Citations

Abstract Patients with multiple myeloma (MM) who do not respond to initial therapy have worse outcomes than primary responders, and effective treatments are lacking in this population. However, the outcomes of primary refractory disease in the modern treatment era have not yet been studied. We reviewed patients with MM treated with triplet/quadruplet therapy at our institution to assess the incidence of primary refractory disease and the impact of salvage therapies in this population. We identified 1127 patients, of whom 1086 were evaluated for hematologic responses after 4 to 6 cycles. Of these, 93.3% (1013) had evidence of response, whereas 6.7% (73) had primary refractory disease. With a median overall survival (OS) of 51.3 months, patients with primary refractory disease had an increased risk of shorter survival in univariable and multivariable analyses (hazard ratio [HR], 3.5 [95% confidence interval (CI), 2.5-4.9]; HR, 4.3 [95% CI, 2.6-6.9], respectively). In the subgroup analysis of patients with primary refractory disease, those who received second-line autologous stem cell transplantation (ASCT) had increased second progression-free survival (20.9 vs 8.1 months; P < .01) and second OS (74.7 vs 31.3 months; P = .02) compared with patients who did not. We conclude that early progression remains a significant factor for shorter OS in the current era, and salvage ASCT could be the most beneficial option for this population.

Background and motivation: Lung computed tomography (CT) techniques are high-resolution and are well adopted in the intensive care unit (ICU) for COVID-19 disease control classification. Most artificial intelligence (AI) systems do not undergo generalization and are typically overfitted. Such trained AI systems are not practical for clinical settings and therefore do not give accurate results when executed on unseen data sets. We hypothesize that ensemble deep learning (EDL) is superior to deep transfer learning (TL) in both non-augmented and augmented frameworks. Methodology: The system consists of a cascade of quality control, ResNet–UNet-based hybrid deep learning for lung segmentation, and seven models using TL-based classification followed by five types of EDL’s. To prove our hypothesis, five different kinds of data combinations (DC) were designed using a combination of two multicenter cohorts—Croatia (80 COVID) and Italy (72 COVID and 30 controls)—leading to 12,000 CT slices. As part of generalization, the system was tested on unseen data and statistically tested for reliability/stability. Results: Using the K5 (80:20) cross-validation protocol on the balanced and augmented dataset, the five DC datasets improved TL mean accuracy by 3.32%, 6.56%, 12.96%, 47.1%, and 2.78%, respectively. The five EDL systems showed improvements in accuracy of 2.12%, 5.78%, 6.72%, 32.05%, and 2.40%, thus validating our hypothesis. All statistical tests proved positive for reliability and stability. Conclusion: EDL showed superior performance to TL systems for both (a) unbalanced and unaugmented and (b) balanced and augmented datasets for both (i) seen and (ii) unseen paradigms, validating both our hypotheses.

e20032 Background: The introduction of anti-CD38 antibodies (daratumumab) in the treatment landscape of multiple myeloma (MM) has increased the progression-free survival (PFS) in recent years. Still, the ideal combination (Dara-IMiD vs. Dara-PI) of these drugs in patients with relapsed MM is unclear. Methods: We reviewed MM patients treated with daratumumab as second-line therapy at our institution from 2016 to 2022 to assess the outcomes with the different combination therapies. We used the Kaplan-Meier method to estimate PFS from the first relapse. Results: We identified 329 patients with a median age of 63.6 years (range: 27.1 – 90.4). Most of the patients received triplet induction therapy (95%) and upfront transplant (67%) as primary therapy. At the time of Dara initiation, 41% were refractory to IMiDs, 22% to PIs, 18% to both, and 18% to neither drug. Patients treated with Dara-IMiD had significantly better PFS than Dara-PI (28.7 months vs. 18.5 months, p < 0.01, respectively). In a subgroup analysis based on the FISH results, high-risk patients had similar PFS outcomes (24.3 vs. 10.2 months, p = NS, respectively), while standard-risk patients benefited significantly from the Dara-IMiD regimen (48.3 vs. 19.2 months, p < 0.01, respectively). Among patients treated with Dara-IMiD, those already refractory to IMiDs from the first line had significantly shorter PFS than those without IMiD resistance (14.9 vs. 43.3 months, p < 0.01, respectively). In addition, the lenalidomide-based combination resulted in significantly better PFS compared to pomalidomide in our cohort (40.5 vs. 19.2 months, p < 0.01, respectively). However, no difference was found in the IMiD refractory population (12.3 vs. 15.1 months, respectively). Finally, no significant differences were seen between Dara-IMiD and Dara-PI in patients only refractory to IMiDs (16.1 vs. 12.8 months, p = NS, respectively). Conclusions: While further studies are required to adjudicate the best combination, we conclude that Dara-IMiD might be more effective in patients with relapsed MM, especially for patients not refractory to IMiD at first line. [Table: see text]

e20011 Background: Drug refractoriness is an important factor determining choice of next therapy and prognosis in multiple myeloma (MM). It is debatable whether refractoriness to a drug varies among patients (pts) with MM progressing on different administration schedules of the drug. Outcomes of MM pts refractory to different doses of lenalidomide have been previously described, but similar data about proteasome inhibitors (PIs) are limited. In this study, we assessed the impact of refractoriness to different schedules of bortezomib (V) on outcomes with subsequent therapies in MM pts. Methods: We retrospectively reviewed MM pts at our center diagnosed between January 2004 and December 2018, who progressed on/ after receiving first line therapy with V containing regimens. At first relapse, we divided pts into 3 groups- MM progressing on or within 60 days of receiving (1) 1-1.3 mg/m2 V once weekly or more frequently (Standard Dose Refractory), (2) 1-1.3 mg/m2 V once every 2 weeks or less frequently (Low Dose Refractory), and (3) MM progressing at least 60 days after stopping V (Not Refractory). In addition to V, we classified pts as being refractory to cyclophosphamide and immunomodulatory drugs (IMiDs) at first relapse. We classified salvage regimens into daratumumab (Dara), PI, IMiD, PI+IMiD, and VDTPACE like regimens. We then assessed progression free survival (PFS) and overall survival (OS) for next line therapy, and for next V containing therapy using Kaplan Meier and Cox models. Results: A total of 302 pts were included. Median age at diagnosis was 62 years and 60% pts were male. Overall, 23%, 31%, and 27% were International Staging System (ISS) stage I, II, and III respectively; 40% had high risk FISH as per mSMART. Of the 302 pts, 134 (44%) pts received another V containing regimen after the first line therapy at any time during the disease course. Using a multivariable Cox model to adjust for ISS, high risk FISH, refractoriness to other drugs, duration of first line therapy, and treatment received at relapse, the Not Refractory Group had a better PFS for both 2nd line therapy and the next V containing line of therapy, while no significant differences were seen between the Standard Dose Refractory and Low Dose Refractory groups. OS differences followed the same trend but were not statistically significant (Table). Conclusions: On multivariable analysis, the PFS did not differ among standard dose vs low dose refractory pts, and both were significantly inferior to pts not refractory to V. These data suggest that refractoriness to V might not be a dose dependent phenomenon. [Table: see text]

8051 Background: Patients with multiple myeloma (MM) that do not respond to initial therapy have worse outcomes compared to primary responders, and effective treatments are lacking in this population. However, the outcomes of primary refractory disease in the modern treatment era have not been studied. Methods: From 2007-2019, we reviewed MM patients treated with triplet/quadruplet therapy in our institution to assess the incidence of primary refractory disease and the impact of salvage therapies in this population. Primary refractory disease was defined as either progressive disease or stable disease at 4-6 cycles. The Kaplan-Meier estimates were used for survival probabilities. PFS and OS were calculated from MM diagnosis to progression, escalation/change of treatment, or death, respectively. For the primary refractory group, the second PFS and second OS were measured from the start of second-line therapy. The Cox regression model was used for multivariable analysis with the following variables (age, ISS, induction therapy, FISH, and bone marrow plasma cells at diagnosis). Results: We identified 1127 patients, of which 1086 were evaluated for hematologic response after 4-6 cycles. Of these, 93.3% (1013) had evidence of response, while 6.7% (73) had primary refractory disease. With a median OS of 50.7 months, patients with primary refractory disease had an increased risk for shorter survival in univariable and multivariable analysis (HR: 3.6, 95% CI = 2.61 – 4.96, HR = 3.89, 95% CI = 2.68 – 5.65, respectively). In subgroup analysis of primary refractory patients, the median second PFS and second OS from the first relapse were 11.9 months (95% CI: 8.7 – 17.3 months) and 43.4 (95% CI: 31.4– 75.9 months), respectively. Patients that received 2nd line ASCT had increased second PFS (20.9 vs. 8.1 months, respectively, p < 0.01) and second OS (74.7 vs. 31.3 months, respectively, p = 0.02) compared to patients that did not. Patients that took daratumumab as the second line did not have any significant survival differences compared to those who did not, p = NS. Conclusions: We conclude that early progression remains a significant factor for shorter OS in the current era, and salvage ASCT could be the most beneficial option for this population. [Table: see text]

8043 Background: Extramedullary Multiple Myeloma (EMM) has an aggressive clinical course and can present at diagnosis [primary EMM, (pEMM)] or relapse [secondary EMM, (sEMM)]. Here, we report clinical characteristics and outcomes of EMM, and predictive factors for the development of sEMM. Methods: We identified 299 patients with biopsy-proven EMM between 01/01/2000 and 12/31/2021. Patients with solitary plasmacytomas, paraskeletal MM and primary plasma cell leukemia were excluded. A 1:1 matched pair analysis (matched for year of diagnosis of MM for comparable follow-up and treatment eras) was performed to compare clinical characteristics and outcomes for patients with and without sEMM. Results: The median follow-up from the diagnosis of MM was 11.5 years (95%CI: 9.3-14.8). Of 299 patients, 204 (68%) patients had sEMM and 95 (32%) had pEMM, with largely comparable baseline characteristics (Table). The median overall survival (OS) from the diagnosis of sEMM was 0.7 (95% CI: 0.6-0.9) years compared to 3.6 (95%CI: 2.4-5.6) years for pEMM (p<0.0001). For sEMM, mutlvariable analysis (MVA) for OS identified visceral organ involvement (eg. liver, lung/pleura, brain, pancreas, etc.) to be associated with inferior OS [HR 1.6 (95% CI: 1.1-2.4), p=0.01]; high-risk cytogenetics (HR-CTG), ISS-3 stage, and age did not impact OS from sEMM. For pEMM, HR-CTG [HR 2.3 (95% CI 1.1 -4.8), p=0.03] and ISS-3 stage [HR 3.1 (95% CI 1.5-6.7), p=0.003] were associated with inferior OS on MVA whereas visceral EMM or age did not impact OS from pEMM. On univariate analysis, patients with sEMM were younger (median age 58.7 vs 61 years, p=0.006), had a higher proportion of 1q duplication (32% vs 17%, p=0.001), t(4;14) [16% vs 9%, p=0.047] and elevated LDH (30% vs 19%, p=0.04), compared to the matched cohort of patients without sEMM. On MVA, younger age at diagnosis, t(4;14) and 1q duplication at diagnosis of MM were independent predictors of development of sEMM. The median OS from diagnosis of MM was 5.4 years (95%CI: 4.1-6.2) for patients with sEMM versus 7.5 years (95% CI: 6.4-10.1; p <0.001) for the matched cohort. Secondary EMM was an independent poor prognostic marker for OS from the diagnosis of MM [HR 1.6 (95%CI: 1.2-2.1); p=0.004] in addition to age, ISS3 disease and HR-CTG. Conclusions: Outcomes of patients with EMM, especially sEMM, remain dismal. Patients with sEMM have higher rates of 1q duplication and t(4;14) at diagnosis of MM and carried an independent adverse prognosis in a matched cohort. [Table: see text]

8022 Background: Extramedullary Multiple Myeloma (EMM) is an aggressive entity with a dismal prognosis. Immune effector therapies (IETs), including chimeric antigen receptor T-cell (CAR-T) therapies and bispecific T-cell engaging antibodies (BsAb), have demonstrated excellent efficacy in relapsed/refractory MM with limited data for efficacy in EMM. Here, we report the treatment outcomes for patients with EMM. Methods: We identified 299 patients with EMM diagnosed between 01/01/2000 and 12/31/2021 after excluding solitary plasmacytomas, paraskeletal MM and primary plasma cell leukemia. The IMWG criteria were used for response definition. Results: Of the 299 patients, 204 (68%) patients had secondary EMM (sEMM) and 95 (32%) patients had primary EMM (pEMM). For sEMM (n=204), the median progression free survival (PFS) with initial therapy was 2.9 (95% CI: 2.4-3.2) months. Initial treatment strategies for sEMM were heterogenous and demonstrated in the table; 44% patients achieved ≥partial response (PR) with initial treatment for sEMM. The median PFS in patients with ≥PR was 5.8 (95%CI: 4.5-6.9) months vs 1.8 (95%CI: 1.4-2; p <0.0001) months for <PR. For patients with pEMM (n=95), the median PFS was 12.9 (95% CI: 6.7-18) months; the median PFS was 17.4 (95%CI: 12.9-25.4) months for patients with ≥PR versus 1 month (95%CI: 0.8-2) in patients with <PR. Thirty patients (all sEMM and triple class refractory) were treated with IET after development of EMM: 18 with BCMA-directed CAR-T, 10 with BsAbs, and 2 with both CAR-T and BsAbs. In patients receiving CAR-T therapy, 75% (15/20) achieved ≥PR with 40% (n=8) achieving MRD negative (by flowcytometry) complete response. Fifteen (75%) patients receiving CAR-T had progressive disease (PD); 7 with systemic and EMM PD and 4 patients each with isolated EMM or systemic PD. The median PFS for patients treated with CAR-T was 4.9 months [3.1- not reached (NR)]. Among patients achieving a ≥PR with CAR-T, the median PFS was 5.8 (95%CI: 4.7-NR) months vs 1.4 months [(95%CI: 0.8-NR), p<0.001] for <PR. Among patients treated with BsAbs (10-BCMA, 1 each against FcRH5 and GPRC5D), 33% (4/12) achieved a ≥PR. Ten (83%) patients had PD on BsAb; 8 with both EMM and systemic PD and 1 patient each with isolated EMM or PD. The median PFS with BsAb was 2.9 months (95%CI: 2.2-NR). Conclusions: Patients with EMM continue to have dismal outcomes with conventional therapies. High response rates were noted with CAR-T therapy, but these tend to be short-lived. [Table: see text]

6036 Background: E3311 is a phase II randomized study of intermediate risk HPV+ OPC patients who underwent TOS by credentialed surgeons and were then treated by pathology-guided, deintensified post-operative treatment. Recent retrospective population data suggest that time to post-operative radiotherapy (PORT) > 6 weeks is associated with worse progression-free survival (PFS). Methods: We retrospectively analyzed demographics, pathologic characteristics, and oncologic outcomes in E3311 by time to initiation of PORT (short vs long, exploring both 6 and 7 weeks postop as the cutpoint). Binary and categorical variables were compared using a chi-square test (or a Fisher’s exact test for small sample sizes). Ordinal variables were compared using a Wilcoxon rank sum test. PFS and overall survival (OS) were estimated using the Kaplan-Meier method and compared using a log-rank test, which was stratified by intermediate- (Arms B/C) vs high-risk (Arm D). Results: Among 321 evaluable patients, the median time to initiation of PORT was 5.1 weeks (range 0.7-15.7); 5.0 (0.7-15.7) for Arm B, 5.1 (2.4-7.0) for Arm C, 5.6 (3.6-7.9) for Arm D. No significant difference in PFS or OS was observed whether PORT commenced by 6 or 7 weeks postoperatively, compared with longer time to initiation of PORT (p = 0.62 and 0.65, p = 0.30 and p = 0.41, respectively). 3-yr PFS rates were 93.5% (95% CI: 89.4%, 96.0%) among those treated ≤6 weeks, and 91.2% (95% CI: 81.4%, 95.9%) among those treated > 6 weeks; 3-yr OS rates were 96.1% (95% CI: 92.7%, 98.0%) among those treated ≤6 weeks, and 94.1% (95% CI: 84.9%, 97.8%) among those treated > 6 weeks. No significant differences in outcomes were observed between intermediate risk treatment arms (deintensified (50Gy) or standard dose (60Gy) RT). Additional demographic and pathologic characteristics, including primary site, margin status, histologic grade, stage, and extranodal extension were not significantly different between those with short vs long time to PORT, whether 6 or 7 weeks was used as the cutpoint. There were more T2 tumors in those treated > 7 weeks postop (84% vs. 49%, p = 0.007). PS = 1 patients were more numerous among those with shorter time to PORT (10% vs 0%, with the 7 week cutpoint). Conclusions: In contrast to retrospective NCDB data, intermediate- and high-risk HPV+ OPC patients enrolled on E3311 had favorable 3-yr PFS and OS rates that were not significantly worse when adjuvant therapy was started > 6 wks or > 7wks. These results argue against a survival advantage for initiation of PORT within 6 or 7 weeks for early-stage HPV+ OPC managed with high-quality transoral resection. The optimal time window to initiate PORT in HPV+ OPC patients remains to be determined. Clinical trial information: NCT01898494 .

3566 Background: According to the various types of clinical trials for metastatic colorectal cancer (mCRC), ETS and DpR are suggested as the surrogates of OS. Whereas associations between ETS/DpR and OS/PPS in the era of chemotherapy (chemo) +anti-EGFR antibody (ab)/bevacizumab (bev) in 1st line therapy have not been elucidated. Methods: From 40,889 Individual patient data (IPD) from 59 studies in ARCAD mCRC database, 2,138 treatment-naïve pts with RAS wild-type (wt) mCRC were selected from 7 randomized studies (PRIME, CAIRO2, CRYSTAL, OPUS, CALGB80495, WJOG4407G, ATOM) of chemo with/without anti-EGFR ab or bev. The ETS was defined as 20%≥ tumor shrinkage at 8 ± 2wks (ETS+: ≥20%, ETS-: <20%). Multivariate Cox regression models for OS/PPS were performed to investigate associations between ETS+ and ETS- by primary tumor location (overall/left-sided/right-sided), adjusting for potential confounders. DpR was defined as tumor shrinkage at nadir. OS/PPS were evaluated across quartile limits of DpR by multivariate Cox regression models. Results: In pts with overall or left-sided RAS wt mCRC, adjusted hazard ratios (HRs) of OS/PPS between ETS- and ETS+ for chemo +anti-EGFR ab, chemo +bev, and chemo alone showed great improvement, respectively (Table). In pts with overall or left-sided RAS wt mCRC, HRs of ETS for OS/PPS are higher in pts treated with chemo +anti-EGFR ab compared to those treated with chemo +bev or chemo alone. Compared to the pts with highest quartile (defined as ≥ 75th percentile (Q3)) of DpR, HRs of OS/PPS for lower quartile groups showed upward trends in pts with RAS wt & overall (Table). Conclusions: ETS/DpR in pts with RAS wt mCRC might be a potential prognostic marker of OS/PPS regardless of targeted therapy and primary tumor location. Superior prognostic values of ETS+ are expected in pts treated with anti-EGFR ab. [Table: see text]

Abstract Reproductive events beginning with pregnancy and ending with remodeling of the breast after cessation of breastfeeding alter breast structure and function and produce dramatic changes in systemic biology. In aggregate, these processes lower overall risk for breast, tubo-ovarian and endometrial cancers, albeit differentially by molecular subtypes of these tumors. Herein, we explore opportunities for research on protective mechanisms operative during this period of the life course, with the goal of encouraging studies to advance cancer prevention. See related article by Getz et al., p. 353