For patients with definite preoperative diagnosis of PH, some scholars suggest only follow-up observation without surgical resection(20). However, PH is often difficult to diagnose before surgery. The preoperative diagnosis of PH mainly based on medical history and image examination. The radiological diagnosis of PH is usually based on CT findings, particularly the detection of popcorn-like calcifications(2) and fat(21). However, typical lesions may be absent, which makes CT insufficient for a benign versus malignant diagnosis. Since the gold criteria of diagnosis still relies on pathological biopsy, preoperative pathological diagnosis is of great significance. For a conclusive diagnosis of hamartoma and to exclude malignancy, cytological or histopathological examination is required.
PTNB provides a definitive diagnosis in most patients with minimal trauma if it is performed properly. PTNB is commonly performed to obtain the diagnosis of pulmonary lesions discovered on imaging. For the patients with high-risk nodules, PTNB has a high diagnostic accuracy and can reduce the resection rate of nonmalignant diseases(22). As the most common benign tumor of the lung, PH remains uncommon in cytologic diagnosis of PTNB specimens, ranging from 0.3%-1.3% of the proportion(10, 11). The diagnostic accuracy of PH was lower than that of malignant tumors, with a specificity of 78%(23).
In our study, the definite histologic diagnosis of PH was obtained by PTNB in 70.2% of the cases, which is consistent with the previous studies(11, 14, 23) and lower than malignant tumors. It is difficult to obtain sufficient diagnostic specimens from cartilage lesions and their nonspecific cytopathology makes it difficult to make a definite diagnosis. The American Pathological Society retrospectively analyzed 19 PTNB tissue samples of PH, and only 26% of pathologists correctly diagnosed PH(23). When studying PTNB, we tend to look for malignant features. The false-positive rate in our study is 12.7%. In one case, the dominance of epithelioid cells led to the diagnosis of carcinoid. The presence of cell pleomorphism, nuclear atypia, and necrosis -even minor and subjective could mislead the diagnosis of cancer, which showed the limitation of PTNB in the diagnosis of PH.
8/49 of the cases were non-diagnostic. It is of challenging to give a benign diagnose because of a high rate of false-negatives. Tongbai et al showed a non-diagnostic rate of 16.7% in their retrospective analysis of 894 PTNBs(24). Fu et al(25) showed that a biopsy result of ‘’chronic inflammation with fibroplasia’’ might indicate the true negatives in nonspecific benign biopsy results, which was confirmed in our study. However, Lee et al showed that nondiagnostic lung biopsies were more likely to develop malignancy and less likely to have insufficient specimens and nonspecific benign categories(26).
Among the patients with a clear preoperative diagnosis of PH, 94.3% of them chose follow-up observation without surgical resection and no significant growth of the PH lesion occurred, indicating that PTNB could be used for preoperative diagnoses and reduce the proportion of surgical intervention. However, 2 patients still underwent surgery, and their PH lesions showed uneven enhancement in CT, which made the choice of follow-up observation need to bear great psychological pressure, Besides, it cannot be ruled out that the tendency of slow progression of lesions may lead to atelectasis, fever and dyspnea. Removal of the lesion can also be done at the patient’s will. The diagnosis of PH in 2 PTNB patients was helpful in guiding surgical methods to avoid lobectomy, regional lymph node dissection or even lung resection(11). 14 patients in our study who were not diagnosed with PH on the basis of PTNB findings were histologically confirmed to have PH by surgical biopsy,, which showed that the false-positive diagnosis and non-diagnosis increased the proportion of unnecessary surgery. Enucleation or segmental resection is required for symptomatic cases and large lesions, PTNB diagnose could help the clinicians to make appropriate therapeutic decisions.
Post biopsy pneumothorax is the most common complication of PTNB, and the reported incidence is 14.6%-23.9%, followed by pulmonary hemorrhage, with a reported incidence of 2.5–15.2%(27). In our study, the pneumothorax rate (14.3%) and hemorrhage rate (4.1%) were similar to reported rates. The dense nature of the of PH lesion could lead to pneumothorax. In Xu’s study(7), the incidence of pneumothorax was related to puncture times and the length of the puncture needle in lung tissues.
PH is composed of mesenchymal tissues including adipocytic component, muscle, cartilage and fibro myxoid tissue. The specific diagnosis of PH relies on the recognition of the FMS or CM in a lesion. In our study, PH was diagnosed in 35 patients, FMS and CM were seen in 65.7% and 57.1% of cases, respectively. As a pathological features of PH, it is not always seen on the PTNB cytology smears. In Wiatrowska’s study(28), CM were seen in only 56% of the smears. CM is of firm texture and the aspiration needle may be deflected or bent when inserted into it. Firm to gelatinous material may provide a clue to the diagnosis of PH(15). In our study, it seems that FMS are more common than CM in PTNB specimens, which was consistent with previous studies(10). It is more helpful to diagnose PH by focusing on FMS. The lack of FMS or CM and the presence of some cellular atypia may prompt the pathologist to question the diagnosis of PH(13). The smears composed of atypical epithelial cells could lead to misdiagnosis of malignant. Cytologic smears of one patients were composed of numerous mucinous epithelial cells(13), the cytologic diagnosis was benign mucinous lesion, but the possibility of low-grade malignancy could not be rule out. This case expanded the cytologic spectrum of PH.
Depending on our results, how to increase the probability of making a benign diagnosis on PH specimens is an important problem to solve. First of all, the prominent FMS/CM seen in the PTNB specimens could lead to the PH diagnosis. However, aspirates of PH are often scanty because of the dense nature of the lesion, resulting in insufficient specimen volume. Meanwhile the dominance of the epithelial component may lead to the misdiagnosis of low-grade neoplasms. Therefore, multiple sampling of the same lesion is of great significance. Second, detailed clinical history of the patients is an important tool, too. it could narrow the number of possible diagnoses. Healthy living habit such as no smoking history and benign radiological characteristics could prevent the pathologist from making the malignant final diagnosis of the PTNB specimens. Finally, the expansion of cytological features of PH could help the pathologists consider the diagnose of PH when encountering unusual smears with non-CM/FMS components and help avoid absolute refusal of radiologically highly suggestive cases(12), possibly helping to decrease the rate of false-positive diagnoses. However, it is necessary for more scholars to conduct relevant studies with larger sample size in the future and find more cytological morphology of PH.
There are some limitations to our study. First, the specification of the needle used for biopsy was different because cases were collected from our hospital and the literature, so that the relationship between this and puncture complications could not be further studied. Second, lack of case data limited the analysis of the results. Third, our research was a retrospective study and could cause selection bias, all our PH lesion were located in the peripheral lung, since it is rarely located in a hilar lesion or an endobronchial lesion. As for these special PH, it is reported in case report that endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) could be used to make diagnose(29, 30), but no studies have been reported in this regard. Last, the collection was not big enough to be convincing so further work needs to be done on providing a more extensive statistical database.