PTTM has been regarded as a rare and severe devastating cancer-related pulmonary complication, the majority of which is mainly induced by poorly differentiated adenocarcinoma arising from gastric mucosa . Although it has been well known that the ante mortem diagnosis of PTTM is very difficult , Miyano et al. reported a PTTM case which could be determined through an ante mortem diagnosis and the patient survived after chemotherapy . This report emphasized the necessity and importance of intensive monitoring and prophylactic intervention, especially in the case of advanced gastric carcinoma, in order to make an early diagnosis and save the life of patients with PTTM.
According to previous research [1, 2], the incidence of PTTM in autopsy cases with gastric carcinoma has been reported ranging from 16.7% (6 PTTM cases among 36 gastric carcinoma patients) to 26.8% (11 PTTM cases among 41 gastric carcinoma patients). Similarly, our study revealed an incidence of 16.2% (6 PTTM cases among 37 gastric carcinoma patients), which is largely consistent with these previous reports. In addition, the most common histological type of gastric carcinoma in the present study was poorly differentiated adenocarcinoma (five of the six cases). This fact is also consistent with the previous studies which reported that the most common tumor associated with PTTM is poorly differentiated adenocarcinoma [1, 2].
We now wish to discuss PTTM in more detail from the viewpoint of our morphometric analysis of pulmonary arteries. Since some patients with PTTM suffered from pulmonary hypertension, we suggested that pulmonary arterial remodeling induced by carcinoma cell adhesion onto the endothelium affected the status of pulmonary hypertension. Therefore, in the present study, morphometric analysis was carried out on pulmonary arteries. Results showed that the degree of luminal narrowing of the pulmonary arteries varied from case to case. We focused particularly on case 1 and 2 because pulmonary arterial pressure was evaluated only in these two cases. Case 1 involved clinically proven pulmonary hypertension and revealed widespread severe stenosis in the pulmonary artery that showed a significant negative association between arterial diameter and luminal stenosis rate, as well as pulmonary arterial hypertension as reported by Heath and Edwards . In contrast, case 2 involved clinically denied pulmonary hypertension and showed a significant negative association between arterial diameter and luminal stenosis rate. However, its stenosis was milder and involved a smaller artery in comparison to case 1. Accordingly, the occurrence of pulmonary hypertension may require a widespread pulmonary lesion with severe luminal narrowing. That is, pulmonary hypertension in a patient with PTTM may occur in the case of which distribution of pulmonary lesion is similar to that shown by a pulmonary hypertension case. Then, we prefer to make a little discussion on the histopathological difference between PTTM and pulmonary arterial hypertension (PAH). The initial event of PAH is understood as luminal stenosis due to a symmetric intimal or medial thickening involving a terminal arteriole with a diameter less than 100 μm . However, quite interestingly, our statistical analysis revealed no significant difference for average stenosis rate between the under 100-μm group and the 100 to 300-μm group in the all PTTM cases of this study. These results suggest that carcinoma cell adhesion onto the endothelium in PTTM tends to start from larger arteries rather than a terminal pulmonary arteriole.
Incidentally, VEGF and/or TF expression by carcinoma cells has been confirmed in many cases of PTTM [2–7]. VEGF has been known as an endothelial cell-specific angiogenetic mitogen and is upregulated by TF . Furthermore, Takahashi et al. reported a case of PTTM with OPN expression  and they suggested that OPN promotes thrombus formation, local activation of coagulatory events, and pulmonary hypertension in the pathogenesis of PTTM. Therefore, we also performed immunohistochemical examinations of both gastric and pulmonary tissue sections from each patient. In the present study, 100% (6/6), 83.3% (5/6) and 50.0% (3/6) of PTTM cases showed positive reactivity for TF, VEGF, and OPN, respectively. It has been reported that 25.1% (52/207) , 76.4% (113/148) , and 69.5% (205/295)  of gastric carcinoma cases show positive reactivity for TF, VEGF, and OPN, respectively. Accordingly, our immunohistochemical examination indicated that gastric carcinoma of PTTM shows a higher TF-positive rate than typical gastric carcinoma. However, it still remains obscuring whether gastric carcinoma inducing PTTM shows a higher immunohistochemical positive rate for VEGF or OPN because typical gastric carcinoma also shows a high immunohistochemical positive rate for these entities. Moreover, no significant difference was found in characteristics of phenotypic expression of VEGF or OPN for gastric adenocarcinoma cells between those with and without PTTM, which represents an unknown and important factor that may influence the development of PTTM.