The term "hamartoma" derives from the Greek word which means “error”. It was first used in 1904 by Albrecht [17] to define certain tumor like malformations resulting from a presumed error in development or mixing of tissues normally present in the involved organ. In 1934, Goldsworthy defined as PH the benign tumor composed of a combination of adipose tissue and cartilage [1, 2]. The significance of PH remains intriguing. Malignant degeneration of PH to carcinoma or sarcoma has been reported in literature, as well as the association between PH and lung cancer. However, these evidences are supported by small series or sporadic case reports, and thus they are still under debate. To evaluate these issues, we planned a multicenter study collecting the largest number of patients with PH undergoing surgical resection published so far.
First, our findings confirmed the previously published data on PH. This tumor occurred most frequently between the sixth and seventh decade of life, with a male preponderance. In most of cases, it appeared as a small, solitary pulmonary nodule, uniformly distributed in the two lungs. Only few patients with endobronchial obstruction presented specific symptoms related to airway obstruction by PH, and in most cases PH was incidentally detected by CT scan performed for other diseases. In 112 cases, there were calcifications on CT findings, and only 33 out of 160 lesions were FDG-avid (SUV uptake > 2.5). Pre-operative FNAB was not routinely performed probably due the difficulty in aspirating adequate cytological or histological samples due to the dense structure of the lesion [18, 19]. In asymptomatic patients the decision for upfront surgical resection was based on (i) the difficulty in distinguishing PH from malignant lesions, particularly in patients with history of malignancy; (ii) the increase in size during follow-up; and (iii) patient’s decision, owing to unclear diagnosis. Lung sparing resections (i.e. enucleation or wedge resection) were the main surgical strategy, while anatomic resections were performed only in selected cases due to tumor extension.
Second, the outcome observed in our series confirmed the benign nature of PH. No case of malignant transformation of PH was found, and only one patient had recurrence (0,2%) 11 months after endoscopic resection. Thus, an incomplete excision likely explained this event. Previously published series in the English literature [1–16], considered collectively with the present in Table 4, confirmed our results. Among 1,733 patients evaluated, recurrence after excision was found in only 4 cases (0.23%). In all cases, it occurred in the same pulmonary segment after enucleation, pointing out that excision with a minimal margin of normal lung was mandatory to prevent the recurrence. No cases of malignant transformation were found suggesting that this event is highly exceptional or impossible. Rare isolated reports showed the possibility of malignant degeneration of PH, but these findings were contradicted by other authors. Hayward and Carabasi [20] found that most papers either lacked evidence that original tumor was a hamartoma or provided weak evidence of malignant change. In histological findings of some reports hamartoma and carcinoma cell lines seemed to be independent from each other, suggesting a coexistence of the two tumors rather than a degeneration of PH [21]. In others, PH involved the pleura and this unusual growth pattern was considered a sign of malignant transformation, but no malignant cells were detected in any of these cases [22]. Furthermore, there was no evidence of malignant transformation in surveillance of patients with non-resected PH. Sinner et al. [23] analyzed 61 patients with asymptomatic FNAB-proven peripheral pulmonary hamartoma. Forty-one patients had a 5-year follow-up (the longest surveillance reported in literature), and they developed no malignant transformation. Similarly, in our series 116 patients were followed up for a mean time of 25 months before resection. The tumor increased in size, but no malignant degeneration was found.
Table 4
Review of the literature regarding recurrence and malignant transformation of PH
Authors | No. of PH | Recurrence | Malignant transformation |
Koutras et al. [1] | 19 | 0 | 0 |
Karasik et al. [2] | 52 | 0 | 0 |
Fudge et al. [3] | 29 | 0 | 0 |
van Den Bosch et al. [4] | 154 | 2 | 0 |
Crouch et al. [5] | 19 | 0 | 0 |
Salminen et al. [6] | 77 | 0 | 0 |
Hansen et al. [7] | 89 | 0 | 0 |
Ribet et al. [8] | 65 | 0 | 0 |
Gjevre et al. [9] | 216 | 0 | 0 |
Lee et al. [10] | 29 | 0 | 0 |
Lien et al. [11] | 62 | 0 | 0 |
Guo et al. [12] | 39 | 1 | 0 |
Çaylak et al. [13] | 20 | 0 | 0 |
Wang [14] | 226 | 0 | 0 |
Ekinci et al. [15] | 73 | 0 | 0 |
Haberal et al. [16] | 24 | 0 | 0 |
Our series | 540 | 1 | 0 |
Total | 1,733 | 4 (0.23%) | 0 |
Third, we found that PH was associated with synchronous or metachronous lung cancer in 14% of our population that is likely higher compared to the incidence of lung cancer in general Italian population considering that about 41,500 new cases of lung cancer have been estimated in 2018 as reported by the Italian Association of Medical Oncology and the Italian Association of Tumor Registries [24]. The association between PH and lung cancer has been previously reported, with an incidence ranging from 1–23% in previous studies [1–9, 15], summarized in Table 5. Furthermore, Karasik et al. [2], and Ribet et al. [8] calculated that the risk for lung cancer in patients with PH were 6.3 and 6.7 times respectively, higher than that expected for the general Israeli and French population. However, it is still under debate whether PH is a real risk factor for lung cancer development or just an associated phenomenon. Karasik et al. [2], and Ribet et al. [8] supported the first hypothesis. In Karasik’s series [2], all four associated lung cancers occurred in the same lobe of the hamartoma. It has been reported that 5–25% of all lung cancer are associated with scars, and the most common example is adenocarcinoma arising from tuberculosis scar. Thus, the authors [2] supposed that the association between PH and lung cancer was another example of the so-called scar carcinomas, because the fibrosing process sustained by PH favored the development of malignancy. By contrast, van den Bosch et al. [4] and Gjevre et al. [9] considered as fortuity this association. In the series of van den Bosch et al. [4] only 45% of associated lung cancer were localized in the same lobe of hamartoma, making likely doubtful any spatial correlation between two tumors. Thus, the authors [4] explained this association by the possibility that the existence of cancer could lead to the discovery of an asymptomatic hamartoma that would have remained unknown otherwise. Furthermore, Gjevre et al. [9] pointed out that PH and lung cancer were associated to the same exposures and risk factors, that could explain their association. Our results were in line with these. In our series only 26% of associated lung cancers were localized in the same lobe of PH, and similar results were also observed in series from literature review reported in Table 5 where associated lung cancer and hamartoma were localized in the same lobe in 42% of cases. Yet, all patients but two were smokers and our logistic regression analysis found that only smoking and advanced age were significantly associated with lung cancer occurrence. Furthermore, also in the Ribet’s series [8] most patients with lung cancer were smokers. Thus, patients with PH should undergo a complete evaluation and to regular follow-up, especially if they are smokers and elderly, as it could lead to the discovery of early stage lung cancer.
Table 5
Review of literature regarding PH associated with lung cancer
Authors | No. of PH | No. Of Associated Lung Cancer | Same lobe (Lung cancer and PH) |
Total number | Synchronous | Metachronous |
Koutras et al. [1] | 19 | 1 (5.2%) | 1 (5.2%) | 0 | 0 |
Karasik et al. [2] | 52 | 4 (7.6%) | 1 (1.9%) | 3 (5.7%) | 4 (100%) |
Fudge et al. [3] | 29 | 5 (17.2%) | 1 (3.4%) | 4 (13.8%) | N/A |
van Den Bosch et al. [4] | 154 | 11 (7.1%) | 6 (3.8%) | 5 (3.3%) | 5 (45.4%) |
Crouch et al. [5] | 19 | 2 (10.5%) | 2 (10.5%) | 0 | 2 (100%) |
Salminen et al. [6] | 77 | 1 (1.2%) | 0 | 1 (1.2%) | N/A |
Hansen et al. [7] | 89 | 1 (1.1%) | 1 (1.1%) | 0 | 1 (100%) |
Ribet et al. [8] | 65 | 3 (4.6%) | 0 | 3 (4.6%) | 0 |
Gjevre et al. [9] | 216 | 45 (1.1%) | 39 (18%) | 6 (2.8%) | 16 (35.5%) |
Ekinci et al. [15] | 73 | 17 (23%) | 13 (17.8%) | 4 (5.2%) | 4 (23.5%) |
Our series | 540 | 76 (14%) | 9 (1.6%) | 67 (12.4%) | 20 (26%) |
Total | 1,333 | 166 (12.4%) | 73 (5.4%) | 94 (7%) | 52 (42%)* |
*This value was calculated on the total of 1,227 cases as in 106 cases it was not reported which lobe was affected by lung cancer. |
Obviously, our results should be evaluated with caution, before drawing definitive conclusions. Due to the retrospective and multicenter nature of the study, there was not a standardized protocol for the timing of surgical resection, the extent of resection, the histological diagnosis and the clinical follow-up modality. The lack of a surveillance control group made impossible to know whether all PHs undergoing upfront surgery in our series would remain stable in size over the time.
In conclusion, resection of PH is a safe procedure. The recurrence and malignant degeneration are likely uncommon and the association between PH and lung cancer seems to be a spurious phenomenon related to common risk factors as smoking and advanced age.