For continuous variables, mean, median, minimum, and maximum will be used for the descriptive purpose. For categorical variables, frequency and percentage will be used for descriptive statistics. The variables of OS will be estimated by the Kaplan-Meier method. Differences between groups will be calculated using the log-rank test for univariate analysis. Cox's proportional hazards model will be employed to test independent prognostic factors. All calculations will be performed using the Statistical Package of Social Sciences software, version 17.0 (SPSS, Inc., Chicago, IL, USA). The level of statistical significance will be set at 0.05 for all tests.

This is an open-label, single arm, prospective, multiple-center phase Ib/II study.

In phase Ib, a maximum of 18 patients will be enrolled in the dose-finding period to determine the RP2D, dependent on the toxicity of the study drugs. The response rates of salvage chemotherapy regimen in R/R DLBCL were reported to be within the range of 35-65%.3 Assuming that the response rate of traditional salvage therapy has an overall response rate of 35% (H0=0.35) and that our experimental regimen would increase it to 50% (H1=0.50), the estimated sample size according to the Simon's two-stage minimal design will be 49 in the second phase.1 Briefly, among the 31 patients recruited during the first stage, the response should be seen in at least 10 of the cases, the lowest threshold for the trial to be moved into the second stage of the phase II part. Furthermore, for the RPR regimen to be deemed effective, a response must be demonstrated in at least 21 of the whole 49 patients enrolled in the phase II period. This would achieve a power of 80%, with a two-sided type I error rate of 10%. An additional 7 patients will be recruited, for an estimated dropout rate of 10%.

Inclusion Criteria:

1. Patients with relapsed/refractory CD20+ diffuse large B-cell lymphoma.
2. Age greater than 20 years and younger than 75 years old.
3. Measurable disease
4. Patients must have an ECOG performance status of less than or equal to 2.
5. Patients must have recovered from toxic effects of all prior therapy before entering onto study.
6. A treatment of drug-free interval of at least 3 weeks since the last dose of chemotherapy is required.
7. More than 4 weeks since prior radiotherapy is required.
8. Adequate bone marrow function
9. Adequate renal function with calculated glomerular filtration rate > 15 mL/min
10. Patients must have adequate liver function
11. All patients must sign a document of informed consent indicating their awareness of the investigational nature and the risks of the study.

Exclusion Criteria:

1. Patients who have prior treatment with ruxolitinib or taxane for DLBCL.
2. Pregnant or breast-feeding females.
3. Active or uncontrolled infection.
4. Life expectancy < 6 months
5. Patients with brain or leptomeningeal metastases.
6. Known hypersensitivity to ruxolitinib or paclitaxel
7. Grade III peripheral neuropathy secondary to prior to therapy
8. Second malignancy, except indolent cancers not on active anti-cancer therapy.