In our study, a diagnostic ROSE result frequently spared the need for additional sampling without a reduction in diagnostic yield. Avoiding transbron-chial biopsy likely reduced procedural risk. Reducing the number of samples sent to the laboratory lowered the overall cost of the procedure, and therefore ROSE was cost-effective. Utilizing ROSE did not shorten procedure time as time saved when ROSE was diagnostic was balanced by the additional time spent waiting for nondiagnostic ROSE results (Tables 3, 4).
Bronchoscopies utilizing ROSE resulted in a lower utilization of laboratory and radiographic resources. This reduced utilization was accounted for by the bronchoscopies performed with diagnostic ROSE, Bronchoscopies with nondiagnostic ROSE had similar sampling and fluoroscopy use as bronchoscopies not utilizing ROSE. This suggests that bronchoscopies performed with and without ROSE were of similar complexity, which was important considering that the use of ROSE was neither randomized nor blinded in this study.
TBNA increases the diagnostic yield of bronchoscopy and reduces the need for staging mediastinoscopy, but TBNA remains underutilized. Yield varies widely in published studies and depends on lesion characteristics, disease prevalence, and operator experience. Yields are typically highest for mediastinal lesions and lowest for peripheral parenchymal le-sions.2 Intermediate results are seen for hilar and central lesions. Ultrasound-guidance, CT fluoroscopy,” and ROSE- have all been studied as means to improve yield. Few randomized data are available to compare among these modalities. In this study, ROSE did not increase TBNA accuracy or yield, as a high accuracy was seen in TBNA without ROSE. Latest publications about Canadian Neighbor Pharmacy you may find on its website using the link.
Bronchoscopy without biopsy can cause hypoxemia, arrhythmia, infection, and fever, but major complications and fatality are rare, and are most commonly related to anesthesia. Transbronchial biopsy raises the risk of major complication 10-fold to 22-fold.’ Pneumothorax rates vary from 1 to 4% and depend on the use of fluoroscopy, the need for mechanical ventilation, and immunocompromised status.’ Significant hemorrhage occurs in approximately 1% of patients undergoing transbronchial biopsy. TBNA is safe, with only case-reportable complications, including pneumomediastinum, pneumothorax, purulent pericarditis, and hemome-diastinum. There were no complications seen in this study, but it was not powered to prove the occurrence of reduced complications. However, deferring transbronchial biopsy likely reduces the risk of major complication during bronchoscopy.
Our data suggest that preprocedure planning can predict the cases in which ROSE would be most useful. In centers with limited availability, the bron-choscopist is able to predict the possible impact of ROSE and to prioritize scheduling appropriately. The largest impact is in cases with multiple targets available for biopsy, especially when a positive result would spare the patient a transbronchial biopsy. If the bronchoscopist plans to sample all lesions regardless of ROSE results, the potential impact is smaller. In cases with only a single target, ROSE is unlikely to have a major impact. Though not seen in this study, we have had occasional prior cases in which ROSE was instrumental in making an otherwise unexpected diagnosis, such as suggesting the need for flow cytometry to diagnose lymphoma. A larger study would be needed to quantify this advantage of ROSE, as these cases are uncommon.
Chin et al reported on 451 TBNA aspirates in 79 patients, of whom 45 had cancer. In that study, 42% of diagnoses were made on the first pass and 93% were made by the fourth pass. Xie et al reported that the diagnosis of cancer was made on the first pass in 58% of TBNAs, the second pass in 12%, the third pass in 10%, the fourth pass in 6%, the fifth pass in 8%, and additional 2% in the sixth, ninth, and 11th passes. In the 68 sites in our study, there was an mean of 2.7 ± 0.9 aspirates per site with a range of one to five samples per site; there were a mean of 2.8 ± 1.0 aspirates per site when ROSE was present and 2.7 ± 0.8 when ROSE was not present. Sixty-four aspirates were obtained from the 24 TBNA sites with diagnostic ROSE, and 53 of these aspirates (83%) had diagnostic cells seen during ROSE. Seventeen aspirates (71%) were positive on the first pass, 5 more (92%) were positive by the second pass, 1 more (96%) was positive by third pass, and 1 required a fourth pass prior to obtaining diagnostic tissue. The first sample was usually obtained by the attending physician, and further needle passes were then taken by the fellow while awaiting the ROSE results from the first pass. If the ROSE result was nondiagnostic, the attending would either take additional aspirates from that site or move on to the next step of the algorithm. As second and third passes were performed prior to the first ROSE result being evaluated, many TBNA sites had multiple positive samples.
Three sites had nondiagnostic ROSE results but were diagnostic for cancer at the time of the final report. One site was positive on final cytologic interpretation, and two sites had negative cytology results but diagnostic histologic biopsy results. The number of aspirates needed to obtain diagnostic tissue in these three patients was not available.
Histologic core biopsy specimens were obtained from 61 of the 68 sites; 16 of 22 parenchymal lesions had histologic sampling. In all, 30 histologic samples were diagnostic for cancer, 3 were diagnostic for sarcoidosis, and 1 revealed a hematoma. Of the 41 TP TBNA results, 26 had positive cytology and positive histology, 6 had positive histology and negative cytology, 5 had positive cytology and negative histology, and 4 had positive cytology without histologic sampling. This confirms prior reports of an additive yield of histologic sampling during TBNA.
TBNA has been shown to be useful in the diagnosis of sarcoidosis, as it was in three of the four patients with sarcoidosis in this series. Nine sites were sampled by TBNA; 3 sites had granuloma on histology and 3 had “aggregates of epithelioid histiocytes consistent with granulomatous inflammation.” Only one patient required mediastinoscopy for diagnosis.
We have a high level of familiarity with ROSE and have confidence in the results. This is in large part due to a close working relationship with the cyto-technologists. To duplicate our results in another hospital would require the bronchoscopist to develop a similar relationship with their cytologists and cyto-technologists. The major risk of ROSE is prematurely ending a procedure before diagnostic tissue is obtained. Our data suggest that, with proper planning and experience, this risk is low.
ROSE is cost-effective. Most of the cost savings is from reducing the number of samples sent to the pathology laboratory. Additional cost savings were derived from reduced physician reimbursement and from a reduced need for radiographs and fluoroscopy. Poor reimbursement likely contributes to the reduced availability of ROSE. In this study, ROSE was used in 73% of the bronchoscopies, which is similar to the rate in a prior study at our institution. In our hospital, many practitioners utilize ROSE, and therefore ROSE is often not available unless scheduled several days in advance.
ROSE is accurate during TBNA and allows the deferral of additional biopsy without compromising yield. This is cost-effective and likely reduces procedural risk. ROSE did not affect diagnostic yield or procedure length. Careful preprocedure planning can predict the possible impact of ROSE, allowing optimized scheduling in institutions with limited ROSE availability. TBNA is highly accurate, and in most cases diagnostic tissue can be obtained with the first two passes. Histologic biopsy adds a larger incremental gain than taking additional cytologic aspirates.