Lim Tien Huey, Gane Edward, Moyes Chris, Borman Barry, Cunningham Chris
Department of Gastroenterology, Middlemore Hospital, Auckland, New Zealand.
New Zealand Liver Transplant Unit, Auckland City Hospital, Private Bag 92024, Auckland, New Zealand.
Hepatol Int. 2016 Sep;10(5):829-37. doi: 10.1007/s12072-016-9709-6. Epub 2016 Mar 8.
HBsAg seroclearance is the most desired endpoint in chronic hepatitis B (CHB) but occurs uncommonly. Recent studies have shown baseline HBsAg levels to be predictive of HBsAg loss up to 10 years. We report the 28-year rates of HBsAg loss and outcomes in the Kawerau study cohort from New Zealand, and assess the predictive value of baseline HBsAg levels to predict long-term HBsAg loss.
The 1984 Kawerau community study identified 572 CHB patients, followed up for 28 years (41 % HBeAg-positive, median age 17 years, range 1-71 years). In 2012, surviving individuals attended a local clinic for an interview, blood tests and transient elastography.
384/218 (74 %) surviving individuals attended the clinic in 2012. Spontaneous HBsAg loss occurred in 145 (33 %) after 12,702 person-years of follow-up (1.14 per 100 person-years). Liver stiffness measurements were significantly lower if HBsAg loss occurred <50 years (mean 6.1 kPa) versus >50 years (mean 11.6 kPa), p = 0.0002. No HCC occurred following HBsAg loss (median follow-up 72 months). Predictors of HBsAg loss were older age and lower baseline HBsAg level (HR for HBsAg loss at 28 years 2.7 (95 % CI 1.7-4.2), 6.7 (95 % CI 3.9-11.4) and 9.4 (95 % CI 5.2-16.9), respectively, for HBsAg 1000-9999, 100-999 and <100 IU/mL compared to HBsAg >10,000 IU/mL at baseline, (p < 0.0001). Baseline HBsAg was a superior predictor of HBsAg loss compared to HBV DNA at all time-points: AUROC at 15 years: 0.87 (95 % CI 0.82-0.93) versus 0.73 (95 % CI 0.66-0.80) (p < 0.0001) and AUROC at 28 years: 0.74 (95 % CI 0.69-0.79) versus 0.67 (95 % CI 0.62-0.72) (p = 0.0007). The optimal cut-off HBsAg level to predict HBsAg seroclearance at 28 years is HBsAg <10,000 IU/mL (sensitivity 72 %, specificity 64 %, NPV 88 %).
Rates of HBsAg loss in our community cohort were high, and occurred earlier than previously reported. Earlier HBsAg loss was associated with less severe liver fibrosis. Baseline HBsAg level was a good predictor of long-term HBsAg loss up to 28 years and superior to HBV DNA.
HBsAg血清学清除是慢性乙型肝炎（CHB）最理想的终点，但并不常见。最近的研究表明，基线HBsAg水平可预测长达10年的HBsAg消失情况。我们报告了来自新西兰卡韦劳研究队列的28年HBsAg消失率及转归情况，并评估基线HBsAg水平预测长期HBsAg消失的价值。
1984年卡韦劳社区研究纳入了572例CHB患者，随访28年（41%HBeAg阳性，中位年龄17岁，范围1 - 71岁）。2012年，存活个体前往当地诊所接受访谈、血液检测和瞬时弹性成像检查。
2012年有384/218（74%）存活个体前往诊所。经过12702人年的随访，145例（33%）出现自发HBsAg消失（每100人年1.14例）。如果HBsAg在50岁之前消失，肝脏硬度测量值显著低于50岁之后消失的情况（平均6.1 kPa对比平均11.6 kPa，p = 0.0002）。HBsAg消失后未发生肝癌（中位随访72个月）。HBsAg消失的预测因素为年龄较大和较低的基线HBsAg水平（与基线HBsAg>10000 IU/mL相比，28年时HBsAg消失的HR分别为：HBsAg 1000 - 9999 IU/mL为2.7（95%CI 1.7 - 4.2），100 - 999 IU/mL为6.7（95%CI 3.9 - 11.4），<100 IU/mL为9.4（95%CI 5.2 - 16.9），（p < 0.0001）。在所有时间点，基线HBsAg比HBV DNA是更好的HBsAg消失预测指标：15年时的AUROC为0.87（95%CI 0.82 - 0.93）对比0.73（95%CI 0.66 - 0.80）（p < 0.0001），28年时的AUROC为0.74（95%CI 0.69 - 0.79）对比0.67（95%CI 0.62 - 0.72）（p = 0.0007）。预测28年HBsAg血清学清除的最佳HBsAg水平临界值为HBsAg<10000 IU/mL（敏感性72%，特异性64%，阴性预测值88%）。
我们社区队列中的HBsAg消失率较高，且比之前报道的出现得更早。更早的HBsAg消失与较轻的肝纤维化相关。基线HBsAg水平是长达28年长期HBsAg消失的良好预测指标，优于HBV DNA。
