Cryptogenic fibrosing alveolitis: response to corticosteroid treatment and its effect on survival.

We have studied retrospectively 220 patients with cryptogenic fibrosing alveolitis (CFA) who were first seen between 1955-73 and had been followed up for at least four years until 1977. Seventy-seven patients had received no treatment and 143 had received corticosteroids. The only clinical difference between the groups was the age at presentation (untreated mean age 61 years±11 SD; treated mean age 56 years±11 SD p <0·001). The influence of corticosteroid treatment has been assessed both in terms of the clinical, radiographic, and physiological response after four to eight weeks and in terms of survival, using a log rank method of analysis. Of 143 treated cases 127 had detailed follow-up information. Seventy-two (57%) had substantial subjective improvement of breathlessness but only 22 (17%) of the total showed additional objective improvement. The only factors significantly distinguishing the 72 subjective responders from the 55 non-responders, and which therefore might be used as predictors of response, were a younger age at presentation (p <0·001) and less dyspnoea (p <0·02). Analysis of survival data demonstrated that treated patients who showed an early good response to steroids had a markedly better survival than non-responders (p <0·001). The survival curve for the untreated group lay between the two steroid groups. Analysis of the survival data suggests that steroid responsiveness relates particularly to a subgroup with an inherently better prognosis, which is characterised by a younger age and probably by a cellular histology not however necessarily typical of desquamative interstitial pneumonia. Nevertheless treatment appears to have an additional influence on survival especially in early disease, which is not observed in the untreated cases. A large number of factors are not related either to steroid responsiveness or to survival and these include the presence of associated “connective tissue” diseases, autoantibodies, and the duration of dyspnoea at presentation. It is concluded that biopsy information is valuable in assessment of potential to respond to steroids and that if steroid treatment is to be helpful it must be used at an early stage. The response is very variable and titration of the steroid dose against an individual patient's response is likely to give better results than adherence to arbitrary dose schedules.