Chinese Gastric Cancer Commonly Used Chemotherapy Regimens To Induce Myelosuppression In A Multicenter, Prospective Clinical Study

Gastric cancer is frequently associated with chemotherapy-induced myelotoxicitydue to the fact that subtotal gastrectomy or total gastrectomy might lead to poorpostoperative nutrition status and conditions making patients become poor-tolerated tohematological toxicity from chemotherapy. Currently doublet and triplet regimenshave been widely used in the treatment for gastric cancer. However, evidence-baseddata on frequency, severity and risk factors of myelotoxicity that is associated withthose commonly used regimens are scarcely available. Additionally, identification ofsubgroup of gastric cancer patients with high risks of developing myelotoxicity eventsand would benefit from early intervention is of utmost importance to improve patients’life quality.The major objectives of our study were to:1. Define the prevalence, severity andrisk factors of myelotoxicity in gastric cancer patients before the initiation ofchemotherapy;2. Explore the frequency, incidence, severity and predictors ofmyelotoxicity in gastric cancer patients undertaking six commonly used regimens (i.e.CF/XP; EC(O)F/EC(O)X; DC(O)F/DC(O)X; PC(O)F/PC(O)X; FOLFOX4;mFOLFOX7/XELOX) and myelotoxicity profile for each regimen;3. Analyze theclinical impacts of myelotoxicity on chemotherapy delivery;4. Examine the clinicalintervention targeting on chemotherapy-induced myelotoxicity among Chinese gastriccancer population.Patients and method: A multicenter, prospective, observational study wasconducted between March2010and January2014, including three major regions inChina (Northeastern region, East coastal region and Southern region). Data of1,285gastric cancer patients undertaking six commonly used regimens with both curativeand palliative intention were prospectively collected in our study. The definition ofanaemia in our study was hemoglobin (Hb)<12.0g/dL, thrombocytopenia wasdefined as platelet (PLT) count<125×109/L, neutropenia was labeled as absoluteneutrophil count (ANC)<2.0×109/L and leukopenia was defined as white blood count(WBC)<4.0×109/L. Myelotoxicity events (i.e. anaemia, thrombocytopenia,neutropenia and leukopenia) were further classified as mild, moderate and severe inaccordance with toxicity grading criteria from the National Cancer Institute.Myelotoxicity was defined as the presence of anaemia, and/or thrombocytopeniaand/or neutropenia and/or leukopenia. Myelotoxicity was further classified based on the most severe myelotoxicity events that patient ever experienced (e.g. If onepresented with mild anaemia, moderate thrombocytopenia, severe neutropenia andmoderate leukopenia during the study period, then he/she was considered toexperience with severe myelotoxicity). All of the statistical analyses were performedusing SPSS version22.0and SAS package version9.1.2. Group comparisons werebased on Chi-square test for categorical variables. Multivariate logistic regressionanalysis was assessed to estimate the odds ratios and identify the predictive factorsindependently associated with the frequency of each myelotoxicity indices.Differences were assumed to be statistically significant when P value was less than0.05.Results:Part1. Epidemiological study of myelotoxicity among Chinese gastric cancerpopulation at baseline and risk factors associated with relevant myelotoxicity indices.The prevalence of anaemia (Hb＜12.0g/dL) before chemotherapy initiation was46.5%. Most patients (33.1%) had mid anaemia; moderate anaemia was recorded for12.4%and severe anaemia for1.1%of all the patients. The prevalence of neutropeniaand leukopenia were14.9%and13.4%, respectively, much lower than that of anaemia.Most patients with neutropenia and leukopenia were mild.4.7%of the whole studypopulation had thrombocytopenia PLT count＜125×109/L,4.5%patients had mildthrombocytopenia and the prevalence of severe thrombocytopenia was0.2%.Multivariate logistic regression analysis showed that patient-specific factorsincluding, female, baseline PLT count (<125×109/L), baseline ANC count<2.0×109/L,baseline KPS<80, comorbidities with Hypertension (HTN)/Diabetes mellitus(DM)/Liver disease(LD)/Chronic obstructive pulmonary disease (COPD), anddisease-specific factors including gastric esophageal junction (GEJ) cancer, poorlydifferentiated/undifferentiated, late stage, multiple metastatic sites were associatedwith the presence of anaemia at baseline. Locally advanced/recurrent/metastaticcancer and previously treated with surgery were the risk factors for severe anaemia.Patients with locally advanced/recurrent/metastatic cancer were more likely todevelop severe anaemia than patients free from tumor (adjusted OR,9.56;95%CI:1.84-49.53, P=0.007). Patients treated with surgery had higher risk to develop severeanaemia than those who did not received surgery before (adjusted OR,15.69;95%CI: 2.37-19.68, P=0.001). Taking802patients who received D2resection into analysis,we found that resection extension and gastric intestinal (GI) construction method wereassociated with moderate-to-severe anaemia. Patients treated with total gastrectomyhad higher risk to develop moderate-to-severe anaemia than those treated withsubtotal gastrectomy (adjusted OR,6.89;95%CI,5.71-8.14, P=0.000). Patients whoreceived Billroth II bypass were more like to develop moderate-to-severe anaemiathan those who received Billroth I bypass (adjusted OR,4.40,95%CI1.92-10.10,P=0.001). Patients with Billroth II bypass also had higher risk to developmoderate-to-severe anaemia than those with Roux-en Y/Interposition Reconstruction(IR) bypass (adjusted OR,7.52;95%CI2.47-11.03, P=0.000). Additionally, riskfactors including previously treated with Platinum-based chemotherapy, previouslyreceived radiotherapy, comorbidities with HTN/DM/LD/COPD, and baseline PLTcount(<125×109/L) were associated with moderate-to-severe anaemia among patientswho received D2resection. Disease status, previous treatment history withradiotherapy, baseline Hb count<12.0g/dL, baseline WBC<4.0×109/L and tumordifferentiation were found to be related with the prevalence of thrombocytopenia atbaseline.Part2. Incidence and frequency of chemotherapy-induced myelotoxicity duringchemotherapy and risk factors associated with the development of myelotoxicityindices.The main epidemiological feature of All-Grade myelotoxicity events for sixregimens was the increase observed during chemotherapy. The frequency ofmyelotoxicity events was higher in patients undertaking triplet regimen than that inthose receiving doublet regimen. Patients treated with palliative intention were morefrequent with myelotoxicity events than those treated with curative intention. Amongthe six established regimens, DC(O)F/DC(O)X and PC(O)F/PC(O)X regimens wereassociated with severe myelotoxicity, while, FOLFOX4and mFOLFOX7/XELOXregimens were better tolerated than others. Multivariate regression analysis found thatbaseline Hb count<12.0g/dL, baseline ANC count<2.0×109/L, baseline KPS<80,primary tumor located at GEJ, previously treated with surgery, administrated withtriplet regimen, number of cycles received≥4, palliative intention and regimens arepredictors related to the development of severe myelotoxicity during chemotherapy.Part3. Clinical impacts of chemotherapy-induced myelotoxicity on treatment delivery among Chinese gastric cancer populationThe frequency of treatment delay≥7days were21.1%and dose reduction≥20%were recorded in16.7%of the whole study population. Patients receiving doubletregimen were significantly less frequent with treatment delay≥7days than thosereceiving triplet regimen (13.4%vs.31.2%, P=0.000). Patients undertaking tripletregimen were found to be more frequent with dose reduction≥20%than thosereceiving doublet regimen (26.2%vs.9.7%, P=0.000).Part4. Treatment rate of chemotherapy-induced myelotoxicity among Chinesegastric cancer population.Treatment rate for chemotherapy-induced myelotoxicity were much lower inChina than that in other western countries. Anaemia treatment rate was30.1%andmost patients with mid-to-moderate anaemia had never received any treatmenttargeting anaemia throughout the study period. The treatment rate forchemotherapy-induced thrombocytopenia was26.0%and treatment rate forchemotherapy-induced neutropenia and leukopenia were83.1%and60.7%,retrospectively.Conclusion:1. Approximately half of the study population had anaemia before the initiationof chemotherapy and most anaemia were mild. The prevalence of thrombocytopenia,neutropenia and leukopenia was low at baseline.2. The main epidemiological feature of All-Grade myelotoxicity events for sixregimens was the increase observed during chemotherapy. The frequency ofmyelotoxicity events was higher in patients undertaking triplet regimen than that inthose receiving doublet regimen. Patients treated with palliative intention were morefrequent with myelotoxicity events than those treated with curative intention. Amongthe six established regimens, DC(O)F/DC(O)X and PC(O)F/PC(O)X regimens wereassociated with severe myelotoxicity, while, FOLFOX4and mFOLFOX7/XELOXregimens were better tolerated than others.3. During the study period, nearly one in five of the patients experiencedtreatment delay≥7days or dose reduction≥20%. Patients undertaking triplet regimenwere found to be significantly more frequent with treatment delay≥7days and dosereduction≥20%than those receiving doublet regimen.4. Treatment rate for chemotherapy-induced myelotoxicity were much lower in China than that in other western countries. More attention should be paid to thepopulation with high risks of developing chemotherapy-induced myelotoxicity. Earlyintervention and management should be optimized in an effort to improve patients’quality of life.