Construction Of Perioperative Risk Assessment For Surgical Treatment Of Chronic Pulmonary Aspergillosis

Part Ⅰ: Patient selection and risk factors for surgery of chronic pulmonary aspergillosisBackground: Chronic pulmonary aspergillosis(CPA)is a chronic invasive infection caused by Aspergillus.It usually forms saprophytic fungal bulbs in preexisting cavities induced by underlying lung lesions,whereas the efficacy of antifungal medication is limited.Recurrent or fatal haemoptysis may occur as PA erodes the intercostal artery.Complete resection,including lobectomy,helps eradicate lifethreatening symptoms and treat the drug resistant lesion.The clinical features and prognosis of PA also require more observation.Methods: Patients who underwent surgery at Shanghai Pulmonary Hospital from January 2004 to December 2019 were collected.CPA was classified according to CT scan images.Simple aspergilloma(SA)refered to isolated thin-walled(≤3 mm)cavity lesions surrounded by normal lung parenchyma,while complex aspergilloma(CA)refered to a lesion within a thick-walled(> 3 mm)cyst invading surrounded lung parenchyma or pleura.Results: A total of 378 patients were enrolled in this study,including 166(43.9%)patients with CA and 212 patients with SA(56.1%).The average lesion size in the CA group was 4.6±1.4 cm,and in the SA group was 2.7±0.9cm(p<0.001).The most common symptoms were hemoptysis in 192(50.8%)cases,49 cases(13%)received preoperative BAE treatment,133 cases(35.2%)were combined with tuberculosis,and more in the CA group than those in the SA group(p < 0.05).There were 3 perioperative deaths(0.8%),all of which occurred in the CA group.A total of 14(3.7%)cases suffered from complication during the follow-up,which was significantly higher in the CA group compared with the SA group(7.8% vs 0.5%,p=0.002).In the SA group,after propensity score matching for lobectomy and sublobectomy,there was no significant difference in the operation time,intraoperative bleeding,hospital stay time,drainage time,perioperative complications and postoperative recurrence.During the follow-up period,there were 16 relapses(4.2%)and 19 deaths(5.0%),both were significantly higher in the CA group than in the SA group(8.4% vs 0.9%,8.4% vs 2.4%,p<0.05).The overall 5-year survival rate was 96.1%,which was 94.1% and 97.6% in the CA group and the SA group,respectively(p=0.028).Cox multivariate analysis suggested that sex(HR [95%CI] =0.10[0.02-0.50])and age(HR [95%CI] =1.09[1.04-1.14])are risk factors for overall survival.Conclusion: For patients with CPA,surgery is an exact and effective treatment with acceptable perioperative mortality,long-term survival and recurrence rates.SA exhibited significantly better prognosis than CA.Female and younger patients exhibited better long-term survival.Selective sublobectomy in patients with SA also obtained good safety and prognosis,while it still need further investigation.Part Ⅱ: Study of perioperative risk assessment for surgical treatment of chronic pulmonary aspergillosisBackground: Surgery is currently one option for treating CPA,whereas it can lead to serious complications and even pose a fatal threat.Data mining can help us extract valuable information from medically complex data and help with clinical decision-making.Methods: CPA patients who underwent surgery at Shanghai Pulmonary Hospital from January 2004 to December 2019 were divided into SA group and CA group.Moreover,the CA group was further divided into destroyed lung subgroup and nondestroyed lung subgroup.Complications and mortality were counted within 30 days after surgery or during the same hospital stay.Machine learning based on statistical methods,including logistic regression,random forest,BP neural network,and decision tree analysis,was performed to calculate the optimal model for predicting postoperative complications of CPA.Results: Of the 378 patients,there were 166(43.9%)in the CA group and 212(56.1%)in the SA group.A total of 67(17.7%)cases suffered from complication,which was obviously higher in the CA group compared with SA group(30.7% vs 7.5%,p< 0.001).The complication rates of massive hemorrhagic drainage(10.2% vs 1.4%,p<0.001),continuous postoperative air leakage(13.8% vs 4.2%,p=0.002),empyema(6.0% vs 0,p=0.001),re-operation(4.8% vs 0.5%,p=0.014)and atelectasis(4.2% vs 0,p=0.008)in the CA group were significantly higher than those in the SA group.CA group presented markedly longer postoperative drainage time(11.4±8.9 vs 5.7±2.6 d,p<0.001)and postoperative length of stay(10.9±6.1 vs 6.4±2.5 d,p<0.001).There were 3 perioperative deaths(0.8%)in the CA group,all of which were postoperative infections.There was no statistical difference in mortality between the two groups(p=0.167).Logistic multivariate analysis suggested that sex(HR [95%CI] =0.45[0.22-0.89])and extensive pleural adhesions(HR [95%CI] =3.70[1.73-8.09])were risk factors for complications.Of the 35 patients with lung damage in the CA group,18 cases(51.4%)underwent total pulmonary resection,which was obviousy higher than that in the ordinary CA subgroup,which only occurred in 8 cases(6.1%,p<0.001),and there was a significant difference in complications between the two groups(48.6% vs 26%,p=0.01).In the four postoperative complication prediction models,the recall value of logistic regression,random forest,BP neural network,and decision tree were 0.75,0.4375,0.3750 and 0.3750,respectively.Moreover,their AUC values were 0.8568,0.8544,0.7397 and 0.7464,respectively.Conclusion: CPA surgery exhibits acceptable perioperative complication,with CA group was significantly higher than SA group.Male and extensive pleural adhesions are high risk factors for complications.CPA patients with destroyed lung suffer from a high proportion of pneumonectomy and perioperative complications,whose surgical selection needs more caution.The risk assessment model established by logistic regression method can effectively predict the occurrence of perioperative complications of CPA.Part Ⅲ: The necessity of antifungal therapy in the postoperative treatment of pulmonary aspergillomaBackground: Pulmonary aspergillosis is an opportunistic pulmonary mycosis induced by lung fungi colonization or infection.Surgery is one of the therapeutic options for CPA.However,the necessity of adjuvant antifungal treatment after pulmonary surgery for CPA is still questionable,especially in terms of regimens,duration and timescale.Antifungal therapy might be needed to decrease recurrence of PA in the cases with indefinitive surgical margin.Nevertheless,there is lack of supportive evidence about the role of adjuvant antifungal agents after complete resection without spillage of fungal material.The aim of this current retrospective study was to assess the necessity of postoperative antifungal therapy for CPA patients who underwent lung resection.Methods: We retrospectively enrolled 335 PA patients who underwent surgery between Jan 2004 and Dec 2018,including 154 cases of CA and 181 cases of SA.Of these,66 patients were prescribed antifungal agents during the postoperative period(group A),as the regimen was oral therapy with itraconazole or voriconazole for at least 4 weeks.The other 269 were treated with surgery alone without any antifungal therapy postoperatively(group B).The perioperative complications and prognosis of the two groups were analyzed and compared after propensity score matching for sex,age,CPA type,pulmonary function,comorbidities,clinical symptoms and preoperative medications.Results: A total of 60 cases in group A(containing 35 cases of CA)were matched with 189 cases in group B(containing 103 cases of CA)by propensity score.The two groups showed similar clinical features,pulmonary functions,immune status,and operative methods.The overall complication rate was 10.0%(25 cases),including 7 cases in group A and 18 cases in group B(11.7% vs 9.5%,p>0.05).During the median follow-up period of 46.7(41.0-52.5)months,there was no recurrence in both groups of SA patients.13 cases(12.6%)of CA patients in group B suffered from recurrence,which was slightly higher than CA patients in group A(1case,2.9%),but lack of statistical difference(p = 0.08).Conclusions: For patients with SA,the postoperative antifungal therapy did not have a significant effect on disease recurrence.On the other hand,for CA patients,postoperative adjuvant antifungal therapy can reduce the risk of recurrence,but it is limited to a small sample size,which needs to be clarified by further research.