Development of a LAG-3 immunohistochemistry assay for melanoma

AIMS

A robust immunohistochemistry (IHC) assay was developed to detect lymphocyte-activation gene 3 (LAG-3) expression by immune cells (ICs) in tumour tissues. LAG-3 is an immuno-oncology target with demonstrable clinical benefit, and there is a need for a standardised, well-characterised assay to measure its expression. This study aims to describe LAG-3 scoring criteria and present the specificity, sensitivity, analytical precision and reproducibility of this assay.

METHODS

The specificity of the assay was investigated by antigen competition and with LAG3 knockout cell lines. A melanin pigment removal procedure was implemented to prevent melanin interference in IHC interpretation. Formalin-fixed paraffin-embedded (FFPE) human melanoma samples with a range of LAG-3 expression levels were used to assess the sensitivity and analytical precision of the assay with a ≥1% cut-off to determine LAG-3 positivity. Interobserver and intraobserver reproducibility were evaluated with 60 samples in intralaboratory studies and 70 samples in interlaboratory studies.

RESULTS

The LAG-3 IHC method demonstrated performance suitable for analysis of LAG-3 IC expression in clinical melanoma samples. The pretreatment step effectively removed melanin pigment that could interfere with interpretation. LAG-3 antigen competition and analysis of LAG3 knockout cell lines indicated that the 17B4 antibody clone binds specifically to LAG-3. The intrarun repeatability, interday, interinstrument, interoperator and inter-reagent lot reproducibility demonstrated a high scoring concordance (>95%). The interobserver and intraobserver reproducibility and overall interlaboratory and intralaboratory reproducibility also showed high scoring concordance (>90%).

CONCLUSIONS

We have demonstrated that the assay reliably assesses LAG-3 expression in FFPE human melanoma samples by IHC.

Endoscopic ultrasound guided fine needle aspiration of malignant pancreatic lesions

PURPOSE

To evaluate the accuracy, safety, and clinical utility of endoscopic ultrasound (EUS) guided fine needle aspiration (FNA) of pancreatic masses.

METHODS

Forty-seven patients were referred for EUS with a pancreatic mass and suspicion of pancreatic cancer based upon painless obstructive jaundice, epigastric abdominal pain plus weight loss/anorexia, or idiopathic pancreatitis. All patients underwent EUS with both radial (Olympus UM20) and linear array (Pentax FG32-UA) systems. After TNM staging by EUS, ultrasound directed FNA of the pancreatic mass was performed using a 23 gauge, 4 cm long needle.

RESULTS

EUS-guided FNA was performed in all 47 patients. Results: successful targeting = 100%, adequate cellularity = 100%,

FINDINGS

adeno Ca = 25, squamous cell Ca = 1, lymphoma = 1, poorly differentiated Ca= 1, atypical cytology or suspicious for carcinoma = 9, no malignant cells = 10. The sensitivity, specificity, positive predictive value and negative predictive value of EUS-guided pancreatic FNA for the diagnosis of malignancy was 64%, 100%, 100% and 16% respectively.

CONCLUSIONS

EUS with FNA is useful for detection of malignancy in a pancreatic mass. The procedure appears to have a complication rate of 2%. Impact of this technique on clinical management of patients needs further evaluation.

Endoscopic ultrasound with fine-needle aspiration in the diagnosis and staging of lung cancer

BACKGROUND

Esophageal endoscopic ultrasonographic (EUS) guidance for fine-needle aspiration (FNA) of mediastinal lymph nodes has been introduced only recently. The utility of EUS/FNA in diagnosing and staging bronchogenic carcinoma is unknown.

METHODS

After a thoracic computed tomographic scan, 27 patients with known or suspected lung cancer underwent EUS. Accessible abnormal mediastinal lymph nodes were aspirated under EUS guidance. Patients with positive cytologic studies did not undergo further testing, whereas the remaining patients underwent mediastinal exploration. The sensitivity, specificity, accuracy, positive predictive value, and negative predictive value were calculated for both chest computed tomography and EUS/FNA:

RESULTS

Twenty-two of 27 patients had mediastinal adenopathy by computed tomography scan. Sixteen patients had positive findings on EUS, 15 with positive FNA (10 non-small cell lung cancer; 5 small cell lung cancer) and 1 with T4 status. Fourteen patients with positive FNA had lymph nodes sampled at level 5, level 7, or both. Of 11 patients with negative EUS/FNA, 2 had positive findings at operation (sensitivity 89%). The diagnosis of lung cancer was established in 7 patients.

CONCLUSIONS

The results showed that EUS/FNA improves the accuracy of computed tomographic scan in the staging of lung cancer. By accessing lymph nodes at levels 5 and 7, EUS/FNA complements mediastinoscopy and is considered the staging modality of choice in these regions. Positive EUS/FNA can obviate the need for further invasive staging.