CALGB 9633 was a randomized trial of observation versus adjuvant chemotherapy for patients with stage IB non-small cell lung cancer (NSCLC). In CALGB 9633, the presence of mucin in the primary tumor was associated with shorter disease-free survival (DFS; hazard ratio (HR)=1.9, p=0.002) and overall survival (OS; HR=1.9, p=0.004).
To validate these results, mucin staining was performed on primary tumor specimens from 780 patients treated on IALT, 351 on JBR.10 and 150 on ANITA. The histochemical technique using mucicarmine was performed. The prognostic value of mucin for DFS and OS was tested in a Cox model stratified by trial and adjusted for clinical and pathological factors. A pooled analysis of all 4 trials was performed for the predictive value of mucin for benefit from adjuvant chemotherapy.
The cross-validation group had 48% squamous, 37% adenocarcinoma and 15% other NSCLC compared with 29%, 56%, and 15% respectively in CALGB. Among 1262 patients with assessable results, mucin was positive in IALT 24%, JBR.10 30%, ANITA 22% compared with 45% in CALGB. Histology was the only significant covariate (p<0.0001) in multivariate analysis with mucin seen more commonly in adenocarcinoma (56%) compared with squamous (5%) and other NSCLC (15%). Mucin was a borderline negative prognostic factor for DFS (HR=1.2 [1.0–1.5], p=0.06) but not significantly so for OS (HR=1.1 [0.9–1.4], p=0.25). Prognostic value did not vary according to histology: HR=1.3 [1.0–1.6] in adenocarcinoma vs. 1.6 [1.2–2.2] for DFS in other histology (interaction p=0.69). Mucin status was not predictive for benefit from adjuvant chemotherapy (test of interaction: DFS p=0.27; OS p=0.49).
Mucin was less frequent in the cross-validation group due to its higher percentage of squamous cell carcinomas. The negative impact of mucin was confirmed for DFS but not for OS. Mucin expression was not predictive of overall survival benefit from adjuvant chemotherapy.