The submitted nonclinical studies were in accordance with regulatory guidelines for registration of a new chemical entity (TAF), and all pivotal safety studies were good laboratory practice (GLP) compliant. The ICH M3 (R2) guideline10 states that toxicity studies are generally not warranted for drug combinations for HIV, and adequate scientific justification was provided for their absence. The FDA and EMA waived requirements for carcinogenicity and peri postnatal studies due to lack of TAF exposure in rats and TgRasH2 mice11, and lower TFV exposure after TAF administration. Most of the virology studies were provided in the clinical dossier.
Tenofovir alafenamide (TAF) is a prodrug for tenofovir (TFV), which is metabolised intracellularly to the active metabolite, TFV-DP. Intracellularly, TFV-DP competes with deoxyadenosine triphosphate for incorporation into nascent DNA, and acts as a DNA chain terminator during HIV reverse transcription.
The lysosomal carboxypeptidase Cat A is essential to the intracellular activation of TAF12. Cat A activity in primary cluster determinant 4 (CD4)+ T cells and monocyte derived macrophages showed similar levels of activity in cells from 13 donors of differing age, gender, ethnicity and sex (Study PC-120-2017). Some viral protease inhibitors may strongly inhibit Cat A, and in vitro assays with purified Cat A (Study PC-120-2001) showed that telaprevir and boceprevir, registered inhibitors of the Hepatitis C virus (HCV) protease, strongly inhibited Cat A mediated hydrolysis of TAF, with respective IC50 values of 0.3 and 0.2 µM, 6 x and 8 x below the clinical Cmax adjusted for plasma protein binding. The more recently registered inhibitors of HCV protease, simeprevir, asunaprevir and paritaprevir, were not tested. The HIV protease inhibitors darunavir, atazanavir, lopinavir and ritonavir, and COBI, did not inhibit Cat A mediated hydrolysis of TAF at concentrations up to 50 µM. The precautions section of the draft Genvoya PI states that ‘Genvoya should not be used in conjunction with protease inhibitors .… due to potential drug-drug interactions including altered and/or suboptimal pharmacokinetics of cobicistat, elvitegravir, and/or the co-administered products’.13
TAF showed anti-HIV-1 activity in vitro against wild type clinical isolates from Group M (subtypes A to G), and Groups N and O, and HIV-2, in human activated PBMCs, with concentration of a compound inhibiting virus replication by 50% (EC50) values ranging from 0.14 to 12.0 nM (Study PC-120-2004). TAF mean EC50 values were 11.0 ± 11.3 nM and 9.7 ± 4.6 in CD4+ T cells and monocyte derived macrophages from diverse human donors (Study PC-120-2017). TAF showed a range of synergistic activity, and no antagonism, when tested in pair wise combinations with nucleoside reverse transcriptase inhibitors (NRTIs) (including emtricitabine), nonnucleoside reverse transcriptase inhibitor (NNRTIs), protease inhibitors and integrase strand transfer inhibitors (INSTIs) (including elvitegravir) in vitro in MT-2 cells infected with HIV-1 IIIB (Study PC-120-2002).
In a dose escalation resistance study, TAF and TFV showed nearly identical resistance profiles, with emergence of the K65R ± S68N mutation in reverse transcriptase, associated with up to 6.5 fold phenotypic resistance to TFV (Study PC-120-2011). TAF and TFV also had identical phenotypic resistance profiles against a panel of 24 patient derived HIV-1 isolates with a wide range of NRTI resistance mutations (Study PC-120-2014). These results are consistent with TAF and TFV having the same active anti-HIV component, TFV‑DP.
Tenofovir alafenamide (TAF) did not exhibit antiviral activity against the majority of a panel of 18 human viruses, other than HIV and Simian immunodeficiency virus (SIV). Tenofovir alafenamide weakly inhibited Herpes simplex virus -2 (HSV-2) KW (EC50 = 424 nM) and human parainfluenza virus (EC50 = 843 nM) (Study PC-120-2003). However, TAF was active against HBV, with an EC50 value of 18 nM.
Mitochondrial toxicity is a recognised potential adverse effect of long-term nucleoside analogue therapy, which may manifest as myopathy, peripheral neuropathy, and hepatic steatosis with lactic acidosis.14 A lack of selectivity for HIV reverse transcriptase (a viral DNA polymerase) may lead to binding to host DNA polymerases, including mitochondrial DNA (mtDNA) polymerase ϒ, which is required for replication of mtDNA. The kinetic inhibition (Ki) constant for TFV-DP against HIV-1 reverse transcriptase of 0.21 µM was significantly lower than its Ki for human DNA polymerases α (5.2 µM), β (81.7 µM) and ϒ (59.5 µM).15 The relative efficiency of incorporation of TFV-DP into DNA by human polymerases α, β and ϒ relative to natural 2’-deoxynucleoside triphosphates (dNTPs) was 1.4%, 1.3% and 0.06%.16 TAF did not deplete mtDNA in HepG217 cells over 10 days culture at concentrations up to 1 µM (Study PC-120-2006). These studies indicate selectivity of TFV-DP for HIV-1 reverse transcriptase in relation to host (patient) polymerases. Furthermore, examination of kidney, liver and skeletal muscle for indicators of mitochondrial toxicity in the monkey 4 week toxicity study showed no effects.
TDF may have effects on bone mineral density and turnover, however the TAF half maximal cytotoxic concentration (CC50) was > 500 µM (TAF plasma Cmax 484 nM after 25 mg dose) in human primary osteoblasts in vitro (Study PC-120-2008). Levels of TAF in primary osteoblasts were similar to those in PBMCs. Cytotoxicity studies in human resting and dividing PBMCs (Study PC-120-2009), lymphoblastoid cell lines MT-2 and MT-4, the hepatic cell line HepG2 (Study PC-120-2007), human myeloid and erythroid progenitor cells from 3 donors (Study PC-120-2016), and HEK293T cells expressing renal organic anion transporters (OAT) OAT1 and OAT3, showed no significant TAF toxicity at clinically relevant concentrations.
The sponsor submitted five studies investigating the safety pharmacology effects of TAF in vitro and in vivo. Rats administered a single oral dose of 1000 mg/kg TAF showed no biologically significant signs of a pharmacological effect on the central nervous system (Study R990188).
In vitro cardiovascular investigations found that TAF (≤ 10 µM) does not statistically significantly inhibit hERG when compared with vehicle control values, with the IC50 inhibitory effect of TAF on hERG potassium current estimated > 10 µM (Study PC-120-2005). An in vivo study in dogs found no treatment related effects on ECG, heart rate or systemic blood pressure at 100 mg/kg (Study D2000006). In a 9 month repeat dose toxicity study in dogs administration of ≥ 6 mg/kg/day produced a statistically significant prolonged PR interval18 (Study TX-120-002). However, mild PR prolongation has occasionally been seen in normal relaxed dogs and a definitive relationship of an effect of PR prolongation with exposure to TAF cannot be confirmed by this study due to the small number of dogs assessed. The dog study observed a statistically significant reduced heart rate with QT prolonging effect at week 39 in animals administered 18/12 mg/kg/day, which were reversed following the 3 month recovery period. A reduction in serum Triiodothyronine (T3) levels at the high dose (HD) might have been associated with the heart effects.
A renal function study in rats (Study R990186) following single oral administration of TAF ≤ 1,000 mg/kg found that although urinary output of calcium was increased at 1,000 mg/kg, there was a correlation with an increase in serum calcium concentration which indicated that the kidneys were functioning in order to reduce the serum calcium load. The no observed adverse effect level (NOAEL) on male rat kidney function was 1,000 mg/kg. TAF 100 mg/kg had no clear effect on gastric emptying whereas administration of 1,000 mg/kg significantly reducing the rate of gastric emptying.
The sponsor has submitted one in vitro and eleven in vivo absorption and plasma pharmacokinetic studies. In vitro permeability of TAF (1,000 µM) was studied in forward (apical to basolateral) and reverse (basolateral to apical) directions using Caco-219 monolayers (Study AD‑120‑2037). Under the study conditions, TAF showed a dose dependent increase in forward permeability and a decrease in efflux ratio indicating saturable efflux transport, with the addition of ciclosporin (CsA) diminishing the efflux ration and increasing forward permeability.
Mouse studies following single dose oral administration of TAF found exposure in plasma was observed in a dose dependent manner, with TFV plasma exposure increasing with increased dose in a greater than dose proportional manner (Studies AD-120-2014 and AD-120-2016). Rat plasma analysis following single oral dose administration found that TAF was rapidly absorbed and converted to TFV, with TFV exposure increasing with the increase in dose that was greater than dose-proportional (Studies R990130, AD-120-2015 and R2000065). Dogs administered a single oral or IV dose of TAF, were observed with no apparent changes in pharmacokinetic parameters following oral administration of TAF, displaying dose dependent pharmacokinetics with TFV observed to be dose proportionality in plasma and oral bioavailability of TAF and TFV was reduced by food (Study 99-DDM-1278-001-PK). TAF was found to be rapidly absorbed following single dose administration in dogs which was followed by the rapid appearance of TFV in plasma and that exposure of TAF and TFV increased in a greater than dose proportional manner with the major metabolites being TFV in plasma and TFV-DP in the liver (Study AD-120-2034). Monkeys administered a single dose of TAF either via oral or IV administration had rapid absorption of TAF followed by the rapid appearance of TFV in plasma where exposure of TAF and TFV increased in a greater than dose proportional manner, with no significant gender based differences (Study P2000087). PBMC assays found that TFV concentration declined more slowly compared to plasma, with a significant portion of TFV related material in PBMCs present in phosphorylated states. Oral repeat dose studies in dogs found that TAF was rapidly absorbed, with TFV and its phosphorylated metabolites efficiently formed post-dose and found to be the major metabolite in plasma and TFV‑DP was efficiently formed in the liver (Study AD‑120-2033 and Study P2000087). Repeat dose studies in monkeys found that peak TAF plasma concentrations were observed within 1 h post-dose and rapidly converted to TFV (Study P2000114-PK).
Overall, the various submitted in vivo absorption studies, single or repeat dose, have demonstrated that TAF generates sufficient exposure in the animal models chosen and is rapidly converted to TFV after administration, with some accumulation of TFV-DP in the liver.
Human and dog plasma was utilised for protein binding in vitro which found that TAF bound moderately (Study AD-120-2026). Mice were used for tissue distribution studies with radio labelled TAF, showing most tissues reached maximum concentration by 1 h post-dose, with blood: plasma concentration ratios suggesting radioactivity was preferentially distributed into the cellular fraction of whole blood (Study AD-120-2011). Distribution analysis in rats following repeat dose administration of TAF found that blood: plasma concentration ratios of radioactivity post‑dose suggested limited distribution of radioactivity into the cellular fraction of whole blood (Study AD‑120-2020). Radioactivity was found to be widely distributed to most tissues. The small percentage of radioactivity eliminated in bile indicated that biliary excretion was a minor route of radioactivity elimination in rats. Studies in dogs found that after reaching maximum concentrations, radioactivity in blood and plasma declined through 24 h post‑dose, with the highest concentrations observed in the liver and kidney, with a high concentration of total radioactivity in bile, suggesting that a substantial portion of the absorbed dose undergoes biliary excretion in dogs (Study AD-120-2009 and Study D990173-BP).
Overall, in vitro analysis found TAF moderately bound in plasma. In vivo studies in mice found TAF was well tolerated in mice and was rapidly distributed into tissues with most reaching maximum concentration by 1 h post-dose. Radioactivity was mainly retained in the liver, which is consistent with high hepatic extraction, and kidneys. Low levels of radioactivity observed in brain and testis in mice indicate that TAF minimally crosses the blood: brain and blood: testis barrier. Results observed in the tissues of pigmented mice show that distribution trends in the pigmented uveal tract of the eye and pigmented skin indicate that TAF is not selectively associated with melanin containing tissues with no melanin binding was observed. Similarly, results indicate no binding of TAF to melanin in rats.
In vivo metabolism studies have been conducted in mice, rats and dogs. Radiolabelled TAF underwent rapid biotransformation in mice via oxidation, hydrolysis, dealkylation, glucuronidation, and acetylation pathways, with hydrolysis of isopropylalanine and phenoxy groups being the major pathway. In mice, the major route of radioactive elimination was via faeces and renal elimination. Metabolism in dogs was indicated to be via oxidation, hydrolysis, dealkylation, and glucuronidation, with a major route elimination of radioactivity in bile duct intact dogs via faeces. Biotransformation of TAF was studied in mice, rats, and dogs and compared to humans, with endogenous purine metabolites including hypoxanthine, xanthine, allantoin, and uric acid observed in all species. TFV was found to be the major metabolite in plasma, urine, and faeces in all species except for human plasma, with no unique human metabolites observed. The major enzymes involved in intracellular conversion of TAF to TFV in PBMCs and primary human hepatocytes are Cat A and carboxylesterase 1 (CES1), respectively.
Absorption and excretion analysis following a single dose of radiolabelled TAF in to bile duct intact and to bile duct cannulated dogs found that bile duct intact dogs, as well as bile duct cannulated dogs eliminated radioactivity rapidly post-dose (Study AD-120-2007). Elimination of a large percentage of radioactivity was via bile in bile duct cannulated dogs, indicating a major route of elimination is through biliary excretion. At the end of the study period, low recoveries of radioactivity remained measurable in urine and faeces, suggesting radioactivity was possibly retained in the carcasses.
The pharmacokinetic profiles in the laboratory animal species (particularly those used in the pivotal repeat dose toxicity studies) were sufficiently similar to allow them to serve as appropriate models for the assessment of drug toxicity in humans.
Pharmacokinetic drug interactions
In vitro studies have been submitted to investigate the potential for TAF and/or TFV to inhibit human CYP mediated drug metabolism, as well as investigating the inhibitory activity of TAF with human liver microsomal CYP isozymes (that is, CYP1A2, CYP2B6, CYP2C8, CYP2C9, CYP2C19, CYP2D6 and CYP3A). TAF weakly inhibited CYP3A mediated oxidation of midazolam or testosterone, and did not inhibit CYP1A2, CYP2C9, CYP2D6, CYP2E1, and CYP3A. TAF incubation produced no time or cofactor dependent inhibition of any enzyme, and cytotoxicity was only observed at a high concentration (100 µM) used in cultured human hepatocytes. At low, non‑cytotoxic concentrations (1 to 10 µM), TAF had no significant effects on increasing messenger RNA (mRNA) levels and/or CYP activities, with no significant induction of P-glycoprotein (P-gp) and/or UGT1A1 mRNA. TAF incubation in cultured human hepatocytes demonstrated little or no potential for CYP induction at clinically relevant concentration (1 µM).
Hepatoma derived cell lines incubated with TAF (50 µM) demonstrated the extent of activation of pregnane X receptor (PXR) being only 23% of the maximal effect of rifampicin, while incubation with lower concentrations found it did not activate aryl hydrocarbon receptors (AhR). In caco-2 cells, TAF increased absorption in the presence of CsA or COBI. Pre-treatment with CsA in dogs increased TAF plasma exposure and oral bioavailability (approximately10 fold), indicating that co-administration of efflux inhibitors increases TAF absorption. TAF was found to be a substrate for OATP1B1 and OATP1B3 (hepatic uptake transporters) but not a substrate for OAT1 or OAT3 (renal transporters).
Overall, the stability of TAF in intestinal subcellular fractions was not markedly affected following incubation with HIV-1 protease inhibitors or CYP inhibitors, indicating that FTC and TFV, EVG and COBI are unlikely to interact with high specificity of the enzymes catalysing the phosphorylation of nucleoside analogues. TAF’s lack of or weak potency for CYP inhibition indicates it will be unlikely to:
Significantly affect hepatic CYP3A activity
Affect the metabolism of EVG or COBI
Activate either PXR or AhR receptors
Contribute to renal tubular cell loading of TFV, since it is not a substrate for renal transporters (OAT1; OAT3)
Be a pharmaceutical agent to mediate transporter drug interactions.
However, TAF is a substrate for intestinal efflux transporters (P-gp; breast cancer resistance protein (BCRP)) and hepatic uptake transporters (OATP1B1; OATP1B3), its exposure may be affected by inhibitors and/or inducers of the intestinal efflux transporters and inhibitors or genetic polymorphisms of OATPs.
Three single dose toxicity studies were conducted in rats and dogs. Rats administered TAF (0 to 1,000 mg/kg) were observed with slight reductions in lymphocyte counts, liver weight and salivation immediately following dose administration (NOEL 6.25 mg/kg (males) and 25 mg/kg (females); NOEL 1,000 mg/kg). A single administration of TAF (30 to 270 mg/kg) in dogs resulted in slight weight loss, reduced food consumption, and microscopic renal changes at 90 and 270 mg/kg (NOEL 30 mg/kg). Overall, the single dose studies in rats and dogs found that TAF was well tolerated. No single dose toxicity studies have been performed with the proposed fixed dose combination (FDC).
Repeat dose toxicity
A series of GLP compliant oral repeat dose toxicity studies in mice (2, 13 weeks), rats (1, 4, 26 weeks), dogs (4, 39 weeks) and monkeys (4 weeks) were submitted. The studies were in accordance with ICH Guidelines on repeat dose toxicity, where the repeat-dose toxicity studies have been conducted in at least two mammalian species (one non-rodent).20 The studies have been well designed and attained high relative systemic exposures (see Relative exposure; Table 3 below).
Exposure ratios have been calculated based on TFV animal: human plasma AUC0–24 h. Human reference values are from clinical study GS-US-292-0104 and GS-US-292-0111. Relative exposure in mice was observed from very low at the lowest concentration of TAF used to very high at the highest TAF dose used, while mice administered TDF in the carcinogenicity study were observed to have extremely high exposure ratios when compared to TFV human steady state values (Table 3 below). A 6 month rat study found that the relative exposure for TAF ranged from very low to extremely high, while the carcinogenicity study using TDF and compared to TFV steady state values in humans found that the relative exposure ranged from moderate to extremely high in the doses used for both sexes. The 9 month study in dogs observed exposure ratios ranging from subclinical to moderate. The exposures observed in the repeat dose toxicity studies have greatly exceeded the human exposure levels observed in the clinical studies of TAF within the FDC. Overall, studies showed that TAF was rapidly converted to TFV, with no substantial plasma accumulation of either TAF or TFV.
Table 3. Relative exposure of TFV after administration of TAF/TDF in repeat dose toxicity and carcinogenicity studies
Study duration (drug)
Study dose (mg/kg/day)
TFV AUCss 0‑24h (ng∙h/mL)
TAF AUCss 0–24h (ng∙h/mL)
TFV/TAF exposure ratio#*
13 weeks (TAF)
2 years (TDF) (carcinogenicity)
6 months (TAF)
2 years (TDF) (carcinogenicity)
9 months (TAF)
4 weeks (TAF)
Human (healthy volunteers)
steady state (TAF)
* Mean steady state AUCSS in pivotal studies for TFV (GS-US-292-0104 and GS-US-292-0111); ** = animal: human plasma AUCss0–24h NC = not calculated, NA = not applicable
Repeat dose toxicity studies identified the kidney (karyomegaly, tubular degeneration in rats and dogs) and bone (atrophy of cancellous bone in rats, increased markers of bone turnover in rats and dogs) as the main targets in the long-term studies.
The 7 day toxicity study in rats showed tubular epithelial cell enlargement and vacuolation enlargement in only one female at the HD of 400 mg/kg/day, might have been indicative of an early toxic effect on tubular cells. In the 4 week study, foci and occasional contiguous areas of cortical tubular basophilia, generally associated with minimal to slight focal nuclear karyomegaly, were observed in both sexes at 400 mg/kg/day, and at 100 mg/kg/day in males. The respective NOAELs were 25 mg/kg/day in males and 100 mg/kg/day in females. In the 26 week study, renal cortical tubular karyomegaly of minimal severity was observed only at the HD of 100 mg/kg/day, with an incidence of 73% in males and 13% in females. At the NOAEL of 25 mg/kg/day the TFV systemic exposure (AUC) ratio was 13.
In the 4 week toxicity study in dogs, tubular karyomegaly was observed in all HD (10 mg/kg/day) dogs, and in 1 male and 1 female at 3 mg/kg/day. In the 9 month study, renal cortical tubular degeneration/regeneration and karyomegaly were observed after 3 months at 6 and 18/12 mg/kg/day. The changes were minimal to slight at 6 mg/kg/day, and mild to moderate at 18/12 mg/kg/day. Similar (minimal) lesions were also observed after 9 months in 2 low dose (LD) (2 mg/kg/day) males. The kidney lesions were still evident after 3 months recovery, but were of reduced incidence and severity. At the NOAEL of 2 mg/kg/day, the TFV exposure ratio was 4.
In the 7 day toxicity study in rats, mean plasma parathyroid hormone (PTH) was increased and mean plasma 1,25-dihydroxyvitamin D3, a marker of bone turnover) was decreased at all doses (-74% at HD), however 25-dihydroxyvitamin D3, mean plasma total calcium and ionised calcium were unaffected. Bone mineral density scans were performed by peripheral quantitative computed tomography in the 4 week and 26 week toxicity studies. In the 4 week study, decreased bone areas and circumferences were consistent with effects on bodyweight, however bone mineral density was decreased, and urinary calcium increased, at the HD of 400 mg/kg/day. There were no significant effects on markers of bone turnover. Dose related decreases in 1,25 dihydroxy vitamin D3 were observed at ≥ 6.25 mg/kg/day. It was suggested that in growing rats the effect on bone mass may be due to a failure to accrue bone and/or fully mineralise bone, rather than removal of existing bone. In the 26 week toxicity study, tibial cancellous bone atrophy, and decreases in bone mineral density and content at the tibial metaphysis and distal femur metaphysis were observed at 100 mg/kg/day. Increases in markers of bone turnover (deoxypyridine-DPD, C telopeptide-CTx) and related hormones (1,25 dihydroxy vitamin D3 and 25 hydroxcy vitamin D3) were observed at 25 and 100 mg/kg/day (TFV exposure ratio 13 at NOAEL of 25 mg/kg/day).
In the 4 week toxicity study in dogs, bone mineral density and markers of bone turnover were generally unaffected at doses up to 10 mg/kg/day. In the 9 month study, decreases in bone density (whole body, femur and lumbar spine) at the HD of 18/12 mg/kg/day were considered secondary to effects on bodyweights. A significant increase in the bone resorption marker collagen type 1 cross-linked N-telopeptide (NTx), and significant decreases in 1,25-dihydroxyvitamin D and 25-hydroxyvitamin D were observed at the HD. There was some evidence of recovery of these parameters 3 months after cessation of treatment, in conjunction with bodyweight gains. At the NOAEL of 2 mg/kg/day, the TFV systemic exposure ratio was 4.
There was no evidence of bone deformities in rats or dogs. Dogs were more sensitive to the bone (and renal) effects than rats. Osteomalacic lesions were reported in a previous study in juvenile monkeys in which TFV was chronically administered by SC injection at 30 mg/kg/day, resulting in very high TFV systemic exposure ratios.
In the 1 month toxicity study in monkeys, there were no treatment related effects on bone morphology, PTH, 1,25-(OH)2 vitamin D2, or bone derived alkaline phosphatase at doses up to 30 mg/kg/day (TFV exposure ratio 20).
Other toxicities included microscopic changes in the nasal mucosa in mice, and minimal histiocyte infiltration in the eye, lung and spleen in dogs. ECG changes observed in dogs are assessed under Safety pharmacology above.
Mice administered TAF ≥ 10 mg/kg/day for 13 weeks showed treatment related microscopic findings in the nasal turbinates (infiltrates of neutrophils in respiratory and olfactory mucosa, as well as exudate in the lumen). Analysis of nasal turbinate sections showed a decrease in the nuclear density/thickness of the olfactory epithelial layer on the ethmoid turbinates, nasal septum, and/or dorsal meatus associated with minimal to moderate severity of olfactory epithelium degeneration, with animals administered ≥ 30 mg/kg/day experiencing greater severely. These effects were not observed in rats, dogs or monkeys, and are unlikely to have clinical relevance.
Some HD dogs showed minimal to slight infiltration of mononuclear cells in the ocular posterior uvea at the HD of 18/12 mg/kg/day in the 9 month study only. An accumulation of macrophages with pigment was also noted in the lungs and spleen of HD dogs, which may have represented accumulation of drug related material. 14C-TAF distribution studies showed only low levels of drug related radioactivity in the eyes of rats and dogs, and no selective binding to melanin. Since this effect was only observed at multiples of the clinical TFV exposure in dogs, was not observed in mice, rats or monkeys, and microscopic and ophthalmic examinations did not reveal any eye abnormalities in any test species, it is unlikely to have clinical relevance.
Combination toxicity studies
No toxicity studies were conducted with the FDC. The ICH M3 (R2) guideline; questions and answers21 states ‘It is accepted that combination studies on advanced cancer, tuberculosis, and HIV products are generallynot warranted unless there is a specific cause for concern under clinically relevant conditions.’ A previously submitted 2 week toxicity study in rats (TX-164-2001) showed no exacerbation of toxicity with the combination of TDF/FTC.
Three genotoxicity studies have been submitted investigating TAF. The gene mutation assays with bacterial strains (Salmonella typhimurium, Escherichia coli), as well as a L5178Y gene mutation assay in mouse lymphoma cells demonstrated that TAF had no genotoxic potential. The studies submitted are appropriate to support the submission and are in accordance to EMEA genotoxicity testing guidelines (CPMP/ICH/174/95).
The sponsor made separate agreements with the FDA (Reference ID: 3161161) and EMA (EMA/CHMP/SAWP/629722/2012, FAL 2410-1-2012), which stated that carcinogenicity studies were not required for TAF registration due to the lack of TAF exposure in rats and TgRasH2 mice, and lower TFV exposure in rats and mice compared to TDF. Therefore, the sponsor has submitted two carcinogenicity studies with TDF in mice and rats. Exposures were approximately10 times (mice) and approximately4 times (rats) those at the therapeutic dose in humans. Both studies were in accordance with EMEA guidance22 for carcinogenicity studies.
The studies in mice investigating the carcinogenic effects of TDF found that under the study conditions female mice showed a low incidence of liver adenomas (600 mg/kg/day) and rats did not show any carcinogenic potential in the long-term study. In comparison with the genotoxicity studies, results suggest that TAF has no genotoxic potential and that there are no carcinogenic potentials identified under the study conditions in animals or possibly humans administered TAF/TDF. No new 104 week carcinogenicity studies have been conducted for EVG, COBI or FTC since they have previously been submitted for the individual agent approvals. No carcinogenicity studies have been submitted for the FDC, which is acceptable.
Reproductive and developmental toxicity
No toxicological interactions were expected with the E/C/F/TAF FDC and further studies with E/C/F/TAF FDC are not required in accordance to the guideline.27Previously submitted studies for the registration of EVG, COBI and FTC demonstrated no significant effects on embryofetal development in rats or rabbits.
Reproductive toxicity studies with TAF/TFV were conducted in two mammalian species:
One study in rats investigating fertility and early embryonic development
Two studies in rats and two in rabbits investigating embryofetal development
One study with TDF in rats investigating perinatal/postnatal reproduction.
The fertility and early embryonic development study in rats administered TAF orally as per clinical use. As stated, the use of the rat model, timing and duration of treatment (males: 10 weeks old and treated from 4 weeks prior to mating for approximately10 weeks; Females: 12 weeks old and treated from 2 weeks prior to mating for, throughout mating and through to gestation day 7 (GD7)), as well as group size per treatment dose was in accordance with ICH guidelines.23 The study was appropriately designed, and observed:
Males: No treatment related effects on reproductive parameters or mean epididymal sperm motility or on sperm concentration. No statistically significant differences in organ weights for the epididymis, prostate, seminal vesicles or pituitary
Females: No treatment related differences in any reproductive parameters measured, macroscopic findings at necropsy. No statistically significant differences in reproductive organ weights or treatment related effects on any of the caesarean section parameters
A dose dependent decrease in mean body weight was observed for animals administered ≥ 80 mg/kg/day. During gestation the mean gestation body weight was found to be dose dependently reduced with statistically significant reductions noted on GD3, 7, and 10 (160 mg/kg/day)
NOAEL (male and fertility): 80 mg/kg/day; NOAEL (reproductive and early embryonic toxicity): 160 mg/kg/day.
Embryofetal development studies found:
Administration of ≥ 100 mg/kg/day was maternally toxic, and administration of 250 mg/kg/day to dams produced a decreased fetal body weight associated with some minor transitory delays in the rate of ossification
No treatment related evidence of embryo lethality or teratogenicity
Maternal clinical signs included higher decreased faecal output compared to control during the treatment period, which was associated with decreased food intake (100 mg/kg/day). Compared to control, maternal body weight gains were observed to be statistically significantly decreased for absolute and percentage gain, as well as an overall decrease in food consumption during the treatment period (100 mg/kg/day). Uterine findings included a pregnancy rate of at least 85% per group with the numbers of corpora lutea, implantation sites, live and dead fetuses and resorptions, the sex ratio, as well as the pre and post implantation losses and fetal weights were comparable to controls at all doses
No evidence of embryo lethality, fetotoxicity or teratogenicity at any dose level
As per separate agreements with the FDA24 and EMA25, a perinatal and postnatal study were not required for TAF registration due to the lack of TAF exposure in rats and TgRasH2 mice and lower TFV exposure in rats and mice compared to the same studies in which TDF was administered. Therefore, a perinatal/postnatal reproduction study in rats orally administered TDF (50 to 600 mg/kg/day) was submitted. The study was appropriately designed and observed:
An increase in pup mortality at maternally toxic doses ≥ 450 mg/kg/day
Maternal (F0): Adverse clinical observations at ≥ 150 mg/kg/day, included reductions in body weight gain during gestation and increases in body weight gain during lactation. Dams administered ≥ 450 mg/kg/day had increased peri/postpartum pup mortality, reduced pup survival and reduced pup body weights
Pups (F1): Treatment related, dose dependent prolongation of sexual maturation was observed. No significant treatment related observations in behavioural evaluations. Mating performance of both sexes was unaffected by treatment
Toxicokinetic milk analysis (Day 11 postpartum) confirmed the presence of excreted TFV which was not concentrated. Pharmacokinetic plasma data and plasma profiles indicate an increased exposure of TFV with increased oral doses of TDF.
Relative exposure ratios to TFV after administration of TAF in rats and rabbits were observed to be extremely high at the highest concentrations used compared human steady state values. Additionally, cross species comparisons of exposure in F0 and F1 generation animals, demonstrate that F0 and F1 generation animals experienced minimal reproductive toxicity at high doses more than 40 to 59 fold (rats) and 93 fold (rabbits) the exposure of TFV compared to the predicted safety margin relative to humans. The exposure to TAF at NOAEL for reproductive toxicity was 1.2 fold (rats) and 53.6 fold greater than the predicted safety margin relative to human steady state AUC (Table 4 below). Overall, studies showed that TAF was rapidly converted to TFV, with no substantial plasma accumulation of either TAF or TFV. The doses used and exposure ratios in the rat and rabbit reproductive toxicity studies are acceptable.
Table 4. Relative exposure of TFV after administration of TAF in embryofetal development studies
Test dose (mg/kg/day)
TFV AUC0–t (ng∙h/mL)
TAF AUC0–t (ng∙h/mL)
TFV/TAF exposure ratio#*
Human -healthy volunteers
steady state *
* Mean steady-state AUCSS in pivotal studies for TAF (GS-US-292-0104 and GS-US-292-0111); *# = animal: human plasma AUC0–24 h NC = not calculated, NA = not applicable
Table 5. TFV/TAF exposure ratios at the NOAELs in reprotox studies
NA = not applicable; NC = insufficient data to calculate a Predicted safety margin for TFV human exposure is based on pooled PK data from E/C/F/TAF Phase III pivotal Studies GS-US-292-104 and GS-US-292-111 where the mean TFV AUCss = 293 ng.h/mL; b Predicted safety margin for TAF human exposure is based on pooled PK data from E/C/F/TAF Phase III pivotal Studies GS-US-292-104 and GS-US-292-111 where the mean TAF AUCss = 206 ng.h/mL; c NOAEL for reproductive endpoints provided; AUC data is for maternal exposure; the peri/postnatal study was conducted with TDF not TAF
The registered drugs EVG, COBI, and FTC have respective pregnancy categories of B226, B127 and B1. TFV has category B328. An Australian pregnancy category of B3 is appropriate for Genvoya.
The sponsor submitted two local tolerance studies, one in rabbits and one using bovine eyes. The adequately designed studies conclude that TAF was a non-corrosive/non severe eye irritant, as well as being non irritating/non corrosive to rabbit skin under semi occluded conditions. Previously submitted studies for registration showed that EVG was not irritating to skin, not a severe irritant to eyes and showed no potential for phototoxicity. Similarly, COBI in previously submitted studies was mildly irritating to skin, not a severe irritant to eyes and showed no potential phototoxicity. No local tolerance studies were conducted for the EVG/COBI/FTC/TAF combination.
The FDC is proposed for paediatric use, for children over 12 years old, however no specific studies in juvenile animals were submitted. Repeat dose toxicity studies in adult animals did not identify systems that maybe targets for toxicity at clinically relevant exposure levels.
Comments on the safety specification of the risk management plan
Results and conclusions drawn from the nonclinical program for Genvoya detailed in the sponsor’s draft Risk Management Plan are in general concordance with those of the nonclinical evaluator.
The nonclinical evaluator made recommendations regarding the PI but these are beyond the scope of the AusPAR.
Nonclinical summary and conclusions
The nonclinical studies submitted were adequate, well designed, and safety studies were GLP compliant and in accordance with ICH guidelines for a new chemical entity (TAF). The guideline8 states that combination toxicity studies with drugs for HCV are generally not warranted, and adequate scientific justification was provided for their absence.
Primary pharmacology: TAF is a prodrug for tenofovir (TFV), which is metabolised intracellularly to the active metabolite, TFV-DP, and acts as a DNA chain terminator during HIV reverse transcription. The lysosomal carboxypeptidase Cat A is essential to the intracellular conversion of TAF to TFV. Some viral protease inhibitors may strongly inhibit Cat A, in vitro assays revealed that telaprevir and boceprevir, inhibitors of the HCV protease, strongly inhibited Cat A mediated hydrolysis of TAF. The HIV protease inhibitors darunavir, atazanavir, lopinavir and ritonavir, and COBI, had no clinically relevant effect.
TAF showed anti-HIV-1 activity in vitro against wild type clinical isolates from Group M (subtypes A to G), and Groups N and O, and HIV-2, in human activated PBMCs, with EC50 values ranging from 0.14 to 12.0 nM. TAF respective mean EC50 values were 11.0 ± 11.3 nM and 9.7 ± 4.6 in CD4+ T cells and monocyte derived macrophages from diverse human donors. TAF showed a range of synergistic activity, and no antagonism, when tested in pair wise combinations with anti-HIV NRTIs, NNRTIs, protease inhibitors and INSTIs in vitro in MT-2 cells infected with HIV-1 IIIB.
Since TAF and TFV are prodrugs for the same active intracellular metabolite, TFV-DP, they showed nearly identical resistance profiles in vitro, with emergence of the K65R ± S68N mutation in reverse transcriptase, associated with up to 6.5 fold phenotypic resistance to TFV in a long-term selection study. TAF and TFV also had identical phenotypic resistance profiles against a panel of 24 patient derived HIV-1 isolates with a wide range of NRTI resistance mutations.
Secondary pharmacodynamics studies showed that TAF had potent activity against HIV-1, SIV, and HBV (EC50 18 nM), but no significant activity against a range of other viruses. In vitro studies indicated that TFV-DP was selective for HIV reverse transcriptase in comparison to human DNA polymerases α, β, and ϒ, and did not deplete mitochondrial DNA in cultured HepG2 cells.
Safety pharmacology investigations showed that TAF had no effect on the central nervous system after a single dose in rats. TAF did not significantly inhibit hERG in vitro. Cardiovascular investigations in dogs found no TAF treatment related effects on the ECG, heart rate or systemic blood pressure. Single oral administration in rats found that TAF increased urinary output of calcium and there was a correlation with an increase in serum calcium concentration which indicated that the kidneys were functioning in order to reduce the serum calcium load. A gastrointestinal study in rats found that TAF 100 mg/kg had no clear effect on gastric emptying whereas administration of 1,000 mg/kg significantly reducing the rate of gastric emptying.
Pharmacokinetics: Absorption studies, whether single or repeat dose showed that following administration of TAF in different animal models, the intact prodrug was rapidly converted to TVF, with some accumulation of TFV-DP in the liver. Distribution studies found that TAF moderately bound in plasma. In vivo studies in mice demonstrated that TAF was well tolerated. Studies showed that radioactivity was mainly retained in the liver and kidneys and minimally crossed the blood: brain and blood: testis barrier. Study results in rats suggest there is possibly no binding of TAF to melanin. In vivo metabolism studies in mice, rats and dogs indicated that 14C-TAF underwent rapid biotransformation via oxidation, hydrolysis, dealkylation and glucuronidation pathways. The studies concluded that clearance of TAF was via faeces as well as renal elimination. Elimination of a large percentage of radioactivity was via bile from bile duct cannulated dogs, which indicated that biliary excretion maybe a major route of elimination of 14C‑TAF‑derived radioactivity.
Single dose toxicity studies in rats showed slight reductions in lymphocyte counts, liver weight and salivation immediately following dose administration. Dogs administered TAF showed slight weight loss, reduced food consumption, and microscopic renal changes at 90 and 270 mg/kg (NOAEL 30 mg/kg).
Repeat dose toxicity studies of 2 and 13 weeks duration in mice, 1, 4 and 26 weeks duration in rats, 4 and 39 weeks duration in dogs, and 4 weeks in monkeys identified the kidney (karyomegaly, tubular degeneration in rats and dogs), and bone (atrophy of cancellous bone in rats, bone mineral loss, increased markers of bone turnover in rats and dogs) as the main targets in the longer studies. In dogs, the bone kidney changes were still present, but with reduced incidence and severity, after 13 weeks recovery. TFV exposure ratios at the respective TAF NOAELs of 25, 2 and 30 mg/kg/day in the 26 week, 39 week and 4 week toxicity studies in rats, dogs and monkeys were 13, 4 and 20.
Mice showed degenerative and inflammatory (neutrophil infiltrate) changes in the nasal mucosa, however these effects were not observed in rats, dogs or monkeys, and are unlikely to have clinical relevance. Minimal infiltration of histiocytes was observed in the eye (choroid plexus, ciliary body), lung and spleen at the high dose in the 39 week study in dogs, but is unlikely to have clinical relevance.
TAF was not genotoxic, and TFV had no relevant carcinogenic potential. Adequate justification was provided for the lack of genotoxicity and carcinogenicity studies with the FDC.
A fertility and early embryonic development study in rats with TAF found no statistically significant treatment related changes in organ weights for the epididymis, prostate, seminal vesicles or pituitary. Females were observed with no treatment related differences in any reproductive parameters measured, or macroscopic findings at necropsy. There were no statistically significant differences in reproductive organ weights or treatment related effects on any of the caesarean section parameters. Embryofetal development studies in rats and rabbits found no treatment related evidence of embryo lethality or teratogenicity. The perinatal/postnatal reproduction study in rats administered TDF found:
No significant treatment related behavioural evaluations of F1 generation pups
Mating performance of both F1 generation sexes was unaffected by treatment of F0 generation dams
Administration of TAF to the F0 generation dams resulted in no treatment related effects in F1 generation female rats. Analysis of caesarean sectioning and litter parameters found that litter averages for corpora lutea, implantations, litter sizes, live fetuses, early and late resorptions, fetal body weights, percent resorbed conceptuses, and percent live male fetuses were comparable to control
All placentae appeared normal
Toxicokinetic milk analysis (Day 11 postpartum) confirmed the presence of excreted TFV, which was not concentrated. Pharmacokinetic plasma data and plasma profiles indicate an increased exposure of TFV with increased oral doses of TDF.
Local tolerance studies found that TAF was a non-corrosive/non severe eye irritant, as well as being non irritating/non corrosive to rabbit skin under semi occluded conditions.
The stability and degradation of EVG, FTC and COBI were reportedly comparable in both Genvoya and Stribild, and there were no unique impurities or degradants in Genvoya. However, there was insufficient data to qualify two TAF impurities.
There were no major deficiencies in the nonclinical data.
Primary pharmacology studies with TAF support its use for the proposed indication. TAF is metabolised intracellularly to the active metabolite, tenofovir diphosphate. The lysosomal carboxypeptidase Cat A is essential to the intracellular conversion of TAF to TFV. The resistance profile of TAF was almost identical to that of TDF. In vitro assays revealed that the HCV protease inhibitors telaprevir and boceprevir strongly inhibited Cat A mediated hydrolysis of TAF. More recently registered HCV protease inhibitors were not tested. This is brought to the attention of the clinical evaluator.
No clinically relevant hazards were identified with secondary pharmacodynamics and safety pharmacology studies.
Repeat dose toxicity studies conducted in mice, rats, dogs and monkeys achieved high tenofovir exposure ratios. The main targets were the kidney (karyomegaly, tubular degeneration) and bone (mineral loss, increased turnover), at multiples of the TFV clinical exposure. There was no evidence of bone deformities.
In accordance with the guideline8, toxicity studies were not conducted with the drug combination, and adequate scientific justification was provided for the absence of combination toxicity studies.
Genotoxicity and carcinogenicity were conducted in accordance with ICH guidelines. TAF was not genotoxic and TFV had no relevant carcinogenic potential.
Reproductive toxicity studies with TAF/TDF were appropriately designed and in accordance to the ICH guidelines, with no evidence of embryo lethality, fetotoxicity or teratogenicity.
All impurities were qualified except for the two TAF degradants.
There are no nonclinical objections to registration; however the qualification of 2 TAF impurities requires further clarification.