Purpose: Despite high NPV of negative iPET in patients with HL about 10% of them relapse/progress after completion of ABVD. Here we present clinical characteristics and outcome of such patients comparing them to patients with positive iPET and interim treatment escalation [iPET(+)&TE]. Patients and Methods: Data of the Polish patients were retrieved from the prospective observational study of the Polish Lymphoma Research Group on the predictive role of iPET after 1 ABVD cycle. Briefly, intermediate and advanced stage (IIB-IVB) HL patients were treated with ABVD followed by iPET after each of the first 2 ABVD cycles. iPET(-) patients continued with ABVD, while some iPET2(+) had at the discretion of a local physician treatment intensification with BEACOPP esc. with a median of 6 (2-8) cycles. iPET(-) patients failing ABVD had second line treatment with ICE(3), IGEV(6), DHAP(7),ESHAP(5), BEACOPP esc(3), ABVD/MOPP(2) Brentuximab(1), IVOX(1), ASCT(1). iPETs were interpreted according to the Deauville scale as negative(-)(score 1,2,3) and positive (+)(score 4,5). The primary endpoints were: overall survival (OS) and time (EFS2) to the second event (no complete remission or allogeneic HCT at the day of last follow-up), respectively. The secondary endpoint: time to first event (2^nd progression for iPET1(-) or 1^st one for iPET(+)&TE after treatment escalation) Results: After a median follow-up of 48.8 (range 15 – 84) months 29 iPET(-) failing ABVD and 26 iPET(+)&TE patients were identified. (Table 1.) The two cohorts were not different with regard to sex, age, stage, histological subtype, presence of: B symptoms, bulky mass and extranodal disease, absolute number of lymphocytes, lymphocyte to monocyte ratio, and type of relapse. Six(21%) iPET(-) and 4(15%) iPET(+)&TE died. However, in the former group deaths occurred during the longer follow-up, whereas in the latter occurred in the 25 months after treatment onset (Figure1). Similarly more patients experienced S2 events in iPET(-) group (12-41% vs. 8-31%) that tended to occur within longer time in iPET(-) patients. This translated to a significant difference (p=0,04) in the probability of EFS2 events between the two cohorts as shown by the landmark analysis @25 months. Conclusion: Patients with iPET(-) interim BEACOPP escalation despite iPET(+) may experience better long-term OS and significantly better EFS than patients with iPET(-) failing ABVD. Markers to improve NPV of iPET are warranted.