Consecutive newly diagnosed DLBCL patients were screened from March 2013 to June 2020 at Zhejiang Cancer Hospital. The DLBCL diagnosis for these patients was confirmed via pathological review as performed by an independent experienced pathologist. Disease stage was judged according to the criteria of Lugano 2014 (11). First-line treatment consisted of at least four cycles of rituximab-containing anthracycline-based chemotherapy. Patients that completed fewer than four cycles were excluded. All patients underwent baseline whole-body PET-CT scans within four weeks before starting therapy, iPET-CT scans after four cycles of chemotherapy, and ePET-CT scans conducted within eight weeks after the completion of chemotherapy. Responses to chemotherapy were evaluated based upon the revised criteria published by Cheson et al. (12). The Deauville score (DS) was employed for measuring ¹⁸F-FDG-uptake in PET-CT (13). A DS 1 to 3 was defined as PET negativity. DS 4 or DS 5 were used to define PET positivity. After completion of first-line chemotherapy, all patients underwent regular follow-up CT scans every 3 months over the first two years, every 6 months for the next three years, and once a year from the sixth year onward. A retrospective analysis of data extracted from patient electronic medical records including demographic information, pathological features, treatment regimens, therapeutic responses to initial or salvage chemotherapy, and survival was performed. The Zhejiang Cancer Hospital ethics committee approved this study, which was consistent with the Declaration of Helsinki.

PFS was calculated from the start of first-line chemotherapy to the first recording of disease progression or disease relapse or death. OS was defined as the period from the start of first-line chemotherapy to the date of death from any cause or the last follow-up. Categorical variables are given as proportions and were analyzed with chi-squared tests and Fisher’s exact test. Continuous variables are given as medians and ranges. PFS and OS were calculated using the Kaplan–Meier survival method and log-rank tests. Univariable and multivariable Cox regression analyses were performed to determine the independent factors affecting PFS or OS. P <0.05 was the threshold of significance. To further explore exact survival differences, survival time distributions in four groups were compared pairwise. A Bonferroni corrected p-value was applied to the multifactorial logistic regression p–values to account for the multiple testing of six different comparisons (corrected α = 0.05/6 = 0.00833). Statistical analyses were performed with Statistical Package for Social Sciences (SPSS), version 24. Survival curves were drawn with GraphPad Prism 8.

Initially, 505 total patients diagnosed with DLBCL were identified, of whom 198 were excluded due to ambiguous diagnoses (n=4), fewer than 4 chemotherapy cycles (n=12), or a lack of available iPET-CT or ePET-CT data (n=182). Therefore, 307 patients were analyzed in this study (Figure 1). Patient baseline clinical characteristics are shown in Table 1.

All 307 patients underwent initial pretreatment PET-CT and iPET-CT scanning (Figure 1). At iPET-CT evaluation, 250 patients achieved complete response (CR) and the proportion of patients with negative metabolic uptake was 81.4% (250/307). Moreover, 45 patients achieved partial response (PR), all of whom continued to complete prior chemotherapy regimens for at least 2 cycles, and 15 of them (33.3%) achieved CR at ePET-CT. Twenty-six patients maintained PR, while 4 patients ultimately exhibited progressive disease (PD). Additionally, 12 patients exhibited SD/PD at iPET-CT, of whom just 3 underwent biopsy and 2 were confirmed to have progressive disease. Of these 12 patients, 10 underwent second-line treatment, while one underwent palliative radiotherapy. The remaining patient did not receive any treatment, and died 5 months later.

At time of ePET-CT evaluation (n=295), 262 patients (88.8%) achieved CR and were considered as negative ePET-CT, whereas 33 patients (11.2%) exhibited ePET-CT positivity, including 26 patients with PR and 7 patients with PD. Among the 26 patients with PR at time of ePET-CT, 10 received second-line chemotherapy and 2 of them underwent subsequent autologous stem-cell transplantation (ASCT) with no evidence of disease. Eight patients received palliative radiotherapy for residual lesions without chemotherapy. Another 8 patients did not receive any treatment, and 7 of them were still alive. All 7 patients with PD at ePET-CT received salvage chemotherapy, but only 3 patients remained alive at last follow-up.

After a median follow-up of 45.1 months (range: 5.1 - 100 months), 81 patients (26.4%) experienced disease progression or relapse, and 36 patients (11.7%) were censored due to death. The 2-year PFS rate and 2-year OS rate for the whole cohort (n=307) were 76.6% (95% confidence interval (CI), 71.8 to 81.4%) and 91.0% (95% CI, 87.7 to 94.2%), respectively.

The iPET-CT and ePET-CT results for these patients were both significantly associated with survival outcomes (Figure 2). The 2-year PFS and 2-year OS for patients with iPET-CT positivity were 50.7% (95%CI, 37.6 to 63.8%) and 76.5% (95%CI, 65.3 to 87.7%), respectively, and were significantly shorter than those for patients with iPET-CT negativity (2-year PFS: 82.7% (95% CI, 78 to 87.4%), p<0.001; 2-year OS: 94.2% (95% CI, 91.3 to 97.1%), p<0.001). The survival outcomes for patients with SD/PD at iPET-CT were extremely poor, with median PFS and OS were only 3.2 months and 11.0 months, respectively. Similarly, patients with ePET-CT positivity had a significantly poorer 2-year PFS (48.1%, 95% CI, 30.9 to 65.3%) and 2-year OS (78.5%, 95% CI, 64.4 to 92.6%) rates compared with those of patients with ePET-CT negativity (2-year PFS: 83.8% (95% CI, 79.3 to 88.3%), p<0.001; 2-year OS: 94.9% (95% CI, 92.2 to 97.6%), p<0.001).

These results suggest that there are significant relationships between PET avidity at different follow-up time points and DLBCL patient survival. In light of these results, we conducted a further examination of the prognostic value of the speed of metabolic response. As patients with SD/PD at iPET-CT began undergoing second-line chemotherapy and lacked available ePET-CT scans, so they were excluded in this section. The remaining 295 patients were divided into the following 4 groups: EMR (early metabolic responders, iPET-CR+ePET-CR, n=247), DMR (delayed metabolic responders, iPET-PR+ePET-CR, n=15), IMR (incomplete metabolic responders, iPET-PR+ePET-PR, n=26), and MP (metabolic progressors, iPET-CR/PR+ePET-PD, n=7). The 2-year PFS rates were significantly different in these four groups (83.7%, 86.2%, 61.1%, and 0%, respectively; p<0.001). The 2-year OS rates were also significantly different in these four groups (94.6%, 100%, 84.3%, and 57.1%, respectively; p<0.001). The survival distribution of the four groups was compared in a pairwise manner. For 2-year PFS rate, there was a significant difference between MP and EMR (p<0.001), DMR (p<0.001), and IMR (p<0.001). There was also a difference between EMR and IMR (p=0.002). Between the other groups, no significant difference was found (p>0.0083). For 2-year OS, there was a difference between MP and EMR (p<0.001), and DMR (p=0.006). No significant difference was found between the other groups (p>0.0083). After Bonferroni correction, results showed a significant prognostic difference between MP and EMR/DMR. In Figure 3, a Kaplan-Meier plot for PFS and OS of the different groups of patients is shown.

The iPET-CT and ePET-CT positivity rates in different international prognostic index (IPI) risk groups were significantly different. Overall, 13.9% (29/209) patients with low/low-intermediate risk exhibited iPET-CT positivity, while 29.6% (29/98) patients with high/high-intermediate risk exhibited iPET-CT positivity (p=0.001). Moreover, 8.8% (18/205) patients with low/low-intermediate risk exhibited ePET-CT positivity, while 16.7% (15/90) patients with high/high-intermediate risk exhibited ePET-CT positivity (p=0.048). Additionally, bulky nodes (> 5 cm) and elevated serum C reactive protein (CRP) were more common in patients with positive iPET-CT (28.2% vs 15.7%, p=0.015; 26.1% vs 14.6%, p=0.013).

Factors including iPET-CT (positivity vs negativity), ePET-CT (positivity vs negativity) and IPI (high/high-intermediate vs low/low-intermediate) were analyzed in univariable and multivariable analysis for potential significance in terms of PFS and OS. In univariable analysis, positive iPET-CT, positive ePET-CT and high/high-intermediate IPI were all associated with inferior PFS and latter two factors were also associated with inferior OS. In multivariable analysis, positive ePET-CT and high/high-intermediate IPI were independent prognostic factors for poor PFS and OS (Table 2).