Secondary Logo

Comparison of Transbronchial and Cryobiopsies in Evaluation of Diffuse Parenchymal Lung Disease

Ramaswamy, Anuradha MD*; Homer, Robert MD; Killam, Jonathan MD; Pisani, Margaret A. MD, MPH*; Murphy, Terrence E. PhD§; Araujo, Katy MPH§; Puchalski, Jonathan MD, MEd*

Journal of Bronchology & Interventional Pulmonology: January 2016 - Volume 23 - Issue 1 - p 14–21
doi: 10.1097/LBR.0000000000000246
Original Investigations
Free
Editor's Choice

Background: Diffuse parenchymal lung diseases (DPLDs) are common. An accurate diagnosis is essential due to differences in etiology, clinicopathologic features, therapeutic options, and prognosis. Transbronchial lung biopsies (TBLBs) are often limited by small specimen size, crush artifact, and other factors. Transbronchial lung cryobiopsies (TBLCs) are under investigation to overcome these limitations.

Methods: We conducted a retrospective study of 56 patients in a single, tertiary-care academic center to compare the yield of both techniques when performed in the same patient. Patients underwent flexible bronchoscopy using moderate sedation with TBLB followed by TBLC in the most radiographically abnormal areas. Clinical data and postprocedural outcomes were reviewed, with a final diagnosis made utilizing a multidisciplinary approach.

Results: The mean age of patients was 60 years and 54% were male. Comorbidities included COPD (14%) and prior malignancy (48%). The number of TBLB specimens ranged from 1 to 10 per patient (mean 4) and size varied from 0.1 to 0.8 cm. The number of TBLC specimens ranged from 1 to 4 per patient (mean 2) and size ranged from 0.4 to 2.6 cm. Both techniques provided the same diagnosis in 26 patients (46%). An additional 11 (20%) patients had a diagnosis established by adding TBLC to TBLB. Compared with TBLB, TBLC had a higher diagnostic yield in patients with hypersensitivity pneumonitis and interstitial lung disease. Only 2 patients required video-assisted thoracoscopic surgery to establish a diagnosis. Complications included pneumothorax (20%) and massive hemoptysis (2%).

Conclusion: TBLC used with TBLB can improve the diagnostic yield of flexible bronchoscopy in patients with DPLD.

*Division of Pulmonary, Critical Care and Sleep Medicine

Departments of Pathology

Radiology

§Program on Aging, Yale University School of Medicine, New Haven, CT

Disclosure: There is no conflict of interest or other disclosures.

Reprints: Jonathan Puchalski, MD, Division of Pulmonary, Critical Care and Sleep Medicine, Yale University School of Medicine, 15 York Street, LCI 100, New Haven, CT 06510 (e-mail: jonathan.puchalski@yale.edu).

Received June 25, 2015

Accepted December 1, 2015

Diffuse parenchymal lung diseases (DPLDs) comprise a heterogenous group of over 200 pulmonary disorders that contribute to diffuse and often patchy involvement of the bilateral lung parenchyma.1 Estimated to occur in up to 15% of the general pulmonary clinic population, DPLDs are characterized by broad differences in etiology, clinicopathologic features, therapeutic options, and prognosis.2 An accurate diagnosis is therefore essential to guide therapy.

A multidisciplinary approach to DPLD is critical. As guidelines suggest, the patients’ clinical presentation, laboratory data, and pulmonary function testing is reviewed in conjunction with a radiologist. In cases where high-resolution computed tomographic scans of the chest are inconclusive, biopsies may be required. Surgical lung biopsies have been long considered the gold standard for diagnosis of DPLD. However, not all patients can safely undergo surgery. Bronchoscopic biopsies, especially cryobiopsies, may fit into the diagnostic algorithm in these scenarios.

Transbronchial forceps have been traditionally used for procuring parenchymal biopsies for diagnostic purposes, despite a lower diagnostic yield for conventional forceps estimated at 29% in 1 study.3 The diagnostic yield is traditionally low in patients with idiopathic pulmonary fibrosis (IPF). Limitations of transbronchial lung biopsies (TBLB) include small specimen size, sampling errors, and artifacts. While surgical lung biopsy remains the gold standard for definitive diagnosis, it cannot be safely performed in many patients due to respiratory impairment and medical comorbidities.

Cryoprobes have been used in the field of bronchoscopy since the 1970s.4 Initially used for therapeutic tumor debulking and later for “cryorecanalization” of airways, cryobiopsies have been increasingly performed over the past decade for obtaining parenchymal specimens.5,6 These biopsies are obtained by utilizing the Joule-Thomson effect, whereby the probe is cooled by release of compressed gas at high flow. The gas, typically carbon dioxide (CO2) or nitric oxide (NO), expands due to the sudden difference in pressure relative to the atmosphere, causing a drop in temperature at the tip of the probe. Adjacent tissue becomes attached to the frozen probe.7 Purported advantages of the biopsy technique include larger specimens with less artifact.8,9 However, the specific etiologies by which cryoprobes routinely improve diagnostic yield are unknown. The aim of this study is to report our experience with transbronchial lung cryobiopsies (TBLC) in the diagnosis of patients with DPLD, and to compare the yield of TBLB and TBLC when performed simultaneously in the same patient.

Back to Top | Article Outline

MATERIALS AND METHODS

We performed a retrospective study on all patients who underwent TBLC at our institution between March 2013 and June 2014. All procedures were performed at Yale-New Haven Hospital, a tertiary care academic medical center. Written informed consent was obtained for bronchoscopy with biopsies from all patients. Approval for the study was obtained from the Yale University Human Investigation Committee (HIC# 1406014160).

Our definition of DPLD included diffuse alveolar filling abnormalities or interstitial disease identified by thoracic CT scans obtained as part of the patient’s clinical evaluation. These included traditionally characterized interstitial lung diseases (ILDs) as well as parenchymal disease felt possibly due to drug-pneumonitis or radiation-pneumonitis, autoimmune disease, and others. Patients were typically referred by hospitalists or other pulmonologists who were guiding the patients’ evaluation. Pertinent clinical data at the time of the procedure included age, sex, medical history, medication use, and smoking status (Table 1). A number of patients had been followed and then referred for biopsy by expert clinicians specializing in ILD. In these cases, the patient often did not have a definitive prior diagnosis but worsening symptoms and radiographs prompted the request for bronchoscopy. Patients underwent flexible bronchoscopy using moderate sedation with fluoroscopy to guide conventional TBLB and TBLC. Biopsies were not performed if a patient’s platelet count was <50,000/μL, the INR was >1.5, or if they were on medications predisposing to an increased risk of bleeding (clopidogrel, warfarin).

TABLE 1

TABLE 1

Back to Top | Article Outline

Procedural Technique

All procedures were performed by a single interventional pulmonologist in a procedure center using conscious sedation with midazolam and fentanyl. Oral anesthesia was accomplished using 1% atomized lidocaine and 2% lidocaine gel. Additional atomized 1% lidocaine was administered with a spray catheter (Olympus PW-6C-1) through the bronchoscope during the procedure. Supplemental oxygen was provided and patients had spontaneous respiration without the support of an artificial airway or assisted ventilation. The nurse and physician monitored the patient throughout the procedure, a technician helped with specimen processing, and anesthesiology was not present.

The bronchoscope (1T180; Olympus, Japan) was introduced orally and an airway inspection was performed. Patients routinely underwent bronchoalveolar lavage (BAL) to assess for infection. Endobronchial biopsies were performed if the mucosa appeared abnormal. Thereafter, fluoroscopy was used to perform parenchymal biopsies. TBLB were performed throughout all lobes in 1 hemithorax for up to 10 visible specimens. These were performed before TBLC to assess for any unexpected significant bleeding caused by performing biopsies. Subsequently, TBLCs were performed in the area of the lung that was most radiographically abnormal, as determined by the bronchoscopist. The cryoprobe (2.4 mm diameter; ERBE, Tubingen, Germany) was advanced through the working channel of the bronchoscope and advanced into the parenchyma similar to TBLB forceps. A 4- to 5-second freeze was performed, after which the bronchoscope and cryoprobe were concurrently withdrawn from the patient. The specimen was removed from the probe into saline and subsequently transferred to formalin. Typically only 1 biopsy was performed, although up to 4 were performed if the specimen was visibly inadequate. The bronchoscope was then readvanced into the airways to assess for and remove blood before completion of the procedure. A postprocedural chest x-ray was routinely performed.

Back to Top | Article Outline

Postprocedural Review

The final clinical diagnosis (Table 2) for each patient was established by a multidisciplinary review of the radiographic diagnosis, the pathologic diagnosis determined by both TBLB and TBLC specimens, and clinical data obtained from the medical records. The patients diagnosed with malignancy in this study did not have a discrete nodule or mass but rather ground-glass opacities, nonspecific nodules, or interstitial changes that would be consistent with lymphangitic spread. For analysis, we separated sarcoidosis, hypersensitivity pneumonitis (HP), organizing pneumonia, malignancy and drug reaction due to their relative numbers, and diagnostic yield noted in this study. The broader general term “ILD” encompasses other types of ILD, such as but not exclusively nonspecific interstitial pneumonia (NSIP) and usual interstitial pneumonia (UIP). In the former group, with the exception of malignancy, steroids are often a mainstay of therapy. In other types of ILD, including fibrotic NSIP and UIP, steroids are often not useful. Thus, this division of subgroups may be helpful clinically.

TABLE 2

TABLE 2

All radiographs were initially read by a number of radiologists. For this study, CT scans were rereviewed by an independent, blinded radiologist. This radiologist was asked to provide a diagnosis based on our subgroups. Although the pathologic diagnosis was initially provided by a number of pathologists, all samples were rereviewed by an independent, blinded pulmonary pathologist. The pathologist was asked to determine whether the ultimate pathologic diagnosis was defined by TBLB specimens alone, TBLC specimens alone, both TBLB and TBLC, or neither TBLB and TBLC. The nondiagnostic specimens, comprising the latter group, were those in which no specific entity was identified using both biopsy techniques. These included normal parenchyma, nonspecific inflammation, or nonspecific fibrosis. The final clinical diagnosis then incorporated the reviews by the radiologist, pathologist, and 2 investigators (A.R., J.P.) who reviewed the patients’ medical records.

Postprocedural complications were reviewed. The presence of a pneumothorax was acutely assessed by postprocedure radiography. Massive hemoptysis was defined as that requiring therapy beyond application of topical iced saline or epinephrine during the procedure. All charts were reviewed 6 months following the bronchoscopy to assess for delayed complications.

Back to Top | Article Outline

RESULTS

A total of 56 patients were identified who fulfilled study criteria. The mean age of patients was 60 years, and 54% of patients were male. Despite cell count, cultures, cytology, and flow cytometry (CD4:CD8 ratio) sent for BAL fluid, results were suggestive of the final clinical diagnosis in 19 cases (34%). The number of TBLB specimens recorded by pathology ranged from 1 to 10 (average 4) and size per biopsy varied from 0.1 to 0.8 cm. The number of TBLC ranged from 1 to 4 per patient (average 2) and the size ranged from 0.4 to 2.6 cm. Our mean of 4 TBLBs obtained per patient is consistent with recommendations for 4 to 6 transbronchial biopsies for optimal yield in patients with DPLDs.10,11

Of the 56 patients, 45 (80%) had a definitive pathologic diagnosis. Both techniques provided the same diagnosis in 26 (46% of all cases). An additional 15 (27%) patients had a diagnosis made by only one of the techniques. Whereas 4/56 (7%) diagnoses were made only by TBLB, an additional 11/56 (20%) diagnoses were established using TBLC. Six patients were diagnosed with infectious processes as a result of bronchoscopy and 4 had nonspecific inflammation. The clinicians caring for the patients determined that all but 2 patients (96%) had clinical diagnoses established by these methods. The other 2 patients (4%) required video-assisted thoracoscopic surgery to establish the definitive diagnosis (UIP/IPF and graft-versus-host disease).

Compared with TBLB, TBLC had a higher diagnostic yield in patients with nonspecific ILD and HP (Fig. 1). Specimens obtained by TBLC were larger and had less crush artifact, similar to previous studies. Complications included pneumothorax in 11/56 patients (20%) and hemoptysis. Given that cryobiopsies were performed immediately after transbronchial biopsies without an intervening chest radiograph, the pneumothorax may have been caused by either biopsy technique. However, this rate is higher than our traditional pneumothorax rate of <2%, suggesting the higher rate was due TBLC. The rate of pneumothorax did not correlate with a prior history of COPD, the number or size of biopsies performed, or the anatomic location of the biopsies. One patient had significant hemoptysis that was managed with iced saline, topical lidocaine with epinephrine, and Fogarty balloon occlusion. No patient required intubation, and there were no procedure-related deaths.

FIGURE 1

FIGURE 1

Back to Top | Article Outline

DISCUSSION

DPLD causes significant morbidity and mortality. When radiographs are insufficient and histology is felt necessary, clinicians are challenged to determine the most appropriate invasive tests. Clinician perceptions are often that TBLB will be insufficient and surgical lung biopsy too risky, thus creating a clinical conundrum. The advent of TBLC has generated enthusiasm as an advance over these techniques, providing a higher diagnostic yield than TBLB with fewer complications than surgery.

In our study, although BAL was routinely performed, it suggested a diagnosis a minority of the time. The reliability of TBLB is felt to be disease specific. Whereas malignancy, sarcoidosis, and posttransplant pathologic processes are often reliably diagnosed with TBLB, the diagnostic yield is lower for drug toxicity, HP, and certain ILDs (UIP and NSIP).12 In contrast, surgical lung biopsies have higher morbidity. Kreider and colleagues reported a 10.7% incidence of hospital readmission, 5.9% incidence of postoperative mechanical ventilation, 4.4% incidence of mortality, and 19.1% overall incidence of complications following video-assisted thoracoscopic surgery for ILD. They performed a meta-analysis and identified a composite 4.5% rate of mortality in 2223 patients. Overall, patients requiring oxygen, preoperative ventilation, and those with pulmonary hypertension had higher complications.13

Previous, mostly retrospective, studies have investigated the use of TBLC for evaluating DPLD, including ILD and the posttransplant population (Table 3). Some investigators use flexible bronchoscopy and moderate sedation, whereas others have used general anesthesia, intubation, or rigid bronchoscopy. Similarly, some studies have used only cryobiopsies and not conventional forceps biopsies, whereas others have not compared forceps to cryobiopsy specimens in the same “per-patient” analysis.

TABLE 3

TABLE 3

Initial experiments determined that activation time and probe size influenced the quantity of tissue obtained by biopsy.22 In general, compared with TBLB, TBLC specimens are larger, contain more alveoli, and have fewer artifacts.15 In a prospective randomized study by Pajares et al,3 the histologic diagnosis was attained significantly more frequently with cryobiopsies than with conventional transbronchial biopsies (74% vs. 34%). Using a multidisciplinary approach to final diagnoses, including radiographic and clinical data, the final diagnostic yield using cryobiopsies was superior to conventional transbronchial biopsies (51% vs. 29%). Surgical biopsies were avoided in at least 50% of cases in this study as a result of TBLC. Similarly, in a study by Hagmeyer and colleagues, 72% of patients were diagnosed using TBLC and did not require surgical lung biopsy. In those who underwent surgical biopsy (n=8), the diagnosis was made in 75%. Following a multidisciplinary review including the surgical biopsies, TBLC showed a strong congruence with the final diagnosis in 78% of patients, an approximate congruence in 19%, and no congruence in 3%.14

Our study contributes to the literature by emphasizing the complementary role of TBLB and TBLC performed using flexible bronchoscopy and moderate sedation. By performing both techniques in the same patient, we attempt to identify which patients may benefit from the use of the combined techniques. Furthermore, by including patients with idiopathic, exposure-related, connective-tissue disease–related, and other types of DPLD, we incorporate many disease processes not previously investigated with the TBLC technique. Finally, we performed an independent review of all radiographs and pathologic specimens while incorporating clinical information to establish a final diagnosis based on multidisciplinary investigation.

Although not powered to determine a statistical significance for individual disease processes, our study suggests that TBLC offers an improved diagnostic yield compared with conventional TBLB for patients with suspected ILD and granulomatous diseases such as HP and sarcoidosis. These results are similar to those obtained by Babiak et al,6 who investigated many types of diffuse parenchymal disease, and those investigating ILD as shown in Table 3. In contrast, the diagnostic yield of TBLC in patients with diffuse malignancy was not different than TBLB in our investigation. Cases of organizing pneumonia or drug-induced pneumonitis were often diagnosed by both techniques, although the additive process improved the overall yield.

As noted previously, many patients in our study were referred by ILD subspecialists for bronchoscopy. A prior diagnosis of ILD was suggested in the history by 16 patients (28% of the patient population in this study). Most patients were now referred for progressive radiographic or clinical symptoms that raised questions about their previous diagnosis. Given that many patients with ILD are initially classified by clinical and radiographic findings without biopsies, and that many diseases are progressive by nature, our study demonstrates the potential role of a nonsurgical biopsy technique in this scenario.

Our complication rate is similar to that published in the literature. The most troublesome complication was pneumothorax given its rate is significantly higher than that historically obtained by TBLB. There are several explanations, including the larger samples obtained and learning curve of the procedure. Whether a shorter freeze and thaw time or smaller probe would lessen this incidence requires clarification in the future. We experienced bleeding that grossly appeared more frequent and potentially more significant than that due to TBLB. One patient experienced significant bleeding requiring Fogarty balloon occlusion. However, no other patients required any intervention beyond iced saline and topical lidocaine with epinephrine. No patients required intubation as a result of hemoptysis. That being said, additional investigation is required to determine whether this procedure is safe when performed by large numbers of bronchoscopists, including those not trained in interventional pulmonology.

As noted previously and demonstrated in Table 3, the approach to performing TBLC varies, including the setting (bronchoscopy arena vs. operating room), sedation (moderate sedation and a cooperative patient vs. a deeply sedated patient), method of bronchoscopy (flexible vs. rigid), and measures to treat complications such as hemoptysis. Although this study did not compare these issues, it was performed using a simple, standard bronchoscopic approach that is easy to replicate. Flexible bronchoscopy performed in a bronchoscopy suite with moderate sedation (fentanyl and midazolam) is less expensive than that performed in an operating room with anesthesiology. Depending on the institution, this approach may be beneficial when scheduling constraints are prominent in the OR. Disadvantages to this approach focus mostly on the risk of complications, namely massive hemoptysis. Although some authors use rigid bronchoscopy and/or preemptively insert endobronchial blockers, we approached TBLC as any other diagnostic procedure. We had nursing, standard monitoring, and equipment available if necessary (intubation equipment, endobronchial blockers, and Fogarty balloons). We found that we did not require additional preparation.

Limitations of this study include its retrospective design, performance in a single institution, number of patients, and number of operators. The use of TBLC was initiated in our institution at the suggestion of pathology and this study reports our initial experience. We did not compare pathologic results to a surgical gold standard and thus base the diagnostic yield on the new technique (TBLC) compared with our conventional method of biopsy (TBLB). Furthermore, the number of specimens obtained was not consistent for every patient and the location of biopsies was based on the bronchoscopists’ interpretation of radiographic disease severity rather than biopsies consistently being performed in a defined lobe or subsegment. Although we performed biopsies in the same general area with both techniques, we cannot guarantee we were in the exact same area. When only 1 sample was obtained by TBLC, the pathologic diagnosis was undoubtedly limited to the location of the biopsy and did not necessarily reflect the diffuse process. Despite these limitations, our institutional practice has changed with the availability of TBLC. Our algorithm for diagnosing DPLD typically includes a multidisciplinary conference. When the clinical evaluation and imaging fail to provide a diagnosis that strongly supports a particular disease process, bronchoscopy with BAL, TBLB, and TBLC is typically performed. Thoracoscopic biopsies are then considered if the patient is a surgical candidate and the diagnosis remains in question. As our experience continues to increase, we hope to answer the many questions that exist regarding TBLC. Future studies need to define patients most likely to benefit, the technical aspects most likely to improve yield while minimizing complications, and the impact of this algorithm on the clinical management of patients with DPLD.

Back to Top | Article Outline

CONCLUSIONS

Patients with DPLD represent a diagnostic challenge. TBLBs are often limited by small size and artifact, whereas the morbidity of surgical procedures makes clinicians hesitant to pursue this approach in many patients. We believe that further refinement in the use of TBLC may improve the yield of flexible bronchoscopy in DPLDs, thus improving clinical care and avoiding potentially more invasive surgical procedures. Following a multidisciplinary review of patients’ clinical and radiographic presentation, we believe that TBLC may be additive to conventional TBLB, but additional studies are required to determine which patients benefit the most from this approach.

Back to Top | Article Outline

REFERENCES

1. Maher TM. A clinical approach to diffuse parenchymal lung disease. Immunol Allergy Clin North Am. 2012;32:453–472.
2. Maher TM. Diffuse parenchymal lung disease. Medicine (Baltimore). 2012;40:314–321.
3. Pajares V, Puzo C, Castillo D, et al.. Diagnostic yield of transbronchial cryobiopsy in interstitial lung disease: a randomized trial. Respirology. 2014;19:900–906.
4. Poletti V, Casoni GL, Gurioli C, et al.. Lung cryobiopsies: a paradigm shift in diagnostic bronchoscopy? Respirology. 2014;19:645–654.
5. Hetzel J, Hetzel M, Hasel C, et al.. Old meets modern: the use of traditional cryoprobes in the age of molecular biology. Respiration. 2008;76:193–197.
6. Babiak A, Hetzel J, Krishna G, et al.. Transbronchial cryobiopsy: a new tool for lung biopsies. Respiration. 2009;78:203–208.
7. Poletti V, Benzaquen S. Transbronchial cryobiopsy in diffuse parenchymal lung disease. A new star in the horizon. Sarcoidosis Vasc Diffuse Lung Dis. 2014;31:178–181.
8. Fruchter O, Fridel L, El Raouf BA, et al.. Histological diagnosis of interstitial lung diseases by cryo-transbronchial biopsy. Respirology. 2014;19:683–688.
9. Sunna RErnst A, Herth FJF. Chapter 33: cryotherapy and cryodebridement. Principles and Practice of Interventional Pulmonology. New York: Springer; 2013.
10. British Thoracic Society Bronchoscopy Guidelines Committee, a Subcommittee of Standards of Care Committee of British Thoracic Society. British Thoracic Society guidelines on diagnostic flexible bronchoscopy. Thorax. 2001;56(suppl 1):i1–i21.
11. British Thoracic Society. The diagnosis, assessment and treatment of diffuse parenchymal lung disease in adults. Thorax. 1999;54(suppl 1):S1–30.
12. Gal AA. Use and abuse of lung biopsy. Adv Anat Pathol. 2005;12:195–202.
13. Kreider ME, Hansen-Flaschen J, Ahmad NN, et al.. Complications of video-assisted thoracoscopic lung biopsy in patients with interstitial lung disease. Ann Thorac Surg. 2007;83:1140–1144.
14. Hagmeyer L, Theegarten D, Wohlschlager J, et al.. The role of transbronchial cryobiopsy and surgical lung biopsy in the diagnostic algorithm of interstitial lung disease. Clin Respir J. January 26, 2015. DOI: 10.1111/crj.12261. E-published ahead of print.
15. Griff S, Schonfeld N, Ammenwerth W, et al.. Diagnostic yield of transbronchial cryobiopsy in non-neoplastic lung disease: a retrospective case series. BMC Pulm Med. 2014;14:171.
16. Hernandez-Gonzalez F, Lucena CM, Ramirez J. Cryobiopsy in the diagnosis of diffuse interstitial lung disease: yield and cost-effectiveness analysis. Arch Bronconeumol. 2015;51:261–267.
17. Casoni GL, Tomassetti S, Cavazza A, et al.. Transbronchial lung cryobiopsy in the diagnosis of fibrotic interstitial lung diseases. PLoS One. 2014;9:e86716.
18. Fruchter O, Fridel L, Rosengarten D, et al.. Transbronchial cryobiopsy in immunocompromised patients with pulmonary infiltrates: a pilot study. Lung. 2013;191:619–624.
19. Kropski JA, Pritchett JM, Mason WR, et al.. Bronchoscopic cryobiopsy for the diagnosis of diffuse parenchymal lung disease. PLoS One. 2013;8:e78674.
20. Fruchter O, Fridel L, Rosengarten D, et al.. Transbronchial cryo-biopsy in lung transplantation patients: first report. Respirology. 2013;18:669–673.
21. Yarmus L, Akulian J, Gilbert C, et al.. Cryoprobe transbronchial lung biopsy in patients after lung transplantation: a pilot safety study. Chest. 2013;143:621–626.
22. Franke KJ, Szyrach M, Nilius G, et al.. Experimental study on biopsy sampling using new flexible cryoprobes: influence of activation time, probe size, tissue consistency and contact pressure of the probe on the size of the biopsy specimen. Lung. 2009;187:253–259.
Keywords:

bronchoscopy; transbronchial biopsy; cryobiopsy; diffuse parenchymal lung disease

Copyright © 2016 Wolters Kluwer Health, Inc. All rights reserved.