Open

Preface

This risk-benefit analysis is part of Forever Healthy's "Rejuvenation Now" initiative that seeks to continuously identify potential rejuvenation therapies and systematically evaluate their risks, benefits, and associated therapeutic protocols to create transparency.

Section 1: Overview 

Motivation

Cellular senescence, a state of essentially irreversible replicative arrest, is one of the hallmarks of aging. Senolytics are drugs that act by selectively facilitating apoptosis of senescent cells by transiently disabling one or more of the senescent cell anti-apoptotic pathways (SCAPs) that enable senescent cells to survive.

Dasatinib & Quercetin (D+Q) were the first senolytic drugs to be discovered and as they have been shown to affect different SCAPs in vitro, targeting different types of senescent cells, they are often employed in combination.

It is supposed that intermittent dosing of D+Q in combination leads to the elimination of senescent cells in humans and by doing so, has the potential to delay, prevent or alleviate multiple age-related diseases and increase the healthy lifespan. 

Key Questions 

This analysis seeks to answer the following questions:

• Which benefits result from D+Q senolytic therapy? 

• Which risks are involved in D+Q senolytic therapy (general and method-specific)?

• What are the potential risk mitigation strategies?

• Which method or combination of methods is the most effective for D+Q senolytic therapy?

• Which of the available methods are safe for use? 

• What is the best therapeutic protocol?

• What is the best treatment monitoring strategy available at the moment?

Impatient readers may choose to skip directly to Section 5 for the presentation of the results. 

Section 2: Methods

Analytic model

This RBA has been prepared based on the principles outlined in A Comprehensive Approach to Benefit-Risk Assessment in Drug Development (Sarac et al., 2012). 

Literature search

A literature search was conducted on PubMed and the Cochrane Library using the search terms shown in Table 1 and includes results available as of April 17, 2020. Titles and abstracts of the resulting studies were screened and relevant articles downloaded in full text. The references of the full-text articles were manually searched in order to identify additional studies that may have been missed by the search terms.

Inclusion criteria: All studies (clinical, preclinical, in vitro) that tested D or Q or the combination as senolytics were included. In order to assess the adverse effects of each compound, we also included studies in humans that were performed for the usual indications of the drugs.

Exclusion criteria: We excluded studies that used combined chemotherapy regimens from our analysis as well as preclinical studies in our assessment of adverse effects. 

Table 1: Literature Search 

Recommended Reading

General Introduction

The following sites offer information on Dasatinib & Quercetin senolytic therapy at a consumer level and are useful as an introduction to the topic:

• The Scripps Research Institute – Dasatinib and Quercetin - lifespan.io

• First-in-human trial of senolytic drugs encouraging Small pilot study points to feasibility of larger trials in age-related diseases - sciencedaily.com

• Senolytics - thelri.org 

• Young anti-aging field takes big step with Mayo Clinic senolytics showcase - Endpts.com
  
  
  

Scientific Overview 

The following scientific reviews provide a more detailed overview of the topic of senolytic therapy:

• Discovery, development, and future application of senolytics: theories and predictions

• Cellular Senescence: A Translational Perspective

Abbreviation list

Section 3: Existing evidence

Summary of results

We screened 3,343 papers and included 156 in our analysis. We identified 118 relevant human studies that used D or Q, 111 of which were related to side-effects or safety. In total, there have only been 3 trials that used D+Q as senolytics in human subjects. Two of the clinical trials were of relatively high quality but were both small, phase I, open-label studies (n= 9,14) on subjects with pre-existing diseases (lung and chronic kidney) (Hickson et al., 2019; Justice et al., 2019). A corrigendum with a reanalysis of data from one of the trials was also included (Hickson et al., 2020). The third trial is a randomized control trial (RCT) of low quality but did have 4 test groups (D+Q, D+placebo, Q+placebo, placebo+placebo) and enrolled healthy participants (Tkemaoadze & Apkjazava, 2019).

A fourth study in which senescent cell markers from skin biopsies were measured retrospectively (dasatinib only) was also chosen for inclusion. Additionally, there are 4 trials listed on clinicaltrials.gov that are expected to publish results over the next 3 years. 

We identified 31 preclinical trials related to the use of D+Q as senolytics, alone or in combination. 12 of the studies investigated the senolytic effects of Q alone. We included another 7 preclinical studies that provided possible mechanisms for side effects encountered in clinical trials. 

Clinical trials

Table 2: Clinical trials

Preclinical trials

Table 3: Preclinical trials

Section 4: Risk-Benefit Analysis

Decision Model

Risk and benefit criteria

The decision profile is made of up risk and benefit criteria extracted from the outcomes of the above-mentioned papers. The benefit criteria are organized by category and include the type, magnitude, and duration of the benefit as well as its perceived importance to the patient. The risk criteria are organized by category, type, severity, frequency, detectability, and mitigation. All are assigned numerical values: 

1 = low

2 = moderate

3 = high

The numerical values for both risk and benefit criteria are then summarized serving as the justification for the weighting in the following column.

Weight

The criteria are weighted on a value scale to enable comparison (based on the relative importance of a difference). Risk and benefit criteria are assigned to either low (1-1.66), medium (1.67-2.33), or high (2.34-3) weighted categories.

Weighting is independent of data sets and the final weights are based on consensus with justification based on the preceding columns of the table.

Score

Each category is assessed according to the performance of D+Q senolytic therapy against the comparator (physiological aging) whereby a numerical value is assigned for each criterion -1 (inferior), 0 (equivalent or non-inferior), and +1 (superior) to the comparator.

Uncertainty

Uncertainty is determined according to the amount and quality of the evidence, whether it came from human or animal studies and whether methodological flaws, conflicting studies, or conflicts of interest (funding) by the authors are present. Evidence that is based on human RCTs or systematic reviews is initially assigned an uncertainty score of 1, evidence from open-label trials is assigned a score of 2, and evidence that is based on observational studies, and preclinical trials is assigned a score of 3. The uncertainty score is then adjusted by upgrading or downgrading using the above-mentioned criteria. 

Weighted score

The weights and scores are multiplied to produce weighted scores that enable direct comparison (-3 → +3) and then adjusted using the uncertainty score. Weighted scores may be upgraded where the uncertainty of the evidence is low or downgraded where the uncertainty of the evidence is high. 

Benefit assessment 

Our analysis identified a total of only 8 benefits that have been documented in human studies and another 46 benefits from preclinical trials (Table 4). Of the 8 benefits in humans, 5 were actually various measurements of markers of senescence or the SASP, hypothesized to translate to clinically beneficial effects. Only 3 benefits had any direct clinical relevance and they were of low magnitude. Based on the current state of evidence, the beneficial effects of D+Q seem to be extremely limited in humans.

However, the benefits identified in the preclinical studies are significant and encompass many organ systems. Of note, several of the benefits only occurred in diseased populations (ie. diabetic mice) and did not extend to control populations that received treatment with D+Q.

Table 4: Benefit assessment

Benefits from clinical trials

Markers of Senescence

An open-label phase 1 clinical trial (n=9) of a 3-day oral course of D+Q (100 mg + 1000 mg) in patients with chronic kidney disease (aged 50-80) was the first to measure a decrease in the number of several key markers of senescence (Hickson et al., 2019).

The number of p16INK4a+ cells was reduced by 35% in adipose tissue biopsies and 20% in the epidermal layer (although the result did not reach statistical significance). Epidermal p16INK4a cells have been associated with cardiovascular disease (CVD) risk and "aging" (Waaijer et al., 2012). 

The number of p21CIP1+ cells was also decreased (Hickson et al., 2019). The raw values were reduced by 17% in adipose tissue biopsies and 31% in the epidermis. That the reductions occurred in both adipose tissue and skin suggests that D+Q treatment works systemically to decrease senescent cell burden.

D+Q also reduced the number of SABgal+ cells by 62% and decreased the number of macrophages per adipocyte by 28% (Hickson et al., 2019). Senescent cells have been shown to attract, activate, and anchor macrophages in adipose tissue (Hickson et al., 2019).

Senescent cells and macrophages contribute to the formation of the "crown-like structures" (CLS) characteristically found in adipose tissue in diabetes and obesity. The study reported 86% fewer CLS per adipocyte following treatment with D+Q (Hickson et al., 2019). 

Senescent and pre-senescent cells have no or limited replicative potential, resulting in increased population doubling times as they accumulate. The trial also found there was an increase in the number of primary adipocyte progenitors which is consistent with the effects of removing senescent cells (Hickson et al., 2019).

While as of yet, there is no ideal marker for senescent cells, the changes in the several markers mentioned above indicate that treatment with D+Q is likely effective as a senolytic in humans. It appears that senolytics work by facilitating apoptosis of senescent cells due to their SASP, not by targeting all cells expressing pINK4a (Hickson et al., 2019).

The changes in multiple tissues (skin, adipose tissue, plasma) suggest that oral administration of D+Q decreases overall senescent cell burden rather than targeting cells within a single organ or structure (Hickson et al., 2019).

Decreases in circulating SASP factors/gene expression

An open-label trial (n=9) found that there was a decrease in circulating SASP factors (plasma IL-1a, IL-2, IL- 6, IL-9 and MMP 2, MMP 9, and MMP 12) following 3 days of senolytic treatment (Hickson et al., 2019). FGF-2, GM-CSF, and IL-1RA also tended to be lower but did not reach significance. A recently published reanalysis of the data found that the composite score of the SASP was significantly reduced despite only MMP-12 being decreased significantly in isolation (Hickson et al., 2020).

A second open-label trial (n=14) in patients with idiopathic pulmonary fibrosis (IPF) found that select SASP proteins including IL-6, MMP-7 and TIMP2 showed a trend towards reduction (8 participants had reductions in circulating amounts) following treatment with D+Q 3 days per week for 3 weeks (Justice et al., 2019).

An analysis of SASP gene signatures in skin biopsies from a trial (n=12) that used D (100 mg) for 169 days to treat systemic sclerosis-associated interstitial lung disease (Martyanov et al., 2019) found that in the subset of patients that responded to D treatment (n=3) SASP levels were both higher at baseline and, significantly lower post-treatment compared with non-improvers. 80.3% (53/66) of the SASP gene signatures showed a decrease in expression post-treatment which was correlated with clinical improvements (vs. 53% (35/66) in non-improvers).

Cardiovascular system

One RCT (n=64) in healthy volunteers (over the age of 36 years) reported a significant reduction in post-exercise systolic blood pressure at 10 and 20 minutes in the group that received treatment with D+Q for 5 days (Tkemaoadze & Apkjazava, 2019). The initial blood pressure in all groups was approximately 115 mmHg and decreased to 108 mmHg in the D+Q group at 10 minutes after the completion of the "stair-ascending test" while the BP of the control group decreased to 112 mmHg. The difference was still significant at 20 minutes post-exercise but reached the same value in all groups by 30 minutes.

Increased physical function in IPF

An open-label trial reported improvements in physical function that included improved 6-min walk distance, 4-m gait speed, and 5-repeated chair-stand times (Justice et al., 2019). These improvements were consistent with preclinical findings of improvements in treadmill endurance and frailty following senescent cell removal in various murine models.

"Lightness" in joints

One RCT (n=64) in healthy volunteers reported that nearly all participants in the D+Q group experienced a feeling of "lightness" in the joints the day after treatment (Tkemaoadze & Apkjazava, 2019).

Benefits in preclinical trials

Health/Lifespan

A trial that used intermittent treatment with D+Q (5 mg/kg + 50 mg/kg) weekly in an accelerated aging mouse model found that healthspan was significantly extended (Zhu et al., 2015). They reported a significant reduction in a composite score of age-related symptoms that included kyphosis, dystonia, tremors, loss of grip strength, coat condition, ataxia, urinary incontinence, impaired gait, hind limb paralysis, and poor body condition. The extension of healthspan was due to both the delay in onset of symptoms and the attenuation of their severity (Zhu et al., 2015).

A second study reported that bi-weekly administration of D+Q (5 mg/kg + 50 mg/kg) starting at 24-27 months of age (equivalent to age 75-90 years in humans) resulted in a 36% higher median post-treatment lifespan and lower mortality hazard (64.9% compared to the control group) (Xu et al., 2018).

Central nervous system

Three preclinical trials in mice reported beneficial effects in the CNS due to the elimination of senescent cells (Ogrodnik et al., 2019; Zhang et al., 2019; Musi et al., 2018). The first trial demonstrated that obesity results in the accumulation of senescent glial cells in the region of the lateral ventricle and that senescent glial cells exhibit excessive fat deposits. Clearance of senescent cells using D+Q (5 mg/kg+ 50 mg/kg) for 5 days every two weeks over 8 weeks restored neurogenesis and alleviated anxiety-related behavior (Ogrodnik et al., 2019).

A second trial (Zhang et al., 2019) found that exposure to amyloid-beta (Aβ) plaques triggered senescence in oligodendrocyte progenitor cells (OPCs) and that short-term treatment with D+Q (12 mg/kg + 50 mg/kg) daily for 9 days reduced SA-BGal activity and levels of Olig2 and p21. When Alzheimer's disease (AD) mice received D+Q over a longer period of 11 weeks, there was a decrease in Aβ load, and neuroinflammation (as evidenced by decreases in IL-1, 6, TNFa) as well as improvements in cognition.

Using AD transgenic mouse models, a third trial (Musi et al., 2018) found that neurofibrillary tangles (NFT), but not Aβ plaques, display a senescence‐like phenotype and that intermittent treatment with D+Q (5 mg/kg+ 50 mg/kg) in 6 sessions over 12 weeks reduced the number of NFT-containing cortical neurons by 35%. The gene expression of the NFT-associated senescence gene array was also reduced. The reduction in NFT-containing neurons corresponded with a decreased ventricular volume pathology of 28% and a reduction in cortical brain atrophy. Aberrant cerebral blood flow was improved to the point that it no longer differed significantly from controls and the D+Q treated mice displayed higher levels of neurogenesis markers (Musi et al., 2018).

Cardiovascular system

Four preclinical studies reported benefits to the cardiovascular system following treatment with D+Q (Roos et al., 2016; Zhu et al., 2015; Kim et al., 2020; Lewis-McDougall et al., 2019).

The first trial, assessed the effect of D+Q ( 5 mg/kg + 10 mg/kg) once per month for 3 months in aged and atherosclerotic mice (Roos et al., 2016). The authors reported a significant reduction in senescent cell markers in the medial layer of the aorta but not in intimal atherosclerotic plaques although intimal plaque calcification was decreased.

A single dose of D+Q (5 mg/kg + 50 mg/kg) has been shown to improve left ventricular ejection fraction in mice by approximately 10% (from 68% baseline up to 78% following treatment) due to improvements in end-systolic cardiac dimensions (Zhu et al., 2015).

D+Q treatment also improved vasomotor function in two trials (Zhu et al., 2015; Roos et al., 2016) as measured by a greater response to stimulation with acetylcholine and nitroprusside (Zhu et al., 2015). The data suggest that senolytic treatment improves nitric oxide signaling in aged mice, however, the molecular mechanisms are unclear. 

In vitro, Q has been shown to alleviate oxidative-stress induced vascular smooth muscle cell senescence through activation of AMPK (Kim et al., 2020). 

Elimination of senescent cardiac progenitor cells (CPCs) using D+Q has been shown in vitro to abrogate the SASP and in vivo, to activate resident CPCs (Lewis-McDougall et al., 2019). 

Improved cardiac diastolic function following D+Q treatment was reported by a study in obese mice (Palmer et al., 2019). 

Genes

Incubation with Q (3-12 µM for 24 hours) has been shown to increase the expression of SIRT1 and thioredoxin in a dose-dependent manner in human kidney cells (Abharzanjani et al., 2017).

Immune system

One trial reported a decrease in the inflammatory aspects of IPF in bronchoalveolar lavage (BAL) fluid following treatment with D+Q. White blood cell counts were significantly increased in vehicle-treated bleomycin-exposed mice, and treatment with D+Q attenuated this increase. Although cytokine levels within the BAL fluid were highly variable, the increases in MCP-1 and IL-6 were diminished following treatment with D+Q (Schafer et al., 2017).

In vitro studies of Q also reported a decrease in the level of reactive oxygen species (ROS) (Geng et al., 2019; Sohn et al., 2018). One trial reported decreased ROS levels and restoration of the heterochromatin architecture in a model of Werner's syndrome in human mesenchymal stem cells (Geng et al., 2019). A second study demonstrated that treatment with Q (5 uM) significantly decreased the relative ROS level when cells were exposed to H202 (Sohn et al., 2018). 

An in vitro study reported that cancer cells became more sensitive to radiation therapy following treatment with D+Q (Wang et al., 2020).

Markers of senescence

As in the human trials, a large number of "benefits" are related to reductions in markers of senescence or increases in cell proliferation capacity. Senescent cells often express p16INK4a, a cyclin-dependent kinase inhibitor, tumor suppressor, and biomarker of aging, which renders the senescence growth arrest irreversible (Coppé et al., 2011). Several in vivo (Nath et al., 2018; Schafer et al., 2017; Kim et al., 2019; Zhu et al., 2015) and in vitro (Parikh et al., 2018; Schafer et al., 2017; Suvakov et al., 2019; Geng et al., 2019; Kim et al., 2020; Yang et al., 2014 ) studies have demonstrated decreased p16Ink4a expression following treatment with or exposure to D+Q. This decrease has been measured in fetal airway cells, veins, lung fibroblasts, mesenchymal stem cells, renal tubular cells, liver, and muscle. 

Several in vivo (Ogrodnik et al., 2019; Xu et al., 2018; Zhu et al., 2015) and in vitro (Chondrogianni et al., 2010; Parikh et al., 2018; Abharzanjani et al., 2017; Geng et al., 2019; Kim et al., 2020; Sohn et al., 2018) studies also reported a decrease in the number of SABgal+ cells, another important marker of senescence. In a mouse model, the decrease in SABGal+ cells in perigonadal adipose tissue was approximately 7% following D+Q treatment (Ogrodnik et al., 2019) while another study reported a decrease of approximately 9.5% in human explanted adipose tissue (Xu et al., 2018). A third study also reported a decrease in SABgal+ cells in the inguinal fat of irradiated mice following a single dose of D+Q (Zhu et al., 2015).

In vitro, treatment of HLF-1 cells with Q resulted in only 13.5% of cells staining positive for SABgal after 55 days (compared to treatment with DMSO or CAP that showed >75% SABgal+ staining) (Chondrogianni et al., 2010). Dose-dependent decreases in SABgal+ cells following treatment with D and/or Q have been seen under various senescence-inducing conditions including hyperglycemia, hyperoxia and chemotherapy (Abharzanjani et al., 2017; Geng et al., 2019; Yang et al., 2014; Parikh et al., 2018). 

Several studies also reported a decrease in p21+ cells following treatment with D+Q (Zhang et al., 2019; Hohmann et al., 2018; Parikh et al., 2018; Geng et al., 2019; Kim et al., 2020; Yang et al., 2014). The mean intensity of p21+ cells decreased from 2800 down to 800 following short term (9 days) of D+Q treatment in AB plaques in a mouse model of Alzheimer's disease (Zhang et al., 2019). Treatment with Q (30 mg/kg intraperitoneally, over a period of 1 or 3 weeks also reduced p21 expression in bleomycin-induced lung injury in aged mice at 14 days (Hohmann et al., 2018).

In vitro studies also showed a decrease in levels of p21 following treatment with Q alone (Geng et al., 2019; Kim et al., 2020) and demonstrated an inhibitory effect on vascular smooth muscle cell (VSMC) senescence via activation of AMPK (Kim et al., 2020). Immunofluorescence analysis of D+Q incubated fetal airway smooth muscle cells showed decreased nuclear co-localization of p21 and p-γH2A.X from 65% down to 45% (Parikh et al., 2018).

Q has also been shown to reduce the expression of p19-ARF in the lungs (Hohmann et al., 2018) and kidneys (Kim et al., 2019) of aged and high-fat diet-fed mice, respectively. 

Three studies on Q also reported a significant decrease in p53 expression following exposure to Q, in oxidative (H202) or high-fat diet-induced metabolic stress (Kim et al., 2019; Kim et al., 2020) and in adriamycin and replicative senescence (Yang et al., 2014 ). However, one study reported an increase in p53 expression following D+Q treatment (Cavalcante et al., 2019).

In vitro, Q has also been shown to reduce markers of DNA damage including yH2AX and 53BP1 (Geng et al., 2019). The relative expression of cells double-positive for both markers decreased from 1 to 0.6 following exposure to Q (Geng et al., 2019). In mice, D+Q treatment has been shown to reduce yH2AX in liver biopsies 17% down to 11% (Ogrodnik et al., 2017). 

Telomeres

An in vitro study demonstrated that telomere length was increased by 70% and cell proliferation was increased by >50% in a Werner Syndrome (WS) model of human mesenchymal stem cells when exposed to Q at a concentration of 100 nmol/L (Geng et al., 2019). 

Telomere-associated foci (TAFs) are sites of DNA damage within telomeres and are believed to be a more specific marker of senescence than SABgal (Xu et al., 2018). Several studies have reported a decrease in TAF cells in various tissues including the brain, aorta, and liver in mice and human explanted tissue (Ogrodnik et al., 2019; Roos et al., 2016; Xu et al., 2018; Ogrodnik et al., 2017). Levels of TAF+ cells were decreased from 34% down to 18% in perigonadal adipose tissue of obese mice (Ogrodnik et al., 2019), from 42% to 22% in the medial layer of the aorta in aged atherosclerotic mice (Roos et al., 2016), and from 16% to 5% in the liver of aged mice (Ogrodnik et al., 2017). Explanted human omental tissue from obese individuals exposed to 1 uM + 20 uM D+Q for 48 hours also showed a reduced number of TAF+ cells compared to controls (Xu et al., 2018).

SASP

Several studies found a decrease in a variety of SASP components in mice (Zhang et al., 2019; Hohmann et al., 2018; Schafer et al., 2017; Palmer et al., 2019), in ex vivo human tissue (Xu et al., 2018; Suvakov et al., 2019; Geng et al., 2019 ) and in vitro. 

Summary of measured SASP components

Based on decreases in the above markers, several studies reported decreases in the number of senescent cell types including HUVECs, lung fibroblasts, mouse embryonic fibroblasts, preadipocytes, bone marrow-derived mesenchymal stem cells, human dermal fibroblasts. 

Metabolic system

Two studies in mice showed improvements in the metabolic system (Ogrodnik et al., 2017; Palmer et al., 2019). Senescence-associated mitochondrial dysfunction reduces cellular fatty acid oxidation capability resulting in increased fat deposition (Ogrodnik et al., 2017). A reduction in hepatic fat deposition was reported (in conjunction with reduced TAF+ markers in hepatocytes) following treatment with D+Q in a mouse model of diabetes and hepatocyte senescence (measured by TAF and p21) was shown to correlate with the severity of non-alcoholic fatty liver disease (NAFLD) (Ogrodnik et al., 2017).

A second study reported significantly improved glucose tolerance and insulin sensitivity following D+Q (5+50 mg/kg) for either 5 consecutive days monthly or 3 consecutive days with 14 days between treatment rounds. HbA1c was 5.1% after D+Q vs. 5.3% in DIO mice (Palmer et al., 2019). The time course of metabolic improvement paralleled that of clearance of p16Ink4a+ cells.

Musculoskeletal system 

Elimination of senescent cells has been shown to both prevent and alleviate physical dysfunction in mice (Xu et al., 2018). When senescent cells were transplanted into young mice, D+Q prevented the decreases in hanging endurance, grip strength, and speed that were seen in vehicle fed mice. A single 5-day course of D+Q also alleviated the effects of transplanting senescent cells after they were already established. In older mice that received D+Q intermittently for 4 months beginning at month 20, physical dysfunction was also alleviated (Xu et al., 2018).

In a model of fibrotic lung disease, mice treated with D+Q ran, on average, >37% further to exhaustion on a graded treadmill test than bleomycin injured, vehicle-treated mice (Schafer et al., 2017). In mice that were irradiated, a single dose of D+Q, resulted in improved exercise time, distance, and total work performed to exhaustion on the treadmill. D+Q showed no effect in sham-irradiated mice. Senescent markers were reduced in muscle and inguinal fat 5 days after treatment. When retested at 7 months after the single treatment, exercise capacity was significantly better in the mice that had been irradiated than in vehicle-treated controls. D+Q-treated animals had endurance essentially identical to that of sham-irradiated controls (Zhu et al., 2015).

D+Q administration has also been shown to affect trabecular bone microarchitecture positively (Farr et al., 2017). Histological examination showed fewer osteoclasts and femur cortical thickness and bone strength were higher in the D+Q group. The number of senescent osteocytes decreased from 12% to 8%.

D+Q administered as a cocktail but not stand alone in irradiated mice, resulted in a significant recovery in the bone architecture of radiated femurs via a reduction in senescent cells as assessed by TIF+ osteoblasts and osteocytes, markers of senescence (p16Ink4a and p21), and key SASP factors (Chandra et al., 2020). 

Renal and reproductive systems

Renal podocytes in a diet-induced obesity mouse model showed increased expression of Wilms tumor protein, a measure of podocyte integrity and function, after D+Q treatment (Palmer et al., 2019). There was no effect in the non-obese group that received D+Q. Furthermore, a decreased urinary albumin to creatinine ratio (ACR), an indicator of renal dysfunction, was reported.

Treatment with Q alone (50 mg/kg) for 5 days every two weeks for 10 weeks was shown to restore creatinine (from 0.5 to 0.35 mg/dl) and urinary microalbumin levels (45 ug/ml to 30 ug/ml) in obese mice (Kim et al., 2019). It also prevented renal cortical hypoxia in obese mice. However, in control mice fed quercetin, the results were not significant (Kim et al., 2019). 

In D+Q treated aged female mice, p53 was upregulated in uterine tissue and profibrotic factor miR34c was significantly reduced suggesting a possible anti-fibrotic effect (Cavalcante et al., 2019). 

Respiratory system

In a mouse model of lung fibrosis, D+Q was shown to increase compliance, almost back to the level of the controls (Schafer et al., 2017). The same study reported that D+Q caused a decrease in enhanced pause, an indirect measure of airway resistance and that bodyweight loss due to bleomycin lung injury was less in D+Q treated mice than in vehicle-treated mice. 

Post-radiation ulcer prevention and healing

Senescent cells accumulate after radiation exposure, which can induce cell and tissue dysfunction and skin or mucous membrane ulcers (Wang et al., 2020). An in vivo rodent study reported that clearance of senescent cells following treatment D+Q mitigated radiation ulcers (Wang et al., 2020).

Risk assessment 

We identified 56 risks that have occurred with D or Q therapy (Table 5). D has been used in humans for over 20 years and its side effect profile is well known. Q is well tolerated and has a very low incidence of adverse effects (Andres et al., 2017). How likely adverse effects are to occur with intermittent combined D+Q treatment is largely unknown.

In the two open-label human pilot trials there was only one serious adverse event reported (bacterial multifocal pneumonia and pulmonary edema superimposed on IPF) and no subjects required drug discontinuation (Hickson et al., 2019; Justice et al., 2019). However, these trials included a total of only 23 participants. Several patients did experience more serious respiratory symptoms (edema, effusion, dyspnea), as well as headache and GI discomfort but as the trials were performed on patients with preexisting disease, it is difficult to discern to what extent D was responsible. 

In the clinical trials, the reported adverse events were mostly mild to moderate in severity, reversible, without sequelae, and consistent with events reported in the placebo arms of RCTs. There was no evident decline in renal or hepatic function or evidence of cell lysis syndrome (Justice et al., 2019).

Table 5: Risk assessment 

Individual risks by system

Central Nervous system

Headache is amongst the most common side effects of D (40% of patients) (Medscape.com) and also occurred in the first human senolytic trials (Justice et al., 2019) as well as many of the cancer trials (Mayer et al., 2011; Yu et al., 2011; Lindauer & Hochhaus, 2018; Hartmann et al., 2009; Kim et al., 2018; Saglio et al., 2010; Huang et al., 2012; Breccia et al., 2016; Shah et al., 2008; Huang et al., 2018; Wong et al., 2018; Martyanov et al., 2017; Apperley et al., 2009; Yu et al., 2009; Takahashi et al., 2011; Kantarjian et al., 2010 ).

It is a common initial side effect and can occur following the first dose. Tyrosine kinase inhibitor (TKI)-induced hypertension should be ruled out as a cause (Steegman et al., 2016). The risk of headache risk can be minimized by taking the first dose at bedtime, drinking plenty of water to stay hydrated, and taking extra magnesium.

Neuropathy was described in a case report but occurred after 6 months. It was suggested to be mediated by an immune mechanism as it responded to treatment with intravenous immunoglobulins and drug discontinuation (Ishida et al., 2017). Other sources report that neuropathy occurs in as many as 31% of patients taking D (Bristol-Myers). 

There is an increased risk of stroke in patients taking D, particularly if they are already "high-risk" for CVD (Assunção et al., 2018). Overall, the risk of stroke is low and incidents occurred during long term chronic use with the first incidence occurring at 1095 days after the start of treatment. 

Severe insomnia was reported as an adverse event in one clinical trial (Schilder et al., 2012; Martyanov et al., 2017). No mention was made of the time insomnia occurred. Insomnia that resulted in only 2-3 hours of sleep was also described in a case report in which the patient was taking a lower dose of dasatinib, 25 mg/day on alternate weeks, although he had taken higher doses in the past (Sami et al., 2014). Initial clinical trials on TKIs reported insomnia in 1-10% of patients (fda.gov). 

Depression/agitation and poor mental health have been reported in approximately 1-10% in early clinical trials of patients taking dasatinib (Sami et al., 2014). We only identified one case report that reported severe depression and agitation (Sami et al., 2014). 

Two case reports involved spontaneous subdural hematomas in patients receiving D. The first patient had a normal platelet count (Mustafa Ali et al., 2014). The second case was bilateral and occurred in a patient shortly after initiation of D who had a reduced platelet count (although not to the point of expecting spontaneous bleeding) (Yhim et al., 2012). Dasatinib is known to cause broad-spectrum inhibition of kinases, including PDGFR-b, a receptor expressed in pericytes that is known to play an important role in angiogenesis and vessel wall formation. Inhibition of PDGFR-b by dasatinib could induce mechanical instability of the capillary wall (Mustafa Ali et al., 2014).

Dizziness was experienced by 13% of patients in a 6-month trial that used D to treat systemic sclerosis-associated interstitial lung disease although the cases believed to be caused by D were only 3.2% (Martyanov et al., 2017). There was no mention of the time of onset. A second trial (n=174) reported that dizziness occurred in 10% of subjects (Apperley et al., 2009) and a third trial (n=54) reported dizziness in 5.4% of patients (Wong et al., 2018). 

Syncope was reported as an adverse event in a trial that used D to treat sarcoma. Again, the time of onset was not mentioned but likely to be within a few months as the trial was on advanced sarcoma and didn't show any benefit (Schuetze et al., 2015).

Cardiovascular system 

Pulmonary arterial hypertension (PAH) has been reported as an adverse event in several clinical trials and case reports (Suh et al., 2017; Gora-Tybor et al., 2015; Huang et al., 2018; Yurttaş & Eşkazan, 2018; Fox et al., 2017; Fox et al., 2017; Lindauer & Hochhaus, 2018; Cortes et al., 2016), mostly as a complication related to chronic D use over years (Suh et al., 2017). One study reported that 6.8% of patients suffered PAH and that the earliest time of onset 10 days after treatment initiation (Kim et al., 2013) though means have been reported between 34 (Yurttaş & Eşkazan, 2018) and 42 months (Weatherald et al., 2017).

Many patients recover after discontinuation of D (Orlikow et al., 2019) but 37% of patients experience persistent PAH (Weatherald et al., 2017). There are currently no known biomarkers or methods to identify patients predisposed to D-induced PAH. PAH is often preceded by pleural effusions (Gora-Tybor et al., 2015) and baseline pretreatment chest x-ray and echocardiography could help rule out pre-existing pleural effusions, pulmonary hypertension, and intracardiac shunts.

One of the main differences between dasatinib and the other TKIs is that it additionally inhibits Src. It is speculated that Src inhibition may play a role in the development of dasatinib-induced PAH. Src tyrosine kinase is expressed abundantly in vascular tissue, and activation of Src appears to play a crucial role in smooth muscle cell proliferation and vasoconstriction.

It has also been shown that dasatinib may cause direct pulmonary endothelial damage in humans and rodents, attenuating hypoxic pulmonary vasoconstriction responses, and increasing susceptibility to PAH (Yurttaş & Eşkazan, 2018 ). An in vitro study found that dasatinib dramatically inhibits endothelial cell tube formation which is essential for proper function and angiogenesis (Gover-Proaktor et al., 2018) providing a possible mechanistic explanation for its effects on the vascular system.

Vascular occlusive events were reported in 4.78% of patients taking second-generation TKIs (n=3000) but no time of onset was reported (Haguet et al., 2016). A second systematic review of 3043 patients found an odds ratio of 3.86 for vascular occlusive events in D compared to Imatinib (Douxfils et al., 2016), a first-generation TKI.

A meta-analysis of cardiac ischemic events (myocardial infarction, angina, coronary artery disease, acute coronary syndrome) in D-treated patients (n=2712) found a frequency of 2-4%. Most events occurred within a year with the majority occurring in the first 6 months (Saglio et al., 2017). That most events occurred in the first 6 months, support the lack of a cumulative drug effect, although additional studies are required to help determine the mechanism for the development of these events early after initiation of therapy. The most distinguishing event was myocardial infarction, where seven patients in the D group and one patient in the placebo arm experienced a heart attack. Approximately 80% of ischemic events occurred in patients who had a history of and/or risk factors for atherosclerosis. A smaller retrospective analysis (n=105) also reported a 4% rate of vascular events (Gora-Tybor et al., 2015).

Palpitations were reported by 10.5% of patients on D in a retrospective analysis (n=90) (Chen et al., 2018). Time of onset was not mentioned but the information on adverse effects was collected at 8 months. Palpitations were also reported in 2 patients in a D trial on sarcoma (Schuetze et al., 2015) and are listed as a potential side effect in the Food & Drug Administration (FDA) information page on D (fda.gov) where it states that it occurred in 7% of patients in clinical trials. 

Chest pain was reported by multiple studies (Chen et al., 2018; Bergeron et al., 2007; Wong et al., 2018). One study reported that 5% (2/40) patients developed chest pain (Bergeron et al., 2007). One of the patients developed abnormalities at 33 days and the other at 463 days. A second study reported 1.8% (1/57) of patients had chest pain (Chen et al., 2018) and a third study (n=54) reported a 6% incidence of chest pain with no mention of the time of onset (Wong et al., 2018). 

Pericardial effusion (+/- cardiac tamponade) has been reported as an adverse effect in several clinical trials and case reports at varying frequencies that appear to be dose-dependent.

Studies reporting pericardial effusion as an adverse effect

NR= not reported

There were two case reports of massive pericardial effusion that progressed to life-threatening cardiac tamponade (Wattal et al., 2017; Rajakariar et al., 2018 ). Both cases resolved with pericardiocentesis, steroid therapy and discontinuation of D. The earliest case of tamponade occurred 3 weeks after initiation of D (Rajakariar et al., 2018). Like other types of effusions, these are likely due to effects on the endothelium. D causes profound, dose-dependent disorganization of the endothelial cell monolayers via the disassembly of cell-cell contacts, altered cell-matrix contacts and altered wound healing (Kreutzman et al., 2017) presenting a likely mechanism for the increased risk of pleural and pericardial effusions and bleeding tendency (Phan et al., 2018).

An increased risk of heart failure for D compared to other TKIs was reported through the analysis of a pharmacovigilance database. The authors found a 2.52 adjusted odds ratio that D is associated with cardiac failure (de Campaigno et al., 2017). Congestive heart failure or cardiac dysfunction was reported in 2% of patients (5/258) after a 1-year year followup (Medeiros et al., 2018). Right-sided heart failure has been reported as soon as a few days following the initiation of D (100 mg/day) (Krauth et al., 2011).

A study in dogs (Izumi-Nakaseko et al., 2019) reported that D decreased the heart rate and cardiac output in a dose-dependent manner. One animal showed impaired left ventricular mechanical function for 45 min. D did not alter pulmonary artery pressure. D-induced heart failure has been correlated to the inhibition of non-receptor type protein kinase ABL1 and ABL2 based on pharmacovigilance data (Izumi-Nakaseko et al., 2019).

An effect on the electric conducting system of the heart has also been reported in several clinical studies. ECGs showed a slight increase in the average QT interval (n=143) (Schuetze et al., 2015). Only two patients had a QT interval that was lengthened to >500 ms. American prescribing information reports a mean change in the QT interval of 7-13 ms and advises caution with D in patients at increased risk for QT prolongation (Medeiros et al., 2018) as 1% of patients in clinical trials had clinically relevant QT interval prolongation. Other studies also reported a prolonged QT interval (Wong et al., 2018; Yu et al., 2009) and Grade 1 ECG changes (Apperley et al., 2009). Patients should be assessed for risk of QT prolongation based on medical history and medication. Low potassium or magnesium levels should be corrected in advance and then monitored (Medeiros et al., 2018). 

Arrhythmias have been reported in several studies. The earliest time of onset in the studies we identified was 21 days (Assunção et al., 2018). A study reported one episode of atrial tachyarrhythmia and one episode of ventricular tachyarrhythmia (Schuetze et al., 2015). In another analysis, one patient (n=100) had a grade 2 arrhythmia (Apperley et al., 2009). D-associated aggravation of a preexisting arrhythmia was also reported (Sprechbach et al., 2013).

The United States FDA approval summary, which is based on safety data from 911 patients, reports two cases of patients with asymptomatic non-sustained ventricular tachycardia and the database of the manufacturer of dasatinib records three cases of nonfatal arrhythmias (Sprechbach et al., 2013). 

Mechanistically, D has been shown to increase the conduction speed in cardiac cells, a feature that can be explained by c-Src tyrosine kinase inhibition (Izumi-Nakaseko et al., 2019). 

Hypotension has been reported as a frequent side effect (1-10%) in the FDA approval sheets (fda.gov) as well as in an open-label trial (Schuetze et al., 2015). 

Endocrine

Thyroid abnormalities were reported in 70% of patients under treatment with D in one small trial (n=10) (Kim et al., 2010). 5 patients presented with hypothyroidism and 2 with hyperthyroidism. Hyperthyroidism occurred earlier, at a mean of 6 weeks whereas hypothyroidism occurred at a mean of 22 weeks. However, the earliest time of onset in both cases was 1 week. The FDA approval documents describe hypo/hyperthyroidism as occurring less than 1% of the time (fda.gov). Due to the transient nature of drug-induced hypo- or hyperthyroidism, as well as the mild clinical course with lack of symptoms in all but two patients, the therapeutic relevance of early diagnosis of hypothyroidism or hyperthyroidism is unclear.

Eye

Three studies reported adverse effects on the eye. Loss of vision deemed possibly-related to D was reported in an open-label trial (n=54) (Wong et al., 2018). The time of onset was not reported. In one retrospective analysis (n=109), papilledema occurred in one patient. A case report describes the development of optic neuropathy 2.5 months after initiation of D that improved with the use of corticosteroids and discontinuation of D (Monge et al., 2015). The mechanisms by which TKIs could produce optic neuropathy remain unclear. 

Fatigue

Fatigue and/or weakness has been reported as a common side effect of D in numerous trials. Most cases were mild-moderate and occur as early as the first day of treatment.  

Studies reporting fatigue as an adverse effect