Journal of Cancer Research and Clinical Oncology

, Volume 137, Issue 1, pp 125–130

Palliative radiotherapy for bleeding from advanced gastric cancer: is a schedule of 30 Gy in 10 fractions adequate?

Authors

    • Division of Radiation OncologyShizuoka Cancer Center Hospital
  • Takayuki Hashimoto
    • Division of Radiation OncologyShizuoka Cancer Center Hospital
  • Hideyuki Harada
    • Division of Radiation OncologyShizuoka Cancer Center Hospital
  • Masashi Mizumoto
    • Division of Radiation OncologyShizuoka Cancer Center Hospital
  • Kazuhisa Furutani
    • Division of Radiation OncologyShizuoka Cancer Center Hospital
  • Noriaki Hasuike
    • Division of EndoscopyShizuoka Cancer Center Hospital
  • Masaki Matsuoka
    • Division of Gastrointestinal OncologyShizuoka Cancer Center Hospital
  • Hiroyuki Ono
    • Division of EndoscopyShizuoka Cancer Center Hospital
  • Narikazu Boku
    • Division of Gastrointestinal OncologyShizuoka Cancer Center Hospital
  • Tetsuo Nishimura
    • Division of Radiation OncologyShizuoka Cancer Center Hospital
Original Paper

DOI: 10.1007/s00432-010-0866-z

Cite this article as:
Asakura, H., Hashimoto, T., Harada, H. et al. J Cancer Res Clin Oncol (2011) 137: 125. doi:10.1007/s00432-010-0866-z

Abstract

Purpose

To evaluate the effectiveness of short-course radiotherapy (RT) with 30 Gy in 10 fractions for bleeding from advanced gastric cancer.

Methods

We reviewed the data for all patients with gastric cancer requiring blood transfusions due to gastric bleeding who were treated with RT at the Shizuoka Cancer Center Hospital between September 2002 and March 2007. Patients with curative-intent chemoradiotherapy or previous irradiation were excluded. RT was planned to deliver a total of 30 Gy at 3 Gy per fraction. We defined RT as effective if the patients did not require blood transfusions for 1 or more months after RT.

Results

Twenty-two out of 30 patients (73%) responded to RT, and rebleeding occurred in 11 (50%) of 22 patients responding to RT. The median actuarial time to rebleeding was 3.3 months. Twelve patients received concurrent chemoradiotherapy and had a significantly lower rebleeding rate than patients undergoing RT alone (P = 0.001). Among patients receiving CRT, 1 with grade 3 non-hematological toxicity and 5 with grade 3–4 hematological toxicity were observed. No Grade 3 or higher adverse events were observed in patients treated with RT alone.

Conclusions

RT with 30 Gy in 10 fractions is an adequate treatment for bleeding from advanced gastric cancer, especially in patients with poor prognosis.

Keywords

Gastric cancerBleedingPalliationRadiotherapyChemoradiotherapy

Introduction

For patients with resectable gastric cancer, the standard treatment with curative intent is radical surgery. Radiotherapy (RT) is mainly used with concurrent chemotherapy as adjuvant therapy in gastric cancer treatment (Xiong et al. 2003; Macdonald et al. 2001; Foukakis et al. 2007). The majority of patients with unresectable gastric cancer are treated with chemotherapy. Many clinical trials have shown the effectiveness of chemotherapy in unresectable gastric cancer (Wöhrer et al. 2004; Wagner et al. 2006; Rivera et al. 2007).

Gastric bleeding from the primary tumor frequently occurs in patients with unresectable gastric cancer. RT has been used to manage hemorrhage from lung cancer, cervical cancer, bladder cancer, and skin cancer (Pereira and Phan 2004). Although RT has also been used for palliation of symptoms including hemorrhaging in patients with unresectable gastric cancer, there have been only a few reports on this topic (Tey et al. 2007; Kim et al. 2008; Hashimoto et al. 2009), and various dose-fractionation regimens were used in these previous studies. Although 30 Gy in 10 fractions is one of the most commonly used dose-fractionation regimens for palliative RT, it has remained unclear whether this regimen is effective in treating bleeding from advanced gastric cancer. The purpose of this study was therefore to evaluate the effectiveness and safety of short-course RT with 30 Gy in 10 fractions for bleeding from advanced gastric cancer.

Methods

Patients

The medical and radiation records of all patients with gastric cancer requiring blood transfusions due to gastric bleeding who were treated with RT at the Shizuoka Cancer Center Hospital between September 2002 and March 2007 were retrospectively reviewed. Patients with curative-intent chemoradiotherapy or previous irradiation were excluded.

Radiotherapy

RT was planned to deliver a total of 30 Gy at 3 Gy per fraction using a linear accelerator with a 6-, 10-, or 18-MV photon beam. Three-dimensional dose calculations were performed using Pinnacle3 software (ADAC, Milpitas, CA) with tissue-density inhomogeneity correction. The treatment planning was based on 3.8- to 5-mm-thick CT scans obtained in the treatment position. The clinical target volume (CTV) was based on clinical examinations including CT scan and endoscopy. Regional lymph nodes were not intended to be included in the CTV. The radiation fields were designed to cover the CTV with an adequate margin. The majority of patients were treated with anteroposterior–posteroanterior opposed beams or oblique opposed beams.

Evaluation

We defined RT as effective if the patients did not require blood transfusions for 1 or more months after the beginning of RT. Treatment toxicities were assessed according to Common Terminology Criteria for Adverse Events version 3.0 (CTCAEv3.0). Overall survival, rebleeding-free survival, and time to rebleeding were estimated from the beginning of RT using the Kaplan–Meier method, and the differences between groups with or without concurrent chemotherapy were compared using the log-rank test. For the estimation of time to rebleeding, patients without rebleeding were censored at the time of death or last follow-up. Patients who did not achieve a 1-month blood transfusion–free survival time or who underwent other concomitant treatments for hemostasis were considered as failure to achieve bleeding control (i.e. rebleeding) at day 0. Comparisons of the amount of blood transfused and the hemoglobin (Hb) levels before and after treatment were carried out using the Wilcoxon signed rank test. Data were considered statistically significant at P values <0.05.

Results

Subjects

During the four and a half years of the study, there were 37 patients at Shizuoka Cancer Center Hospital who required blood transfusions due to bleeding from gastric cancer and were treated with RT. In each patient, the diagnosis was confirmed histopathologically. A total of 30 patients were analyzed, and the other 7 patients were excluded from the analysis because of curative-intent chemoradiotherapy (n = 6) or previous RT (n = 1).

The patient characteristics are shown in Table 1. There were 21 men and 9 women with a median age of 69 years (range, 36–82 years). Twenty-one patients received chemotherapy before RT. The number of chemotherapy regimens carried out before RT was one in 7 patients, two in 9 patients, and three or more in 5 patients. All patients underwent upper gastrointestinal endoscopy prior to radiotherapy, which confirmed the bleeding from the primary tumors. Seven patients had been treated for bleeding by endoscopic interventions within 1 month before starting RT and had not achieved hemostasis. Twenty-six patients had symptoms including melena or hematemesis.
Table 1

Patient characteristics (n = 30)

 

No. (%)

Age (year)

 Range

36–82

 Median

69

Sex

 Male

21 (70)

 Female

9 (30)

Performance status (ECOG)

 0–1

9 (30)

 2

9 (30)

 3

11 (37)

 4

1 (3)

Lowest Hb level before RT (g/dl)

 Range

3.8–8.5

 Median

5.1

Distant metastasis

 No

2 (4)

 Yes

28 (96)

Previous chemotherapy

 No

9 (30)

 Yes

21 (70)

Radiation dose

 21 Gy/7 fr

1 (3)

 27 Gy/9 fr

2 (7)

 30 Gy/10 fr

27 (90)

Treatment field

 Local

24 (80)

 Whole stomach

6 (20)

Concurrent chemotherapy

 No

18 (60)

 Yes

12 (40)

Additional chemotherapy after RT

 No

11 (37)

 Yes

19 (63)

Treatment

Twelve patients received concurrent chemoradiotherapy (CRT). The concurrent chemotherapy regimens were as follows: S-1/cisplatin in 6 patients, S-1 in 1 patient, methotrexate/5-fluorouracil in 2 patients, 5-fluorouracil in 2 patients, and paclitaxel in 1 patient. Seven of 9 patients without previous chemotherapy received CRT. Nineteen patients received additional chemotherapy after completing RT. Ten of 12 patients with CRT received additional chemotherapy. Only 1 patient never received previous, concurrent, or additional chemotherapy.

Twenty-seven patients received a total of 30 Gy in 10 fractions. One patient was scheduled to receive a total of 27 Gy in 9 fractions due to the machine operating schedule resulting from a long holiday period. Two patients discontinued RT due to a poor general condition at a total dose of 21 Gy and 27 Gy, respectively. The CTV was determined as the whole stomach in 6 of the 30 patients.

Efficacy

The median follow-up from the beginning of RT for the 30 patients was 3.5 months (range, 0.5–19.6 months). Figure 1 shows the overall survival curve for all 30 patients in the study. Twenty-nine (97%) of 30 patients died, and 1 patient was lost to follow-up. The median survival time from the beginning of RT was 3.6 months. Twenty-two (73%) of 30 patients responded to RT (i.e., patient did not require blood transfusion for 1 or more months). Six patients did not achieve a 1-month blood transfusion–free survival time, including 3 patients who died due to metastatic disease within 1 month of the beginning of RT, and 2 patients underwent other concomitant treatments for hemostasis (one patient was treated with an endoscopic argon plasma coagulator system and the other received arterial embolization). Of the 26 patients with melena or hematemesis, 20 (77%) showed an improvement in their symptoms. Figure 2a shows the amounts of blood transfused for 22 patients responding to RT before and after completion of RT. The mean amount of total blood that corresponds to transfused red blood cells in the 1-month periods before and after completion of RT were 2236 and 273 ml, respectively (P < 0.0001). Figure 2b shows the Hb levels of 22 patients responding to RT before and after treatment. The median Hb level 1 month after the last blood transfusion was 8.2 g/dl (range, 6.1–11.4 g/dl), compared to the median value of the lowest Hb level before the last blood transfusion of 4.9 g/dl (range, 3.8–8.5 g/dl) (P < 0.0001).
https://static-content.springer.com/image/art%3A10.1007%2Fs00432-010-0866-z/MediaObjects/432_2010_866_Fig1_HTML.gif
Fig. 1

Overall survival for the entire patient population

https://static-content.springer.com/image/art%3A10.1007%2Fs00432-010-0866-z/MediaObjects/432_2010_866_Fig2_HTML.gif
Fig. 2

a Comparison of the amount of blood transfused in the 1 month before and after completion of RT for 22 patients responding to RT. Each circle on this plot represents the individual amount of total blood that corresponds to transfused red blood cells. The black bars are the mean amounts of total blood that correspond to transfused units across patients. b The Hb levels of 22 patients responding to RT before and after treatment. Each circle on this plot represents the individual Hb level, and black lines connect the Hb levels before and after treatment for each patient

Figure 3a shows the rebleeding-free survival curve. The median rebleeding-free survival time from the beginning of RT was 2.6 months. The rebleeding-free survival curves in those with and without concurrent chemotherapy are shown in Fig. 3b. The median rebleeding-free survival was 1.7 months for RT alone compared with 5.5 months for CRT (P = 0.002). The median actuarial time to rebleeding was 3.3 months (Fig. 4a). The cumulative incidence curves for rebleeding in those with and without concurrent chemotherapy are shown in Fig. 4b. The 3-month cumulative incidences of rebleeding were 60% for RT alone compared with 17.5% for CRT (P = 0.001).
https://static-content.springer.com/image/art%3A10.1007%2Fs00432-010-0866-z/MediaObjects/432_2010_866_Fig3_HTML.gif
Fig. 3

a Rebleeding-free survival for the entire patient population. b Rebleeding-free survival of patients with and without concurrent chemotherapy

https://static-content.springer.com/image/art%3A10.1007%2Fs00432-010-0866-z/MediaObjects/432_2010_866_Fig4_HTML.gif
Fig. 4

a Actuarial rebleeding rate for the entire patient population. b Actuarial rebleeding rate of patients with and without concurrent chemotherapy

Of the 22 patients responding to RT, rebleeding occurred in 7 of 11 patients receiving RT-alone and 4 of 11 patients receiving CRT, respectively.

Toxicities

Most patients experienced only minimal toxicity. A total of four grade 3 and two grade 4 treatment-related adverse events were observed in 4 patients receiving CRT. Of these patients, one experienced grade 3 gastric bleeding as a late toxicity. No grade 3 or higher adverse events were observed in patients treated with RT alone (Table 2).
Table 2

Toxicity (CTCAEv3.0 Grade ≥ 2)

 

RT alone

(n = 18)

CRT

(n = 12)

Total

Leukocytes

Grade 2

1

2

3

 

Grade 3

0

3

3

 

Grade 4

0

1

1

Platelets

Grade 2

0

0

0

 

Grade 3

0

0

0

 

Grade 4

0

1

1

Nausea/vomiting/anorexia/heartburn

Grade 2

5

5

10

Hemorrhage, GI-stomach

Grade 3

0

1

1

Discussion

Apart from three recent studies (Tey et al. 2007; Kim et al. 2008; Hashimoto et al. 2009) on the topic, there have been few reports regarding palliative RT for gastric cancer. The rates of bleeding control ranged from 57 to 70% in these studies. However, various dose fractionation regimens were used in each study, and there were also differences in the evaluation of treatment efficacy among the studies. These differences should be considered when comparing the bleeding control rates in the present study and other publications. In the present study, all patients received blood transfusions before RT, and the majority of patients received blood transfusions during RT, so RT was assumed to be effective when the patient did not require blood transfusions for 1 month or longer. Hashimoto’s study and our study used similar methods for the evaluation of treatment efficacy, and treatment success was observed in 68% (13/19) of patients in their paper. A biologically effective dose (BED), calculated by the linear quadratic formula with an α/β ratio of 10 Gy, of 50 Gy10 or more was significantly correlated with treatment success compared with a BED of <50 Gy10, and so they recommended a dose of 40 Gy in 16 fractions. However, the dose-fractionation regimen of 30 Gy in 10 fractions used in our study corresponds to 39 Gy10, and the rate of bleeding control was 73% on an intention-to-treat basis in the present study. As Hashimoto et al. mentioned in their discussion, patients treated with a BED of <50 Gy10 included more than a few patients who discontinued RT, and the dose–response relationships in their study might also have been biased.

A dose-fractionation regimen of 30 Gy in 10 fractions is commonly used for palliative radiotherapy such as for brain metastasis and bone metastasis, and also showed a high response rate of 73% for palliating bleeding from advanced gastric cancer in our study. This schedule appears to have the benefit of being able to reduce the treatment period relative to 40 Gy in 16 fractions or 35 Gy in 14 fractions, doses which have been used for a large portion of patients in previous studies regarding palliative RT for gastric cancer. In fact, of the patients in this study, 2 received RT on an outpatient basis, and the majority of patients who needed to be hospitalized were discharged from the hospital after achieving hemostasis. Reduction of the treatment period would be beneficial for these patients. Although 73% of patients responded to RT, the median actuarial time to rebleeding was 3.3 months in the current study. Kim et al. (2008) have reported that 6-month local control is better in patients treated with a BED of ≥41 Gy10. The longer the patient lives after the end of RT, the greater the risk of rebleeding. Survival time will be influenced by various factors such as the time from the initial treatment, previous chemotherapy, performance status, and the existence of metastasis. Further clinical studies are required to assess the optimal dose-fractionation regimen, especially for patients expected to survive for a relatively long period.

The rate of patients free from rebleeding was better among the CRT group than the RT-alone group. However, there are limitations to this study, in that it was retrospective with a small sample size. Since systemic chemotherapy is the standard treatment for metastatic or unresectable gastric cancer, 70% (21/30) of patients received chemotherapy as the first-line treatment, and gastric bleeding occurred as a result of the disease progression in these patients in the present study. The CRT group tended to include patients without previous chemotherapy (7 of 12 patients) compared to the RT-alone group (2 of 18 patients). Selection biases might be relevant to an evaluation of this study. Considering that short-course RT with concomitant use of chemotherapy was feasible in the present study, indications for CRT should be determined based on whether there are chemotherapeutic regimens effective for the patients, by assessing the treatment history of the patients.

In conclusion, RT with 30 Gy in 10 fractions was found to be an effective and safe treatment option for bleeding from advanced gastric cancer. Seventy-three percent of patients achieved hemostasis in this study. Considering the median actuarial time to rebleeding (3.3 months), this dose-fractionation regimen appears to be adequate, especially for patients with poor prognosis.

Conflict of interest statement

We declare that we have no conflict of interest.

Copyright information

© Springer-Verlag 2010