in supraclavicular lymph nodes (N3). A positive sample means stage IV disease to be managed with systemic treatment while a negative biopsy will lower the stage designation and may direct the patient to potentially curative surgery.

As low volume tissue and cytologic material are provided by the techniques used in sampling distant metastases, obtaining suf cient material for diagnosis, immunohistochemistry, and mutational analysis should be the aim. Suspected metastatic sites should be sampled rst rather than the suspected primary lesion. If there are multiple sites of metastases (M1c), the site that can be sampled by the safest or easiest approach is preferred for pathologic con rmation [61].

Depending on the local expertise, CTor US-guided percutaneous FNA/CNB, or EUS-­ FNA can be used to sample suspected metastatic lesions in liver, or adrenal gland. Brain biopsy for suspected metastatic lesions is rarely performed when there are neurological symptoms/signs and a focal abnormality on imaging. As it is a high-­ risk procedure, potential bene ts and harms of the procedure in the individual patient should be weighed by neurosurgical consultation. For suspected bone metastasis, CT-guided percutaneous CNB can be performed but practices may show differences according to the institution and physician. Owing to the need for decalci cation, molecular analysis of bone biopsy specimens are usually suboptimal. Thus, if present, biopsy of another metastatic site is preferred. Supraclavicular or scalene lymph node metastasis should be the rst to sample among other metastases. Percutaneous FNA or CNB is used to identify lung cancer involvement of these lymph nodes. USor CT-guidance can be used during FNA or CNB if required [64].

Sputum Cytology: A Noninvasive

Method

As a noninvasive method, sputum cytology can be used in patients with suspected lung cancer but not able or willing to go under other diagnostic procedures. However, it does not provide staging information and ideal specimens for

immunohistochemical or molecular studies. It has a pooled sensitivity of 66% (42–97%) in the diagnosis of NSCLC. The sensitivity is higher in central and large tumors than in peripheral and small tumors. Its utility is quite limited as every patient cannot produce sputum, a negative test does not exclude lung cancer and a second diagnostic procedure will be required, and additional testing for staging will be needed even if sputum cytology is positive in the presence of suspected mediastinal or distant metastasis. Thus, sputum cytology can be useful in patients with multiple comorbidities, contraindications to invasive procedures, and metastatic disease and can permit palliative or other treatment avoiding the risk of biopsy. If sputum cytology is positive, chest CT to search for the tumor, and further diagnostic and staging procedures targeted at the primary or metastatic lesions will be required. In cases with negative chest imaging, white light bronchoscopy with or without autofuorescence bronchoscopy will be indicated [61].

Tissue Acquisition, Handling and Processing

Implications of Tissue Acquisition

The above-mentioned techniques for tissue acquisition can help the clinician in diagnosing and choosing the appropriate treatment for the lung cancer patient: primary resection, neoadjuvant­ chemotherapy and/or radiation, or palliative chemotherapy and/or radiotherapy. Combining diagnostic or staging techniques strategically provides more successful yields and better outcomes in the management, and may possibly be more cost-effective (e.g. combination of EBUS-­TBNA and EUS-FNA, and if negative, followed by mediastinoscopy) [65]. However, there had been signi cant underuse of multimodality diagnosis and/or staging but it has been increasing since 1998. Using a multimodality approach improves survival irrespective of the lung cancer stage. Furthermore, approximately 15–40% of the patients will not be chosen for a therapy given with curative intent if only imaging

techniques are used for staging. Therefore, in a case with suspected lung cancer, ndings of primary and metastatic tumor on imaging should be corroborated by cytological or histological studies. Unfortunately, it is not uncommon to see inadequate pathological evaluation particularly for staging (e.g. not identifying nodal disease). The most important consequence of such an approach will most probably be reduction in survival related to lung cancer [2, 66].

Guideline Recommendations for Tissue Acquisition in Mediastinal Staging

The guidelines of the several major international societies such as ACCP and ESTS share similar recommendations regarding indications and procedures for diagnosing and staging lung cancer [1, 33].

Sampling a single enlarged lymph node or random sampling is considered inadequate in surgical staging. This approach against random sampling can be extrapolated to staging by minimally invasive procedures [2]. Although recommendations for appropriate staging include stations 2R, 2L, 4R, 4L and 7, sampling of nodes smaller than 5 mm is not easy and will more likely provide insuf cient amounts of tissue for diagnosis and staging. Sampling should also be performed from the clinically suspicious lymph nodes (with a short axis diameter of 1 cm or larger, or FDG-avid). ESTS guidelines and two other guidelines from the UK and Canada recommend a systematic assessment of at least three mediastinal node stations including station 7 (subcarinal) [33, 67, 68].

Diferences in Sample Acquisition

and Processing: Cytology Versus

Histology

Having an appreciation of cytological/histological processing and evaluation of small biopsies obtained by minimally invasive procedures is required. For establishing a diagnosis of malig-

nancy, subclassifying cancer reliably by using immunohistochemical stains, and for molecular analysis to determine targetable driver mutations, the obtained cytologic or histologic (small biopsy) specimens should be suf cient in quality and amount. Over the past 2 decades, the information obtained from these small specimens has increased signi cantly.

Specimens obtained by minimally invasive means provide limited cellular material. Endobronchial biopsy, TBB, and transthoracic CNB can obtain samples with tissue architecture that is useful in differentiating invasive carcinoma from in-situ or lepidic patterns. However, limitations in sampling can be a problem for these specimens. Cytological aspirates obtained by procedures such as EBUS-TBNA or EUS-­FNA usually do not have tissue architecture but fragments of tissue can occasionally be obtained and recognized on smears or cell block preparations. As cytological features of malignancy are usually not dif cult to recognize, diagnosing malignancy on cytological specimens is rarely an issue. Whereas biopsy specimens are almost always formalinxed and paraf n-embedded, processing and evaluation of cytologic specimens can be performed through several ways: cell block preparations, alcoholxed or air-dried direct smears or touch prints of tissue biopsies, and alcohol-xed liquid-based concentration methods. Cell block preparation collects the cellular material into a pellet which is formalin-­ xed and paraf n- embedded and thus, creates a specimen similar to a tissue biopsy. From this tissue-biopsy-like specimen, multiple serial slides can be cut and used for immunohistochemical stains and molecular testing. Indeed, the difference between a small biopsy specimen and a cytology specimen, particularly a cell block preparation, is not so clear anymore as both types of specimens can provide speci c histologic diagnosis and can be used for molecular testing [2].

According to the 2015 WHO classi cation of lung tumors [69] and the 2011 IASLC/ATS/ERS classi cation of lung carcinomas on small biopsy/ cytology specimens [70], a panel of immunostains should be performed judiciously and in a focused manner to preserve cellular material for down-