The proportion of gastric cancer patients among the annually registered cancer patients in Korea is as high as approximately 20%.(1) It is very important to assess the presence or absence of metastasis to establish the treatment plans for gastric cancer, which occurs at such a high rate, and to predict the prognosis. Gastric cancer generally metastasizes to the peritoneal membrane, liver, lymph nodes, etc., and it may metastasize to the spleen, adrenalin, ovary, lung, brain and skin.

Bone metastasis generally occurs in patients with prostate cancer, breast cancer and lung cancer, and bone metastasis in gastric cancer patients has been shown to be very rare.(2-4) Bone metastasis is usually associated with disseminated vascular coagulation, hemolytic anemia and other hematological complications, and the prognosis is very poor.(5) The diverse incidence and prognosis of bone metastasis from gastric cancer have been reported in numerous studies, but this has not been sufficiently established in Korea. Therefore, we conducted this study to examine the clinical characteristics of the patients diagnosed with gastric cancer by gastroscopy as well as histological tests, and these patients had bone metastasis detected simultaneously or later, and we also wanted to assess their prognosis.

Bone metastasis frequently occurs in patients with breast cancer, lung cancer, renal cancer, prostate cancer, bladder cancer and other primary cancers. In comparison, it has been shown that bone metastasis that originates from malignant tumors of the gastrointestinal tract is rare.(2-4) In 1983, Yoshikawa and Kitaoka(6) reported that the incidence of bone metastasis is 1~20%. In 1987, Nishidoi and Koga(7) have reported that in 246 gastric cancer patients, bone metastasis was associated in 33 patients (13.4%). In our study, among the 2,150 patients diagnosed with gastric cancer from June 1992 to August 2010, bone metastasis was associated in 19 patients for a frequency of 0.9%. In other words, depending on the research institutions and investigators, the incidence of bone metastasis varies greatly. Seto et al.(8) have reported that bone scintigraphy was performed on 60 patients, and bone metastasis was suspected in 25% of these patients. Bone metastasis may occur more frequently in cases with primary cancer in the body of stomach, poorly differentiated adenocarcinoma and in cases with abundant lymph node metastasis in the vicinity. In our study, like as shown in the study reported by Seto et al., the undifferentiated histological types (14/19, 73%) were more abundant than the differentiated types (5/19, 27%). However, the survival period of the two groups was not significantly different. Thus, the survival period was not different according to the histological types. In 1995, Choi et al.(9) performed bone scintigraphy on 234 gastric cancer patients, and based on the presence or absence of hot uptake lesions, the frequency and pattern of bone metastasis were inferred and reported. According to this, there were 106 patients who showed hot uptake lesions, the estimated incidence of bone metastasis was 45.3% and an elevated serum ALP value was associated with bone metastasis. In our study, similarly, among the 19 patients, the serum ALP value was elevated in 12 patients at the time of diagnosis. Similarly, in 2006, Kusumoto et al.(10) reported on the clinical characteristics of 9 patients with bone metastasis from gastric cancer, and it was observed that the serum ALP level was elevated in all 9 patient, and the serum ALP value was an important factor for making the diagnosis of bone metastasis. Therefore, it is thought that in patients diagnosed with gastric cancer, if the ALP value is atypically elevated, then evaluation of bone metastasis is required.

In our study, bone scintigraphy was applied most frequently for the diagnosis of bone metastasis (15/19, 78%). In addition, positron emission tomography-computed tomography and MRI were used alone or in combination. On account of the use of a gamma camera and the development of test agents such as Tc-99m MDP, the sensitivity of bone scintigraphy is particularly high, and thus it is known to be the most useful screening test. Bone scintigraphy could detect the change of abnormal blood flow within the bone that developed metastasis, and so it could detect bone metastasis at the time approximately 3 months earlier than that with using plain X-rays.(11,12) Nevertheless, hot uptake lesions may be detected in Paget's disease and other metabolic bone diseases, degenerative arthritis, fractures, infectious bone diseases and other benign bone diseases or primary bone tumors, and so it has limitations of low specificity. In our study, among the 19 patients, there were only 5 patients with a solitary lesion. Several studies have examined the rate of detecting bone metastasis with a solitary lesion as detected by bone scintigraphy, and only approximately 50~55% of the cases were diagnosed as having bone metastasis.(13-16) In other words, bone scintigraphy has shortcomings that it is difficult to diagnose bone metastasis from malignant tumors in the cases with a hot uptake lesion. Therefore, it is considered that for cases with a solitary hot uptake lesion detected by bone scintigraphy, together with accurate history taking for fracture and other past medical problems, the rate of a false diagnosis of bone metastasis from gastric cancer could be decreased by making an accurate differential diagnosis from infectious bone diseases, metabolic bone diseases and other benign diseases. In our study, for the cases that a solitary lesion was detected by bone scintigraphy, efforts were made to increase the accuracy of the diagnosis by additionally performing positron emission tomography-computed tomography or magnetic resonance imaging. The accuracy of the diagnosis may be increased more by performing additional bone marrow tapping or bone marrow histological tests in parallel. Choi et al.(9) have reported that the area where bone metastasis was frequently diagnosed by bone scintigraphy was in the order of the vertebrae (66%), the costa (59%), the pelvic bone (43%), the femur (30%) and the scapula and clavicle (17%). In our study, among the 19 patients, metastasis in the vertebrae was detected in 17 cases (89%), and so similar results were obtained. Other metastases associated with bone metastasis were also examined, and particularly, 3 of the 19 patients (15%) were associated with liver cancer. In general, for gastric cancer and including early gastric cancer, the most frequently metastasized area by hematogenous metastasis is the liver.(10) In our gastric cancer patients, the rate of liver metastasis was lower than the general rate of liver metastasis. This could be explained by the difference of metastasis routes. In the cases with well differentiated tumors, metastasis is developed through the portal vein. On the other hand, in the cases with poorly differentiated cancer, bone metastasis primarily occurs through the vertebral vein system.(8,17) Several studies have reported that the vertebrae are the site where bone metastasis occurs most frequently, which supports that the vertebral vein system is the major route of bone metastasis.(9,12,18) Kusumoto et al.(10) have reported that in patients with gastric cancer, bone metastasis is more prevalent in the young age groups, but in our study, age was not found to be correlated with bone metastasis. In bone metastasis from gastric cancer, the cancer cells diffusely proliferate in the bone marrow and this can cause disseminated carcinomatosis; they also proliferate rapidly and thus induce bone destruction as well as hematological complications. Yet the developmental mechanism of bone destruction has not yet been elucidated. More studies on this are required.(10)

Yoshikawa and Kitaoka(6) and Nishidoi and Koga(7) have reported that bone metastasis from gastric cancer is associated with Borrmann's types 3 and 4, which arereferred to as scirrhous carcinoma. In our study, similarly, Borrmann's types 3 and type 4 were found in 17 cases of the entire 19 cases (89%), and these results are similar to those of the above studies. In regard to the association with the disease stage of primary gastric cancer, the 11 patients who underwent surgery (61%) were stage IV. In other words, in the Borrmann type 3 and 4 patients, bone metastasis occurs frequently in the cases with primary cancer of a high disease stage. Nonetheless, the macroscopic type and disease stage were not significantly correlated with the survival period after the diagnosis of bone metastasis. Yet the number of cases in each disease stage was too small, and a comparative analysis of the survival period of each disease stage could not be performed. Analysis of a larger number of cases is required for a clearer interpretation of the results of the analysis of our study(that the survival period according to the disease stages was not different. Among the 19 patients diagnosed with bone metastasis, 5 cases could not be treated because of the deterioration of their general condition due to the progression of gastric cancer, and the remaining 14 patients were treated. On the univariate analysis of the survival period of the treatment group and the non-treatment group, the survival period of the treatment group was 4.4±0.7 months and it was significantly longer than the 2.0±0.4 survival period of the untreated group. The most common clinical symptoms and complications of bone metastasis are bone pain, pathologic fracture and spinal cord compression. In our study, similarly, in the 19 subject patients, the majority or 16 patients (84%) were positive for bone pain. Clinically, the cause of the worst suffering of patients with bone metastasis is bone pain. Therefore, pain management for patients with bone pain is very important. In regard to radiation therapy for such bone pain, Murai et al.(19) have reported that 68 patients with bone metastasis were classified according to the type of primary tumors, and radiation therapy was effective in 73% of the lung cancer patients, 100% of the breast cancer patients and 75% of the gastric cancer patients. In addition, Yoshikawa and Kitaoka(20) have reported that in 23 gastric cancer patients associated with bone metastasis, radiation therapy was effective for the amelioration of bone pain, yet chemotherapy was not effective for this. McQuay et al.(21) have reported that in regard to the effectiveness of radiation therapy for the amelioration of bone pain, 1 month after the treatment, the bone pain was resolved completely in 1/4 of the patients, and at least 50% pain amelioration could be obtained in 1/3 of the patients. The survival period of was recently improved by treatment with MTX+5FU, S1+paclitaxel, S1+cisplatin and other chemotherapy,(22,23) and in addition it was effective on the amelioration of pain.(24) Bisphosphonate(25) has recently been used for the treatment of clinical symptoms caused by bone metastasis and the complications. It suppressed the production of proliferative factors by bones through the suppression of the reabsorption of bone, and thus it suppressed the proliferation of cancer cells.(26,27) Therefore, in patients with bone metastasis, it reduces pain and it also suppresses or delays complications caused by bone metastasis. Nonetheless, large scale studies on bone metastasis patients have not been conducted, and studies are needed on the indications for the use of Bisphosphonate and to prove its treatment effectiveness. Chung et al.(28) reported on 4 patients with bone metastasis caused by gastric cancer, and all 4 patients died within 4 months after the diagnosis of bone metastasis. In our study, similarly, the period from the diagnosis of bone metastasis to death was on average 3.8±2.6 months, and it was found that the prognosis was very poor.

In conclusion, the prognosis of bone metastasis caused by gastric cancer is very poor and the prognosis of patients may be worsened due to a delayed diagnosis. Hence, tests to assess for bone metastasis are required for gastric cancer patients at the time of the initial diagnosis and the postsurgical follow-up observation, and bone scintigraphy with its high sensitivity appears to be most useful. In such a manner, it is anticipated that the survival period as well as the quality of life may be improved with making a rapid diagnosis and by administering radiation therapy and other appropriate treatments. Studies with a large number of larger subjects are required to assess the effect ofimproving the survival period according to different treatment methods.