Maintenance therapy promotes profound organ and haematologic response in light-chain amyloidosis patients not undergoing autologous stem cell transplantation.

Introduction

Primary light-chain amyloidosis (AL) is a plasma cell disorder with organ dysfunction caused by monoclonal light chains of κ or λ type deposition, and it is the most common form of systemic amyloidosis [ 1 ]. Autologous haematopoietic stem cell transplantation (ASCT) remains the primary treatment for eligible patients with AL amyloidosis [ 2 ] because of its remarkable efficacy in reducing amyloidogenic light chains. However, it has been reported that only 20% of newly diagnosed primary AL patients have a chance to receive this intensive form of treatment [ 3 ]. At the same time, most patients do not qualify for ASCT due to severe organ dysfunction, for example, heart failure. Thus, the combination of chemotherapy and newer drugs was included in the treatment algorithm of primary AL patients. Recent studies confirmed that proteasome inhibitors (including Bortezomib and Ixazomib) [ 4–6 ] and anti-CD38 antibody, Daratumumab [ 7 ], are capable of remarkably improving AL patient outcomes. However, there is still debate on the role of maintenance therapy after upfront therapy in the treatment of AL patients not undergoing ASCT.

Maintenance therapy, which aims to prevent disease progression, is a widely accepted therapy following prior therapy for multiple myeloma (MM) [ 8 ]. Unlike MM, primary AL has a low burden on plasma tumour cells but severe organ involvement [ 9 ]. Studies exploring the pros and cons of maintenance therapy in primary AL patients with considerable haematologic responses are limited, with only a few studies focusing on patients undergoing ASCT [ 10–12 ]. In this study, we reviewed newly diagnosed primary AL patients treated with or without maintenance therapy at Zhongshan Hospital, Fudan University, and attempted to depict the clinical benefits of maintenance therapy in AL patients not undergoing ASCT in real-world practice.

Methods

Study design and patients

This was a retrospective, observational study that included patients diagnosed with primary AL from December 2008 to June 2023 at Zhongshan Hospital, Fudan University, aiming to identify the role of maintenance therapy in AL patients not undergoing ASCT in real-world practice. All cases were biopsy-proven and showed Congo red-positive amyloid deposits staining for one light chain isotype on immunofluorescence. Eligible patients aged over 18 years had completed their PI-based therapy (often Bortezomib-Cyclophosphamide-Dexamethasone, CyBorD) and had a response assessment by the 4th month. Patients who had achieved a considerable haematologic response after frontline induction therapy (at least a very good partial response (VGPR) by the 4th month) were included, as shown in the flow chart ( Figure S1 ). After the completion of induction therapy, patients would receive a dose-adjusted/frequency-reduced/toxicity-limited regimen for continuing medication as maintenance therapy or enter the observational phase depending on the physician’s choice. Patients with subsequent maintenance therapy were usually recommended to continue treatment until they had completed 24 months or disease progression (PD). Blood and urine tests were closely monitored to detect haematologic/organ relapse during monthly follow-ups for all patients.

All patients or their guardians provided written informed consent for the use of their anonymized medical data. The study was performed in compliance with the Declaration of Helsinki, and the clinical ethics committee of Zhongshan Hospital Fudan University approved the study procedure(B2023-185R). Data were subjected to peer-based quality checks for completeness and internal consistency.

Data collection

By reviewing historical records from the electronic data capture system, the baseline characteristics of patients were collected, including age, sex, amyloid light chain subtype, Mayo staging system, tumour burden (e.g. difference between involved and uninvolved free light chain (dFLC), bone marrow plasma cells (BMPCs)), organ involvement, and cytogenetic abnormalities (CAs, including Del 17p, Del 13q14, Amp/Gain 1q21, t(4;14), and t(11;14)). Bone marrow was evaluated for the presence of minimal residual disease (MRD) at an accuracy of 10 −6 using flow cytometry every two cycles of therapy, as reported in our previous study [ 13 ]. For the patients entering the observational phase, they were treated with regular outpatient follow-up and usually recommended to have MRD detection every 4–6 months. The type of treatment received, duration, and changes in depth of response were also recorded.

Response assessment and outcome

Responses were assessed every cycle of induction therapy and during the maintenance/observational phase until PD. Haematological and organ responses were evaluated according to the consensus criteria of the International Society of Amyloidosis published in 2005 [ 14 ] and updated in 2012 [ 15 ]. The cardiac response was graded by the change in N-terminal prohormone of brain natriuretic peptide (NT-proBNP) among those with a baseline NT-proBNP over 650 ng/L as reported [ 16 ]. Positive renal response was defined as a ≥ 30% decrease in proteinuria or a drop of proteinuria below 0.5 g/24h in the absence of renal progression [ 17 ]. Progression-free survival was defined as the time from the end of induction therapy to death, haematologic progression, or major organ deterioration (heart, kidney or liver). Haematologic progression was defined based on the increase in free light-chain or M-protein levels, and organ progression was defined based on the increase in NT-proBNP, 24-hour urine protein, serum creatinine and alkaline phosphatase according to consensus guidelines in 2012 [ 18 ]. Overall survival was calculated from the end of induction therapy until the date of the last follow-up or death if it occurred. Duration of response (DOR) was calculated from the date of the first evidence of haematologic response to death, haematologic progression, or organ progression. Treatment-related adverse events (TRAEs) were coded by the National Cancer Institute’s Common Terminology Criteria for Adverse Events v5.0.

Statistical analysis

Means with standard deviations (SD) or medians with interquartile ranges (IQR) were calculated for continuous variables, whereas frequencies with percentages were reported for categorical variables. Descriptive statistics were used to summarize the patient and disease characteristics across the two groups. The variables were compared using Fisher’s exact test for categorical variables and the Wilcoxon rank-sum test for continuous variables. Time-to-event curves were plotted using the Kaplan-Meier method, and comparisons were made using the log-rank test. All statistical analyses were conducted using SPSS software (version 25.0) and the R package in R/Bioconductor (version 3.6.1). p < 0.05 was considered statistically significant.

Results

Baseline clinical information

After receiving 3 to 6 cycles of induction therapy, the AL population was divided into two groups according to the intervention. There were 24 and 44 patients in the observational and maintenance groups. Clinical information was compared between the two groups, as seen in Table 1 . The median age at diagnosis was 60.4 ± 8.4 vs. 60.3 ± 8.0 years ( p = 0.961) in the observational group and the maintenance group. The majority of patients in both groups were males and had λ light chain restriction. Patients had similar baseline characteristics of tumour burden, disease staging, and organ involvement in the two distinct groups. The most commonly involved organs were the heart and kidneys. There were 87.5% ( n = 21) patients having cardiac involvement and 54.2% ( n = 35) patients having renal involvement in the observational group, while there were 79.6% ( n = 13) and 50.0% ( n = 22) in the maintenance group. Del13q14 was detected by fluorescent in situ hybridization at a rate of 9.1% ( n = 2) in the observational group while 38.5% ( n = 15) of patients in the maintenance group had del13q14 ( p = 0.031). Other CAs, especially t(11;14), had a comparable positive rate between the two sets of patients.

Induction therapy and maintenance therapy

Of the 68 patients involved in this study, 100% were exposed to frontline PI therapy (treatment regimens are shown in Table S1 ). With a median of four (range, 3–6) cycles in the observational group and six (range, 3–6) cycles of induction in the maintenance group, patients had similar haematologic complete response (hemCR) and VGPR rates from induction therapy. In addition, there were 8 in 23 patients achieved haematologic very good partial response (hemVGPR) plus dFLC < 10 mg/L in the maintenance group and 6 in 8 patients in the observation group at the end of induction. Among those receiving regular bone marrow MRD inspection, there were 9 (37.50%) and 17 (38.64%) patients having detectable MRD ( p = 0.965) in the observational and maintenance group, respectively. After the last cycle of first-line therapy, patients in the maintenance group received subsequent medication including IMiD, PI, a combination of IMiD and PI, daratumumab-based therapy, or alkylator for maintenance(details seen in Table S2 ) according to their induction therapy. The median duration of maintenance therapy was 7.2(IQR, 4.7–18.6) months ( Table 2 ). Common any-grade and grade 3/4 TRAEs are reported in Table S3 . There were 30 haematologic AEs and 60 non-haematologic AEs of any grade during maintenance therapy. Grade 3/4 events of anaemia and fatigue were most frequently reported.

Haematologic and organ response

The improvement in haematologic and major organ responses before and during the maintenance/observational phase was recorded. The hemCR response rate increased from 47.7% to 68.2% with maintenance regimens. However, none of the patients referred for observation achieved a response improvement based on the haematologic criteria ( Figure 1 ). Meanwhile, three patients assessed hemVGPR after induction were confirmed to acquire dFLC decrease to <10mg/L through maintenance therapy.

Cardiac and renal AL patients who received maintenance therapy experienced more profound organ responses than those being just monitored ( Table 3 ). Among patients with cardiac involvement at baseline, a numerically large proportion of patients with improved cardiac response was witnessed, with 14(31.8%) patients in the maintenance group compared to 3(12.5%) in the observational group ( Figure 2 ). The increase in cardiac CR (CarCR), VGPR, and partial response (PR) rates were 8.5%, 11.5% and 8.6% among the maintenance group. Those with cardiac no-response (CarNR) at frontline therapy tended to achieve improved cardiac response from maintenance therapy. A median time from maintenance initiation to response improvement was 5.3 (range, 1.5–33.6) months. The elevation of CarCR and CarVGPR rates were 10.0% and 4.7%, respectively, in the observation group. Patients who had CarVGPR or CarPR at frontline therapy were observed to have cardiac response improvement. It took a median of 2.6 (range, 2.0–3.1) months from the end of induction therapy to response improvement for patients to exhibit cardiac response. Those who had CarNR in the observational group at the frontline did not show an improvement in cardiac response.

Between-group differences in renal response improvement were also observed (1(4.2%) patient versus 5(11.4%), respectively) during follow-up ( Figure 2 ). In addition, there were five (11.4%) patients who had an MRD conversion benefit from positive to negative through subsequent continued treatment while no one in the observation group had conversion to MRD negativity. ( Table 3 ).

Patient outcomes

With a median respective follow-up among survivors of 33.9 ± 13.0 vs. 31.0 ± 2.0 months, there were 7 PFS and 2 OS events recorded in the observational group and 10 and 3 events in the maintenance group at the last follow-up. The median OS of the patients who received maintenance or observation was not reached. The median PFS in the observational group was 43.8 months compared to that in the control group (not reached, p = 0.528). Statistically significant differences in OS and PFS were not observed between the maintenance and observational groups ( Figure 3 ). Moreover, upon stratification by t(11;14), BMPC percentage, dFLC burden, and MRD status, there was no significant difference in PFS between the two groups ( Figure S2 ).

Discussion

Maintenance therapy following frontline induction has been a consensus in patients with MM not undergoing ASCT, while evidence is sparse in primary AL [ 2 ]. Owing to the application of the M-protein screening system in our institution [ 19 ], there has been an increase in the diagnosis rate of AL. Though ASCT remains an important part of frontline therapy [ 3 ], most patients in our centre are unable to undergo ASCT due to advanced Mayo staging, poor cardiac function, and advanced age with frailty. Other reasons included financial burden, concerns of side effects and medical preference. Given the tumour burden of AL, our institution initiated dose-adjusted/frequency-reduced/toxicity-limited maintenance therapy or randomly transitioned patients into an observation phase, aiming to explore effective therapeutic approaches for amyloidosis. Here, we retrospectively analysed the characteristics, treatment regimens, responses and outcomes of two types of newly diagnosed patients after achieving a favourable haematologic response (seen Figure 4 ). This is the first study to demonstrate the effectiveness of maintenance therapy among patients with primary AL who are not accessible to ASCT and report both haematologic and major organ response details of maintenance therapy. In conclusion, our study shows the benefits of maintenance therapy in haematologic response, organ function and MRD negativity conversion improvement for patients with primary AL who had received at least a VPGR haematologic response, although it did not translate into a survival advantage.

One of the reasons why survival benefits were not observed in the maintenance group is that the good haematologic response rate that participants had already acquired from frontline therapy resulted in excellent PFS and OS. As shown in a previous study [ 20 ], the median OS of patients with VGPR plus dFLC <10 mg/L at 3 months was 9.3 years, which was much better than that of those who did not achieve VGPR. Moreover, a shorter number of induction cycles implies rapid haematologic remission among patients in the observation group, which suggests a better patient outcome [ 20 ]. The limited follow-up time and sample size, compared with prolonged survival time, may be another reason for the failure to detect survival differences.

To be noted, the improvement in responses and MRD conversion in our study holds an indicative value, which the survival curve cannot explain. A previous study [ 12 ] showed that there were 9 (36%) patients with AL or light-chain deposition disease who achieved at least hemPR before induction had haematologic response improvement after receiving a four-week daratumumab consolidation therapy. The deepened haematologic response rate was 20.5% ( n = 9) through maintenance therapy among AL patients who had already achieved at least hemVGPR response in our study, which provides additional evidence that extended therapy can benefit AL patients. Meanwhile, the results of TRAEs showed that maintenance therapy is safe and tolerable for patients with AL.

Considering the delayed organ responses in AL [ 21 ], differences in cardiac and renal response rates between the maintenance and observational groups were compared. Fourteen(31.8%) patients in the maintenance group and 3(12.5%) in the observational group had a better cardiac response, with the majority of them achieving CarVGPR and CarCR ( Figure 2 ). Meanwhile, five(11.4%) and one (4.2%) patient showed an improved renal response. Organ dysfunction is thought to be a life-threatening matter for patients with AL. The depth of cardiac response has been repeatedly shown to be associated with reduced mortality risk [ 22 , 23 ], whereas renal response is related to a significantly reduced risk of dialysis [ 24 ]. A deeper organ response can also provide patients with a high quality of life [ 25 ]. In addition, for MRD detection, there were 5 (11.4%) patients having with MRD conversion from positive to negative. MRD negativity has been reported to be clinically relevant to the patients with AL, mainly leading to the improved organ response and patient outcome [ 26 ]. In summary, maintenance therapy has the potential to be an important development in the treatment of AL patients not undergoing ASCT.

Bortezomib is the backbone of therapy for AL amyloidosis. Most patients included in our study received CyBorD or bortezomib-based regimens as induction therapy. Though novel agents were increasingly used as primary therapy for AL, bortezomib-containing regimens are still considered as ‘standard therapy’ in many centres in Europe, the US and Asia-Pacific [ 27 ]. This real-world practice proved that continued therapy after standard therapy could bring clinical benefits to AL patients with considerable haematologic therapy, especially for those receiving PI-based induction therapy.

Although we evaluated the effectiveness of maintenance therapy for AL, this retrospective study had some limitations. One is the small size of the observational group, which potentially limits the statistical value of the two groups. There were more del13q CAs present in patients with maintenance therapy. Though del13q was not found to be of independent prognostic significance among AL patients in several studies [ 10 , 28 , 29 ], the effect of CAs on organ involvement and survival needs to be clarified thoroughly later. Moreover, patients in the two groups exhibited different hematological CR rates post-induction therapy, though not statistically significant, which could have biased our results. Further investigations in larger cohorts are necessary to confirm the findings of our study.

In conclusion, maintenance therapy following PI-based primary therapy showed an advantage on the promotion of MRD negativity conversion, haematologic and organ response in patients with primary AL not accessible to ASCT.

Supplementary Material

supplemantary materials.docx

Acknowledgments

The authors would like to thank all the patients who participated in this study and all the nursing and clinical staff who assisted with this study at the Department of Haematology, Zhongshan Hospital Fudan University.

Authors contributions

Conception and design: Peng Liu, Jing Li. Management of patient and data collection: Wenjing Wang, Tianhong Xu,Yang Yang, Chenqi Yu, Pu Wang, Chi Zhou. Data analysis and manuscript writing: Wenjing Wang. Manuscript revising: Jing Li, Peng Liu. All authors agree to be responsible for the study.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The raw data supporting the conclusions of this article are available from the corresponding author upon reasonable request.

References

1 Huang X , Fu C , Chen L , et al. Combination of bortezomib in the induction, conditioning and consolidation with autologous hematopoietic stem cell transplantation in patients with immunoglobulin light chain amyloidosis . Am J Hematol . 2019 ; 94 ( 4 ): E101 – E104 . doi: 10.1002/ajh.25404 . 30652338

2 Gertz MA. Immunoglobulin light chain amyloidosis: 2024 update on diagnosis, prognosis, and treatment . Am J Hematol . 2024 ; 99 ( 2 ): 309 – 324 . doi: 10.1002/ajh.27177 . 38095141

3 Wechalekar AD , Cibeira MT , Gibbs SD , et al. Guidelines for non-transplant chemotherapy for treatment of systemic AL amyloidosis: EHA-ISA Working Group . Amyloid . 2023 ; 30 ( 1 ): 3 – 17 . doi: 10.1080/13506129.2022.2093635 . 35838162

4 Sanchorawala V , Boccadoro M , Gertz M , et al. Guidelines for high dose chemotherapy and stem cell transplantation for systemic AL amyloidosis: EHA-ISA working group guidelines . Amyloid Int. J. Exp. Clin. Investig. Off. J. Int. Soc. Amyloidosis . 2022 ; 29 ( 1 ): 1 – 7 . doi: 10.1080/13506129.2021.2002841 . 34783272

5 Cohen OC , Sharpley F , Gillmore JD , et al. Use of ixazomib, lenalidomide and dexamethasone in patients with relapsed amyloid light-chain amyloidosis . Br J Haematol . 2020 ; 189 ( 4 ): 643 – 649 . doi: 10.1111/bjh.16401 . 31984481

6 Muchtar E , Gertz MA , LaPlant BR , et al. Phase 2 Trial of ixazomib, cyclophosphamide, and dexamethasone for previously untreated light chain amyloidosis . Blood Adv . 2022 ; 6 ( 18 ): 5429 – 5435 . doi: 10.1182/bloodadvances.2022007781 . 35737873 PMC9631712

7 Kastritis E , Palladini G , Minnema MC , et al. Daratumumab-based treatment for immunoglobulin light-chain amyloidosis . N Engl J Med . 2021 ; 385 ( 1 ): 46 – 58 . doi: 10.1056/NEJMoa2028631 . 34192431

8 Cowan AJ , Green DJ , Kwok M , et al. Diagnosis and management of multiple myeloma: a review . JAMA . 2022 ; 327 ( 5 ): 464 – 477 . doi: 10.1001/jama.2022.0003 . 35103762

9 Kumar SK , Callander NS , Adekola K , et al. Systemic light chain amyloidosis, Version 2.2023, NCCN clinical practice guidelines in oncology . J. Natl. Compr. Cancer Netw. . 2023 ; 21 ( 1 ): 67 – 81 . doi: 10.6004/jnccn.2023.0001 . 36652935

10 Ozga M , Zhao Q , Benson D , et al. AL amyloidosis: the effect of maintenance therapy on autologous stem cell transplantation outcomes . J Clin Med . 2020 ; 9 ( 11 ): 3778 . doi: 10.3390/jcm9113778 . 33238501 PMC7700492

11 Al Saleh AS , Sidiqi MH , Sidana S , et al. Impact of consolidation therapy post autologous stem cell transplant in patients with light chain amyloidosis . Am J Hematol . 2019 ; 94 ( 10 ): 1066 – 1071 . doi: 10.1002/ajh.25572 . 31273808 PMC7377298

12 Kastritis E , Rousakis P , Kostopoulos IV , et al. Consolidation with a short course of daratumumab in patients with AL amyloidosis or light chain deposition disease . Amyloid . 2021 ; 28 ( 4 ): 259 – 266 . doi: 10.1080/13506129.2021.1971192 . 34468250

13 Wang Y , Xu J , Xu B , et al. The prognostic role of 1q21 gain/amplification in newly diagnosed multiple myeloma: the Faster, the worse . Cancer . 2023 ; 129 ( 7 ): 1005 – 1016 . doi: 10.1002/cncr.34641 . 36704927

14 Gertz MA , Comenzo R , Falk RH , et al. Definition of ­organ involvement and treatment response in immunoglobulin light chain amyloidosis (AL): a consensus opinion from the 10th International Symposium on Amyloid and Amyloidosis, Tours, France, 18–22 April 2004 . Am J Hematol . 2005 ; 79 ( 4 ): 319 – 328 . doi: 10.1002/ajh.20381 . 16044444

15 Palladini G , Dispenzieri A , Gertz MA , et al. New criteria for response to treatment in immunoglobulin light chain amyloidosis based on free light chain measurement and cardiac biomarkers: impact on survival outcomes . J Clin Oncol . 2012 ; 30 ( 36 ): 4541 – 4549 . doi: 10.1200/JCO.2011.37.7614 . 23091105

16 Muchtar E , Dispenzieri A , Wisniowski B , et al. Graded cardiac response criteria for patients with systemic light chain amyloidosis . J Clin Oncol . 2023 ; 41 ( 7 ): 1393 – 1403 . doi: 10.1200/JCO.22.00643 . 36215675 PMC10489422

17 Palladini G , Hegenbart U , Milani P , et al. A staging system for renal outcome and early markers of renal ­response to chemotherapy in AL amyloidosis . Blood . 2014 ; 124 ( 15 ): 2325 – 2332 . doi: 10.1182/blood-2014-04-570010 . 25115890

18 Comenzo RL , Reece D , Palladini G , et al. Consensus guidelines for the conduct and reporting of clinical trials in systemic light-chain amyloidosis . Leukemia . 2012 ; 26 ( 11 ): 2317 – 2325 . doi: 10.1038/leu.2012.100 . 22475872

19 Wang W , Li J , Yang Y , et al. Update on the outcome of M-protein screening program of multiple myeloma in China: a 7-year cohort study . Cancer Med . 2024 ; 13 ( 1 ): e6859 . doi: 10.1002/cam4.6859 . 38133437 PMC10807560

20 Kastritis E , Fotiou D , Theodorakakou F , et al. Timing and impact of a deep response in the outcome of patients with systemic light chain (AL) amyloidosis . Amyloid . 2021 ; 28 ( 1 ): 3 – 11 . doi: 10.1080/13506129.2020.1798224 . 32713209

21 Kastritis E , Leleu X , Arnulf B , et al. Bortezomib, melphalan, and dexamethasone for light-chain amyloidosis . J Clin Oncol . 2020 ; 38 ( 28 ): 3252 – 3260 . doi: 10.1200/JCO.20.01285 . 32730181

22 Muchtar E , Dispenzieri A , Leung N , et al. Depth of ­organ response in AL amyloidosis is associated with ­improved survival: grading the organ response criteria . Leukemia . 2018 ; 32 ( 10 ): 2240 – 2249 . doi: 10.1038/s41375-018-0060-x . 29581546

23 Sidana S , Milani P , Binder M , et al. A validated composite organ and hematologic response model for early assessment of treatment outcomes in light chain amyloidosis . Blood Cancer J . 2020 ; 10 ( 4 ):41. doi: 10.1038/s41408-020-0306-5 . PMC7156647 32286270

24 Basset M , Milani P , Ferretti VV , et al. Prospective urinary albumin/creatinine ratio for diagnosis, staging, and ­organ response assessment in renal AL amyloidosis: ­results from a large cohort of patients . Clin Chem Lab Med . 2022 ; 60 ( 3 ): 386 – 393 . doi: 10.1515/cclm-2021-0912 . 35018751

25 Sanchorawala V , Palladini G , Minnema MC , et al. Health-related quality of life in patients with light chain amyloidosis treated with bortezomib, cyclophosphamide, and dexamethasone +/- daratumumab: results from the ANDROMEDA study . Am J Hematol . 2022 ; 97 ( 6 ): 719 – 730 . doi: 10.1002/ajh.26536 . 35293006

26 Palladini G , Paiva B , Wechalekar A , et al. Minimal residual disease negativity by next-generation flow cytometry is associated with improved organ response in AL amyloidosis . Blood Cancer J . 2021 ; 11 ( 2 ):34. doi: 10.1038/s41408-021-00428-0 . PMC7887224 33594045

27 Palladini G , Milani P. Diagnosis and treatment of AL amyloidosis . Drugs . 2023 ; 83 ( 3 ): 203 – 216 . doi: 10.1007/s40265-022-01830-z . 36652193

28 Cuenca I , Alameda D , Sanchez-Vega B , et al. Immunogenetic characterization of clonal plasma cells in systemic light-chain amyloidosis . Leukemia . 2021 ; 35 ( 1 ): 245 – 249 . doi: 10.1038/s41375-020-0800-6 . 32203144 PMC7787969

29 Warsame R , Kumar SK , Gertz MA , et al. Abnormal FISH in patients with immunoglobulin light chain amyloidosis is a risk factor for cardiac involvement and for death . Blood Cancer J . 2015 ; 5 ( 5 ): e310 – e310 . doi: 10.1038/bcj.2015.34 . 25933374 PMC4423220