic factors (e.g., positron emission tomography [PET] scan uptake). According to intrathoracic radiographic characteristics (including both the primary tumor and the mediastinum), however, patients with known or suspected lung cancer can be separated into four groups to help guide the choice for these subsequent diagnostic studies.⁴ These groups include the following:

Our patient fits into group B because he had a 1.7-cm LN in station 4R. Although contrast CT is very accurate in detecting LN enlargement, the clinical relevance of LN enlargement for staging is poor, because large nodes may be benign and small nodes may contain metastases in up to 20% of cases.⁵ A diameter larger than 1 cm in the short axis is generally considered suspicious. In a review article, pooled data on CT performance characteristics for detecting mediastinal involvement showed a sensitivity of 57%, a specificity of 82%, a positive predictive value (PPV) of 56%, and a negative predictive value (NPV) of 83%, with marked heterogeneity across individual studies.⁶ This performance is insufficient for clinical decision making, and in many instances, it is inappropriate to rely solely on CT scan for nodal staging.

Noninvasive staging of lung cancer is further enhanced by the use of positron emission tomography with 18F-fluoro-2-deoxy-d-glucose (FDG-PET). A large number of accuracy studies and meta-analyses have demonstrated that PET is superior to CT for mediastinal LN staging in potentially operable non–small cell lung carcinoma (NSCLC).⁷ Sensitivities and NPV were comparable for PET and mediastinoscopy (sensitivity ≈80%, NPV ≈90%). However, the PPV and the specificity of the FDG-PET scan are lower than those of mediastinoscopy owing to the fact that FDG is also taken up by inflammatory processes such as sarcoidosis, fungal disease, or anthracosilicosis. Because of the high NPV of the PET scan, invasive staging procedures generally can be omitted in patients with clinical stage I NSCLC with negative mediastinal PET images. Implementing this strategy warrants caution in patients with central tumors, in those with central hilar N1 disease on CT scan or broncho-alveolar cell carcinoma, in situations with low FDG uptake in the primary tumor, and in cases where mediastinal PET-negative LNs are larger than16 mm on CT scan, as in our case. For example, in one meta-analysis, a post-test probability for N2 disease of 21% was found in patients with PET-negative nodes larger than 16 mm.⁸ In cases of positive mediastinal PET, tissue confirmation is still needed to confirm LN metastasis because the PPV of the PET scan is only 79%. One report that evaluated the role of PET scanning as compared with mediastinoscopy for mediastinal staging in NSCLC found a 26% false positivity rate of PET, a false negativity rate of 25%, sensitivity of 74%, specificity of 73%, and accuracy of only 74%. The NPV of mediastinoscopy was 94%, PPV 100%, sensitivity 84%, specificity 100%, and accuracy 95%. The authors concluded that PET results do not provide acceptable accuracy rates.⁹ Therefore, in our patient with an LN larger than 16 mm in the setting of confirmed adenocarcinoma from CT-guided biopsy of the primary lesion, invasive or minimally invasive mediastinal staging procedures are still necessary.

Comorbidities

This patient had COPD with a forced expiratory volume in 1 second (FEV₁) of 49% predicted (moderate, GOLD [Gold Initiative on Obstructive Lung Disease] stage II) but was asymptomatic from a pulmonary standpoint. His hypertension was controlled. No evidence was found of concurrent comorbid conditions such as coronary heart disease, stroke, or obstructive sleep apnea, which could preclude or complicate procedures performed under moderate sedation or general anesthesia.

Support System

Results of studies show that appropriate lung cancer care is affected by sociodemographic factors. In patients with early-stage NSCLC, comorbidities, older age, and low educational level all were found to be associated with a lower probability of receiving surgery. These same factors, as well as being unmarried, were associated with a higher probability of receiving other noncurative care only. Comorbidities and low educational level did not seem to affect the more effective patterns of care in the advanced-stage group. When initial patterns of care were controlled for in the early-stage group, age older than 75 years and being unmarried were negative prognostic factors, and survival was completely independent of educational level. Among patients at an advanced stage of disease, only comorbidities had a negative impact on survival.¹⁰ Another study suggested that socioeconomically disadvantaged groups with NSCLC received less-intensive care, but low education remained an independent predictor of poor survival only in women with early-stage disease. The exact underlying mechanisms of these social inequalities are unknown, but differences in access to care, comorbidities, and lifestyle factors all may contribute.¹¹ Early discussions about quality of life and symptom concerns are justifiable in that patients will commonly develop pain, dyspnea, cough, and fatigue during the course of their illness.¹² These conversations took place repeatedly during initial and subsequent clinical encounters with our patient.

Patient Preferences and Expectations

Evidence indicates that discrepancies exist between preferred and actual roles in decision making for patients suffering from a variety of cancers, including NSCLC.¹³ When interviewed, patients wanted a more shared or an active role, across all cancer types, patients wanted more participation than what actually occurred. Role preferences are dynamic, however, and vary greatly during decision making, requiring repeated clinical assessments to help providers meet patients’ expectations and improve patient satisfaction with treatment decisions. At the time of our first encounter, this patient had clearly expressed a desire for diagnosis and treatment. He was prepared to consider all available diagnostic options for staging, including conventional TBNA, EBUS-guided TBNA, endoscopic ultrasound (EUS)-guided fine-needle aspiration (FNA), mediastinoscopy, and video-assisted thoracic surgery (VATS).

Indications

For this patient with discrete mediastinal lymph node enlargement and no distant metastases, invasive confirmation of the radiographic stage was recommended, regardless of results of PET scan findings for mediastinal nodes.⁴ The status of the mediastinum is, in fact, the most crucial factor in selecting an optimal treatment strategy. It is also fundamental for estimating prognosis. Patients with tumors in clinical stage III are a heterogeneous group, in whom the extent of LN involvement before and after induction therapy determines outcome.¹⁴ If our patient’s nodal station 4R is positive for malignancy, his clinical stage is III A-N2. In this subgroup of patients, induction chemotherapy, combined with surgery and/or radiotherapy, has proved effective. If surgical combined-modality treatment is being considered, complete resection is essential in the potential for cure.

If complete resection is unlikely, a nonsurgical multimodality approach is preferred. Therefore, an accurate preoperative and postinduction nodal evaluation is mandatory. For both primary staging and restaging, not every technique is available in every center, and different techniques are advocated in different countries and in different institutions. In many situations, however, an invasive test can provide confirmation of the diagnosis and confirmation of the stage at the same time, thus illustrating the importance of not immediately pursuing a diagnostic test in patients, but rather first thinking through the presumptive diagnosis, the presumptive stage, and the need for additional confirmatory staging tests.

Although the sensitivity of various invasive mediastinal staging tests in clinical N2 and N3 patients appears to be similar, a strict comparison is not justified, because patients undergoing these procedures are not comparable owing to differences in how they are selected for a particular procedure (e.g., the location of the nodes). The primary issue, therefore, is the variability in false-negative rates. If a needle aspiration (NA) technique is chosen, it must be remembered that a negative result is not completely reliable, but at the same time, NA may well be a good first choice because it is less invasive than mediastinoscopy. Sampling station 4R nodes in our patient is essential because the incidence of occult N2* disease in NSCLC patients with negative mediastinal uptake of 18FDG on PET-CT is considered to be approximately 16% (25 of 153 patients). The highest incidence of occult N2 involvement is seen in station 7 (subcarinal) (16 of 25 patients [64%]) followed by station 4 (lower paratracheal) (7 of 25 patients [28%]).¹⁵ Mediastinal LN size is directly related to metastatic involvement. One meta-analysis showed that the prevalence of malignant involvement ranged from 9% to 42% for nodes measuring 10 to 15 mm, from 19% to 75% for nodes measuring 16 to 20 mm, and from 27% to 100% for nodes measuring larger than 20 mm. The probability of nodal metastasis in mediastinal nodes measuring 10 to 15 mm in the short axis on CT is 29% and is consistently twofold higher in larger ones.⁸

Some physicians routinely perform a flexible bronchoscopy before considering curative resection, because evidence indicates that 8% of patients with a small noncalcified pulmonary nodule up to 3 cm in diameter might have an endobronchial lesion that could potentially alter staging and management.¹⁶ Flexible bronchoscopy would allow complete airway inspection to ascertain the absence of endobronchial lesions and to sample the 4R lymph node at the same setting. Although this could be done with the use of conventional TBNA, EBUS-TBNA may increase diagnostic yield at the level 4R compared with conventional sampling techniques (71% to 94% vs. 66%).^17,18

Contraindications

No absolute contraindications to bronchoscopy with EBUS-TBNA were noted. This procedure, however, is often performed under general anesthesia, which in our patient with COPD and baseline CO₂ retention could predispose to bronchospasm, air trapping, and prolonged ventilatory depression, especially if opioids are used.¹⁹ These concerns would have to be addressed with the anesthesiologist before the procedure is undertaken.

Expected Results

We usually perform EBUS-TBNA under general anesthesia with a large endotracheal tube (8.5 or 9 mm) to accommodate the EBUS-TBNA scope (outside scope diameter is 6.2 mm). The EBUS scope would be introduced through the tube once the bite block is secured in place around the tube. Laryngeal mask airway and moderate or deep sedation are feasible alternatives to intubation and general anesthesia. Regardless, in case of general anesthesia, after the airway is secured, complete airway examination would be performed. Station 4R sampling should be accompanied by a complete sonographic mediastinal and hilar nodal assessment because the tumor may be upstaged to N3 (stage IIIB), in case EBUS identifies contralateral lymph nodes from which aspirates are positive for malignancy.²⁰ The lymph nodes would be systematically visualized by starting with N1 lymph nodes followed by N2 nodes and finally N3 nodes. EBUS-TBNA is then performed first from N3 nodes, followed by N2 nodes, and, if necessary, N1 nodes. If N3 nodes were found to be positive for malignancy on rapid on-site cytologic evaluation, the procedure could be terminated. For this patient, if complete EBUS mediastinal and hilar evaluation reveals no other nodes, then only the CT-documented 4R node should be sampled. In one small study, the diagnostic rate of EBUS-TBNA for station 4R was only slightly better than that of conventional TBNA (71% vs. 66%).²¹ However, larger studies have shown high diagnostic yields of 86% to 94% with EBUS for station 4R.^17,22

Team Experience

A team familiar with the techniques and the equipment is necessary because of the particularities of TBNA, EBUS, and EBUS-TBNA.²³ Because of its availability at our institution, we routinely use on-site cytology to determine the adequacy of specimens obtained and to identify an immediate diagnosis. Experience with sampling this station is high because station 4R and the subcarinal nodes (station 7) are the most commonly biopsied nodes during conventional TBNA and EBUS-TBNA.^17,18

Risk-Benefit Analysis

Although EBUS-TBNA has a high yield and is a safe procedure,²⁴ rare clinically significant complications include pericarditis and pneumothorax requiring chest tube drainage.²⁵ Even when it is performed using moderate sedation, patient satisfaction is high, and no complications occur that might compromise diagnostic yield.²⁶

Diagnostic Alternatives

For patients with discrete mediastinal lymph node enlargement and no distant metastases, as in this case, another nonbronchoscopic NA technique (i.e., endoscopic ultrasound-guided needle aspiration [EUS-NA] or transthoracic needle aspiration [TTNA]), mediastinoscopy, or VATS could be performed instead of EBUS or conventional TBNA. In choosing an invasive staging test, the availability of different procedures has to be considered, because some of these procedures require specialized experience and skill, and those who perform these procedures only occasionally may be unable to achieve performance characteristics comparable with those published in studies performed at high-volume institutions. The location of the suspicious node is important as well, because nodes in one location may be accessible only through a particular approach. Advantages and disadvantages of each of the procedural modalities should be discussed with the patient and his spouse.

Cost-Effectiveness

Based on a review of 12 studies performed between 2004 and 2008, it seems that the sensitivity, specificity, false positivity (FP), and false negativity (FN) of EBUS-TBNA for mediastinal lymph node staging are 93%, 100%, 0%, and 9%, which may be better than findings for cervical mediastinoscopy, for which these values are 78%, 100%, 0%, and 11%, respectively. Sensitivity and false-negative rates for videomediastinoscopy, however, are 90% and 7%, respectively; these rates are similar to those reported for EBUS-TBNA. Three published cost-effectiveness analyses performed in the United States, the United Kingdom, and Singapore, based on mathematical modeling, show that EBUS may be less expensive than mediastinoscopy.²⁸

Informed Consent

After they had been advised of all of diagnostic alternatives, the patient and his wife elected to proceed with EBUS-TBNA under general anesthesia. They were informed of the potential for failure to obtain a definitive diagnosis despite successful LN visualization and sampling. Feedback communication techniques were used to confirm patient understanding of the disease. Both the patient and his spouse were able to accurately describe the proposed procedure, alternatives, and potential complications.

Anesthesia and Perioperative Care

EBUS-TBNA can be performed under conscious (moderate), deep sedation or under general anesthesia using laryngeal mask airway (LMA) or endotracheal tubes. Moderate sedation offers the advantage of performing the procedure in the bronchoscopy suite and may result in better cost savings/safety ratios when compared with general anesthesia.²⁹ Because of the relatively small size of the node in this patient with COPD, moderate sedation could have resulted in excessive cough and respiratory movements, significant artifacts, and suboptimal ultrasound image acquisition. The procedure was performed with the patient under general anesthesia in the operating room. LMA No. 4 or 4.5 is preferred by some operators to establish a secure airway,³⁰ but LMAs may not be appropriate in severely obese patients or in patients with severe untreated gastroesophageal reflux.³¹