| Background:Up to 2015,an estimation of 257 million people suffered from chronic HBV(hepatitis B virus)infection.Untreated patients with chronic HBV infection can progress to life-threatening complications,including cirrhosis and hepatocellular carcinoma.HBV replication is the most important biomarker for predicting the risk of disease progression and long-term prognosis in patients with chronic HBV infection.Long-term use of nucleos(t)ide analogs(NAs)can effectively inhibit HBV replication,thereby preventing disease progression.At present,the chronic hepatitis B(CHB)management guidelines(or guidance)developed in China and other countries all recommend long-term use of nucleos(t)ide analogues.The rationale for discontinuing antiviral therapy should be based on the following terms:the disappearance or seroconversion of hepatitis B e antigen(HBeAg),even the disappearance of hepatitis B surface antigen(HBsAg),and a long duration of consolidation therapy at the same time.Due to the long-term use of nucleos(t)ide analogs,drug safety is particularly important.Adefovir dipivoxil(ADV)is a nucleotide analogue that is a precursor to adefovir.It was rapidly absorbed and transferred into adefovir after oral absorption into the blood.The latter passed through the glomerular filtration and active secretion of renal tubules,and was excreted through the kidneys.Previous studies have proved the safety of long-term use of ADV in CHB patients.With the prolonged use of ADV andPart Ⅰ:Analysis of clinical characteristics,intervention and prognosis of renal hypophosphatemia and osteomalacia induced by adefovir dipivoxil the increase of the population,several reports on side effects such as hypophosphatemia,osteomalacia and even Fanconi’s syndrome caused by ADV have been reported worldwide.In December 2014,the State Food and Drug Administration also issued the 64th "Notice of Adverse Drug Reactions",suggesting that attention should be paid to the risk of hypophosphatemia and osteomalacia caused by ADV.ADV has been used in a large number of CHB patients in China,it had more specific clinical significance that how to identify ADV-related renal hypophosphatemia and osteomalacia,individualized intervention,and prognosis.Objective:This study aimed to summarize the clinical manifestations of ADV renal side effects and prognosis,proposed relatively specific treatment measures during a long time of follow-up.Methods:The patients enrolled from October 2013 to December 2018 in our hospital,and met the inclusion criteria of "adefovir dipivoxil-related renal hypophosphatemia and osteomalacia" clinical observation.Baseline characteristics were collected.The following indicators were tested:biochemical indices,liver function indices,kidney function and renal injury indices,parathyroid hormone assay,HBV serological test,HBV DNA quantitative test,bone density,whole body bone imaging examination.According to the liver disease status(CHB or cirrhosis)at the time of diagnosis,the history of antiviral medication,drug resistance and mutation sites,renal function and bone mineral density,ADV was replaced with telbivudine(LdT)or entecavir(ETV).According to the severity of the patient’s condition at the time of diagnosis and the patient’s bone pain,calcitriol,phosphorus and calcium supplementation were given.For patients with moderate to severe hypophosphatemia with bone pain and/or bone mineral density suggesting osteoporosis(bone density T value<-2.5),intravenous or oral phosphorus supplementation was given.After the serum phosphorus returned to normal,oral supplementation of calcium carbonate D3 600～1200mg/day was given.Oral calcitriol was administered 0.25-0.50 μg/day for patients with a bone density T value<-2.5 or-1 to-2.5 but significant bone pain.All confirmed patients were told to consume foods rich in phosphorus and calcium.Follow-up observations were performed on patients who agreed to participate.The HBV DNA quantitative test and abnormal indices mentioned above were detected,recorded and processed in time at 2 weeks,1 month,3 months,6 months,9 months and 12 months.Bone mineral density was measured every 3 months and changes in symptoms were recorded at any time.Results:This study included 62 patients with ADV related renal hypophosphatemia,the median age was 55 years old,range from 27 to 76 years old.Forty-two patients were male.There was 34 CHB patients and 28 patients with cirrhosis.Forty-four patients were diagnosed as osteomalacia.The median duration of ADV was 66 months,range from 18 to 116 months.67.7%of patients were treated for 36-96 months.The number of patients with ADV treatment-naive,monotherapy and no resistant was 28.ADV sequential treatment was found in 13 patients with lamivudine(LAM)resistance or poor response.The median serum phosphorus level of 62 patients was 0.675(range:0.41 to 0.79)mmol/L,of which 27 cases(43.5%)were moderate hypophosphatemia;12 patients had hypocalcemia,and 7 patients had hypokalemia.Elevated serum alkaline phosphatase was found in 35 patients,20 of whom had an elevation for 11.5(range:1 to 55)months before diagnosis.Serum parathyroid hormone level was normal in all patients.The abnormal rate of eGFR was higher than that of creatinine(χ2=37.922,P=0.000),and the abnormal rate of cystatin C was higher than that of serum creatinine(χ2=14.762,P=0.000).Twenty-six patients developed bone pain,mainly in the bilateral ribs,heel,femur,knee joint and lumbosacral region.The bone pain was confirmed to last 9.5(1 to 24)months.A decrease in bone density occurred in 44 patients(71%),and the abnormal T value was-3(-5.5 to-1.1).Of the 44 patients with decreased bone density,24 of them had a T value of less than-2.5 and were osteoporotic.Eight patients with severe bone pain underwent a whole-body bone imaging examination with the consent of the patients.Eight patients all showed abnormal bone metabolism in different parts,mainly concentrated in the ribs,lumbar vertebrae,hip joints,femur,knee joints and Ankle joints and other parts.Sixty patients agreed for intervention.Mild hypophosphatemia was found in the other two patients who did not agree for intervention.According to the individualization principle mentioned above,there were 50 patients who discontinued ADV,switched to LdT,ETV or TDF;eight patients reduced the dose of ADV to 10 mg once every other day with LdT adding on;one patient was given LdT combined with ADV.ADV was discontinued in one patient who met the criteria for discontinuation.Twenty-eight patients were given oral calcitriol at the same time.Among them,19 patients received intravenous or oral phosphorus supplementation,and the serum phosphorus level was normal and then supplemented with calcium carbonate D3 600-1200 mg per day.Forty-five patients were followed up after intervention.All patients were followed up for at least 6 months.The number of patients who completed 1 year,2 years,3 years,and 4 years of follow-up were 50,38,31,and 18,respectively.Serum phosphorus was recurring to a normal level in a median of 1(range:0.25 to 12)months.Twenty patients with bone pain agreed to follow-up.Bone pain began to ease at a median of 1(range:0.25 to 4)months,and completely ease at a median of 2(range:0.5 to 6)months.Forty-five patients were followed up for liver function as described above.There was no fluctuation in serum ALT and AST.Of the 35 patients with elevated serum ALP,6 patients were not followed up.One patient was only followed for 3 months,and the serum ALP level did not return to normal.The other 28 patients were followed up for 3,6,9 and 12 months,3、9、14 and 20 patients had normal serum ALP,respectively.8 patients who had not recovered to normal ALP during the follow-up of 12 months,the follow-up was continued until 24 months.Decreased serum ALP level was found in 4 patients at 24 months,which was still abnormal.Of the 54 follow-up patients,44 patients with a baseline eGFR decline had no significant change in eGFR at 1 year of follow-up(Wilcoxon signed rank sum test,Z=-1.491,P=0.136).However,eGFR was improved at the time of follow-up until 2 years(Wilcoxon signed rank sum test,Z=-1.991,P=0.046).The bone density test T value did not improve at 3 months.At 6 months,osteoporosis improved for bone loss was found in 21.4%(3/14)of patients;no further improvement was found at 9 months or 12 months.One patient had no improvement in bone mineral density at 12 months,but a review of whole-body bone imaging revealed abnormal bone metabolism.Eight patients with abnormal bone metabolism were ordered to a re-examination of whole-body bone imaging at 12 months,which indicated that abnormal bone metabolism disappeared.Conclusion:Blood phosphorus can prompt the occurrence of ADV-related renal side effects earlier than serum creatinine;ADV withdrawal and switching to other antiviral drugs or adefovir reduction is the key for this medicine side reaction.Experience of patients’ previous antiviral drugs and drug resistance should be considered for individualized intervention.After individualized treatment,hypophosphatemia and bone pain can be completely corrected,and renal function can be slowly improved.Part Ⅱ:Establishment of predictive model of adefovir related renal hypophosphatemiaBackground:Hepatitis B virus(HBV)is a small,enveloped and closed circular DNA virus of the hepatotropic virus family that replicates and assembles specifically in liver cells in humans.Nucleos(t)ide analogs compete for binding to the active site of HBV DNA polymerase in hepatocytes,inhibit the synthesis of DNA polymerase,and terminate the elongation of viral DNA strands,thereby inhibiting viral replication.Nucleos(t)ide analogs have low levels of activity against human mitochondrial DNA polymerase(mtDNA).When the concentration of the nucleos(t)ide analog exceeds a critical threshold,mtDNAy polymerase activity is inhibited resulting in a decrease of mitochondrial replication,that can lead to mitochondrial dysfunction,including renal toxicity.Long-term use of adefovir dipivoxil(ADV)in patients with chronic hepatitis B(CHB)can lead to kidney injury.Multiple clinical case reports and clinical studies have been confirmed and written into the guidelines.The guidelines for the management of hepatitis B developed by the Asian Pacific Association for the Study of the Liver(APASL)in 2015 and the guidelines for the prevention and treatment of chronic hepatitis B issued in China in 2015 pointed out that patients taking ADV should regularly monitor serum creatinine and phosphorus.In our previous study,patients with hypophosphatemia was more observed than those with elevated serum creatinine.Hypophosphatemia may suggest an ADV related renal dysfunction earlier.We searched PubMed,EMBASE,WanFang,CNKI and other databases until February 2019,it was found that few researches have been done on the incidence of hypophosphatemia induced by ADV.Multiple studies have reached ditferent conclusions in the study of risk factors for ADV related renal dysfunction.If all possible risk factors are analyzed one by one,the process will be complicated,which will increase the complexity of research.Principal component analysis(PCA)method is adopted to reduce the dimensionality of high-dimensional data of multiple variables,which can reduce complex factors into several main variables(principal components),simplify the problem,and ensure that these variables are independent of each other,effectively minimize the loss of information and increase the explanatory value of results.The PCA research method is applicable to a variety of numerical data types.In recent years,it has been widely used in many fields such as physics,protein molecular dynamics,and drug discovery.Linear discriminant analysis(LDA)is a classical linear learning method with supervised learning.It is one of the most widely used and extremely effective methods in dimension reduction and pattern classification.After LDA analysis,the variance between the two groups was the largest,and the variance within the group was the smallest,thus achieving the purpose of distinguishing two groups of different subjects.LDA is widely used,such as machine learning,face recognition,biomedical research and earth science.Objective:This study aimed to study the incidence of ADV related renal hypophosphatemia,used PCA method to study the risk factors of hypophosphatemia,and discussed the establishment of a risk model for hypophosphatemia preliminarily.Methods:Patients were enrolled that received ADV treatment-naive or rescued therapy from January 2010 to December 2018.The general data of patients who started using ADV were collected,as well as the duration of ADV when occurring hypophosphatemia,history of antiviral therapy and drug resistance.The clinical data of patients were collected meanwhile:liver function indices,kidney function and renal injury indices,serum phosphorus,HBV markers and HBV DNA quantitative test.Principal component analysis(PCA)was used to determine the main risk factors for hypophosphatemia.Linear discriminant analysis(LDA)was used to calculate the occurrence of hypophosphatemiaResults:248 patients were enrolled in the study,146 patients with CHB and 102 patients with cirrhosis;73%of them(181 patients)were male in the study.The mean age of patients was 49.6±12.6 years.The median duration of ADV was 53.5 months(range:3 to 103 months);there were 122 patients taking ADV monotherapy,no drug resistance,no other antiviral drugs combining,and 83 patients taking lamivudine(LAM)combined with ADV(50 of them received rescued therapy as a result of LAM resistance).There were 31 patients taking telbivudine(LdT)combined with ADV and 12 patients receiving entecavir(ETV)combined with ADV.There were 13 patients with hypophosphatemia,and males accounted for the majority(11 cases,84.6%).There were 5 patients with cirrhosis and 8 patients with CHB.The mean age of 13 patients with hypophosphatemia was 49.5±11.7 years,and the median duration of ADV was 49 months(range:18 to 96).One of them had a history of drinking and had stopped drinking when receiving antiviral therapy.Two patients with hypertension had stable blood pressure control after taking antihypertensive drugs.Eleven patients received ADV monotherapy and no drug resistance occurred,and 2 patients were treated with LAM combined with ADV,one of them received ADV rescued therapy as a result of LAM resistance.In 13 patients with hypophosphatemia,the mean serum phosphorus level was 1.0±0.08mmol/L and the mean serum creatinine was 71.3±13.5μmol/L.One patient developed after taking ADV for only 18 months,and most of the other 12 patients(66hypophosphatemia.7%,8/12)occurred after 3 years of ADV therapy.The 5-year cumulative incidence of hypophosphatemia was 12.0%,and it was 42.3%in 9 years.Compared with patients who did not develop hypophosphatemia,patients with hypophosphatemia had lower serum phosphorus before ADV therapy(P=0.015).However,there was no statistically significant difference in gender,age,cirrhosis,combined hypertension,combined lamivudine or serum creatinine between the two groupsPCA method was used to analyze factors that may affect hypophosphatemia:gender,age,cirrhosis,whether or not hypertension,combined use of lamivudine,treatment of baseline serum creatinine,and serum phosphorus.The eigenvalues of the first three variables were all more than 1,but their cumulative contribution rate was only 60%.The three variables above could not be used to predict the occurrence of hypophosphatemia.When the first six variables were included,the cumulative contribution rate was over 90%.There was no obvious separation trend between the two groups,and most cases overlapped.We could not directly determine whether hypophosphatemia would occur based on these six risk factors by the PCA method.Then we used all variables that can collect complete data including age,gender,cirrhosis or not,whether or not hypertension,combination with LAM treatment,treatment of baseline serum creatinine and serum phosphorus,and history of drinking,etc.LDA method was used to establish a linear classification model for hypophosphatemia in the R language environment.The 27 variables above for each patient were brought into the model.The AUROC of subjects was 0.940(95%CI 83.8-94.7).The sensitivity was 88.89%and the specificity was 81.8%of the model.The principle of cross-validated was applied to verify the individual attributes in the cohort.The accuracy of this model for predicting hypophosphatemia was 73%.Conclusions:Long-term therapy of ADV in patients with CHB could cause renal hypophosphatemia.The PCA method could be used to identify some characteristics for people at high risk of hypophosphatemia.The LDA method could be used to establish a model for predicting the occurrence of hypophosphatemia.The accuracy was high through cross-validation. |
PDF Full Text Request