To our knowledge, this is the first prospective study to evaluate the diagnostic accuracy of the new 1.7-mm cryoprobe for PPLs. This study demonstrated the feasibility of transbronchial cryobiopsy using the new cryoprobe, suggesting its potential as a tool with excellent diagnostic accuracy. We achieved a diagnostic yield of 94% (95% CI, 83.5–98.8%), which exceeded the expected yield of 85%, and the lowest limit for 95% CI exceeded the threshold yield of 70% that was estimated on the basis of the findings of previous studies on conventional devices. Nevertheless, a certain percentage of patients had factors that were generally considered poorly diagnosed using bronchoscopy, such as small lesions less than 20 mm in diameter and a negative bronchus sign.4, 6, 7, 10, 12 The new cryoprobe may also be effective in diagnosing lesions that are difficult to access or have a poor diagnostic yield with conventional devices.
Numerous studies reported that the diagnostic yield of forceps biopsy for PPLs was higher (73–84%) when the R-EBUS probe was within the lesion, as opposed to being adjacent to it (43–61%).5 This is possibly due to the design of the forceps that enables grasping a tissue only in the forward area, but not the lateral area. In contrast, our study showed that cryobiopsy wherein the R-EBUS probe was positioned adjacent to the lesions had comparable diagnostic yields to cryobiopsy wherein the R-EBUS probe was positioned within the lesions. Similar results were reported for conventional cryoprobes.20, 23 These suggested that cryobiopsy could be useful even when the R-EBUS probe was adjacent to the lesions. This was attributed to the fact that the cryoprobe allowed biopsy sampling from lateral areas because the tissue surrounding the tip of the probe was frozen. This is another advantage of cryoprobes over forceps, in addition to obtaining better quality and quantity of tissues.
Although the difference between the number of diagnostic and non-diagnostic results was small, the presence of the bronchus sign was significantly associated with an increased diagnostic yield. This finding is consistent with those of previous studies showing that the CT bronchus sign is a strong predictor of successful diagnosis for every bronchoscopic technique.1, 5, 11, 22 In terms of other factors, the diagnostic yield was not significantly affected by the following variables: lobe, location, distance from the costal pleura, and related bronchial generation. The conventional cryoprobe is thick and stiff and poses difficulties when approaching some locations, such as the subpleural areas and upper lobes. Moreover, these features may lead to inaccurate placement of the probe when the route to the lesion contains multiple branches. In contrast, the new thin and flexible cryoprobe could be extended to the distal bronchus, thus enabling easier and more appropriate placement of the probe at the same site confirmed using R-EBUS even in previously difficult-to-reach areas, as demonstrated in our representative case (Figure 4).
Despite these advantages, using the new cryoprobe poses some challenges. The biopsy specimen still has to be removed together with the endoscope, as it cannot be retracted through the working channel as with the conventional cryoprobe. This necessitates re-navigation and re-insertion of the bronchoscope and cryoprobe even in cases of bleeding from the biopsy site, thereby raising the possibility that the cryoprobe cannot be consistently introduced to the correct location. This may explain why approximately one-third of the biopsy specimens were nondiagnostic in this study. Furthermore, we calculated the cumulative yield to be 80%, 86%, 92%, and 94% for the first, second, third, and fourth biopsy specimens, respectively. These results suggest that three biopsies are necessary to optimise the diagnostic value of the new cryoprobe.
Cytological examination, with a diagnostic yield for stamp and liquid cytology of 74.0% and 68.0%, respectively, was not very effective for detecting malignancy. Meanwhile, two lesions were identified as atypical based on tissue samples but were identified as malignancies on stamp cytology. Similar findings have been reported in a study using a conventional cryoprobe.21 Owing to the large size of the tissue samples obtained using cryobiopsy, the sectioned area may not contain tumour cells if their distribution is uneven. Although little evidence exists regarding this discrepancy between histological and cytological findings, cytological findings may be important to increase the chances of a diagnosis.
Another major concern with cryobiopsy is the possibility of complications, mainly bleeding and pneumothorax. Although there was heterogeneity in the grading of bleeding severity and the method used to control bleeding, cryobiopsy reportedly induced more bleeding than did forceps biopsy; mild to moderate bleeding occurred in 47–72% of PPLs, whereas the rate of severe bleeding was 0.5%.18–23 Similarly, the most frequently observed complication in this study was mild to moderate bleeding. Nevertheless, it could be controlled in all patients, including a case of severe bleeding, by using the two-scope method without further interventions.29 Pneumothorax occurred in one patient (2%), and its incidence did not differ from that of forceps biopsy (0–5.1%).1, 4 However, the larger specimen obtained using the cryoprobe may have increased the rate of chest tube placement due to more damage to the pleura; for forceps biopsy, the reported rate was only 0.4%.
Our study has some limitations. First, this study was performed at a single institution with a relatively small sample size calculated on the basis of the width of the 95% CI rather than statistical power. Thus, our results may have limited generalizability. Nevertheless, all the procedures were performed according to protocol to ensure an objective analysis. Second, this pilot study did not compare the new cryoprobe with other sampling devices. Further large-scale trials in a randomised setting are required to establish the true diagnostic value of cryoprobes.