Confidence intervals for odds ratio from multistage randomized phase II trials.

A multi-stage randomized trial design can significantly improve efficiency by allowing early termination of the trial when the experimental arm exhibits either low or high efficacy compared to the control arm during the study. However, proper inference methods are necessary because the underlying distribution of the target statistic changes due to the multi-stage structure. This article focuses on multi-stage randomized phase II trials with a dichotomous outcome, such as treatment response, and proposes exact conditional confidence intervals for the odds ratio. The usual single-stage confidence intervals are invalid when used in multi-stage trials. To address this issue, we propose a linear ordering of all possible outcomes. This ordering is conditioned on the total number of responders in each stage and utilizes the exact conditional distribution function of the outcomes. This approach enables the estimation of an exact confidence interval accounting for the multi-stage designs.

Estimation of the odds ratio from multi-stage randomized trials.

A multi-stage design for a randomized trial is to allow early termination of the study when the experimental arm is found to have low or high efficacy compared to the control during the study. In such a trial, an early stopping rule results in bias in the maximum likelihood estimator of the treatment effect. We consider multi-stage randomized trials on a dichotomous outcome, such as treatment response, and investigate the estimation of the odds ratio. Typically, randomized phase II cancer clinical trials have two-stage designs with small sample sizes, which makes the estimation of odds ratio more challenging. In this paper, we evaluate several existing estimation methods of odds ratio and propose bias-corrected estimators for randomized multi-stage trials, including randomized phase II cancer clinical trials. Through numerical studies, the proposed estimators are shown to have a smaller bias and a smaller mean squared error overall.

Timing of Adjuvant Immunotherapy in Stage III Melanoma, Does it Matter?

BACKGROUND: The optimal time to initiate adjuvant immune checkpoint inhibitors (ICI) following resection remains undefined. Herein, we investigated the impact of time to adjuvant ICI on survival in patients with stage III melanoma. METHODS: Patients with resected stage III melanoma receiving adjuvant immune therapy were identified within a multi-institutional retrospective cohort. Patients were stratified by time to adjuvant ICI: within 6 weeks, 6-12 weeks, and greater than 12 weeks from surgery. Recurrence-free survival (RFS) was compared among time strata with Kaplan-Meier and Cox proportional hazards methods in the multi-institutional cohort. RESULTS: Altogether, 626 patients were identified within the multi-institutional cohort: 39% of patients initiated adjuvant ICI within 6 weeks, 42.2% within 6-12 weeks, and 18.8% greater than 12 weeks from surgery. In a multivariate Cox model, adjusting for histology, nodal tumor burden, and pathologic stage, we found that increased time to adjuvant ICI was associated with improved RFS. Patients who initiated adjuvant ICI within 6 weeks of surgery had worse RFS. These findings were preserved in a conditional landmark analysis and separate subgroups of patients with (1) new melanoma diagnoses, (2) occult stage III disease, and (3) those receiving anti-PD-1 monotherapy. CONCLUSIONS: Outcomes for patients with stage III melanoma are not compromised when adjuvant ICI is initiated beyond 6 weeks from resection. Additional work is needed to better understand the underlying mechanisms and implications of timing of adjuvant ICI on long-term outcomes.

Survival and tumor characteristics of patients presenting with single primary versus second primary melanoma lesions.

BACKGROUND: Patients with single primary melanomas have an increased risk of developing subsequent melanomas. Secondary tumors diagnosed within and after 3 months are termed "synchronous" and "asynchronous," respectively. OBJECTIVE: To compare tumor distributions and survival characteristics between patients with second primary melanomas and those with single primary melanomas. METHODS: Retrospective cohort study. Data were collected from an institutional database from 14,029 patients with a diagnosis of a primary melanoma seen between 1970 and 2004. RESULTS: The synchronous and asynchronous cohorts demonstrated significantly improved survival probabilities compared with the single primary cohort (P = .04 and .002, respectively). Single primary lesions (2.2 ± 2.3 mm) were significantly thicker than the first-identified synchronous (2.0 ± 1.7 mm) and asynchronous (1.7 ± 1.3 mm) lesions. Synchronous lesions were more likely to be anatomically concordant compared with asynchronous lesions (55.7% vs 38.2%, P < .001). LIMITATIONS: Single-center study design and incomplete records for second primary melanoma Breslow depth and histopathology. CONCLUSION: Patients with second primary melanomas demonstrated a significant survival advantage and thinner lesions compared with those with single primary melanomas. Our reported tumor distributions support the role of full body skin examinations, with attention to the region of initial diagnosis.

Optimal designs for phase II clinical trials with heterogeneous patient populations.

We consider single-arm phase II cancer clinical trials with tumor response as the primary outcome. Oftentimes, the patient population of a phase II clinical trial consists of subpopulations with different expected response rates. A well-accepted design in this case is to specify the response rate and the prevalence of each subpopulation, to compute the response rate of the whole population using the weighted (by prevalence) average of the response rates across subpopulations, and to find a standard phase II design, such as Simon's minimax or optimal design, for testing on the response rate of the whole population based on the unstratified binomial test. In such trials, while the response rate is the primary parameter and the prevalence of each subpopulation is a nuisance parameter, the validity of an unstratified statistical test for deciding acceptance or rejection of the experimental treatment is influenced by observed prevalence. In order to avoid bias due to the discrepancy between observed and specified values of the nuisance parameter, we have to use stratified test for such trials. In this paper, we propose optimal and minimax designs for stratified binomial test. We also develop a user-friendly interactive software to visualize the optimal designs and help users make correct statistical decisions.

Antibiotic Prophylaxis for Megaprosthetic Reconstructions: Drug and Dosing May Matter More than Duration.

In orthopedic oncology, the implant of a megaprosthetic device is standard of care after large-scale tumor resection involving segmental removal of bone. Infection remains the leading cause of implant failure, often resulting in major morbidity. Perioperative antibiotic practices for megaprosthetic reconstructions are not standardized and are based on guidelines for conventional joint arthroplasties. This study aims to evaluate the efficacy of current prophylactic strategies for megaprosthetic reconstructions. We conducted a retrospective review of megaprosthetic reconstructions performed at Duke University from 2001 to 2021. Logistic regression with GEE was used to assess whether a prolonged course of postoperative antibiotics is associated with infection risk. We assessed the microbial profile and corresponding susceptibilities of megaprosthetic infections through record review. Additionally, we designed a pharmacokinetic subgroup analysis using liquid chromatography-tandem mass spectrometry to quantify antibiotic concentrations in surgical tissue. Wilcoxon rank-sum tests were used to correlate tissue concentrations with infection risk. Out of 184 cases, 23 (12.5%) developed infection within 1 year. Extended postoperative antibiotics were not significantly associated with infection risk (P = 0.23). Among 18 culture-positive cases, 4 (22.2%) were caused by cefazolin-susceptible organisms. Median bone and muscle concentrations of cefazolin among cases that developed postoperative infection (0.065 ng/mL and 0.2 ng/mL, respectively) were significantly lower than those of cases that did not (0.42 ng/mL and 1.95 ng/mL, P < 0.01 and P = 0.03). This study is the first to comprehensively assess aspects of perioperative prophylaxis for megaprosthetic reconstructions. Extending postoperative antibiotics did not reduce infection risk. We detected a high frequency of cefazolin nonsusceptible organisms among postoperative infections. Additionally, intraoperative antibiotic tissue concentrations may be predictive of later infection. Future studies ought to examine optimal drug choices and dosing strategies.

Group sequential testing for cluster randomized trials with time-to-event endpoint.

We propose group sequential methods for cluster randomized trials (CRTs) with time-to-event endpoint. The alpha spending function approach is used for sequential data monitoring. The key to this approach is determining the joint distribution of test statistics and the information fraction at the time of interim analysis. We prove that the sequentially computed log-rank statistics in CRTs do not have independent increment property. We also propose an information fraction for group sequential trials with clustered survival data and a corresponding sample size determination approach. Extensive simulation studies are conducted to evaluate the performance of our proposed testing procedure using some existing alpha spending functions in terms of expected sample size and maximal sample size. Real study examples are taken to demonstrate our method.

Investigating readmission rates for patients undergoing oncologic resection and endoprosthetic reconstruction for primary sarcomas and tumors involving bone.

BACKGROUND: Little is known about the drivers of readmission in patients undergoing Orthopaedic oncologic resection. The goal of this study was to identify factors independently associated with 90-day readmission for patients undergoing oncologic resection and subsequent prosthetic reconstruction for primary tumors involving bone. METHODS: This was a retrospective comparative cohort study of patients treated from 2008 to 2019 who underwent endoprosthetic reconstruction for a primary bone tumor or soft tissue tumor involving bone, as well as those who underwent a revision endoprosthetic reconstruction if the primary endoprosthetic reconstruction was performed for an oncologic resection. The primary outcome measure was unplanned 90-day readmission. RESULTS: A total of 149 patients were identified who underwent 191 surgeries were for a primary bone or soft tissue tumor. The 90-day readmission rate was 28.3%. Female gender, depression, higher tumor grade, vascular reconstruction, longer procedure duration, longer length of stay (LOS), multiple surgeries during an admission and disposition to a Skilled Nursing Facility were associated with readmission (p < 0.05). In a multivariate analysis, female sex, higher tumor grade and longer procedure duration were independently associated with risk of readmission (p < 0.05). CONCLUSIONS: Readmission rates are high following endoprosthetic reconstruction for Orthopaedic oncologic resections. Further work is necessary to help minimize unplanned readmissions.

Diagnostic Performance of an Artificial Intelligence System in Breast Ultrasound.

OBJECTIVES: We study the performance of an artificial intelligence (AI) program designed to assist radiologists in the diagnosis of breast cancer, relative to measures obtained from conventional readings by radiologists. METHODS: A total of 10 radiologists read a curated, anonymized group of 299 breast ultrasound images that contained at least one suspicious lesion and for which a final diagnosis was independently determined. Separately, the AI program was initialized by a lead radiologist and the computed results compared against those of the radiologists. RESULTS: The AI program's diagnoses of breast lesions had concordance with the 10 radiologists' readings across a number of BI-RADS descriptors. The sensitivity, specificity, and accuracy of the AI program's diagnosis of benign versus malignant was above 0.8, in agreement with the highest performing radiologists and commensurate with recent studies. CONCLUSION: The trained AI program can contribute to accuracy of breast cancer diagnoses with ultrasound.

Sample size calculation for clustered survival data under subunit randomization.

Each cluster consists of multiple subunits from which outcome data are collected. In a subunit randomization trial, subunits are randomized into different intervention arms. Observations from subunits within each cluster tend to be positively correlated due to the shared common frailties, so that the outcome data from a subunit randomization trial have dependency between arms as well as within each arm. For subunit randomization trials with a survival endpoint, few methods have been proposed for sample size calculation showing the clear relationship between the joint survival distribution between subunits and the sample size, especially when the number of subunits from each cluster is variable. In this paper, we propose a closed form sample size formula for weighted rank test to compare the marginal survival distributions between intervention arms under subunit randomization, possibly with variable number of subunits among clusters. We conduct extensive simulations to evaluate the performance of our formula under various design settings, and demonstrate our sample size calculation method with some real clinical trials.

Bacteremia in solid organ transplant recipients as compared to immunocompetent patients: Acute phase cytokines and outcomes in a prospective, matched cohort study.

We undertook a prospective, matched cohort study of patients with Staphylococcus aureus bacteremia (SAB) and gram-negative bacteremia (GNB) to compare the characteristics, outcomes, and chemokine and cytokine response in transplant recipients to immunocompetent, nontransplant recipients. Fifty-five transplant recipients (GNB n = 29; SAB n = 26) and 225 nontransplant recipients (GNB n = 114; SAB n = 111) were included for clinical analysis. Transplant GNB had a significantly lower incidence of septic shock than nontransplant GNB (10.3% vs 30.7%, p = .03). Thirty-day mortality did not differ significantly between transplant and nontransplant recipients with GNB (10.3% vs 15.8%, p = .57) or SAB (0.0% vs 11.7%, p = .13). Next, transplant patients were matched 1:1 with nontransplant patients for the chemokine and cytokine analysis. Five cytokines and chemokines were significantly lower in transplant GNB vs nontransplant GNB: IL-2 (median [IQR]: 7.1 pg/ml [7.1, 7.1] vs 32.6 pg/ml [7.1, 88.0]; p = .001), MIP-1ß (30.7 pg/ml [30.7, 30.7] vs 243.3 pg/ml [30.7, 344.4]; p = .001), IL-8 (32.0 pg/ml [5.6, 53.1] vs 59.1 pg/ml [39.2, 119.4]; p = .003), IL-15 (12.0 pg/ml [12.0, 12.0] vs 12.0 pg/ml [12.0, 126.7]; p = .03), and IFN-α (5.1 pg/mL [5.1, 5.1] vs 5.1 pg/ml [5.1, 26.3]; p = .04). Regulated upon Activation, Normal T Cell Expressed and Secreted (RANTES) was higher in transplant SAB vs nontransplant SAB (mean [SD]: 750.2 pg/ml [194.6] vs 656.5 pg/ml [147.6]; p = .046).

Characterization of Sentinel Lymph Node Immune Signatures and Implications for Risk Stratification for Adjuvant Therapy in Melanoma.

BACKGROUND: Although sentinel lymph node (SLN) biopsy is a standard procedure used to identify patients at risk for melanoma recurrence, it fails to risk-stratify certain patients accurately. Because processes in SLNs regulate anti-tumor immune responses, the authors hypothesized that SLN gene expression may be used for risk stratification. METHODS: The Nanostring nCounter PanCancer Immune Profiling Panel was used to quantify expression of 730 immune-related genes in 60 SLN specimens (31 positive [pSLNs], 29 negative [nSLNs]) from a retrospective melanoma cohort. A multivariate prediction model for recurrence-free survival (RFS) was created by applying stepwise variable selection to Cox regression models. Risk scores calculated on the basis of the model were used to stratify patients into low- and high-risk groups. The predictive power of the model was assessed using the Kaplan-Meier and log-rank tests. RESULTS: During a median follow-up period of 6.3 years, 20 patients (33.3%) experienced recurrence (pSLN, 45.2% [14/31] vs nSLN, 20.7% [6/29]; p = 0.0445). A fitted Cox regression model incorporating 12 genes accurately predicted RFS (C-index, 0.9919). Improved RFS was associated with increased expression of TIGIT (p = 0.0326), an immune checkpoint, and decreased expression of CXCL16 (p = 0.0273), a cytokine important in promoting dendritic and T cell interactions. Independent of SLN status, the model in this study was able to stratify patients into cohorts at high and low risk for recurrence (p < 0.001, log-rank). CONCLUSIONS: Expression profiles of the SLN gene are associated with melanoma recurrence and may be able to identify patients as high or low risk regardless of SLN status, potentially enhancing patient selection for adjuvant therapy.

Outcomes of adjunctive therapies post hedgehog inhibitors in the management of locally advanced basal cell carcinoma: A systematic review and pooled analysis.

Management of patients with locally advanced basal cell carcinoma (laBCC) with traditional strategies has yielded suboptimal outcomes. Targeted treatments including hedgehog inhibitor therapy (HHIT) present limitations when utilized as monotherapy. Herein, we report evidence-based outcomes from available literature on multimodality treatments adjuvant to HHIT in laBCC management. Utilizing a systematic search strategy in PubMed, we identified studies published from inception to April 15, 2020, screened for definitive inclusion/exclusion criteria, and performed individual study quality assessment and pooled analysis to assess impact of adjunctive treatment-based responses post-HHIT on clinical response and recurrence outcomes. Twenty-nine studies (n = 103) were included. Primary findings include a complete response (CR) rate of 90.5%, the median follow-up of 12 months post-HHIT completion. The recurrence rate was 10.8% with 12-month median time to recurrence. Mohs micrographic surgery (MMS) had 100% CR post-HHIT, while no difference was observed between surgery and radiation therapy (RT). MMS and surgery had comparable 2-year recurrence free rates (RFR) at 87% and 86% respectively, while RT had the lower 2-year RFR at 67%. Male gender portended a more advanced stage at diagnosis and worse outcomes. In a subset analysis, periorbital laBCCs with orbital involvement had a CR rate of 81.8% versus 100% in those without orbital involvement, with similar rates of recurrence. Limited available quantitative data and possible publication bias were limitations. Pooled analysis of observational data supports use of adjunctive therapies post-HHIT to improve treatment response in patients with laBCC. Longer-term follow-up is needed to study recurrence rates after combination therapy.

Survival outcomes in sinonasal carcinoma with neuroendocrine differentiation: A NCDB analysis.

BACKGROUND: Sinonasal carcinoma with neuroendocrine differentiation (SCND) is a rare group of tumors with poor prognosis. Treatment and sequence of therapies are still unclear. The goal of this study is to analyze treatment outcomes in SCND using a national database. METHODS: The National Cancer Database was queried for SCND from 2004 to 2014. Patient demographics, tumor characteristics and treatment paradigms were tabulated. Multivariable Cox proportional hazards regression was performed for statistical analysis of treatment regimen on overall survival (OS). RESULTS: A total of 415 patients were identified. Most patients were male (61.2%), with a median age of 58 years and the most common primary site was the nasal cavity (52.5%). T4 tumors were observed in 67.7% of cases. Unimodality (41.9%) and bimodality (43.9%) therapies were the most common treatment modalities. Radiation therapy was the only treatment administered in 30% of the patients, while 27.2% received definitive chemoradiation (CRT) and 11.6% had surgery with adjuvant CRT. In our Cox-PH model, age (HR = 1.04, p < 0.001), T4 (HR = 2.6, p = 0.004) and N2/N3 (HR = 2.18, p = 0.001) were associated with worse survival. Trimodality (HR = 0.49, p = 0.005) and bimodality (HR = 0.65, p = 0.009) therapies had a better OS compared to unimodality. Patients treated with definitive CRT or surgery with adjuvant CRT had a significant increase in OS (p = 0.01 and 0.002 respectively). CONCLUSION: SCND appears to be best treated using a multimodality approach with definitive CRT or surgery followed by CRT. Neoadjuvant chemotherapy could be helpful in selecting the best treatment strategy.

The Contribution of Material, Behavioral, Psychological, and Social-Relational Factors to Income-Related Disparities in Cardiovascular Risk Among Older Adults.

BACKGROUND: Understanding the factors underlying health disparities is vital to developing strategies to improve health equity in old age. Such efforts should be encouraged in Korea. OBJECTIVE: This study explored how material, behavioral, psychological, and social-relational factors contribute to income-related disparities in cardiovascular risk among Korean adults 65 years and older. METHODS: This was a secondary analysis of Korean National Health and Nutrition Examination Survey data (2013-2017), targeting 7347 older adults (≥65 years). Socioeconomic position, defined as income, was the primary indicator. The outcome was binary for predicted cardiovascular risk (<90 vs ≥90 percentile). Disparities were measured using relative index of inequality (RII). The contributions of material, behavioral, psychological, and social-relational factors were estimated by calculating percentage reduction in RII when adjusted for these factors. RESULTS: Among men aged 65 to 74 years and women 75 years or older, the largest reductions in RII were achieved after adjusting for social-relational factors. Among men 75 years or older and women aged 65 to 74 years, adjusting for material factors resulted in the largest reductions in RII. Adjustments for behavioral factors also reduced RII for both genders aged 65 to 74 years. CONCLUSIONS: Improving the social, material, and behavioral circumstances of lower-income older adults may help address income-related disparities in cardiovascular risk in old age.

CALGB 50604: risk-adapted treatment of nonbulky early-stage Hodgkin lymphoma based on interim PET.

A negative interim positron emission tomography/computerized tomography (PET/CT) after 1 to 3 cycles of doxorubicin, bleomycin, vinblastine, and dacarbazine (ABVD) in patients with newly diagnosed, nonbulky stage I or II Hodgkin lymphoma (HL) predicts a low relapse rate. This phase 2 trial was designed to determine if a population of patients with early-stage disease can be treated with short-course ABVD without radiation therapy (RT) on the basis of a negative interim PET/CT, thereby limiting the risks of treatment. Between 15 May 2010 and 21 February 2013, 164 previously untreated patients with nonbulky stage I/II HL were enrolled, and 149 were included in the final analysis. Patients received 2 cycles of ABVD followed by PET. Deauville scores 1 to 3 were negative (≤ liver uptake) based on central review. PET- patients received 2 more cycles of ABVD, and PET+ patients received 2 cycles of dose-intense bleomycin, etoposide, doxorubicin, cyclophosphamide, vincristine, procarbazine, and prednisone (escalated BEACOPP) plus 3060-cGy involved-field RT. The primary objective was to determine 3-year progression-free survival (PFS) for the PET- group. One hundred thirty-five patients (91%) were interim PET-, and 14 patients (9%) were PET+ With median follow-up time of 3.8 years, the estimated 3-year PFS was 91% for the PET- group and 66% for the PET+ group (hazard ratio, 3.84; 95% confidence interval, 1.50-9.84; P = .011). There was 1 death as a result of suicide. Four cycles of ABVD resulted in durable remissions for a majority of patients with early-stage nonbulky HL and a negative interim PET. This trial was registered at www.clinicaltrials.gov as #NCT01132807.

Nomogram to predict lymph node metastasis in patients with early gastric cancer: a useful clinical tool to reduce gastrectomy after endoscopic resection.

BACKGROUND: The indications for endoscopic dissection have been expanded to improve the quality of life of patients with early gastric cancer (EGC). This study aimed to develop a nomogram to predict the status of lymph node metastasis with the aim of avoiding unnecessary gastrectomies. METHODS: We reviewed the clinicopathological data of 10 579 patients who underwent curative resection for EGC. The nomogram was developed by multivariate analysis and was evaluated by external validation. Overall, disease-free and recurrence-free survival were compared between the gastrectomy group of 6641 patients and the endoscopic dissection group of 999 patients to show the efficacy of the nomogram. RESULTS: Multivariate analyses revealed that age, tumor size, lymphatic invasion, depth of invasion, and histologic differentiation were all significant prognostic factors for lymph node metastasis. The nomogram had good discriminatory performance, with a concordance index of 0.846. This was supported by the external validation point of 0.813. For patients with low risk of lymph node metastasis on the nomogram (≤ 3 % of the provisional value in this study), the endoscopic dissection and gastrectomy groups had comparable rates of overall (P = 0.32), disease-free (P = 0.47), and recurrence-free (P = 0.09) survival. CONCLUSIONS: We developed and validated a nomogram that predicts the risk of lymph node metastasis in EGC based on a large database. This precision nomogram is useful to avoid unnecessary gastrectomy after endoscopic dissection, which may ultimately improve the quality of life of patients with EGC.

Sample size calculation for cluster randomization trials with a time-to-event endpoint.

Cluster randomization trials randomize groups (called clusters) of subjects (called subunits) between intervention arms, and observations are collected from each subject. In this case, subunits within each cluster share common frailties, so that the observations from subunits of each cluster tend to be correlated. Oftentimes, the outcome of a cluster randomization trial is a time-to-event endpoint with censoring. In this article, we propose a closed form sample size formula for weighted rank tests to compare the marginal survival distributions between intervention arms under cluster randomization with possibly variable cluster sizes. Extensive simulation studies are conducted to evaluate the performance of our sample size formula under various design settings. Real study examples are taken to demonstrate our method.

Clinical Factors Affecting Survival in Pediatric Vascular Tumors of the Head and Neck.

Clinical factors associated with the behavior and outcomes of nonbenign, head, and neck vascular tumors in children are not well described. Our aim is to provide descriptive information and identify prognostic factors associated with lower overall survival for children with these types of tumors. A retrospective cohort study was performed using the SEER database (years 1973-2015). Children aged 18 years and under with the diagnosis of a vascular tumor with locally aggressive or borderline and malignant behavior, classified by ICD-O-3, within the head and neck were included. Vascular tumors with benign behavior as classified by ISSVA were excluded. One hundred forty-eight children were identified. Mean age was 9.9 years (SD = 6.4). A gender predilection was noted with more males affected, female (37.8%) and male (62.2%), P = 0.0031. Majority of the children were white (79.4%) and angiosarcoma was the most common histologic subtype (68.2%). Children had a significantly better overall survival than adults with head and neck vascular tumors, P < 0.0001. Univariate and multivariate analysis did not reveal any factors with significant association to overall survival. Vascular tumors in the head and neck in children most commonly affect males and white children, with angiosarcoma as the most common histologic subtype. Children seem to have a more favorable overall survival compared to adults.

Randomized trial of ofatumumab and bendamustine versus ofatumumab, bendamustine, and bortezomib in previously untreated patients with high-risk follicular lymphoma: CALGB 50904 (Alliance).

BACKGROUND: This multicenter, randomized phase 2 trial evaluated complete responses (CRs), efficacy, and safety with ofatumumab and bendamustine and with ofatumumab, bendamustine, and bortezomib in patients with untreated, high-risk follicular lymphoma (FL). METHODS: Patients with grade 1 to 3a FL and either a Follicular Lymphoma International Prognostic Index (FLIPI) score of 2 with 1 lymph node >6 cm or an FLIPI score of 3 to 5 were randomized to arm A (ofatumumab, bendamustine, and maintenance ofatumumab) or to arm B (ofatumumab, bendamustine, bortezomib, and maintenance ofatumumab and bortezomib). RESULTS: One hundred twenty-eight patients (66 in arm A and 62 in arm B) received treatment. The median age was 61 years, and 61% had disease >6 cm; 29% had an FLIPI score of 2, and 71% had an FLIPI score of 3 to 5. In arm A, 86% completed induction, and 64% completed maintenance. In arm B, 66% and 52% completed induction and maintenance, respectively. Dose modifications were required in 65% and 89% in arms A and B, respectively. Clinically significant grade 3 to 4 toxicities included neutropenia (A, 36%; B, 31%), nausea/vomiting (A, 0%; B, 8%), diarrhea (A, 5%; B, 11%), and sensory neuropathy (A, 0%; B, 5%). The estimated CR rates were 62% (95% confidence interval [CI], 50%-74%) and 60% (95% CI, 47%-72%) in arms A and B, respectively (P = .68). With a median follow-up of 3.3 years, the estimated 2-year progression-free survival (PFS) and overall survival (OS) rates were 80% and 97%, respectively, for arm A and 76% and 91%, respectively, for arm B. CONCLUSIONS: The CR rates, PFS, and OS were not improved with the addition of bortezomib to ofatumumab and bendamustine in patients with high-risk FL. Although grade 3 to 4 toxicities were similar, more patients treated with bortezomib required dose modifications and early discontinuation.