Objective To evaluate haematogenesis of complex Shenlu granule in treating myelodysplastic syndrome (RA/RCMD) due to kidney yang deficiency. Methods The qualified cases were randomly divided into the complex Shenlu granule group and the Andriol group. During treatment, indexes such as peripheral hemogram, T cell subgroup, NK cell and immunological phenotypic expression of BMC were tested. Results Compared with prior treatment, Rbc, Hb and Plt have significantly increased in Shenlu group [prior treatment: (2.39±0.99)×1012/L, (84.47±28.68) g/L, (81.13±96.85) ×109/L; post-treatment (2.80±0.98)×1012/L, (94.87±25.63) g/L, (98.67±107.9)×109/L, and the difference was statistically significant (t =4.0359, P =0.001, t =2.7009, P =0.0172, t =2.2573, P = 0.0405). Andriol group has only increased Hb [prior treatment: (71.93±27.53) g/L; post-treatment (80.07±26.03) g/L] was statistically significant (t =2.3125,P =0.0365). After treatment, CD+4, CD+8, CD+4/CD+8, and NK were at or near normal value in Shenlu group (37.9±5.9) %, (24.0±5.8) %, 1.75±0.83, (13.0±6.9) %, compared with prior treatment (29.3±11.7) %, (29.6±5.8) %,1.12±0.59, (8.8±5.7) %. The difference was statistically significant (t =2.6194, P =0.0202, t = 2.6595, P =0.0187,t =2.6581, P =0.0187, t =2.2288, P =0.0427, P <0.05), and CD+8, CD+4/CD+8 close to normal in Andriol group [(22.1 ±7.5) %, 1.50±0.74], prior treatment [(26.6±7.5) %, 1.18±0.55]. The difference was statistically significant (t =2.2377, P =0.0420, t =2.9352, P =0.0109, P <0.05). After treatment, CD+4 [Shenlu group:(37.9±5.9) %, Andriol group: (30.5±12.6) %] difference was statistically significant (t =2.1738, P =0.0474) in two groups; Shenlu group control abnormal expression of bone marrow cells, which were CD+13, CD+33, CD+34, CD+64,CD+117. The effect was better than Andriol group (the first three: u =2.76, u =3.39, u =2.85, P <0.01, the latter two: u =2.17, u =2.46, P <0.05). Conclusion clomplex Shenlu granule can effectively control normal hematopoietic function for MDS (RA/RCMD) due to kidney Yang deficiency.

Objective To study a rational method of pancreaticojejunostomy after pancreatoduodenectomy.Methods Retrospective analysis was made on the clinical material of 108 cases of bundle-invagination(pancreaticojejunostomy) after pancreatoduodenectomy performed from Jan 1996 to Dec 2003 in our department.Results The operative time was 3.5~4.5 hours.No perioperative death nor pancreatic fistula occurred.One diabetic patient developed biliary fistula on the 8~(th) post-operative day and it healed spontaneously after drainage.Two cases had wound dehiscence that healed after tension relaxation suture.Two cases of(postoperative) gastroparesis recovered after conservative treatment.Conclusions The bundle-invagination(pancreaticojejunostomy) can more effectively prevent pancreatic fistula and have less operation time.