Search

Aim: We propose a method for screening full blood count metadata for evidence of communicable and noncommunicable diseases using machine learning (ML). Materials & methods: High dimensional hematology metadata was extracted over an 11-month period from Sysmex hematology analyzers from 43,761 patients. Predictive models for age, sex and individuality were developed to demonstrate the personalized nature of hematology data. Both numeric and raw flow cytometry data were used for both supervised and unsupervised ML to predict the presence of pneumonia, urinary tract infection and COVID-19. Heart failure was used as an objective to prove method generalizability. Results: Chronological age was predicted by a deep neural network with R2: 0.59; mean absolute error: 12; sex with AUROC: 0.83, phi: 0.47; individuality with 99.7% accuracy, phi: 0.97; pneumonia with AUROC: 0.74, sensitivity 58%, specificity 79%, 95% CI: 0.73–0.75, p < 0.0001; urinary tract infection AUROC: 0.68, sensitivity 52%, specificity 79%, 95% CI: 0.67–0.68, p < 0.0001; COVID-19 AUROC: 0.8, sensitivity 82%, specificity 75%, 95% CI: 0.79–0.8, p = 0.0006; and heart failure area under the receiver operator curve (AUROC): 0.78, sensitivity 72%, specificity 72%, 95% CI: 0.77–0.78; p < 0.0001. Conclusion: ML applied to hematology data could predict communicable and noncommunicable diseases, both at local and global levels., Lay abstract Identifying and monitoring an infectious disease within a community requires sampling of both symptomatic and asymptomatic individuals. Hematology data from a full blood count contains significantly more information than used clinically. We applied machine learning to hematology data to predict the presence of both communicable and noncommunicable diseases.

Introduction Hereditary hemochromatosis with serum ferritin (SF) greater than 1000µg/L at diagnosis has been associated with symptoms and end-organ damage from iron overload. However, clinical manifestation of iron overload is much less frequently observed when only a moderate iron overload of SF 1000µg/L or less is present at diagnosis. Data to guide optimal management with venesection of patients with moderate overload is currently limited and many clinical guidelines do not distinguish the degree of iron overload at presentation in their management recommendations. Since 2013, patients with hemochromatosis who present to our center with a SF of less than 1000µg/L without end organ dysfunction have been venesected to a target of Method Patients referred to our center for management of HFE mutated haemochromatosis between 1 January 2013 and 31 December 2017 were included (recent cohort). These patients were discharged after initial iron reduction. A second cohort of patients managed between 1 December 2001 to 31 December 2003 was also included for comparison (historic cohort). This historic cohort initially received maintenance venesections before being discharged in 2013, following a change in our policy. Patients were identified using the departmental electronic database and data was extracted from the patient electronic record. Time to re-referral was measured from the date of last venesection prior to discharge to the date of re-commencing venesection or referral to other services for iron related complications. Patients who had not been re-referred for resuming venesection or to other services for iron-overload complications were censored on 1 July 2020. Result A total of 109 patients were included for analysis (historic cohort: 54; recent cohort: 55). The median age was 45 (range: 45 - 81), and 74 (67.9%) patients were male. Homozygote C282Y mutation was the most common genotype (74.3%), followed by compound heterozygotes (C282Y/H65D or C282Y/S63C, 23.9%). The median SF prior to venesection was 539µg/L (range: 146 - 968), and a median of 8 venesections were required initially to achieve a SF of 500µg/L), and number of initial venesections needed for de-ironing (quartiles) were not predictive of re-referral. Conclusion Hereditary hemochromatosis patients who present with only a moderate iron overload at the time of diagnosis could be discharged to primary care following initial iron reduction. A significant proportion of patients do not accumulate significant levels of iron following cessation of venesection. The rate and pattern of iron re-accumulation was similar irrespective of whether patients received a period of maintenance venesection prior to discharge. Figure 1 Disclosures Chan: Amgen: Other: TRAVEL, ACCOMODATIONS, EXPENSES (paid by any for-profit health care company); Celgene: Other: TRAVEL, ACCOMODATIONS, EXPENSES (paid by any for-profit health care company); Roche: Other: TRAVEL, ACCOMODATIONS, EXPENSES (paid by any for-profit health care company); AbbVie: Membership on an entity's Board of Directors or advisory committees; Janssen: Membership on an entity's Board of Directors or advisory committees, Other: TRAVEL, ACCOMODATIONS, EXPENSES (paid by any for-profit health care company), Research Funding, Speakers Bureau.

Introduction Acute myeloid leukaemia (AML), high risk myelodysplastic syndrome (MDS) and chronic myelomonocytic leukaemia (CMML) are bone marrow disorders that predominately affect the elderly population. The treatment options for this patient group have been limited and those unable to undergo intensive chemotherapy have historically had a poor prognosis. Low dose cytarabine (LDAC) is an accepted treatment option in patients who are ineligible for intensive therapy. LDAC has been shown to achieve a complete response (CR) and complete response with incomplete count recovery (CRi) in 11-18% of patients with a median overall survival of 5- 12 months. Recent evidence suggests that venetoclax (VEN) has a synergistic effect with LDAC with a phase I/II trial reporting a CR +CRi of 33/61 (54%) with this combination. Antifungal agents are commonly used in AML treatment protocols due to the increased risk of invasive fungal infections in this population. Fluconazole, a moderate CYP3A4 inhibitor, increases the maximum concentration (Cmax) of VEN resulting in a 2-5-fold increase in the area under the curve (AUC), thereby significantly increasing drug exposure. When using an azole antifungal, a dose reduction of VEN is recommended to maintain safe therapeutic levels. Here, we present our experience with using a lower dose of VEN in combination with fluconazole and LDAC (VeLDAC-F) in patients with AML, high risk MDS and CMML. Methods Patients with a diagnosis of AML, high risk MDS or CMML that required treatment but were not candidates for high intensity chemotherapy were offered the combination of VeLDAC-F. In cycle 1, VEN was started at 100mg daily and ramped up over the first 4 days to 200mg. This dose was continued from days 4-10. Fluconazole was started on day 3 and continued at 200mg until day 10. On subsequent cycles, VEN was administered at 200mg daily on days 1-10 with 200mg of fluconazole. Subcutaneous LDAC 20 mg/m2/day was given on days 1-10 of each cycle. Treatment cycles were repeated every 4-6 weeks. Patients were not routinely admitted for initiation of chemotherapy, unless the circulating blast count was high. Patients who received at least one dose of VeLDAC-F were included. Event-free survival was the time to either treatment termination, disease progression or death; overall survival was the time to death from all cause. Patients who achieved a haemoglobin > 100g/L, platelets >100 x 109/L and neutrophils > 1 x 109/L were said to have reached a modified haematological response (HR), and time to modified haematological response was analysed using competing risk model with death as the competing factor. All statistical analysis was done using IBM SPSS version 20 and R Statistics. Results Nineteen patients received at least one dose of VeLDAC-F between the 1st of June 2017 and the 1st of June 2018. The majority of patients were male (89.5%). 14/19 (73.7%) had an ECOG performance status of 0 or 1 and the median Charleston co-morbidity index was 6. The median age at diagnosis was 77 years (range 64.4- 87.7 years). The cohort included 9 patients with AML (47.4%), 7 with high risk MDS (36.8%) and 2 with CMML (15.8%). Twelve patients were still alive at the time of analysis (63.2%). The median follow-up was 182 days. Ten patients achieved a modified HR (52.6%). The median duration of response in these patients was 210 days and the 100 day probability of achieving a modified HR was 50.1%. The majority of patients who achieved a modified HR did so within 2 cycles of treatment with only one patient achieving a modified HR after 100 days. Seven patients died over the follow-up period (36.8%) with one patient dying within 30 days of commencing treatment (5.3%). Four patients died of progressive disease, 2 from sepsis and 1 from catastrophic bleeding. The median OS had not been reached at the time of analysis. The 180-day OS was 61%. The median event free survival was 217 days. There were no cases of clinical tumour lysis and three cases of biochemical tumour lysis (15.8%). Only 10 patients were admitted to hospital with neutropenic fever over the course of the follow up period (52.6%). Discussion The combination of VeLDAC-F appears to be an effective regimen for elderly patients with AML, high grade MDS and CMML who are otherwise ineligible for intense chemotherapy. The risk of tumour-lysis is low in this real-world cohort, but ongoing follow-up and further clinical trials are needed to establish the longer-term outcomes of this regime. Figure. Figure. Disclosures Chan: Amgen: Honoraria; Karyopharm: Research Funding. Simpson:Pharmacyclics LLC, an AbbVie Company: Research Funding; Acerta: Research Funding; Merck: Honoraria, Research Funding; MSD: Honoraria; BeiGene: Research Funding; Sanofi: Research Funding; Bristol-Myers Squibb: Other: TRAVEL, ACCOMMODATIONS, EXPENSES; Celgene: Honoraria, Other: TRAVEL, ACCOMMODATIONS, EXPENSES; Novartis: Other: TRAVEL, ACCOMMODATIONS, EXPENSES; Roche: Honoraria, Research Funding; Abbvie: Honoraria, Research Funding; Amgen: Research Funding, TRAVEL, ACCOMMODATIONS, EXPENSES; Janssen: Honoraria, Research Funding.

Abstract 4049 Background Bortezomib became funded for frontline therapy of symptomatic multiple myeloma (SMM) in New Zealand on 1 May 2011. From this date all new patients requiring treatment at our institution, both transplant eligible (TE) and ineligible (TIn), were initially treated with CyBorD consisting of weekly Cyclophosphamide 300mg/m2 PO, Bortezomib 1.6mg/m2 SC(2.5mg/ml) and Dexamethasone 40mg PO. One cycle was arbitrarily defined as 28 days, TE patients received 4–6 cycles CyBorD before ABMT (using 200mg/m2 melphalan), then 3 cycles of VTD (Bortezomib 1.6mg/m2 weekly, Thalidomide 100mg daily, Dexamethasone 40mg weekly) as post transplant consolidation. TIn patients received 4 cycles of CyBorD then 4 cycles of VTD. In June 2011 we changed all patients to subcutaneous bortezomib due to ease of administration and reduced toxicity. Methods The records of all untreated SMM patients initiating treatment with Bortezomib at our institution from 1 May 2011 were analyzed for response. Results We have treated 25 patients with SMM, 14 TE and 11 TIn. The median age was 68 yrs (range 45 to 88 yrs). The treatments were well tolerated. There was one Bortezomib dose reduction due to toxicity (burning feet on VTD). Injection site skin redness was common, but local reactions were not severe, one patient changed to IV Bortezomib because of this. There were no dose delays due to thrombocytopenia, and no patient had a pre-dose thrombocytopenia of Response after each cycle Conclusions Weekly subcutaneous Bortezomib is well tolerated, and has high patient acceptance. Dose limiting neuropathy is rare, and responses are high when used in combination with CyBorD or VTD. Disclosures: No relevant conflicts of interest to declare.

Angioimmunobalstic T cell lymphoma (AILT) is a relatively rare subtype of lymphoma, accounting for approximately 1 – 2% of all cases of NHL. The pathogenesis is thought due to clonal expansion of follicular T-helper cells which subvert normal germinal centre function. It is characterised by systemic disease and significant immune dysregulation with frequent atypical infections and autoimmune disease. It has a poor prognosis with a median survival of less than 3 years using CHOP or CHOP-like regimens. Dose intensification, or anthracycline use, appears to be ineffective. Thalidomide is an immunomodulatory agent with isolated case reports of activity in this disease. We report 3 cases of AILT where thalidomide was effective at inducing or providing a sustained remission. Case 1 was an 86 year old woman who presented with systemic symptoms, widespread lymphadenopathy and a polyclonal increase in immunoglobulins. AILT was diagnosed from both lymph node biopsy and bone marrow biopsy. Symptoms improved only temporarily with prednisone. She commenced thalidomide at 200mg daily with improvement in symptoms and resolution of lymphadenopathy and after 2 months the dose was reduced to 100mg daily. She continued the thalidomide at the same dose and remained in remission until her death from unrelated causes 4 years later Case 2 was a 72 year old man who initially presented with autoimmune haemolytic anaemia and red cell aplasia which was unresponsive to steroids, intravenous immunoglobulins (IVIG), cyclosporine and splenectomy. He then developed cervical and axillary lymphadenopathy, biopsy of which showed AILT. Bone marrow examination showed no evidence of AILT. He commenced thalidomide at 150mg daily with resolution of the haemolysis and red cell aplasia within 12 days and reduction in all lymphadenopathy. He remains in remission 3 months after commencing thalidomide and is currently on 100mg daily. Case 3 was a 57 year old woman presented with ITP with platelets < 10, systemic symptoms, widespread lymphadenopathy and hepatosplenomegally. AILT was diagnosed from lymph node biopsy and bone marrow biopsy. She commenced on prednisone and IVIG for the ITP, then proceeded to CHOP chemotherapy. Her first cycle was complicated by Pneumocystis jirovecii infection. Following the chemotherapy, her systemic symptoms subsided and the lymphadenopathy and organomegaly resolved. She commenced thalidomide at 100mg daily and has remained in remission 2 months later. There are now several case reports in the literature indicating that thalidomide is effective in AILT which is reflected by the cases reported here. Given the poor prognosis of AILT using conventional chemotherapeutic regimens, such as CHOP, prospective clinical trials using thalidomide, or the newer derivatives such as lenalidomide, are warranted.